Studies of the kinetics of two parallel reactions: Ammonia decomposition and nitriding of iron catalyst

Article abstract of DOI:10.1021/jp8079759

The reaction of ammonia decomposition and nitriding reaction as an example of the parallel reactions were studied. A surface reaction was assumed as the rate limiting step. The experiments were carried out in the range of temperatures from 623 to 723 K. M

Full text of DOI:10.1021/jp8079759

J. Phys. Chem. A 2009, 113, 411–416
411
Studies of the Kinetics of Two Parallel Reactions: Ammonia Decomposition and Nitriding of
Iron Catalyst
Walerian Arabczyk and Rafat Pelka*
Szczecin UniVersity of Technology, Institute of Chemical and EnVironment Engineering, Puaskiego 10,
70-322 Szczecin, Poland
ReceiVed: September 8, 2008; ReVised Manuscript ReceiVed: October 27, 2008
The reaction of ammonia decomposition and nitriding reaction as an example of the parallel reactions were
studied. A surface reaction was assumed as the rate limiting step. The experiments were carried out in the
range of temperatures from 623 to 723 K. Mixtures of different iron nitrides (γ′-Fe4N, ꢀ-Fe3-2N) were obtained.
Differential tubular reactor with thermogravimetric (TG) measurement and analysis of the gas phase composition
in the reaction volume was used. Reacting gases flowing through the reactor were mixed. Effective reactor
volume was determined. The rate constants for ammonia decomposition and ammonia adsorption process at
critical point between R-Fe and γ′-Fe4N phases were estimated. The number of collisions and the sticking
coefficient of ammonia over R-Fe phase were also assessed.
Introduction
dX
dt
E ) kA(T)f(XB) + kC(T)f(XB)
(4)
The phase composition of the iron-nitrogen-hydrogen
system is very well-known.1-3 As a result of the reaction
between iron and ammonia, iron nitrides are formed (γ′-Fe4N,
ꢀ-Fe3N, ꢀ-Fe2N). Ammonia, hydrogen and nitrogen interactions
with iron surface were analyzed in depth from the point of view
of studying the elementary steps of catalytic ammonia synthesis
as defined by Ertl’s and Somorjai’s researchers group and as
described in reference papers.4,5
Nanocrystalline iron nitriding process can be regarded as an
example of a system of parallel reactions, where, apart from
iron nitriding reaction, catalytic ammonia decomposition also
proceeds. During a nanocrystalline iron (A) nitriding reaction
with ammonia (B), solid (C) as well as gaseous (D) products
are formed. At the same time, catalytic ammonia decomposition
takes place while nanocrystalline iron and iron nitrides are the
catalyst of this reaction. Gaseous products (D, E) are formed.
The above-mentioned parallel reactions, generally, can be
described according to the following scheme:
where X ) concentration, f(X) ) concentration function, k )
reaction rate constant, t ) time, T ) temperature, and subscripts
A, B, C, D, and E refer to substances A, B, C, D, and E,
respectively.
While studying the nitriding process of thin iron foils,
Grabke6,7 observed that the surface chemical reaction is the rate
limiting step only in the initial stage of the process. The
influence of diffusion through the product layer in case of
nanocrystalline iron nitriding can be neglected.8 It implies that
there is no nitrogen concentration gradient in the iron nanoc-
rystallites. Dissociative adsorption of ammonia is the rate
limiting step during the whole iron nitriding process.
When the equilibrium state between the gas and solid phases
is reached, it is only the effect of catalytic ammonia decomposi-
tion reaction on iron saturated with nitrogen and/or iron nitrides
which can be measured. In this regard nitrogen contents of
created nitrides depend on temperature and gaseous phase
ammonia concentrations.8,9 However, while the stationary state
is present, three reactions i.e. nitrides formation and ammonia
decomposition on iron and/or iron nitrides take place. Then,
concentrations of reactants present in the gas phase change as
a result of the conducted chemical reactions. Thus, in order to
perform kinetics analysis of a considered system, besides
specifying the solid mass changes, it is necessary to evaluate
the composition of reacting gases. Such an analysis can be
performed using a differential reactor with thermogravimetric
(TG) measurement combined with continuous gas phase com-
position measurement. By applying experimental conditions (a
lack of temperature gradient in the bed of solid, mixing of
reactants in a gas stream, a lack of concentration gradient in
the interface between phases as well as in the bulk of phases10,11
so that the process is conducted in the kinetic region) one can
directly obtain the reaction rate values.
k
A + BfC + D
A,C,kA,kC
B
f D + E
Three parallel chemical reactions between solid and gaseous
substrate result in the formation of solid and gaseous products.
