Chlorination of lanthanum oxide

Article abstract of DOI:10.1021/jp210457r

The reactive system La₂O₃(s)-Cl₂(g) was studied in the temperature range 260-950 °C. The reaction course was followed by thermogravimetry, and the solids involved were characterized by X-ray diffraction, scanning electron microscopy, and energy dispersive spectroscopy. The results showed that the reaction leads to the formation of solid LaOCl, and for temperatures above 850 °C, the lanthanum oxychloride is chlorinated, producing LaCl₃(l). The formation of the oxychloride progresses through a nucleation and growth mechanism, and the kinetic analysis showed that at temperatures below 325 °C the system is under chemical control. The influence of diffusive processes on the kinetics of production of LaOCl was evaluated by studying the effect of the reactive gas flow rate, the mass of the sample, and the chlorine diffusion through the boundary layer surrounding the solid sample. The conversion curves were analyzed and fitted according to the Johnson-Mehl-Avrami description, and the reaction order with respect to the chlorine partial pressure was obtained by varying this partial pressure between 10 and 70 kPa. The rate equation was obtained, which includes the influence of the temperature, chlorine partial pressure, and reaction degree.

Full text of DOI:10.1021/jp210457r

Article
pubs.acs.org/JPCA
Chlorination of Lanthanum Oxide
Juan P. Gaviría,* Lucas G. Navarro, and Ana E. Bohe
,†,‡
†
†,‡,§
́
†
Departamento de Fisicoquímica y Control de Calidad, Complejo Tecnolog
Nacional de Energía Atomica, Av. Bustillo km 9500 (8400), S.C. de Bariloche, Río Negro, Argentina
Consejo Nacional de Investigaciones Científicas y Tecnicas, Argentina
́ ́ ́
ico Pilcaniyeu, Centro Atomico Bariloche, Comision
́
‡
́
§
Centro Regional Universitario Bariloche, Universidad Nacional del Comahue, Argentina
ABSTRACT: The reactive system La O (s)−Cl (g) was studied in the
2
3
2
temperature range 260−950 °C. The reaction course was followed by
thermogravimetry, and the solids involved were characterized by X-ray
diffraction, scanning electron microscopy, and energy dispersive spectros-
copy. The results showed that the reaction leads to the formation of solid
LaOCl, and for temperatures above 850 °C, the lanthanum oxychloride is
chlorinated, producing LaCl (l). The formation of the oxychloride
3
progresses through a nucleation and growth mechanism, and the kinetic
analysis showed that at temperatures below 325 °C the system is under
chemical control. The influence of diffusive processes on the kinetics of
production of LaOCl was evaluated by studying the effect of the reactive
gas flow rate, the mass of the sample, and the chlorine diffusion through
the boundary layer surrounding the solid sample. The conversion curves
were analyzed and fitted according to the Johnson−Mehl−Avrami description, and the reaction order with respect to the
chlorine partial pressure was obtained by varying this partial pressure between 10 and 70 kPa. The rate equation was obtained,
which includes the influence of the temperature, chlorine partial pressure, and reaction degree.
8
,15−19
1
. INTRODUCTION
chlorinated hydrocarbons, toxins of the environment.
Its
nanomaterial can be used as a luminescent material and gas
Direct chlorination is a potential dry method for the production
of rare earth chlorides, which allows further isolation of metals
from ores and concentrates. The combination of chlorination
and chemical vapor transportation is a method for recovery and
separation of rare earths from oxide mixtures and sludges.
20−22
sensor.
. MATERIALS AND METHODS
The materials used were a lanthanum hydroxide [La(OH) ]
3
1
2
2
,3
Although there are several studies about chlorination
powder 99.98% (Fluka), chlorine (Cl ) 99.8% purity (Indupa,
́
Bahıa Blanca, Argentina), and argon 99.99% purity (AGA,
Buenos Aires, Argentina). The particle size distribution of the
2
4
,5
6−10
with NH Cl and CCl4
of lanthanum oxide and other
4
lanthanum compounds, none have paid special attention to the
kinetics, mechanism, and structural changes and there are no
published reports about direct chlorination of lanthanum oxide
with gaseous chlorine.
La(OH) powder measured by laser diffraction showed that
3
9
4
0% v/v of the particles have a size lower than 10 μm and
5% v/v have sizes between 2 and 6 μm (Mastersizer, Malvern
The aim of this work is to continue with the kinetic research
Instruments Limited).
Solids were analyzed by X-ray diffraction (Philips PW 1310-01,
with Ni filtered Cu Kα radiation), scanning electron microscopy
11−14
on direct chlorination of rare earth oxides,
and to study
the chlorination of La O in order to determine the rate
2
3
equation, the mechanism of the reaction, and the kinetics
parameters.
An electro-balance was used to follow the course of the
reaction, and the resulting solids were analyzed with X-ray
diffraction (XRD), scanning electron microscopy (SEM), and
energy dispersive spectroscopy (EDS).
(SEM 515, Philips Electronic Instruments), and energy dispersive
spectroscopy (EDAX Genesis 2000).
