Mechanism of Molten-Salt-Controlled Thermite Reactions

Abstract

The present work was undertaken to study the chemistry and phase formation mechanism in the salt-controlled MoO3 + Mg + NaCl thermite reaction. It was found that the structure and phase formation mechanism in the studied system primarily depend on the salt content in the initial mixtures. In salt-poor mixtures, nucleation of product particles takes place in the molten magnesium, whereas under salt-rich conditions, products are mainly formed in molten sodium chloride. Analyses of combustion temperature profiles and product microstructures and thermal analysis of reacting mixtures suggested that the molybdenum oxide reacts with the salt at early stages of the process. The formed intermediate molybdenum oxychloride and sodium molybdate then react with magnesium, yielding Mo, MgO, and NaCl phases. The low value of the activation energy (50 kJ/mol) of the combustion process also suggests that gaseous (liquid) intermediates play an important role in the phase formation mechanism.

Full text

Published: August 29, 2011
r2011 American Chemical Society 10982 dx.doi.org/10.1021/ie2003544 |Ind. Eng. Chem. Res. 2011, 50, 10982–10988
ARTICLE
pubs.acs.org/IECR
Mechanism of Molten-Salt-Controlled Thermite Reactions
Khachatur V. Manukyan,*
,†,‡
Khachatur G. Kirakosyan,
†
Yeva G. Grigoryan,
†
Ofik M. Niazyan,
†
Armenuhi V. Yeghishyan,
†
Artavazd G. Kirakosyan,
†
and Suren L. Kharatyan
†,‡
†
Laboratory of Kinetics of SHS Processes, A. B. Nalbandyan Institute of Chemical Physics, National Academy of Sciences of the Republic
of Armenia (NAS RA), 5/2, P. Sevak Street, Yerevan 0014, Armenia
‡
Department of Inorganic Chemistry, Yerevan State University, 1, A. Manoogian Street, Yerevan 0025, Armenia
ABSTRACT: The present work was undertaken to study the chemistry and phase formation mechanism in the salt-controlled
MoO
3
+ Mg + NaCl thermite reaction. It was found that the structure and phase formation mechanism in the studied system
primarily depend on the salt content in the initial mixtures. In salt-poor mixtures, nucleation of product particles takes place in the
molten magnesium, whereas under salt-rich conditions, products are mainly formed in molten sodium chloride. Analyses of
combustion temperature profiles and product microstructures and thermal analysis of reacting mixtures suggested that the
molybdenum oxide reacts with the salt at early stages of the process. The formed intermediate molybdenum oxychloride and sodium
molybdate then react with magnesium, yielding Mo, MgO, and NaCl phases. The low value of the activation energy (50 kJ/mol) of
the combustion process also suggests that gaseous (liquid) intermediates play an important role in the phase formation mechanism.
1. INTRODUCTION
The term “thermite reaction”is used to describe a class of
reactions that involves a metal reacting with a metal or nonmetal
oxide. This form of oxidationreduction reaction can be written
in general form as
MþAO ¼MO þAþΔH
where M (typically Mg, Al, Ti, Zr, Zn, etc.) is a metal, A (MoO
3
,
WO
3
,Fe
2
O
3
,Cr
2
O
3
, TiO
2
, SiO
2
, CuO, etc.) is either a metal or a
nonmetal, MO and AO are their corresponding oxides, and ΔHis
the heat generated by the reaction.
14
Because of the large ex-
othermic effect, thermite reactions can generally be initiated locally
and become self-sustaining, a feature that makes their use ex-
tremely energy-efficient. Many thermite reactions yield a molten
product consisting of a heavier metallic phase and a lighter oxide
phase that can be separated by gravity and surface tension forces.
3
The latter makes these reactions potentially useful in a variety of
metallurgical applications.
47
More recently, thermite reactions
have become important in the synthesis of refractory ceramics,
8,9
composite materials,
913
and metal powders.
