HZSM-5 supported silica-zirconia and its application in one-step synthesis of methyl isobutyl ketone

Article abstract of DOI:10.1016/j.catcom.2009.10.022

In order to adjust the acidity and texture property of HZSM-5, SiO₂-ZrO₂ was introduced via a modified sol-gel method. Pd-loaded catalyst was prepared and its catalytic activity was tested in one-step synthesis of MIBK. The catalyst exhibited a significant activity of acetone conversion up to 33.4% and MIBK selectivity up to 88.1% under atmospheric pressure. The acidity and textural properties of support material, as well as palladium dispersion of Pd-loaded catalyst were characterized. The results indicated that the proper acid strength and textural property of the support, together with the high dispersion of palladium resulted in good catalytic performance.

Full text of DOI:10.1016/j.catcom.2009.10.022

Catalysis Communications 11 (2010) 322–325
Contents lists available at ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
HZSM-5 supported silica–zirconia and its application in one-step synthesis
of methyl isobutyl ketone
Yanxia Liu , Kunpeng Sun , Xianlun Xu a,*, Xiaolai Wang
a,b
a
a,*
a
State Key Laboratory for Oxo Synthesis and Selective Oxidation, Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences, Lanzhou 730000, China
Graduate University of Chinese Academy of Science, Beijing 100039, China
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 21 June 2009
Received in revised form 26 September
In order to adjust the acidity and texture property of HZSM-5, SiO –ZrO was introduced via a modified
2
2
sol–gel method. Pd-loaded catalyst was prepared and its catalytic activity was tested in one-step synthe-
sis of MIBK. The catalyst exhibited a significant activity of acetone conversion up to 33.4% and MIBK selec-
tivity up to 88.1% under atmospheric pressure. The acidity and textural properties of support material, as
well as palladium dispersion of Pd-loaded catalyst were characterized. The results indicated that the
proper acid strength and textural property of the support, together with the high dispersion of palladium
resulted in good catalytic performance.
2
009
Accepted 22 October 2009
Available online 27 October 2009
Keywords:
Condensation
Ó 2009 Elsevier B.V. All rights reserved.
Methyl isobutyl ketone
Bifunctional catalyst
HZSM-5
2 2
SiO –ZrO
1
. Introduction
dehydration [19], alcohol oxidation [20] and cycloalkane oxidation
[
21]. However, researches of Zr–Si oxide used in condensation
Methyl isobutyl ketone (MIBK) is a kind of important product
reactions have not been reported. Among different preparation
techniques, sol–gel was often used for its good control of the prop-
erties of the resulted material. Here, the mixed oxide and HZSM-5
were compounded by a modified sol–gel method in order to obtain
an effective support material for synthesis of MIBK at atmospheric
pressure. The Pd-loaded catalysts were prepared and demon-
strated excellent catalytic activity in one-step synthesis of MIBK.
which is extensively used as a solvent in paints, resins and coat-
ings, as well as an extracting agent [1]. Traditionally, MIBK is man-
ufactured via a three-step process: (1) aldol condensation of
acetone to diacetone alcohol (DAA), (2) dehydration of DAA to
mesityl oxide (MO), (3) hydrogenation of MO to MIBK [2]. All the
three reactions can be carried out in a single step over bifunctional
catalysts, which have attracted extensive interests [3–5]. One-step
synthesis of MIBK reaction was typically studied using fixed-bed
tubular reactors at atmospheric pressure and temperatures that
rarely exceed 473 K [6]. The investigated bifunctional catalysts
were consisted of Pd or Pt as hydrogenation component and acid
or base support material which included various kinds of molecu-
lar sieves (ZSM-5 [7–9], X-zeolite [10,11], SAPO-11 [12]), mixed
oxides [1,13,14], calcium hydroxyapatite [15], active carbon
2
. Experimental
2.1. Catalyst preparation
Before utilization, the as-received HZSM-5 (labeled as HZ) was
calcined at 823 K for 5 h. The Zr–Si sol was prepared as follows:
solution A containing 0.06 mol tetraethyl orthosilicate (TEOS) and
[
6,16], etc.
