Effect of preparation condition on performance of Ag-MoO 3/ZrO2 catalyst for direct epoxidation of propylene by molecular oxygen

Article abstract of DOI:10.1016/j.molcata.2005.01.040

The Ag-MoO₃/ZrO₂ catalyst for an epoxidation of propylene by molecular oxygen was prepared by the precipitation/impregnation method, and characterized by N₂ adsorption at low temperature, XRD, NH₃-TPD and CO

Full text of DOI:10.1016/j.molcata.2005.01.040

Journal of Molecular Catalysis A: Chemical 232 (2005) 165–172
Effect of preparation condition on performance of Ag–MoO /ZrO2
3
catalyst for direct epoxidation of propylene by molecular oxygen
∗
Guojie Jin, Guanzhong Lu , Yanglong Guo, Yun Guo, Junsong Wang,
Weiying Kong, Xiaohui Liu
Lab for Advanced Materials, Research Institute of Industrial Catalysis, East China University of Science and
Technology, 130 Meilong Road, Shanghai 200237, PR China
Received 31 March 2004; received in revised form 10 December 2004; accepted 26 January 2005
Abstract
The Ag–MoO
and characterized by N
on its physicochemical properties and catalytic performance were studied. The results show that suitably low surface area, big pore size, and
modest and weak surface acidity–basicity for the Ag–MoO /ZrO catalyst are in favor of the selective formation of propylene oxide. High
surface area, abundant micro-pore structure and strong acidity–basicity or acidity-free are liable to the deep oxidation of propylene to CO
and H O. As the support of the Ag–MoO catalyst for an epoxidation of propylene, the monoclinic phase of ZrO is preferable to tetragonal
catalyst are beneficial to improve the selectivity to
3
/ZrO
2
catalyst for an epoxidation of propylene by molecular oxygen was prepared by the precipitation/impregnation method,
2
adsorption at low temperature, XRD, NH -TPD and CO -TPD. The effects of preparation condition of ZrO support
3
2
2
3
2
2
2
3
2
one. Suitably larger particles of Ag and highly dispersed MoO
propylene oxide.
3 3 2
on the Ag–MoO /ZrO
©
2005 Elsevier B.V. All rights reserved.
Keywords: Propylene; Epoxidation; Silver–molybdenum catalyst; Zirconia support; Preparation condition
1
. Introduction
but high energy consumption makes it not be commercial-
ized up to date. For the catalytic epoxidation of propylene by
O2/H2 over Au/TiO2 or Pd–Pt/TS-1, the selectivity to PO is
higher than 90%, but the conversion of propylene is 1–2%
only [3–5], in which large quantities of H2 is consumed to
react with O2 to H2O and a careful handling is strictly re-
quired for the explosive mixture of H2–O2. Therefore, in view
of economic and green chemistry, the direct epoxidation of
propylene by molecular oxygen over a solid catalyst is a most
ideal technique of producing PO.
Attempts to directly epoxidize propylene, similar to what
is done in the epoxidation of ethylene over the silver cata-
lyst, have been tried and some good results have been ob-
tained. Using silver supported on CaCO3, the PO selectiv-
ity of 64.0% was attained with the propylene conversion of
Propylene oxide (PO) is an important intermediate for the
organic chemistry industry and there is an increasing demand
for it in marketplace. Currently, PO is commercially pro-
duced by two conventional processes of chlorohydrin and
hydroperoxide. Chlorohydrin method has the disadvantages
of producing stoichiometric amount of waste salts and some
chlorinated byproducts; the disadvantage of hydroperoxide
method is that the co-product (styrene or tert-butylalcocol)
is produced in a fixed amount. The direct synthesis of PO
from propylene and oxygen has long been desired and consid-
ered as one of the most important and challenging chemical
reactions to remain unsolved by catalysis [1]. The molten-
salt method can directly epoxidize propylene by oxygen [2],
1.5% when NO, EtCl and CO2 additives were added in the
feedstock [6]. Lu and Zuo prepared the unsupported silver
catalyst containing NaCl as the promoter, over which 54.0%
C3H6 conversion and 26.3% selectivity to PO were achieved
∗
Corresponding author. Tel.: +86 21 64253703; fax: +86 21 64253703.
E-mail address: gzhlu@ecust.edu.cn (G. Lu).
