1
60
G. Munteanu et al. / Thermochimica Acta 329 (1999) 157±162
leads further to the blockage of the hematite reduction
to magnetite.
For the reaction rate we have used the relationship
[7,8]:
dnox
In the TPR curve of sample B there are only two
peaks, their areas having the ratio 1.6:8. This fact
means that the ®rst peak, denoted by 1 in Fig. 1(B), is
not due only to the reduction of hematite to magnetite.
It hides an additional process that is probably due, in
our opinion, to the hydrogenation of sulfate anions on
the surface of sample B.
v �
dt
2
n0D
0
VrꢀT0=Tꢀp=p0 ꢀ2D0=noxk0 expꢀE=RT
5)
where nox stands for the concentration of the oxidic
form, n is the hydrogen concentration at the entrance
(
The hydrogen consumption originating in the
reduction of hematite to magnetite is regarded as an
internal standard. The ratio of hydrogen consump-
tion assigned to the elimination of sulfate anions
from the two (A and B) samples was determined
as 1:2.4. Taking into account this ®nding, as well as
that the sulfur contents in the two samples are in
the ratio 1:4.3, we have concluded that in sample B
the surface ``sulfate anions'' have a structure close to
SO . Consequently, to eliminate the bonded SO in
0
of the reactor, D is the ¯ow rate in normal conditions
0
of pressure and temperature, V is the volume of the
r
reaction space, p and T are the pressure and tempera-
ture in the reaction space, respectively, p and T are
the values of the same parameters in normal condi-
tions, k is the pre-exponential factor, E is the activa-
tion energy, R is the gas constant.
The concentrations of the oxidic species, of sulfate
anions as well as of sulfur atoms, as functions of
temperature, were determined by numerical integra-
tion of the reaction rate equation, Eq. (5), using the
fourth-order Runge±Kutta method [9]. The total
hydrogen consumption has been calculated by means
0
0
0
2
2
sample B, we have considered the following two
processes:
(3)
of the continuity equation:
X
The second peak from the TPR curve of sample B,
denoted by 2 in Fig. 1(B), has been assigned to the
reduction of magnetite to metallic iron.
Án Vr
where v stands for the reaction rate of species i from
vi=D
(6)
i
the surface and D for the ¯ow rate.
4
. Simulation of the TPR curves
5
. Results and discussion
The hydrogen consumption due the elimination of
The values of the kinetic parameters characteristic
of the above mentioned processes were evaluated by
tting the computed TPR spectra to the experimental
the OH groups from the surface was neglected,
although both the samples had not been dehydroge-
nated before performing the TPR measurements.
The two reduction processes, described by
Eq. (1), as well as those speci®c to the elimination
of sulfate anions, described by Eqs. (2) and (3),
were approximated by second-order reactions. We
adopted this procedure because the elimination of
any oxygen atom, whether it belongs to an iron oxide
or to a sulfate anion, may be regarded as reaction of the
type:
®
ones. In such a procedure one has always to take into
account the physical signi®cance of the results
obtained. It is possible to obtain a good ®t using for
the kinetic parameters values without any physical
signi®cance. For example, if one assumes that the ®rst
peak from curve B (Fig. 1(B)) is due to only one
hydrogen consuming process (neglecting the above
conclusion that this peak is due to at least two different
processes) we obtain a value for the activation energy
�
1
of 774Æ0.4 kJ mol . This value is, of course, abnor-
mally high, suggesting again that this peak is due to
the superposition of, at least, two hydrogen consuming
processes.
H fOg ! f&g H O
(4)
2
S
2
where {O} denote a surface oxygen atom and {&}
S
the oxygen vacancy on the surface.