Synthesis of Sn-Beta with Exclusive and High Framework Sn Content

Article abstract of DOI:10.1002/cctc.201403050

Sn-Beta zeolite was prepared by acid dealumination of Beta zeolite, followed by dehydration and impregnation with anhydrous SnCl₄. The formation of extraframework Sn (EFSn) species was prevented by the removal of unreacted SnCl₄ in a methanol washing step prior to calcination. The resulting Sn-Beta zeolites were characterized by X-ray diffraction, Ar physisorption, NMR, UV/Vis, and FTIR spectroscopy. These well-defined Lewis acid zeolites exhibit good catalytic activity and selectivity in the conversion of 1,3-dihydroxyacetone to methyl lactate. Their performance is similar to a reference Sn-Beta zeolite prepared by hydrothermal synthesis. Sn-BEA zeolites that contain EFSn species exhibit lower catalytic activity; the EFSn species also catalyze formation of byproducts. DFT calculations show that partially hydrolyzed framework Sn-OH species (open sites), rather than the tetrahedral framework Sn sites (closed sites), are the most likely candidate active sites for methyl lactate formation.

Full text of DOI:10.1002/cctc.201403050

DOI: 10.1002/cctc.201403050
Full Papers
Synthesis of Sn-Beta with Exclusive and High Framework
Sn Content
William N. P. van der Graaff,[a] Guanna Li,[a] Brahim Mezari,[a] Evgeny A. Pidko,*[a, b] and
Emiel J. M. Hensen*[a]
Sn-Beta zeolite was prepared by acid dealumination of Beta
zeolite, followed by dehydration and impregnation with anhy-
drous SnCl4. The formation of extraframework Sn (EFSn) spe-
cies was prevented by the removal of unreacted SnCl4 in
a methanol washing step prior to calcination. The resulting Sn-
Beta zeolites were characterized by X-ray diffraction, Ar physi-
sorption, NMR, UV/Vis, and FTIR spectroscopy. These well-de-
fined Lewis acid zeolites exhibit good catalytic activity and se-
lectivity in the conversion of 1,3-dihydroxyacetone to methyl
lactate. Their performance is similar to a reference Sn-Beta zeo-
lite prepared by hydrothermal synthesis. Sn-BEA zeolites that
contain EFSn species exhibit lower catalytic activity; the EFSn
species also catalyze formation of byproducts. DFT calculations
show that partially hydrolyzed framework Sn-OH species (open
sites), rather than the tetrahedral framework Sn sites (closed
sites), are the most likely candidate active sites for methyl lac-
tate formation.
Introduction
Dwindling reserves of fossil fuels and the growing environmen-
tal concerns associated with their combustion necessitates the
development of new chemical processes based on renewable
feedstock. Readily available lignocellulosic biomass that does
not compete with food production will play an increasingly im-
portant role in sustainable chemical processes for the produc-
tion of chemicals and fuels.[1–4] The chemistry needed to up-
grade lignocellulosic biomass is very different from the chemis-
try currently used to convert hydrocarbon-based raw materials.
The selective and atom-efficient conversion of biomass to
value-added compounds requires the development of new
technologies and catalyst systems.
acetone and glyceraldehyde;[7–10] an alternative, performed on
a commercial scale is the fermentation of glucose.[5] To develop
viable catalytic systems for these transformations, the develop-
ment of benign chemocatalytic routes employing stable and
versatile catalysts is desirable. Recently, the development of
active Lewis acid chemocatalysts for the conversion of carbo-
hydrates into value-added chemicals has become an important
research area in the field of heterogeneous catalysis. In particu-
lar, Sn-modified large-pore Beta zeolites are highly active and
selective catalysts for the conversion of carbohydrates.[11–16] To
date, Sn-Beta zeolite is the preferred catalyst in terms of activi-
ty and selectivity for the retro-aldol reaction of glucose to-
wards lactic acid, as well as in the isomerization of glucose to
fructose, representing a key step in selective sugar-valorization
paths.[11–15]
Several intermediates that can be derived from biomass and
can be used in different applications have already been identi-
fied.[4–6] Among such platform chemicals, lactic acid is promis-
ing because it is already widely used in the food industry and
it can also serve as a monomer for biopolymers, such as poly-
lactic acid, and as an intermediate in the production of other
industrially important compounds, such as acrylic acid or
green solvents.[5] Lactates can be obtained from biomass by
conversion of glycerol-derived trioses, such as 1,3-dihydroxy-
It is generally believed that the active sites in Sn-Beta cata-
lysts for such conversions are framework Sn atoms (FSn). Direct
incorporation of Sn into the zeolite framework by hydrother-
mal synthesis is difficult and typically requires the use of fluo-
ride media,[17,18] which presents considerable environmental
drawbacks. This problem is accompanied by the formation
large crystallites owing to the long synthesis times, although
methods have been developed to reduce these draw-
backs.[13,19] Furthermore, the employed Sn precursor has an
effect on the morphology and the catalytic performance of the
zeolite.[20] In addition to the lengthy synthesis, the hydrother-
mal synthesis method yields relatively large zeolite crystals,
which may lead to mass-transfer limitations when relatively
bulky biomass fragments are to be converted. Therefore, the
introduction of Sn in smaller zeolite crystals is one way to pre-
vent these diffusion limitations. Another problem with the
fluoride-assisted method is the relatively low Sn content of the
framework (<2 wt%). At a high Sn content in the gel, extrafra-
[a] W. N. P. Van der Graaff, Dr. G. Li, B. Mezari, Dr. E. A. Pidko,
Prof. Dr. E. J. M. Hensen
Inorganic Materials Chemistry Group
Schuit Institute of Catalysis
Eindhoven University of Technology
PO Box 513, 5600 MB Eindhoven (The Netherlands)
E-mail: e.j.m.hensen@tue.nl
[b] Dr. E. A. Pidko
Institute for Complex Molecular Systems
PO Box 513, 5600 MB Eindhoven (The Netherlands)
E-mail: e.a.pidko@tue.nl
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201403050.
