In Situ Controlled Promotion of Catalyst Surfaces via NEMCA: The Effect of Na on the Pt-Catalyzed NO Reduction by H2

Article abstract of DOI:10.1006/jcat.1997.1551

It was found that the catalytic activity and selectivity of Pt for NO reduction by H₂ can be markedly and reversibly affected by depositing polycrystalline Pt films on β″-Al₂O₃, a Na⁺ conductor, and applying external currents or potentials to supply or remove Na to or from the Pt catalyst surface. The increase in the rate of NO reduction to N₂ is typically 10³-10⁵ times larger than the rate of supply of Na and up to 30 times larger than the umpromoted rate. The use of the β″-Al₂O₃ support permits precise in situ control of the Na coverage on the Pt surface. Sodium coverages of 0.06 cause up to 1450% increases in the total rate of NO reduction and enhance the selectivity to N₂ from 30 to 75%. The promotional effect of Na is due to enhanced NO chemisorption and dissociation on the Pt surface.

Full text of DOI:10.1006/jcat.1997.1551

JOURNAL OF CATALYSIS 166, 218–228 (1997)
ARTICLE NO. CA971551
In Situ Controlled Promotion of Catalyst Surfaces via NEMCA: The Effect
of Na on the Pt-Catalyzed NO Reduction by H2
O. A. Marina,�,1
�
�
,2
,3
I. V. Yentekakis, C. G. Vayenas, A. Palermo,† and R. M. Lambert†
�
Department of Chemical Engineering, University of Patras, GR-26500 Patras, Greece; and †Department of Chemistry, Cambridge University,
Cambridge CB2 1EW, England
Received April 17, 1996; revised November 11, 1996; accepted December 2, 1996
great technological and environmental interest. It was re-
00
It was found that the catalytic activity and selectivity of Pt for cently found that the use of � -Al2O3 to reversibly dose
NO reduction by H2 can be markedly and reversibly affected by Na on Pt leads to very significant enhancement in the rate
0
0
+
depositing polycrystalline Pt films on � -Al2O3, a Na conductor,
and applying external currents or potentials to supply or remove
Na to or from the Pt catalyst surface. The increase in the rate of
and N2 selectivity of the Pt-catalyzed NO reduction by CO
23–25), C2H4 (11), and C3H6 (24, 25). The present paper is
(
a logical extension of our work on the electrochemical pro-
motion of environmentally important catalytic reactions.
The use of hydrogen as a reducing agent for NO has been
discussed in the literature for about 20 years; the process
is also of importance in automotive catalysts (26). One of
the key issues is the relatively poor selectivity of many cata-
3
5
NO reduction to N2 is typically 10 –10 times larger than the rate of
supply of Na and up to 30 times larger than the umpromoted rate.
The use of the � -Al2O3 support permits precise in situ control of
the Na coverage on the Pt surface. Sodium coverages of 0.06 cause
up to 1450% increases in the total rate of NO reduction and enhance
the selectivity to N2 from 30 to 75%. The promotional effect of Na
00
is due to enhanced NO chemisorption and dissociation on the Pt lystswhich tend to produce substantialamountsofN2O and,
surface.
�c 1997 Academic Press
worse, NH3. Recently, there has been renewed interest in
the NO/H2 reaction as a result of improved selectivity ex-
hibited by Pd and Pd/Mo on alumina catalysts (27, 28). Our
results demonstrate that Na promotion greatly enhances
both the activity and the selectivity for N2 production.
Platinum group metals are used widely for the catalytic
removal of NO from effluent gases by means of NO re-
duction with light hydrocarbons, H2, NH3, or CO. Because
of its ability to chemisorb NO dissociatively Rh is in gen-
eral the most active and selective catalyst for NO reduc-
tion (29). Due to its scarcity, increased Rh replacement by
Pt in automotive exhaust catalytic converters is generally
considered desirable (30) despite recent fluctuations in the
relative market supply of noble metals.
INTRODUCTION
The effect of non-Faradaic electrochemical modification
of catalytic activity (NEMCA) (1–9) or electrochemical
promotion (10, 11) or in situ controlled promotion (12, 13)
hasbeen studied for over 40catalyticreactionsofPt, Pd, Rh,
Au, Ag, Ni, and IrO2 surfaces. Work in this area has been
extensively reviewed. (14–17). In brief it has been found
that the catalytic (1–17) and chemisorptive (18, 19) prop-
erties of polycrystalline metal films interfaced with solid
2�
electrolytes, such as yttria-stabilized zirconia (YSZ), an O
0
0
+
�
conductor, � -Al2O3, a Na conductor, CaF2, a F conduc-
tor, or TiO2, a mixed electronic–ionic conductor, can be af-
fected dramatically and reversibly by electrically polarizing
the metal–solid electrolyte interface. More recently the ef-
fect has been demonstrated using protonic conductors (20,
The activity and selectivity of the NO + H2 reaction on Pt
surfaces is largely determined by the nature of the adsorbed
states (molecular or dissociative NO chemisorption) which
is strongly dependent on surface structure (31). Both sup-
ported Pt catalysts and polycrystalline Pt wires and foils ex-
hibit a low activity for NO dissociation because after high
2
1) and aqueous alkaline solutions (22).
