Selective catalytic reduction of nitric oxide with hydrogen over Pd-based catalysts

Article abstract of DOI:10.1016/j.jcat.2005.11.025

Catalytic reduction of NO with hydrogen in the presence of excess oxygen was studied using Pd-based catalysts. High steady-state conversion of NO (>80%) is achieved, using 1%Pd/TiO₂/Al₂O₃ as the catalyst and at temperatures ranging from 140 to 180 °C. The addition of V₂O₅ to this catalyst not only increases NO conversion, but also widens the reaction temperature window to 250 °C. The effects of CO, O₂, and H₂ concentrations on the NO conversion were also investigated. It was found that the reaction order is nearly first order with respect to NO, 0.6 with respect to H₂ and -0.18 with respect to CO. Increasing the O₂ concentration from 2 to 10% produced only a slight decrease in NO conversion. FTIR studies showed that a significant amount of NH4+ was formed over the V₂O₅-containing catalyst, during the reaction, at temperatures >200 °C, compared with almost no NH4+ formation over the V₂O₅-free sample. The higher activity of 1%Pd-5%V₂O₅/TiO₂-Al ₂O₃ is partly attributed to the in situ formation of NH4+.

Full text of DOI:10.1016/j.jcat.2005.11.025

Journal of Catalysis 237 (2006) 381–392
www.elsevier.com/locate/jcat
Selective catalytic reduction of nitric oxide with hydrogen
over Pd-based catalysts
Gongshin Qi a, Ralph T. Yang a,∗, Fabrizio C. Rinaldi b
a
Department of Chemical Engineering, University of Michigan, Ann Arbor, MI 48109-2136, USA
b
Tenneco Automotive, Exhaust Engineering Center, Grass Lake, MI 49240, USA
Received 12 September 2005; revised 14 November 2005; accepted 21 November 2005
Abstract
Catalytic reduction of NO with hydrogen in the presence of excess oxygen was studied using Pd-based catalysts. High steady-state conversion
◦
of NO (>80%) is achieved, using 1%Pd/TiO /Al O as the catalyst and at temperatures ranging from 140 to 180 C. The addition of V O
2
2
3
2
5
2
◦
to this catalyst not only increases NO conversion, but also widens the reaction temperature window to 250 C. The effects of CO, O , and H
2
concentrations on the NO conversion were also investigated. It was found that the reaction order is nearly first order with respect to NO, 0.6 with
respect to H and −0.18 with respect to CO. Increasing the O concentration from 2 to 10% produced only a slight decrease in NO conversion.
2
2
+
FTIR studies showed that a significant amount of NH
>200 C, compared with almost no NH formation over the V O -free sample. The higher activity of 1%Pd–5%V O /TiO –Al O is partly
attributed to the in situ formation of NH
was formed over the V O -containing catalyst, during the reaction, at temperatures
4
2 5
◦
+
.
4
2
5
2
5
2
2 3
+
4
2005 Elsevier Inc. All rights reserved.
Keywords: NO reduction by hydrogen; Selective catalytic reduction of NO; Diesel engines; Pd-based catalysts; Lean burn engines; Pd–V O /TiO –Al O
2 3
x
2
5
2
1. Introduction
Similar SCR technology also has been effectively applied
to mobile sources, where ammonia is usually generated by the
thermal decomposition of urea. However, there are some com-
mercial and logistic drawbacks, namely (1) a separate tank and
injection system, (2) issues relating to NH3 slip (i.e., unreacted
ammonia), (3) handling of urea solutions during cold condi-
tions, and (4) as of yet, no real infrastructure to widely deploy
the needed urea solution. These factors indicate that the devel-
opment of an active NOx, reduction catalyst that makes use of
other reductants is desirable. Three-way catalysis is very effec-
tive on emissions from gasoline engines, where narrow-band
oxygen sensors afford a tight closed-loop control of the air:fuel
ratio (∼14.7). In comparison, diesel engines operate very lean
and with a wider band of air:fuel ratio (14–24) [4]. Diesel en-
gines have considerable benefits over gasoline engines, includ-
ing better fuel economy, less CO2 production, higher torque,
cooler exhaust temperatures, and overall greater longevity [4].
However, due to the nature of diesel fuel and of the compres-
sion ignition combustion process, diesel engines emit a high
quantity of particulate matter (PM) and NOx emissions. These
pose a significant air quality issue, accounting for about one-
Nitrogen oxides (NO, NO2, and N2O) resulting from com-
bustion processes continue to be a major source of air pollution.
They contribute to photochemical smog, acid rain, ozone de-
pletion, ground-level ozone, and greenhouse effects. More than
95% of NOx emissions are derived from two sources, ∼49%
from mobile (vehicles) and ∼46% from stationary (power
plants) [1]. Currently the commercially available technology
for reducing NOx (x = 1, 2) emissions from stationary sources
is selective catalytic reduction (SCR). Ammonia is widely ac-
cepted as the reducing agent of choice. Many catalyst formu-
lations have been reported to be active for this reaction. These
can be subdivided into two main groups, mixed oxides and ion-
exchanged molecular sieves [2,3]. Commercially available cat-
alyst formulations contain, among others, vanadia (doped with
MoO3 or WO3) supported on titania.
*
Corresponding author. Fax: +1 734 763 0459.
E-mail address: yang@umich.edu (R.T. Yang).
0021-9517/$ – see front matter
doi:10.1016/j.jcat.2005.11.025
2005 Elsevier Inc. All rights reserved.

