Carbochlorination of samarium sesquioxide

Article abstract of DOI:10.1016/S0040-6031(03)00047-9

The chlorination of Sm₂O₃ in the presence of carbon using the gaseous mixture C12(g) + Ar(g) has been studied by thermogravimetry. The effects of both the temperature between 200 and 950 °C and the total gas flow rate between 2.1 and

Full text of DOI:10.1016/S0040-6031(03)00047-9

Thermochimica Acta 403 (2003) 207–218
Carbochlorination of samarium sesquioxide
M.R. Esquivela,∗, A.E. Bohéb, D.M. Pasquevicha,b
Comisión Nacional de Energ´ıa Atómica, Centro Atómico Bariloche, 8400 San Carlos de Bariloche, R´ıo Negro, Argentina
a
b
Consejo Nacional de Investigaciones Cient´ıficas y Técnicas, Centro Atómico Bariloche,
8400 San Carlos de Bariloche, R´ıo Negro, Argentina
Received 5 November 2002; accepted 19 December 2002
Abstract
The chlorination of Sm2O3 in the presence of carbon using the gaseous mixture Cl2(g) + Ar(g) has been studied by
thermogravimetry. The effects of both the temperature between 200 and 950 ◦C and the total gas flow rate between 2.1 and
7.9 l h−1 on the reaction rate were analyzed. The starting temperature of reaction, the stoichiometry and kinetic regimes
of the reaction were obtained. Reactants and products were analyzed by X-ray diffraction (XRD) and electronic dispersive
spectroscopy (EDS). The temporal evolution of the solid microstructure was followed by scanning electron microscopy (SEM).
© 2003 Elsevier Science B.V. All rights reserved.
Keywords: Samaria; Thermogravimetry; Gas–solid; Carbochlorination
1. Introduction
produce its chlorides is achieved in the presence of
carbon.
This compound plays two roles in the reaction:
Like the other lanthanide elements, Samarium pos-
sesses a stable trivalent state [1,2] which determines
its main stable oxide, Sm2O3 [3,4]. These compounds
are among the most stable in nature [4]. This feature,
along with the chemical similarities exhibited by all
the lanthanide sesquioxides [1,2,5], makes difficult ei-
ther their extraction from the mineral [5,6] or the sep-
aration from synthetic mixtures [5,6]. Since the direct
chlorination of the lanthanide sesquioxides is thermo-
dynamically feasible over 400 ◦C [7], it provides a
suitable method for the production of the chloride [5].
But the global reaction releases oxygen which may
react with the obtained chloride to form an oxychlo-
ride instead. Therefore, the chlorination of Sm2O3 to
(1) Thermodynamic: Carbon provides a low oxygen
potential atmosphere, diminishing the potential of
the oxide and favoring the formation of the chlo-
ride [8].
(2) Kinetic: Carbon is thought to favor the formation
of reaction intermediates producing an increment
in both the maximum reaction degree achieved at
a given time and an increase on the reaction rate
[8,9].
Changes on the carbon surface are observed if car-
bon is reacted in a given oxide–chlorine system [8,9].
No changes on carbon are seen if the direct chlorina-
tion between chlorine and oxide is achieved.
The chlorination of Sm2O3 in the presence of car-
bon to produce its chlorides represents an important
technological subject and a method of separation of
lanthanide oxides at high temperatures (1200 ◦C) has
∗
Corresponding author. Tel.: +54-29-4444-5397;
fax: +54-29-4444-5100.
E-mail address: esquivel@cab.cnea.gov.ar (M.R. Esquivel).
0040-6031/03/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0040-6031(03)00047-9

