A transient kinetic study of the mechanism of the NO + H2 reaction over Pt/SiO2 catalysts: 1. Isotopic transient kinetics and temperature programmed analysis

Article abstract of DOI:10.1006/jcat.1999.2566

The NO + H₂ reaction has been studied over a supported Pt catalyst using temperature programmed reaction (TPR_xn), temperature programmed desorption (TPD), and steady state and non-steady state isotopic transient kinetic analysis (SSITKA and NSSITKA). N₂, N₂O, and NH₃ formation are noted during the TPR_xn. N₂O is the most prevalent species formed at lower temperatures while N₂ is most prevalent at higher temperatures. NH₃ is formed in relatively low yields at intermediate temperatures. The SSITKA shows that N₂O is the isotopically first product and that N₂ is isotopically second. The concentration of sites producing N₂O and N₂ increases with temperature, and the increase in the effective activity of the sites is greater for those sites producing N₂ than those sites producing N₂O. The chemisorption of NO is also found to be a reversible process under actual reaction conditions. Non-steady state transient kinetic experiments show that after removal of the gas phase NO, there is a conversion of some of the residual NO on the surface into N₂ precursors following a short treatment in a H₂ whereas N₂O precursors on the surface are destroyed by this treatment. It is concluded that the formation of N₂O requires the presence of gas-phase or very weakly adsorbed NO.

Full text of DOI:10.1006/jcat.1999.2566

Journal of Catalysis 186, 353–361 (1999)
Article ID jcat.1999.2566, available online at http://www.idealibrary.com on
A Transient Kinetic Study of the Mechanism of the NO + H2 Reaction
over Pt/SiO2 Catalysts
1. Isotopic Transient Kinetics and Temperature Programmed Analysis
R. Burch, A. A. Shestov,1 and J. A. Sullivan
Catalysis Research Centre, Department of Chemistry, The University of Reading, Whiteknights, Reading, RG6 6AD, United Kingdom
Received February 4, 1999; revised May 11, 1999; accepted May 11, 1999
in the case of the NH3 selective catalytic reduction process
(2)) or already present in the exhaust stream (in the case of
a three-way catalyst NOx reduction (1)).
The NO + H2 reaction has been studied over a supported Pt cata-
lyst using temperature programmed reaction (TPRxn), temperature
programmed desorption (TPD), and steady state and non-steady
state isotopic transient kinetic analysis (SSITKA and NSSITKA).
N2, N2O, and NH3 formation are noted during the TPRxn. N2O is
the most prevalent species formed at lower temperatures while N2
is most prevalent at higher temperatures. NH3 is formed in rela-
tively low yields at intermediate temperatures. The SSITKA shows
that N2O is the isotopically first product and that N2 is isotopically
second. The concentration of sites producing N2O and N2 increases
with temperature, and the increase in the effective activity of the
sites is greater for those sites producing N2 than those sites produc-
ing N2O. The chemisorption of NO is also found to be a reversible
process under actual reaction conditions. Non-steady state tran-
sient kinetic experiments show that after removal of the gas phase
NO, there is a conversion of some of the residual NO on the surface
into N2 precursors following a short treatment in a H2 whereas N2O
precursors on the surface are destroyed by this treatment. It is con-
cluded that the formation of N2O requires the presence of gas-phase
c
The removal of NOx under net oxidising conditions is
still a problem for mobile power sources, e.g., lean-burn
gasoline and diesel engines (10). Many catalyst formula-
tions and hydrocarbon reductants have been tested in the
NO/hydrocarbon/O2 reaction (11–13), but supported Pt
catalysts remain among the most promising (14–17).
H2 is a potential reductant for NO since the internal com-
bustion engine can be tuned to produce significant amounts
of H2 from gasoline. This might be an attractive solution
to the NO emission problem under cold start conditions
and for some specific diesel applications where the exhaust
temperature is low. Supported Pt catalysts are active for the
production of N2 and N2O from the NO + H2 reaction both
in the presence and in the absence of O2 (18). However, the
production of N2O also creates environmental problems
(19).
or very weakly adsorbed NO.
1999 Academic Press
Steady state isotope transient kinetic analysis (SSITKA),
developed by Happel (20), Bennett (21), and Biloen (22),
has previously been used to clarify the mechanisms of
several reactions, e.g., Fischer–Tropsch synthesis (23), CO
hydrogenation (24), oxidative coupling of CH4 (25), and
methanol synthesis (26). As a general rule the reactions
which provide the clearest information are those in which
the surface contains large reservoirs of adsorbed interme-
diates which are slowly transformed to products.
SSITKA of the CO oxidation and CO/NO reactions over
three-way catalysts (27) and aged Pt–Rh catalysts (28) have
been published. Efstathiou et al. (29) and Janssen et al. (30)
have studied the NO/NH3/O2 reaction over V2O5/TiO2 cata-
lysts. Ozkan et al. (31) have studied the NO/CH4/O2 reac-
tion over Pd/TiO2 catalysts. We have recently completed an
analysis of the NO/C3H6/O2 reaction over Pt/SiO2 catalysts
(32, 33).