14,15
However, in-
tense gas evolution due to the decomposition/vaporization of initial
oxides and/or reducing elements coupled with high reaction
temperatures make it difficult to control the microstructure of
the obtained materials. Therefore, some approaches have been
adopted to soften violent reaction conditions and tune the mor-
phology of the products. One of the most recognized methods is the
application of so-called inert diluents. Addition of diluents to
thermite mixtures effectively reduces the combustion tempera-
ture and reaction rate because of the production of less heat and
the longer transport distances between reactants. A modified pro-
cess of conventional thermite reactions with halide salt additives is
known as molten salt-controlled combustion synthesis.
1315
The basic precursors for the process are known higher oxides
of transition metals such as WO
3
,Ta
2
O
5
,MoO
3
,andTiO
2
. Metallic
magnesium and zinc are frequently used as reduction agents.
14,15
For certain oxides (WO
3
,MoO
3
), sodium azide (NaN
3
), and
sodium boron hydride (NaBH
4
) can also be used as reducing
agents.
15
Recently, it was shown that one can use this method to
synthesize not only nanopowders of pure metals but also different
carbides (e.g., TiC,
16
WC
17
), silicides (e.g, MoSi
213
), and complex
compositions such as WCCo.
16
Two main factors are important in controlling the microstruc-
ture of the products in salt-controlled thermite reactions. The
first factor is mild reaction conditions, such as low temperatures,
which prevent intense grain growth. Second is the presence of a
molten “inert phase”in the reaction zone. Because of the heat
generated by self-sustaining reaction, the salt melts at about
800 °C, and further nucleation of product particles occurs in the
molten salt environment, which protects them from agglomera-
tion and grain growth. In all published works, however, the effect
of sodium chloride on the chemistry of combustion process was
not studied, and salt was always considered as only an inert
diluent. This does not rule out the possibility that, in the initial
stages of the reaction, metal oxides might react with salt yielding
various intermediates. For instance, it is well-documented
18,19
that MoO
3
reacts intensely with NaCl at 400800 °C, forming
MoO
2
Cl
2
and Na
2
MoO
4
. Early research
20,21
on the interaction
of transition metal oxides, namely, Ta
2
O
5
,WO
3
,MoO
3
,andTiO
2
,
with sodium chloride showed intense weight loss at 650950 °C,
which is conditioned by evaporation of sodium chloride, volatile
initial metal oxide, metal chlorides, and oxyclorides. The aqueous
solutions obtained after water treatment of the metal oxide
NaCl reaction products contains oxyanions and chloride species.
The concentrations of soluble metal species varied from several
Received: February 21, 2011
Accepted: August 23, 2011
Revised: August 2, 2011

10983 dx.doi.org/10.1021/ie2003544 |Ind. Eng. Chem. Res. 2011, 50, 10982–10988
Industrial & Engineering Chemistry Research ARTICLE
hundred parts per million to several thousand parts per million.
The soluble metal species in the solutions were in the form of
either metal chloride or MeO
2
and MeO
22
. Products formed
from the MoO
3
+ NaCl interaction showed relatively high
concentrations of soluble species (e.g., Na
2
MoO
4
). Significant
losses of tungsten conditioned by volatile tungsten oxychloride
vaporization during the WO
3
electroreduction in molten salt
media at 900 °C were also reported.
22
Therefore, studying the influence of sodium chloride on the
initial stages of the process in the salt-controlled thermite re-
actions is of special interest. The present work focused on studying
the effects of sodium chloride on the chemistry and phase forma-
tion mechanism of salt-controlled combustion reaction in the
MoO
3
+ Mg + NaCl system. This system was selected because
earlier research
1921
showed that molybdenum trioxide reacts with
sodium chloride more intensely than other transition metal oxides.
2. EXPERIMENTAL PROCEDURE
2.1. Combustion Experiments. The precursors used in this
study included MoO
3
(technical condition of manufacturing no.