208
M.R. Esquivel et al. / Thermochimica Acta 403 (2003) 207–218
imental lines with those contained on PDF-1 (1996)
Nomenclature
diffusion coefficient (m2 s−1
Eap apparent activation energy (kJ mol−1
[13] using PC Identify program (PW1776). Masses
of these reactants were weighed and mixed mechani-
cally to obtain a mixture of Sm2O3–C 10.58 wt.% of
carbon.
The measurements were performed on a thermo-
gravimetric analyzer based on a Cahn 2000 electrobal-
ance adapted to work with corrosive atmospheres. The
system is described elsewhere [14]. The mass change
resolution of the system was of 10 ␮g under the ex-
perimental conditions [8,11,14]. Samples were placed
inside a quartz crucible suspended by a quartz fiber
inside a furnace. Isothermal and non-isothermal mea-
surements were made.
D
)
)
L
m0
characteristic dimension of the sample (m)
initial mass sample (mg)
ꢀm mass variation (mg)
ꢀM balance mass variation (mg)
N
P
r
R
t
molar flow of chlorine (mol s−1
pressure (kPa)
)
reaction rate (mol s−1
)
reaction rate (s−1
time (s)
)
T
W
temperature (◦C)
formula weight (kg mol−1
)
In the first ones, samples were heated for 1 h in
Ar atmosphere at the desired operation temperature.
After that, Cl2 was injected into the system reaching a
Greek symbols
α
β
ν
Sm2O3 reaction degree (dimensionless)
stoichiometric coefficient (dimensionless)
pCl = 30.3 kPa under a overall pressure of 101.3 kPa.
2
The total gas flow rate was varied between 2.1 and
7.9 l h−1
.
kinematic viscosity (m2 s−1
)
In the second ones, samples were heated from room
temperature to 950 ◦C at 2.6 ◦C min−1 at a pCl
=
2
been reported [10]. Despite that and the academic
importance, the effect of temperature and total gas
flow rate on the carbochlorination of Sm2O3 has not
been studied to the best of the author’s knowledge.
This situation has aimed the elaboration of the present
paper.
30.3 kPa in flowing Ar–Cl2 under a overall pressure
of 101.3 kPa. In both cases, mass changes were mea-
sured and apparent mass changes were corrected as
explained elsewhere [8,11,14]. Since SmCl3 is hy-
groscopic [1,15], auxiliary measurements were made
[6,7]. Samples obtained were isolated and handled in-
side a glove box [6,7].
2. Experimental
2.2. Expression of results
2.1. Materials and procedure
For non-isothermal runs, TG data is expressed as
ꢀM versus T, where ꢀM is the experimental mass
change observed and T is the temperature in degree
Celsius (◦C). For isothermal runs, TG data is ex-
pressed as the fractional oxide mass loss given by
α = ꢀm(Sm2O3)/m0(Sm2O3), where ꢀm(Sm2O3)
is the Sm2O3 mass loss and m0(Sm2O3) is the ini-
tial Sm2O3 mass. Sm2O3 consumption is related
to the observed mass change by ꢀm(Sm2O3) =
β ꢀM.
Gases used were Ar 99.999% purity (AGA, Ar-
gentina) and Cl2 99.8% (Indupa, Argentina). Solid re-
actants utilized were Sm2O3 and carbon. Both sub-
stances were previously characterized [6,11]. Carbon
was obtained from the calcination of sucrose (Fluka
Chemie AG) in inert atmosphere at 980 ◦C during 48 h
and sieved to a size of −200 to 275 mesh (ASTM) [11].
Carbon characteristics are given elsewhere [11]. Since
Sm2O3 absorbs either CO2 or H2O from atmosphere
[12], it was previously burned at 950 ◦C [3,4]. Sm2O3
possesses a particle distribution between 5 and 20 ␮m
as observed by scanning electron microscopy (SEM)
Where β and ꢀM stand for the different stoichiome-
tries involved and observed mass loss, respectively.
Therefore, the reaction degree is transformed to:
ꢀ
ꢁ
and a BET surface area of 6.571 0.0389 m2 g−1
Sm2O3 structure was verified by comparing the exper-
.
β ꢀM
α =
(1)
m0(Sm2O3)

