Propylene epoxidation with O2 and H2: A high-performance Au/TS-1 catalyst prepared via a deposition-precipitation method using urea

Article abstract of DOI:10.1039/c3cy00339f

Gold nanoparticles could only be highly dispersed onto the as-synthesized TS-1 zeolite (TS-1-syn: with template) rather than the calcined one (TS-1-calc: template removed), via a deposition-precipitation method using urea. The Au-1/TS-1-syn (Au loading: 1 wt%) is promising for propylene epoxidation with O₂ and H₂, delivering a conversion of 8.1% with ～82% selectivity to propylene epoxide (PO). The pores of TS-1 may not be essential for PO synthesis as they are blocked by the organic template.

Full text of DOI:10.1039/c3cy00339f

Catalysis
Science & Technology
View Article Online
View Journal | View Issue
COMMUNICATION
Propylene epoxidation with O and H : a high-
2
2
performance Au/TS-1 catalyst prepared via a
deposition–precipitation method using urea†
Cite this: Catal. Sci. Technol., 2013, 3,
906
2
Received 16th May 2013,
Accepted 14th June 2013
Xinnan Lu, Guofeng Zhao and Yong Lu*
DOI: 10.1039/c3cy00339f
www.rsc.org/catalysis
8
Gold nanoparticles could only be highly dispersed onto the
as-synthesized TS-1 zeolite (TS-1-syn: with template) rather than
the calcined one (TS-1-calc: template removed), via a deposition–
precipitation method using urea. The Au-1/TS-1-syn (Au loading:
Ba(NO ) ), and the preparation of Au-particles with controlled
3
2
size and location.
Among these catalysts, the Au supported on TS-1 showed
promising catalytic activity with high selectivity and relatively
1
wt%) is promising for propylene epoxidation with O
2
and H
2
,
good stability in the epoxidation of propylene with O
presence of H
2
2
in the
. Delgass et al. have studied the influence of
the precipitant, the pretreatment conditions and the support
9
delivering a conversion of 8.1% with ∼82% selectivity to propylene
epoxide (PO). The pores of TS-1 may not be essential for PO syn-
thesis as they are blocked by the organic template.
1
0
in the Au/TS-1 catalyst system, the kinetics and the forma-
tion of hydrogen peroxide over Au cluster by density func-
1
1–13
Propylene epoxide (PO) is an important organic chemical
intermediate which is mainly used for the production of
polyether polyols and propylene glycol. In principle, PO can
be produced through stoichiometric chlorohydrin and organic
tional theory (DFT) calculation.
They showed that the
supported Au catalyst can reach a higher activity when TS-1
1
4
with poor crystallinity was used as the support, which indi-
cated that the morphology and some structural defects of TS-1
have an important impact on the performance of the catalyst.
Recently, the interest for Au catalysis has moved toward Au
cluster catalysts that are obtainable with very low Au loadings
and offer much higher intrinsic activities compared to the tra-
2 2
peroxide (later H O ) processes. However, the environmental
issues (e.g., waste and toxic chemical production) and unsat-
isfactory cost-efficiency (e.g., multi steps and market mismatch
of co-products) accompanying these processes are exigently
calling for much greener and more atom-efficient methods
1
5,16
ditional supported Au catalysts.
Haruta et al. compared
that adopt solid catalysts and O as an oxidant.
the Au/TS-1 catalysts prepared by the deposition–precipitation
(DP) and solid grinding (SG) methods, and they indicated that
the Au clusters deposited on the exterior surface were mainly
2
2
One-step catalytic epoxidation of propylene with O in the
presence of H using solid catalysts has been attracting particu-
2
1
7
lar attention as a promising route for PO production. In 1998,
Haruta et al. reported for the first time a nano-gold catalyst
responsible for PO synthesis, as they might be more active
than the tiny Au clusters incorporated into the microporous
9
supported on TiO that can selectively catalyse the epoxidation
channels.
2
1
18–20
of propylene with H and O to produce PO. From then on,
DP urea and NaOH methods
are well established tech-
2
2
supported Au NP catalysts for propylene epoxidation have been
extensively studied in order to reach the estimated industrial
targets (10% conversion and 90% selectivity), with most of the
attention focused on the screening of Ti-containing supports
niques for the deposition of either Au nanoparticles (Au NPs)
or Au clusters onto various oxide supports with point of zero
charge (PZC) in the range 6–8 rather than on ones with PZC
of ∼2. The DP urea (DPU) method has been receiving growing
2
,3
4,5
4,6
7
20–22
(including TS-1, Ti-MCM-41, Ti-MCM-48, and Ti-TUD ),
interest for the preparation of supported Au catalysts.
the tuning of chemical properties by additives (such as
Since the hydrolysis of urea can gradually release hydroxyl
ions, the pH value of the suspension can rise by degrees. As a
result, the Au can be deposited onto the support adequately
Shanghai Key Laboratory of Green Chemistry and Chemical Processes,
Department of Chemistry, East China Normal University, Shanghai 200062, China.