382
G. Qi et al. / Journal of Catalysis 237 (2006) 381–392
third of the nation’s NOx emissions and one-quarter of the
PM emissions from mobile sources. In the United States, the
Environmental Protection Agency (EPA) has addressed such
concerns by imposing increasingly stringent emission regula-
tions, the latest wave of which are Tier 2 for light duty vehicles
and both the 2004 Rule and 2010 Highway Rule for heavy-duty
diesel engines (all mobile, on-road applications). Due to these
stringent emission regulations, urea SCR has received renewed
attention as an NOx abatement tool for mobile sources [5–7],
together with NOx traps and exhaust gas recirculation (EGR).
Although urea-SCR is very effective and has high N2 selectiv-
ity, the earlier-mentioned drawbacks remain.
Recently, hydrogen has been used in an effort to regenerate
effectively NOx traps. Its long-term prospects (fueling internal
combustion engines, fuel cell vehicles, on-board reforming of
diesel fuel, producing essentially only water vapor) make hy-
drogen the most likely candidate as a reductant in a SCR reac-
tion to abate NOx emissions in an excess oxygen environment,
such as diesel engine exhaust. The following main reactions oc-
cur in the NO/H2/O2 system:
catalysts. H2 is available from fuel processors for fuel cell ap-
plications. Moreover, CO is available both within raw diesel
engine exhaust and as part of a reformation process. Because
it bonds strongly to catalysts, CO could have a significantly
detrimental effect on the low-temperature oxidation of H2 by
both NO and O2 [13–15]. Macleod and Lambert [16] investi-
gated the influence of CO on both the activity and selectivity of
Pt- and Pd-based catalysts and showed that for the Pt catalysts,
the presence of CO did indeed have a significantly detrimen-
tal effect. Macleod and Lambert [17] reported a highly active
Pd/TiO2/Al2O3 catalyst capable of >90% NO conversion (at a
high space velocity) in the presence of excess oxygen at 140–
180 ◦C with N2 selectivity up to 80%. This catalyst would not
be suitable in a diesel engine exhaust stream, because a wider
operating temperature window is required.
The objective of this work is to investigate and develop SCR
catalyst formulations that use hydrogen as the reducing agent
and can lower the NOx content of a 2004-compliant heavy-duty
engine exhaust to the 2010 Highway Rule levels. The catalyst
developed in this work must have high activity and a wider tem-
perature window than previously developed catalysts, both of
which are critical for practical applications. Another aspect of
our study is steady-state reactions, as opposed to the transient
reactions studied by others, such as Macleod and Lambert [17].
We also aim to clarify the function and role of V2O5 in the Pd-
based catalysts.
2NO + 4H2 + O2 → N2 + 4H2O,
2NO + 3H2 + O2 → N2O + 3H2O,
O2 + 2H2 → 2H2O.
A number of groups have investigated the use of hydrogen as
a reductant over Pt-based catalysts [8–12]. Yokota et al. [8] in-
vestigated NO reduction with H2 and CO in the presence of
excess oxygen and found that the Pt/zeolite catalyst has high
activity with a N2 selectivity of 50%. An improved catalyst, Pt–
Mo–Na/SiO2, also developed by Yokota et al. [8], has a wider
temperature window for NOx conversion and a lower selectiv-
ity to the byproduct, N2O, than the conventional Pt catalysts.
Burch and Coleman [11] investigated NO reduction with H2 in
the presence of excess oxygen over Pt/SiO2 and Pt/Al2O3 cata-
lysts and found the SiO2 supported catalysts to be more active at
lower temperatures than the Al2O3-supported catalyst. Costa et
al. [12] reported an improved Pt-based catalyst with the use of
mixed-oxide La0.5Ce0.5MnO3 supports. The improved catalyst
performance was attributed to the presence of oxygen vacan-
cies on the support surface adjacent to small platinum clusters.
These oxygen vacancies were thought to participate in the re-
action mechanism. Ueda et al. [10] investigated the H2/NO/O2
reaction on Pt- and Pd-based catalysts and found two conver-
sion maxima at 373 and 573 K in the reduction of NO with H2
over supported Pd catalysts. The appearance of two conversion
maxima can be explained by a switch in the reaction pathways:
direct reduction of NO by H2 at approximately 373 K and re-
duction of in situ generated NO2 by H2 at approximately 573 K.
Lee and Gulari [18] studied H2-SCR in the presence of CO on
Pd/Al2O3 prepared by PdCl2 as a precursor and found it to be
more active than Pd/Al2O3 prepared by Pd(NO3)2 as a precur-
sor. More recently, Nanba et al. [19] investigated the reduction
of NO by H2 in the presence of O2 over physical mixtures of
Pt/ZrO2 and secondary catalysts active for the SCR of NO by
ammonia.
2. Experimental
2.1. Preparation of catalysts
The 20 wt%TiO2-on-γ -Al2O3 support was prepared by hy-
drolysis of a solution of Ti[O(CH2)3CH3]4 in the presence of
γ -Al2O3 (PSD-350 grade, BET surface area approximately
350 m2/g, 60–100 mesh; Alcoa). The solid sample was then
dried in air at 500 ◦C for 6 h. The 5%V2O5/20 wt%TiO2-γ -
Al2O3 was prepared by impregnation in 20%TiO2-γ -Al2O3
with an aqueous solution of NH4VO3 in oxalic acid. The
same procedure was used to prepare 5%V2O5/Al2O3 and
5%V2O5/TiO2 (P25, Degussa, BET surface area = 30.6 m2/g).
After impregnation, these catalysts were dried at 120 ◦C for
12 h, then calcined at 500 ◦C in oxygen for 12 h to decompose
the ammonium salt into the corresponding oxide. Palladium
was subsequently impregnated in 5%V2O5/20 wt%TiO2-γ -
Al2O3 and 20wt%TiO2-γ -Al2O3 using a Pd(NH3)4Cl2 aque-
ous solution. The catalyst was then dried at 120 ◦C for 12 h and
calcined at 500 ◦C for 6 h in oxygen.
2.2. Catalytic activity measurement
The activity measurement was carried out in a fixed-
bed quartz reactor. The typical reactant gas composition was
500 ppm NO, 4000 ppm H2, 0–2000 ppm CO (when used), 5%
O2, and the balance He. A 100-mg sample was used in each
run. The total flow rate was 200 ml/min (under ambient condi-
tions). The premixed gases (1.01% NO in He, 5.00% H2 in He,
and 1.0% CO in He) were supplied by Matheson. Water vapor
In summary, hydrogen has proven to be a promising reduc-
tant for NO under lean burn conditions over Pt- and Pd-based