The chlorination reactions were carried out using a thermo-
23
gravimetric system. It consists of an electro-balance (Cahn
2
000, Cahn Instruments, Inc., Cerritos, CA) adapted to work
with corrosive atmospheres, a vertical tube furnace, a gas line,
and a data acquisition system. The sensitivity of the system
is ±0.01 mg while operating at 1000 °C under a gas flow rate of
The studies showed that LaOCl is the only existing solid
phase after direct chlorination of La O at the working
2
3
temperature range. This compound has several potential uses
as it is becoming of great interest within various industries due
to its capability and potential to be used as a catalyst in the
oxidative methane coupling, oxidative dehydrogenation of
ethane, oxidative chlorination of methane, conversion of ethane
and ethylene to vinyl chlorides and in the destruction of
9
L/h (measured at normal temperature and pressure). Each
Received: July 26, 2011
Revised: January 25, 2012
Published: January 25, 2012
©
2012 American Chemical Society
2062
dx.doi.org/10.1021/jp210457r | J. Phys. Chem. A 2012, 116, 2062−2070

The Journal of Physical Chemistry A
Article
sample was placed in a cylindrical silica glass crucible (7.8 mm
inner diameter and 3.3 mm deep), which hangs from one of the
arms of the electro-balance through a silica glass wire. The
reactor is a hang-down silica glass tube (4.6 cm inner diameter)
group, P-3m1; space group number, 164). These results are in
25
good agreement with those obtained by Neumann et al.
The sample of La O obtained was left in the reactor and
2
3
subsequently chlorinated.
which carries the gases (Ar/Cl ) to the sample. The tem-
perature was measured using a Pt−Pt (10% Rh) encapsulated
thermocouple which was placed 5 mm below the crucible.
2
3. RESULTS AND DISCUSSION
3
.1. Thermodynamic Considerations. The following
Flows of Ar and Cl were controlled by flow meters, and they
2
reactions are possible in the La O −Cl reactive system:
2
3
2
were dried by passing them through silica gel and CaCl2,
respectively. The La O samples (product of the La(OH)
3
La
O
(s) + Cl
(g) → 2LaOCl(s) + 1/
O
(g)
(4)
2
3
2
3
2
2
2
decomposition) were heated in flowing Ar until the reaction
temperature was reached. After temperature stabilization, chlo-
rine was admitted into the reactor while mass changes were
monitored every 2.5 s. The acquired data were carefully analy-
zed to determine the reaction zero time.
1
2
1
/ La O (s) + Cl (g) → / LaCl (s, l) + / O (g) (5)
3
2 3
2
3
3
2 2
1
LaOCl(s) + Cl (g) → LaCl (s, l) + / O (g)
(6)
2
3
2 2
1
1
2
.1. Production of La O by La(OH) Decomposition.
LaOCl(s) → / La O (s) + / LaCl (s, l)
2 3
2
3
3
3
3
3
(7)
The lanthanum oxide used in the chlorinations was obtained by
The physical properties reported for lanthanum compounds are
La(OH) thermal decomposition under Ar atmosphere. The
3
26
the following: La O (T = 2313 °C); LaOCl (T = 934 °C
thermal treatment was a heating of 40 min until 850 °C and
2
3
m
d
27
26
(
in vacuum)); LaCl (T = 858 °C, T = 1747 °C). T , T ,
3 m v m d
20 min at this temperature. Figure 1 shows a thermogravimetric
and T are the melting, decomposition, and vaporization
v
temperatures.
Figure 2 shows the temperature dependence of the standard
Gibbs free energy for reactions 4−7 between 0 and 1000 °C.
Figure 1. Thermo-gravimetric curve of La(OH) decomposition.
3
curve of the La(OH) decomposition. The curve indicates three
3
steps of weight loss, with these steps being associated with the
following reactions:
Figure 2. Standard Gibbs free energy (per mole of chlorine) vs
temperature for reactions involved in the La O −Cl (g) system. T is
La(OH) (s) → LaOOH(s) + H O(g)
2
3
2
m
3
2
the melting point of LaCl (858 °C).
3
[
≈240−360°C]
(1)
(2)
2
8
The calculations were performed with HSC 6.12 Software,
and the reactions were taken per mole of chlorine.
1
1
LaOOH(s) → / La O (s) + / H O(g)
2
2 3
2 2
The results can be compared, assuming that the reactions are
independent from each other. The curves show that the
chlorination reactions of La O to produce LaOCl and LaCl
3
have negative ΔG values, and the formation of the oxychloride
is thermodynamically more feasible to occur. Then, LaOCl
could be chlorinated (reaction 6) or decomposed (reaction 7).
[≈>440°C]
2
3
La O CO (s) → La O (s) + CO (g)
[≈>630°C]
2
2
3
2 3
2
0
(3)
The presence of a small amount of lanthanum oxide carbonate
La O CO ) in the starting material and the formation of
0
(
The chlorination of LaOCl has negative ΔG values for
2
2
3
0
LaOOH during the decomposition was verified by XRD
temperatures below 580 °C, and has lower ΔG values than the
2
4
(
La O CO : reference pattern, 37-804; crystal system,
LaOCl decomposition reaction (compared per mole of
LaOCl).
2
2
3
hexagonal; space group, P6 /mmc; space group number, 194.
LaOOH: reference pattern, 19-656; crystal system, mono-
clinic; space group, P2 /m; space group number, 11).
The final relative mass loss of the curve corresponds to the
global reaction of La O production by La(OH) decom-
position (≈0.142). The final product of the treatment was
analyzed by XRD obtaining only the diffraction lines of La O
reference pattern, 73-2141; crystal system, hexagonal; space
3
24
The phase stability diagram of the La−O−Cl system at
28
400 °C is shown in Figure 3. The axes are the partial pressure
1
of O and Cl . The diagram indicates that La O and LaCl do
2
2
2
3
3
not have a thermodynamic equilibrium, and LaOCl has to be
formed previous to the formation of LaCl from La O .