1
7 mL ethanol was added dropwise to solution B which was con-
sisted of 0.02 mol ZrOCl
Á8H O and 17 mL distilled water at
13 K. Then HZ powder (3.621 g) was added in after aging for
0 min. The mixture became very ropy after 5 h. At last, it was
HZSM-5 was studied as support in this reaction with various
kinds of modifications. Good results were obtained at high pres-
sure when treated with alkaline metal cations [17] or organic acid
18] in which acid strength or pore structure was modified. Zr–Si
mixed oxide has attracted great research attentions for using as
catalyst or support in acid-catalyzed reactions, such as alcohol
2
2
3
3
[
dried at 383 K for 12 h and calcined at 823 K for 4 h. The obtained
material was denoted as ZrSiHZ. SiO –ZrO was synthesized by
sol–gel method. Firstly, Zr–Si sol was prepared. After aging for
0 min, the temperature was increased to 343 K and gelation oc-
curred immediately. The gel was dried and calcined. For compari-
son, physical mixture (PM) of SiO –ZrO and HZ with the same
2
2
3
*
Corresponding authors. Tel./fax: +86 931 4968217 (X. Xu).
E-mail address: xlxu@lzb.ac.cn (X. Xu).
2
2
1
566-7367/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.catcom.2009.10.022

Y. Liu et al. / Catalysis Communications 11 (2010) 322–325
323
ratio of ZrSiHZ was prepared by grinding. Palladium-loaded cata-
lysts were prepared by the conventional wetness impregnation
method using PdCl solution with palladium loading of 0.3 wt%.
2
The material was then dried at 393 K and calcined at 823 K for 4
h. Before reaction, raw catalysts were formed into particles of size
the framework of amorphous SiO
2
. When SiO
2
–ZrO
2
was synthe-
sized on surface of HZ by sol–gel method, the mixed oxide was
proved to be X-ray amorphous for the high mixing degree. In addi-
tion, the dealumination of HZSM-5 in ZrSiHZ which was proved by
2
7
Al-NMR spectra (not shown) may also relate to the decreased dif-
0
.29–0.84 mm and reduced by hydrogenation at 573 K for 4 h.
fraction intensity. Besides a well-resolved peak at about 55 ppm
due to tetrahedral framework aluminum (FAL), there was a small
peak near 0 ppm attributed to octahedral extra-framework alumi-
num (EFAL) [23] in the Al-NMR spectra of HZ and ZrSiHZ. A simple
integration of the peaks was applied to determine the relative
amount of the two species. The ratios of FAL to EFAL were: HZ-
1:0.02 and ZrSiHZ-1:0.05. A stronger signal of EFAL existed in spec-
trum of ZrSiHZ, which implied that dealumination occurred.
2.2. Catalyst characterization
2
7
X-ray diffraction (XRD) patterns were obtained by Philips X Pert
Pro X diffractometer operated with a Ni-filtered Cu K
Physisorption of N at 77 K was carried out on ASAP 2020 appara-
tus. NH -TPD was performed on AutoChem II 2920 instrument.
Typically, a sample of 100 mg was first pretreated in helium at
23 K for 30 min. Then temperature was cool to 373 K and the
a radiation.
2
3
The results from N
Table 1. SiO –ZrO was proved to be a kind of mesoporous material
and possessed only a few micro-pores. Compared with HZ, intro-
duction of SiO –ZrO led to the decrease of surface and pore vol-
2
physisorption of the samples are given in
8
2
2
sample was saturated with ammonia. Desorption of ammonia
was then carried out from 372 to 973 K under flowing helium
2
2
(
30 mL/min) using a temperature ramp of 10 K/min. High Resolu-
ume of ZrSiHZ and PM. Both the surface area and pore volume of
ZrSiHZ were lower than that of PM which may be related to the
amorphous SiO –ZrO in ZrSiHZ. Combined with the above men-
2 2
tion Transmission electron microscopy (HRTEM) was performed
on JEM-2010 apparatus operating at 200 kV.
tioned dealumination, the blockage of the EFAl species in the pore
channels was also accounts for the decreased pore volume, espe-
cially in micro-pores [24].