E-mail: ylu@chem.ecnu.edu.cn; Fax: +86 21-62233424; Tel: +86 21-62233424
20
with rare losses. In addition, DPU can avoid metal contami-
+
nants such as Na residues. However, no success has been
reported in highly and uniformly dispersing Au NPs on TS-1
zeolites with the DPU method.
†
Electronic supplementary information (ESI) available: Fig. S1 to S7, Table S1.
2
0,21
See DOI: 10.1039/c3cy00339f
2906 | Catal. Sci. Technol., 2013, 3, 2906–2909
This journal is © The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Communication
Table 1 Epoxidation of propylene with O
2
and H
2
over Au-1/TS-1a
Fig. S5 in ESI† shows the conversion and selectivity for the
d
epoxidation of propylene against the Au loading. When 1 wt
Catalyst
C
3
H
6
conv. (%)
PO select. (%)
PO yield (%)
%
Au (relative to the mass of the TS-1-syn support) was
b
e
Au-1/TS-1-syn
Au-1/TS-1-calc
8.1
0.6
81.5
6.5
0.4
c
loaded, the catalyst showed a good balance between conver-
sion and selectivity. Lower Au loadings of <1 wt% favoured
selectivity but conversion was too low while higher Au load-
ings of >1 wt% promoted the conversion but the selectivity
was seriously degraded due to over-oxidation.
65.7
a
Catalyst preparation conditions: light-free DPU, DPU time 4.5 h,
DPU temperature 90 °C, urea–Au molar ratio of 100 and calcination
in air at 300 °C for 4 h. Catalyst testing conditions: catalyst 0.3 g, 150 °C,
C
(
3
H
6
: O
GHSV) 12 000 ml h
and they were pre-reduced in H
2 2 2
: H : N = 1 : 3 : 9 : 12 (vol/vol), total gas hourly space velocity
−
1
−1
g . Au loading was 1 wt% for both catalysts
b
To gain insight into the nature of the high activity and
selectivity of the TS-1-syn supported catalyst, the real Au
loadings were determined using inductively coupled plasma
atomic emission spectroscopy (ICP-AES). It is clear from
Table S1 (ESI†) that, under optimal DPU conditions (urea–Au
molar ratio of 100, at 90 °C for 4.5 h), over 95% of Au could
be deposited onto the TS-1-syn zeolite support even for a high
Au loading of 3 wt%. In contrast, a relatively low value of
86% was obtained when depositing 1 wt% Au onto TS-1-calc
zeolite support. We wonder whether Au loss was the main
cause for such a big difference in activity and selectivity
between Au-1/TS-1-syn and Au-1/TS-1-calc. To address this
question, a 90% deposition of a 1 wt% Au loading (relative to
the support mass) onto TS-1-syn was obtained by tuning the
DPU parameters (i.e., using a low DPU temperature of 50 °C
or a short DPU time of 1 h; Table S1, ESI†), which is very
close to the 86% obtained for Au-1/TS-1-calc. However, these
two samples still delivered much better reactivity than the
Au-1/TS-1-calc. (0.6% conversion, 65.7% selectivity): 5.8%
conversion and 84.3% selectivity for 50 °C DPU temperature
and 4.3% conversion and 81.1% selectivity for 1 h DPU time
(Fig. S4 (ESI†) and Table 1). Moreover, over the Au-0.5/TS-1-syn
(Au loading of 0.5 wt%, with 100% Au deposition as shown
in Table S1, ESI†), a conversion higher than 5% could also
be obtained with over 80% selectivity (Fig. S5, ESI†). It is thus
rational to infer, from the above results, that Au loss during
the DPU preparation was not the main cause for the reactivity
difference between Au-1/TS-1-syn and Au-1/TS-1-calc.
2
at 250 °C for 1 h before testing. Using
as-synthesized TS-1 zeolite (TS-1-syn, with template) as the support.
c
Using calcined TS-1 zeolite (TS-1-calc, template removed) as the sup-
d
e
2 2
port. Byproducts were propanol, acetone and CO . H utilization effi-
ciency, defined as PO produced per molar amount of hydrogen converted,
was estimated to be 6%.
