Effect of ZnO additives and acid treatment on catalytic performance of Pt/WO3/ZrO2 for n-C7 hydroisomerization

Article abstract of DOI:10.1016/j.jcat.2006.08.015

The effect of WO₃ (5-50 wt%) and ZnO (0.7-22 wt%) on the catalytic properties of Pt/WO₃/(ZnO)-ZrO₂ for n-heptane (n-C₇) hydroisomerization was investigated. The optimized WO₃ and ZnO contents are 20 wt% and 3.4 wt%, respectively. The catalytic performance is achieved at 81% n-C₇ conversion and 89% C₇ isomer selectivity at 250 °C, which is reproducible and can be kept constant over 82 h under reaction conditions. Both WO₃ and ZnO can stabilize the tetragonal phase of ZrO₂. The Bronsted acid-to-Lewis acid ratio should be optimized to achieve high catalytic performance. The activity for Pt/WO₃/ZrO₂ using Zr(OH)₄ as the catalyst support (n-C₇ conversion, 88% at 250 °C) is much higher than that for Pt/WO₃/ZrO₂ with ZrO₂ as the support (n-C₇ conversion, 9% at 250 °C) with the same Pt and WO₃ loadings. BET, SEM-EDX, and pyridine-FTIR analyses show that acid treatment can successfully enhance the surface area (from 73 to 91 m²/g), increase the number of Bronsted acid sites, and lower the surface Zn:Zr ratio (from 0.43 to 0.15) for ZnO-ZrO₂ with 22 wt% ZnO. The yield of C₇ isomers is increased from nil to 47% at 300 °C on Pt/WO₃/ZnO-ZrO₂ catalyst after acid treatment. It is suggested that n-heptane hydroisomerization activity is related to acidity, surface area, and crystalline phase of ZrO₂.

Full text of DOI:10.1016/j.jcat.2006.08.015

Journal of Catalysis 244 (2006) 17–23
www.elsevier.com/locate/jcat
Effect of ZnO additives and acid treatment on catalytic performance
of Pt/WO /ZrO for n-C hydroisomerization
3
2
7
a,∗
b
b
a
a
a
a
Yan Liu , Yejun Guan , Can Li , Juan Lian , Geok Joo Gan , Eng Chew Lim , Fethi Kooli
a
Institute of Chemical and Engineering Sciences, 1 Pesek Road, Jurong Island, Singapore 627833
State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Dalian 116023, China
b
Received 16 August 2006; revised 18 August 2006; accepted 19 August 2006
Available online 26 September 2006
Abstract
The effect of WO (5–50 wt%) and ZnO (0.7–22 wt%) on the catalytic properties of Pt/WO /(ZnO)–ZrO for n-heptane (n-C ) hydroisomer-
3
3
2
7
ization was investigated. The optimized WO and ZnO contents are 20 wt% and 3.4 wt%, respectively. The catalytic performance is achieved at
3
◦
8
1% n-C conversion and 89% C isomer selectivity at 250 C, which is reproducible and can be kept constant over 82 h under reaction condi-
7
7
tions. Both WO and ZnO can stabilize the tetragonal phase of ZrO . The Brønsted acid-to-Lewis acid ratio should be optimized to achieve high
3
2
◦
catalytic performance. The activity for Pt/WO /ZrO using Zr(OH) as the catalyst support (n-C conversion, 88% at 250 C) is much higher
than that for Pt/WO /ZrO with ZrO as the support (n-C conversion, 9% at 250 C) with the same Pt and WO loadings. BET, SEM-EDX, and
3
2
4
7
◦
3
2
2
7
3
2
pyridine-FTIR analyses show that acid treatment can successfully enhance the surface area (from 73 to 91 m /g), increase the number of Brønsted
acid sites, and lower the surface Zn:Zr ratio (from 0.43 to 0.15) for ZnO–ZrO with 22 wt% ZnO. The yield of C isomers is increased from nil to
2
7
◦
7% at 300 C on Pt/WO /ZnO–ZrO catalyst after acid treatment. It is suggested that n-heptane hydroisomerization activity is related to acidity,
3 2
4
surface area, and crystalline phase of ZrO .