3.2. Nonisothermal Chlorination Experiments. Non-
isothermal chlorination experiments were carried out in order
2
3
3
3
2
3
2
3
24
(
2
063
dx.doi.org/10.1021/jp210457r | J. Phys. Chem. A 2012, 116, 2062−2070

The Journal of Physical Chemistry A
Article
Figure 3. Phase stability diagram of the La−O−Cl system at 400 °C.
−
6
Experimental conditions: pCl , 35 kPa (0.35 bar); pO , 10 bar
2
2
(
1 ppm).
Figure 5. Diffraction patterns for (1) unreacted La O ; (2) product of
the nonisothermal chlorination; (3) LaOCl reference pattern.
2
3
24
to determine both the reactivity with temperature and the
reaction products. Figure 4 shows the relative mass change (Δm/m0)
temperature range (260−950 °C) showed that LaOCl was the
only phase present. This behavior is indicating that conversion
vs temperature for this system under Ar−Cl atmosphere from
2
of LaOCl to LaCl is kinetically limited.
100 to 1000 °C.
3
Therefore, the results indicate that the reaction taking place
is reaction 4:
1
La O (s) + Cl (g) → 2LaOCl(s) + / O (g)
(4)
2
3
2
2 2
The conversion degree is defined as
m − m
0
tLa O
2 3
α = α
=
La O
2
3
m
(I)
0
where m and m
are the weights of La O at zero time and
2 3
0
tLa
O
2 3
time t, respectively.
The mass change measured in the thermo-balance (Δm)
includes the weights of La O and LaOCl. After relating these
2
3
weights with conversion, an expression is obtained for α as a
function of Δm:
⎛
⎜
⎝
⎞
Figure 4. Nonisothermal thermo-gravimetry of La O chlorination.
Δm
2
3
⎟
⎛
≅ ⎜
⎝
⎞
⎟
⎠
m ⎠
Δm
0
1
0
α =
The sample undergoes a mass gain starting from a
temperature of about 260 °C. The mass change rate increases
quickly until the mass change reaches a maximum value of
⎛
⎝
⎞
⎠
m
0.169
PM
LaOCl
⎜
⎜
2
− 1⎟
⎟
PM
La O
2
3
(II)
Δm/m = 0.18, very close to the relative mass change expected
0
for reaction 4. Figure 5 shows the XRD patterns for the
unreacted La O (obtained by La(OH) decomposition), the
where 0.169 is the final relative mass change of reaction 4.
The formation of LaOCl as the intermediate solid reaction
product in the production of LaCl by La chlorination with
NH Cl and CCl has been observed by several authors
(NH
.3. Isothermal Thermo-Gravimetric Experiments. The
isothermal chlorination experiments performed at temperatures
between 260 and 950 °C (m0,La O3, 10 mg; pCl , 35 kPa; QAr−Cl2,
2
3
3
residue after nonisothermal chlorination, and the LaOCl
3
O
2 3
24
pattern (reference pattern, 8-477; crystal system, tetragonal;
space group, P4/nmm; space group number, 129). The solid
residue only showed the LaOCl diffraction lines, and the
analyses performed by EDS were consistent with the chemical
formula of this compound. The EDS analyses were performed
over three samples chlorinated at 350, 850, and 950 °C. The
average composition, taking into account only La and Cl, for six
measurements (two for each sample) was 52 ± 3 for La and
4
4
4
,5
6−9
Cl; CCl
).
4
4
3
2
2
4 L/h) showed that the system has different kinetic behavior
depending on the temperature range (Figure 6). For temper-
atures below 325 °C, the conversion curves have a sigmoidal
shape, which is typical for nucleation and growth transformations
(Figure 14). This type of process has been observed for the
formation of REOCl by RE O chlorination (RE: rare eath)
4
8 ± 3 for Cl (% at). Lanthanum oxychloride is a not hygro-
29
scopic white powder and is stable when exposed to air.
The thermodynamical analysis showed that LaOCl is the
compound thermodynamically more feasible to be formed from
La O chlorination (Figure 2) and LaCl is the more stable final
2
3
1
3
14,30,31
(samarium, yttrium
). For temperatures above 350 °C,
the initial reaction rate is the fastest of the overall reaction,
which is a different behavior compared to the reactions per-
formed at lower temperatures where the reaction rate at the
2
3
3
product under the experimental conditions used (Figure 3).
The DRX analyses of the final condensed products in the whole
2
064
dx.doi.org/10.1021/jp210457r | J. Phys. Chem. A 2012, 116, 2062−2070

The Journal of Physical Chemistry A
Article
1
LaOCl(s) + Cl (g) → LaCl (l) + / O (g)
(6)
(8)
2
3
2 2
LaCl (l) = LaCl (g)
3
3
Reactions 6 and 8 occur at temperatures above 850 °C. Reac-
tion 8 is faster than reaction 6, causing LaCl to not be present
3
as a condensed product after the chlorinations.