G. Qi et al. / Journal of Catalysis 237 (2006) 381–392
383
was generated by passing He through a heated saturator con-
taining deionized water. The NO and NO2 concentrations were
continually monitored using a chemiluminescent NO/NOx an-
alyzer (Thermo Environmental Instruments model 42C). The
products were analyzed using a gas chromatograph (Shimadzu
model 8A) with a 13X molecular sieve column for H2, CO,
and N2 separation and a Porapak Q column for N2O. Ammo-
nia formation was monitored by FTIR. In no case was ammonia
detected by FTIR under lean burn conditions, consistent with
the report by Shelef et al. [20]. The catalytic activity was based
on the calculated NOx (NO + NO2) conversion using the fol-
lowing formula:
inlet NOx(ppm) − outlet NOx(ppm)
NOx conversion =
inlet NOx(ppm)
× 100 (%).
The N2 selectivity was calculated as follows:
[N2]
N2 selectivity =
× 100 (%).
Fig. 1. XRD patterns of the catalysts: (a) Al O , (b) 20%TiO /Al O ,
2
3
2
2 3
[N2] + [N2O]
(c) 5%V O /20%TiO /Al O , (d) 1%Pd–5%V O /TiO /Al O , (e) 1%Pd/
2
5
2
2
3
2
5
2
2 3
20%TiO /Al O , (f) 1%Pd–5%V O /Al O .
Because the reactions were carried out at relatively low tem-
perature, part of the decrease in NO concentration could be
attributed to the adsorption of NO onto the catalysts. Thus, to
ensure that this did not occur, at the beginning of each exper-
iment the catalyst was purged with reactant gas until the inlet
and outlet NO concentrations were equal (i.e., 500 ppm). Sub-
sequently, the temperature was raised to the desired level. At
each reaction temperature, NO conversion and product analysis
was performed after allowing the reaction to reach steady state
(usually 1–2 h, depending on the reaction).
The nitrogen balance was calculated for each step using
the following equation: inlet [NO] = outlet {[NO] + 2[N2] +
2[N2O]}. This was found to be >95% for all experiments.
Steady-state kinetic studies for the NO reduction by H2 in
the presence of CO and O2 were also carried out for the 1%Pd–
5%V2O5/TiO2/Al2O3 catalyst, using a fixed-bed quartz flow
reactor, with 5 mg of catalyst used in each run. The NO con-
centration in an exhaust was simulated by blending different
gaseous reactants. The typical reactant gas composition was 0–
5000 ppm H2, 100–500 ppm NO, 0–500 ppm CO, 1–5% O2,
and the balance He. The total flow rate was 500 ml/min (un-
der ambient conditions). The same instrumentation as described
above was used throughout.
2
2
3
2
5
2 3
was monitored using a thermal conductivity detector. The water
produced during the reduction was trapped in a 5-Å molecular
sieve column.
Infrared spectra were recorded on a Nicolet Impact 400
FTIR spectrometer with a TGS detector. The sample was pre-
pared as a self-supporting wafer 1.3 cm in diameter. This was
achieved by compressing 15 mg of the sample. The wafer was
then loaded into the IR cell (BaF2 windows). The wafers were
pretreated at 573 K in a flow of high-purity O2/He for 0.5 h,
then cooled to room temperature. At each temperature step,
the background spectrum was recorded in flowing O2/He. This
spectrum was subsequently subtracted from the sample spec-
trum obtained at the same temperature step. Thus the IR absorp-
tion features that originated from the structure vibrations of the
catalyst were eliminated from the sample spectra. IR spectra
were recorded by accumulating 100 scans at a spectra resolu-
tion of 4 cm−1
.
3. Results
3.1. Characterization of catalysts
2.3. Catalyst characterization
The XRD patterns of the catalysts are shown in Fig. 1. Crys-
talline PdO phases were not detected in any samples, indicating
that Pd was highly dispersed on the support. Four diffraction
peaks with 2θ = 25.3, 37.5, 39.4, and 48 were observed in the
titania-containing samples. These peaks originated from the
anatase form of titania.
Powder X-ray diffraction (XRD) measurements were carried
out on the catalysts using a Rigaku Rotaflex D/Max-C system
with a Cu-Kα (λ = 0.1543 nm) radiation source. The samples
were loaded with a depth of 1 mm.
In each H2 temperature-programmed reduction (TPR) ex-
periment, 50 mg of sample was loaded into a quartz reactor
and then pretreated with O2/He (100 ml/min) flow at 500 ◦C
for 0.5 h. The sample was then cooled to room temperature
in an O2/He flow. Sample reduction was carried out starting at
room temperature to 600 ◦C in a 5.32%H2/N2 flow (40 ml/min)
with a temperature ramp of 10 ◦C/min. The consumption of H2
H2-TPR profiles of the 1%Pd–5%V2O5/TiO2–Al2O3 and
1%Pd/TiO2–Al2O3 catalysts are shown in Fig. 2. All samples
demonstrate one main peak, which can be assigned to the re-
duction of Pd (II) to Pd(0). The reduction peak temperature is
46 ◦C higher on the V2O5-containing sample than on the V2O5-
free sample. It is believed that V2O5 retards the reduction of Pd
oxides.

384
G. Qi et al. / Journal of Catalysis 237 (2006) 381–392
Fig. 2. TPR profiles of 1%Pd–5%V O /TiO /Al O (upper) and 1%Pd/TiO /
2
5
2
2
3
2
Fig. 4. H conversion as a function of temperature over various catalysts. Reac-
2
Al O catalysts (lower).
2
3
tion conditions: 0.1 g catalyst, total flow rate = 200 ml/min, [NO] = 500 ppm,
[O ] = 5%, [H ] = 4000 ppm, He = balance.
2
2
the addition of V2O5 widens the high NO conversion tem-
perature window to 140–250 ◦C (80% NO conversion). The
1%Pd/TiO2/Al2O3 catalyst exhibited a narrower NO conver-
sion temperature window of 140–180 ◦C. The NO conversion
order was 1%Pd–5%V2O5/TiO2/Al2O3 > 1%Pd/TiO2/Al2O3
> 5%V2O5/TiO2/Al2O3, based on the investigated temperature
range.
Hydrogen conversions as a function of temperature on var-
ious TiO2/Al2O3 supported catalysts are shown in Fig. 4. The
H2 conversion for each catalyst reached 100% at the temper-
atures where the maximum NO conversion was observed; on
the 1%Pd/TiO2/Al2O3 catalyst, this was achieved at 150 ◦C. In
contrast, on the 1%Pd–5%V2O5/TiO2/Al2O3 catalyst, 98% H2
conversion was reached at 150 ◦C. The 5%V2O5/TiO2/Al2O3
catalyst, on the other hand, reached complete H2 conversion
only at 240 ◦C.
For the best catalyst, 1%Pd–5%V2O5/TiO2/Al2O3, Fig. 5
shows the NO–H2–O2 reaction activities, including NO conver-
sion, H2 conversion, and N2 selectivity, at various space veloc-
ities. Generally, the space velocity significantly influenced the
low-temperature NO conversion but not the high-temperature
NO conversion. As the space velocity was increased (from
1.0 × 105 to 1.8 × 106 h−1), NO conversion decreased, and
maximum conversion shifted toward higher temperatures (150–
200 ◦C). H2 conversion exhibited a similar trend—namely, as
the space velocity increased, H2 conversion decreased. The cor-
responding 100% H2 conversion temperature also increased.
No significant space velocity effects were visible on the N2 se-
lectivity at temperatures below 200 ◦C. Significantly decreased
N2 selectivity was observed in the temperature range 200–
300 ◦C.
Fig. 3. NO conversion as a function of temperature over various catalysts. Reac-
tion conditions: 0.1 g catalyst, total flow rate = 200 ml/min, [NO] = 500 ppm,
[O ] = 5%, [H ] = 4000 ppm, He = balance.
2
2
3.2. NOx reduction on Pd-based catalysts by H2 in the
presence of oxygen
The performance of a range of Pd-based catalysts is shown
in Fig. 3. A maximum NO conversion of 98% was achieved
at 150 ◦C over the 1%Pd–5%V2O5/TiO2/Al2O3 catalyst. This
reaction displayed only one NO conversion peak. Under the
same conditions, the 1%Pd/TiO2/Al2O3 catalyst also showed
high NO conversion. However, two separated NO conversion
peaks at 150 and 240 ◦C were observed. These findings are
consistent with the reports of Ueda et al. [10] and Macleod
and Lambert [16,17]. Whereas a maximum NO conversion
of 80% was obtained over the 5%V2O5/TiO2/Al2O3 cata-
lyst, the peak conversion temperature was the same as the
second NO conversion peak for the 1%Pd/TiO2/Al2O3 cata-
lyst. This could indicate that the second NO conversion peak
may not depend on the Pd concentration. Fig. 3 shows that
adding V2O5 to the 1%Pd/TiO2/Al2O3 catalyst increased the
NO conversion, especially at temperatures around 200 ◦C. Thus
Fig. 6 shows the conversions of NO and H2 as well as
the N2 selectivity of various supported Pd catalysts as a
function of temperature. It can be seen that at low temper-
atures (<170 ◦C), the 1%Pd–5%V2O5/TiO2 catalyst had a
very low activity, whereas 1%Pd–5%V2O5/TiO2/Al2O3 cata-
lyst and 1%Pd–5%V2O5/Al2O3 catalyst displayed relatively
high activity under the same conditions. In the high-temperature