2
©
2006 Elsevier Inc. All rights reserved.
Keywords: Pt/WO –ZrO catalysts; ZnO; Acid treatment; n-Heptane hydroisomerization; Pyridine-FTIR
3
2
1
. Introduction
Unlike C4–C6 isomerization, alkane conversion above C7
to their multibranched components in the selective alkane iso-
merization is rather difficult. This is attributed to the ease of
cracking by β-scission. These components have large numbers
of methylene species that, in the presence of Lewis acids, can
form carbenium ions by abstraction of hydride ions [7] or, in
strong Brønsted acids, can form carbonium ions [8]. It is highly
desirable to obtain environmentally clean catalysts that can be
used to isomerize long-chain paraffins with minimum cracking.
Recently, metal oxides promoted by sulfur compounds
Environmentally driven regulations are requiring significant
improvement in the quality of diesel and gasoline in many parts
of the world [1]. Reducing the concentrations of aromatic com-
pounds has been recognized as one way to meet the requirement
of “clean fuels” [2], which may be accompanied by a drop
in octane number. One potential solution is to isomerize the
straight-chain paraffins with low octane number, which are a
major constituent in some crude oil fraction, into the branched
isomers with high octane number as octane-enhancing compo-
nents. Consequently, the isomerization of C7+ has reemerged as
the researchers’ interest and is becoming increasingly popular
from both scientific and industrial standpoints [3–6].
2−
were studied as strong solid acid catalysts, particularly SO
-
4
promoted zirconia containing noble metal to inhibit catalyst
deactivation [9]. However, the low isomerization selectivity for
C7+ alkanes, poor stability, and tendency to form volatile sulfur
compounds during catalysis and regeneration limit their further
applicability.
Tungsten-related solid acids, especially Pt/WO3–ZrO2, with
mild acidity, show good activity and selectivity for alkane
isomerization and have been investigated in n-heptane isomer-
*
Corresponding author. Fax: +65 63166182.
E-mail address: liu_yan@ices.a-star.edu.sg (Y. Liu).
0021-9517/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2006.08.015

18
Y. Liu et al. / Journal of Catalysis 244 (2006) 17–23
ization [10–16]. Iglesia et al. showed that Pt/SO2 /ZrO2 and
−
pressed into self-supporting wafers and treated at 400 C for
◦
4
3
Pt/WOx/ZrO2 exhibited similar activity for low-temperature
200 C) n-heptane isomerization, but Pt/WOx/ZrO2 was sig-
nificantly more active at high reaction temperature (250 C)
17]. This was attributed to the greater stability of Pt/WOx/ZrO2
1 h in vacuo or in H gas (flow rate, 80 cm /min) in a quartz
IR cell. Once the samples were cooled to room temperature,
a background was recorded. After being exposed to pyridine
vapor at ambient temperature for 10 min, the wafers were out-
gassed at 150 C for 1 h, and then IR spectra were collected
2
◦
(
◦
[
◦
to deactivation by sublimation or decomposition of the acid
sites at higher reaction temperatures. These authors also ob-
served that Pt/WOx/ZrO2 has much higher isomerization selec-
at room temperature. All spectra were differences between the
spectra of samples with adsorbed pyridine and the backgrounds.
tivity than Pt/SO2 /ZrO2. This was attributed to the loss of a
−
4
hydrogen activating metal function due to sulfur poisoning of
2
.3. Catalytic test conditions
the Pt particles in Pt/SO2 /ZrO2 [18].
−
4
The isomerization reaction is a bifunctional catalytic process
19]. The balance between the hydrogenation–dehydrogenation
Catalytic activity analysis was carried out at atmospheric
pressure in a tubular quartz glass, fixed-bed reactor. Each run
used approximately 300 mg of catalyst in the form of 30–
[
function and acidity is crucial for determining activity, stability,
and product selectivity [20,21]. In the present study, ZnO base
additives and acid treatment method were varied to adjust the
acidity of the catalysts. The resulting catalysts, Pt/WO3/(ZnO)–
ZrO2, was investigated for n-C7 hydroisomerization. We found
that an appreciable amount of ZnO can stabilize the tetragonal
phase of ZrO2 and affect acidity, and that acid treatment can
remove extra ZnO on the surface and further enhance catalytic
performance.