2.3. Catalytic activity test
The reactions were carried out in a fixed-bed micro-reactor
Fig. 2 shows the NH
were two peaks on NH
peak (l-peak) found at about 450 K and high-temperature peak
(h-peak) found at about 670 K. The spectrum of SiO –ZrO exhib-
ited a low broad peak indicating of less acid sites due to formation
of Zr–O–Si [19]. Two desorption peaks were also observed on
curves of ZrSiHZ and PM. The h-peak of PM found at 650 K sug-
gested the decreased acid strength, which was due to dilution of
3
-TPD traces of different materials. There
(
H
6 mm i.d.) using 0.6 g catalyst at ambient pressure. Acetone and
were fed into the reactor from the top and mixed in preheating
section. The acetone (0.02 mL/min) and H were introduced with
the molar ratio of H and acetone fixed at 1. Before reaction, the
catalyst was pretreated with hydrogen (20 mL/min) in situ at
53 K for half an hour and the experiment was carried out at this
temperature for 2 h. The products were collected in a cold trap
held in an ethanol-dry ice bath) and quantified by GC equipped
with a flame ionization detector.
3
-TPD spectrum of HZ: low-temperature
2
2
2
2
2
4
(
2 2
strong acid sites by SiO –ZrO . Compared with that of PM, acid
strength of ZrSiHZ, expressed by the corresponding temperature
of peaks, especially strong acid strength, weakened distinctly while
3
. Results and discussion
Table 1
3.1. Characterizations
Textural properties of materials.
a
2
b
3
c
3
Sample
HZ
SiO -ZrO
2 2
ZrSiHZ
PM
S
BET (m /g)
V
pore (cm /g)
Vmicr (cm /g)
The XRD patterns of different samples are shown in Fig. 1. Differ-
ent from that of PM, the pattern of ZrSiHZ just exhibited diffraction
peaks of HZSM-5 with much lower intensity which suggested that
373
82.2
187
263
0.229
0.046
0.098
0.137
0.123
0.004
0.065
0.095
SiO
that high degree mixing of Zr–Si mixed oxide often led to absence
of XRD pattern for ZrO [22]. The broad peak around 22° (2h) seen
from the smaller inset could be attributed to insertion of Zr into
2 2
–ZrO was highly dispersed on surface of HZ. It was reported
a
b
c
BET surface area.
Single point total pore volume at P/P0 = 0.97.
Cumulative volume by H-K method.
2
Fig. 1. XRD patterns of materials: (A) HZ, (B) SiO
2
–ZrO2, (C) ZrSiHZ and (D) PM.
3
Fig. 2. NH -TPD profiles of samples: (A) HZ, (B) SiO
2
–ZrO2, (C) ZrSiHZ and (D) PM.

3
24
Y. Liu et al. / Catalysis Communications 11 (2010) 322–325
weak acid sites and its strength changed insignificantly. Since the
EFAL did not contribute to the acidity [25], the weak of strong acid
strength of the ZrSiHZ can be ascribed to the dealumination in pre-
paring procedure. Meanwhile, the coverage of surface of HZ by
[26]. Large palladium particles cannot be found in images of Pd/
ZrSiHZ and Pd/PM, which implied that palladium species dispersed
better than that of Pd/SiO –ZrO . But some palladium aggregated
2 2
on surface of Pd/PM, which resulted in uneven distribution of
palladium.
2 2
amorphous SiO –ZrO may be another reason for the reduction of
the acid sites.
Typical HRTEM images of the catalysts are shown in Fig. 3. The
palladium was highly dispersed on surface of HZ. Whereas image
3.2. Acetone condensation
of Pd/SiO
particles (2–7 nm) in meso-pores. The presence of worm-like
meso-pores confirmed the mesoporous texture of the SiO –ZrO
2
–ZrO
2
showed that palladium was distributed as large
Activities of support materials and catalysts were tested under
the same experiment conditions and the representative results
were listed in Table 2. Pd/ZrSiHZ exhibited acetone conversion
2
2
2
Fig. 3. HRTEM images of catalysts: (A) Pd/HZ, (B) Pd/SiO –ZrO2, (C) Pd/ZrSiHZ, (D) Pd/PM and (E) Pd/ZrSiHZ after used for 12 h.