Herein, we demonstrate, for the first time to our knowl-
edge, the successful high and uniform dispersion of Au NPs
(4–6 nm) onto an as-synthesized TS-1 zeolite by employing a
modified DPU method. Such catalysts provide a nice activity
and selectivity for the direct epoxidation of propylene with
O and H . In addition, whether or not the pores of TS-1 zeo-
2 2
lite are essential is also tentatively discussed.
As shown in Table 1, the conversion and selectivity of the
Au/TS-1 catalysts were strongly dependent on whether we
kept the organic template in the TS-1 zeolite or removed it by
calcination before using TS-1 as the support. Note that both
catalysts were activated and tested under optimal conditions,
as shown in Fig. S1–S3 and Table S1 (ESI†). The representa-
tive Au-1/TS-1-syn catalyst (1 wt% Au loading), obtained by
using as-synthesized TS-1 zeolite, could deliver a propylene
conversion of 8.1% with a PO selectivity of ∼82%, but it was
associated with a low H utilization efficiency of 6% (com-
2
1
6
pared to a reported number of ∼20% ). In contrast, the
Au-1/TS-1-calc obtained by using calcined TS-1 was neither
active nor highly selective.
The influence of the DPU parameters on the catalytic per-
formance of the TS-1-syn supported Au (1 wt%) catalysts was
carefully investigated as well as the catalyst calcination tem-
perature, with the results shown in Fig. S4 (ESI†). It is clear
that the catalyst activity was very sensitive to both the DPU
temperature and DPU time (Fig. S4A and B, ESI†) but not
to the urea–Au molar ratio (Fig. S4C, ESI†). The DPU tem-
perature of 90 °C, higher than previously reported ones
It is widely accepted that particle size of the Au catalyst is
critical for endowing them with an outstanding catalytic
activity. After ruling out the possibility of Au loss as the main
cause for the reactivity difference, we employed transmission
electron microscopy (TEM) to check the Au particle size
over as-synthesized Au/TS-1 catalysts, with the results shown
in Fig. 1. Interestingly, a high dispersion of Au could be
observed on the Au-1/TS-1-syn catalyst with a narrow Au parti-
cle distribution centred at ∼5 nm (Fig. 1A and a). In contrast,
larger Au particles were observed on the Au-1/TS-1-calc cata-
lyst with a dual size particle distribution (∼10 and ∼30 nm;
Fig. 1B and b). Clearly, a good correlation between the Au
particle size and the catalyst activity–selectivity for propylene
epoxidation using O and H can be obtained; a high activity
2
0
(
≤80 °C), was mostly preferred for activity promotion; but
a high DPU temperature of 100 °C caused a decline of the
activity (Fig. S4A, ESI†). The optimal DPU time length was
around 5 h rather than previously reported times of ≤4 h.
2
0
Increasing the DPU time from 1 to 6 h would almost double
the propylene conversion (Fig. S4B, ESI†). Thus, the opti-
mized DPU conditions were absence of light, DPU tempera-
ture 90 °C, DPU time 4.5 h and a urea–Au molar ratio of
2
2
can only be achieved when the Au particle size was reduced
to ∼5 nm. Recently, several research groups have indicated
that the activity can be visibly promoted again by further
1
00. In addition, the catalyst calcination temperature was
1
5–17
found to be most crucial for both the activity and selectivity.
The optimal calcination temperature was identified to be
reducing the Au particle size to ∼1 nm (Au cluster).
Our
observation together with the above mentioned literature
results solidly indicates that the Au/TS-1 catalysts feature a
3
00 °C (Fig. S4D, ESI†).
This journal is © The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 2906–2909 | 2907

View Article Online
Communication
Catalysis Science & Technology
catalyst was observed (Fig. 1C and c). Notably, the Au-3/TS-1-syn
catalyst also showed a unique and high dispersion of Au
NPs, with a narrow particle size distribution and an average
particle size comparable to that of the Au-1/TS-1-syn catalyst
(∼5 nm), even though the Au loading was tripled (Fig. 1D and d).
However, the Au-3/TS-1-syn catalyst provided very poor selec-
tivity for the epoxidation of propylene. For instance, at 150 °C,
a selectivity of only ∼11% was obtained with a high conver-
sion of ∼15% (Fig. S5, ESI†). This is likely due to the high
density of Au NPs which might significantly promote the deep
oxidation of H and propylene as well as the PO product with
2
2 2 x
O to form H O and CO while releasing a lot of heat. Indeed,
under identical reaction conditions, a remarkable increase of
the catalyst bed temperature was observed when using the
Au-3/TS-1-syn catalyst, compared to the catalyst bed packed
with Au-1/TS-1-syn.