G. Qi et al. / Journal of Catalysis 237 (2006) 381–392
385
(a)
(a)
(b)
(b)
(c)
Fig. 5. Effect of space velocity on NO (a), H conversions (b), and N
2
2
selectivity (c), for H -SCR over 1%Pd–5%V O /TiO /Al O catalyst: ",
2
2
5
2
2 3
5
−1
5
−1
6
−1
1.0 × 10
h
; 2, 5.2 × 10
h
; !, 1.8 × 10
h . Reaction conditions:
(c)
Fig. 6. NO (a), H conversions (b), and N selectivity (c), as functions of tem-
[NO] = 500 ppm, [H ] = 4000 ppm, [O ] = 5%, He = balance.
2
2
2
2
perature over various supported Pd based catalysts. Reaction conditions: 0.05 g
range (>240 ◦C), all of the samples showed similar activity.
The order of increasing activity of these Pd-based catalysts
was 1%Pd–5%V2O5/TiO2/Al2O3 > 1%Pd–5%V2O5/Al2O3 >
1%Pd–5%V2O5/TiO2. The same trend was observed for H2
conversion. Fig. 6 also shows the N2 selectivity of different cat-
alysts; all samples had a similar N2 selectivity.
catalyst, [NO] = 500 ppm, [H ] = 4000 ppm, [O ] = 5%, He = balance, total
2
2
flow rate = 500 ml/min.
and 1%Pd–5%V2O5/TiO2–Al2O3 (at 150 ◦C). At 150 ◦C, the
1%Pd–5%V2O5/TiO2–Al2O3 catalyst had a high NO conver-
sion rate to reach steady state in only 1 h. A similar trend was
observed for the 1%Pd–5%V2O5/Al2O3 catalyst at 170 ◦C. At
150 ◦C, however, the 1%Pd–5%V2O5/Al2O3 catalyst displayed
Fig. 7 displays NO conversions as a function of reaction
time for catalysts 1%Pd–5%V2O5/Al2O3 (at 150 and 170 ◦C)

386
G. Qi et al. / Journal of Catalysis 237 (2006) 381–392
(a)
◦
Fig. 7. NO conversion as a function of time on stream at 170 C over
◦
1%Pd–5%V O /Al O ("), at 150 C over 1%Pd–5%V O /Al O (Q), and
2
5
2
3
2
5
2 3
1%Pd–5%V O /TiO –Al O (!) catalysts for H -SCR. Reaction conditions:
2
5
2
2
3
2
0.05 g catalyst, [NO] = 500 ppm, [H ] = 4000 ppm, [O ] = 5%, He = balance,
2
2
total flow rate = 500 ml/min.
a slow increase in NO conversion and reached only 60% NO
conversion after 3 h. The 1%Pd–5%V2O5/TiO2–Al2O3 cata-
lyst had a higher conversion rate; thus it can be concluded that
modification of Al2O3 by titania had a positive effect on the
catalyst conversion rates. Fig. 7 also demonstrates that the H2-
SCR reaction needs to be studied after it has reached steady
state and that transient study results must be interpreted with
caution.
(b)
3.3. NOx reduction on Pd-based catalysts by H2 + CO in the
presence of oxygen
The presence of CO in diesel engine exhaust gas prompted
an investigation of the effects of CO on the H2-SCR process.
Fig. 8 shows the influence of CO on the H2-SCR reaction
over the 1%Pd–5%V2O5/TiO2–Al2O3 catalyst with the total
reductant concentration at 4000 ppm. Clearly, the presence of
CO actually decreased NO conversion, especially in the low-
temperature range. This result was different from that reported
by Macleod and Lambert [16,17], who observed enhancement
by CO on Pd/TiO2/Al2O3. As the CO concentration was in-
creased, the temperature of maximum NO conversion shifted
from 150 to 200 ◦C. The presence of CO clearly inhibited the
H2 oxidation reaction, because the temperature to reach com-
plete oxidation of H2 increased from 170 to 200 ◦C as CO was
added to the feed gas. In the low-temperature range, CO has a
complex effect on N2 selectivity, whereas at high temperatures,
N2 selectivity was almost independent of the presence of CO.
(c)
Fig. 8. NO (a), H , CO conversions (b) and N selectivity (c) as functions of
2
2
temperature over 1%Pd–5%V O /TiO –Al O catalyst on different feed com-
2
5
2
2 3
positions. H conversion (solid line), CO conversion (dashed line). Reaction
2
3.4. Kinetics studies for H2-SCR in the presence of CO and
excess O2 in a differential reactor
conditions: 0.1 g catalyst, total flow rate = 200 ml/min, [NO] = 500 ppm,
[O ] = 5%, [H ] = 2000–4000 ppm, [CO] = 0–2000 ppm, He = balance.
2
2
concentration of H2 was varied between 1000 and 5000 ppm.
Because the flow rate was 500 ml/min and only 5 mg of cat-
alyst was used, <20% NO conversion was obtained at 200 ◦C
in these experiments. Therefore, the reactor can be treated as
a differential reactor. The experimental results on the rate of
To determine the reaction order with respect to NO, the con-
centrations of H2, CO, and O2 were kept constant while the
concentration of NO was varied from 100 to 500 ppm. Sim-
ilarly, to determine the reaction order with respect to H2, the
concentration of NO and CO were kept constant while the