6-09-4471-77, Pobedit Co., Vladikavkaz, Russia, purity 99.5%,
particle size <5 μm), magnesium (MPF-3, AVISMA, Verkhnaya
Salda, Russia, particle size 150300 μm), and sodium chloride
(high grade, Michailovskii Factory of Chemical Reagents, Michai-
losvk, Russia, purity 99.9%, particle size <20 μm) powders. The
precursors were thoroughly hand-mixed in the desired ratio in a
ceramic mortar for 1 h to ensure homogeneity of the reaction
medium, which was then uniaxially cold-pressed into cylindrical
pellets (diameter 20 mm, height 40 mm) at a pressure of 20 MPa
to relative densities in the range of 0.450.5. Two thermocouple
holes (2 mm in diameter, 10 mm deep) were drilled into each
specimen perpendicular to the cylinder axis at a spacing of ∼10 mm.
Combustion experiments were conducted in a laboratory constant-
pressure reactor (CPR-3 L, Sapphire Co., Abobyan, Armenia).
Before reaction initiation, the reaction chamber was sealed,
evacuated, and purged with argon (purity 99.8%, oxygen content
no more than 0.1%) for three cycles and finally filled with argon
to the desired pressure (1 MPa). A tungsten coil, positioned at
the upper surface of the sample, was electrically heated until the
reaction was initiated locally, after which the power was immedi-
ately turned off, as the reaction wave propagated along the sample.
The temperaturetime distributions (temperature profile)
at given points of the reacting samples were recorded by two
0.1-mm-diameter C-type thermocouples. To ensure the stability
of the measurements, the thermocouples first were sputtered by a
thin layer of boron nitride. The output signals of the thermo-
couples were transformed by a multichannel data acquisition
board at a rate of 2 kHz and were recorded on a computer. The
maximum combustion temperatures (T
c
) were determined from
the maxima of the temperature profiles. The average values of
combustion velocity (U
c
) were calculated from the physical dis-
tance between the thermocouples and the temporal distance be-
tween the signals of the thermocouples. All experimental data
points for the combustion parameters were determined as averages
of at least three measurements. The standard measurement
errors for T
c
and U
c
were (20 °C and 5%, respectively. Average
heating rates of initial reagents in the combustion wave were
determined from the temperaturetime profiles.
2.2. Characterization of Materials. After the combustion
process, reacted samples were kept in the reactor to complete
cooling. Typical purification operations for crushed solid products
include treatment with warm deionized water and dilute hydro-
chloric acid (5 wt %). After the water treatment the content of
metals (Mo, Mg, Na, Ca, etc.) in the obtained solution were
analyzed by ELAN-9000 ICP mass spectrometer. The drying of
purified material was performed in a vacuum furnace at a tem-
perature of 80 °Cfor4h.
The combustion products were studied by XRD analysis with
Cu Kαradiation (diffractometer DRON-3.0, Burevestnik, Russia).
XRD analyses of samples were performed at 25 kV and 10 μA.
Nova 230 and Hitachi S4800 field-emission scanning electron
microscopes were used to study the microstructure and compo-
sition of the produced materials by the energy-dispersive spec-
troscopy (EDS) method. The spatial resolution of the EDS
analysis was 1.52μm. The specific surface areas of the ob-
tained molybdenum powders were determined by the Brunauer
EmmettTeller (BET) method using nitrogen adsorption
(Gasometer, GKh-1). The oxygen content in the molybdenum
powder was determined with a LECO TC400 analyzer.
Differential thermal analysis (DTA) of initial reagents and
reacting mixtures was performed using a Derivatograph Q1500
instrument (MOM, Budapest, Hungary). DTA investigations
were conducted in argon flow (7 mL/s) at 20 °C/min heating
rate. Oxidation of the prepared molybdenum powder in air was
investigated by the DTA technique at a heating rate of 20 °C/min.
3. EXPERIMENTAL RESULTS
3.1. Thermodynamic Considerations. Before experiments,
ISMAN-THERMO software
23
was used to consider the thermo-
dynamics in the MoO
3
+ 3Mg + nNaCl (where nis the number of
moles of salt) system. This software is specially designed to
calculate adiabatic temperature (T
ad
) and product equilibrium
compositions in heterogeneous chemical processes. The inert gas
pressure in these calculations was kept at 1 MPa. The results
(Table 1) suggest that the calculated T
ad
value for the diluent-free
mixture (n= 0) is about 3500 °C. Increasing nto 12 decreases the
reaction temperature to 800 °C. The equilibrium products at T
ad
<
2000 °C consist of Mo, MgO, andNaCl, as well as small amounts of
different molybdenum oxychlorides. At elevated temperatures, the
amounts of molybdenum oxychlorides (MoOCl
2
and MoO
2
Cl
2
)
in the products are relatively higher (Table 1). Some amounts of
MoO
3
and MoO
2
are also present in the products.