1
381-1169/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2005.01.040

1
66
G. Jin et al. / Journal of Molecular Catalysis A: Chemical 232 (2005) 165–172
for the epoxidation of propylene by air [7]. The epoxidation
of propylene with air over the similar NaCl-modified silver
catalyst was also recently reported [8]. The mechanism for an
oxidation of propylene by molecular oxygen over the silver
catalyst has been studied in detail over the past years; it is
thought that the routes leading to PO and CO2 coexist on the
surface of silver catalyst [9–11]. It was reported that alkenes
wasoxidizedtoepoxidebymolecularoxygenoverthehetero-
geneousmolybdenumcatalyst[12]. Recently, NaCl-modified
VCexCu1−x mixed oxide [13], Ti-modified high silica zeolite
its pH value was about 12. The precipitate was aged at room
temperature for 12 h, filtered and washed repeatedly with dis-
tilled water until the pH value of filtrate was about 7, and
then dried at 110 C for 24 h and calcined at 500 C for 4 h to
obtain the Ag–ZrO2 precursor. This precursor was crushed
and impregnated by an equal-volume aqueous solution of
◦
◦
(NH4) Mo7O24 (Ag:ZrO2:MoO3 = 20:76:4%, w/w/w), and
6
◦
◦
then it was dried at 110 C for 4 h and calcined at 300 C for
4 h to obtain the 4%MoO3/20%Ag–ZrO2 catalyst named as
M/AZ(500). AZ represents Ag–ZrO2 and 500 within bracket
◦
[14] and Ti-modified sulfated zirconia [15] were reported to
is the calcination temperature ( C) of Ag–ZrO2.
be effective for direct epoxidation of propylene with molec-
ular oxygen.
2.2. Characterization of catalyst
We have reported the supported Ag–Mo catalyst for di-
rect epoxidation of propylene by molecular oxygen [16–18],
and found that the promoter and support have a very im-
portant influence on the performance of catalyst, MoO3 in
catalyst plays the roles of electric- and structure-type bifunc-
tional promoter, and ZrO2 is an excellent support. In this
paper, by changing preparation condition of ZrO2 support,
the Ag–MoO3/ZrO2 catalysts with different physicochemi-
cal properties were prepared. Based on correlating the epox-
idation performance with the physicochemical properties of
catalyst, the effect of preparation condition or physicochemi-
cal properties of ZrO2 on the epoxidation performance of the
Ag–MoO3/ZrO2 catalyst is discussed.
The BET surface area and pore size distribution of catalyst
were measured with ST-03A Instrument of Surface Area and
Pore Size Distribution (Beijing Analytic Instrument Plant) at
liquid-N2 temperature. The crystalline structure of catalyst
was identified by XRD with Rigaku D/max/2550/PC, Cu K␣
radiation. The NH3-TPD and CO2-TPD experiments were
conducted in the conventional flow apparatus equipped with
a thermal conductivity detector (TCD). The sample (300 mg)
◦
was pretreated under helium stream (30 ml/min) at 500 C
for 2 h, and then with the mixture of NH3–He (or CO2–He)
◦
◦
at 50 C for 1 h and with helium stream at 50 C for 1 h to
remove physisorbed NH3 (or CO2). TPD of adsorbed NH3
◦
(
or CO2) was measured from 50 to 450 C at the heating rate
◦
of 8 C/min.
2
2
2
. Experimental
2.3. Activity testing of catalyst
.1. Preparation of catalyst
The catalytic epoxidation of propylene by molecular oxy-
.1.1. Method A
gen was performed in the micro-reactor-GC system. The
stainless steel reactor in the fixed-bed is Ø5 mm × 300 mm
and0.5 mlcatalyst(20–40 mesh)wasfilled. Thereactioncon-
Zr(NO3)4 aqueous solution was slowly added in 10 M
NaOH aqueous solution under stirring and at room temper-
ature, and its pH value was about 12. After aged at room
−
1
◦
◦
dition is 0.1 MPa and 7500 h of space velocity at 350 C.
temperature for 12 h (or at 100 C for 160 h), the precipi-
The feedstock gas consisted of 22.7%C3H , 9.0%O2 and bal-
tate was filtered and washed repeatedly with distilled water
until the pH value of filtrate was about 7, and then dried
6
ance N2, without any additive such as nitrogen oxide, alkyl
halide, carbon dioxide or hydrogen. The composition of feed-
stock and effluent gas was analyzed by two on-line gas chro-
matographs with three packed columns (407 porous poly-
◦
◦
at 110 C for 24 h and calcined at 500 C for 4 h to ob-
tain ZrO2. ZrO2 was impregnated by a stoichiometric aque-
ous solution of AgNO3 (Ag:ZrO2 = 20:76%, w/w) under ul-
trasonic stirring and at room temperature, and then dried
˚
mer, silica gel and 5 A zeolite), in which TCD was used. The
◦
◦
method of carbon balance was used to verify the consumption
of propylene during the reaction. As O2 was insufficient in
at 110 C for 4 h, calcined at 450 C for 4 h. After that it
was impregnated with an equal-volume aqueous solution of
comparison with C3H in feedstock, the activity of catalyst
(
NH4) Mo7O24 (Ag:ZrO2:MoO3 = 20:76:4%, w/w/w), dried
6
6
◦
◦
was expressed by the conversion of O2 rather than that of
at 110 C for 4 h and calcined at 300 C for 4 h to obtain the
0%Ag–4%MoO3/ZrO2 catalyst named as AM/Z(500) (or
C3H hereinafter.