Additionally, solid substrate A and solid product C are the
catalyst of a reaction taking place in the gas phase. The reaction
rate equations concerning particular constituents can be given
as follows:
dX
-
B ) k(T)f(XB)f(XA) + kA(T)f(XB) + kC(T)f(XB) (1)
dt
dX
C ) k(T)f(XB)f(XA)
(2)
dt
dX
D ) k(T)f(XB)f(XA) + kA(T)f(XB) + kC(T)f(XB) (3)
dt
Experimental Methods
The analyzed iron was obtained by fusion of magnetite with
structured promoters of iron - calcium oxide (2.8%) and
* Corresponding author. Telephone: +48 91 449 47 30. Fax +48 91
449 46 86. E-mail address: rpelka@ps.pl.
10.1021/jp8079759 CCC: $40.75
2009 American Chemical Society
Published on Web 12/16/2008

412 J. Phys. Chem. A, Vol. 113, No. 2, 2009
Arabczyk and Pelka
Figure 2. Gas phase composition changes. System without chemical
reaction.
Nitriding processes were conducted under pure ammonia
atmosphere at isothermal conditions, at various temperatures
in the range of 623-723 K at ammonia load 500 cm3 min-1
gmat-1 until the analyzed sample’s mass was stable, which could
be seen on thermogravimetric (TG) curves as horizontal lines.
The reacting gas mixture composition was determined. Gas
samples were collected directly over the catalytic bed.
Figure 1. Experimental equipment: 1, sample holder; 2, single layer
of grains; 3, reactor furnace; 4, reactor wall; 5, thermocouple.
aluminum oxide (3.3%). Then, the alloy was crushed and a
fraction of grains of the size in the range 1.0-1.2 mm was
separated. The reduction process of the alloy was performed
stepwise from the temperature of 623 K (2 h), then 673 K (4
h), 723 K (24 h), and finally 773 K (24 h) at pure hydrogen
load (99.999%) 10 dm3 g-1 h-1, under atmospheric pressure.
Then, the sample was overheated at 1073 K for 5 h. The material
obtained in such a way contained iron nanocrystallites of a mean
size 42 nm (determined by means of XRD method). The specific
surface area determined by means of BET method (a method
for surface area determination based on Brunauer-Emmet-Teller
theory) was 5.5 m2/g.
Results and Discussion
The results of the gas phase composition measurements
performed over and under the solid bed during nitriding process
prove, that a mixing of the components of the reacting gas
mixture occurs in the reactor. It is due to an intensive convection.
The modeling of the gas phase composition changes, observed
without any chemical reaction, was performed, basing on the
mass balance of a reactor with an ideal mixing
In order to perform investigations of the kinetics of parallel
nanocrystalline iron nitriding and catalytic ammonia decomposi-
tion reactions an analytical system equipped with a flow tubular
reactor was used (Figure 1). The system made it possible to
conduct continuous thermogravimetric measurements combined
with analyses of gas phase compositions. The reactor’s con-
struction enabled one to perform processes under aggressive
conditions by means of applying special construction elements
resistant to high temperature and gases such as ammonia. The
flow rates of the applied gases, i.e., hydrogen, ammonia, and
nitrogen, were determined by means of mass flow controllers.
Hydrogen concentration in the reacting gas mixture was
determined by means of an electronic system, based on a specific
heat evaluation of gas mixture containing hydrogen and other
gases. Gas samples were collected over and under a platinum
basket containing a solid sample (Figure 1). The analytical
system was made of materials resistant to aggressive chemical
reagent. The experimental condition (temperature of measure-
ment) was selected so that the value of heat conductivity was
the same for nitrogen as that for ammonia. Thus, it is possible
to specify hydrogen content in a gas phase. The mass and
temperature of the analyzed solid samples as well as data
concerning gas concentrations in the gas mixture were recorded
digitally. In a platinum basket hanging on a balance arm the
analyzed substance (ca. 1 g) was placed in a form of single
layer of grains. The platinum from a sample holder did not affect
the chemical processes taking place in the reactor. Measurements
for a system without chemical reactions were performed. The
reactor was filled in with nitrogen until all air was entirely
removed. Then, the reactor’s content was heated to 473 K,
nitrogen flow was turned off and hydrogen flow was turned on
(9 dm3/h). Gas phase composition changes vs time were
determined.