Key Words: Pt catalysts; NO reduction; reaction mechanism;
TPD; TPRxn; SSITKA.
INTRODUCTION
The removal of nitrogen oxides (NOx defined as NO +
NO2) from exhaust streams of stationary and mobile power
sources has been a topic of much discussion in the litera-
ture for the past 25 years (1–6) and the detrimental effects
of NOx on the environment and to personal health are well
known (7, 8). Therefore the removal of NO from the ex-
hausts of combustion sources remains a challenging scien-
tific and technical problem. No NO decomposition catalyst
is yet active enough at the temperatures of interest to be an
efficient deNOx agent (9). Current technologies for the re-
moval of NOx rely on reducing agents either added to (e.g.,
In the present work, the NO + H2 reaction is studied over
1 On leave from Boreskov Institute of Catalysis, Russian Academy of a supported Pt catalyst (5% Pt/SiO2) using temperature
Sciences, Pr. Akademika Lavrentieva, 5, Novosibirsk, 630090, Russia.
programmed reaction/desorption and isotopic transient
353
0021-9517/99 $30.00
c
Copyright
1999 by Academic Press
All rights of reproduction in any form reserved.

354
BURCH, SHESTOV, AND SULLIVAN
kinetic analysis (both steady state and non-steady state) out-switched NO can be different due to the back-pressure
in an attempt to understand more about the nature of the caused by the presence of the catalyst in the line containing
surface intermediates in the formation of N2 and N2O.
the unlabelled NO/Ar mixture. To overcome this problem
the pressure between these two gas lines was equalised by
creating a back-pressure in the labelled line using a fine
needle valve at the outlet. The pressure created in both
lines, and the difference in pressure between the lines, was
monitored using two “Wika” pressure transducers.
EXPERIMENTAL
One hundred and thirteen milligrams of a sample of 5%
Pt/SiO2 with a metal dispersion of 31% was placed in a
tubular quartz reactor. The catalyst was preceded by 100 mg
of quartz chips which serve to preheat the incoming gas.
The reactor was configured to minimise the dead volume
in front of the catalyst. The sample was pretreated in an H2
stream at 450 C prior to any measurements being taken.
For the temperatureprogrammed reaction experiment the
catalyst was held in the reaction mixture at 300 C for 1 h and
cooled in the mixture to room temperature. The reaction
mixture contained 1.8% NO (from a cylinder of 5% Ar in
Once the 14NO/Ar mixture was removed, the decay in
14N-containing products (NH3, N2, N2O) could be moni-
tored by a mass spectrometer (Gaslab 300) which was con-
nected to the exhaust of the reactor by a continuously evac-
uated heated capillary. The formation and decay of mixed
labelled products (14N15N and 14N15NO), as well as the for-
mation of doubly labelled species (15N2O), could also be
followed. The formation of 15N2 cannot be followed as the
mass of this species overlaps with that of 14NO. 14NH3 and
15NH3 can be discerned after the contributions of various
water fragments to their respective signals have been fac-
tored out (34). This is made more straightforward by virtue
of the fact that the H2O signal is a constant before and after
the switch. The decrease in the 14NO and the correspond-
ing increase in the 15NO following the switch can also be
followed and this gives information about the reversibility
of the reactant adsorption. All the generated profiles can
be compared with the Ar profile which itself is indicative
of the gas-phase holdup of the system. The lag time for the
Ar signal was approximately 1.5 s.
Further non-steady state transient kinetic analysis was
also carried out. This involved 15NO switching experiments
where the catalyst was not maintained at steady state. In
these reactions the catalyst was brought to steady state in
the 14NO/H2 mixture and the 14NO was removed from the
stream and replaced with Ar/He at the same flow rate. The
catalyst was then held in the H2/Ar mixture for a definite
time and then the Ar was replaced with 15NO. All product
and reactant signals were followed after this switch, but the
ones of interest are those of the mixed labelled 14N15N and
14N15NO (or 15N14NO) profiles. What we are attempting to
measure in these experiments is the titration of the remain-
ing 14N species from the surface by the switching-in of the
15NO.
NO) and 1.8% H2 (from a cylinder of 3% H2 in He), and it
was balanced using He to give a total flow of 103 cm3 min
.
1
Gas flow rates were controlled using Bronkhurst mass flow
controllers. The temperature was ramped at 5 C min to
1
350 C and NO, NH3, N2, N2O, and H2O were monitored as
a function of temperature.
The percentage conversions of reactant from NO into
NH3, N2, and N2O were calculated from observed mass
spectrometer partial pressures and calibration plots of ppm
versus partial pressure.
Temperature programmed desorption (TPD) experi-
ments were carried out on a prereduced Pt/SiO2 catalyst.
One hundred and thirteen milligrams of catalyst was pre-
treated in the reducing gas (10% H2 in He, 100 cm3 min
)
1
at 500 C for 1 h, cooled to room temperature, dosed in NO
(2% NO in He 100 cm3 min 1) for 1 h, and purged in He
at the same temperature, again for 1 h. The temperature
was then ramped (20 C min 1) while the catalyst was held
in an He flow (100 cm3 min 1) and the effluent monitored
by a mass spectrometer. Concentrations of NO, NO2, N2O,
NH3, and N2 were continuously analysed as a function of
temperature.