Table 1. Adiabatic (T
ad
) and Measured Maximum (T
c
)
Combustion Temperatures, Amounts of Molybdenum Oxy-
chlorides Formed (C), Heating Rates of Reagents in the
Combustion Wave (V
h
), and Flame Propagation Velocities
(U
c
) for MoO
3
+ 3Mg + nNaCl Mixtures
n
(mol)
T
ad
(°C)
C(MoO
x
Cl
y
)
(mol)
T
c
(°C)
V
(°C/s)
U
c
(cm/s)
0 3500 0  
2 2260 0.025 2050 1500 0.33
3 1970 0.04 1760 350 0.15
4 1970 0.05 1330 50 0.07
4.5 1880 0 930 35 0.04
5 1780 0 770 20 0.03
6 1600 0 no combustion
8 1270 0
10 1030 0
12 800 0

10984 dx.doi.org/10.1021/ie2003544 |Ind. Eng. Chem. Res. 2011, 50, 10982–10988
Industrial & Engineering Chemistry Research ARTICLE
Equilibrium concentrations of the products for the MoO
3
+NaCl
mixture as a function of temperature were calculated as well. Figure 1
shows that, at T<770°C, no reaction occurred between the salt and
molybdenum trioxide. In the 770 < T<1200°Crange,liquid
Na
2
MoO
4
and gaseous MoO
2
Cl
2
were the main products of the
reaction between sodium chloride and molybdenum oxide. Thus,
these calculations predicted that, for molten-salt-assisted combustion
reactions, sodium chloride cannot be considered as an inert medium.
3.2. Combustion of MoO
3
+ Mg + NaCl Mixtures. Fol-
lowing the thermodynamic analysis, combustion processes in the
MoO
3
+ 3Mg + nNaCl mixtures were studied under 1 MPa inert
gas pressure. Combustion temperature profiles for these mix-
tures are shown in Figure 2. Inspection of all of the data suggests
(Table 1 and Figure 2) that the maximum combustion tempera-
ture (T
c
) decreased dramatically as the salt content grew. The salt
amount was found to have a critical point (n= 5) above which
the combustion wave did not propagate throughout the sample.
The measured maximum temperature at this point was about
770 °C. Note that the experimentally measured values for the
combustion temperature were significantly lower than the ther-
modynamically calculated adiabatic temperatures (Table 1). This
disagreement is a result of heat losses that usually occur during
the combustion process. In all cases, the combustion wave prop-
agated within the sample in the steady-state regime.
Another combustion feature that is significantly affected by the
salt amount is the heating rate (V,°C/s) of the reagent in the
reacting zone, determined from the temperature profiles
(Table 1). This parameter for n= 2 is about 1500 °C/s. For
n=3,V= 350 °C/s. Unusually low heating rates, from 20 to
50 °C/s, for combustion synthesis processes were observed at
high n. The temperature profiles (Figure 2) recorded for most n
values contained constant-temperature plateaus at 450510,
650, and 790820 °C. As can be seen, these plateaus were
well-defined, especially at low heating rates (high nvalues).
The salt concentration strongly influences the flame propagat-
ing velocity (U
c
) as well. U
c
fell by more than a factor of 10,
from 0.33 to 0.03 cm/s, as the salt content increased to 5 mol
(Table 1). U
c
exhibited a dramatic dependence on T
c
, which
provides a basis for the determination of the apparent activation
energy of the combustion process. The relationship correlating
combustion temperature, flame velocity, and activation energy is
expressed by the simplified equation
24
ln Uc
Tc

¼constant Eeff
2RTc
Figure 1. Calculated equilibrium compositions of products for the
MoO
3
+ NaCl mixture.
