Simple and scalable preparation of highly active lewis acidic Sn-β

Article abstract of DOI:10.1002/anie.201206193

Tin goes in: The solid Lewis acid Sn-zeoliteβ is obtained in a simple and scalable procedure (see scheme). A high metal content can be obtained, without undesirable side-effects. The space-time-yields of the resulting catalysis are over one order-of-magnitude larger than those of the state-of-the-art materials for the Baeyer-Villiger oxidation of cyclohexanone and the synthesis of ethyl lactate from the triose dihydroxyacetone. Copyright

Full text of DOI:10.1002/anie.201206193

.
Angewandte
Communications
DOI: 10.1002/anie.201206193
Zeolites
Simple and Scalable Preparation of Highly Active Lewis Acidic Sn-b**
Ceri Hammond, Sabrina Conrad, and Ive Hermans*
Lewis acids are a versatile class of catalysts that exhibit
remarkable activity for a number of essential transformations,
concept, we focus on the preparation of Sn-b. This material
[4]
has tremendous potential, equivalent to TS-1, but industrial
implementation is currently hampered by the tedious syn-
thesis procedure.
[
1]
including oxidations and isomerizations. Although homo-
genous analogues are well-established, heterogeneous cata-
lysts offer several advantages for the development of more
sustainable technology in terms of facile downstream proc-
essing and process intensification. Of particular interest are
Lewis acid doped zeolites, some of which exhibit remarkable
Preliminary work focused on the efficient dealumination
of a parent Al-b zeolite. Although steaming is a well-known
3
+
method for removing framework Al , it has the disadvantage
of leaving behind ill-defined extra-framework Lewis acidic
[
2]
3+
activity, selectivity, and lifetime for a number of processes.
Al species, which could affect the catalytic performance of
IV
The development of TS-1 (a Ti -doped MFI-type zeolite) is,
for instance, viewed as one of the greatest breakthroughs in
sustainable chemistry in the last few decades, having resulted
the material. In view of this problem, an acidic pre-treatment
À1
with HNO (13m, 1008C, 20 h, 20 mLg ) was performed to
3
[
5]
extract and remove the aluminum quantitatively. The results
in Table 1 show indeed that nearly all Al can be removed,
[3]
3+
in a “greener” process for the epoxidation of propylene,
amongst others. Promising results have also been obtained in
IV
the development of Sn -doped zeolite b, which has shown
[
a]
Table 1: Physicochemical properties of the materials.
unparalleled activity and selectivity for the isomerization of
glucose to fructose and the Baeyer–Villiger oxidation of
ketones to lactones using H O as green oxidant.
Entry Catalyst
Treatment
S
Vmicro
SiO2
SiO2
BET
2
À1
À3 À1
[3,4]
[m g
]
[cm
g
]
/Al
2
O
3
/SnO
2
2
2
Lewis acid doped zeolites are typically obtained by direct
incorporation of the Lewis acid into the framework during
A
B
C
D
H-b
deAl-b
Sn/deAl-b H /SSIE
Sn/deAl-b H /SSIE
–
H
600
620
610
600
0.17
0.18
0.17
0.17
25
>1900
>1900 32
>1900 16
–
–
+
+
[
2–4]
hydrothermal synthesis.
However, it remains a challenge
+
to obtain a significant number of isolated sites within the
structures, without the undesirable formation of metal oxide
particles which are significantly less active. Moreover, even
under optimized conditions, the incorporation of large Lewis
[a] H-b=commercial H-b-zeolite, deAl-b=dealuminated b-zeolite, Sn/
deAl-b=Sn-b-zeolite, prepared by SSIE, SBET =Brunauer-Emmett-Teller
surface area, VMicro =micropore volume.