In summary, a promising Au/TS-1 catalyst for the epoxida-
2 2
tion of propylene with O and H has been successfully devel-
oped by using the DPU method with only as-synthesized TS-1
zeolite as support. The presence of the organic template results
in a raised isoelectric point, which is crucial for high and uni-
form dispersion of Au on the TS-1 support. Most notably, the
pores of TS-1 may not be essential for PO synthesis as the pores
of the Au/TS-1-syn catalyst are blocked by the organic template.
We anticipate that our findings will initiate attempts to develop
high-performance gold catalysts for direct propylene epoxida-
tion and to understand gold catalysis in more detail.
Fig.
1
TEM images and Au particle size distributions. (A) Au-1/TS-1-syn,
at 250 °C for 1 h, (D) Au-3/
This work was funded by the “973 program” (2011CB201403)
from the MOST of China, and the NSF of China (21273075,
21076083, 20973063).
(B) Au-1/TS-1-calc, (C) reduced Au-1/TS-1-syn in H
2
TS-1-syn (Au: 3 wt%). Note: the statistics were based on hundreds of Au particles
in the samples except for Au-1/TS-1-calc.
Au-size-dependent activity in the direct propylene epoxidation
process.
Experimental section
Catalyst preparation
Based on the TEM results (Fig. 1A and B), we can draw the
conclusion that the organic template filling the pores of TS-1
zeolites plays a determinant role in inducing a high and uni-
form dispersion of Au onto the external surface of the TS-1
crystals. A comprehensive understanding of the positive effect
of the presence of the organic template in the zeolite pores is
particularly desirable. As noted previously, the TS-1-calc has an
The hydrothermal synthesis of TS-1 zeolite with a Si–Ti molar
ratio of 40 was performed under static conditions at 170 °C
for 48 h, basically according to the literature, from a gel
having the following molar composition: 1TEOS : 0.025TBOT :
2
3
0.18TPAOH : 18H O. Calcined TS-1 (TS-1-calc: template removed)
2
was obtained by heating the as-synthesized TS-1 (TS-1-syn)
sample in air at 550 °C for 6 h with a heating rate of 2 °C min .
8
−1
isoelectric point of 2–3, not matching the value (5–6) required
by DP methods for Au NP deposition on oxides from the pre-
Fig. S7 (ESI†) shows the XRD patterns for both as-synthesized
and calcined TS-1 samples, indicating the presence of a highly
crystalline phase evidenced by typical separated peaks.
Gold deposition onto the TS-1 zeolite was carried out by
the DPU method as previously described elsewhere. The fol-
lowing is the detailed procedure: TS-1-syn or TS-1-calc zeolite
support (1.5 g) was added to a solution (80 ml) with the
desired amount of HAuCl4 and urea. The suspension was
then heated at 50–100 °C under stirring for an appointed
period of time. The DPU process proceeded under protection
2
0
cursor HAuCl4. Interestingly, the presence of the organic tem-
plate in TS-1-syn shifted the isoelectric point close to 5 (Fig. S6,
ESI†) thereby permitting the DPU method to work efficiently.
Additionally, the organic template occupied the TS-1 zeolite
pores and therefore blocked the deposition of Au into them.
Based on the above results, it is rational for us to infer that the
pores of TS-1 may not be essential for this reaction. This agrees
with the results from Haruta in that Au clusters on the exterior
2
2
1
7
surface are mainly responsible for PO synthesis. Notably,
Delgass et al. recently showed that gold clusters inside the TS-1
channels are active for the PO reaction.
from light. After that, the solid was collected by filtration and
1
5
−
washed 5 times to thoroughly remove residual Cl (AgNO test).
3
After undergoing a pre-reduction with H2 at 250 °C for
h, no significant growth of Au NPs on the Au-1/TS-1-syn
The obtained catalyst samples were dried at 100 °C overnight
and calcined in air at 100–700 °C for 4 h.
1
2908 | Catal. Sci. Technol., 2013, 3, 2906–2909
This journal is © The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Communication
Catalyst characterization
6 B. S. Uphade, T. Akita, T. Nakamura and M. Haruta,
J. Catal., 2002, 209, 331–340.
J. Lu, X. Zhang, J. J. Bravo-Suárez, K. K. Bando, T. Fujitani
and S. T. Oyama, J. Catal., 2007, 250, 350–359.