A potentially important application of electrochemical
promotion is in enhancing the catalytic activity and selec-
tivity of metals for the reduction NO to N2, a reaction of
�
temperature treatment (T > 300 C) their surfaces consist
predominantly of Pt(111) and hex-Pt(100) crystal planes
which adsorb NO only molecularly (32, 33).
Alkali metals exhibit a strong promotional effect for NO
dissociation on transition metal surfaces (34). Neverthe-
less, the subject has been essentially overlooked in regard
1
Present address: Boreskov Institute of Catalysis, Novosibirsk 630090,
Russia.
2
To whom correspondence should be addressed.
3
to practical NO reduction catalysts. Our work serves to
Present address: INTEMA-CONICET, Universidad Nac. Mar del
Plata 7600, Mar del Plata, Argentina.
x
underline the surprisingly neglected opportunities for NOx
218
0021-9517/97 $25.00
Copyright �c 1997 by Academic Press
All rights of reproduction in any form reserved.

PROMOTION OF CATALYST SURFACE VIA NEMCA
219
reduction offered by Na promotion of Pt metals. In addi-
tion, the electrochemical technique enables us to probe the
effects of promoter addition in a unique and controllable
way that provides useful insight into the reaction mecha-
nism. Sodium, in particular, can be deposited on Pt sur-
faces either from the gas phase by dosing from an evap-
oration source or by interfacing the catalyst film with a
+
00
Na -conducting solid electrolyte, such as � -Al2O3 (3, 5,
1, 12, 23–25, 35). In the latter case by appropriate voltage
1
application (� 1 V) one can control, in situ, the coverage of
the alkaline promoter on the active metal surface. This has
been recently confirmed by a variety of techniques includ-
ing work function measurements (3), X-ray photoelectron
spectroscopy (XPS), (24, 25, 36), and scanning tunneling
microscopy (STM) (37).
It is worth noting that a similar approach employing
borosilicate glasses, instead of solid electrolytes, and very
large applied potentials (� 5–10 kV) has been explored to
control the coverage of Na on semiconductor catalysts such
as TiO2 and ZnO (38, 39).
EXPERIMENTAL
The experiments were carried out in a continuous flow
well-mixed “single-pellet” type reactor described previ-
0
0
ously (6, 12) (Fig. 1) The � -Al2O3 disc was suspended in the
quartz reactor with the three electrodes (catalyst, counter,
and reference), all exposed to the reacting gas mixture. The
3
reactor volume was 25 cm .
00
The Pt catalyst film was deposited on one side of the � -
Al2O3 disc (Ceramatec) following the method described
previously (12–14), i.e., by using a thin coating of Engel-
FIG. 1. Single pellet reactor and electrode configuration.
hard Pt A1121 paste followed by calcination in air first at film and Au reference and counter electrodes under the
� �
4
00 C for 1 h and then at 750 C for 20 min. The true surface same experimental conditions have shown that Au, which
area of the Pt catalyst was determined via the isothermal does not adsorb NO and H2 dissociatively, is catalytically
surface titration technique described in detail elsewhere inert for the NO + H2 reaction. Thus all the current- and
(
14, 16). The reactive oxygen uptake, NO, of the catalyst potential-induced changes in catalytic activity described be-
�
at 375 C obtained via isothermal oxygen titration with CO low can be attributed to the Pt catalyst only.
(
1
�
7
12, 14) was 3.8 � 10 mol O, which corresponds (12, 14) to
Reactants were L’ Air Liquide certified standards of 3%
2
50 cm . Figure 2 shows the isothermal oxygen titration re- NO in He and 2% H2 in He. They were further diluted in
sults, i.e., it depicts the amount, N, of previously adsorbed ultrapure He (99.999% ) and were used without additional
reactive oxygen (forming CO2 after exposure to CO) re- purification. Reactants and products were analyzed by on-
maining on the surface after isothermal desorption in ul- line gas chromatography (Perkin-Elmer Sigma 2) with a TC
�
trapure He for various times tHe. NO is determined by ex- detector. A Porapak Q column at 40 C was used to separate
�
trapolation to tHe = 0 (12, 14, 16). The triangle on the same N2O and a molecular sieve 5A column at 110 C was used to
figure shows the value of NO estimated from the literature separate H2, N2, and NO. Nitrogen, nitrous oxide, and wa-
value of the initial dipole moment of Na on Pt(111) in con- ter were the only measurable products. Ammonia has also
junction with Helmholz’s equation (12) as described below. been reported as a product of the NO–H2 reaction on poly-
This estimate is a factor of 2 smaller than the value obtained crystalline Pt under reducing conditions (43). No ammonia
via surface titration.
was detected in the present work due to the relatively high
The Au counter and reference electrodes were deposited NO/H2 ratios used.