IV
acidic centers, such as Sn , typically leads to a significant
retardation of the zeolite nucleation, and hence long synthesis
timescales (up to 40 days) and unfavorably large crystals. To
facilitate the crystallization process, additives, such as HF, are
also commonly added to the synthesis gel, posing additional
practical and environmental limitations. This, in combination
with the limited amount of active metal that can be
incorporated into the structures (under 2 wt.%) currently
limits the large-scale applicability of these otherwise promis-
ing materials.
without destruction of the framework or significant alter-
ations to its textural properties. The removal of framework Al
is exemplified by the loss of the Brønsted acidity, that is, the
À1
sharp IR signal at 3610 cm (Figure 1A,B). In its place,
À1
a broad absorbance around 3500 cm appears, confirming the
successful formation of silanol nests and vacant T-sites for the
IV [6]
incorporation of Sn .
With these limitations in mind, we aimed to develop
a convenient post-synthetic route for the incorporation of
various Lewis acid centers into zeolitic frameworks. An
attractive route involves the incorporation of the desired
transition-metal ions into the vacant tetrahedral (T)-sites of
a pre-dealuminated zeolite. Not only does this approach avoid
the long synthesis times associated with the conventional
hydrothermal synthesis routes, but it also allows for the
synthesis of a material with significantly smaller crystallite
sizes than possible through direct synthesis. As a proof of
[
*] Dr. C. Hammond, S. Conrad, Prof. Dr. I. Hermans
Department of Chemistry and Applied Bio-Sciences, ETH Zurich,
Wolfgang-Pauli-Strasse 10, 8093 Zurich (Switzerland)
E-mail: hermans@chem.ethz.ch
Homepage: http://www.hermans.ethz.ch
Figure 1. IR spectra of A) H-b, B) dealuminated b (deAl-b), C) 5 wt%
Sn/deAl-b, and D) 10 wt% Sn/deAl-b (see Table 1).
[
**] This work was supported by ETH Grant ETH-38 12-1.
1
1736
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 11736 –11739

Angewandte
Chemie
IV
Subsequently, methods of introducing Sn into the vacant
T-sites were considered. Although liquid-phase routes (e.g.
impregnation) offer a convenient and scalable route for the
deposition of tin, the solvation shell surrounding the tin
cations could potentially lead to significant diffusion limita-
tions within the zeolite micropores, negatively affecting the
incorporation of tin and resulting in poor dispersion. Fur-
thermore, hydrolysis of the tin precursor may also result in the
partial formation of bulk oxide species in the final material.
Although traditional gas–solid deposition (e.g. chemical
vapor deposition; CVD) offers exciting possibilities of
incorporating “naked” metal atoms into the T-sites, the low
volatility of typical Sn precursors (e.g. SnCl ) results in
4
difficulties grafting sufficient quantities of Sn into the final
material. Moreover, previous work has demonstrated that
a multitude of Sn species are formed by SnCl grafting, the
4
majority of which appear to be extra-framework species or
[
7]
bulk tin oxides. The catalytic activity of such materials is
therefore rather poor.
Solid-state ion-exchange (SSIE) was therefore considered
to be an interesting route for incorporating the desired
amount of metal into the structure, with both high dispersion
and homogeneity. The procedure simply involves mechanical
grinding of dealuminated zeolite and the appropriate pre-
cursor, in this case tin(II)acetate, prior to calcination at 5508C
for removal of the residual organic species.
A combination of spectroscopic techniques was used to
determine the nature of the Sn species generated post-
synthetically. The incorporation of Sn into the dealuminated
framework can be detected by IR spectroscopy as a result of
the closure of the silanol nests (Figure 1C,D). Diffuse
reflectance spectroscopy also reveals a sharp UV absorbance
around 216 nm, even for very high initial loadings of
tin(II)acetate (Figure 2a, spectrum B); this signal is charac-
Figure 2. a) UV/Vis spectra of A) conventional Sn-b (1.5 wt%),
B) 10 wt% Sn/deAl-b, and C) 10 wt% Sn(O )/deAl-b. b) Raman spectra
of 1) deAl-b, 2) 5 wt% Sn/deAl-b, 3) 10 wt% Sn/deAl-b, and 4) bulk
2
SnO2.