J. Lu, X. Zhang, J. J. Bravo-Suárez, T. Fujitani and
S. T. Oyama, Catal. Today, 2009, 147, 186–195.
J. Huang, E. Lima, T. Akita, A. Guzmán, C. Qi, T. Takei and
M. Haruta, J. Catal., 2011, 278, 8–15.
The catalysts were characterized by X-ray diffraction (XRD;
Rigaku Ultima IV diffractometer [Cu-Kα]) and transmis-
sion electron microscopy (TEM; JEOL-JEM-2010 instrument
at 200 kV). The Au loading of the catalyst samples was
determined by inductively coupled plasma atomic emission
7
8
9
spectroscopy (ICP-AES) on
a Thermo IRIS Intrepid II
XSP spectrometer. Electrophoresis measurements were
undertaken using a Powereach JS94H instrument (China).
Each value was calculated as an average of five parallel
measurements.
1
1
1
1
1
0 E. E. Stangland, B. Taylor, R. P. Andres and W. N. Delgass,
J. Phys. Chem. B, 2005, 109, 2321–2330.
1 B. Taylor, J. Lauterbach, G. E. Blau and W. N. Delgass,
J. Catal., 2006, 242, 142–152.
2 D. H. Wells, W. N. Delgass and K. T. Thomson, J. Am. Chem.
Soc., 2004, 126, 2956–2962.
3 D. H. Wells, Jr., W. N. Delgass and K. T. Thomson, J. Catal.,
Reactivity test
The gas-phase epoxidation of propylene was carried out in a
fixed-bed quartz reactor with an inner diameter of 16 mm.
The test was carried out under atmospheric pressure and
2
004, 225, 69–77.
4 B. Taylor, J. Lauterbach and W. N. Delgass, Catal. Today,
007, 123, 50–58.
0
.3 g of the catalyst was charged. The catalysts were pre-
2
reduced with a H flow at 200–350 °C for 0.5–5 h, prior to the
2
15 W.-S. Lee, L.-C. Lai, M. C. Akatay, E. A. Stach, F. H. Ribeiro
reaction test. The gas effluent was quantitatively analysed by
using two on-line HP 5890 GCs equipped with TCD (Plot Q
and MS 5A parallel capillary column (DIKMA); for analysing
and W. N. Delgass, J. Catal., 2012, 296, 31–42.
1
6 J. Chen, S. J. A. Halin, E. A. Pidko, M. W. G. M. Verhoeven,
D. M. P. Ferrandez, E. J. M. Hensen, J. C. Schouten and
T. A. Nijhuis, ChemCatChem, 2013, 5, 467–478.
2 2 2 2
O , N , CO, CO and H O) and FID (60 m FFAP capillary
column; for analysing organic and O-containing organic
chemicals) detectors.
1
7 J. Huang, T. Takei, T. Akita, H. Ohashi and M. Haruta, Appl.
Catal., B, 2010, 95, 430–438.
1
8 M. Haruta, Catal. Today, 1997, 36, 153–166.
1
9 S. Tsubota, D. A. H. Cunningham, Y. Bando and M. Haruta,
in Stud. Surf. Sci. Catal., (Preparation of catalysts VI), ed. G.
Poncelet, P. Grange and P. A. Jacobs, Elsevier Science,
Amsterdam, 1995, vol. 91, pp. 227–235.
Notes and references
1
2
3
4
5
T. Hayashi, K. Tanaka and M. Haruta, J. Catal., 1998, 178,
66–575.
T. A. Nijhuis, B. J. Huizinga, M. Makkee and J. A. Moulijn,
Ind. Eng. Chem. Res., 1999, 38, 884–891.
5
20 R. Zanella, L. Delannoy and C. Louis, Appl. Catal., A,
2005, 291, 62–72.
21 S. Ma, G. Li and X. S. Wang, Chem. Eng. J., 2010, 156,
532–539.
22 G. Zhao, M. Ling, H. Hu, M. Deng, Q. Xue and Y. Lu, Green
Chem., 2011, 13, 3088–3092.
23 A. Thangaraj, M. J. Eapen, S. Sivasanker and P. Ratnasamy,
Zeolites, 1992, 12, 943–950.
N. Yap, R. P. Andres and W. N. Delgass, J. Catal., 2004, 226,
1
56–170.
C. Qi, T. Akita, M. Okumura, K. Kuraoka and M. Haruta,
Appl. Catal., A, 2003, 253, 75–89.
B. S. Uphade, M. Okumura, S. Tsubota and M. Haruta, Appl.
Catal., A, 2000, 190, 43–50.
This journal is © The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 2906–2909 | 2909