ChemCatChem 0000, 00, 0 – 0
1
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
mework Sn species (EFSn), in the form of dispersed intrazeolite
Sn species as well as bulk tin oxides, are formed. It has been
speculated that such EFSn species decrease the performance
of Sn-Beta in carbohydrate-isomerization reactions.[21,22]
more active for the conversion of 1,3-dihydroxyacetone (DHA)
to methyl lactate (ML) in methanol than the previously report-
ed postsynthetically prepared Sn-Beta zeolites[33] and also the
Sn-Beta prepared by alternative hydrothermal routes.[13] The
materials were characterized by pyridine-FTIR, UV/Vis spectros-
copy, solid-state NMR spectroscopy, Ar physisorption, X-ray dif-
fraction, and elemental analysis. Mechanistic insight was ob-
tained by exploring the importance of open versus closed Sn
sites by DFT calculations of the rate-limiting H-shift step.
In the view of these limitations, there has been quite some
interest in the development of more-facile preparation meth-
ods for Sn-Beta catalysts.[23–26] Specifically, indirect/postsynthe-
sis methods that aim to introduce Sn in the framework of an
already synthesized Beta zeolite have been explored
(Scheme 1). In addition to their simplicity, higher amounts of
Results and Discussion
As a precursor for the Sn-containing zeolites, nanosized crystal-
line Al-containing Beta with varying Si/Al ratios (BEA-25, BEA-
50, and BEA-100) were prepared according to the procedure
reported by Mintova et al.[35]
Scheme 1. Postsynthesis functionalization of zeolites with Sn.
The X-ray diffraction patterns shown in Figure 1 confirm the
exclusive formation of Beta zeolite. Direct calcination of the
material at 5508C (heating rate 18Cminꢀ1), either statically in
Sn can be introduced into the framework than by direct hydro-
thermal synthesis. A common postsynthesis strategy involves
the introduction of Sn after the removal of framework hetero-
atoms, such as B and Al.[22,27–34] The direct introduction of B
and Al into the Beta lattice is easier than that of Sn. The silanol
nests arising from acid deboronation or dealumination provide
anchor points for Sn. For instance, Li et al. used chemical vapor
deposition (CVD) of SnCl4 on dealuminated zeolites.[27] Al-
though more Sn can be introduced into the framework by this
method, it remains difficult to limit the amount of EFSn spe-
cies.[27,28] Jin et al. reported that the presence of mesoporosity
reduces the formation of EFSn species.[29] Solid-state ion ex-
change has also been explored by several groups.[22,30,31] By
thorough grinding of a solid Sn precursor [e.g. Sn(OAc)2 or
Sn(Me2Cl2)] and the zeolite followed by careful calcination, Sn
is dispersed in the zeolite micropores and anchored to the
vacant T-sites of the dealuminated parent zeolite. Although
high Sn contents of up to 10 wt% have been achieved by this
method, the contribution of EFSn species increases as the Sn
content is increased. Other methods include the impregnation
of a dealuminated zeolite with a Sn-containing precursor, fol-
lowed by calcination.[32–34] Although facile, there is little control
over Sn speciation and, usually, a considerable amount of EFSn
species is formed.[32,34] These methods and their most impor-
tant limitations are briefly summarized in the Supporting Infor-
mation, Table S1. Despite the promise of these methods com-
pared with the direct synthesis of Sn-BEA, a simple and repro-
ducible approach to prepare Sn-Beta with exclusive framework
Sn is still desirable.
Figure 1. X-Ray diffraction patterns of the as-synthesized BEA-25, BEA-50,
and BEA-100 and the corresponding Sn-treated materials. Sn-BEA-HF was
added as reference.
air or in an artificial air flow (He/O2 =4:1), resulted in partial
collapse of the zeolite lattice.[36] The structural damage follows
from the severe line broadening in the diffraction patterns (see
the Supporting Information Section S2). Therefore, a milder
two-step calcination procedure was employed involving heat-
ing of the as-synthesized zeolite to 4008C for 10 h, followed by
a second calcination step at 5508C for 10 h by using heating
rates of 18Cminꢀ1 in an artificial air flow. By this approach, zeo-
lite-framework decomposition was almost completely sup-
pressed, as evidenced by the XRD results (Figure 1). The Si/Al
ratio in the as-synthesized zeolite was different from the gel
Si/Al ratio (Table 1). The final Si/Al ratios for BEA-25, BEA-50,
and BEA-100 were 14, 20, and 33, respectively. These findings
agree well with those in the original work of Bein and co-work-
ers.[35]
Herein, we report the synthesis, characterization, and catalyt-
ic properties of a Sn-Beta zeolite by a simple and reproducible
protocol for the introduction of large amounts (>5 wt%) of Sn
into the Beta framework without the formation of EFSn spe-
cies. The dealuminated Beta zeolite is impregnated with anhy-
drous SnCl4 under an inert atmosphere and excess SnCl4 is re-
moved by washing with methanol. This step is necessary to
avoid the formation of EFSn species during the subsequent
calcination step. The resulting Sn-Beta zeolite is substantially
The textural properties of the synthesized materials were an-
alyzed by Ar physisorption (Table 2). BEA-25, BEA-50, and BEA-
100 zeolites have similar BET surface areas. The external surface
&
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
2
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
previous reports,[37] the acid leaching of lattice Al resulted in
a small increase of the mesopore volume at the expense of
the micropore volume (Table 1).
Table 1. Elemental composition of the prepared materials[a]
Material
Si/Al[b]
Si/Al[c]
Sn content Si/Sn Substitution
(Al-BEA) (Sn-BEA) [wt%]
efficiency [%]
The dealuminated zeolites (Deal-BEA-X) were placed in
a Schlenk flask and dehydrated under vacuum at 1708C for
3 h. This step was necessary to avoid uncontrolled hydrolysis
of the added Sn precursor. Such hydrolysis would lead to the
formation of EFSn species in the micropores. After cooling to
1008C, the dehydrated zeolite powder was impregnated with
excess anhydrous liquid SnCl4 under an inert Ar atmosphere.
The impregnated zeolite was then incubated for 16 h at 1008C
to allow diffusion of the grafting agent into the pores. The
condensation of Sn into the framework was achieved by calci-
nation at 5508C (18Cminꢀ1) for 5 h. Prior to the calcination
step, the zeolites were thoroughly washed with methanol to
remove excess SnCl4. This step was essential to avoid the for-
mation of EFSn sites.