V(Xi - X)
dX
dt
)
(5)
(6)
n
and after integration
V
n
X ) Xi 1 - exp - t
[
(
)
]
where n ) total number of moles, V ) total molar flow rate,
and subscript i refers to the feed stream.
A comparison of the calculated and experimental data,
concerning hydrogen and nitrogen concentrations in the system
without chemical reaction (N2 and H2 were introduced into the
reactor one after the other), is presented in Figure 2. The
experimental and calculated concentration values enabled one
to determine the effective reactor volume which was 250 cm3.
The accordance of experimental gas composition data with the
calculated data makes it possible to assume, that in the effective
reactor volume an almost ideal mixing occurs under the
experimental conditions.
The steady values of conversion degree of ammonia during
catalytic ammonia decomposition reaction with all the applied
grain diameters at lower temperature (724 K) (Figure 3) show
that at this temperature the process was conducted in the kinetic
reaction region, and that any internal diffusion effects could be
neglected. At a higher temperature (773 K) an internal diffusion
effect was observed only at grain diameters greater than 1.6
mm. It implies, that at grain diameters (1.0-1.2 mm) selected
for further consideration the internal diffusion effects do not
influence the catalytic ammonia decomposition reaction and
nitriding reaction at all the experimental temperatures. Therefore,
the applied reactor could be regarded as a differential one.

Studies of Kinetics of Two Parallel Reactions
J. Phys. Chem. A, Vol. 113, No. 2, 2009 413
reaction runs slower at lower temperaturessthe inflection point
occurs later in timesbut ammonia concentration needed to start
the formation of a new phase, richer in nitrogen, is reached when
the nitrogen content in iron is lowersthe inflection point occurs
at lower and lower nitrogen concentrations in iron.
The hydrogen concentration in stationary states increased with
temperature (Figure 4b). At temperatures greater than 673 K, it
was possible to measure some amount of hydrogen even when
the nitriding reaction was finished. It implied that the catalytic
ammonia decomposition reaction took place on the iron surface.9
Therefore, a decrease of ammonia concentration observed in a
stationary state resulted only from a catalytic ammonia decom-
position reaction.
Figure 3. Ammonia conversion degree vs grain size.
On the basis of thermogravimetric analyses DTG curves were
obtained (Figure 5), informing about the nitriding reaction rate.
DTG curves also provide information about the amount of
hydrogen resulting only from the nitriding reaction.
In order to determine the values of ammonia decomposition
reaction rate the following algorithm was applied. Two tem-
peratures were used as an example, namely 673 and 723 K.
Initially, one calculated values of concentration occurring while
hydrogen was replaced by ammonia without chemical reaction
inside the reactor, XNH3c, XH2c (Figures 6 and 7)a and by means
of mass balance of stirred tubular reactor (eq 6) one can write:
Ammonia concentration changes observed in the reacting gas
mixture are due to a physical mixing process proceeding
between ammonia feeding the reactor and products of nitriding
reaction and ammonia catalytic decomposition reaction. Am-
monia is also consumed during the chemical processes men-
tioned above. Thus, the nitriding process’s rate, and so the rate
of nitriding reaction and the rate of simultaneously proceeding
catalytic ammonia decomposition reaction, taking into account
gas mixing process, can be described by an expression derived
from Equation (1):
dX cNH3 V(XNH3,i - X Nc H3
)
n+d
dX
V(XNH3,i - X cNH3
)
)
(8)
NH3
dt
n
)
- kn(T)f(XNH3) - kd(T)f(XNH3
)
dt
n
Then, using calculated curves of ammonia and hydrogen
concentration changes, respectively, without chemical reaction
as well as applying DTG curves, the gas phase composition
was calculated, which could be observed, only if a nitriding
reaction took place in the system, XNH3n, XH2n (Figure 6.a) and
7.a)). It can be written as follows:
(7)
where the following are denoted: d refers to a catalytic ammonia
decomposition reaction; i refers to a feed stream; n refers to a
nitriding reaction; NH3 refers to ammonia; superscripts n + d
refer to a nitriding process (nitriding + ammonia decomposition
reaction); and superscript c refers to calculated values without
a chemical reaction.