G. Qi et al. / Journal of Catalysis 237 (2006) 381–392
387
(a)
(a)
(b)
(b)
Fig. 10. Dependence of NO conversion rate (upper) and reaction order (lower)
◦
Fig. 9. Dependence of NO conversion rate (upper) and reaction order (lower)
on H concentration for 1%Pd–5%V O /TiO /Al O catalyst at 200 C. Re-
2
2
5
2
2 3
◦
on NO concentration on 1%Pd–5%V O /TiO /Al O catalyst at 200 C. Re-
action conditions: 5 mg catalyst, [NO] = 500 ppm, [CO] = 500 ppm, [O ] =
2
5
2
2
3
2
action conditions: 5 mg catalyst, [H ] = 4000 ppm, [CO] = 500 ppm, [O ] =
5%, He balance, total flow = 500 ml/min.
2
2
5%, He balance, total flow = 500 ml/min.
slightly. Low O2 concentrations (<1%) were not investigated,
due to the propensity to form ammonia. The reaction order (m)
with respect to O2 was calculated as −0.04.
NO conversion as a function of NO, CO, H2, and O2 concen-
trations are presented in Figs. 9–12. Fig. 9 shows that the rate
of NO conversion increased linearly as a function of NO con-
centration. The reaction rate of NO conversion as a function of
reactant concentrations can be expressed as
According to the foregoing results, the H2-SCR reaction in
the presence of CO and excess O2 can be considered approx-
imately first order with respect to NO, 0.6-order with respect
to H2, −0.18-order with respect to CO, and −0.04-order with
respect to O2. The rate of NO conversion can be expressed as
follows:
x
y
z
m
rNO = −ka[NO] [H2] [CO] [O2] ,
(1)
where rNO is the SCR rate, ka is the apparent rate constant, and
x, y, z, and m are reaction orders for NO, H2, CO, and O2,
respectively. According to Fig. 9, the reaction order (x) with
respect to NO was 0.92, thus making this reaction close to first
order.
0.92
0.6
rNO = −ka[NO] [H2] [CO]−0.18[O2]−0.04
.
(2)
Figs. 9–12 also show the corresponding reaction order for each
reactant at different concentrations. For all reactants, the reac-
tion order increased as the concentration was increased.
Fig. 10 shows that the rate of NO conversion increased with
increasing H2 concentration. The reaction order (y) with re-
spect to the H2 concentration was calculated as 0.6. In contrast,
Fig. 11 shows that the rate of NO conversion decreased with in-
creasing CO concentration. Thus CO inhibits the reaction, due
to the competitive adsorption between NO and H2, as discussed
in detail later. The reaction order (z) with respect to CO con-
centration was calculated to be −0.18. Fig. 12 shows the rate
of NO conversion as a function of O2 concentration (500 ppm
NO, 4000 ppm H2, 500 ppm CO). With an increase in O2 con-
centration from 2 to 10%, the NO consumption rate decreased
3.5. Effects of H2O and SO2 on the H2-SCR reaction for the
Pd–V2O5/TiO2–Al2O3 catalyst
Water vapor, a main component of diesel engine exhaust,
often leads to catalyst deactivation. Therefore, resistance of the
NOx abatement catalyst to deactivation by water vapor is a very
important factor. Fig. 13 shows the effect of H2O on SCR ac-
tivity of the Pd–V2O5/TiO2–Al2O3 catalyst. Note that the SCR

388
G. Qi et al. / Journal of Catalysis 237 (2006) 381–392
(a)
(a)
(b)
(b)
Fig. 11. Dependence of NO conversion rate (upper) and reaction order (lower)
◦
Fig. 12. Dependence of NO conversion rate (upper) and reaction order (lower)
on CO concentration for 1%Pd–5%V O /TiO /Al O catalyst at 200 C. Re-
2
5
2
2 3
◦
on O concentration for 1%Pd–5%V O /TiO /Al O catalyst at 200 C. Re-
action conditions: 5 mg catalyst, [H ] = 4000 ppm, [NO] = 500 ppm, [O ] =
2
2
5
2
2 3
2
2
action conditions: 5 mg catalyst, [H ] = 4000 ppm, [NO] = 500 ppm, [CO] =
5%, He balance, total flow = 500 ml/min.
2
500 ppm, He = balance, total flow = 500 ml/min.
reaction was allowed to stabilize for 1 h at 200 ◦C before water
was added.
The addition of 2.3% H2O produced a barely detectable de-
crease in NO conversion. After the water vapor was removed,
activity was quickly restored to its original level.
The effect of SO2 on SCR activity is another important fac-
tor in the H2-SCR reaction due to the presence of very small
amounts of sulfur in diesel fuel. Fig. 14 indicates the effect of
SO2 on SCR activity of the Pd–V2O5/TiO2–Al2O3 catalyst.
Results indicate that a 20-ppm concentration of SO2 at
200 ◦C rapidly decreased the NO conversion rate within the first
2 h of the reaction. Subsequently, the reaction slowly stabilized
to yield a 46% conversion (down from 85%) after 4 h. Removal
of the SO2 feed restored the activity to ∼75%.
Fig. 13. Effect of H O on NO conversion over 1%Pd–5%V O /TiO /Al O
2 3
2
2
5
2
◦
3.6. FTIR studies
catalyst at 200 C. Reaction conditions: 0.05 g catalyst, [H ] = 4000 ppm,
2
[NO] = 500 ppm, [O ] 5%, He = balance, [H O] = 2.3%, total flow =
2
2
500 ml/min.
Fig. 15 shows the FTIR spectra obtained at 150, 200, and
240 ◦C (4000 ppm H2 + 500 ppm NO + 5% O2) over the
1%Pd/TiO2–Al2O3 catalyst. At 150 ◦C, two broad peaks are
seen at 1600 and 1300 cm−1. The various bands observed at
1600–1550 and 1300 cm−1 may be assigned to the asymmet-
ric and symmetric stretching modes of variously coordinated
nitrates [27–30,37]. The bands at 1904 and 1848 cm−1 can be
attributed to gas phase or weakly adsorbed NO [31–33]. The