Sn-BEA-25
Sn-BEA-50
Sn-BEA-100
Sn-BEA-50*
Sn-BEA-HF
Sn-MCM-41
14
20
33
20
–
>1500
>1500
>1500
>1500
–
2.81
5.07
3.49
5.88
1.95
6.82
62
34
51
29
95
26
23
59
65
n/a
n/a
n/a
–
–
[a] Determined by ICP-AES. [b] Si/Al ratio in the Al-containing zeolites
after calcination. [c] Si/Al ratio in the postsynthesized Sn-BEA materials
after dealumination, SnCl4 treatment, and calcination.
Table 2. Physicochemical properties of the synthesized materials.[a]
Material
SBET
Vmicro
[m3 gꢀ1
Vmeso
[m3 gꢀ1
Sexternal
The UV/Vis spectra (Figure 3) of Sn-BEA-25, Sn-BEA-50, and
Sn-BEA-100 contained only very weak absorption features at
260 nm, characteristic for EFSn species.[27] Incomplete removal
of unreacted Sn precursor from the zeolite pores, owing to in-
sufficient washing with methanol, led to the formation of EFSn
upon calcination (Sn-BEA-50*). This formation was evident
from the intense and broad absorption band at higher wave-
lengths (250–300 nm). The UV/Vis spectrum of Sn-BEA-50* re-
sembles that of Sn-BEA materials prepared by high-tempera-
ture chemical vapor deposition of SnCl4.[27,28] Grafting of SnCl4
to silanol nests in zeolites followed by extensive washing with
methanol resulted in more-homogeneous Sn speciation. Com-
pared with other Sn-silicate materials reported in litera-
ture,[18,19,22,27,35,38] the ligand-to-metal charge transfer
(O2ꢀ!Sn4+) band, which is usually assigned to tetrahedrally
coordinated FSn, was not observed at 210 nm but at lower
wavelength (<200 nm). This shift to shorter wavelengths has
been observed before for Sn-BEA.[21]
[m2 gꢀ1
]
]
]
[m2 gꢀ1
]
Al-BEA-25
Al-BEA-50
542
573
602
508
473
477
461
483
485
436
450
0.12
0.16
0.21
0.07
0.14
0.11
0.07
0.15
0.15
0.11
0.16
0.54
0.14
0.09
0.47
0.16
0.11
0.50
0.19
0.12
0.11
0.03
304
254
211
357
188
263
303
202
211
238
153
Al-BEA-100
Deal-BEA-25
Deal-BEA-25
Deal-BEA-100
Sn-BEA-25
Sn-BEA-50
Sn-BEA-100
Sn-BEA-50*
Sn-BEA-HF
[a] Determined by Ar physisorption.
area increased with the Al content. This trend indicates that
the zeolites with a higher Al content consist of smaller crystals.
This is likely to be because of the higher nucleation rate when
the Al content in the gel is increased. TEM analysis confirms
that smaller particles were ob-
tained for Al-rich gel composi-
tions (Figure 2). The obtained
zeolite particles are agglomer-
ates of smaller crystals of ap-
proximately 5 nm. This morphol-
ogy is particularly pronounced
for BEA-25 and BEA-50 (Fig-
ure 2a,b). BEA-25 contains parti-
cles smaller than 50 nm (Fig-
ure 2a). The particle size of BEA-
100 is approximately 150 nm
(Figure 2c).
The Al-BEA zeolites were deal-
uminated by treatment with 6m
aqueous HNO3 at 1108C for 16 h
(50 mLgꢀ1). This procedure re-
sulted in the complete removal
of Al (Table 2), without affecting
the crystallinity of the zeolites
(see the Supporting Information
Figure 2. TEM images of a) BEA-25, b) BEA-50, c) BEA-100, d) Sn-BEA-25, e) Sn-BEA-50, and f) Sn-BEA-100. Scale
Section S1). In agreement with bars=50 nm.
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
3
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Figure 3. Diffuse reflectance UV/Vis absorption spectra of the synthesized
materials. Absorption bands owing to framework and extraframework Sn
species are indicated with f and ef, respectively.
Additional evidence for the selective incorporation of Sn
into the vacant silanol nests is provided by FTIR spectroscopy.
The intensity of the broad band in the range n˜ =3500–
3600 cmꢀ1 associated with silanol hydroxyl nests in the Deal-
BEA zeolites decreased after the introduction of SnCl4
(Figure 4).
The BEA-50 materials before and after dealumination and
SnCl4 grafting were further characterized by 1H, 29Si, and
119Sn NMR spectroscopy. 1H and 29Si NMR spectra (Figure 5A
and B, respectively) further demonstrate that the majority of
the silanol nests reacted with SnCl4. The reaction of Deal-BEA-
50 with SnCl4 led to a strong decrease in the relative intensity
Figure 5. A) 1H magic angle spinning (MAS) NMR spectra of dehydrated
parent Al-Beta (H-BEA-50, blue line), Deal-BEA-50 (red line) and Sn-BEA-50
(green line). The spectra are normalized by weight. B) 29Si direct excitation
MAS NMR spectra of a) parent H-BEA-50, b) dealuminated H-(Al-)BEA-50 and
c) Sn-BEA-50.
1
of the H NMR signal at d=2.0 ppm corresponding to internal
py (see Figure S4) showed that Sn species in hydrated and de-
hydrated materials displayed characteristic signals at d=ꢀ700
and ꢀ430 ppm, respectively, associated with Sn in tetrahedral
positions of the zeolite lattice.[21,27,39,40]
silanol nests (Figure 5A). The changes in the intensity of the
signal at d=1.8 ppm, owing to silanol groups on the external
surface of the zeolite crystals, was much less pronounced. The
occupation of internal silanols by Sn is further underpinned by
the 29Si NMR spectra shown in Figure 5B. The respective Q3-
signal at d=ꢀ101 ppm increased upon dealumination of the
parent BEA-50. This signal was reduced to its original intensity
after treatment of the zeolite with SnCl4. 119Sn NMR spectrosco-
The Sn content of the modified zeolites (Table 3) was in the
range of 2.81–5.07 wt% (236–427 mmolgꢀ1). These values are
lower than the values computed on the basis of the initial Al
content. It is also seen that the efficiency of Sn incorporation
increased when the parent zeolite was more siliceous (Table 3).