0
0
on the other side of the � -Al2O3 disk (Fig. 1) using
The kinetics of the reaction were studied at atmospheric
�
Demetron M8032 Au paste followed by calcination at pressure and over the temperature range 300–430 C. The
50 C. Blank experiments performed with the Au catalyst NO partial pressure was varied between 0 and 1.6 kPa and
�
7

2
20
MARINA ET AL.
(
Na supply to the catalyst) on catalyst potential VWR, and
rates of N2 and N2O production. Figure 3a also shows a
VWR transient obtained in presence of He, when no catalytic
reaction is taking place on the Pt catalyst.
The VWR vs time transient shown in Fig. 3a is very sim-
ilar to those obtained in previous electrochemical promo-
tion studies of C2H4 (5, 35) and CO (12) oxidation and NO
00
reduction by CO (23–25) on Pt/� -Al2O3. These previous
studies have established that the galvanostatic VWR tran-
sients are very useful for two purposes:
I. Establishing the relationship between VWR and Na cov-
erage, �Na, under reaction conditions.
II. Obtaining information about the dipole moment,
PNa, of Na species on the catalyst surface under reaction
conditions.
The analysis proceeds as follows: first the Na coverage is
computed via Faraday’s Law,
FIG. 2. Effect of desorption time tHe on the amount, N, of reactive
�
= � It/F NO ,
[2]
Na
oxygen adsorbed on the unpromoted Pt surface; N is measured by the
total amount of CO2 formed upon exposure to CO; the maximum oxygen
reactive uptake, NO, is obtained by extrapolating to tHe = 0. (᭡) NO value
estimated on the basis of the initial dipole moment analysis; see text for
discussion.
where t is the time of current application I, F is Faraday’s
constant, and NO is the number of available Pt sites
�
7
(
3.8 � 10 mol Pt) independently measured via surface
titration (Fig. 2). This permits construction of the �Na ab-
the H2 partial pressure was varied between 0 and 1 kPa. The scissa in Fig. 3a. Thus the VWR transient establishes the rela-
3
total gas flow rate was high, typically 300–600 cm /min in or- tionship � (VWR) between Na coverage and VWR. (Fig. 4).
Na
der to minimize the conversion of the reactants and to elim- This relationship is in general dependent on gaseous com-
inate mass and heat transfer effects. The NO and H2 conver- position but the dependence is relatively weak (12) because
sions were also thus kept below 25 and 40% , respectively.
the catalyst work function e8 varies according to
In order to clean the Pt catalyst surface from Na which
0
0
inevitably migrates on the Pt surface from the � -Al2O3
e1VWR = 1(e8)
[3]
�
structure during sintering at 750 C, a potential difference
VWR = + 400 mV was applied between the catalyst and ref- and is determined primarily by the coverage of Na, due to
�
erence electrodes at T = 400 C until the current (I > 0) be- the large dipole moment of Na (� 5 Debye) (12) on the Pt
tween the catalyst and the counter electrode vanished. This surface. The latter, PNa, can be computed from the VWR
current corresponds to the reaction
transient via the differentiated Helmholz equation (12, 16,
4
0–42),
+
00
�
Na(Pt) � Na (� -Al2O ) + e .
[1]
edVWR
d(e8)
dt
PNa � I
εo � A
3
=
=
,
[4]
dt
Our recent XPS studies have demonstrated directly the
0
0
� 12
2
reversible pumping of Na to/from beta alumina from/to the where εo = 8.85 � 10
C /Jm and A is the Pt film surface
�
2
2
� 7
Pt catalyst (25). They have also shown that the Pt catalyst area (1.5 � 10 m ) computed from NO = 3.8 � 10 mol Pt
1
9
2
can indeed be entirely cleaned of Na by the application of and the Pt(111) atom density, d = 1.53 � 10 atoms/m .
a positive potential.
Thus from the initial slope of the VWR transient upon cur-
All results reported here were obtained in the poten- rent application (Fig. 3a) one computes via Eq. [4] an initial
o
� 29
tiostatic mode of operation, i.e., by setting VWR at constant dipole moment for Na, P , of 3.7 � 10
C � m (11.2 De-
Na
values. Galvanostatic(constant current) operation wasused bye) which is a factor of two larger than the literature value
only in the transient experiments.