The first item of the right side of the above equation
corresponds to a mixing and gas exchange process in the reactor.
Nevertheless, the second and the third items represent the
nitriding reaction rate and ammonia catalytic decomposition
reaction rates, respectively.
A solid sample mass gain, observed as TG curves, corre-
sponds to an increase of nitrogen concentration in iron.
An iron nitriding process caused a change of gas phase
composition, which was recorded against the background of
physical mixing and gas exchange processes taking place in
the reactor. Hydrogen concentration in the reaction volume,
XH2n+d, was directly determined. The concentration changes
resulted from ammonia load feeding the reactor as well as from
nitriding reaction and catalytic ammonia decomposition reaction
taking place in the reactor.
The results of the selected experiments are presented in the
form of TG curves (Figure 4a) and hydrogen concentration
changes (Figure 4b). The mass gain reached its maximal value
and remained unchanged at every temperature for a twice longer
time than that shown in Figure 4a, despite prolonged exposure
to the nitriding atmosphere. As a result of the reaction, various
iron nitrides were formed viz. γ′-Fe4N, ꢀ-Fe3-2N. One inflection
point on the obtained TG curves, corresponded to a phase
transition between R-Fe(N) and γ′-Fe4N. At this point two
phases (R-Fe(N) and γ′-Fe4N) exist simultaneously. The inflec-
tion point always appears below the stoichiometric concentration
of nitrogen in γ′-Fe4N. Moreover, the lower temperature the
later the inflection point occurs and at a lower nitrogen
concentration in a solid sample. This is because the nitriding
dX nNH3 V(XNH3,i - X Nc H3
)
)
- kn(T)f(XNH3
)
(9)
dt
n
Comparing results of hydrogen concentration measurement
- aggregating influence of nitriding reaction and catalytic
n+d
ammonia decomposition reaction, XH2
(Figures 6b and
7b)swith the results of calculations concerning hydrogen
content in the reaction mixture for nitriding reaction, XH2n, the
amount of hydrogen was calculated, which was generated as a
result of catalytic ammonia decomposition reactionsdashed area
in Figures 6b and 7b. According to catalytic ammonia decom-
d
position reaction stoichiometry, the amount of nitrogen, XN2
(Figure 6b and 7b) is 3-fold lower than the amount of hydrogen
generated during catalytic ammonia decomposition reaction,
n+d
(XH2
- XH2n). Then, the amount of ammonia remaining in
the reactor after nitriding reaction, XNH3n, was decreased by
hydrogen and nitrogen amounts which were generated during
catalytic ammonia decomposition process, namely by the
amount (XH2n+d - XH2n) and XN2d respectively. Thus, ammonia
concentration changes occurring during nanocrystalline iron
nitriding process were determined, XNH3n+d, (Figure 6b and 7b).
The difference between concentrations XNH3n and XNH3n+d results
from catalytic ammonia decomposition reaction. On the basis
of eq 7, the following relation can be written:
V(XNH3,i - X cNH3
)
dX
NH3
n+d
kd(T)f(XNH3) )
- kn(T)f(XNH3) -
n
dt
(10)
However, the observed decrease of ammonia concentration
was influenced by a decrease of ammonia amount because

414 J. Phys. Chem. A, Vol. 113, No. 2, 2009
Arabczyk and Pelka
Figure 4. (a) TG curves for nitriding process of nanocrystaline iron. (b) Gas phase composition changessnitriding process.
of nitrogen became higher and higher till the critical bulk
concentration was achieved. Nitrogen solution in iron was
obtained. Nitriding reaction (B region) started after a minimal
concentration of ammonia needed for nitriding reaction to be
commenced, equal to ca. 0.3 mol NH3/mol, was reached. After
the phase transition of smaller nanocrystallites, the nitriding
process took place on two phases - R-Fe(N) and γ′-Fe4N.
Nitrogen concentration on γ′-Fe4N surface was very low in
comparison to surface nitrogen concentration on R-Fe phase.
Thus, while the surface of R-Fe decreased, the total surface
nitrogen concentration decreased, and ammonia decomposition
rate decreased as well. Before the phase transformation of
R-Fe(N) to γ′-Fe4N was almost completed, the nitriding reaction
slowed down and its rate reached its minimum value. At this
moment ammonia decomposition reaction rate was nearly 0.