The substitution efficiency, defined as the amount of Sn incor-
porated relative to the amount of removed Al, ranged from
23% for Sn-BEA-25, to 59% for Sn-BEA-50, and 65% for Sn-
BEA-100. A tentative explanation is that the removal of a larger
Table 3. Acid site density of the synthesized materials. Quantification was
performed by using the absorption coefficients reported by Datka et al.[41]
[c]
Material
BAS[a]
[mmolgꢀ1
LAS[a]
[mmolgꢀ1
Sn content[b]
[mmolgꢀ1
]
ALAS,300/ALAS,100
]
]
Sn-BEA-25
Sn-BEA-50
Sn-BEA-100
Sn-BEA-50*
Sn-BEA-HF
16.0
17.2
16.7
3.6
281
530
306
392
118
236
427
292
495
166
0.26
0.36
0.23
0.28
0.13
11.8
Figure 4. FTIR spectra of the OH region of the dealuminated and Sn-func-
tionalized materials after dehydration. The band generally assigned to the
formation of internal silanols after dealumination is indicated in the spec-
trum.
[a] Lewis (LAS) and Brønsted (BAS) acid site density determined at 2008C.
[b] Number of LAS based on ICP Sn content. [c] Ratio of the band areas
of the n˜ ꢁ1450 cmꢀ1 signal.
&
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
4
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
fraction of framework sites for the Al-rich zeolites led to more
extensive restructuring of the lattice. Such large defects might
be more difficult to heal upon treatment with SnCl4 compared
with materials derived from the zeolites with a higher Si/Al
ratio.
Infrared spectroscopy of adsorbed pyridine (Pyads IR) was em-
ployed to evaluate the strength and number of Lewis and
Brønsted acid sites in the Sn-modified samples (Figure 6). The
Figure 7. Sn content versus pyridine adsorbed. The displayed values for the
pyridine adsorption were determined at 2008C.
tively. The amount of LAS correlated linearly with the Sn con-
tent in the postsynthetically modified Sn-BEA zeolites
(Figure 7). The LAS densities determined from the pyridine IR
measurements are higher than Sn content of the zeolites
(Table 3). The extinction coefficients used to estimate the LAS
content were taken to be equal to the extinction coefficients
for Al LAS.[41] We surmise that the different binding of pyridine
to Sn sites may be the cause of the considerable deviations.
Figure 7 also shows that Sn-BEA-50* does not follow the trend
observed for the other samples. Despite its higher Sn content
(Table 1), Sn-BEA-50* contained much less LAS (Table 3,
Figure 7). This result suggests that some of the Sn species in
this material are not accessible to pyridine, possibly because of
blockage of the pores by EFSn species. The reference Sn-BEA-
HF contained slightly less LAS than expected from the correla-
tion.
The ratio of intensities of the pyridine signals owing to LAS
measured at 300 and 1008C (ALAS,300/ALAS,100, Table 3) can be
used as a qualitative measure of the strength of the Lewis acid
sites. The results presented in Table 3 suggest that the LAS
strength of the postsynthesized Sn-BEA samples was consider-
ably higher than that in the reference Sn-BEA-HF. We speculate
that these strong LAS are partially hydrolyzed open Sn cen-
ters.[42]
Figure 6. FTIR spectra of pyridine adsorbed on a) the synthesized Sn-modi-
fied materials and outgassed at 2008C and b) Sn-BEA-50 followed by evacu-
ation at 100, 200, and 3008C. Different sites are denoted as SL (strong Lewis
acid), H (H-bonded pyridine), B (Brønsted acid), and BL (Brønsted/Lewis
acid).
spectra were dominated by very strong signals at n˜ =1450 and
1610 cmꢀ1, corresponding to pyridine adsorbed to strong
Lewis acidic sites (SL). In addition, a weak band at n˜ =
1545 cmꢀ1 corresponding to Brønsted acid sites was observed.
The intensity of this band rapidly decreased upon evacuation
of the samples at elevated temperatures, indicating the weak
acidity of these sites. We surmise that these sites might be un-
occupied silanol nests or partially hydrolyzed framework Sn
sites.
The intensities of the IR bands at n˜ =1450 and 1545 cmꢀ1
were used to estimate the amount of Lewis and Brønsted acid
sites (LAS and BAS, Table 3). Sn-BEA-50 has the highest LAS
content among the well-defined materials. The amount of ad-
sorbed pyridine correlated well with the Sn content. For the
materials Sn-BEA-25, Sn-BEA-50, and Sn-BEA-100, the amounts
of adsorbed pyridine were 281, 530, and 306 mmolgꢀ1, respec-
We compared the catalytic performance of the most Lewis
acidic Sn-BEA-50 zeolite with Sn-BEA-HF and Sn-BEA-50* in the
conversion of DHA to methyl lactate at 708C. The reaction pro-
files are shown in Figure 8. The Sn-BEA-50 and Sn-BEA-HF zeo-
lites that predominantly contain framework Sn displayed com-
parable activities. The catalytic performance of Sn-BEA-50* was
much lower. The complete conversion of DHA into methyl lac-
tate was achieved after 2 h for Sn-BEA-50 and Sn-BEA-HF. On
the contrary, the reaction time to achieve complete DHA con-
version was 4 h for Sn-BEA-50*. The methyl lactate yield for
this catalyst was also lower (86%). One of the side products
was 2-hydroxy-3-methoxypropionaldehyde, which is the prod-
uct of the methylation of DHA, prior to the isomerization to
glyceraldehyde. Turnover number (TON), turnover frequency
(TOF), and space–time yield (STY) values are given in the Sup-
porting Information (Table S9). Sn-MCM-41, an amorphous
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
5
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Figure 9. Consecutive recycling experiments for Sn-BEA-HF and Sn-BEA-50.
Conditions: T=708C, catalyst (40 mg), 0.25m DHA in MeOH, 2.5 mL, t=1 h).
The reusability of Sn-BEA-50 and Sn-BEA-HF was determined
by recycling experiments under different conditions (1 h, 708C,
Figure 9). It can be seen that both catalysts retain their activity
during consecutive recycles.