2
6
AM/Z*(500)). Z and Z* represents ZrO2, the former is aged
at room temperature for 12 h and the latter marked star is aged
◦
3
. Results and discussion
at 100 C for 160 h, and 500 within bracket is the calcination
◦
temperature ( C) of ZrO2.
3.1. Characterization of catalyst
2
.1.2. Method B
A mixed aqueous solution of AgNO3 and Zr(NO3)4
3.1.1. BET surface area
(
Ag:ZrO2 = 20:76%, w/w) was slowly added in 10 M NaOH
The BET surface areas of the catalysts prepared are listed
in Table 1. The results show that the preparation condition
aqueous solution under stirring and at room temperature, and

G. Jin et al. / Journal of Molecular Catalysis A: Chemical 232 (2005) 165–172
167
Table 1
Performances of the catalysts for epoxidation of propylenea
Catalystb
S (m /g)c
2
Conversion (%)
O2
Selectivity (%)
PO
C3H6
AC
ACR
HC
CO2
AM/Z*(500d)
AM/Z*(600)
AM/Z*(700)
AM/Z*(850)
AM/Z*(950)
AM/Z(500)
AM/Z(600)
AM/Z(700)
AM/Z(850)
M/AZ(500)
M/AZ(600)
M/AZ(700)
M/AZ(850)
128
117
105
21.7
3.6
44.0
26.4
24.1
17.1
59.9
39.7
24.9
18.2
100
97.9
43.6
7.3
15.7
15.3
6.8
6.4
3.5
67.2
54.0
14.3
2.9
9.3
9.1
4.5
1.0
1.9
1.8
1.0
0.9
0.6
6.3
5.3
1.7
0.5
1.8
2.1
7.8
40.3
20.8
20.6
39.3
41.5
57.9
2.8
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3.2
3.4
7.1
2.2
7.9
5.5
4.5
3.4
0
3.8
5.6
4.3
2.6
95.0
94.5
85.1
57.5
71.3
73.9
56.2
55.1
42.1
93.4
89.0
69.8
43.6
5.4
25.9
53.8
PO: propylene oxide, AC: acetone, ACR: acrolein, HC: hydrocarbons of C1 + C2 + C3 + C4 + C5 + C6.
a
b
c
d
◦
−1
Reaction condition: 350 C, 0.1 MPa and 7500 h of space velocity.
AM/Z* and AM/Z 20%Ag–4%MoO3/ZrO2; M/AZ: 4%MoO3/20%Ag–ZrO2.
S: BET surface areas.
◦
Calcination temperature ( C).
of ZrO2 has an obvious influence on the surface area of the
catalyst. With an increase of the calcination temperature of
Z, Z* or AZ, the surface areas of the AM/Z, AM/Z* and
M/AZ catalysts decrease continuously. For the AM/Z* cata-
lysts, when the calcination temperature of Z* increased from
◦
5
00 to 700 C, the surface area of AM/Z* decreased slowly
2
◦
from 128 to 105 m /g; after Z* was calcined at 850 C (or
50 C), the surface area of AM/Z* decreased sharply to 21.7
or 3.6 m /g). As the calcination temperature of Z and AZ
increased from 500 to 850 C, the surface area of AM/Z de-
creased from 44.0 to 17.1 m /g, and that of M/AZ decreased
from 59.9 to 18.2 m /g.
◦
9
(
2
◦
2
2
The difference between Z* and Z is that their precursor’s
(Zr(OH)4 precipitate) aging condition is different. The pre-
cursor of Z* was aged for longer time (160 h) at higher tem-
perature (100 C), made AM/Z* have higher surface area;
Fig. 1. Pore diameter distribution of AM/Z*(500).