IV
teristic of isolated, tetrahedral Sn species within the zeolite
framework. No SnO2 could be detected with UV/Vis or
Raman spectroscopy (Figure 2b). It should be noted that the
UV/Vis spectra of the post-synthetically prepared samples are
almost identical to that of a reference Sn-b sample prepared
by direct hydrothermal synthesis (Figure 2a, spectrum A).
This spectroscopic data indicates that Sn sites similar to
those of Sn-b are obtained even at around eight-times the
total loading.
Table 2: Catalytic activity of various tin-containing samples for the
Baeyer–Villiger oxidation of cyclohexanone.
[a]
IV
Entry
Catalyst
Conv.
%]
Selec.
[%]
Yield
[%]
[
The catalytic efficiency of the prepared materials was
evaluated for the Baeyer–Villiger oxidation of cyclohexanone
1
2
3
4
5
5 wt% Sn/deAl-b
10 wt% Sn/deA-lb
10 wt% Sn/deAl-b (3ꢀ)
35.0
41.1
39.8
0
75
93
91
–
26.2
38.2
36.2
0
[
b]
IV
with H O (Table 2). As can be seen, the SSIE of Sn into
2
2
10 wt% Sn(O )/deAl-b
2
dealuminated b (entries 1 and 2) results in the formation of an
active catalyst, comparable in both yield and turnover number
10 wt% Sn/b
0
–
0
[a] Cyclohexanone in 1,4-dioxane (0.33m), H O (30 wt% solution,
2 2
(
TON) to that reported for hydrothermally synthesized
H O /ketone=1.5), Sn content=1 mol%, 908C, 4 h. [b] Third use of the
catalyst.
[4b]
2
2
materials.
Interestingly, the 5 and 10 wt% Sn/deAl-b
IV
show comparable activity per Sn site, but lead to a different
lactone selectivity. We attribute this selectivity to the presence
of un-closed silanol nests in the 5 wt% sample, resulting in the
presence of residual Brønsted acid sites. Such sites are
capable of catalyzing the formation of 6-hydroxycaproic
acid by acid-catalyzed hydrolysis, as confirmed by lactone
stability studies. At higher Sn loadings, a larger fraction of the
silanol nests are closed, thus decreasing the number of
available Brønsted acid sites and resulting in increased
selectivity. Importantly, the catalyst was also found to be
reusable (entry 3), confirming that the Sn sites and catalyst
structure are highly stable under the reaction conditions.
To further substantiate the hypothesis that isolated,
tetrahedral Sn species are indeed the active sites formed
in the post-synthetic material, we verified that the dispersion
of SnO2 into/onto the dealuminated framework (Table 2,
IV
Angew. Chem. Int. Ed. 2012, 51, 11736 –11739
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11737

.
Angewandte
Communications
entry 4) leads to zero catalytic activity. Similarly, the dis-
persion of tin(II)acetate into/onto a non-dealuminated frame-
work (Table 2, entry 5) also leads to an inactive material;
block towards renewable and biodegradable solvents and
[
9]
polymers. Recently, it has been reported that tin(IV)-
containing compounds are amongst the most active and
IV
[10]
clearly, when the formation of framework Sn sites is
selective materials for these reactions, and that a combina-
prohibited (either by an inappropriate precursor or the lack
of vacant framework sites), an inactive catalyst is obtained.
To confirm the heterogeneous nature of the catalytic
reaction, a hot-filtration test was performed, demonstrating
that removal of the catalyst after 1 h (i.e., after ca. 15%
conversion) leads to complete termination of the reaction
tion of Lewis acidity and mild Brønsted acidity is advanta-
geous for this reaction. As can be seen from the results in
Table 4, along with proceeding at exceptional levels of
[
11]
[8]
Table 4: Conversion of dihydroxyacetone into ethyl lactate.
(
Figure 3).
[
a]
Entry Catalyst
S(lactate) (%) TON STY
Ref.