Our data show that not all framework Sn sites are active in
isomerization catalysis. Following the speculation of Boronat
et al.[42] that only open sites are active, we undertook a compu-
tational study to compare the reaction pathway for open and
closed framework Sn sites. The conversion of DHA to methyl
lactate or lactic acid is generally believed to consist of three
steps: dehydration of DHA to pyruvaldehyde, addition of
water, and the subsequent H-shift to the lactate.[26,43] The iso-
merization of 1-hydroxy-1-methoxypropanone (HMP), the
hemiketal intermediate in DHA conversion, involves a H-shift
reaction.[12,26] This step has previously been considered as the
rate-determining step in the overall conversion of DHA to lac-
tates.[43] This important elementary reaction step is also rele-
vant to the Bayer–Villiger oxidation reaction,[42] and the H-shift
reaction in the isomerization of larger carbohydrates.[44] Recent-
ly, an elegant experiment by Davis and co-workers involving
the exchange of the slightly acidic protons at the open Sn-OH
site for Na+ has shown the importance of such open sites for
glucose isomerization.[44] We investigated HMP isomerization to
methyl lactate (Scheme 2) by DFT calculations. Reaction energy
Figure 8. a) Conversion of DHA (dashed line) to methyl lactate (solid line) as
a function of the reaction time (T=708C, catalyst (40 mg), 0.25m DHA in
MeOH, 2.5 mL). b) Methyl lactate yield as a function of reaction time under
the same conditions.
mesoporous Sn-containing silica with a similar Sn content as
Sn-BEA-50, exhibited the lowest activity and methyl lactate se-
lectivity (see the Supporting Information, Figure S6). The
methyl lactate yield after 4 h was only 54%. The main byprod-
uct for Sn-MCM-41 was pyruvaldehyde dimethyl acetal (PADA).
Stroobants et al. reported the formation of pyruvaldehyde di-
ethyl acetal during DHA conversion in ethanol for Sn-MCM-
41.[26] We verified that dealuminated BEA-50 (Deal-BEA-50) did
not show any conversion of dihydroxyacetone after 2 h under
the same conditions (708C, 2.5 mL 0.25m DHA in methanol,
40 mg catalyst).
It should be noted that the performance of all of the well-
defined postsynthesized zeolites was similar, regardless of the
Sn content. Accordingly, the TOFs for the zeolites with a higher
Sn content will be lower at the same weight-based rate (Table
S9). Such activity trends have been reported before for the ap-
plication of Sn-Beta in catalysis.[34,40] The lower activity at high
Sn content is usually attributed to the negative effect of EFSn
species. Our work shows that such effects also occur with
EFSn-free Sn-Beta, therefore, we speculate that not all Sn spe-
cies are active in the reaction. This has also been suggested by
Boronat et al. who studied the nature of the active Sn-sites for
the Bayer–Villiger oxidation of adamantanone.[42] This finding
may be due to the different activities of open and closed Sn
framework sites.
Scheme 2. H-shift of HMP to methyl lactate.
diagrams were constructed for the partially hydrolyzed SnOH
[open site, (SiFOFH)3SnOH] and for the closed site [black,
Sn(OFSiF)4]. The results of these calculations, including the opti-
mized geometries of key intermediates and transition states,
are given in Figures 10 and 11.
The first step is the adsorption of HMP by the carbonyl
group (O2) to the Lewis acidic Sn center. For the closed SnIV
site, this results in the formation of a hydrogen bond between
&
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
6
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
only 40 kJmolꢀ1. In this case, a water molecule is formed that
binds to the Sn center. Hydrogen bonding between the coordi-
nated water and the O2 carbonyl group further stabilizes this
intermediate. The reaction proceeds by a concerted C1–C2 H-
shift and protonation of the O2 site by the coordinated water
molecule (Figure 11). A similar mechanism was previously pro-
posed as the most favorable pathway for the aldose–ketose
isomerization reaction by Sn-containing zeolites.[45] The con-
certed one-step conversion of the O2-deprotonated HMP to
methyl lactate is exothermic by ꢀ55 kJmolꢀ1 and proceeds
with an activation barrier of 72 kJmolꢀ1. This value is in good
agreement with the apparent activation energy determined for
DHA conversion for Sn-BEA-50 zeolite (Supporting Information,
S8).
Figure 10. The comparison of DFT-calculated reaction energy diagrams for
the isomerization of HMP to ML following the direct Lewis acid catalyzed
mechanism over the closed Sn site (black, Sn(OFSiF)4) and the cooperative
OH-assisted mechanism over the open SnOH site (red, (SiFOF)3SnOH).
Conclusions
An improved method for the preparation of nanosized Sn-Beta
zeolite containing almost exclusively framework Sn-sites is re-
ported. The method consists of the impregnation of dehydrat-
ed dealuminated Beta-zeolite with excess anhydrous SnCl4 in
an inert atmosphere, followed by careful washing with metha-
nol and calcination. The washing step is essential to avoid the
formation of EFSn species. A further advantage is that larger
amounts of Sn can be introduced in Sn-Beta than by other
methods, such as direct fluoride-mediated Sn-Beta synthesis.
The strong Lewis acid site density correlated well with the Sn
content for the well-defined Sn-Beta samples. Such samples
were able to rapidly convert DHA into methyl lactate without
byproduct formation, attaining productivity similar to hydro-
thermally synthesized Sn-Beta. TOF values, however, decreased
with increasing Sn-content. When EFSn sites were present, the
performance of the postsynthesized materials was much lower
and byproducts were formed. Nanosized Sn-Beta with exclu-
sive framework Sn centers showed similar performance com-
pared to Sn-Beta prepared by conventional methods and
much better performance compared to Sn-MCM-41. DFT calcu-
lations supported the proposal that the open framework Sn
sites act in a cooperative manner with adjacent proton-donat-
ing groups to catalyze the isomerization of 1-hydroxy-1-
methoxypropanone to methyl lactate. The activation barrier for
this H-shift step is lower than that for the closed site, which
does not contain an hydroxyl group.
Figure 11. Optimized structures and key geometrical parameters of the reac-
tion intermediates and transition states for the isomerization of HMP to
methyl lactate over the open SnOH site in a periodic Sn-BEA model (only
atoms of the active site complex with DHA and those in its first coordination
sphere are shown).
the O1ꢀH hydroxyl and a vicinal lattice oxygen, which bridges
between the framework Sn and Si atoms. Deprotonation of
O1ꢀH by the weakly basic lattice O site is endothermic (DE=
52 kJmolꢀ1), and it results in a cationic intermediate, bidentate-
ly bound to the Sn center by the O1 and O2 atoms. This spe-
cies can then be transformed to methyl lactate by a C1–C2 H-
shift (DE=ꢀ78 kJmolꢀ1
)
and O2 reprotonation (DE=
ꢀ51 kJmolꢀ1). Because of the high endothermicity of the first
deprotonation step, the overall barrier for the methyl lactate
Experimental Section
formation is 113 kJmolꢀ1
.