For the steady state isotopic transient kinetic analysis stud-
ies the catalyst was held in the reaction mixture at 300 C
for 1 h, cooled to room temperature, and then heated (at
5 C min 1) to the reaction temperature. The sample was
held at the reaction temperature until steady state reaction
conditions were attained (usually 20 min was sufficient al-
though longer times were needed around the temperature
of H2 “light-off”).
RESULTS
Temperature Programmed Reaction
The NO/Ar mixture in the stream was then removed and
Figure 1 shows the temperature programmed reaction
replaced with a flow of 15NO (at the same flow rate and profiles seen for the NO + H2 reaction over the 5% Pt/SiO2
pressure). This switch was carried out when the catalyst was catalyst. The profile for NH3 is deduced from the m/e = 17
operating at steady state. Therefore, information about the profile after taking into account the contribution for H2O at
surface of the catalyst under steady state conditions can be m/e = 18. It can be seen that the NO conversion profile goes
determined from the measured profiles of the reactant and through a maximum at ca. 100 C and then rises to almost
product molecules. The pressures of the in-switched and 100% at 210 C. Three products are noted. N2O is first seen

THE NO + H2 REACTION OVER Pt/SiO2 CATALYSTS, 1
355
(14N2 and 14N2O) are found to decrease from normalised
values of 1 to zero after the switch. Simultaneously, the
mixed labelled species (14N15N, 14N15NO, and 15N14NO) are
seen to rise to a maximum, and then return to the back-
ground level, while the doubly labelled species (15N2 and
15N2O) rise from values of zero to their normalised values
of 1.
Figure 3 shows the analysis of these experimental results
in termsof , where representsthe fraction ofheavyatoms
in the gasphase in a particular molecule. For example, (N2)
is defined as
[15N2] + 0.5[15N 14N]
(N2) =
.
[1]
[15N2] + [14N 15N] + [14N2]
It can be seen that the profiles for N2O and for NO
coincide, while the profile for N2 rises much more slowly.
The fact that the profiles for (NO) and (N2O) are the
same indicates that N2O molecules which contain a 14-
labelled N atom can only continue to be formed from 14NO
while 14NO remains in contact with the catalyst. This shows
that an Eley–Rideal-type mechanism might operate here.
The presence of a relatively large reservoir of N2O forming
intermediates is indicated by the difference between the
(N2O) profile and that of the Ar. The difference between
FIG. 1. Temperature programmed reaction between NO and H2 over
5% Pt/SiO2. Reaction conditions: 1.8% NO, 1.8% H2, total flow 103 cm3
min 1, catalyst mass 113 mg, ramp rate 5 C min 1. Conversion to N2 (᭿),
N2O (᭹), and NH3 (᭜), and conversion of NO (᭺).
at 50 C. N2 is observed at 70 C, and NH3 at 90 C. The pro-
duction of N2O goes through a maximum at a temperature
of 90 C and then decreases smoothly. N2 production goes
through a maximum at a slightly higher temperature. The
amount of N2 then decreases to a minimum at 125 C, be-
fore increasing to reach a plateau at 210 C. Conversion to
NH3 is first observed when the conversion to N2O reaches
its maximum, and the amount of NH3 remains steady until a
temperature of 180 C, at which point it quickly decreases
to zero.
Steady State Isotopic Transient Kinetic Analysis
Three temperatures (50, 60, and 73 C) were chosen for
SSITKA under differential conditions. Conversion to N2O
was far more prevalent than conversion to N2, and there
was no discernible conversion to NH3. For comparison, ex-
periments were also carried out at 125 and 185 C, where
the conversion of NO is very high and therefore the reactor
was operating under nondifferential conditions but this did
allow us to investigate the formation of NH3. Values for the
rates of production of N2, N2O, and NH3, as well as the flow
rate of unreacted NO, are presented in Table 1.
FIG. 2. Normalised product responses following the replacement of
14NO/Ar with 15NO in the reaction stream. Reaction conditions: 1.8% NO
and H2, total flow 103 cm3 min 1, catalyst mass 113 mg, T = 60 C. 14N2 (᭡),
14N15N (᭹), 15N2 (᭜), 14N2O (4), 14N15NO (or 15N14NO) (᭺), 15N2O (♦),
Ar and inverse Ar (ᮀ).
Figure 2 shows a typical product profile obtained after the
switch from 14NO to 15NO at 60 C. The unlabelled products

356
BURCH, SHESTOV, AND SULLIVAN
actor. N2 and N2O product molecules do not interact in
this manner with this oxidised portion of the catalyst and
thus the (NO) profile is delayed longer than the (N2) and
(N2O) profiles.
The (NH3) profile isfurther delayed again relative to the
(NO), which is consistent with NH3 being the isotopically
third product, although, as with NO, a slow release of NH3
would give the same profile.