3
.4. Morphological Analysis. Figure 8 shows SEM images
with 10000× magnification of (a) starting La(OH) , (b) La O
3
3
2
obtained by La(OH) decomposition, (c) LaOCl produced at
3
3
25 °C, and (d) LaOCl produced at 950 °C. It can be observed
that there are few morphological differences between La(OH)3
and La O . The sample of LaOCl formed at 325 °C has a
2
3
smaller grain size than the La O , which is in accordance with
2
3
the behavior expected for a nucleation and growth mechanism
(several LaOCl nuclei are formed inside the La O grains). The
Figure 6. Isothermal chlorination of La O .
2
3
2
3
morphology of the lanthanum oxychloride is quite different
at temperatures below and above the melting point of LaCl3
(858 °C). At temperatures below 858 °C, the LaOCl grains are
smaller than those obtained at higher temperatures (Figure 8c
and d).
beginning of the reaction is nearly zero (transformations which
follow a nucleation and growth mechanism have a theoretical
initial rate equal to zero).
For temperatures in the range 350−500 °C, the beginning of
the reaction proceeds through a fast stage and the formation of
LaOCl is subsequently completed by a slower kinetic stage. For
temperatures above 600 °C, the production of LaOCl is fully
performed by the fast stage, being rate independent of the
temperature.
3.5. Effect of the Gas Phase Mass Transfer. The kinetics
in gas−solid reactions can be affected by two gas phase
phenomena: convective mass transfer through the boundary layer
32
and reacting gas supply. The first process is analyzed by
comparing the calculated chlorine molar flow diffusing through
the boundary layer surrounding the solid sample with the measured
chlorine molar flow (obtained from thermo-gravimetric curves and
assuming a given reaction stoichiometry). The second one can be
evaluated by studying the influence of the total flow rate (supply
rate of chlorine) on the reaction rate.
Chlorination experiments performed at temperatures above
8
50 °C showed that after the formation of LaOCl the system
progresses with mass loss (Figure 7). The XRD analyses of the
solid residues of these reactions showed that LaOCl is the only
compound present.
The molar flow of Cl by diffusion (and assuming no
2
starvation) can be estimated by using the Ranz−Marshall
3
3,34
correlation
(the development of eq III is detailed in ref 31):
1/3
dn
D
(2 + 0.6NRe1/2
N
)
Cl
Ar−Cl2
Sc
2
=
dt
L
(
P
− P
)
Cl ,S
2
Cl ,0
2
A
R T
g
(III)
where dn /dt is the chlorine molar flow, D
is the binary
Cl
2
Ar−Cl
2
diffusion coefficient for Ar−Cl , N = U·L/ν represents the
2
Re
Reynolds number (U is the linear fluid velocity, and ν is the
kinematic viscosity), N = ν/D is the Schmidt number, L
Sc
Ar−Cl
2
is the sample characteristic dimension (taken as the crucible
diameter), A is the sample external surface (considered equal to
the crucible transversal area), R is the gas constant, P is the
Figure 7. Isothermal chlorinations of La O above 850 °C.
2
3
g
Cl
2
chlorine partial pressure, and the subscripts 0 and S represent
the bulk gas stream and sample surface, respectively.
Samples of LaOCl treated at 950 °C in argon atmosphere for
h showed no mass changes. After this treatment, the samples
were analyzed by XRD, obtaining only the LaOCl diffraction lines.
These results are indicating that weight loss is caused by a
4
The values of dn /dt calculated by eq III and those obtained
Cl
2
from thermo-gravimetric curves, for temperatures between 300
and 700 °C, are shown in Table 1. Reaction 4 was assumed as
irreversible, since the values of the equilibrium constants are
28
reaction between LaOCl and Cl , as can be reaction 6. The
2
lanthanum trichloride formed is in liquid state, with the vapor
17
7
−
5
−4
7.07 × 10 and 5.48 × 10 at 300 and 800 °C, respectively.
Therefore, the value of PCl2,S employed in the calculation is zero.
pressures being 4.7 × 10 and 3.7 × 10 bar at 850 and
28
9
50 °C, respectively. The reaction is carried out under a
continuous flow which removes the vapor in equilibrium with
the liquid.
The ν and D
Enskog theory.
values were obtained from the Chapman−
The experimental conditions considered were
Ar−Cl
5,36
2
3
Then, reactions involved in the La O −Cl system are the
2
3
2
pCl = PCl2,0, 35 kPa; QAr−Cl2, 4 L/h.
2
following:
Equation III gives approximate values of dn /dt. This equa-
Cl
2
tion is valid for a pellet in a freely flowing gas and not for a
sample contained in a crucible. Hills and Hakvoort con-
1
La O (s) + Cl (g) → 2LaOCl(s) + / O (g)
(4)
32
37
2
3
2
2
2
2
065
dx.doi.org/10.1021/jp210457r | J. Phys. Chem. A 2012, 116, 2062−2070

The Journal of Physical Chemistry A
Article
Figure 8. SEM images of (a) starting La(OH) ; (b) La O produced by La(OH) decomposition; (c) LaOCl formed at 325 °C; (d) LaOCl formed
3
2
3
3
at 950 °C.
Table 1. Comparison between Calculated and Measured
a
Molar Flow of Chlorine
b
c
T
°C)
(1) dnCl
2
/dt [calculated]
(2) dnCl /dt [measured]
2
(
(mol of Cl /s)
(mol of Cl /s)
(1)/(2)
2
2
4.05 × 10−
4.31 × 10−
6
2.72 × 10
−9
1487
131
19
3
00
50
00
00
00
6
−8
3
3.30 × 10
−
6
6
6
−7
4
5
7
4.55 × 10
2.40 × 10
5.02 × 10−
5.88 × 10−
1.22 × 10
−6
4.1
−6
1.68 × 10
3.5
a
2
Data used for the calculations: L = 0.74 cm; A = 0.43 cm ; pCl =
2
b
c
3
5 kPa; QAr−Cl2 = 4 L/h. Calculated from eq III. Obtained from the
thermogravimetric curves at the higher slope (stoichiometry of
reaction 4).