◦
above show that the sample with abundant micropores (such
as AM/Z*(500)) possesses a large surface area, and the sam-
ple with larger pores (such as the AM/Z*(850)) possesses
a smaller surface area. For the AM/Z and M/AZ samples,
that of Z was aged for shorter time (12 h) at lower temper-
◦
ature (25 C), made AM/Z have smaller surface area. Such
2
as, the surface area of AM/Z*(500) is 128 m /g and that of
AM/Z(500) is 44 m /g. The surface area of M/AZ(500) is
0 m /g. With a rise of the calcination temperature of Z*, Z
2
2
6
and AZ, the difference of surface area between AM/Z*(850),
AM/Z(850) and M/AZ(850) is reduced gradually. When Z*,
◦
Z and AZ was calcined at 850 C, AM/Z*(850), AM/Z(850)
and M/AZ(850) have the near surface areas, 21.7, 17.1 and
2
1
8.2 m /g, respectively.
3
.1.2. Pore diameter distribution
The pore diameter distributions of AM/Z*(500) and
AM/Z*(850) are shown in Figs. 1 and 2, respectively. As
can be seen, the pore diameter of AM/Z*(500) is 2–10 nm
and its most probable pore diameter is 5.23 nm, and that
of AM/Z*(850) is 2–50 and 14.9 nm, respectively. With in-
creasing calcination temperature of Z*, the pore diameter of
AM/Z* increases and its surface area decreases. The results
Fig. 2. Pore diameter distribution of AM/Z*(850).

1
68
G. Jin et al. / Journal of Molecular Catalysis A: Chemical 232 (2005) 165–172
hold does not obviously change and consists of main tetrago-
nal phase and minor monoclinic phase. After ZrO2 is calcined
◦
at 850–950 C, its diffraction peaks obviously narrow, that is
to say, the particle (or of crystal) dimension of ZrO2 has in-
creased. But the diffraction peaks of Ag and MoO3 are hardly
affected by the calcination temperature of ZrO2. Comparing
with the very poor diffraction peaks of MoO3, the diffraction
peaks of Ag are strong and narrow, which suggests that Ag
distributes on the surface of ZrO2 with larger particles, and
MoO3 is highly dispersed on ZrO2.
The results in Fig. 4 show that, ZrO2 in the AM/Z catalyst
consists of the monoclinic and metastable tetragonal phases.
With increasing calcination temperature of ZrO2 from 500 to
◦
8
50 C, the metastable tetragonal phase reduces and the mon-
◦
oclinic phase gradually increases. After calcined at 850 C,
the tetragonal ZrO2 has transformed to the monoclinic phase
for the most part. The diffraction peaks of Ag and MoO3
change hardly with the calcination temperature of ZrO2, and
Ag has strong diffraction peaks and that of MoO3 is hardly
observed, which is similar to the situation of the AM/Z* cat-
alysts.
◦
Fig. 3. XRD patterns of the AM/Z* catalysts (500–950 C is the calcination
temperature of Z*; M, monoclinic ZrO2; T, tetragonal ZrO2; *, MoO3; ꢀ,
Ag).
their smaller surface areas are corresponding to their larger
pores.
For the M/AZ catalysts, ZrO in M/AZ(500) exists mainly
2
in the metastable tetragonal phase and very weak monoclinic
3
.1.3. X-ray diffraction (XRD)
phase (Fig. 5). The diffraction peaks of monoclinic ZrO2
The XRD patterns of AM/Z*, AM/Z and M/AZ are shown
strengthen and that of tetragonal ZrO weaken with an in-
2
in Figs. 3–5, respectively. There are four diffraction peaks for
crease of the calcination temperature, but there is still ap-
◦
◦
◦
◦
all the catalysts at 2θ = 38.1 , 44.3 , 64.4 and 77.4 , that are
corresponding to the crystal faces of Ag (1 1 1), (2 0 0), (2 2 0)
and (3 1 1), respectively. The very poor diffraction peaks at
preciable proportion of tetragonal ZrO in the M/AZ(850)
2
catalyst, unlike the AM/Z(850) catalyst in that the tetragonal
phase of ZrO has nearly entirely transformed to the mono-
2
◦
◦
2
θ = 27.0 and 33.3 are attributed to the MoO3 crystal. The
clinic phase. With an increase of the calcination temperature
◦
◦
diffraction peaks at 2θ = 28.2 and 31.4 are assigned to mon-
of Ag–ZrO , the diffraction peaks of Ag strengthen and the
2
oclinic ZrO2, and the diffraction peak of tetragonal ZrO2 is
at 2θ = 30.2 .
very poor peaks of MoO hardly change.