Figure 2. Temperaturetime distributions at combustion of MoO
3
+ 3Mg + nNaCl mixtures for n= (a) 2, (b) 3, (c) 4, (d) 4.5, and (e) 5.

10985 dx.doi.org/10.1021/ie2003544 |Ind. Eng. Chem. Res. 2011, 50, 10982–10988
Industrial & Engineering Chemistry Research ARTICLE
where E
eff
is the effective value of activation energy of the re-
action responsible for the combustion wave propagation rate and
Ris the universal gas constant. As a result, the slope of a plot of
ln(U
c
/T
c
) versus 1/T
c
can provide the effective value of the
activation energy of the process, as depicted in Figure 3. From the
slope of the best-fit line of all of the data in Figure 3, a value for
the activation energy of 50 kJ/mol was obtained.
3.3. Characterization of Combustion Products. XRD anal-
ysis showed that materials obtained from the MoO
3
+ 3Mg +
nNaCl mixtures always consisted of Mo, MgO, and NaCl. Figure 4
shows XRD patterns of products formed at n= 4 before and after
purification. The XRD pattern of the leached solid material
contains only the diffraction lines of molybdenum.
The molybdenum content in the solutions obtained by water
extraction of the reacted materials varied from 600 to 800 mg/L.
This result suggests that the reacted sample contained water-
soluble molybdenum species. The calculated amounts of water-
soluble molybdenum in the reacted samples varied from 1.25 to
1.45 wt %.
Most of the reacted samples contained defined zones formed
from the solidification of molten salt, as shown in Figure 5a.
Molybdenum, magnesium, and oxygen identified in these zones
by microanalysis suggest that at least some portion of the products
formed in the molten salt medium. Note that such zones were
most frequently observed for the salt-rich samples. The other
type of morphology observed in salt-poor samples is shown in the
Figure 5b. In the magnified image (Figure 5c) of the same area,
white particles with sizes of 0.31μm represent molybdenum.
It can be seen that these particles predominantly formed at the
grain boundaries of several-micrometer-sized magnesia crystals.
The combustion products also contained whiskers with dia-
meters of 0.51μm and lengths on the order of tens of
micrometers (Figure 5d). The whiskers consisted of molybde-
num, sodium, chlorine, magnesium, and oxygen, as estimated by
microanalysis. The high-magnification micrograph shows the
whiskers covered by particles with dimensions of 0.050.3 μm.
The morphology of the purified product obtained at n=4is
presented in Figure 5e,f. Here, two characteristic fractions of
particles can be clearly distinguished. The sizes of the particles in
the first fraction vary from 0.3 to 1 μm, whereas the second
fraction contains fine particles with dimensions of 0.050.3 μm.
Note that the purified material did not contain whiskers.
The specific surface area of molybdenum powder was mea-
sured to be 1.35 m
2
/g. The oxygen content in the molybdenum
powder was about 0.2 wt %. The oxidation onset of the powder in
ambient air, as determined by the DTA method, was 400 °C.
3.4. DTA Investigation of MoO
3
+ Mg + NaCl Mixtures. To
track the dynamics of phase formation in the studied system, the
thermal behaviors of individual reagents and their various com-
binations were studied by DTA as well. The DTA trace for MoO
3
shows two endotherms (Figure 6). The first endotherm, ob-
served at 400450 °C, corresponds to the molybdenium trioxide
αfβpolymorphic trasformation.
25
The second endotherm
starting at 750 °C corresponds to the sublimation of MoO
3
,as
TG curve displays intense weight loss. Figure 6 also displays DTA
traces for Mg and NaCl. The expressed endotherms at ∼650 °C
(for Mg) and ∼800 °C (for NaCl) correspond to the melting
points of the reagents. Sublimation of melted magnesium and salt
started at 770 and 880 °C, respectively.