(12) of 5.3 Debye for Na on clean Pt(111) and with the value
of 5 � 0.4 Debye extracted in previous electrochemical
promotion studies on polycrystalline Pt films using the same
approach (12). Alternatively by using the literature value of
RESULTS
Transient Effect of Applied Constant Current
5
.3 Debye and the initial VWR slope one can obtain an esti-
Figure 3 shows a typical galvanostatic transient, i.e., it mate of the catalyst surface area A and thus its reactive oxy-
depicts the effect of applying a constant negative current gen uptake NO, as previously noted (Fig. 2). The computed

PROMOTION OF CATALYST SURFACE VIA NEMCA
221
FIG. 3. Transient effect of applied constant negative current (Na supply to the catalyst) on catalyst potential (a) under reaction conditions (solid
line) and in a He atmosphere (dashed line) and on the rates of formation of N2 and N2O (b). Potentiostatic restoration of the initial rates; see text for
discussion.
�
7
value of 1.8 � 10 mol O is a factor of two smaller than at t = 0 the galvanostat is used to impose a constant current
�
7
+
that obtained by titration (NO = 3.8 � 10 mol O).
I = � 20 �A; Na is now pumped to the catalyst surface at a
�
10
Referring to Figs. 3a and 3b we first concentrate on the rate I/F = 2.07 � 10 mol Na/s. The corresponding Na cov-
solid VWR and rate and lines which correspond to an ini- erage, �Na, starts increasing according to Eq. [2], causing a
tial potentiostatically imposed value of 0.2 V (t <� 1 min). pronounced decrease in the catalyst potentialVWR, and thus
The potentiostat is then disconnected (I = 0, t = � 1 min) (3) catalyst work function e8, and a pronounced increase
and VWR relaxes to � 0.0 V, i.e., to the value imposed by the in the rate of N2 and N2O production, which go through
gaseous composition and corresponding surface coverages a maximum when VWR � � 0.5 V. At this point rN has in-
2
of NO and hydrogen. Similar to the steady-state results dis- creased by a factor of 20 with a significant enhancement in
cussed below this decrease in catalyst potential from 0.2 to the selectivity to N2. This transient behavior is qualitatively
0
V causes a six fold enhancement in the rate of N2 forma- similar to the steady state rate vs VWR behavior discussed
tion and a 50% increase in the rate of N2O formation. Then below.

2
22
MARINA ET AL.
by a factor of 14.5. The promotion index of Na, defined as
1
rNO /ro
NO
PINa
=
,
[7]
1
�Na
is as high as 6000 for low (<0.005) Na coverages (Fig. 5b),
the second highest (11) value reported in NEMCA studies.
Effect of Gaseous Composition
Figures 6a and 6b show the effect of the hydrogen partial
pressure, PH , at constant PNO on the rates of formation of
2
N2 and N2O, respectively, at three fixed values of catalyst
potential VWR. For VWR = 0.4 V the Pt surface is Na-free
and both rN (Fig. 6a) and rN O (Fig. 6b) are low. Both rates
2
2
�
exhibit a maximum at P � 0.1 kPa (hereafter denoting the
H2
PH value which maximizes rN and rN O ). Decreasing VWR
2
2
2
to 0.0 and � 0.4 V, which corresponds to increasing the Na
coverage to approximately 0.001 and 0.04 (Fig. 4), causes
FIG. 4. Dependence of Na coverage on catalyst potential VWR for the
conditions of Fig. 3 (PNO = 0.63 kPa, PH2 = 0.3 kPa, T = 375 C).
�
�
a pronounced increase in the rates and a shift of PH2 to
0
.3 kPa. At this hydrogen partial pressure value, �N and
2
�
N O take the values of 11 and 6 for VWR = 0 V and 18 and
2
Upon current interruption at t = 4 min, VWR relaxes to
1
1 for VWR = � 0.4 V. SN is practically unaffected by PH on
2
2
�
0.3 V, causing rN and rN O to return to their correspond-
2 2
the Na-free surface (� 50% ) and increases with P from
H2
ing values at this potential during the previous galvanos-
tatic transient. Potentiostatic imposition of the initial VWR
value ( = 0.2 V) restores both rates to their initial, un-
promoted, values showing the reversibility of the system
5
5 to 80% on the Na-promoted surface (Fig. 6c).
The dependence of rN2 and rN O on PNO is shown in
2
Figs. 7a and 7b for the same three fixed values of catalyst
potential. Decreasing catalyst potential, and thus increas-
ing Na coverage, causes again a pronounced increase in
both rN and rN O and a decrease in the partial of NO, PNO,
(
Fig. 3).