G. Qi et al. / Journal of Catalysis 237 (2006) 381–392
389
Fig. 14. Effect of SO on NO conversion over 1%Pd–5%V O /TiO /Al O
2 3
2
2
5
2
◦
catalyst at 200 C. Reaction conditions: 0.05 g catalyst, [H ] = 4000 ppm,
2
Fig. 16. FTIR spectra obtained at different temperatures in a flow containing
[NO] = 500 ppm, [O ] 5%, He = balance, [SO ] = 20 ppm, total flow =
2
2
4000 ppm H + 500 ppm NO + 5% O over 1%Pd–5%V O /TiO –Al O .
2
2
2
5
2
2 3
500 ml/min.
◦
(a) 150, (b) 200, and (c) 240 C.
+
4
Fig. 17. Changes in the FTIR spectra of NH and NH
upon switching the
3
Fig. 15. FTIR spectra obtained at different temperatures in a flow containing
gas flow from (4000 ppm H + 500 ppm NO + 5% O ) to (500 ppm NO +
2
2
4000 ppm H + 500 ppm NO + 5% O over 1%Pd/TiO –Al O . (a) 150,
2
2
2
2 3
5% O ) over 1%Pd–5%V O /TiO –Al O catalyst. Time after switching gas
◦
2
2
5
2
2 3
(b) 200, and (c) 240 C.
flow: (a) 0, (b) 1.5, (c) 3, (d) 10 min.
band at 1610 cm−1 can be assigned to adsorbed NO2 [30,31].
The band at 1270 cm−1 is attributed to the deformation modes
of adsorbed NH3 [34–36]. Another band at 1620 cm−1 due to
the deformation modes is barely detected because it is over-
lapped by the band for nitrate. A very weak band at 1740 cm−1
may be assigned to Pd0–NO [34–36]. All of the bands asso-
ciated with nitrates and nitrite decreased in intensity as the
temperature increased from 150 to 240 ◦C, whereas the band
at 1270 cm−1 due to adsorbed NH3 intensified and dominated
at high temperatures.
Fig. 16 shows the corresponding FTIR spectra obtained
over 1%Pd–5%V2O5/TiO2–Al2O3. At 150 ◦C (spectrum a
of Fig. 16), the bands due to Pd0–NO (1740 cm−1), gas-
phase or weakly adsorbed NO (1904 and 1837 cm−1), NO2
(1611 cm−1), and nitrate (1583, 1348, and 1300 cm−1) are also
observed, similar to the corresponding spectra in Fig. 15. Two
new weak bands observed at 1460 and 1680 cm−1 were as-
signed to the symmetric and asymmetric bending modes of
+
NH4 [16,17,37–40,43,44]. With increasing temperature, the
bands related to nitrates/nitrite decreased significantly, whereas
the bands related to ammonium and ammonia first increased
(at 200 ◦C) and then decreased slightly at 240 ◦C. As the tem-
perature was increased from 150 to 240 ◦C, two significant
differences were observed: (1) The intensity of the bands due
to NH4+ increased markedly, and (2) a new band at 1510 cm−1
due to amide (NH2) [41,42] was fo+rmed. Figs. 15 and 16 show
that a significant amount of NH4 was formed at tempera-
tures above 200 ◦C on 1%Pd–5%V2O5/TiO2–Al2O3, which
was the reason that the activity was so high at 200 ◦C on 1%Pd–
5%V2O5/TiO2–Al2O3 compared with 1%Pd/TiO2–Al2O3.
+
To study the reactivity of NH4 and NH3 with NO in the
presence of O2, FTIR spectra (Fig. 17) were obtained over
the 1%Pd–5%V2O5/TiO2–Al2O3 catalyst at 200 ◦C after the
switch from a feed gas containing 4000 ppm H2, 500 ppm NO,
and 5% O2 to one containing 500 ppm NO and 5% O2. After