M.R. Esquivel et al. / Thermochimica Acta 403 (2003) 207–218
209
The reaction rate (s−1) is then:
ꢀ
ꢁ
ꢂ
ꢃ ꢀ
ꢁ
da
dt
β
dM
dt
R =
= −
(2)
(3)
m0(Sm2O3)
The reaction rate (mol Cl2 s−1) is:
ꢀ
ꢁ
ꢂ
ꢃ
dn(Cl2)
dt
m0(Sm2O3)
W(Sm2O3)
r =
=
R
where n(Cl2) are moles of Cl2 and W(Sm2O3) is the
formula weight of Sm2O3.
3. Results and discussion
Fig. 1. Non-isothermal curve of the carbochlorination of Sm2O3.
Selected temperatures of analysis are indicated by arrows. XRD
and EDS techniques are indicated by hollow and full circles,
respectively. Thermodynamic calculations and SEM observations
are indicated by full and hollow squares, respectively.
3.1. Thermodynamics
In the 20–950 ◦C temperature range, SmCl2(s),
SmCl3(s) and SmOCl(s) [15–17] are the possible re-
action products of the carbochlorination of Sm2O3.
Liquid and gaseous chlorides are also probable be-
cause SmCl3 and SmCl2 have mp of 681 and 752 ◦C,
respectively [17]. These compounds can be produced
along with O2(g), CO(g) and CO2(g) during the
carbochlorination.
ing temperatures: 300, 550, 650, 750 and 950 ◦C. The
main reaction products are shown in Table 1.
3.2. Preliminary analysis
A thermodynamic analysis was performed [18] to
estimate the probable reaction products. Calculations
considering masses of chlorine and carbon in excess
over the stoichiometric consumption of Sm2O3 in a
molar ratio of 4/4/1 were performed at five represent-
Non-isothermal exploratory measurements were
made to determine both the reactivity with temper-
ature and the main products of the Sm2O3–C–Cl2
system. The thermogravimetric curve is displayed in
Fig. 1. Corrections to the curve were made as indicated
Table 1
Reaction products with carbon and Cl2 in excess (Sm2O3–C–Cl2 in a initial 4/4/1 molar ratio)
Compounds in equilibrium
300 ◦C
550 ◦C
650 ◦C
750 ◦C
950 ◦C
Condensed phase (mol)
C(s)
2.499
2.364
2.015
1.454
1.031
SmCl3(s)
1.934
1.409
1.084
1.000
1.000
SmCl2(s)
–
–
–
–
–
SmOCl(s)
Sm2O3(s)
SmCl2(l)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
SmCl3(l)
0.065
0.509
0.916
0.999
0.999
Gaseous phase (mol)
CO(g)
CO2(g)
0.001
1.449
1.000
–
–
–
0.271
1.364
1.000
–
–
–
0.968
1.015
1.000
–
–
–
2.090
0.455
1.000
–
–
–
2.937
0.031
1.000
–
–
–
Cl2(g)
SmCl3(g)
SmCl2(g)
O2(g)

210
M.R. Esquivel et al. / Thermochimica Acta 403 (2003) 207–218
elsewhere [14]. From 200 to 825 ◦C, the mass is incre-
mented as temperature is increased. Flat and steeped
zones are alternatively seen along the curve. Different
condensed products of reaction are formed in each
zone. Over 825 ◦C, a mass loss associated to either a
gasification reaction [7] or a vaporization process of
a condensed phase [7] is observed.
3.3. Effect of the temperature: products of reaction
and stoichiometry
To determine the identity of the compounds ob-
served in Fig. 1, isothermal experiments at five charac-
teristic temperatures, 300, 550, 650, 750 and 800 ◦C,
were made. These are indicated by solid arrows. The
samples were led to react at different reaction degrees
and analyzed by X-ray diffraction (XRD), electronic
dispersive spectroscopy (EDS) and SEM as indicated
in Fig. 1 by hollow circles, full circles and full squares,
respectively. Hollow squares correspond to the ther-
modynamic calculations indicated in Table 1.
Fig. 2. Diffraction patterns of the reaction products: (A) 300 ◦C;
(B) 550 ◦C; (C) reference pattern of SmOCl [19].
No diffraction lines other than SmOCl are observed.
A SEM image of the products at this temperature
is shown in Fig. 3. The plate-like particles grouped
into chunks correspond to SmOCl identified by EDS
microanalysis. Smooth steeped particles correspond
to carbon. These particles are labeled in the figure.
No attacks on carbon surface, such as channeling or
pitting are observed [8,9]. Mass balances calculated
from TG curves performed on the 200–375 ◦C range
are in agreement with the formation of solid SmOCl
◦
3.3.1. Analysis from 200 to 375 C
A XRD diffractogram of the products of reaction
at 300 ◦C is shown in Fig. 2A. For comparison, the
reference pattern of SmOCl [19] is shown in Fig. 2C.
Fig. 3. SEM image of the reaction products at 300 ◦C. Carbon surface presents no attacks. SmOCl, plate-like particles grouped in chunks,
are identified by EDS.