3
◦
It was reported that the initial temperature of phase
The XRD patterns of AM/Z* in Fig. 3 show that, with
transformation from the metastable tetragonal phase to
◦
the calcination temperature of ZrO2 increasing from 500 to
the monoclinic phase for pure ZrO is about 600 C, and
2
◦
◦
7
00 C, the crystalline structure of Z* in AM/Z* that they
this transformation can end at 650–700 C. Moreover, this
◦
◦
Fig. 4. XRD patterns of the AM/Z catalysts (500–850 C is the calcination
temperature of Z; M, monoclinic ZrO2; T, tetragonal ZrO2; *, MoO3; ꢀ,
Ag).
Fig. 5. XRD patterns of the M/AZ catalysts (500–850 C is the calcination
temperature of AZ; M, monoclinic ZrO2; T, tetragonal ZrO2; *, MoO3; ꢀ,
Ag).

G. Jin et al. / Journal of Molecular Catalysis A: Chemical 232 (2005) 165–172
169
transformation is affected sensitively by the impurities or ad-
ditives, which can stabilize usually the metastable tetragonal
form to higher temperature [19,20].
In the AM/Z* catalyst, the crystalline structure and phase
transformation of ZrO2 are thoroughly different from the re-
sults reported [19,20], this is probably because of preparing
ZrO2 with different procedure. ZrO2 in AM/Z* is a stable
tetragonal phase, unlike ZrO2 in AM/Z (or M/AZ) that is
a metastable tetragonal phase, its crystalline phase transfor-
mation with the calcination temperature is similar to that re-
◦
ported [19,20] and does not finish until about 850 C. The
fact that there are still some metastable tetragonal ZrO2 in
the M/AZ(850) catalyst suggests that the presence of the ac-
tive Ag particles can retard the transformation of crystalline
phase of ZrO2 from the metastable tetragonal phase to the
stable monoclinic phase.
◦
Fig. 7. CO2-TPD profiles of the AM/Z* catalysts (500–950 C is the calci-
nation temperature of Z*).
3
.1.4. NH3-TPD and CO2-TPD
The acid–base properties of catalyst were investigated by
NH3-TPD and CO2-TPD. The NH3-TPD and CO2-TPD pro-
files of AM/Z*, AM/Z and M/AZ are shown in Figs. 6–11.
The results show that the acidic and basic sites co-exist on
the surface of all the catalysts except for AM/Z*(950).
For AM/Z* (Figs. 6 and 7), with increasing the calcina-
tion temperature of Z* (ZrO2), the strength of acidity and
basicity (top temperature of NH3 and CO2 desorption peak)
on the surface of catalyst decreases, and the concentration
(peak area) of acidic and basic sites decreases quite slowly
for the AM/Z*(500–700) samples and the peak area sharply
decreases for AM/Z*(850). It suggests that there are abun-
dant and stronger acidic and basic sites on the surface of
AM/Z*(500–700), and some acidic and basic sites on the
surface of AM/Z*(850). There are few acidic and basic sites
on AM/Z*(950).
◦
There are two peaks of NH3-desorption at 50–180 C in
◦
Fig. 8. NH3-TPD profiles of the AM/Z catalysts (500–850 C is the calci-
the NH3-TPD curves of AM/Z (Fig. 8), which corresponds to
two kinds of acidic sites. Their peak areas of NH3 desorption
nation temperature of Z).
◦
Fig. 9. CO2-TPD profiles of the AM/Z catalysts (500–850 C is the calci-
nation temperature of Z).
◦
Fig. 6. NH3-TPD profiles of the AM/Z* catalysts (500–950 C is the calci-
nation temperature of Z*).

1
70
G. Jin et al. / Journal of Molecular Catalysis A: Chemical 232 (2005) 165–172
3.2. Epoxidation of propylene
The evaluation results of the catalysts for propylene epoxi-
◦
−1
,
dation at 350 C, 0.1 MPa and the space velocity of 7500 h
are presented in Table 1. The byproducts are mainly hydrocar-
bons (C1–C ) and CO2. Over the AM/Z*(500–700) catalysts,
6
with increasing calcination temperature of Z*(ZrO2), the se-
lectivity to PO increases from 1.8 to 7.8% and the O2 con-
version decreases from 100 to 43.6%. Over the AM/Z*(850)
catalyst, the selectivity to PO sharply increases to 40.3% and
the O2 conversion decreases to 7.3%. Using Z* calcined at
◦
9
50 C as the support of AM/Z*(950), the selectivity to PO
decreases to 20.8%.