An increase in the rates of both reactions was observed after a
complete phase transformation. Nevertheless, nitriding reaction
rate quickly reached the next maximum, and then stabilization
of nitriding reaction rate took place (C region). Then, nitriding
reaction rate went down, but ammonia catalytic decomposition,
after reaching the next maximum value, reached a stable state.
Only catalytic ammonia decomposition reaction determined the
gas phase composition after stationary state was reached, viz.,
when the nitriding reaction was completed. The participation
of catalytic ammonia decomposition reaction in the entire
nitriding process is the higher the higher process temperature
is maintained.
Figure 5. DTG curves for nitriding process of nanocrystaline iron.
of catalytic decomposition reaction as well as by an increase
of reactant volume because of hydrogen and nitrogen
formation, which on account of gas phase mixing caused
ammonia dilution. Hence, making use of the determined
concentrations of particular gas phase components, the
amount of ammonia amount which really was decomposed
at any moment was determined.
Figures 8 and 9 present the results in the form of DTG,
regarding nitriding reaction, as well as calculated values of
ammonia decomposition reaction rate for temperatures 673 and
723 K respectively. Initially, (A region) ammonia replaced
hydrogen inside the reactor. One could observe only catalytic
ammonia decomposition. The rate of this reaction increased on
R-Fe, because surface nitrogen concentration increased. The rate
of this reaction reached its maximum value at the moment when
nitriding reaction just started. Also, the iron bulk concentration
In Figure 10, the sum of rates of the nitriding reaction and
ammonia decomposition is presented.
Considering nanomaterials, a diffusion in the bulk can be
neglected. It implies, that the surface chemical reaction can be
Figure 6. Gas phase composition changes at 673 K: (a) calculated concentrations values and nitriding reaction; (b) gas phase composition occurring
during nitriding process.

Studies of Kinetics of Two Parallel Reactions
J. Phys. Chem. A, Vol. 113, No. 2, 2009 415
Figure 7. Gas phase composition changes at 723 K: (a) calculated concentrations values and nitriding reaction; (b) gas phase composition occurring
during nitriding process.
Figure 9. Reaction rates at 723 Ksnitriding reaction (DTG) and
Figure 8. Reaction rates at 673 Ksnitriding reaction (DTG) and
catalytic ammonia decomposition reaction (rNH3).
catalytic ammonia decomposition reaction (rNH3).
the rate limiting step. Then, between reactants on the surface
and dissolved in the bulk the thermodynamic equilibrium holds.
The relationship between the concentration of nitrogen in iron
bulk, Xb, and the surface coverage degree, θ, is described by
the McLean-Langmuir equation:12
Xb
θ
∆G
RT
)
exp -
(11)
(
)
1 - θ 1 - Xb
where ∆G ) free Gibbs enthalpy of segregation, R ) gas
constant, and T ) temperature.
Figure 10. The rate of nitriding process (nitriding reaction and
ammonia decomposition) at temperature 673 and 723 K.
Considering the critical point between R-Fe(N) and γ′-Fe4N
phase and assuming, that for R-Fe phase ∆G ) -110 kJ/mol13
and Xb ) 0.0012 mol/mol,2 it was calculated that θ ) 0.999998
at temperature 673 K. At this point the rate of nitriding reaction
was 0. Thus we have the Langmuir equation14
The number of collisions can be described as follows:
1
N
p
mol
r ) pkNH3,adsS(1 - θ) - kNH3,decSθ2 [mol/s]
(12)
z )
(15)
2
[
]
√
m s
2 π m k T
where p ) gas partial pressure, kads ) adsorption constant,
kNH3,dec ) ammonia decomposition constant, and S ) total
surface area, can be transformed as follows:
Here N ) Avogadro’s number, m ) mass of particle hitting
the surface, and k ) the Boltzmann constant.
It was found that at the critical point z ) 1230 mol m-2 s-1
.
pkNH3,ads(1 - θ) ) kNH3,decθ2 [mol/s]
(13)
The number of collisions resulting from the total surface area
(S ) 5.5 m2/g) and mass of the solid sample (1 g) used during
the investigations was z′ ) 6765 mol s-1. Thus the sticking
coefficient, s0, for θ ) 0 is 0.12. The corresponding literature
values are s0 ) 0.16 on the Fe(100) surface at -153 °C15 and
0.03 for Fe(111) surface contaminated with oxygen at 27 °C.16
The above presented calculations were carried out assuming
that the whole surface of the solid sample took part in the
nitriding process. However one should take into consideration
that only ca. 50% of the whole surface is available for gas
reactants.17,18
Taking into account that θ ) 0.999998 and p ) 0.297 bar,
one can obtain
k
NH3,ads ) 1683500
(14)
1
bar
kNH3,dec
From Figure 8, for 1 g of solid sample, it is known that kNH3,dec
S θ2 ) 0.00162 mol/s. Assuming, that S ) 5.5 m2/g, it was
calculated that the value of kNH3,dec ) 0.00029 mol g m-2 s-1
and then kNH3,ads ) 2685 mol bar-1 s-1
.