M.R. Esquivel et al. / Thermochimica Acta 403 (2003) 207–218
211
according to:
1
2
Sm2O3(s) + Cl2(g) → 2SmOCl(s) + O2(g)
(4)
Therefore, carbon is not involved on the reaction.
The stoichiometry of Eq. (4) corresponds to the direct
chlorination of Sm2O3 [20]. This system is analyzed
elsewhere [20].
◦
3.3.2. Analysis from 400 to 625 C
Five hundred and fifty degree Celsius was selected
as the temperature representative of this range. The
XRD diffractogram of the reaction products at that
temperature is shown in Fig. 2B. SmOCl [19] is the
reaction product observed. A SEM image detailing
the carbon surface is displayed in Fig. 4. The small
chunks resting on the carbon surface are SmOCl par-
ticles. The carbon surface does present chemical at-
tacks [8,9,21]. Then, its participation on the reaction
is evidenced [8,9,21]. This is in agreement with pre-
vious observations indicating that this type of carbon
is feasible of being oxidized over 400 ◦C [11,22]. The
mass balances performed on the TG curves in the
400–625 ◦C range are in agreement with the following
reaction:
Fig. 5. Effect of the temperature for 2 mg of Sm2O3 + C in the
400–625 ◦C temperature range.
The corresponding TG curves are shown in Fig. 5.
The carbochlorination at all temperatures is fully
achieved. The reaction rate is increased as temperature
is raised. Carbochlorination is fulfilled in the order
of 1 × 103 and 25 s at 400 and 625 ◦C, respectively.
As observed in Fig. 5, the TG curves become closer
to each other as temperature is incremented over
450 ◦C.
◦
3.3.3. Analysis from 650 to 725 C
1
Sm2O3(s) + C(s) + Cl2(g)
2
Fig. 6 is a SEM image of the reaction products at
650 ◦C. Pitting is observed at carbon surface [8,9,21]
1
→ 2SmOCl(s) + CO2(g)
(5)
2
Fig. 4. SEM image of the reaction products at 550 ◦C. Attacks on carbon surface are observed. Small particles are SmOCl identified by EDS.

212
M.R. Esquivel et al. / Thermochimica Acta 403 (2003) 207–218
Fig. 6. SEM image of the reaction products at 650 ◦C. Pitting on carbon surface is indicated by the letter “P”. SmCl3 identified by EDS
is also indicated.
and is indicated by the letter “P”. Particles of SmCl3
identified by EDS are also indicated by arrows. The
isothermal TG curves performed in this temperature
range are shown in Fig. 7. Unlike the lower tempera-
tures, the curves present significant discontinuities at-
tributed to two stages. The first one is quickly achieved
in the first 20 s of reaction at all temperatures. The
diffractogram showing the reaction products at this
stage at 650 ◦C is displayed in Fig. 8A. For compari-
son, the reference pattern of SmOCl [19] is displayed
in Fig. 8B The mass balances performed in this stage,
are in agreement with the formation of SmOCl ac-
cording to Eq. (5).
The second stage rate is slower than the first, i.e. the
first stage is fully achieved in 18 s and the second stage
is finished at 5.5 × 103 s at 700 ◦C. A diffractogram
of the products of reaction at this stage at 650 ◦C is
displayed in Fig. 8C. For comparison, the reference
pattern of SmCl3 [23] is displayed on Fig. 8D. By
comparing the experimental diffractograms and those
of the reference is inferred that the product of reaction
is a mixture of the SmOCl formed at the first stage
and SmCl3 produced in the second stage. Over 650 ◦C
and up to 700 ◦C, the global mass balances including
both the first and second stage are in agreement with
the following reaction:
3
Sm2O3(s) + C(s) + 3Cl2(g)
2
3
→ 2SmCl3(s, l) + CO2(g)
(6)
2
Mass balances performed at 725 ◦C indicates that
the reaction at this temperature presents a global sto-
ichiometry with values closer to both that of Eq. (6)
and that of the following:
Fig. 7. Effect of the temperature for 2 mg of Sm2O3 + C in
the 650–725 ◦C temperature range. Techniques used to analyze
samples at each stage are indicated by its symbol.
Sm2O3(s) + 3C(s) + 3Cl2(s)
→ 2SmCl3(s, l) + 3CO(g)
(7)

M.R. Esquivel et al. / Thermochimica Acta 403 (2003) 207–218
213
Fig. 8. Diffraction patterns of the reaction products at 650 ◦C; (A) 20 s of reaction; (B) reference pattern of SmOCl [19]; (C) 1.6 h of
reaction; (D) reference pattern of SmCl3 [23].
◦
3.3.4. Analysis from 750 to 950 C
A SEM image showing the reaction products at
750 ◦C is displayed on Fig. 10. A strong chemical at-
tack is clearly observed at the carbon surface where
both pitting [8,9,22] and channeling [8,9,22] are indi-
cated by arrows. Letters “P” and “C” stand for pitting
and channeling, respectively. The granular mass ob-
served behind the carbon is molten SmCl3 identified
by EDS. Mass balances performed on the TG curves
led to the conclusion that at temperatures between 750
and 850 ◦C the stoichiometry of the first stage is given
by Eq. (5).
The isothermal thermogravimetric curves corre-
sponding to this temperature range are displayed on
Fig. 9. Two successive stages are detected. Like the
previous temperature range analyzed, the first one is
quickly and fully achieved in the order of 20 s at all
temperatures. The second one is slower and fulfilled
in 150 s at 950 ◦C, the higher temperature.
The diffraction patterns of the reaction products are
not shown here, but the products of reaction are also
SmOCl and SmCl3 in the first and second stages, re-
spectively.
At temperatures higher than 850 ◦C, both the first
stage and SmOCl are detected but no exact stoichiom-
etry calculated by mass balances can be associated.
In all temperature ranges, mass balance indicated that
the global stoichiometry corresponds to Eq. (7).
3.4. The effect of the total gas flow rate
As usual in solid–gas reactions, the effect of the to-
tal gas flow rate on the reaction rate was analyzed [24].
Starvation [7,24] was studied by changing the flow
rate from 2.1 to 7.9 l h−1 at two different temperatures:
400 and 950 ◦C. The TG curves are shown in Fig. 11A
and B, respectively. No significant changes on the re-
action rate are noticed at 400 ◦C when the total flow
rate is increased. A slight increment is observed in both
stages when the flow rate is incremented at 950 ◦C.
Fig. 9. Effect of the temperature for 2 mg of Sm2O3 + C in the
750–950 ◦C temperature range.