8
0 mesh particles. Before catalytic measurements, the cata-
◦
lysts were reduced at 400 C using H2 (30 ml/min) for 2 h.
n-Heptane was introduced to the reactor by a carrier gas of
◦
H2 flowing through a saturator maintained at 16 C. The flow
rate was 30 ml/min. Catalytic tests were performed at a tem-
◦
perature range of 250–350 C, measured using a thermocouple
projecting into the catalyst bed. The product gas mixture was
analyzed periodically (one analysis every 22 min) on-line with a
HP 6890N gas chromatograph equipped with a flame ionization
detector and a HP-PONA (50 m × 0.2 mm) capillary column.
2
. Experimental
2
.1. Catalysts
3
. Results and discussion
Zr(OH) and Zn(OH) –Zr(OH) materials were prepared
4
2
4
3
.1. XRD analysis
using a precipitation method. Briefly, an alkaline slurry (pH 10)
containing ZrOCl2 (with or without ZnCl2) and ammonium hy-
◦
Fig. 1 displays the XRD analysis of the calcined WO3/ZrO2
droxide was heated to 90 C for 4 h. The product was filtered,
◦
−
mixture with different WO3 loadings at 800 C. The intensity
of the monoclinic phase of ZrO2 decreased gradually with the
amount of WO3 loading, indicating that WO3 can stabilize the
tetragonal phase of ZrO2, in agreement with previous reports
washed with excess water until it was Cl free, and then dried
overnight at 110 C. ZrO2 and ZnO–ZrO2 were obtained by cal-
◦
cination at appropriate temperature and time.
◦
ZnO–ZrO2 with 22 wt% ZnO loading calcined at 500 C was
treated in dilute HCl solution (0.3 M) at room temperature for
4
h. Then the powder was filtered, washed with excess water,
◦
and dried overnight at 110 C.
WO3 was introduced by impregnation with a precalculated
amount of (NH4)6W12O39 solution overnight, then dried at
◦
◦
1
10 C for 24 h, followed by calcination at 800 C for 3 h.
Zn(OH)2–Zr(OH)4, Zr(OH)4, ZrO2, and ZnO–ZrO2 before and
after acid treatment were used as the supports unless indicated
otherwise. H4PtCl6 solution was used for Pt impregnation on
WO3/(ZnO)–ZrO2. Then the impregnated sample was dried at
◦
◦
1
10 C for 24 h, followed by calcination at 500 C for 3 h.
2
.2. Catalyst characterization
Powder X-ray diffraction (XRD) data were collected using
CuKα radiation on a Bruker D8 Advance diffractometer. An
automated adsorption apparatus (Autosorb-6B, Quantachrome)
was used for specific surface area and pore size distribution
analysis. XPS analysis was carried out using an ESCALAB
Fig. 1. XRD patterns of WO /ZrO with different WO loadings calcined at
3
2
3
2
50 spectroscope (Thermo VG Scientific). FTIR experiments
◦ ◦ ◦ ◦ ◦
00 C (2θ at 30.3 , 35.4 , 50.2 , and 60 are assigned to tetragonal phase,
8
◦
◦
◦
◦
◦
◦
◦
◦
◦
◦
were investigated on a Thermo Nicolet NEXUS 470 FTIR spec-
while 2θ at 24 , 28.3 , 31.5 , 34.3 , 36.4 , 41 , 49.4 , 54.1 , 55.4 , and 65.7
−
1
are due to monoclinic phase of ZrO ).
trometer with 64 scans at 4 cm resolution. All samples were
2

Y. Liu et al. / Journal of Catalysis 244 (2006) 17–23
19
ing the monoclinic phase of ZrO2 decreased with increasing
ZnO contents.
3
.2. Catalytic activity of Pt/WO3/ZrO2 with different WO3
loadings
Pt/WO3/ZrO2 with different WO3 loadings was investigated
for n-heptane isomerization in a fixed-bed reactor under at-
mospheric pressure. Pure ZrO , WO , and Pt/ZrO showed
2
3
2
◦
no activity at 250 C. When WO3 was introduced to Pt/ZrO2,
conversion and C7 isomer selectivity improved significantly
(Fig. 4). The highest conversion of n-heptane (88%) was
achieved with a 20-wt% WO addition. Meanwhile, the high-
3
est selectivity of 90% was obtained with a 15 wt% loading
of WO . (Ignore the selectivity for the catalyst with 5 wt%
3
WO3, because the conversion is too low at 0.1%.) The yields
for the catalysts with 15 and 20 wt% WO3 loadings did not dif-
fer significantly (58 vs 57%). Thus, unless indicated otherwise,
◦
Fig. 2. XRD patterns of ZrO calcined at 500 C (a), ZnO–ZrO (with 22 wt%
2
2
◦
ZnO) calcined at 500 (b), 600 (c), 650 (d), 700 (e), 800 (f), and 900 C (g).