2
2
Steady-State Effect of Catalyst Potential
which maximizes rN . Note that rN O develops a rate maxi-
2
2
Figure 5 shows the steady-state effect of catalyst potential mum only for VWR = � 0.4 V and that there is practically no
VWR (Fig. 5a) and corresponding Na coverage, �Na (Fig. 5b), difference in rN O for VWR = 0.4 and 0.0 V.
2
�
on rN and rN O , and on the selectivity to N2 at 375 C and
Increasing Na coverage enhances SN over the entire PNO
2
2
2
fixed partial pressures of NO and H2, PNO, and PH , respec- range (Fig. 7c). The effect is more pronounced for interme-
2
tively. The selectivity to N2, SN , is defined as
diate PNO values for which SN exhibits a maximum as a
2
2
function of PNO (Fig. 7c).
SN = rN /(rN + rN O ).
[5]
2
2
2
2
Effect of Catalyst Potential on Activation Energies
As previously noted, VWR values above 0.2 V correspond
to a Na-free catalyst surface. Decreasing VWR and thus in-
creasing �Na causes a dramatic enhancement in rN and rN O .
Figures 8a and 8b show Arrhenius plots for rN and rN O ,
2
2
respectively, obtained at fixed PH2 and PNO and various
fixed values of catalyst potential VWR. These plots show
again the very strong promotional effect on Na and the very
high values of �N and �N O (up to 30 and 12, respectively).
2
2
The rate of N2 formation, rN , in particular increases by a
2
factor of 30 at VWR = � 0.6 V. The rate enhancement ratio
2
2
�
N defined as
2
The apparent activation energies for N2 and N2O forma-
tion, EN2 and EN2O , respectively, are 12 and 13 kcal/mol
on the Na-free Pt surface. Decreasing VWR and thus in-
creasing �Na causes pronounced changes in EN and EN O
o
�
N = rN /r ,
[6]
2
2
N2
where ro is the unpromoted rN value, is thus 30. This is
N2
2
2
2
the second largest (11) � value obtained in NEMCA stud-
0
0
with the appearance of a break in the Arrhenius plots so
ies with � -Al2O3. The N2O production rate increases by
a factor of 6 and goes through a maximum with decreas-
ing VWR, thus increasing �Na. The selectivity to N2, SN2 ,
increases from 36% on the clean Pt surface to 75% for
VWR = � 0.6 V (�Na = 0.06).
that EN and EN O have two values each for every VWR � 0.
2
2
(
N2
I)
(I)
N2O
(II)
N2
(II)
We denote these values as E , E
and E , E
for
N2O
the high (I) and low (II) temperature range, respectively.
(
N2
I)
(II)
(I)
N2O
Figure 9 shows the dependence of E , EN2 , E
, and
(II)
(II)
(II)
Figure 5 demonstrates the strong promotional action of
E
on the catalyst potential VWR. Both EN2 and E
N2O N2O
Na for the reduction of NO, the total rate of which increases go through a maximum at VWR = � 0.1 V and then remain

PROMOTION OF CATALYST SURFACE VIA NEMCA
223
FIG. 5. Effect of catalyst potential (a) and corresponding Na coverage (b) on the rates of formation of N2 and N2O and on the selectivity to N2.
practically stable at 22 and 8 kcal/mol, respectively, for at VWR = � 0.1 V and stabilizes to a constant negative value
(
I)
more negative potentials. E also goes through a maxi- (� 14 kcal/mol) for lower VWR values.
N2
mum at VWR = 0 V and then remains practically constant
Such activation energy maxima with VWR have also
goes through a pronounced minimum been obtained in the NEMCA study of CO oxidation on
(
I)
at 7 kcal/mol. E
N2O

2
24
MARINA ET AL.
FIG. 6. Effect of PH2 on the rates of formation of N2 (a) and N2O (b) and on selectivity to N2 (c) for fixed PNO and various fixed VWR values.
0
0
Pt/� -Al2O3 and have been explained quantitatively on the enhances dramatically both the rate of NO reduction and
basis of Langmuir–Hinshelwood type kinetics by taking the selectivity to N2.
into account the variation of adsorption equilibrium con-
stants with potential (12).
As in all previous electrochemical promotion studies of
NO reduction on Pt, all the observed features can be ex-
Figure 9a also shows the effect of VWR on the tempera- plained by two factors:
ture, Tb, of the break in the Arrhenius plots of rN of rN O .
2
2
1
. Sodium enhances NO chemisorption vs H2 chemi-
Decreasing VWR, i.e., increasing Na coverage, decreases Tb
for both reactions.