214
M.R. Esquivel et al. / Thermochimica Acta 403 (2003) 207–218
Fig. 10. SEM image of the reaction products at 750 ◦C. Both pitting and channeling are observed on the carbon surface and indicated by
the letters “P” and “C”. SmCl3 identified by EDS is also indicated.
Therefore, starvation is thought to affect the reaction
at the higher temperatures [7,24].
tions [7,24] such as Ranz–Marshall [7,24]. The com-
parison is performed at two different reaction degrees
at four different temperatures in Table 2. The reac-
tion stoichiometries, D, and ν, the diffusion coeffi-
cient and the kinematic viscosity used to calculate the
Ranz–Marshall equation, are also shown [7,24]. The
experimental values at the low reaction degree become
closer to those of the theoretical equation as tempera-
ture is raised over 550 ◦C. But the experimental values
at the higher reaction degree, α = 0.5, are at least two
orders lower than the theoretical ones at temperatures
higher than 650 ◦C. Therefore, the first stage of reac-
tion is affected by convective mass transfer at temper-
atures over 550 ◦C. The second stage is not noticed to
be influenced, not controlled, by convection at tem-
peratures as high as 950 ◦C since its theoretical val-
ues are at least one order higher than the experimental
ones. Despite that, this stage is affected by starvation
as observed in Fig. 11B.
By comparing the effect of the total gas flow rate at
both temperatures (Fig. 11) it is inferred that the effect
of total gas flow rate is not uniform in all temperature
ranges involved. It is reasonable because starvation
is only one of the effects involved in external mass
transfer.
The other effect to be studied is convective mass
transfer [7,24]. It is done by comparing the experi-
mental reaction values to those of theoretical equa-
3.5. The activation energy and regimes of reaction
3.5.1. The formation of SmOCl
As observed on Table 1, SmCl3 is the most favored
product of the carbochlorination at all temperatures
[18]. Nevertheless, SmOCl is the only obtained prod-
uct at temperatures below 650 ◦C. It may be due to the
Fig. 11. Effect of the total gas flow rate on the carbochlorination
of 2 mg of Sm2O3 + C: (A) at 400 ◦C; (B) at 950 ◦C.