Figure 9. Effect of Ar−Cl2 flow rate on the kinetics of La O
chlorination at 350 °C. Two curves of 4 L/h are presented to show
the reproducibility of the system.
cluded that values obtained using the Ranz and Marshall
correlation are more than 1 order of magnitude higher than
mass transfer rates for samples contained within crucibles. The
criterion used in this work to conclude that the convective mass
transfer of chlorine to the surface of the sample is not influencing
the reaction rate is when the measured chlorine flow is at least 2
orders of magnitude smaller than the one obtained by using eq III.
The results shown in Table 1 indicate that for temperatures below
2
3
temperature. Two curves of 4 L/h are presented to show the
reproducibility of the system.
3.6. Mass Transfer through the Interparticle Voids.
Diffusion of the gas species (reactants and products) through
pores between solid particles can affect the reaction rate. This
process was analyzed by studying the effect of the reaction bed
thickness, which was varied by changing the mass of La O . The
3
50 °C the reaction kinetics are not affected by this diffusive
process.
Figure 9 shows the influence of the total flow rate on the rate
of La O chlorination at 350 °C, using Q between 2 and
2
3
reaction rate decreases as the bed thickness increases when the
interparticle void diffusion is playing a role in the rate control.
Figure 10 shows thermo-gravimetric curves of chlorination
experiments performed at 325 °C and initial masses of 2 and
10 mg. The reaction rate does not depend on the sample size at
this temperature; therefore, the mass transfer through the
2
3
Ar−Cl
2
8
L/h. The difference between the conversion curves is similar
to the experimental scattering, and it can be concluded that the
supply rate of chlorine is not influencing the kinetics at this
2
066
dx.doi.org/10.1021/jp210457r | J. Phys. Chem. A 2012, 116, 2062−2070

The Journal of Physical Chemistry A
Article
interparticle voids is not affecting the overall reaction rate and
hence an initial mass of 10 mg was used in the kinetic analysis.
produced in the La O −Cl interaction. The chlorine radicals react
2
3
2
readily with the oxide to form the oxychloride. As the reaction
thickness of the bed increases, the reaction probability before the
recombination to form Cl increases. When the reaction thickness
2
diminishes, the radicals have a lower residence time in the bed and
the probability of escaping from the sample is higher. Although the
behavior can be explained with the proposal of the chlorine radical
formation, the presence of this species should be demonstrated.
38,39
Pasquevich et al.
observed the same kinetic behavior in the
carbochlorination of ZrO (enhancement of the conversion with
2
the increase of the initial mass) and verified by modulated-beam
mass spectroscopy the formation of chlorine radicals.
3
.8. Kinetic Analysis and Determination of the
Equation Rate for T ≤ 325 °C. The intrinsic kinetic param-
eters of La O chlorination that produces LaOCl and O (reaction 4)
2
3
2
are determined in this section. The results analyzed in previous
sections showed that the reaction at temperatures below 325 °C
follows a unique kinetic behavior and the reaction rate is not
affected by gas phase and interparticle void mass transfers (QAr−Cl2,
Figure 10. Effect of the initial mass of La O on reaction rate at
2
3
325 °C.
4
L/h; m , 10 mg). Therefore, the kinetics are under chemical control
0
and the reaction rate can be expressed as a function of separate
variables (T, p, α):
3
.7. Kinetic Analysis for T ≥ 350 °C. For temperatures
above 350 °C, the beginning of the reaction proceeds by a fast
stage. The kinetics of this stage are affected by the gas phase
mass transfer (Table 1, section 3.5); therefore, it is not possible
to obtain intrinsic kinetic parameters.
dα
dt
=
k(T)F(pCl )G(α)
2
(IV)
The extension of the fast stage is raised as the temperature
Figure 6) and initial mass are increased. Figure 11 shows the
where k(T) refers to an Arrhenius equation, F(pCl ) expresses the
2
(
dependence of reaction rate on pCl (it is related to the adsorption/
2
effect of the initial mass of La O on the kinetics of chlorination
desorption phenomena), and G(α) is a function that describes the
morphological evolution of the solid reactant.
2
3
at 450 °C. The extension of the fast stage is α = 0.19, 0.74, and
.97 for m = 10, 30, and 40 mg, respectively. The same
0
3.8.1. Reaction Order with Respect to pCl . Figure 12
0
2
behavior was observed at 500 °C. Figure 11 also shows that the
shows the effect of the chlorine partial pressure on the reaction
rate at 300 °C, varying pCl between 10 and 70 kPa.
2
Figure 11. Effect of the mass sample on the kinetics of La O3
2
chlorination at 450 °C.
Figure 12. Effect of chlorine partial pressure on the reaction rate of
La O chlorination at 300 °C.
2
3
time necessary to reach a given conversion depends on the
initial mass. For example, a value of α = 0.91 is reached in 300
and 62 s for 10 and 40 mg, respectively.