�
sorption. This is apparent from the observed increase in P
and decrease in P with increasing Na coverage (Figs. 6
H2
�
NO
and 7) and is consistent with the general rule used to ra-
tionalize all previous electrochemical promotion studies,
DISCUSSION
The present results show that the catalytic activity and i.e., that increasing coverage of electropositive promoters,
selectivity of Pt for the NO reduction by H2 can be altered and thus decreasing catalyst work function, strengthens
dramatically and reversibly by dosing Na on the Pt catalyst the chemisorptive bond of electron acceptor adsorbates
surface. The measured values of the promotion index of (e.g., NO) and weakens the chemisorptive bond of electron
Na, PINa, is up to 6000 for the rate of N2 formation. Sodium donor adsorbates (e.g., H).

PROMOTION OF CATALYST SURFACE VIA NEMCA
225
FIG. 7. Effect of PNO on the rates of formation of N2 (a) and N2O (b) and on the selectivity to N2 (c) for fixed PH2 and various fixed VWR values.
2
. Sodium enhances NO dissociation on the Pt surface.
2H(a) + O(a) � H2O(g).
[13]
This is a consequence of the first factor and is apparent
from the observed dramatic enhancement in selectivity to
N2 with increasing Na coverage (Fig. 5).
The role of H2 is thus to scavenge atomic oxygen resulting
from the NO dissociation step [9]. Increasing PH decreases
2
One can thus rationalize all the observed features of the the coverage, �O, of atomic oxygen, thus increasing the ratio
present study by considering the following reaction scheme: �N/�NO of the coverages of N and NO, respectively, accord-
ing to the equilibrium [9] thus enhancing the selectivity to
N2 (Fig. 6c).
NO(g) � NO(a)
NO(a) � N(a) + O(a)
NO(a) + N(a) → N2O(g)
N(a) + N(a) → N2(g)
H2(g) � 2H(a)
[8]
The enhanced chemisorption of NO vs H with decreasing
VWR is clearly manifested by observing the effect of VWR
on the kinetic behavior with respect to PH (Figs. 6a and
[9]
2
6
b) and with respect to PNO (Figs. 7a and 7b). In the for-
[10]
mer case decreasing VWR shifts the PH value which maxi-
2
�
[11] mizes the rate, P , to higher values, indicating a weakening
in the chemisorptive bond of atomic hydrogen. In the lat-
[12] ter case, decreasing VWR causes the appearance of a rate
H2

2
26
MARINA ET AL.
FIG. 8. Arrhenius plots of the rates of formation of N2 (a) and N2O (b) at various fixed VWR values.
maximum with respect to PNO, which shifts to lower values can rationalize the observed behavior (Figs. 8 and 9) as
with decreasing VWR, which shows a significant strengthen- follows:
ing in the chemisorptive bond of NO. This strengthening in
In a previous paper (12) we analyzed the variation in
the chemisorptive bond of NO with decreasing VWR leads apparent activation energy with temperature and different
eventually to extensive NO dissociation and this causes the levels of promoters affecting the adsorption equilibrium
observed pronounced enhancement in the selectivity to N2 constants for bimolecular Langmuir–Hinshelwood kinetics.
(
Figs. 5a, 6c, and 7c).
The analysis was applied to the case of CO oxidation on Pt
The observed breaks in the Arrhenius plots (Figs. 8a and and was found to be in almost quantitative agreement with
b) manifest the transition from a H-rich surface to a NO- experiment (12). One can write more generally Eq. [18] of
8
and N-rich surface. Thus for the gaseous composition of Ref. (12) for the apparent activation energy E of any simple
the Arrhenius plots and for high VWR values, both rates, bimolecular reaction,
rN and rN O , are first order in NO and negative order in
2
2
A + B → C,
[14]
H2 (Figs. 6a and 7a). Under these conditions the surface,
promoted at low T or unpromoted, is predominantly cov-
ered with atomic H. Upon decreasing VWR both rates ap-
proach their maximum values (Figs. 6a and 7a) which indi-
cates roughly equal coverages of H and NO on the surface.
This, as analyzed below, causes a break in the Arrhenius
which follows Langmuir–Hinshelwood kinetics as
KA PA QA + KB PB QB
E = E � Q � Q + 2
,
[15]
R
A
B
1
+ KA PA + KB PB
plot and a decrease in the apparent activation energy with where ER is the true activation energy for the surface re-
increasing T.
action between adsorbed A and B, QA and QB are the
It is worth noting that decreasing VWR causes a decrease heats of adsorption of A and B, and KA and KB are
in the temperature, Tb, of the break which is consistent the adsorption equilibrium constants of A and B, which
with the enhanced NO (vs H) chemisorption with decreas- are both temperature- and promoter-dependent. Thus E
ing VWR and with a higher heat of adsorption of hydrogen changes from ER + QA + QB at low temperatures (KAPA,
(
vs NO) on the Pt surface. It is also worth noting that, for KBPB � 1) to ER� QA� QB at high temperatures (KAPA,
any fixed temperature, the VWR value corresponding to the KBPB � 1) and can easily become negative. If the rate is
Arrhenius plot break is lower for rN than for rN O (Fig. 9, at the maximum (KAPA � KBPB) and reactant adsorption
2
2
top). This is because decreasing VWR increases the ratio is strong (KAPA � 1), then it follows from Eq. [15] that
�
N/�NO which is proportional to the ratio rN /rN O accord- E = ER.