1
2
3
4
3.9 wt% Sn/MCM-41 98 30 195
[10]
[12]
[11]
1.6 wt% Sn-b
Sn-carbon-silica
10 wt% Sn/deAl-b
>99
100
120 70
350 200
>99
250 1050 this work
À1
À1
[
a] Calculated as “g (lactone produced)kg (catalyst)h ”.
selectivity, the 10 wt% Sn/deAl-b catalyst is significantly
more productive than the benchmark materials in terms of
turnover numbers and space-time-yield.
In conclusion, a convenient route for the preparation of
Lewis acidic Sn-b has been developed. In addition to
exhibiting comparable or higher levels of catalytic activity
and selectivity to the state-of-the-art materials, significantly
higher space-time-yields can be obtained through the prep-
aration of a high-metal-content material. Furthermore, the
developed procedure requires significantly less synthesis time
and produces no toxic waste in comparison to the benchmark
process. We expect such a straightforward approach to
facilitate the utilization of the same or similar materials on
a large scale.
Figure 3. A) Time-dependent formation of e-caprolactone in the pres-
ence of 10 wt% Sn/deAl-b, and B) hot-filtration test for the catalyst
(
conditions, see Table 2).
Comparing the SSIE synthesized material with the hydro-
thermally prepared material, shows both give comparable
levels of lactone- and H O -based selectivity (Table 3).
2
2
[
a]
Table 3: Comparison of post-synthetic and hydrothermal routes.
Experimental Section
[
b]
Entry Catalyst
S(lactone) [%] S(H O ) [%] STY
Ref.
[4b]
2
2
Commercial zeolite H-b (ZeoChem) was dealuminated by treatment
À1
1
3
1.5 wt% Sn-b
10 wt% Sn/deAl-b 93
>98
>95
>95
374
in HNO solution (13m) at 1008C for 20 h (20 mLg (zeolite)). SSIE
3
1075 this
work
was performed by grinding the appropriate amount of tin(II)acetate
with the required amount of dealuminated zeolite for 15 min.
Samples were calcined in an air flow at 5508C. Conventional Sn-b
was synthesized according to the original procedure of Corma et al.
FT-IR spectroscopy was performed on a self-supporting wafer using
[
a] Comparison under identical experimental conditions; 0.33m ketone
in 1,4-dioxane, 908C, 3 h. [b] Space-time-yield calculated as “g (lactone
[4b]
À1
À1
produced)kg (catalyst)h ”.
a Bruker Alpha Spectrometer in transmission mode (resolution of
À1
2
cm ). Intensities were normalized to the Si-O-Si overtones of the
zeolite framework. UV/Vis analysis was performed with an Ocean
Optics UV/Visible Spectrophotometer in diffuse reflectance mode.
Si, Al, and Sn contents were determined by ICP-AES. Porosimetry
measurements were performed on a Micromimetics 2000 apparatus.
The samples were degassed prior to use (2758C, 3 h). Adsorption
isotherms were obtained at 77 K and analyzed using BET and t-plot
methods.
Baeyer–Villiger oxidation of cyclohexanone was carried out in
a 50 mL round-bottomed flask equipped with a reflux condenser. The
vessel was charged with the reactant solution (5 mL, 0.33m cyclo-
hexanone in 1,4-dioxane) and the desired amount of catalyst
However, the Sn/deAl-b is significantly more productive, as
shown by the approximately three-fold increase in the
observed space-time-yield (STY). This increase is predom-
inantly due to the successful incorporation of such high
IV
quantities of Sn , without causing significant unwanted side
effects, such as agglomeration, increased crystallite sizes, and
leaching. This improved reaction efficiency, in combination
with the facile and scalable synthesis procedure is expected to
promote the industrial applicability of Sn-b.
To explore the general applicability of the post-synthetic
material, we studied the conversion of the triose sugar
dihydroxyacetone into ethyl lactate, a convenient building
(corresponding to 1 mol% Sn relative to ketone). The vessel was
heated to the reaction temperature (908C) for 15 min, prior to the
addition of H O at a final concentration of 0.5m (H O /ketone = 1.5)
2
2
2
2
and stirred vigorously for the required reaction period. Samples were
taken periodically and quantified by GC-FID against a biphenyl
1
1738 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 11736 –11739

Angewandte
Chemie
internal standard (30 m FFAP column). H O concentrations were
Domine, L. Nemeth, S. Valencia, J. Am. Chem. Soc. 2002, 124,
3194 – 3195.