390
G. Qi et al. / Journal of Catalysis 237 (2006) 381–392
1.5 min, the band at 1460 cm−1 due to NH4 almost disap-
peared (due to reaction with NO + O2), whereas the band at
1270 cm−1 due to+NH3 exhibited a slight variation. The result
indicates that NH4 is more reactive than NH3 in this SCR re-
action; a similar result was reported by Centeno et al. [36], who
used a V2O5/Al2O3 catalyst for the NH3 SCR reaction. After
10 min, both NH4+ and NH3 peaks disappeared, indicating that
both species can react with NO + O2 at different rates.
dation of Pd. In the absence of oxygen or at very low oxygen
partial pressure (<1%), the main products of the H2-SCR are
nitrogen and ammonia [14,20,50]. This investigation focused
on oxygen concentrations ꢀ2%, to ensure that no ammonia was
produced. It became apparent that O2 concentrations >2% have
a negative impact on the reaction. Similar results were reported
by Frank et al. [9] for H2-SCR under lean burn conditions on a
Pt–Mo–Co/Al2O3 catalyst.
The reaction order with respect to NO was found to be ∼1,
which is consistent with a conventional NH3-SCR reaction. The
effect of H2 concentration was positive. It has been reported that
NO dissociation is promoted by atomic hydrogen adsorbed on
platinum [45–47], and the same process would be expected to
occur on palladium. Hecker and Bell [47] found a positive order
dependence of the NO/H2 reaction on H2 and proposed that
NO dissociation proceeded with the assistance of an adsorbed
H atom via the following reaction:
+
4. Discussion
The present work has shown that a 1%Pd–5%V2O5/TiO2/
Al2O3 catalyst offers high NO conversion using H2 as a reduc-
tant in the presence of excess oxygen and at a very high space
velocity. Compared with 1%Pd/TiO2/Al2O3, this catalyst also
has higher NO reduction activity, as well as a wider operating
temperature window.
As shown in Fig. 3, the activity of NO reduction by H2 in the
presence of O2 is very low at 100 ◦C, whereas at 150 ◦C, very
high NO conversion was observed on 1%Pd/TiO2–Al2O3 and
1%Pd–5%V2O5/TiO2–Al2O3 catalysts. As the reaction tem-
perature was increased, NO conversion decreased on both the
1%Pd/TiO2–Al2O3 and 1%Pd–5%V2O5/TiO-Al2O3 catalysts
but increased on the 5%V2O5/TiO2–Al2O3 catalysts. Two dis-
tinct maximum NO conversion peaks were observed on the
1%Pd/TiO2–Al2O3 catalyst, as previously reported by Ueda
et al. [10] and Lambert et al. [16,17,20]. Conversely, only
one maximum NO conversion peak was observed with the
5%V2O5/TiO2–Al2O3 and 1%Pd–5%V2O5/TiO2–Al2O3 cata-
lysts. Clearly, the modifications to the basic catalyst imposed
by the addition of V2O5, significantly widened the operating
temperature window. A similar trend (Pd/V2O5/Al2O3) was ob-
served by Macleod and Lambert [22]. Burch and Coleman [24]
also reported that adding molybdenum as a promoter resulted
in increased NO conversion and N2 selectivity for the H2-SCR
reaction over a Pt/Al2O3 catalyst in the presence of oxygen.
Our results indicate that the low-temperature NO conver-
sion peak was always obtained at the H2 light-off temperature.
This also occurred for the 5%V2O5/TiO2–Al2O3 catalyst. Fig. 4
shows that H2 light-off occurred at a relatively higher temper-
ature for the 1%Pd–5%V2O5/TiO2/Al2O3 catalyst compared
with 1%Pd/TiO2–Al2O3.
It was reported that the low temperature NO conversion
peak was related to the adsorption and dissociation of NO and
H2 on palladium metal [21–24,26]. The adsorbed NO disso-
ciates to adsorbed atomic oxygen and atomic nitrogen at above
100 ◦C [25]. A probable explanation for our results is that V2O5
captures some form of reduced nitrogen compound (e.g., NH3
or a NHx-type species) that may be formed either on V2O5 or
abstracted from the nearby Pd sites. A similar reaction pathway
was reported by Burch and Coleman [24], who made use of the
Mo oxides on a Pt-Mo/Al2O3 catalyst and confirmed this trend
by FTIR.
NOad + Had → Nad + OHad.
To elucidate the role of V2O5 on the reaction, the FTIR spec-
tra for both V2O5-containing and V2O5-free catalysts were
studied in this work. Fig. 15 shows that over the 1%Pd/TiO2–
Al2O3 catalyst, NH3 species were detected at all tempera-
tures with different intensities. At low temperature (150 ◦C),
nitrate was the dominating species on the surface with a small
amount of NH3 (spectrum (a) in Fig. 15). At higher temperature
(>200 ◦C), bands due to NH3 dominated. At all the tempera-
+
tures investigated, no bands due to NH4 were detected. The
1%Pd–5%V2O5/TiO2–Al2O3 catalyst gave significantly differ-
ent spectra (Fig. 16). At 150 ◦C, nitrate species were prevalent
on the surface with a small fraction of NH3 and a trace amount
+
of NH4 (spectrum (a) of Fig. 16). This is similar to what
occurred over the 1%Pd/TiO2–Al2O3 catalyst. When the tem-
perature is increased above 200 ◦C, the bands due to the nitrate
+
species markedly decreased with the NH4 bands dominating.
Thus, the addition of V2O5 to the catalyst results in the for-
mation of NH4+, which is also the intermediate produced in
the NH3-SCR reaction over V2O5/TiO2 [39] and Fe-ZSM-5
[43,44] type catalysts. The rate of reaction between NH4+/NH3
and NO + O2 was also investigated with FTIR, shown in
+
Fig. 17. The results illustrate that NH4 is much more reac-
tive than NH3 in the SCR reaction. The FTIR results indicate
a switch of reaction mechanism between 150 and >200 ◦C. At
low temperatures, NO dissociation is believed to be the cru-
cial step dominating the overall reaction; at high temperatures,
the NH3-SCR reaction is the crucial step. Similar results have
been discussed previously [9,16,17]. One difference between
this system and the standard NH3-SCR reaction is that in this
system some NO dissociation was still present at high temper-
atures, due to the high level of N2O formed.
More recently, Macleod et al. [48,49] studied the H2/NO/O2/
CO reaction on Pd/TiO2–Al2O3 and Pd/Al2O3 catalysts and ex-
amined the synergy of titania and alumina by DRIFTS, XPS,
HREM, and XRD. They concluded that titania was critical for
the formation of NCO and for the alumina-promoted hydrol-
ysis of NCO to ammonia. Here we focused on the H2/NO/O2
reaction on V2O5-containing Pd/TiO2–Al2O3 catalyst because
A kinetic study carried out at a relatively high temperature
(200 ◦C) demonstrated a decreasing rate of NO conversion with
increasing partial pressure of oxygen. This effect was attributed
to the competing H2/O2 reaction as well as the probable oxi-