These normalised plots can be analysed to provide val-
ues for (the mean surface residence time for intermediates
leading to products). This is calculated from the difference
in area under plots of and the inert tracer (Ar). For ex-
ample, in the case of N2,
Z
∞
=
˜(N2) dt,
[2]
(N2)
0
where ˜ represents the (1
) profile in the case of the
14NO to 15NO switch and the profile in the case ofthe 15NO
to 14NO switch. The “mean surface lifetime” of the spe-
cies leading to product as well as the rates of formation of
the various products are also shown, as a function of tem-
perature, in Table 1. These valuescan be used in conjunction
with each other to yield values for the average concentra-
tion of adsorbed surface species leading to product and of
adsorbed reactant which simply desorbs. For example, in
FIG. 3. Profile of the functions following the replacement of 14NO/
Ar with 15NO in the reaction stream. Reaction conditions: 1.8% NO and
H2, total flow 103 cm3 min 1, catalyst mass 113 mg, T = 60 C. N2 (᭿),
N2O (᭹), NO (᭺), and inverted Ar (ᮀ) profile.
the (NO) profile and the Ar profile indicates a steady state
desorption of NO from the catalyst under working condi-
tions.
Figure 4 shows a similar plot in which the experiment
was carried out at 125 C, where NH3 is formed. It must be
remembered that under these conditions the reactor is not
behaving differentially as the conversion of NO is 86% .
In this profile it is seen that the N2O is still “isotopically
first” and N2 is “isotopically second.” The (NO) profile
lags behind the N2O and N2 profiles, a situation that is un-
usual in SSITKA but which can be explained in terms of
a nondifferential reactor. Such a large delay in the (NO)
profile, i.e., a situation where (NO)
(N2), (N2O), has
been seen previously in SSITKA of the NO/C3H6/O2 re-
action over similar Pt/SiO2 catalysts when the catalyst was
operating under nondifferential conditions (conversion of
C3H6 = 100% ) (33). In that case the delay was ascribed to
the change between a reduced surface (presumably at the
beginning of the catalyst bed) where N2 and N2O formation
takes place and an oxidised Pt surface (presumably lower
down the catalyst bed) where no N2 or N2O formation takes
place.
FIG. 4. Profile of the functions following the replacement of 14NO/
Ar with 15NO in the reaction stream. Reaction conditions: 1.8% NO and
H2, total flow 103 cm3 min 1, catalyst mass 113 mg, T = 125 C. N2 (᭿),
N2O (᭿), NO (᭺), NH3 (᭜), and inverted Ar (ᮀ).
On the oxidised Pt surface the NO can adsorb more
strongly (as NO2 or ONO-type species), and thus it takes
longer for the 14NO molecules to be removed from the re-

THE NO + H2 REACTION OVER Pt/SiO2 CATALYSTS, 1
357
TABLE 1
a rough measure. The k = 1/ relationship only truly holds
for a first-order unidirectional step. However, for compar-
ing the activity of sites producing N2 (or N2O) at different
temperatures, it is a useful guide. It is not useful for compar-
ing the “activity” of sites producing N2 with those producing
N2O.
Rates of Formation of N2, N2O, and NH3 and the Flow Rate of
Unreacted NO ( mol g 1 s 1) over a 5% Pt/SiO2 Catalyst at Various
Temperatures in an NO + H2 Stream
T( C) r(N2) (N2) r(N2O) (N2O) r(NH3) (NH3) r(NO) (NO)
Thus, the reason for the increased rate ofproduction ofN2
is the increased concentration of sites which form N2 (going
from 3 to
in the temperature the number of sites again increases (ris-
ing to 13.6), but now the reactivity is also significantly higher
(with the value falling from 34 to 22 s).
50 0.04
60 0.09
73 0.3
125 1.6
185 1.9
37
34
22
1.6 2.9
1.7 1.4
0.25
0.59
1.6
1.7
1.3
1.2
0.7
0.8
—
—
—
0.62
0.83
—
—
—
35
19
10.6
9.8
7.0
1.5
3.8
1.7
1.3
1.4
3.3
8.1
6
mol g 1 (see Table 2)). On further increase
Note. Values for (the mean surface residence time (s)) for the products
At even higher temperatures the concentration of sites
for N2 production decreases again, but the “activity” is
markedly increased and thus the rates of N2 production
are greater. Note, however, that under these conditions the
reactor is not differential and thus the concentrations of sur-
face intermediates (N ( mol g 1)) and metal surface cov-
erages (2), calculated later, simply reflect average values
over the whole of the catalyst bed.
For N2O production Table 1 shows that, within experi-
mental error (0.5 s), there is no change in the values, as
the temperature is raised between 50 and 73 C. Combina-
tion of Tables 1 and 2 then shows that the increase in the
rate of production between 50 and 73 C is due to an in-
crease in the concentration of the sites which are active for
this reaction.
NH3 production is only observed at the higher tempera-
tures. However, between the two temperatures studied the
concentration of sites decreases by 30% while the “re-
activity” of the sites doubles ( goes from 35.4 to 18.8).
This results in the rate of formation of NH3 increasing from
0.62 mol g 1 s 1 to 0.83 mol g 1 s 1. It must be remem-
bered that these values reflect an average over the whole
of the catalyst bed.