The reaction order is obtained from the integration of eq IV
between α = 0 and α = α and t = 0 and t = t. The temperature is
constant (300 °C), and it is assumed that dependence on α and
This behavior is not the most common in gas−solid kinetics.
An increment in the reaction bed thickness generally causes the
following, depending on the relative importance of the diffusion
through the sample pores: (1) a diminution in the reaction rate
for systems affected by this diffusion; (2) no change in the kinetics
for systems which are not influenced by this process transfer. The
increase of the fast stage extension and the reaction rate with the
initial mass could be related to the formation of highly reactive gas
intermediates. The chlorine radical (Cl•) is a probable species to be
pCl can be separated.
2
dα
G(α)
=
k(T)F(pCl ) dt
2
∫
∫
(V)
g(α = α) − g(α = 0) = k(T)F(pCl )t(α)
(VI)
2
2
067
dx.doi.org/10.1021/jp210457r | J. Phys. Chem. A 2012, 116, 2062−2070

The Journal of Physical Chemistry A
Article
where g(α) is commonly called the integrated conversion function
and t(α) is the time to reach a conversion degree α at a given
temperature and reactive gas partial pressure.
Table 2. Values of n and K Obtained from a Nonlinear Least
Squares Fitting of Conversion Curves with eq XI
K (s−1)
(2.148 ± 0.001) × 10−
(3.901 ± 0.002) × 10−
(6.577 ± 0.001) × 10
(8.804 ± 0.003) × 10−
r2
x
T (°C)
60
270
80
290
n
Assuming that F(pCl ) = B·pCl :
2
2
5
5
5
5
2
1.471 ± 0.001
1.525 ± 0.001
1.396 ± 0.0005
1.377 ± 0.001
1.540 ± 0.002
1.457 ± 0.002
0.9988
0.9981
0.9998
0.9976
0.9958
0.9981
x
g(α) − g(0) = k(T)BpCl t(α)
(VII)
2
−
2
k(T)B
g(α) − g(0)
x
1
/t(α) =
pCl
2
−4
3
3
00
25
(1.266 ± 0.007) × 10
(3.817 ± 0.003) × 10−
(VIII)
4
⎡
⎤
k(T)B
g(α) − g(0)⎦
−
ln t(α) = ln⎢
⎥ + x ln pCl2
⎣
(IX)
The function inside brackets, at constant temperature, depends
only on α:
−
ln t(α) = H(α) + x ln pCl2
(X)
The reaction order for a given α and T is obtained by plott-
ing −ln t(α) vs pCl and calculating the slope of the curve.
2
Figure 13 shows these plots, obtaining an average value of 0.23.
It was considered an experimental scattering error of 10% in the
time to reach a given conversion.
Figure 14. Fits of the conversion curves with the JMA equation. Line
graphs, experimental curves; scatter graphs, calculated curves.
The average value of n obtained was 1.461 ± 0.066. The
value of n is dependent on the nucleation and growth
mechanisms. The types of growth considered in the JMA
description are volume diffusion controlled and interface
controlled. Interface controlled can occur in the absence of
compositional changes (e.g., in cases of allotropic phase
transformations), and volume diffusion can occur upon phase
transformations where long-range compositional changes
Figure 13. Plot of eq X at different conversions of curves shown in
Figure 12.
44
take place. The transformation analyzed in this work is the
nucleation and growth of LaOCl from the chlorination of
La O . This process involves compositional changes; therefore,
3.8.2. Reaction Model. The conversion curves were fitted
2
3
according to the Johnson−Mehl−Avrami (JMA) model, which
we can assume the LaOCl nuclei are growing controlled by volume
diffusion.
40−43
has been developed for isothermal transformations.
This
model proposes that the reaction degree has the following
dependence on the reaction time:
The nucleation models taken into account in the JMA
description are continuous nucleation and site saturation. Kempen
44
α = 1 − exp[−K(T)t]n
et al. demonstrated numerically that intermediate values of
the JMA exponents are possible for combination of the different
nucleation models and volume diffusion growth: (I) between 1/2
(one-dimensional growth) and 3/2 (three-dimensional growth)
for site saturation and (II) between 3/2 (one-dimensional growth)
and 5/2 (three-dimensional growth) for continuous nucleation.
The obtained value for the JMA exponent could be con-
sistent with a three-dimensional growth and site saturation.
Figure 15 shows a magnification of Figure 8c, showing the
detail of the LaOCl nuclei. It can be observed that these nuclei
had a three-dimensional growth. The nucleation mechanism of
site saturation assumes that the number of supercritical nuclei
does not change during transformation; all nuclei are present at
zero time. A possible mechanism, coherent with the results
analyzed in section 3.7, is the following: at the beginning of the
(XI)
⎛
⎞
E
a
⎜
⎟
⎟
⎠
K = K exp −
0
⎜
R T
⎝
g
(XII)
where K(T) is the global rate constant (different from k(T) in
eq IV, which is the intrinsic rate constant), K is the pre-
0
exponential factor, E is the effective activation energy, and n is
a
the JMA exponent. The parameters K , E , and n depend on the
0
a
nucleation and growth mechanisms.