the WO and Pt loadings were fixed at 20 wt% and 0.5 wt%
3
for subsequent studies. Further addition of WO3 led to a de-
crease in both conversion (to 18%) and C7 isomer yield (to
1
5%) for the catalyst with 50 wt% WO3 loading. BET analy-
sis (Fig. 5) indicated that the surface area of Pt/WO3/ZrO2
25 wt%) was still comparable to those of catalysts with lower
(
WO3 loadings. However, when WO3 loading reached 50 wt%,
surface area decreased significantly. Correlating these findings
with the XRD analysis results discussed in Section 3.1 sug-
gests that the isomerization reaction is catalyzed by certain W
atoms working synergistically [11]. Too much W (crystalline
WO ) will have a negative effect on the n-C hydroisomer-
3
7
ization reaction as well as on the surface area. Cortés-Jácome
et al. reported that WO species distributed homogeneously on
x
the surface of the tetragonal ZrO2 are the active species partic-
ipating in the reaction [11,12], and that well-crystallized WO3
phase with large particles is less active for n-hexane isomeriza-
tion [24].
◦
Fig. 3. XRD patterns of WO /ZnO–ZrO (WO : 20 wt%) calcined at 800 C
3
2
3
with different ZnO contents: (a) 0, (b) 0.7, (c) 3.4, (d) 6.8, and (e) 22 wt%.
[
22–24]. When WO3 content exceeds 25 wt%, the XRD pattern
shows the presence of WO3 crystalline phase.
As an additive, ZnO-modified ZnO–ZrO2 was calcined at
different temperatures and characterized by XRD. As shown
◦
in Fig. 2, ZrO2 with 500 C calcination was a mixture of
tetragonal and monoclinic phases, with monoclinic the primary
phase. However, when 22 wt% of ZnO was introduced, at the
◦
same calcination temperature (500 C), ZnO–ZrO2 exhibited
the pure tetragonal phase of ZrO2, which can be maintained
◦
up to 650 C. This suggests that ZnO can also stabilize the
tetragonal phase of ZrO2. At calcination temperatures above
◦
6
50 C, ZnO–ZrO2 showed a mixture of tetragonal and mon-
oclinic phases of ZrO2.
Further XRD studies on Pt/WO3/ZnO–ZrO2 with varying
ZnO content suggested that ZnO can stabilize the tetragonal
phase of ZrO2, as shown in Fig. 3. The XRD peaks represent-
Fig. 4. Catalytic activity of n-C7 hydroisomerization on Pt/WO /ZrO with
3
2
different WO loadings (catalyst: 300 mg; Pt: 0.5 wt%; reduction: H ,
3
2
◦
◦
3
0 ml/min, 400 C, 2 h; reaction: 250 C, 0.1 MPa, H /n-C7 = 30:1).
2

20
Y. Liu et al. / Journal of Catalysis 244 (2006) 17–23
Fig. 7. n-Heptane hydroisomerization activity of Pt/WO /ZnO–ZrO with
3
2
3
.4 wt% ZnO loading (catalyst: 300 mg; Pt: 0.5 wt%; WO : 20 wt%; reduction:
3
◦
◦
Fig. 5. BET analysis for Pt/WO /ZrO with different WO contents (Pt: 0.5
H , 30 ml/min, 400 C, 2 h; reaction: 250 C, 0.1 MPa, H /n-C = 30:1).
3
2
3
2
2
7
wt%) and Pt/WO /ZnO–ZrO with different ZnO contents (Pt: 0.5 wt%; WO :
3
2
3
2
0 wt%).
◦
action temperatures as low as 200 C, both Pt/WO3/ZrO2 and
Pt/WO3/ZnO–ZrO2 with ZnO loadings of 0.7–13.6 wt% gave
low n-C7 conversion with quite high C7 isomer selectivity and
a trace amount of cyclo products, which are all clean blender
for gasoline production.