Y. Liu et al. / Catalysis Communications 11 (2010) 322–325
325
Table 2
ladium particles preferred to the hydrogenation of C@O of acetone
to IPA in the competition between hydrogenation and condensa-
tion. This phenomenon was consistent with the Ref. [5] in which
high Pd-loading resulted in high selectivity of IPA. Moreover, the
absence of micro-pores can explain the higher yield of DIBK which
was often formed in large pores.
a
Acetone conversion and yield of the products.
Sample
Conversion (%)
Yield (%)b
MIBK
IPA
MO
DIBK
ZrSiHZ
PM
Pd/HZ
Pd/ SiO
Pd/ZrSiHZ
Pd/PM
5.7
8.0
9.2
24.8
33.4
28.5
0.9
0.2
8.2
13.7
29.4
23.8
–c
3.3
5.0
0.5
0.6
–
0.1
–
0.1
3.2
0.9
0.9
0.5
0.1
5.6
0.2
0.3
The stability of Pd/ZrSiHZ was evaluated and the result was
shown in Fig. 4. The catalytic activity of Pd/ZrSiHZ declined in
2
-ZrO
2
1
0 h. From the HRTEM image of the used catalyst (Fig. 3E), the
0.2
aggregation of palladium may be the main reason for the declined
activity. The selectivity of IPA was increased for the enlargement of
palladium. This result was consistent with high yield of IPA on Pd/
a
Reaction conditions: 453 K, 2 h, H
MIBK: methyl isobutyl ketone, IPA: isopropanol, MO: mesityl oxide, DIBK: dii-
2
/actone = 1, atmospheric pressure.
b
sobutyl ketone. Other byproducts: C5 (2-methyl-pentane), 2,6,8-trimethylnonane-
2 2
SiO –ZrO . Effective methods of doping the palladium were in re-
4
-one, isophorone and 4,6-dimethyl-2-heptanone, etc.
Can not determined.
c
search and in hope of improving the stability of the catalyst.
4
. Conclusion
A kind of potential carrier ZrSiHZ was prepared by compound-
ing HZSM-5 and SiO –ZrO via a modified sol–gel method. The
2 2
and MIBK yield of 33.4 and 29.4%. This result compared well to
those reported by other Refs. [2,4,11,13,15]. As the product of
MO or MIBK condensed with another molecular of acetone, dii-
sobutyl ketone (DIBK) was the major byproduct in gas-phase con-
densation [27]. Our catalyst was superior to others in terms of low
yield of DIBK in gas phase acetone condensation.
When support materials were tested in absence of palladium,
the condensation product MO was the major product with low ace-
tone conversion, which proved the composed material acted
mainly as acid catalysts. The Pd/HZ exhibited a low acetone con-
version of 9.2%. Introduction of mixed oxide by different method
resulted in the decreased acid strength of PM and ZrSiHZ. Accord-
ingly, their Pd-loaded catalysts exhibited acetone conversion of
Pd/ZrSiHZ showed significant catalytic activity in one-step synthe-
sis of MIBK. The results demonstrated that the catalytic perfor-
mance of the bifunctional catalyst for synthesis of MIBK strongly
depended on the textural properties and the acidity of the support
material. It was considered that the selectivity of the IPA was re-
lated to the particle size of the supported palladium. Although
the stability of the catalyst needed to be further improved, a good
catalytic performance of Pd/ZrSiHZ had been realized. The conse-
quent researches were on the way.
Acknowledgements
2
8.5 and 33.4%, respectively. This suggested that the relative weak
The authors thank Professor Yongxiang Zhao (Shanxi Univer-
sity) and Ying Qian (Lanzhou Petrochemical Research Center, Ptro-
China) for their help on NH3-TPD experiment.
acid strength was favorable for this reaction. Although ZrSiHZ
showed lower surface area than that of PM, a higher MIBK yield
of 29.4% was achieved by Pd/ZrSiHZ. The proper acidity of ZrSiHZ
and more effective palladium dispersion of Pd/ZrSiHZ accounted
for this phenomenon. In addition, SiO
together in random in PM. Unlike that, the amorphous SiO
dispersed on surface of ZrSiHZ which was favorable to the mass
transfer.
Pd/SiO
of 5.6% isopropanol (IPA) and 3.2% DIBK. The high yield of IPA may
be due to the large palladium particles of Pd/SiO –ZrO . Large pal-
References
2
–ZrO
2
and HZ were mixed
–ZrO
2
2
[
[
1] A.A. Nikolopoulos, J.J. Spivey, Appl. Catal. A. 296 (2005) 128.