416 J. Phys. Chem. A, Vol. 113, No. 2, 2009
Arabczyk and Pelka
Conclusions
References and Notes
(1) Lehrer, E. Z. Elektrochem. 1930, 36 (6), 383–392.
(2) Kunze, J. Nitrogen and Carbon in Iron and SteelsThermodynamics;
Physical Research Volume 16; Akademie Verlag: Berlin, 1990.
(3) Kooi, B. J.; Somers, M. A. J.; Mittemeijer, E. J. Metall. Mater.
Trans. A 1996, 27 A, 1064.
(4) Jennings, J. R., Ed.; Catalytic Ammonia Synthesis; Plenum Press,
New York, 1991,
A combination of thermogravimetric measurement of mass
changes of solid samples, during reaction between nanocrys-
talline iron and ammonia, with gas phase composition analysis
makes it possible to estimate rates of parallel reactions at any
moment of the investigated process.
The obtained results concerning catalytic ammonia decom-
position reaction rate imply, that when nanocrystalline iron
nitriding process is conducted one can observe reaction rate
changes of nitriding reaction as well as of catalytic ammonia
decomposition reaction vs time. Nitriding reaction rate reaches
the highest value during R-Fe nitriding reaction. Catalytic
ammonia decomposition reaction rate reaches the highest value
upon R-Fe as well, although the reaction rate values of this
reaction are clearly lower upon recently created nitrides. This
is due to lower nitrogen concentration on the iron nitride surface
in comparison to nitrogen concentration on iron surface.
On the basis of the calculations performed for the critical
point between R-Fe and γ′-Fe4N phases, the rate constants of
ammonia decomposition and ammonia adsorption process were
estimated. The number of collisions and the sticking coefficient
were also assessed.
(5) Aika, K.; Christiansen, L. J.; Dybkjaer, I.; Hansen, J. B.; Nielsen,
P. E. H.; Nielsen, A.; Stolze, P.; Tamaru, K. Ammonia Catalysis and
Manufacture; Springer Verlag, Berlin/Heidelberg, 1995.
(6) Grabke, H. J. Ber. Bunsen-Ges. Phys. Chem. 1968, 4, 533–543.
(7) Grabke, H. J. Arch. Eisenhu¨ttenwesen 1973, 44, 603–608, no. 8.
(8) Wro´bel, R.; Arabczyk, W. J. Phys. Chem. A 2006, 110, 9219–
9224.
(9) Arabczyk, W.; Zamynny, J. Catal. Lett. 1999, 60, 167.
(10) Baranski, A.; Dziembaj, R.; Kotarba, A.; Jagan, J. M.; Jojewska,
J.; Pieprzyk, E.; Baerns, M.; Mleczko, L. Solid State Ionics 1997, 101-
103, 677–680.
(11) Perego, C.; Peratello, S. Catal. Today 1999, 52, 133–145.
(12) Benard, J. Stud. Surf. Sci. Catal. 1983, 13, 261.
(13) Grabke, H. J. Z. Phys. Chem. Neue Folge 1976, 100, 185.
(14) Langmuir, I. J. Am. Chem. Soc. 1918, 40, 1361.
(15) Weiss, M.; Ertl, G.; Nietsche, F. Appl. Surf. Sci. 1979, 2, 614.
(16) Gay, I. D.; Textor, M.; Mason, R.; Iwasawa, Y. Proc. R. Soc.
(London) 1977, A356, 25.
(17) Arabczyk, W.; Narkiewicz, U.; Moszynski, D. Appl. Catal. A: Gen.
1999, 182, 379.
(18) Boisen, A.; Dahl, S.; Norskov, J. K.; Christensen, C. H. J. Catal.
2005, 230, 309–312.
Acknowledgment. The work has been financed as Research
Project No. NN 507 4461 33.
JP8079759