M.R. Esquivel et al. / Thermochimica Acta 403 (2003) 207–218
215
Table 2
Comparison between r (experimental) and N (theoretical) rates
T (◦C)
D (cm s−2
)
ν (cm s−2
)
N (mol Cl2 s−2
4.10 × 10−7
)
α
r stoichiometry
r (mol Cl2 s−2
)
550
0.68
0.61
0.2
0.5
Eq. (5)
Eq. (5)
3.56 × 10−7
2.33 × 10−7
650
750
850
950
0.83
1.00
1.18
1.21
0.74
0.89
1.04
1.21
4.57 × 10−7
5.28 × 10−7
5.90 × 10−7
7.50 × 10−7
0.2
0.45
Eq. (5)
Eq. (6)
3.76 × 10−7
7.37 × 10−10
0.2
0.5
Eq. (5)
Eq. (7)
5.25 × 10−7
2.00 × 10−9
0.2
0.5
Eq. (5)
Eq. (7)
5.86 × 10−7
8.04 × 10−9
0.2
0.5
Eq. (7)
Eq. (7)
7.32 × 10−7
1.78 × 10−8
Values of D and ν at various temperatures for pCl = 30.3 kPa. N values are calculated according to the Ranz–Marshall equation and
2
corrected as suggested [7,25,28]. In this equation L = 0.10. The experimental values of r are obtained for 2 mg of Sm2O3 + C under a
pCl = 30.3 kPa and a total gas flow rate of 7.9 l h−1 at each temperature. The equation stoichiometry used to calculate r is indicated in
2
the third column.
differences on the kinetics of formation of the oxy-
chloride and the chloride. It appears to be consistent
with the fact that the formation of SmCl3 is accom-
plished after that of SmOCl at 650 ◦C. Between 400
and 625 ◦C, the further chlorination of the SmOCl to
SmCl3 would demand much more time than those at
higher temperatures. Therefore, the product of the car-
bochlorination is SmOCl.
of the order of 20 10 kJ mol−1. Although only those
curves of α = 0.1, 0.3, 0.5 and 0.7 are shown, the
calculus were performed for α values between 0.1 and
0.9. This low Eap value suggests external mass transfer
control [25,26]. This conclusion is in agreement with
Figs. 7 and 9 where the reaction rate is practically
independent of the temperature effect.
The diminution of the Eap values from low tem-
peratures to high temperatures is consistent with an
Arrhenius behavior [27,28]. The Eap values found are
in agreement with a change in the controlling reac-
tion regime from chemical–mixed at low temperatures
From 650 to 850 ◦C, the formation of SmOCl
becomes the first step of the carbochlorination of
the Sm2O3. It is achieved with the stoichiometry of
Eq. (5). Over 850 ◦C, the SmOCl is also detected as
the first reaction product but no exact stoichiometry
is associated.
The calculation of the activation energy of the for-
mation of SmOCl in the 400–850 ◦C range is displayed
in Fig. 12. Between 400 and 525 ◦C, the lines are par-
allel at all reaction degrees indicating that the kinetic
regime is the same. Despite only those of 0.1, 0.3, 0.5,
0.7 are shown, the apparent activation energy values
(Eap) obtained were calculated at α values between 0.1
and 0.9. The Eap value found is of 120 5 kJ mol−1
.
This value suggests either mixed or chemical control
[25]. Although the direct chlorination of Sm2O3 is also
possible at this temperature range [20], its Eap value
between 375 and 950 ◦C is 12 7 kJ mol−1 [20], one
order lower than that found for the carbochlorination.
At 525 ◦C, the Eap lines present a break in Fig. 12.
The Eap values calculated between 525 and 850 ◦C are
Fig. 12. Plot of ln t vs. T−1 at various conversions for 2 mg of
Sm2O3+C between 400 and 850 ◦C. The stoichiometry considered
corresponds to Eq. (5). β coefficient is 7.131 for this equation.

216
M.R. Esquivel et al. / Thermochimica Acta 403 (2003) 207–218
Table 3
and achieving the global stoichiometry shown in
Eq. (7).
Activation energy values of the formation of SmOCl and SmCl3
Process
Temperature
range (◦C)
α range
Eap (kJ mol−1
)
As explained earlier, the stages proposed are con-
firmed by mass balances on the TG curves and XRD
identification of the products. Although SmOCl is de-
tected at temperatures higher than 850 ◦C and chem-
ical attacks are observed at carbon surface at low
reaction degrees, the processes influencing the reac-
tion are competing for control and no exact stoichiom-
etry can be assigned to the first step. Despite that, the
complete formation of SmCl3 is obtained in agreement
to Eq. (7).
The calculation of the Eap values assuming the
global stoichiometry of (7) and (6) for the lower
temperatures is displayed on Fig. 13. The values are
calculated from 675 to 950 ◦C. Only those of α = 0.4,
0.6, 0.7 and 0.9 are shown. The Eap values found
and their temperature and α ranges are resumed on
Table 3.
SmOCl formation 400–525
525–850
0.10–0.90 120
5
8
0.10–0.90
0.10–0.30
0.40–0.90 140
0.40–0.90 110
20
20
SmCl3 formation
675–950
675–800
825–950
8
5
10
to external mass transfer at high temperatures [25].
Although the direct chlorination of Sm2O3 presents
similar Eap values in the 525–850 ◦C range releasing
SmOCl as a common product [20], neither the mass
balances on the TG curves nor the attacks of the car-
bon surface confirm the influence of the direct chlori-
nation on the carbochlorination. A resume of the Eap
values found is displayed in Table 3.
3.5.2. The formation of SmCl3
3.5.2.1. Eap values at α between 0.1 and 0.3. At
these α values, the Eap value is 20 8 kJ mol−1. It cor-
responds to the formation of SmOCl in the first stage.
The Eap values are similar to that of the formation of
SmOCl at temperatures between 525 and 850 ◦C. This
low Eap value indicates external mass transfer con-
trol [25,26]. This conclusion is supported by Fig 11B
where the effect of starvation at low α values is ob-
served at 950 ◦C and by the comparison of Table 2
The formation of SmCl3 occurs as a second stage af-
ter that of SmOCl at temperatures higher than 650 ◦C.
At this temperature, no total conversion of SmOCl to
SmCl3 is achieved. A mixture of both products is ob-
tained, as observed in Fig. 8D.
At higher temperatures, between 675 and 950 ◦C,
the chloride formed would be liquid, since the mp
of SmCl3 is 681 ◦C [18]. This would favor the mass
transport of species what would make possible a more
rapid and complete formation of SmCl3 from SmOCl.
At lower temperatures, the solid SmCl3 formed over
the solid SmOCl might not favor the transport of gas
species and no complete reaction can be achieved.
Between 650 and 725 ◦C, the proposed two-stage
scheme is given by Eq. (5) and
2SmOCl(s) + C(s) + 2Cl2(g)
→ 2SmCl3(s, l) + CO2(g)
(8)
and achieving the global stoichiometry shown in
Eq. (6).At temperatures between 725 and 850 ◦C, the
first stage coincides with the stoichiometry indicated
on (5) but mass balances on the second stage are in
agreement with the following reaction:
Fig. 13. Plot of ln t vs. T−1 at various conversions for 2 mg of
Sm2O3 + C between 650 and 950 ◦C. The stoichiometries consid-
ered correspond to Eqs. (6) and (7). β coefficients are 2.71 and
2.37, respectively.
SmOCl(s) + C(s) + Cl2(g) → SmCl3(s, l) + CO(g)
(9)