3
.3. Catalytic activity of Pt/WO3/ZnO–ZrO2 with different
ZnO loadings
Fig. 6 displays n-C7 conversion, selectivity, and yield of
◦
C7 isomers versus ZnO loading at 250 C. Adding ZnO en-
After reaction, Pt/WO3/ZnO–ZrO2 catalysts with different
ZnO loadings were characterized by XPS. The results indicate
that all of the catalysts investigated here gave the same band-
ing energy for either Pt (71.32 and 74.35 eV) or W (37.76 and
35.68 eV) (data not shown), suggesting that ZnO loading did
not affect the reducibility of Pt and WO3 during either reduction
activation or the reaction process for n-C7 hydroisomerization.
The stability of a specific catalyst is an important is-
sue in catalyst design. Long-reaction activity was tested for
hanced both the selectivity and the yield of C7 isomers. The
yield reached a maximum of 72% with 3.4 wt% of ZnO loading,
compared with the 57% C7 isomer yield for ZnO-free catalyst.
However, further addition of ZnO up to 6.8 wt% led to a de-
creased C7 isomer yield of 39%. The ZnO content at 6.8 wt%
is also a critical determinant of surface area, as shown in Fig. 5.
No activity was achieved for a high ZnO loading of 22 wt%,
which also led to a quite low surface area. As reported, only
a specific optimal balance of metal and acid function at suit-
able reaction conditions can suppress cracking to achieve high
isomerization selectivity for long-chain alkanes [25,26]. At re-
Pt/WO /ZnO–ZrO with a 3.4 wt% ZnO loading. As shown
in Fig. 7, n-C conversion remained stable at 81% for at least
82 h at 250 C, with selectivity to C isomers of around 89%.
3
2
7
◦
7
Reproducibility testing for Pt/WO3/ZnO–ZrO2 catalyst with a
3
.4 wt% ZnO loading was also carried out; as shown in Fig. 8,
the activity data for the reprepared catalysts are quite repro-
ducible to the former.
3
.4. Acidity characterization for Pt/WO3/ZnO–ZrO2 with
different ZnO loadings
Acidity was characterized by in situ FTIR using pyridine as
a probe molecule [27,28]. Fig. 9 shows the FTIR spectra of
◦
−1
pyridine desorbed at 150 C in the range of 1700–1400 cm
for catalysts with different ZnO loadings. The bands related to
pyridine adsorbed on Lewis acid sites were recorded at 1451
−
1
and 1611 cm , whereas the pyridine adsorbed on the Brønsted
−
1
acid sites was detected at 1536 and 1640 cm , respectively.
−
1
The band at 1490 cm was attributed to pyridine associated
with both Lewis and Brønsted acid sites [29,30]. The acid prop-
erties of the catalyst are presented using the Brønsted acid-to-
Lewis acid ratio from their specific bands, namely an IR band
Fig. 6. Catalytic activity of n-C7 hydroisomerization on Pt/WO /ZnO–ZrO
3
2
with different ZnO loadings (catalyst: 300 mg; Pt: 0.5 wt%; WO : 20 wt%;
3
◦
◦
−1
−1
reduction: H , 30 ml/min, 400 C, 2 h; reaction: 250 C, 0.1 MPa, H /n-C7
at 1536 cm for Brønsted acid and an IR band at 1611 cm
for Lewis acid. These values were 0.85, 0.72, 0.69, and 0.43 for
2
2
=
30:1).

Y. Liu et al. / Journal of Catalysis 244 (2006) 17–23
21
Table 1
Effects from impregnation procedures for Pt/WO –ZrO
3
2
Solids for
impregnation
T
Conversion
(%)
Selectivity
(%)
Isomer distribution
◦
( C)
MB
DB
TB
Zr(OH)
250
88
99
100
66
4
0
66
56
0
32
40
0
1
4
0
4
300
50
3
ZrO
250
9
68
95
87
76
25
100
78
66
0
22
32
0
0.4
2
2
300
50
3
Catalyst: 300 mg; Pt: 0.5 wt%; WO : 20 wt%; reduction: H , 30 ml/min,
3
2
◦
◦
◦
◦
4
00 C, 2 h; reaction: 250 C, 300 C, 350 C, 0.1 MPa, H /n-C7 = 30:1; MB:
2
monobranched C7 isomers; DB: dibranched C7 isomers; TB: tribranched C7
isomers.