The conversion curves at T ≤ 325 °C were fitted with eq XI,
2
and Table 2 shows the values of n, K, and r obtained from a
nonlinear least-squares fitting. Figure 14 shows the experi-
mental curves (line graphs) and the fitted curves (scatter
graphs), showing the good fit between them.
reaction, the Cl molecule is destructively adsorbed and some
2
2
068
dx.doi.org/10.1021/jp210457r | J. Phys. Chem. A 2012, 116, 2062−2070

The Journal of Physical Chemistry A
Article
3.8.3. Activation Energy. The activation energy in the
temperature range of 260−325 °C was calculated from the
values of the global rate constant K(T) obtained by the appli-
cation of the JMA model (Table 2). Figure 16 shows the plot of
ln K vs 1/T, from which a value for the intrinsic activation energy
of 113 ± 5 kJ/mol was obtained. Table 3 shows the comparison
with the intrinsic activation energies obtained for the same
reaction using other rare earths. These reactions were studied
in the same experimental system. The lower E belongs to the
a
lanthanum oxide chlorination, being consistent with the use of
this compound for the destruction of chlorinated hydrocarbons.
4
. CONCLUSIONS
The chlorination of La O begins at temperatures above 260 °C
2
3
and leads to the formation of solid LaOCl. At temperatures
above 850 °C, the lanthanum oxychloride is chlorinated to
lanthanum trichloride, but the rate for the LaCl (l) production
3
Figure 15. SEM image of LaOCl obtained from La O chlorination at
2
3
is lower than its evaporation rate and thermo-gravimetric
experiments show a loss of mass. Kinetic analysis showed that
the reaction is under chemical control for temperatures below
325 °C.
3
25 °C. The formation of LaOCl proceeds through a nu-
cleation and growth mechanism, and the conversion curves
were analyzed with the Johnson−Mehl−Avrami description.
The JMA parameters obtained were n = 1.46 and E = 113 kJ/mol.
a
According to the JMA description and SEM images, the value
of n is consistent with a three-dimensional growth and site
saturation nucleation of the LaOCl phase. An overall rate
equation was obtained, which includes the effects of the
parameters analyzed (a pre-exponential factor was obtained for
each temperature by combining eqs IV, XI, and XII and
calculating the average):
⎛
⎞
dα
dt
112 kJ/mol
6
0.23 −1
⎜
⎟
⎟
⎠
=
1.2 × 10 (seg·kPa
)
exp −
⎜
⎝
R T
g
× pCl 0 {1.46(1 − α)[−ln(1 − α)]}
.23
0.3
Figure 16. Plot of ln K vs 1/T. The K values were obtained from the
JMA analysis.
2
AUTHOR INFORMATION
Corresponding Author
*E-mail: gaviriaj@cab.cnea.gov.ar. Phone: +54 02 94 444 5100.
Fax: +54 02 94 444 5293.
■
Table 3. Activation Energies for Reaction RE O (s) + Cl (g) =
2
3
2
1
2
REOCl(s) + / O (g) (RE: Rare Earth)
2 2
RE
Ea (kJ/mol)
temperature range (°C)
reference
Notes
The authors declare no competing financial interest.
La
Sm
Y
113 ± 5
130 ± 5
187 ± 3
260−325
270−350
575−800
this work
12, 13
14
ACKNOWLEDGMENTS
The authors thank the Agencia Nacional de Promocion
■
́
Cientıf
́
ica y Tecnolog
Investigaciones Cientı
́
ica (ANPCyT), Consejo Nacional de
icas y Tecnicas (CONICET), and
f
́
́
of the generated Cl atoms diffuse into the La O bulk. These
atoms react and produce all LaOCl nuclei, which subsequently
grow.
2
3
Universidad Nacional del Comahue for the financial support.
REFERENCES
Lanthanum oxide is widely used as a reactive substrate in the
low temperature destruction of chlorinated hydrocar-
■
(
1) Gimenes, M.; Oliveira, H. Metall. Trans. B 2001, 32, 1007−1013.
6
−10,16,17
(2) Sun, Y-H; Wang, Z-C; Guo, L. J. Alloys Compd. 1999, 285,
3−76.
bons.
A feature of these reactions is the facile oxygen
7
(
diffusion from the bulk to the surface to undergo O/Cl
exchange.
The function G(α) can be obtained by combining eqs IV and
XI, and the average value of n in the range of temperature
analyzed was 1.46:
3) Murase, K.; Machida, K-I; Adachi, G.-Y. J. Alloys Compd. 1995,
8
2
(
17, 218−225.
4) Zhu, G-C; Li, F-P; Xiao, M.-G. Trans. Nonferrous Met. Soc. China
2003, 13, 1454−1458.
(5) Shi, W-Z; Zhang, X.; Zhao, Y-H; Wang, J-Y; Zhu, G.-C. Chin.
J. Process Eng. 2005, 5, 23−28.
G(α) = n(1 − α)[−ln(1 − α)](n−1)/n
(6) Weckhuysen, B. M.; Rosynek, M. P.; Lunsford, J. H. Phys. Chem.
Chem. Phys. 1999, 1, 3157−3162.
(n = 1.46)
XIII)
(
(7) Weckhuysen, B. M. Phys. Chem. Chem. Phys. 2003, 5, 4351−4360.
2
069
dx.doi.org/10.1021/jp210457r | J. Phys. Chem. A 2012, 116, 2062−2070

The Journal of Physical Chemistry A
Article
(
8) van der Avert, P.; Podkolzin, S. G.; Manoilova, O.; de Winne, H.;
Weckhuysen, B. M. Chem.Eur. J. 2004, 10, 1637−1646.
9) Van der Avert, P.; Weckhuysen, B. M. Phys. Chem. Chem. Phys.
004, 6, 5256−5262.