With respect to NO we can see that the concentration of
sites from which NO desorbs without reacting under steady
state conditions decreases as the temperature is raised. At
the highest temperatures, both the mean surface residence
and reactant as determined from SSITKA are also given. The reactant
1
stream contained 1.8% NO, 1.8% H2, balanced with He to 103 cm3 min
over 113 mg catalyst.
the case of N2 or N2O this is calculated from the relation-
ship (35)
2
r(N
= N(N
,
or N2O)
2
[3]
(total)
or N2O)
2
where is the mean surface residence time (s), r is the rate
of production of the product molecules ( mol g 1 s 1), and
N is the concentration of surface intermediates leading to
the corresponding products ( mol atoms g 1). The number
2 represents the number of atoms of labelled element per
molecule (2 in the case of N2 and N2O, 1 in the case of NO
and NH3).
15
In our case, for the switch 14NO + H2 → NO + H2,
where the products of the reaction are N2, N2O, and NH3,
we can calculate the total concentration of N-labelled sur-
face intermediates from the equation
Z
Z
∞
∞
out
NO
q
C
(1
NO ) dt + 2 q Cout
(1
) dt
N2
N2
0
0
Z
Z
∞
∞
out
N2O
+ 2 q C
(1
N O ) dt + q Cout
(1
) dt
NH3
2
NH3
0
0
= mcat NTot
,
[4]
where C refers to the outlet concentration of labelled re-
actants or products (mol dm 3), q represents the total flow
rate (dm3 s 1), m represents the catalyst mass, and N rep-
resents the average total concentration of all surface inter-
mediates. A more general expression for the calculation of
the concentration of surface intermediates from the pro-
files for an arbitrary reaction has been presented previously
(33).
TABLE 2
Average Concentrations of Co-adsorbed Species, N ( mol atoms
g
1), Leading to the Corresponding Products, on the Surface of 5%
Pt/SiO2 as a Function of Temperature during the NO + H2 Reaction
as Determined by Eqs. [2] and [3]
P
T( C)
N(N2)
N(N2O)
N(NH3)
N(NO)
N
Table 2 shows the average “individual” and total concen-
tration of surface sites as determined following such analy-
sis. At the lower temperatures (50 or 60 C) the qualitative
“reactivity” of the sites producing N2, as estimated from
the reciprocal of the value (see Table 1), does not in-
crease dramatically as the temperature is raised. It must be
50
60
73
3.0
6.1
13.1
5.1
0.9
1.5
3.8
4.0
2.2
0
0
0
21.9
15.6
18.0
12.7
9.8
5.0
30.8
21.9
20.3
26.7
36.0
55.1
125
185
6.5
Note. Reactant stream contained 1.8% NO, 1.8% H2, balanced with He
remembered that the reactivity—as so estimated—is only to 103 cm3 min 1 over 113 mg catalyst.

358
BURCH, SHESTOV, AND SULLIVAN
TABLE 3
remainder of the surface may be covered with Oads, Hads
,
OH, or N-containing species which simply act as spectators.
Coverages (2) of Surface Intermediates on 5% Pt/SiO2 as a
Function of Temperature during the NO + H2 Reaction as Mea-
sured by SSITKA
Non-steady State Transient Kinetic Analysis
To examine whether gaseous NO (or the weakly ad-
sorbed NO with which it is in rapid equilibrium) has a
direct influence on the production of N2O the following
non-steady state isotopic transient kinetic (NSSITKA) ex-
periments were performed. First, the reaction was brought
to steady state in NO/Ar + H2. The 14NO/Ar mixture was
then replaced by an equivalent flow of Ar/He and the cata-
lyst was left in this H2/Ar/He stream. Finally the 15NO was
switched into the reactant stream after various periods of
time.
Figure 5 and Table 4 show the profiles and the total
amounts for the mixed N2 (Fig. 5a) and mixed N2O (Fig. 5b),
species following five such experiments. The first profile is
obtained from a “true” SSITKA switch, i.e., with no delay
between the removal of the 14NO and the introduction of
the 15NO. Subsequent profiles correspond to delays of 30,
60, 300, and 600 s between the removal of the 14NO and
the switching-in of the 15NO. These experiments represent
a titration of the “active” 14N species remaining on the Pt
surface.
100 2(N2) 100 2(N2O) 100 2(NH3) 100 2(NO) 100 2(Tot)
50
60
73
125
185
3.8
7.8
16.7
6.5
1.1
1.9
4.8
5.1
2.8
0
0
0
27.8
19.8
22.9
16.1
12.5
6.4
27.8
25.8
33.9
45.7
70.0
8.3
39.1
Note. Reactant stream contained 1.8% NO, 1.8% H2, balanced with He.
times and the concentration of sites from which unreacted
NO desorbs increase dramatically. This has been mentioned
above and ascribed to a change in the surface state of the
Pt, i.e., between a metal and an oxygen-covered metal. NO
adsorbs on the metal simply as NOads (probably through
the N atom) whereas on the O-covered metal surface it can
adsorb as an ONO or an NO2ads species which can be bi-
or tri-dentate to the surface. In this way the interaction is
no longer an NO–Pt interaction but rather a stronger NO–
O–Pt interaction. We can correlate this behaviour to an
increase in the Oads level on the catalyst.