M.R. Esquivel et al. / Thermochimica Acta 403 (2003) 207–218
217
where both experimental and theoretical reaction rate
values are of the same order [7,24].
Eq. (6) between 750 and 950 ◦C. The Eap value found
was 140 5 kJ mol−1 between 675 and 800 ◦C. At
higher temperatures, between 800 and 950 ◦C, the
Eap values of the second stage diminishes up to
110 10 kJ mol−1. Although both Eap values indicate
chemical–mixed controlling regime [25], the change
on the values is related to an increasing influence of
external mass transfer as temperature is incremented
on the mixed–chemical controlled step [27].
The complexity of this system is related to the
intrinsic kinetics of the carbochlorination reactions
[8,9,21,22]. These reactions release intermediates of
reaction which mechanism of formation are not com-
pletely established. The influence of the carbon con-
tent on the reaction rate will be the next step on the
study of the Sm2O3–C–Cl2 system.
3.5.2.2. Eap values at α between 0.4 and 0.9. The
Eap value found is 140 5 kJ mol−1 between 675 and
800 ◦C. The increment from 20 8 kJ mol−1 at low
reaction degrees to this value at high reaction degrees
is associated to a change in the mechanism of reac-
tion [25,29]. This change is related to the passage
from the first to the second stage of reaction. This
Eap value at high reaction degrees is in agreement
with a chemical–mixed control regime [25]. Never-
theless, the Eap value is diminished as temperature
is raised over 800 ◦C achieving a value of 110
10 kJ mol−1 as observed in Fig. 13. This value also
indicates chemical–mixed control regime [25]. There-
fore, the diminution of the Eap values observed is due
to an increasing influence of external mass transfer
on this chemical–mixed controlled stage [27]. This is
consistent with a typical Arrhenius behavior [27]. It is
supported by the conclusions obtained from Fig. 11B
where this stage is found to be influenced by starva-
tion [24] at the higher temperature and high α values.
Acknowledgements
The authors wish to thank to the Agencia Nacional
´
de Promoción Cientıfica y Tecnológica (ANPCyT) of
Argentina for the financial support of this work by
PICT 00-109984 project.
4. Conclusions
References
The findings in this paper have improved the scarce
knowledge on the carbochlorination of Sm2O3. The
Sm2O3–C–Cl2 is a complex system. Nevertheless, the
systematic TG measurements allowed to find the stoi-
chiometries of reaction in all the analyzed range. The
effect of total gas flow rate and convective mass trans-
fer were analyzed to estimate their influence on the
reaction rate.
[1] S. Cotton, Lanthanides and Actinides, Macmillan, London,
1991, p. 23.
[2] L. Mors, J. Less Common Met. 93 (1983) 301–321.
[3] T. Sato, Thermochim. Acta 148 (1989) 249–260.
[4] R.G. Haire, L. Eyring, Comparisons of the binary oxides,
Handbook of the Physics and Chemistry of the Rare Earths,
vol. 18, North-Holland, Amsterdam, 1994, Chapter 125,
p. 439.
The effect of the temperature on the chlorination
rate of Sm2O3(s)–C(s) mixture was evaluated from
200 to 950 ◦C. The starting temperature of carbochlo-
rination is obtained at 400 ◦C. Between this tempera-
ture and 625 ◦C, one stage is obtained. At higher tem-
peratures, the two consecutive stages of the reaction
are evidenced. The first one leads to the formation of
SmOCl and the second one to the formation of SmCl3.
The Eap value involved on the formation of SmOCl
is of 120 5 kJ mol−1 between 400 and 625 ◦C. At
[5] C.K. Gupta, N. Krishnamurthy, Int. Mater. Rev. 37 (5) (1992)
197.
[6] M.R. Esquivel, A.E. Bohé, D.M. Pasquevich, in: J.P. Barbosa,
A.J. Dutra, R. Melamed, R. Trindade (Eds.), Proceedings
of the sixth Southern Hemisphere Meeting on Mineral
Technology, vol. 2, 2001, pp. 468–474.
[7] M.R. Esquivel, A.E. Bohé, D.M. Pasquevich, Thermochim.
Acta, in press.
[8] J.J.A. Gamboa, A.E. Bohé, D.M. Pasquevich, Thermochim.
Acta 334 (1999) 131–139.
[9] D.M. Pasquevich, J.J.A. Gamboa, A. Caneiro, Thermochim.
Acta 209 (1992) 209–222.
[10] W. Brugger, E. Greinacher, JOM (1967) 32–35.
[11] J. Gonzalez, M. Ruiz, A. Bohé, D.M. Pasquevich, Carbon 37
(1999) 1979–1988.
higher temperatures, the Eap value is 20 8 kJ mol−1
.
The second stage produces SmCl3 with the stoi-
chiometries of Eq. (7) between 650 and 725 ◦C and