The reaction proceeds in a more favorable manner when it
is catalyzed by the open Sn-OH site. The most significant dif-
ference is observed in the energetics of the initial O1ꢀH depro-
tonation. This difference relates to the higher basicity of the
terminal Sn-OH compared with that of bridging lattice oxygen
atoms. Thus, initial deprotonation of HMP by Sn-OH is exother-
mic (ꢀ23 kJmolꢀ1) and proceeds with an activation barrier of
Chemicals
Ludox SM-30 (30 wt% SiO2 in H2O, Aldrich), Al(OiPr)3 (>98%, Al-
drich), tetraethylammonium hydroxide (TEAOH, 35 wt% in H2O, Al-
drich), anhydrous SnCl4 (99%, Aldrich), tetraethylorthosilicate
(TEOS, synthesis grade, Merck), HF (40% in H2O, Merck), SnCl4·5H2O
(98%, Acros), and 1,3-dihydroxyacetone (DHA, dimeric form, 97%,
Aldrich) were used as received without further purification.
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
7
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
3008C for 30 min, respectively, with a heating rate of 58Cminꢀ1
.
Synthesis of Sn-containing materials
FTIR spectra were recorded in situ at these temperatures. Back-
ground correction was performed by subtracting the spectrum of
the dehydrated wafer recorded at 1008C. Integration of the signals
for determining the amount of Lewis and Brønsted acid sites was
performed by using the Fityk curve fitting program. 1H NMR and
119Sn MAS NMR spectra were recorded by using a Bruker DMX-500
NMR spectrometer with a 4 mm zirconia rotor at a spinning rate of
Sn-Beta zeolite synthesized in fluoride media (Sn-BEA-HF)[18] and
Sn-MCM-41[46] were synthesized according to established proce-
dures. Nanosized Al-Beta zeolite precursors with synthesis Si/Al
ratio of 25, 50, and 100, denoted as BEA-25, BEA-50, and BEA-100,
respectively, were prepared according to a modified procedure re-
ported by Mintova et al.[35] A solution of TEAOH in water (28.1 g)
and Al(OiPr)3 were added Ludox SM-30 (37 g) and the mixture was
stirred for 20 h, after which a clear to slightly opaque (for higher
amounts of Al precursor) solution was formed. The mixture was
subsequently transferred to a Teflon-lined steel autoclave and
placed in an oven at 1008C for 18 days under static conditions.
The resulting zeolites were collected by centrifugation, washed
thoroughly with water, and dried in air. The zeolites were calcined
at 400 and 5508C (both 18Cminꢀ–1, 10 h), successively. Dealumina-
tion was performed by heating in 65% HNO3 (50 mLgꢀ1) at 1108C
overnight. Residual HNO3 was removed by thorough washing with
demineralized water until the pH of the washings was neutral. The
resulting materials are denoted as Deal-BEA-25, Deal-BEA-50, and
Deal-BEA-100. Prior to modification with Sn, the dealuminated BEA
zeolites were dried in vacuo at 1708C for 3 h in a Schlenk flask. Sn
was incorporated by adding an excess of approximately three sila-
nol nest equivalents of anhydrous SnCl4 at 1008C under an Ar at-
mosphere and the mixtures were stirred overnight. To remove un-
reacted SnCl4 from the zeolite pores, the materials were thorough-
ly washed with methanol at least six times and dried in air. To
remove residual chloride and to complete the condensation of the
Sn centers into the framework, the materials were recalcined at
5508C (18C minꢀ1, 5 h). The final Sn-modified catalysts are denoted
as Sn-BEA-25, Sn-BEA-50, and Sn-BEA-100.
1
10 kHz. H chemical shifts were referenced to Si(CH3)4, 119Sn chemi-
cal shifts were referenced to SnO2 (d=ꢀ604 ppm).
Catalytic activity measurements
DHA was dissolved in methanol (0.25m) and 2.5 mL of the solution,
together with the catalyst (40 mg), was added to a 4 mL thick-
walled glass vial. Reactions were performed at 708C under stirring
(500 rpm) and mixtures were analyzed by GC-FID with n-decane as
an external standard on a Shimadzu GC-17 A apparatus equipped
with a Restek Rxi-5 ms column (30 mꢁ0.25 mm, df =0.25 mm). Re-
actions used to determine the activation energy of the reaction
were performed at 50, 60, and 708C. Identification of unknown
compounds was performed on
a Shimadzu QP5050 GC-MS
equipped with a Stabilwax column (30 mꢁ0.32 mm, df =0.5 mm).
Recycle experiments were performed in a 12 mL screw-capped
glass vial. The catalyst was recovered by centrifugation after each
cycle, washed twice with methanol, and dried in air.
Computational details
DFT calculations were performed by using the Vienna ab initio Sim-
ulation Package (VASP 5.2).[47] The generalized gradient exchange-
correlation PBE functional was used.[48] The electron–ion interac-
tions were described by the projected augmented wave (PAW)
method.[49] Geometry optimization was performed with a plane-
wave basis set with a cut-off of 400 eV. The calculations were as-
sumed to be converged, when the forces on each atom were less
than 0.05 eVꢂꢀ1. Brillouin zone-sampling was restricted to the G
point. Optimized cell parameters for all-silica zeolite Beta are a=
b=12.66 ꢂ, c=26.40 ꢂ. Two candidate active sites for Sn-Beta
were considered. One is the closed SnIV site (Sn(OFSiF)4, OF and SiF
represent the framework oxygen and silicon atoms, respectively).
The other is generated by H2O dissociation over the framework Sn-
O-Si moiety, resulting in the open (SiFOF)3SnOH···HOFSiF active site.