The behaviors of the 3Mg + 5NaCl, MoO
3
+ 3Mg, and MoO
3
+
5NaCl binary mixtures were also studied by DTA. The DTA
trace for the 3Mg + 5NaCl mixture shows two endothermic
effects coinciding with melting of the individual reagents (Figure 7).
Intense exothermic reaction followed by Mg melting for binary
MoO
3
+ 3Mg mixture corresponds to Mg reduction of molybdnum
trioxide. The exothermic effect for the binary MoO
3
+5NaCl
mixture accompanied by weight loss of the reacting sample at
450650 °C (Figure 7) is conditioned by the evaporation of the
formed MoO
2
Cl
2
.
13,14
Note that the XRD pattern of rapidly
quenched material recorded after this experiment showed dif-
fraction lines for NaCl, MoO
3
, and Na
2
MoO
4
(MoO
3
)
x
com-
pounds (Figure 8).
Finally, the results of DTA for the MoO
3
+ 3Mg + 5NaCl
mixture are displayed in Figure 9. Here, melting of the reducer
coincides with the start of the intense exothermic reduction of
MoO
3
by Mg. The DTA trace shows that the exothermic reaction
and melting of sodium clroride are slightly overlapped. Most
likely, the exothermic MoO
3
+ NaCl reaction was not detected in
this DTA curve because of reletively high concentration of sodium
chloride. Nevertheless, some small weight loss was recorded in the
TG curve at 470530 °C.
4. DISCUSSION
To date, numerous investigations of salt-controlled thermite
reactions have been performed using different metals and metal
oxides, studying burning rates, types and amounts of salt used,
influences of particle size and gas pressure, and so on. However,
the mechanisms of salt-controlled thermite reactions are still far
from being completely understood.
Figure 3. ln(U
c
/T
c
)1/T
c
plot for calculation of the effective
activation energy.
Figure 4. XRD patterns of the (a) reaction products of MoO
3
+ 3Mg +
4NaCl mixture and (b) purified metal.

10986 dx.doi.org/10.1021/ie2003544 |Ind. Eng. Chem. Res. 2011, 50, 10982–10988
Industrial & Engineering Chemistry Research ARTICLE
Addition of salts of alkali metals (e.g., NaF, KF, NaCl, and KCl),
alkaline earth metals (e.g., CaF
3
, MgF
2
), and cryolite (NaAlF
6
)
can increase the combustion rate of thermite mixtures with
aluminum reducers.
2
The highest combustion rate has been
found for compositions containing aluminum fluoride and
cryolite.
2,26
It has been proposed that such salts reduce the
temperature at which the reaction between the oxide and the
aluminum commences. The oxide film on the aluminum particle,
which acts as a barrier to the interaction, can be disintegrated by
fluorides at temperatures significantly lower than the ignition
temperature of the thermite, and consequently, the ignition
temperature of the thermite mixture with salt addition is notably
reduced.
Alkali metal chlorides (e.g., NaCl) are mainly used as control-
ling agents in salt-controlled thermite reactions with magnesium
as the reducer.
13,1517
In such processes, the heat generated by
the self-sustaining reaction causes the salt to melt at about 800 °C.
Further nucleation of product particles occurs in the molten salt,
which protects them from agglomeration and growth. In all cases,
salt addition significantly reduces the combustion rate and max-
imum combustion temperature in the Mg-thermite reactions.
Therefore, the salt was always considered as an inert medium,
Figure 5. Microstructure of the products obtained: (a) product of the MoO
3
+ 3Mg + 4NaCl mixture, (b,c) product of the MoO
3
+ 3Mg + 2NaCl
mixture, (d) whiskers formed as the product of the MoO
3
+ 3Mg + 4NaCl, (e,f) purified Mo particles of the initial MoO
3
+ 3Mg + 4NaCl mixture.

10987 dx.doi.org/10.1021/ie2003544 |Ind. Eng. Chem. Res. 2011, 50, 10982–10988
Industrial & Engineering Chemistry Research ARTICLE
and the effects of the salt on the reaction mechanism have not yet
been considered. The thermodynamic analysis in the present
study predicts that, under combustion mode, MoO
3
reacts with
NaCl, yielding Na
2
MoO
4
and MoO
2
Cl
2
.