The first point to note is that when there is an interme-
diate treatment in H2, no 14NO, 14N2, or 14N2O is released
when the sample is then contacted with 15NO. Also, the total
Table 3 shows the surface coverage of Pt with species
leading to N2, N2O, and NH3, as well as the surface coverage
of adsorbed unreacted NO as a function of temperature. At
the three lower temperatures the Pt surface is never more
than 35% covered with N-containing species which go on
to form either N2 or N2O or desorb as unreacted NO. The
amount of 14N species on the surface is greatly decreased
1
by any treatment in H2 (from 20 mol g to
4
mol
g
1 (compare Table 2 at 60 C and Table 4)). It can be seen
FIG. 5. Mixed labelled N2 (a) and N2O (b) responses following the final switch in the reaction sequence NO/Ar + H2 → Ar + H2 (for a variable
time) → 15NO + H2 over 5% Pt/SiO2 at 60 C. Times in the H2/Ar stream 0 (♦), 30 (᭞), 60 (4), 300 (᭺), and 600 s (ᮀ). Part (c) shows an example of the
transient increased production of N2 and N2O when 15NO is switched into the gas stream in place of Ar following 60 s in the H2 stream: inverted Ar
(᭹), normalised N2 (ᮀ), normalised N2O (᭺). Also shown are profiles describing the isotopic composition of N2, fraction of 14N15N (4), and fraction
of 15N2 (᭡).

THE NO + H2 REACTION OVER Pt/SiO2 CATALYSTS, 1
359
TABLE 4
Total Production of Mixed Labelled N2 and N2O ( mol g 1) Fol-
lowing the Gas Delivery Sequence 14NO + H2 → H2 (t) → 15NO +
H2 where the Intermediate Time (t) in the Absence of Any Gaseous
NO Is Varied as Indicated (T = 60 C)
Amount of N N
Amount of N NO
1
1
Time (t) in H2 flow
(
mol g
)
(
mol g )
None
30 s
60 s
300 s
600 s
0.45
3.8
3.9
3.1
3.7
0.39
0.17
0.19
0.17
0.12
that the concentration of the mixed-labelled N2O species
decreases drastically once there is a time interval between
the removal of the 14NO and the introduction of the 15NO.
On the other hand, it is seen that the concentration of the
mixed-labelled N2 actually increases by about a factor of 8.
Table 4 shows that the increases in the amounts of
14N15N formed following the introduction of 15NO are much
greater than the decreases in the amounts of 14N15NO (or
15N14NO formed). Thus, there must be some other source
of 14N on the catalyst surface that becomes available for N2
formation after the removal of gaseous NO. This indicates
(a) that there are N2-forming intermediates on the catalyst
surface which are very stable at these temperatures and
(b) that the presence of gaseous NO during steady state
reaction causes these to form N2O or to simply desorb.
It should also be noted that the time spent in the H2 purge
results in a more active N2/N2O forming catalyst. This is
seen by the fact that there is a transient rise in the total
N2/N2O profiles once the 15NO is switched into the stream
(Fig. 5c). There is a fast and also a relatively slow relaxation
process for the N2 transient, while the relaxation of the N2O
transient is a single fast process. Figure 5c also shows the
fractional isotopic composition of the N2 species (14N15N
(4) and 15N2 (᭡)), and this reveals that almost 100% of the
N2 formed directly following the switch was 15N2. Shortly af-
ter the switch the fraction containing a 14-N atom becomes
significant. This decreases slowly with time following the
switch as the reservoir of 14N species on the surface is de-
pleted.
FIG. 6. Temperature programmed desorption ofNO from 5% Pt/SiO2
following a pretreatment in H2 at 500 C, an NO dose at room temperature,
1
and an He purge at room temperature for 1 h. Ramp rate 20 C min
,
catalyst mass 113 mg, total He flow 100 cm3 min 1: N2 (᭿), N2O (ᮀ), and
NO (᭺).
and the temperatures of the peak maxima are listed in
Table 5. N2 and N2O formation, as well as NO desorption,
were seen during the TPD. There was no desorption of NO2
or O2 noted. There are two desorption ranges for N2 and
N2O formation. One is centred at around 150 C, and the
second is seen as sharp peaks at 180 C.
The lower temperature formation of N2 and N2O can
be related to the increased availability of sites as the NO
desorption progresses, i.e., as NO desorbs then some NO
TABLE 5
Absolute Amounts Desorbed and Temperatures
of Desorption maxima following a TPD of NO from
a 5% Pt/SiO2 which Had Been Prereduced (Fig. 6)
Thistransient increase associated with an increased expo-
sure to H2 can be explained by the removal of Oads creating
dissociation sites and possibly also as a result of increas-
ing the Hads coverage. This “adsorbate assisted reaction”
of NO, assisted by the presence of Hads, has been reported
previously on Rh/SiO2 and Pt/SiO2 (37, 38).
1
Tmax ( C)
mol g
N2
N2
NO
NO
N2O
N2O
144
278
104
225
135
278
4.0
15.9
32.3
36.9
5.2
Temperature Programmed Desorption
5.1
Note. Ramp rate, 20 C min 1; catalyst mass, 113 mg;
The TPD-NO profiles from the prereduced catalyst are
given in Fig. 6. The absolute amounts of desorbed species
1
total flow, 100 cm3 min
.