Using the AM/Z catalysts, with the calcination tempera-
◦
ture of ZrO2 increasing from 500 to 850 C, the selectivity
to PO increases from 20.6 to 57.9% and the O2 conversion
decreases from 15.3 to 3.5%.
◦
Fig. 10. NH3-TPD profiles of the M/AZ catalysts (500–850 C is the calci-
nation temperature of AZ).
Using the M/AZ catalysts, with the calcination tempera-
◦
ture of AZ (Ag–ZrO2) increasing from 500 to 850 C, the
decrease with a rise of the calcination temperature of Z. In
the CO2-TPD curve (Fig. 9) of AM/Z(500), there is one peak
of CO2 desorption (one kind of basic site). With increasing
calcination temperature of Z, this peak (one kind of basic
site) divides into two peaks (two kinds of basic sites) and
their peak areas of CO2 desorption decrease.
The NH3-TPD and CO2-TPD profiles of M/AZ shown
in Figs. 10 and 11 do not resemble those of AM/Z* and
AM/Z, in which the concentration of basic sites is far smaller
selectivity to PO increases from 2.8 to 53.8% and the con-
version of O2 decreases from 67.2 to 2.9%.
The above results show that the preparation condition of
ZrO2 support has a great influence on the physicochemical
characteristics of the Ag–MoO3/ZrO2 catalyst and its sur-
face, consequently, their epoxidation performance is affected
markedly by different preparation condition of ZrO2.
3.3. Discussion
than that of acidic sites. In the NH3-TPD profiles of M/AZ
◦
(
Fig. 10), there are two peaks at ∼150 C (␣ peak) and
3.3.1. Relationship between the surface area, pore size,
crystalline form and acidity–basicity
◦
2
90–340 C (␤ peak), and ␣ peak shifts to low temperature
and ␤ peak disappears gradually with an increase of the calci-
nation temperature of AZ (Ag–ZrO2). In the CO2-TPD pro-
files of M/AZ(500) (Fig. 11), two peaks (two kinds of basic
sites) are observed, and both peaks incorporate to one peak
and then disappear with increasing the calcination tempera-
The above results show that, increasing calcination tem-
perature of ZrO2 or Ag–ZrO2 makes the pore size of catalyst
increase, the surface area and acidity–basicity of catalyst de-
crease, and the crystalline form of ZrO2 support transform.
At higher temperature, the phase transformation and grain
growth of ZrO2 is a main reason of leading the surface area
of catalyst to decrease [19,20]. For the AM/Z* catalyst, the
decrease of its surface area mainly results from the grain
growth of ZrO2; for the AM/Z and M/AZ catalysts, the de-
crease of their surface area mainly results from the crystalline
phase transformation.
◦
ture of AZ from 500 to 850 C.
Correlating the pore size, surface area and acidity–basicity
of the Ag–MoO3–ZrO2 catalysts, it can be found that the
catalyst having abundant micro-pores and high surface area
possesses abundant acidic and basic sites on its surface and
vice versa.
The studies of the acid–base properties of ZrO2 by the
adsorption of CO2 and NH3/pyridine showed that there are
Lewis acid sites and no Br o¨ nsted acid sites on the surface
of ZrO2 [21,22]. On the Ag–MoO3–ZrO2 catalyst, the acidic
and basic sites co-exist independently, and this situation is
similar to that of ZrO2 reported in Ref. [19]. But the prepara-
tion condition of catalyst has a great influence on the acidic
◦
and basic properties on the surface of the Ag–MoO –ZrO
Fig. 11. CO2-TPD profiles of the M/AZ catalysts (500–850 C is the calci-
nation temperature of AZ).
3
2
catalyst.

G. Jin et al. / Journal of Molecular Catalysis A: Chemical 232 (2005) 165–172
171
3
.3.2. Relationship between the epoxidation
ture of ZrO2 with the epoxidation performance of catalyst
shows that the Ag–MoO3 catalyst supported on monoclinic
ZrO2 has higher selectivity to PO than the catalyst supported
tetragonal ZrO2. It is possible that the monoclinic ZrO2 can
make Ag keep the metallic properties to reduce its electron
loss, which leads the Oad on Ag to have the stronger elec-
trophilic properties. This situation is similar to the Pt catalyst
supported on ZrO2 promoted by tungstate, i.e., the crystalline
structure of ZrO2 has an obvious effect on the metallic prop-
erties of the Pt catalyst [23,24].