10) Podkolzin, S. G.; Manoilova, O. V.; Weckhuysen, B. M. J. Phys.
Chem. B 2005, 109, 11634−11642.
11) Esquivel, M. R.; Bohe, A. E.; Pasquevich, D. M. Thermochim.
Acta 2003, 398, 81−91.
12) Esquivel, M. R.; Bohe,
Technol. 2005, 170, 304−309.
13) Esquivel, M. R.; Bohe,
(
2
(
(
́
(
́
A. E.; Pasquevich, D. M. J. Mater. Process.
́
A. E.; Pasquevich, D. M. Mater. Sci. Eng.,
(
A 2005, 397, 310−313.
(
(
14) Gaviría, J. P.; Bohe,
́
A. E. Metall. Trans. B 2009, 40, 45−53.
15) Julbe, A.; Chanaud, P.; Larbot, A.; Guizard, C.; Cot, L.;
Mirodatos, C.; Borges, H. Key Eng. Mater. 1991, 61/62, 65−70.
16) van der Heijden, A. W. A. M.; Belliere, V.; Espinosa Alonso, L.;
Daturi, M.; Manoilova, O. V.; Weckhuysen, B. M. J. Phys. Chem. B
005, 109, 23993−24001.
17) van der Heijden, A. W. A. M.; Mens, Ad J. M.; Bogerd, R.;
Weckhuysen, B. M. Catal. Lett. 2008, 122, 238−246.
18) Podkolzin, S. G.; Stangland, E. E.; Jones, M. E.; Peringer, E.;
Lercher, J. A. J. Am. Chem. Soc. 2007, 129 (9), 2569−2576.
19) Afanasiev, P.; Aouine, M.; Deranlot, C.; Epicier, T. Chem. Mater.
010, 22, 5411−5419.
20) Chen, Y.; Qian, Q.; Liu, X.; Xiao, L.; Chen, Q. Mater. Lett. 2010,
4, 6−8.
21) Marsal, A.; Dezanneau, G.; Cornet, A.; Morante, J. R. Sens.
Actuators, B 2003, 95, 266−270.
22) Marsal, A.; Rossinyol, E.; Bimbela, F.; Tellez, C.; Coronas, J.;
Vornet, A.; Morante, J. R. Sens. Actuators, B 2005, 109, 38−43.
23) Pasquevich, D. M.; Caneiro, A. M. Thermochim. Acta 1989, 156,
75−283.
24) Joint Committee for Powder Diffraction Standards, Powder
Diffraction File, International Center for Diffraction Data, Swarthmore, PA.
25) Neumann, A.; Walter, D. Thermochim. Acta 2006, 445, 200−
04.
26) Barin, I. Thermochemical Data of Pure Substances; VCH Verlags
Gesellschaft: Weinheim, Germany, 1993.
27) Baev, A. K.; Novikov, G. I. Russ. J. Inorg. Chem. 1965, 10 (11),
337−1341.
28) HSC 6.12, Chemistry for Windows, Outukumpu Research Oy,
Pori, Finland, 2006.
29) Peringer, E.; Tejuja, C.; Salzinger, M.; Lemonidou, A. A.;
Lercher, J. A. Appl. Catal., A 2008, 350, 178−185.
30) Gaviría, J. P.; Bohe, A. E. Thermochim. Acta 2010, 509,
00−110.
31) Gaviría, J. P.; Fouga, G. G.; Bohe,
(
̀
2
(
(
(
2
(
6
(
(
(
2
(
(
2
(
(
1
(
(
(
́
1
(
́
A. E. Thermochim. Acta 2011,
5
(
17, 24−33.
32) Hills, A. W. D. Metall. Trans. B 1978, 9, 121−128.
33) Ranz, W. E.; Marshall, W. R. Jr. Chem. Eng. Prog. 1952, 48 (3),
(
1
41−146.
34) Ranz, W. E.; Marshall, W. R. Jr. Chem. Eng. Prog. 1952, 48 (4),
73−180.
35) Geiger, G. H.; Poirier, D. R. Transport Phenomena in Metallurgy;
Addison-Wesley: Reading, MA, 1973.
36) Szekely, J.; Evans, J. W.; Sohn, H. Y. Gas-Solid Reactions;
Academic Press: New York, 1976.
(
1
(
(
(
(
37) Hakvoort, G. Thermochim. Acta 1994, 233, 63−73.
38) Pasquevich, D. M. Ph.D. Thesis, Facultad de Ciencias Exactas de
la Universidad Nacional de La Plata, La Plata, Argentina, 1990.
39) Pasquevich, D. M.; Amorebieta, V. Ber. Bunsen-Ges. Phys. Chem.
992, 96 (4), 534−541.
(
1
(
40) Avrami, M. J. Chem. Phys. 1939, 7 (12), 1103−1113.
41) Avrami, M. J. Chem. Phys. 1940, 8 (2), 212−224.
42) Avrami, M. J. Chem. Phys. 1941, 9 (2), 177−184.
43) Johnson, W. A.; Mehl, R. F. Trans. Am. Inst. Min. Metall. Eng.
(
(
(
1
(
2
939, 135, 416−427.
44) Kempen, A. T. W.; Sommer, F.; Mittemeijer, E. J. J. Mater. Sci.
002, 37 (2), 1321−1332.
2
070
dx.doi.org/10.1021/jp210457r | J. Phys. Chem. A 2012, 116, 2062−2070