2
2
4
+
determined by titration against Ce . The conversion of dihydrox-
yacetone (DHA) into ethyl lactate was performed in an autoclave
reactor at 20 bar N2 pressure. The vessel was charged with DHA
solution (25 mL; 0.4m in EtOH plus 0.3 mol% Sn, relative to DHA),
and the reaction performed at 908C. The reactant and product were
quantified against a biphenyl internal standard by means of GC-FID
[3] F. Cavani, J. H. Teles, ChemSusChem 2009, 2, 508 – 534.
[4] a) M. Moliner, Y. Roman-Leshkov, M. E. Davis, Proc. Natl.
Acad. Sci. USA 2010, 107, 6164 – 6168; b) A. Corma, L. T.
Nemeth, M. Renz, S. Valencia, Nature 2001, 412, 423 – 425; c) M.
Renz, T. Blasco, A. Corma, V. Fornes, R. Jensen, L. Nemeth,
Chem. Eur. J. 2002, 8, 4708 – 4717.
(30 m FFAP column) and HPLC (nucleodur 100–5 NH2-RP column).
[
[
[
[
[
5] S. Dzwigaj, M. J. Peltre, P. Massiani, A. Davidson, M. Che, T.
Sen, S. Sivasanker, Chem. Commun. 1998, 87 – 88.
6] P. Wu, T. Komatsu, T. Yashima, J. Phys. Chem. 1995, 99, 10923 –
Received: August 2, 2012
Published online: October 8, 2012
10931.
7] P. Li, G. Liu, H. Wu, Y. Liu, J.-g. Jiang, P. Wu, J. Phys. Chem. C
2011, 115, 3663 – 3670.
8] R. A. Sheldon, M. Wallau, I. Arends, U. Schuchardt, Acc. Chem.
Res. 1998, 31, 485 – 493.
9] M. S. Holm, S. Saravanamurugan, E. Taarning, Science 2010, 328,
602 – 605.
Keywords: Lewis acid catalysis · selective oxidation ·
.
sustainable chemistry · triose isomerization · zeolite synthesis
[
1] a) Y. Romꢀn-Leshkov, M. E. Davis, ACS Catal. 2011, 1, 1566 –
580; b) A. Corma, H. Garcia, Chem. Rev. 2003, 103, 4307 – 4365.
1
[10] L. Li, C. Stroobants, K. Lin, P. A. Jacobs, B. F. Sels, P. P.
Pescarmona, Green Chem. 2011, 13, 1175 – 1181.
[11] 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. Pescarmona, P. A. Jacobs, B. F. Sels, J. Am.
Chem. Soc. 2012, 134, 10089 – 10101.
[
2] a) C. Hammond, M. M. Forde, M. H. Ab Rahim, A. Thetford, Q.
He, R. L. Jenkins, N. Dimitratos, J. A. Lopez-Sanchez, N. F.
Dummer, D. M. Murphy, A. F. Carley, S. H. Taylor, D. J. Willock,
E. E. Stangland, J. Kang, H. Hagen, C. J. Kiely, G. J. Hutchings,
Angew. Chem. 2012, 124, 5219 – 5223; Angew. Chem. Int. Ed.
2
012, 51, 5129 – 5133; b) P. Wu, T. Tatsumi, T. Komatsu, T.
[12] E. Taarning, S. Saravanamurugan, M. S. Holm, J. Xiong, R. M.
West, C. H. Christensen, ChemSusChem 2009, 2, 625 – 627.
Yashima, J. Catal. 2001, 202, 245 – 255; c) A. Corma, M. E.
Angew. Chem. Int. Ed. 2012, 51, 11736 –11739
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11739