218
M.R. Esquivel et al. / Thermochimica Acta 403 (2003) 207–218
[12] S. Bernal, F.J. Botana, J. Pintado, R. Garc´ıa, J.M. Rodriguez
Izquierdo, J. Less Common Met. 110 (1985) 439–443.
[13] Joint Committee for Powder Diffraction Standards, Powder
Diffraction File, International Center for Diffraction Data,
Swarthmore, PA, 1996 (Card number 150813).
[21] F.C. Gennari, A.E. Bohé, D.M. Pasquevich, Thermochim.
Acta 3271 (1997) 1–9.
[22] J.R. Gonzalez, F.C. Gennari, M. Ruiz, A.E. Bohé, D.M.
Pasquevich, Trans. Inst. Miner. Metall. C 107 (1998) C130–
C138.
[14] D.M. Pasquevich, A. Caneiro, Thermochim. Acta 156 (2)
(1989) 275–283.
[15] D. Brown, Halides of the Lanthanides and Actinides, Wiley,
London, 1968, p. 156.
[16] H.A. Heick, Lanthanide and actinide halides, Handbook of the
Physics and Chemistry of Rare Earths, vol. 18, North-Holland,
Amsterdam, 1994 (Chapter 124).
[23] Joint Committee for Powder Diffraction Standards, Powder
Diffraction File, International Center for Diffraction Data,
Swarthmore, PA, 1996 (Card number 120789).
[24] A.W. Hills, Metall. Trans. B 9B (1978) 121.
[25] F. Habashi, Principles of Extractive Metallurgy, vol. 1,
Gordon and Breach, New York, 1969, Chapter 7, pp. 143–
144.
[17] S. Boghosian, G.N. Papatheodorou, Handbook of the Physics
and Chemistry of Rare Earths, vol. 23, North-Holland,
Amsterdam, 1996, Chapter 157, pp. 450–451.
[26] L.K. Doraiswamy, M.M. Sharma, Heterogeneous Reactions:
Analysis, Examples and Reactor Design, vol. 1, Wiley, New
York, 1984, Chapter 7, p. 143.
[18] A. Roine, Outokumpu HSC Chemistry for Windows,
93001-ORGT, Version 2.0, Outokumpu Research Oy
Information Service, Finland, 1994.
[27] L.K. Doraiswamy, M.M. Sharma, Heterogeneous Reactions:
Analysis, Examples and Reactor Design, vol. 1, Wiley, New
York, 1984, Chapter 7, p. 149.
[19] Joint Committee for Powder Diffraction Standards, Powder
Diffraction File, International Center for Diffraction Data,
Swarthmore, PA, 1996 (Card number 120790).
[28] J. Szekely, J. Evans, H.Y. Sohn, Gas–Solid Reactions,
Academic Press, New York, 1976, Chapter 4, pp. 109–
110.
[20] M.R. Esquivel, A.E Bohé, D.M. Pasquevich, Unpublished
work.
[29] O. Levenspiel, Chemical Reaction Engineering, Wiley, New
York, 1979, p. 32.