Sn atoms were introduced at T6 tetrahedral site of the BEA frame-
work. The R-HMP and R-methyl lactate were used as the initial and
final states to study the mechanism of the intramolecular hydride-
shift reaction. The minimum reaction energy path and correspond-
ing transition state were determined by using the nudged-elastic
band method (NEB) with improved tangent estimate.[50] The maxi-
mum energy geometry along the reaction path obtained with the
NEB method was further optimized by using a quasi-Newton algo-
rithm. In this step, only the extraframework atoms and the atoms
in the first coordination sphere of Sn were relaxed. Frequency anal-
ysis of the stationary points was performed by means of the finite
difference method as implemented in VASP. Small displacements
(0.02 ꢂ) were used to estimate the numerical Hessian matrix. The
transition states were confirmed by the presence of a single imagi-
nary frequency corresponding to the reaction coordinate. To ac-
count for the van der Waals (vdW) interactions between zeolite
voids and carbohydrates, all calculations were carried out with the
DFT-D2 method[51] as implemented in VASP 5.2.
Characterization
X-ray diffraction patterns were recorded on a Bruker D4 Endeavor
diffractometer by using CuKa radiation in the 2q range 5–608 with
a scanning speed of 0.018sꢀ1. Elemental analyses were carried out
by using a Spectro Ciros CCD ICP optical emission spectrometer
with axial plasma viewing. For analysis of the materials HF/HNO3/
H2O 1:1:1 was used as matrix. Thermogravimetric analyses to deter-
mine the water content of the solid samples were performed on
a Mettler TGA/DSC-1 apparatus using 70 mL alumina crucibles.
Helium was used at a gas flow rate of 40 mLminꢀ1. Argon physi-
sorption measurements were performed at ꢀ1868C on a Micromer-
itics ASAP2020 apparatus in static measurement mode. In a typical
experiment, a zeolite sample (typically 100 mg) was outgassed at
2008C for 8 h prior to the measurement. The Brunauer–Emmett–
Teller equation was used to calculate the specific surface area from
the adsorption data in the p/p0 range 0.05–0.35. The mesopore
volume was calculated by applying the Barrett–Joyner–Halenda
(BJH) method on the adsorption branch of the isotherm and the
micropore volume was calculated by using the t-plot method. FTIR
spectra of pyridine adsorbed to the zeolites were recorded by
using a Bruker Vertex V70v system. The spectra were acquired as
an average of 32 scans at a resolution of 2 cmꢀ1 The samples were
prepared as thin self-supporting wafers of 5–15 mgcmꢀ2 density
and placed inside a controlled-environment infrared transmission
cell. Prior to pyridine adsorption, the catalyst wafer was heated to
5508C at a rate of 28Cminꢀ1 under vacuum (p<5ꢁ10ꢀ6 mbar) and
maintained at this temperature for 3 h. After cooling to 1008C, pyr-
idine was introduced in the cell in static vacuum for 30 min. Physi-
sorbed pyridine was removed in vacuo for 30 min. The evacuated
sample containing chemisorbed pyridine was subjected to a tem-
perature-programmed desorption sequence at 100, 200, and
&
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
8
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
[24] P. Y. Dapsens, M. J. Menart, C. Mondelli, J. Pꢄrez-Ramꢅrez, Green Chem.
2014, 16, 589–593.
Acknowledgements
[25] F. de Clippel, M. Dusselier, R. Van Rompaey, P. Vanelderen, J. Dijkmans, E.
Makshina, L. Giebeler, S. Oswald, G. V. Baron, J. F. M. Denayer, P. P. Pes-
carmona, P. A. Jacobs, B. F. Sels, J. Am. Chem. Soc. 2012, 134, 10089–
10101.
[26] L. Li, C. Stroobants, K. Lin, P. A. Jacobs, B. F. Sels, P. P. Pescarmona, Green
Chem. 2011, 13, 1175–1181.
[27] P. Li, G. Liu, H. Wu, Y. Liu, J.-g. Jiang, P. Wu, J. Phys. Chem. C 2011, 115,
3663–3670.
[28] M. Liu, S. Jia, C. Li, A. Zhang, C. Song, X. Guo, Chin. J. Catal. 2014, 35,
723–732.
This work has been funded by the Joint Scientific Thematic Re-
search Programme of the Netherlands Organization for Scientific
Research (NWO) and the Chinese Ministry of Science and Technol-
ogy, as well as supported by the European Union Project NOVA-
CAM (Novel cheap and abundant materials for catalytic biomass
conversion, FP7-NMP-2013-EU-Japan-604319. NWO and SurfSARA
are acknowledged for providing access to supercomputer facili-
ties.
[29] J. Jin, X. Ye, Y. Li, Y. Wang, L. Li, J. Gu, W. Zhao, J. Shi, Dalton Trans.
2014, 43, 8196–8204.
[30] P. Wolf, C. Hammond, S. Conrad, I. Hermans, Dalton Trans. 2014, 43,
4514–4519.
[31] B. Tang, W. Dai, G. Wu, N. Guan, L. Li, M. Hunger, ACS Catal. 2014, 4,
2801–2810.
Keywords: carbohydrates
mechanisms · tin · zeolites
·
green chemistry
·
reaction
[32] P. Y. Dapsens, C. Mondelli, J. Jagielski, R. Hauert, J. Pꢄrez-Ramꢅrez, Catal.
Sci. Technol. 2014, 4, 2302–2311.
[33] Q. Guo, F. Fan, E. A. Pidko, W. N. P. van der Graaff, Z. Feng, C. Li, E. J. M.
Hensen, ChemSusChem 2013, 6, 1352–1356.
[34] J. Dijkmans, D. Gabriꢆls, M. Dusselier, F. de Clippel, P. Vanelderen, K.
Houthoofd, A. Malfliet, Y. Pontikes, B. F. Sels, Green Chem. 2013, 15,
2777–2785.
[1] J. J. Bozell, G. R. Petersen, Green Chem. 2010, 12, 539–554.
[2] G. W. Huber, A. Corma, Angew. Chem. Int. Ed. 2007, 46, 7184–7201;
Angew. Chem. 2007, 119, 7320–7338.
[3] G. W. Huber, S. Iborra, A. Corma, Chem. Rev. 2006, 106, 4044–4098.
[4] A. Corma, S. Iborra, A. Velty, Chem. Rev. 2007, 107, 2411–2502.