360
BURCH, SHESTOV, AND SULLIVAN
can adsorb in a bi-dentate manner (through both the N and are exactly the same, indicating that the incorporation of a
the O). These species might then interact with molecular 14N atom into an N2O molecule does not take place when
NO to form N2 and N2O. As this concentration of molecular the gas phase 14NO is fully removed from the system. This is
NO decreases (due to desorption), the production of N2 and not the case for the formation of N2 since 14N15N is formed
N2O through this route reaches a maximum and also begins for a considerable time after all traces of gaseous 14NO are
to decrease.
removed. This may indicate the requirement for gas-phase
The higher temperature formation of N2 and N2O might NO (or adsorbed undissociated NO) in the formation of
then be caused by a second mechanism. At these temper- N2O. Further mathematical analysis of the properties of the
atures NO can dissociate on the surface, without the in- experimental profiles will clarify the nature of the reaction
teraction of a second molecule of NO, forming Nads and mechanism (36).
Oads species. The combination of the Nads species with each
The non-steady state isotopic transient kinetic analysis
other, or with another molecular NO species, could gener- (“stopped-flow”) NO switching experiments allow us to ex-
ate the N2 and N2O products, respectively. This leads to the amine the nature ofthe intermediateson the surface. Lessof
sharp production of N2 and N2O centred at 280 C.
the mixed-labelled N2O is formed from the switch of 15NO
This also explans the shape of the NO desorption pro- over the catalyst after an interval in H2 than is formed from
file, i.e., the fact that the profile has two maxima (104 and the true SSITKA experiment. Therefore, we conclude that
225 C). The first maximum coincides with the beginning the treatment in H2 results either in the desorption of N2O
of production of the low-temperature N2/N2O peaks. The precursors or in their conversion into N2 precursors. We
reason for the decreased NO desorption at these temper- can also say that some spectator species on the surface are
atures is that the NO can interact with the proposed “bi- converted into N2 precursors by the H2 treatments because
dentate” NO on the surface, and generate N2 and N2O. In the decrease in N NO formation does not fully account for
the case of the second maximum this is due to the fact that the increase in N N formation.
the NO molecules (at these higher temperatures) now have
A preliminary analysis of the SSITKA experiments
sufficient energy to dissociate on the surface and form Nads shows that in the formation of N2 and N2O at least one
intermediates which can go on to form N2 and N2O.
of the surface intermediates is not used in the formation
of both products. Also, comparing the SSITKA with the
“stopped-flow-switching” experiments we conclude that in-
termediates leading to N2O are short lived and not stable
on the surface while those leading to N2 are more stable.
DISCUSSION
The temperature programmed reaction results (Fig. 1)
show that the production of N2/N2O occurs in two stages.
At low temperature the dissociation of NO is unfavourable;
otherwise, the rate of N2 production would be higher. In
fact, N2O is the most abundant product under these con-
ditions. As the temperature is increased the conversion to
N2O reaches a maximum and then begins to decrease. This
CONCLUSIONS
For the production ofN2O on our Pt catalyst we conclude:
after a peak production at 100 C, selectivity to N2O
can be understood in terms of a competition between an in- decreases as the temperature of reaction is raised (as deter-
creasing probability for the dissociation of adsorbed NOads
,
mined from the TPRxn experiment);
and the decreasing concentration of adsorbed NO. The pre-
the production of N2O seems to require gas-phase, or
dominant reactions at the higher temperatures are the for- weakly adsorbed, NO;
mation of N2 or NH3.
the rate of N2O production increases with temperature
The SSITKA results show for all the product molecules primarily because the concentration of sites active for this
(N2, N2O, and NH3) that the trend is for the residence reaction on the catalyst increases; i.e., an increase in the ef-
time to decrease as the temperature increases. On the other fective activity of the sites is not as important as an increase
hand, the concentration of sites that produce N2 and N2O in the concentrations of sites;
increases by a factor of 4 between 50 and 73 C (see
N2O is produced isotopically first, so therefore there
Table 2). Thus, the catalyst produces more N2O mainly due are surface intermediates which do not go on to form N2O,
to an increase in the concentration of sites which are active i.e., the 14N species that produce N2;
for this reaction, while the increase in N2 production arises
some intermediate in the chain for the production of
both because of an increase in the concentration of sites N2O is not stable under an H2 atmosphere at 60 C (as de-
active for N2 formation and an increase in their effective termined from the NSSITKA experiments).
“activity.”
For N2 formation we can conclude:
Figure 3 shows changes in the fraction of heavy atoms
in the product molecules following the switch of 14NO for
N2 is produced with more selectivity at higher temper-
15NO. Note that the (N2O) profile and the 15NO profile atures (from TPRxn);

THE NO + H2 REACTION OVER Pt/SiO2 CATALYSTS, 1
361
the rate of production of N2 increases with tempera- 14. Burch, R., and Watling, T. C., J. Catal. 169, 45 (1997).