2
2
ing to the above reaction scheme [8] to [13].
In the present case the situation is more complex to ana-
After establishing that the observed breaks in the lyze due to the surface NO dissociation step. By analogy to
Arrhenius plots (Figs. 8a and 8b) are related to a change the above analysis for the reaction scheme [8] to [13] one
(
N2
II)
(I)
(II)
N2O
(I)
N2O
in the reaction orders with respect to NO and H2, one may conclude that differences E � EN2 and E
� E

PROMOTION OF CATALYST SURFACE VIA NEMCA
227
catalyst surfaces is of considerable theoretical and practical
importance.
ACKNOWLEDGMENTS
We thank the CEC JOULE Programme and the EPET II and PENED
Programmes of the Hellinic Secretariat of Research and Technology for
partial financial support.
REFERENCES
1. Vayenas, C. G., Bebelis, S., and Neophytides, S. G., J. Phys. Chem. 92,
5083 (1988).
2
3
. Bebelis, S., and Vayenas, C. G., J. Catal. 118, 125 (1989).
. Vayenas, C. G., Bebelis, S., and Ladas, S., Nature (London) 343, 625
(
1990).
4
5
6
7
8
9
. Politova, T. I., Sobyanin, V. A., and Belyaev, V. D., React. Kinet. Catal.
Lett. 41, 321 (1990).
. Vayenas, C. G., Bebelis, S., and Despotopoulou, M., J. Catal. 128, 415
(
1991).
. Cavalca, C. A., Larsen, G., Vayenas, C. G., and Haller, G. L., J. Phys.
Chem. 97, 6115 (1993).
. Mar’ina, O. A., Sobyanin, V. A., Belyaev, V. D., and Parmon, V. N.,
Catal. Lett. 13, 567 (1992).
. Varkaraki, E., Nicole, J., Plattner, E., Comninellis, Ch., and Vayenas,
C. G., J. Appl. Electrochem. 25, 978 (1995).
. Pliangos, C., Yentekakis, I. V., Verykios, X. E., and Vayenas, C. G.,
J. Catal. 154, 124 (1995).
1
1
1
0. Pritchard, J., Nature (London) 342, 592 (1990).
1. Harkness, I. R., and Lambert, R. M., J. Catal. 152, 211 (1995).
2. Yentekakis, I. V., Moggridge, G., Vayenas, C. G., and Lambert, R. M.,
J. Catal. 146, 292 (1994), and references therein.
FIG. 9. Effect of catalyst potential VWR on the temperature of the
break in the Arrhenius plots, Tb (top), and on the activation energies of
rN2 (circles) and rN2O (squares) in the high T (I) and low T (II) ranges,
respectively.
1
1
3. Cavalca, C. A., and Haller, G. L., J. Catal., in press (1996).
4. Vayenas, C. G., Bebelis, S., Yentekakis, I. V., and Lintz, H.-G., Catal.
Today 11, 303 (1992).
5. Vayenas, C. G., Ladas, S., Bebelis, S., Yentekakis, I. V., Neophytides,
S., Jiang,Yi, Karavasilis, Ch., and Pliangos, C., Electrochim. Acta
1
1
39, (11/12), 1849 (1994).
(
16 and 19 kcal/mol, respectively) relate to the heat of ad-
(
6. Vayenas, C. G., Jaksic, M. M., Bebelis, S. I., and Neophytides, S. G., The
electrochemical activation of catalytic reactions, in “Modern Aspects
of Electrochemistry” (J. O’ M. Bockris, B. E. Conway, and R. E. White,
Eds.), Vol. 29, pp. 57–202. Plenum, New York, 1995.
7. Vayenas, C. G., and Yentekakis, I. V., Electrochemical Modification
of Catalytic Activity, in “Handbook of Heterogeneous Catalysis” (G.
Ertl, H. Kn o¨ tzinger, and J. Weitkamp, Eds.), in press. VCH Publishers,
1997.
8. Bebelis, S., and Vayenas, C. G., J. Catal. 138, 570 (1992).
9. Neophytides, S., and Vayenas, C. G., J. Phys. Chem. 99, 17063 (1995).
0. Tsiplakides, D., Neophytides, S., Enea, O., Jaksic, M. M., and Vayenas,
C. G., J. Electrochem. Soc., in press (1997).
I)
(I)
sorption of hydrogen, while the difference E � E
and
N2
N2O
(
II)
(II)
E
� E
(20 and 15 kcal/mol, respectively) relates to the
N2
N2O
activation energy for NO dissociation on the promoted Pt
surface.