[5] M. Dusselier, P. Van Wouwe, A. Dewaele, E. Makshina, B. F. Sels, Energy
Environ. Sci. 2013, 6, 1415–1442.
[35] S. Mintova, V. Valtchev, T. Onfroy, C. Marichal, H. Knçzinger, T. Bein, Mi-
croporous Mesoporous Mater. 2006, 90, 237–245.
[36] W. Kim, J.-C. Kim, J. Kim, Y. Seo, R. Ryoo, ACS Catal. 2013, 3, 192–195.
[37] J. C. Groen, J. A. Moulijn, J. Pꢄrez-Ramꢅrez, J. Mater. Chem. 2006, 16,
2121–2131.
[38] P. S. Niphadkar, A. C. Garade, R. K. Jha, C. V. Rode, P. N. Joshi, Micropo-
rous Mesoporous Mater. 2010, 136, 115–125.
[39] W. R. Gunther, V. K. Michaelis, M. A. Caporini, R. G. Griffin, Y. Romꢃn-Lesh-
kov, J. Am. Chem. Soc. 2014, 136, 6219–6222.
[40] P. Wolf, M. Valla, A. J. Rossini, A. Comas-Vives, F. NuÇez-Zarur, B. Mala-
man, A. Lesage, L. Emsley, C. Copꢄret, I. Hermans, Angew. Chem. Int. Ed.
2014, 53, 10179–10183; Angew. Chem. 2014, 126, 10343–10347.
[41] J. Datka, A. M. Turek, J. M. Jehng, I. E. Wachs, J. Catal. 1992, 135, 186–
199.
[42] M. Boronat, P. Concepcion, A. Corma, M. Renz, S. Valencia, J. Catal.
2005, 234, 111 – 118.
[43] R. S. Assary, L. A. Curtiss, J. Phys. Chem. A 2011, 115, 8754–8760.
[44] R. Bermejo-Deval, M. Orazov, R. Gounder, S.-J. Hwang, M. E. Davis, ACS
Catal. 2014, 4, 2288–2297.
[45] G. Li, E. A. Pidko, E. J. M. Hensen, Catal. Sci. Technol. 2014, 4, 2241–
2250.
[46] A. Corma, M. T. Navarro, L. Nemeth, M. Renz, Chem. Commun. 2001,
2190–2191.
[47] G. Kresse, J. Hafner, Phys. Rev. B 1993, 48, 13115–13118.
[48] J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996, 77, 3865–
3868.
[49] P. E. Blçchl, Phys. Rev. B 1994, 50, 17953–17979.
[50] G. Mills, H. Jꢇnsson, G. K. Schenter, Surf. Sci. 1995, 324, 305–337.
[51] S. Grimme, J. Comput. Chem. 2006, 27, 1787–1799.
[6] H. Rçper, Starch/Stꢀrke 2002, 54, 89–99.
[7] R. Garcia, M. Besson, P. Gallezot, Appl. Catal. A 1995, 127, 165–176.
ˇ
[8] E. Farnetti, J. Kaspar, C. Crotti, Green Chem. 2009, 11, 704–709.
[9] N. Wçrz, A. Brandner, P. Claus, J. Phys. Chem. C 2010, 114, 1164–1172.
[10] H. J. Kim, J. Lee, S. K. Green, G. W. Huber, W. B. Kim, ChemSusChem
2014, 7, 1051–1054.
[11] Y. Romꢃn-Leshkov, M. Moliner, J. A. Labinger, M. E. Davis, Angew. Chem.
Int. Ed. 2010, 49, 8954–8957; Angew. Chem. 2010, 122, 9138–9141.
[12] M. S. Holm, C. M. Osmundsen, S. Dahl, E. Taarning, Proc. R. Soc. London
Ser. A 2012, 468, 2000–2016.
[13] C.-C. Chang, Z. Wang, P. Dornath, H. J. Cho, W. Fan, RSC Adv. 2012, 2,
10475–10477.
[14] M. S. Holm, S. Saravanamurugan, E. Taarning, Science 2010, 328, 602–
605.
[15] E. Nikolla, Y. Romꢃn-Leshkov, M. Moliner, M. E. Davis, ACS Catal. 2011, 1,
408–410.
[16] M. Moliner, Dalton Trans. 2014, 43, 4197–4208.
[17] A. Corma, L. T. Nemeth, M. Renz, S. Valencia, Nature 2001, 412, 423–
425.
[18] A. Corma, M. E. Domine, S. Valencia, J. Catal. 2003, 215, 294–304.
[19] a) Z. Kang, X, Zhang, H. Liu, J. Qiu, K. L. Yeung, Chem. Eng. J. 2013,
218, 425–432; b) Z. Kang, X. Zhang, H. Liu, J. Qiu, W. Han, K. L. Yeung,
Mater. Chem. Phys. 2013, 141, 519–529.
[20] S. Tolborg, A. Katerinopoulou, D. D. Falcone, I. Sꢃdaba, C. M. Osmund-
sen, R. J. Davis, E. Taarning, P. Fristrupa, M. S. Holm, J. Mater. Chem. A
2014, 2, 20252–20262.
[21] R. Bermejo-Deval, R. Gounder, M. E. Davis, ACS Catal. 2012, 2, 2705–
2713.
[22] C. Hammond, S. Conrad, I. Hermans, Angew. Chem. Int. Ed. 2012, 51,
11736–11739; Angew. Chem. 2012, 124, 11906–11909.
[23] J. Wang, J. Ren, X. Liu, J. Xi, Q. Xia, Y. Zu, G. Lu, Y. Wang, Green Chem.
2012, 14, 2506–2512.
Received: December 23, 2014
Published online on && &&, 0000
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
9
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

FULL PAPERS
W. N. P. Van der Graaff, G. Li, B. Mezari,
E. A. Pidko,* E. J. M. Hensen*
Only in the framework! Extraframe-
work Sn (EFSn) species in Sn-Beta
zeolites have a negative impact on the
activity and selectivity of the catalyst for
the conversion of 1,3-dihydroxyacetone
into methyl lactate. The reported
method is specifically designed to avoid
EFSn formation during the synthesis of
Sn-Beta zeolites. In this way, the activity
and selectivity of the catalysts are signif-
icantly improved.
&& – &&
Synthesis of Sn-Beta with Exclusive
and High Framework Sn Content
&
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
10
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!