performance and physicochemical properties of catalyst
Correlating the epoxidation performance with the surface
area of catalyst, we can see that the selectivity to PO in-
creases and the O2 conversion decreases with the decrease of
the surface area of the Ag–MoO3–ZrO2 catalysts, except for
AM/Z*(950) having no any acidic and basic properties on its
surface. For the M/AZ catalyst, with its surface area decreas-
ing slowly (from 59.9 to 18.2 m /g), the selectivity to PO in-
creases sharply (from 2.8 to 53.8%); for the AM/Z* catalyst,
when its surface area decreases from 127.8 to 21.7 m /g, the
2
2
The XRD results show that the particle size of Ag on
AM/Z* and AM/Z does not vary obviously with the calci-
nation temperature of ZrO2 support. But with an increase
of calcination temperature of Ag–ZrO2, the Ag particles
on M/AZ increase sharply and its surface area and surface
acidity–basicity decrease slowly, meanwhile the epoxidation
performance of catalyst is improved sharply. Based on the
above facts, it can be inferred that the particle size of Ag
on catalyst is also a more important factor of affecting the
selectivity to PO, and the catalyst having the Ag particles
with properly larger size behaves a higher performance for
the epoxidation of propylene. In the epoxidation of ethylene,
it has been found that the large Ag particles are more effec-
tive than small Ag crystallites [25,26]. This size effects result
from a change of the surface Ag structure, the small Ag par-
ticles have higher surface energy and more surface defects,
on which the adsorbed oxygen behaves stronger activity to
result in the deep oxidation of propylene to CO2 and H2O.
On the basis of above discussion, it can be found that
the epoxidation performance of the catalyst is affected by
many factors that should include the particle size of the active
silver, the crystalline form of ZrO2 support, and the surface
acidity–basicity, the surface area and pore size of catalyst,
which can be put into practice by optimizing the preparation
conditions of the ZrO2 support and catalyst.
selectivity to PO increases and the O2 conversion decreases;
when its surface area decrease further to 3.6 m /g, on the
2
contrary, the selectivity to PO obviously decreases and the
O2 conversion increases. So there is no parallel relationship
between the epoxidation performance and the surface area of
catalyst. But the suitably low surface area and large pores is
in favor of a diffusion of PO out of the catalyst, to avoid PO
to be deeply oxidized to CO2 and H2O and to improve the
selectivity to PO.
For the AM/Z and M/AZ catalysts, the weak acidity and
basicity on its surface makes the selectivity to PO increase.
Using the AM/Z*(500–700) catalyst with the abundant acidic
and basic sites, the PO selectivity is quite low; using the
AM/Z*(850) catalyst containing a small quantity of weaker
acidic sites and basic sites, the high selectivity to PO can
be obtained; for the AM/Z*(950) catalyst without the acidic
and basic sites, its epoxidation performance is very poor. The
above phenomena show that, the acidity and basicity on the
surface of catalyst has an obvious influence on its epoxidation
performance, the modest and weaker acidity and basicity is
beneficial to form PO selectively. If there are stronger acidity
and basicity or acidity-free on the surface of catalyst, the
deep oxidation of propylene or PO to CO2 and H2O will
occur advantageously.
When there are modest Lewis acidic sites on the
Ag–MoO3/ZrO2 catalyst, those acidic sites can acquire elec-
trons from the oxygen adsorbed (O ) on the Ag sites to make
4. Conclusions
ad
Oad possess stronger electrophilic character, which is bene-
ficial to the olefinic carbons of propylene react with Oad to
The physicochemical properties of the Ag–MoO3/ZrO2
catalyst are significantly affected by the preparation condi-
tions of the ZrO2 support. When ZrO2 is prepared by the
precipitation method, a long aging time at higher tempera-
ture is liable to form the tetragonal phase, and a shorter aging
time at room temperature is liable to form the monoclinic
phase. With the increase of the calcination temperature of
ZrO2 or Ag–ZrO2, the pore sizes of the catalyst enlarge, the
surface area and acidity and basicity of the catalyst decrease,
and the crystalline phase of ZrO2 will transform. The suitably
low surface area and big pore diameter, modest and weaker
acidity and basicity on the surface of catalyst are beneficial to
form PO selectively; the high surface area and abundant mi-
cropore structure, strong acidity and basicity or acidity-free
are liable to the deep oxidation of propylene to CO2 and H2O.
ThemonoclinicZrO2 isanexcellentsupportoftheAg–MoO3
catalyst for an epoxidation of propylene. The properly larger
produce PO, and inhibits the acid–base reaction between al-
+
lylic H (H properties) and O to a certain extent [11]. If
ad
the acidity on the catalyst is excessively strong, the action
between propylene and Oad is too strong to crack propylene
to form CO2 and H2O. If there are many strong basic sites
on the catalyst, the acid–base reaction (or dehydrogenation)
between allylic H and basic sites can occur, thus propylene
is also completely oxidized to CO2 and H2O, as a result,
the conversion of O2 increases and the selectivity to PO de-
creases.