Fig. 8. The reproducibility investigation for conversion, selectivity and prod-
ucts distribution for C7 isomers for Pt/WO /ZnO–ZrO (catalyst: 300 mg; Pt:
3
2
◦
.5 wt%; WO : 20 wt%; ZnO: 3.4 wt%; reduction: H , 30 ml/min, 400 C,
3 2
0
2
2
◦
h; reaction: 250 C, 0.1 MPa, H /n-C7 = 30:1; 2,3-C : 2,3-dimethyl-pentane;
2
5
,4-C : 2,4-dimethyl-pentane; 2,2,3-C : 2,2,3-trimethyl-butane; 3,3-C : 3,3-di-
5
4
5
methyl-pentane; 2-C : 2-methyl-hexane; 2,3-C : 2,3-dimethyl-pentane; 3-C :
6
5
6
3
-methyl-hexane; 2E-C : 2-ethyl-pentane).
5
Fig. 10. Pyridine-FTIR characterizations for Pt/WO /ZrO using either
3
2
Zr(OH) or ZrO for impregnation.
4
2
was achieved when Zr(OH)4 was used as a starting support pre-
cursor for catalyst preparation (by impregnation of W and Pt),
compared with that achieved when using ZrO2 as a starting
◦
support, as reported previously [31]. At 250 C, the n-C7 con-
version for the former reached 88%, compared with only 9%
Fig. 9. Pyridine-FTIR characterizations of Pt/WO /ZrO with different ZnO
loadings.
3
2
◦
for the latter. The activity for the latter at 300 C is only compa-
◦
rable with that for the former at 250 C. However, according to
the isomerization thermodynamic equilibrium [3], the lower re-
action temperature gives high multibranched isomer selectivity.
To investigate the reason for the significant difference in ac-
tivity due to the impregnation procedure used, related character-
izations were carried out. As shown in Fig. 10, pyridine-FTIR
characterization gave Brønsted acid site-to-Lewis acid site ra-
tios of 0.85 and 0.83 for Pt/WO3/ZrO2 using Zr(OH)4 and
ZrO2 for impregnation, respectively. The only slight difference
indicates that acid sites can be created from both impregna-
tion procedures, contrary to findings reported previously [31].
That earlier report claimed that superacid sites are created by
impregnation not on the crystallized oxide, but rather on the
amorphous form, with subsequent calcination converting it to
crystalline form. Fig. 11 displays the crystalline phase differ-
ence between these two samples. Pt/WO3/ZrO2 using Zr(OH)4
Pt/WO3/ZrO2 and Pt/WO3/ZnO–ZrO2 with 0.7 wt%, 3.4 wt%,
and 6.8 wt% ZnO loadings, respectively. When the ZnO loading
reached 22 wt%, pyridine-FTIR analysis detected no Brønsted
acid sites. As shown in Fig. 6, the highest yield of C7 isomers
was achieved for the catalyst with 3.4 wt% ZnO loading, and
C7 isomer yield decreased when ZnO loading was ꢀ6.8 wt%.
This finding suggests that the n-C7 hydroisomerization requires
a specific ratio of Brønsted acid sites to Lewis acid sites on the
catalyst surface.
3
.5. Effects of impregnation procedures for Pt/WO3/ZrO2
The effect of the impregnation procedure was also inves-
tigated. Table 1 presents catalytic data for the two catalysts
prepared with different support precursors. Higher conversion

22
Y. Liu et al. / Journal of Catalysis 244 (2006) 17–23
Fig. 12. SEM micrographs of ZnO–ZrO after acid treatment.
2
Fig. 11. XRD patterns of Pt/WO /ZrO impregnated with either Zr(OH) or
3
2
4
ZrO .