1
The very similar values of the activation energies for both
reactions on the unpromoted surface (Figs. 8a and 8b) indi-
cate that in this case both rates are probably controlled by
the same rate limiting reaction,
1
1
2
H + NO → N + OH,
[16] 21. Makri, M., Buekenhoudt, A., Luyten, J., and Vayenas, C. G., Ionics 2
1996).
(
2
2. Neophytides, S. G., Tsiplakides, D., Stonehart, P., Jaksic, M. M., and
Vayenas, C. G., Nature (London) 370, 45 (1994).
followed by rapid reaction of OH with H to form H2O.
Under these conditions the extent of NO dissociation must
by very small and the nondissociative mechanism must be
dominant.
In conclusion the observed very pronounced promo-
tional role of Na for this reaction system can be mainly
attributed to the pronounced enhancement of NO dissoci-
ation (step [9] of the reaction scheme). The ability of solid
23. Palermo, A., Lambert, R. M., Harkness, I. R., Yentekakis, I. V.,
Mar’ina, O. A., and Vayenas, C. G., J. Catal. 161, 471 (1996).
24. Lambert, R. M., Tikhov, M., Palermo, A., Yentekakis, I. V., and
Vayenas, C. G., Ionics 1, 366 (1995).
2
5. Palermo, A., Tikhov, M. S., Filkin, N. C., Lambert, R. M., Yentekakis,
I. V., and Vayenas, C. G., Stud. Surf. Sci. Catal. 101, 513 (1996).
6. Masel, R. I., Catal. Rev. 28, 335 (1986).
2
2
7. Stenger, H. G., Jr., and Hepburn, J. S., Energ. Fuels 1, 412 (1987).
electrolytes to tune precisely the coverage of promoters on 28. Halasz, I., Brenner, A., and Shelef, M., Catal. Lett. 18, 289 (1993).

2
28
MARINA ET AL.
2
9. Yentekakis, I. V., Pliangos, C. A., Papadakis, V. G., Verykios, X. E., 36. Cavalca, C., Ph.D. thesis, Yale Univ., 1995.
and Vayenas, C. G., in “Catalysis and Automotive Pollution Control 37. Makri, M., Vayenas, C. G., Bebelis, S., Besocke, K., and Cavalca, C.,
III. Vol. 96. Studies in Surface Science and Catalysis” (A. Frennet and
J.-M. Bastin, Eds.), p. 375. Elsevier, Amsterdam, 1995.
0. Engler, B. H., Lindler, D., Lox, E. S., Sch a¨ fer-Sindlinger, A., and
Surf. Sci. 369, 351 (1996).
38. Mikheeva, E. P., Devyatov, V. G., and Kolchanova, V. M., Kinet. Katal.
24, 418 (1983).
3
Ostgathe, K., in “Catalysis and Automotive Pollution Control III. 39. Kolchanova, V. M., and Mikheeva, E. P., React. Kinet. Catal. Lett. 24,
Vol. 96. Studies in Surface Science and Catalysis” (A. Frennet and
J.-M. Bastin, Eds.), p. 441. Elsevier, Amsterdam, 1995.
1. Masel, R. I., Catal. Rev. Sci. Eng. 28, 335 (1986).
2. Gohndrone, J. M., and Masel, R. I., Surf. Sci. 209, 44 (1989).
3. Zemlyanov, D. Y., Smirnov, M. Y., Gorodetskii, V. V., and Block, J. M.,
Surf. Sci. 329, 61 (1995).
207 (1984); 26, 339 (1984).
40. Ladas, S., Bebelis, S., and Vayenas, C. G., Surf. Sci. 251/252, 1062
3
3
3
(1991).
41. H o¨ lzl, J., and Schulte, F. K., in “Solid Surface Physics” (G.
H o¨ hler and E. Niekisch, Eds.), pp. 1–150. Springer-Verlag, Berlin,
1979.
3
4. Kiskinova, M.P., in “Studies in Surface Science and Catalysis,” Vol. 70, 42. Gundry, P. M., and Tompkins, F. C., in “Experimental Method in Cata-
Sect. 6.3. Elsevier, Amsterdam, 1992.
5. Harkness, I. R., Hardacre, C., Lambert, R. M., Yentekakis, I. V., and
Vayenas, C. G., J. Catal. 160, 19 (1996).
lytic Research” (R. B. Anderson, Eds.), pp. 100–168. Academic Press,
NY, 1968.
43. Pirug, G., and Bonzel, H. P., J. Catal. 50, 64 (1977).
3