For the AM/Z*(850), AM/Z(850) and M/AZ(850) cata-
◦
lysts calcined at 850 C, they have the near surface areas
2
(
21.7, 17.1 and 18.2 m /g, respectively) and similar surface
acidity and basicity (such as AM/Z*(850) and AM/Z(850)),
but their crystalline structure of ZrO2 and epoxidation perfor-
mance are obviously different. Relating the crystalline struc-

1
72
G. Jin et al. / Journal of Molecular Catalysis A: Chemical 232 (2005) 165–172
particle size of Ag should be kept for improving the epoxi-
dation performance of the catalyst. The suitable preparation
conditions of ZrO2 are that the Zr(OH)4 precipitate is aged
[5] R. Meiers, U. Dingerdissen, W.F. H o¨ lderich, J. Catal. 176 (1998)
76.
6] B. Cooker, A.M. Gaffney, J.D. Jewson, W.H. Onimus, US Patent
,965,480 (1999).
3
[
5
◦
at room temperature for 12 h, dried at 110 C for 24 h and
[
[
[
7] G. Lu, X. Zuo, Catal. Lett. 58 (1999) 67.
8] J. Lu, M. Luo, H. Lei, C. Li, Appl. Catal. A 237 (2002) 11.
9] H. Nakatsuji, Progr. Surf. Sci. 54 (1997) 1.
◦
calcined at 850 C.
[
[
[
[
10] Z. Hu, H. Nakai, H. Nakatsuji, Surf. Sci. 401 (1998) 371.
11] J.T. Roberts, R.J. Madix, W.W. Crew, J. Catal. 141 (1993) 300.
12] P.S.E. Dai, J.H. Lunsford, J. Catal. 64 (1980) 184.
Acknowledgements
13] J. Lu, M. Luo, H. Lei, X. Bao, C. Li, J. Catal. 211 (2002) 552.
This study was financially supported by National Natu-
ral Science Foundation of China (No. 20173015), Commis-
sion of Science and Technology of Shanghai Municipality
[14] K. Murata, Y. Kiyozumi, Chem. Commun. (2001) 1356.
[15] K. Murata, K.E. Bere, T. Hayakawa, S. Hamakawa, K. Suzuki, K.
Matano, React. Kinet. Catal. Lett. 72 (2001) 57.
[
[
16] G. Jin, G. Lu, X. Wang, Chin. J. Catal. 23 (2002) 405.
17] G. Jin, G. Lu, Y. Guo, Y. Guo, J. Wang, X. Liu, Catal. Lett. 87
(
No. 02DJ14028, No. 03DJ14006) and Ph.D. Program Spe-
cific Foundation by Ministry of Education of China (No.
0010251006).
(
2003) 249.
[18] G. Jin, G. Lu, Y. Guo, Y. Guo, J. Wang, X. Liu, Catal. Today 93/95
2004) 171.
2
(
[
[
[
[
19] T. Yamaguchi, Catal. Today 20 (1994) 199.
20] C. Chang, R. Gopalan, Y. Lin, J. Membr. Sci. 91 (1994) 27.
21] B. Xu, T. Yamaguchi, K. Tanabe, Chem. Lett. (1988) 1663.
22] Y. Nakano, T. Iizuka, H. Hattori, K. Tanabe, J. Catal. 57 (1979) 1.
References
[
1] J.F. Roth, Chemtech (1991) 357.
[23] J. Yori, C. Pieck, J. Parera, Appl. Catal. A. 181 (1999) 5.
[24] J. Yori, J. Parera, Catal. Lett. 65 (2000) 205.
[25] S. Goncharova, E. Paukshtis, B. Bal’zhinimaev, Appl. Catal. A. 126
(1995) 67.
[26] D. Minahan, G. Hoflund, W. Epling, D. Schoenfeld, J. Catal. 168
(1997) 393.
[2] T.A. Nijhuis, S. Musch, M. Makkee, J.A. Moulijn, Appl. Catal. A
196 (2000) 217.
[
[
3] T. Hayashi, K. Tanaka, M. Haruta, J. Catal. 178 (1998) 566.
4] G. Jenzer, T. Mallat, M. Maciejewski, F. Eigenmann, A. Baiker,
Appl. Catal. A 208 (2001) 125.