2] I.V. Kozhevnikova, E.F. Kozhevnikova, J. Catal. 238 (2006) 286.
[3] P. Yang, N. Xu, J. Wang, T. Wu, Catal. Commun. 8 (2007) 997.
[
4] R.D. Hetterley, R. Mackey, J.T.A. Jones, Y.Z. Khimyak, A.M. Fogg, I.V.
Kozhevnikov, J. Catal. 258 (2008) 250.
2 2
–ZrO exhibited acetone conversion of 24.8% with yield
[
5] P.P. Yang, Z.L. Wang, M.P. Xu, Q.S. Liu, X.W. Yang, Catal. Commun. 6 (2005) 107.
2
2
[6] O.R.G. Waters, B. Kraushaar-Czarnetzki, Ind. Eng. Chem. Res. 45 (2006) 6111.
[
[
7] L. Melo, A. Llanos, M. Mediavilla, D. Moronta, J. Mol. Catal. A 177 (2002) 281.
8] R. Morales, L. Melo, J. Brito, A. Llanos, D. Moronta, L. Albornoz, E. Rodrguez, J.
Mol. Catal. A 203 (2003) 277.
[
9] L.M.M. Mediavilla, Y. Díaz, A. Albornoz, A. Llanos, J.L. Brito, Micropor. Mesopor.
Mater. 116 (2008) 627.
[
[
[
[
[
[
10] L.V. Mattos, F.B. Noronha, J.L.F. Monteiro, J. Catal. 209 (2002) 166.
11] A.C.C. Rodrigues, J.L.F. Monteiro, Appl. Catal. A 362 (2009) 185.
12] S.M. Yang, Y.M. Wu, Appl. Catal. A 192 (2000) 211.
13] F. Al-Wadaani, E.F. Kozhevnikova, I.V. Kozhevnikov, J. Catal. 257 (2008) 199.
14] Y.Z. Chen, H.R. Tan, K.L. Shen, Appl. Catal. A 205 (2001) 61.
15] N. Cheikhi, M. Kacimi, M. Rouimi, M. Ziyad, L.F. Liotta, G. Pantaleo, G.
Deganello, J. Catal. 232 (2005) 257.
[
[
16] V.A.S. Thotla, S.M. Mahajani, Ind. Eng. Chem. Res. 46 (2007) 8371.
17] P.Y. Chen, C.C. Chen, N.S. Chang, W.C. Lin, T.K. Chuang, Preparation of MIBK,
USP (1991) 5059724.
[
18] P.Y. Chen, S.J. Chu, K.C. Wu, W.C. Lin, Preparation of MIBK, USP (1997)
5084207.
[
[
19] Q. Zhuang, J.M. Miller, Appl. Catal. A. 209 (2001) L1.
20] Y.-C. Lin, C.-H. Chang, C.-C. Chen, J.-M. Jehng, S.-G. Shyu, Catal. Commun. 9
(
2008) 675.
[
[
[
[
[
21] C. Palazzi, L. Oliva, M. Signoretto, G. Strukul, J. Catal. 194 (2000) 286.
22] Z.-G. Wu, Y.-X. Zhao, D.-S. Liu, Micropor. Mesopor. Mater. 68 (2004) 127.
23] F. Kollmer, H. Hausmann, W.F. Holderich, J. Catal. 227 (2004) 398.
24] Y. Fang, X. Bao, X. Lin, G. Shi, H. Liu, J. Phys. Chem. B 110 (2006) 15411.
25] L. Rodriguez-Gonzalez, F. Hermes, M. Bertmer, E. Rodriguez-Castellon, A.
Jimenez-Lopez, U. Simon, Appl. Catal. A. 328 (2007) 174.
[
[
26] A. Tarafdar, A.B. Panda, P. Pramanik, Micropor. Mesopor. Mater. 84 (2005) 223.
27] Y. Watanabe, Y. Izumi, Y. Mizutani, Bull. Chem. Soc. Jpn. 47 (1974) 2922.
Fig. 4. Acetone conversion and product selectivity versus time on stream over Pd/
ZrSiHZ.