2
3
.6. Acid treatment for ZnO–ZrO2
As discussed in Sections 3.1 and 3.3, Pt/WO3/ZnO–ZrO2
as the catalyst support is a mixture of tetragonal and monoclinic
phases, with tetragonal the dominant phase. However, the cata-
lyst with ZrO2 as the support is a pure monoclinic phase. As
mentioned in Section 3.2, WOx species homogeneously dis-
tributed on the surface of the tetragonal ZrO2 are the active
species for the isomerization reaction [11,12,24]. This may be
why the latter showed low activity for n-C7 hydroisomerization
even though their Brønsted acid site-to-Lewis acid site ratios
were comparable. Another reason for this finding may be that
the surface area for Pt/WO3/ZrO2 using ZrO2 for impregna-
with 22 wt% ZnO maintained the tetragonal phase of ZrO2
but showed no activity for n-C7 hydroisomerization. Indeed,
pyridine-FTIR characterization (Fig. 9) showed the absence of
Brønsted acid sites for this catalyst, due to the basic character
of ZnO, which neutralized all of the Brønsted acid sites from
WO3/ZrO2. To recover the Brønsted acid sites, ZnO–ZrO2 was
treated by diluted hydrochloride acid. The acid-treated catalysts
were then characterized using SEM-EDX, BET, and pyridine
FTIR. It is found that the morphology of the surface (Fig. 12)
was highly coarse compared with the nontreated catalyst, ac-
2
tion was only 44 m /g, lower than that for the catalyst using
2
Zr(OH)4 as the catalyst support (60 m /g). This led to differ-
2
2
ent WOx surface densities, 11.8 and 8.6 W/nm , respectively.
companied by an increase in surface area from 73 to 91 m /g.
It has been reported that WOx surface density influences the
size and electronic properties of WOx domains and the rate of
catalytic reaction [32]. No obvious difference was revealed on
XPS analysis.
The superficial Zn:Zr ratio dropped to 0.15, much lower than
the original Zn:Zr ratio of 0.43. The pyridine-FTIR charac-
terization shown in Fig. 13 indicates that extra Brønsted acid
sites were regenerated after acid treatment; these acid sites re-
Fig. 13. Pyridine-FTIR characterizations before (A) and after (B) in situ H reduction for Pt/WO /ZnO–ZrO with and without acid treatment.
2
3
2

Y. Liu et al. / Journal of Catalysis 244 (2006) 17–23
23
tivity (89%). The reactive crystalline phase is proposed to be
the tetragonal phase, which is stabilized by adding WO3 and
ZnO. For the high-ZnO content (22 wt%) Pt/WO3/ZnO–ZrO2
◦
catalyst, acid treatment at 300 C greatly improved the C7 iso-
mer yield, from nil to 47%.
Acknowledgments
Y. Liu thanks Drs. Keith Carpenter and P.K. Wong for their
support in this study (project code ICES/02-112001).
References
[
1] T.G. Kaufmann, A. Kaldor, G.F. Stuntz, M.C. Kerby, L.L. Ansell, Catal.
Today 62 (2000) 77.
[
[
2] I.E. Maxwell, J.E. Naber, Catal. Lett. 12 (1992) 105.
3] Y. Rezgui, M. Guemini, A. Tighezza, A. Bouchemma, Catal. Lett. 87
Fig. 14. Acid treatment effects on Pt/WO /ZnO–ZrO catalytic activity for
3
2
(
2003) 11.
n-C7 hydroisomerization (catalyst: 300 mg; Pt: 0.5 wt%; WO : 20 wt%; ZnO:
3
◦ ◦
2 wt%; reduction: H , 30 ml/min, 400 C, 2 h; reaction: 300 C, 0.1 MPa,
2
[4] N. Bouchenafa-Saïb, R. Issaadi, P. Grange, Appl. Catal. A 259 (2004) 9.
5] T. Okuhara, J. Jpn. Petrol. Inst. 47 (2004) 1.
6] J.F. Kriz, T.D. Pope, M. Stanciulescu, J. Monnier, Ind. Eng. Chem. Res. 37
2
[
[
H /n-C7 = 30:1).
2
(1998) 4560.
mained detectable after in situ H2 reduction. XRD analysis (not
shown here) indicated no significant difference between these
two samples; both are mixtures of tetragonal and monoclinic
phases.
[7] T.K. Cheung, B.C. Gates, Chemtech. Sep. (1997) 28.
[8] G.A. Olah, J. Am. Chem. Soc. 94 (1972) 808.
[9] E. Iglesia, S.L. Soled, G.M. Kramer, J. Catal. 144 (1993) 238.
[
10] J.G. Santiesteban, D.C. Calabro, W.S. Borghard, C.D. Chang, J.C. Vartuli,
Y.P. Tsao, M.A. Natal-Santiago, R.D. Bastian, J. Catal. 183 (1999) 314.
[11] D.G. Barton, S.L. Soled, G.D. Meitzner, G.A. Fuentes, E. Iglesia, J. Catal.
181 (1999) 57.
ZnO and ZrO before and after acid treatment were investi-
2
gated for n-C7 hydroisomerization after loading of WO3 and Pt.