15. Captain, D. K., Robberts, K. L., and Amaridis, M. D., Catal. Today 42,
93 (1998).
16. Sasaki, M., Hamada, H., Kintaichi, Y., Ito, Y., and Tabata, M., Catal.
Lett. 15, 297 (1992).
17. Burch, R., and Ottery, D., Appl. Catal. B Environ. 9, L19 (1996).
18. Burch, R., and Coleman, M., Appl. Catal. B Environ., in press.
19. Armor, J. N., Appl. Catal. B Environ. 1, 221 (1992).
20. Happel, J., Chem. Eng. Sci. 33, 1567 (1978).
21. Bennett, C. O., in “Catalysis Under Transient Conditions” (A. T. Bell,
and L. L. Hegedus Eds.), ACS Symposium Series, Vol. 178, p. 1. Am.
Chem. Soc. Washington, DC, 1982.
ture as the concentration and effective activity of the sites
leading to its production both increase;
N2 is produced from more longer lived intermediates
in the reaction sequence than N2O; i.e., it is isotopically
second (from SSITKA);
N2 intermediates are more stable on the surface than
N2O intermediates, and the latter (or NOads spectator
species) can be transformed into the former by a reductive
treatment (from the NSSITKA experiments).
22. Biloen, P., J. Mol. Catal. 21, 17 (1982).
23. Hanssen, K. F., Blekkan, E. A., Schanke, D., and Holmen, A., Stud.
Surf. Sci. Catal. 109, 193 (1997).
ACKNOWLEDGMENTS
24. Bajusz, I. G., Kwik, D. J., and Goodwin, J. G., Catal. Letts. 48(3-4), 151
(1997).
25. Efstathiou, A. M., and Verykios, X. E., Appl. Catal. A General 151(1),
109 (1997).
We are grateful to the EPSRC for supporting this research through
contract GR/K70403. A.A.S. thanks NATO and The Royal Society for
providing a fellowship (NATO/96A).
26. Ali, S. H., and Goodwin, J. G., J. Catal. 170(2), 265 (1997).
27. Oukaci, R., Blackmond, D. G., Goodwin, J. G., Jr., and Gallagher,
G. R., in “Catalytic Control of Air Pollution,” Ch. 5, pp. 61–72. Am.
Chem. Soc., Washington, DC, 1992.
28. Frost, J. C., Lafyatis, D. S., Rajaram, R. R., and Walker, A. P., “4th In-
ternational Congress on Catalysis and Automotive Pollution Control,
Brussels, April 1997,” Vol. 1, p. 129, O14, 1997.
REFERENCES
1. Taylor, K. C., Catal. Rev. Sci. Eng. 35(4), 457 (1993). [and refs. therein]
2. Bosch, H., and Janssen, F., Catal. Today 2, 4 (1988). [and refs. therein]
3. Ansell, G. P., Golunski, S. E., Hayes, J. W., Burch, R., and Millington,
P. J. , Stud. Surf. Sci. Catal. 96, 577 (1995).
4. Iwamoto, M., and Mizuno, N., J. Automobile Eng. 207, 23 (1993).
5. Burch, R., and Watling, T. C., Catal. Lett. 43(1-2), 19 (1997).
6. Burch, R., and Watling, T. C., J. Catal. 169(45) (1997).
7. Shelef, M., Otto, K., and Gethi, H., Atmos. Environ. 3, 107 (1969).
8. “Ozone in the United Kingdom 1993,” third report of the United
Kingdom Photochemical Oxidants Review Group, Dept. of the Envi-
ronment, 1993.
9. d’Itri, J. L., and Schatler, W. M. H., Appl. Catal. B Environ. 15, 298
(1992).
10. Hums, E., in “214th ACS National Meeting, Las Vegas Nevada, Sept.
1997,” p. 1, 1997.
29. Efstathiou, A. M., and Fliatoura, K., Appl. Catal B Environ. 6(1), 35
(1995).
30. Janssen, F. J. J. G., Van Den Kerkhof, F. M. G., Bosch, H., and Ross,
J. R. H., J. Phys. Chem. 91(27), 6633 (1987).
31. Kumthekar, M. W., and Ozkan, U. S., J. Catal. 171(1), 54 (1997).
32. Burch, R., and Sullivan, J. A., J. Catal. 182(2), 489 (1999).
33. Burch, R., Shestov, A. A., and Sullivan, J. A., J. Catal. 182(2), 497
(1999).
34. “Eight Peak Index of Mass Spectra,” 3rd ed. Unwin Brothers, Surrey,
1983.
35. Shannon, S. L., and Goodwin, J. G., Chem. Rev. 95, 677 (1995).
36. Burch, R., Shestov, A. A., and Sullivan, J. Catal. 186, 363 (1999).
37. Hecker, W. C., and Bell, A. T., J. Catal. 92(2), 247 (1985).
38. Burch, R., and Watling, T. C., Catal. Letts. 37, 51 (1996).
11. Burch, R., and Scire, S., Appl. Catal. B Environ. 3, 295 (1994).
12. Catal. Today, 26 (1995). [R. Burch, Ed.]
13. Catal. Today, 22 (1994). [M. Iwamoto, Ed.]