G. Qi et al. / Journal of Catalysis 237 (2006) 381–392
391
Fig. 18. Reaction scheme of H -SCR of NO on Pd/V O /TiO –Al O catalysts.
2
2
5
2
2 3
of the negative effect of CO found in this study. In the present
study, a mechanism of H2-SCR of NO on Pd/V2O5/TiO2–
Al2O3 was proposed (Fig. 18). The reaction starts with the
dissociation of H2 into H atoms and the dissociation of NO into
N and O atoms on the Pd metal surface. N2 was produced by the
reaction of N to another N, whereas N2O was produced by the
reaction of N to adsorbed NO. The fact that the reaction order
of NO was nearly first order indicates that NO dissociation and
adsorbed NO were rapidly equilibrated with gaseous NO. NH3
was produced by consecutive hydrogenation of N, and then
the formed NH3 reacted with (possibly involving spillover) the
Brönsted acid sites to form NH4+. On the Brönsted acid sites
of the catalyst, N2 was produced by ammonia and ammonium
reacted to NO and O2 (as in the ammonia-SCR reaction stud-
ied extensively in the literature). From this mechanism, it can
be concluded that this reaction is basically a bifunctional reac-
tion mechanism on Pd and support. Based on the activity results
(Fig. 3), at low temperature, NO conversion was related mainly
to the Pd metal, because very low NO conversion was reached
on Pd-free samples. In the high-temperature range (>240 ◦C),
NO conversion did not differ much on Pd-containing and Pd-
free samples, which means that the supports including V2O5
have played a significant role. Based on different NO conver-
sions on Pd/V2O5/TiO2–Al2O3 and V2O5/TiO2–Al2O3, it can
be concluded that at 150 ◦C, most of the NO conversion oc-
curred on Pd alone; at 240 ◦C, most of the NO conversion
occurred on supports; and at 200 ◦C, nearly half of the NO con-
version occurred on Pd alone and the supports.
reactive than NH3 in these SCR reactions, to which the higher
activity of the 1%Pd–5%V2O5/TiO2–Al2O3 catalyst was at-
tributed.
Acknowledgments
This work was supported by Tenneco Automotive, Exhaust
Engineering Center.
References
[1] L.R. Raber, Chem. & Eng. News April 14 (1997) 10.
[2] H. Bosch, F. Janssen, Catal. Today 2 (1988) 369.
[3] G. Busca, L. Lietti, G. Ramis, F. Berti, Appl. Catal. B 18 (1998) 1.
[4] R.M. Heck, R.J. Farrauto, Catalytic Air Pollution Control, Van Nostrand
Reinhold, New York, 1995.
[5] M. Koebel, M. Elsener, M. Kleemann, Catal. Today 59 (2000) 335.
[6] M. Koebel, M. Elsener, T. Marti, Combust. Sci. Tech. 121 (1996) 85.
[7] J.H. Baik, S.D. Yim, I.S. Nam, Y.S. Mok, J.H. Lee, B.K. Cho, S.H. Oh,
Top. Catal. 30 (2004) 37.
[8] K. Yokota, M. Fukui, T. Tanaka, Appl. Surf. Sci. 121/122 (1997) 273.
[9] B. Frank, G. Emig, A. Renken, Appl. Catal. B 19 (1998) 45.
[10] A. Ueda, N. Takayuki, A. Masashi, T. Kobayashi, Catal. Today 45 (1998)
135.
[11] R. Burch, M.D. Coleman, Appl. Catal. B 23 (1999) 115.
[12] C.N. Costa, V.N. Stathopoulos, V.C. Belessi, A.M. Efstathiou, J. Ca-
tal. 197 (2001) 350.
[13] K.C. Taylor, R.L. Klimisch, J. Catal. 30 (1973) 478.
[14] T.P. Kobylinski, B.W. Taylor, J. Catal. 33 (1974) 376.
[15] W. Dabil, S.J. Gentry, H.B. Holland, A. Jones, J. Catal. 53 (1978) 164.
[16] N. Macleod, R.M. Lambert, Appl. Catal. B 35 (2002) 269.
[17] N. Macleod, R.M. Lambert, Catal. Commun. 3 (2002) 61.
[18] Y.-W. Lee, E. Gulari, Catal. Commun. 5 (2004) 499.
[19] T. Nanba, K. Sugawara, S. Masukawa, J. Uchisawa, A. Obuchi, Ind. Eng.
Chem. Res. 44 (2005) 3426.
5. Conclusion
The catalytic reduction of NO with hydrogen in the presence
of excess oxygen was studied over Pd-based catalysts for NOx
abatement applications in diesel engine exhaust. A high steady-
state conversion of NO (>80%) is achieved at reaction temper-
atures in the range 140–250 ◦C over the 1%Pd–5%V2O5/TiO2–
Al2O3 catalyst. Adding V2O5 to the initial 1%Pd/TiO2/Al2O3
catalyst increases NO conversion, especially at temperatures
around 200 ◦C. The reaction order is nearly first order with re-
spect to NO, 0.6-order with respect to H2, and −0.18-order with
respect to CO. Increasing the O2 concentration from 2 to 10%
produced only a slight decrease in NO conversion. FTIR stud-
[20] J.H. Jones, J.T. Kummer, K. Otto, M. Shelef, E. Weaver, Environ. Sci.
Tech. 5 (1971) 791.
[21] N. Macleod, R. Cropley, R.M. Lambert, Catal. Lett. 86 (2003) 69.
[22] N. Macleod, R.M. Lambert, Catal. Lett. 90 (2003) 111.
[23] B. Wen, Fuel 81 (2002) 1841.
[24] R. Burch, M.D. Coleman, J. Catal. 208 (2002) 435.
[25] M. Slinko, T. Fink, T. Löher, H.H. Madden, S.J. Lombardo, R. Imbihl, G.
Ertl, Surf. Sci. 264 (1992) 157.
[26] R. Burch, P.J. Millington, A.P. Walker, Appl. Catal. B 4 (1994) 65.
[27] W.S. Kijlstra, D.S. Brand, E.K. Poels, A. Bliek, J. Catal. 171 (1997) 208.
[28] F.C. Meunier, J.P. Breen, V. Zuaniuk, M. Olsson, J.R.H. Ross, J. Catal. 187
(1999) 493.
[29] S. Kameoka, Y. Ukisu, T. Miyadera, Phys. Chem. Chem. Phys. 2 (2000)
367.
[30] J. Laane, J.R. Ohlsen, Prog. Inorg. Chem. 28 (1986) 465.
[31] K. Hadjiivanov, H. Knözinger, Phys. Chem. Chem. Phys. 2 (2000) 2803.
ies showed significant NH4 formation above 200 ◦C over the
+
+
V2O5-containing catalyst, compared with n+early no NH4 ob-
served on the V2O5-free sample. The NH4 specie was more

392
G. Qi et al. / Journal of Catalysis 237 (2006) 381–392
[32] A. Martínez-Arias, J. Soria, J.C. Conesa, X.L. Seoane, A. Arcoya, R.
Cataluña, J. Chem. Soc., Faraday Trans. 91 (1995) 1679.
[33] G. Ramis, L. Yi, G. Busca, Catal. Today 28 (1996) 373.
[34] G. Ramis, G. Busca, V. Lorenzelli, P. Forzatti, App. Catal. 64 (1990) 243.
[35] G. Ramis, L. Yi, G. Busca, M. Turco, E. Totur, R.J. Willy, J. Catal. 157
(1998) 523.
[36] M.A. Centeno, I. Carrizosa, J.A. Odriozola, Appl. Catal. B 19 (1988) 67.
[37] K.I. Hadjiivanov, Catal. Rev. Sci. Eng. 42 (2002) 71.
[38] R.Q. Long, R.T. Yang, J. Catal. 196 (2000) 73.
[41] W.S. Kijlstra, D.S. Brands, H.I. Smit, E.K. Poels, A. Bliek, J. Catal. 219
(1997) 171.
[42] G. Qi, R.T. Yang, J. Phys. Chem. B 108 (2004) 15738.
[43] G. Qi, R.T. Yang, Appl. Catal. B 60 (2005) 13.
[44] G. Qi, R.T. Yang, Appl. Catal. A 287 (2005) 25.
[45] M. Uchida, A.T. Bell, J. Catal. 60 (1979) 204.
[46] G. Pirug, H.P. Bonzel, J. Catal. 50 (1977) 64.
[47] W.C. Hecker, A.T. Bell, J. Catal. 92 (1985) 247.
[48] N. Macleod, R. Cropley, J.M. Keel, R.M. Lambert, J. Catal. 221 (2004)
20.
[39] N.Y. Topsøe, H. Topsøe, J.A. Dumesic, J. Catal. 151 (1995) 226.
[40] H. Schneider, S. Tschudin, M. Schneider, A. Wokaun, A. Baiker, J. Ca-
tal. 147 (1994) 5.
[49] N. Macleod, R.M. Lambert, Appl. Catal. B 46 (2003) 483.
[50] R. Burch, A.A. Shestov, J.A. Sullivan, J. Catal. 188 (1999) 69.