As shown in Fig. 14, the activity was greatly enhanced at rel-
atively high temperatures, suggesting that the weak and mild
[
12] S. Zhang, Y. Zhang, J.W. Tierney, I. Wender, Appl. Catal. A 193 (2000)
155.
[
[
13] M.A. Arribas, F. Márquez, A. Martínez, J. Catal. 190 (2000) 309.
14] S. Zhang, Y. Zhang, J.W. Tierney, I. Wender, Fuel Process. Technol. 69
acid sites were recovered first. The C isomer yield reached
7
◦
4
4
7% at 300 C (with conversion and C7 isomer selectivity of
9% and 96%, respectively). At this temperature, no activity
(
2001) 59.
[15] P. Pérez-Romo, C. Potvin, J.-M. Manoli, G. Djéga-Mariadassou, J. Catal.
05 (2002) 191.
2
was obtained for the nontreated catalyst Pt/WO /ZnO–ZrO .
3
2
[
[
16] W. Hua, J. Sommer, Appl. Catal. A 232 (2002) 129.
17] E. Iglesia, D.G. Barton, S.L. Soled, S. Moseo, J.E. Baumgartner, W.E.
Gates, G.A. Fuentes, G.D. Meitzner, Stud. Surf. Sci. Catal. 101 (1996)
This suggests that the novel surface properties created after
acid activation (e.g., relatively higher surface area and increased
number of Brønsted acid sites, lower the surface Zn:Zr ratio) do
533.
increase the isomerization activity for Pt/WO /ZnO–ZrO cat-
[18] E. Iglesia, S.L. Soled, G.M. Kramer, J. Catal. 144 (1993) 238.
3
2
[
[
[
[
19] E. Blomsma, J.A. Martens, P.A. Jacobs, J. Catal. 159 (1996) 323.
20] J.W. Ward, Stud. Surf. Sci. Catal. 16 (1983) 587.
21] I.E. Maxwell, W.H.J. Stork, Stud. Surf. Sci. Catal. 58 (1991) 571.
22] R.A. Royse, E.I. Ko, J. Catal. 171 (1997) 191.
alysts. After acid treatment, the activity for Pt/WO3/ZnO–ZrO2
was comparable even to that for Pt/WO3/ZrO2 (using ZrO2 for
impregnation; Table 1). Acid treatment had no significant affect
on the activity for ZrO -supported catalysts.
2
[23] J.G. Santiesteban, J.C. Vartuli, S. Han, R.D. Bastian, C.D. Chang, J. Catal.
68 (1997) 431.
1
[24] M.A. Cortés-Jácome, J.A. Toledo, C. Angeles-Chavez, M. Aguilar, J.A.
Wang, J. Phys. Chem. B 109 (2005) 22730.
4
. Conclusion
[
25] T. Bécue, J.-M. Manoli, C. Potvin, C. Djéga-Mariadassou, J. Catal. 170
Pt/WO3/ZrO2 tested with a WO3 loading optimized at
(1997) 123.
2
0 wt% showed good reactivity toward n-heptane hydroiso-
[26] M.J. Girgis, Y.P. Tsao, Ind. Eng. Chem. Res. 35 (1996) 386.
[27] S. Triwahyono, T. Yamada, H. Hattori, Appl. Catal. A 242 (2003) 101.
◦
merization at relatively high temperatures (around 250 C).
Adding ZnO to the catalyst is an alternative approach to achiev-
ing the highest C isomer yield (72%) at 250 C, with a ZnO
content of 3.4 wt% in this study. This catalyst exhibited a con-
[28] S. Triwahyono, T. Yamada, H. Hattori, Appl. Catal. A 250 (2003) 75.
◦
[29] C.A. Emeis, J. Catal. 141 (1993) 347.
7
[30] Y. Li, W. Zhang, L. Zhang, Q. Yang, Z. Wei, Z. Feng, C. Li, J. Phys. Chem.
B 108 (2004) 9739.
stant n-heptane hydroisomerization performance for at least
[31] K. Arata, M. Hino, Mater. Chem. Phys. 26 (1990) 213.
8
2 h with high n-C conversion (81%) and C isomer selec-
[32] C.D. Baertsh, S.L. Soled, E. Iglesia, J. Phys. Chem. B 105 (2001) 1320.
7
7