The obtained experimental results suggest that the reaction
parameters in the MoO
3
+ Mg + NaCl system depend strongly
on the salt content. The process temperature can be reduced to
800 °C as the salt content in the reacting mixture is increased.
The salt amount also dramatically affects the heating rate of re-
agents in the combustion front. Temperature profile analysis
enables the observation of well-defined plateaus at 480550,
650, and 800 °C. Thermal analysis of the precursors and their
mixtures helps clearly identify these temperature plateaus. Parti-
cularly at 450650 °C, molybdenum oxide reacts exothermically
with salt according to the reaction
2MoO3ðsÞþ2NaClðsÞ¼MoO2Cl2ðgÞþNa2MoO4ðs, lÞ
ð1Þ
The molybdenum detected in the water extracts of reacted
samples and XRD analysis of quenched samples also confirm the
mechanism of reaction 1. Temperature profile analyses and DTA
investigations suggest that the following reaction starts just after
melting of the reducer
MoO3ðsÞþ3MgðlÞ¼MoðsÞþ3MgOðsÞð2Þ
Microstructure analysis shows that, under salt-poor condi-
tions, the nucleation of product particles mainly takes place in the
molten magnesium. It is obvious that molybdenum particles are
separated from the melt upon cooling and appear at the grain
boundaries of the formed magnesia crystals. In salt-rich mixtures,
the nucleation of product particles mainly takes place at molten
NaCl media. Whiskers formed by a vaporliquidsolid mechanism
indicate that the following reaction might occur during flame
propagation
Na2MoO4ðs, lÞþMoO2Cl2ðgÞþ6Mgðl, gÞ
¼2MoðsÞþ6MgOðsÞþ2NaClðg, lÞð3Þ
It is assumed that fine Mo particles (0.050.3 μm) in the
purified product (see Figure 5f) are mainly formed by reaction 3.
The low activation energy (50 kJ/mol) of the combustion
reaction might provide additional evidence of the important role
of gaseous (liquid) intermediates in phase formation processes.
5. CONCLUSIONS
The thermodynamic analysis and experimental results of this
work show that the main combustion parameters (temperature,
Figure 6. DTA and TG curves of initial precursors: (1,1*) MoO
3
, (2,2*)
NaCl, and (3,3*) Mg.
Figure 7. DTA analyses of (1,1*) MoO
3
+ 3Mg, (2,2*) 3Mg + 5NaCl,
and (3,3*) MoO
3
+ 5NaCl binary mixtures.
Figure 8. XRD pattern of the queched material obtained by DTA of the
MoO
3
+ 5NaCl mixture.
Figure 9. DTA results for the MoO
3
+ 3Mg + 5NaCl mixture.

10988 dx.doi.org/10.1021/ie2003544 |Ind. Eng. Chem. Res. 2011, 50, 10982–10988
Industrial & Engineering Chemistry Research ARTICLE
velocity, heating rate) in the combustion of MoO
3
+ Mg + NaCl
mixtures depend strongly on the salt content. It was shown that
the structure of the combustion products also depends on the salt
content in the initial mixtures. In salt-poor mixtures, nucleation
of the product particles takes place in the molten magnesium,
whereas under salt-rich conditions products form in molten
sodium chloride. The results obtained suggest that the 2MoO
3
(s) +
2NaCl(s) = MoO
2
Cl
2
(g) + Na
2
MoO
4
(s,l) reaction proceeds at
early stages of the combustion process. The formed intermedi-
ates molybdenum oxychloride and sodium molybdate then react
with magnesium, yielding Mo, MgO, and NaCl phases. The cal-
culated low activation energy of the process also confirms
the dominant role of gaseous (liquid) intermediates in the phase
formation mechanism.
’AUTHOR INFORMATION
Corresponding Author
*E-mail: khachat@ichph.sci.am. Tel.: 00374 10 28-16-10. Fax:
00374 10 28-16-34. Address: 1 Manoogian Street, Yerevan 0025,
Armenia.
’ACKNOWLEDGMENT
The authors acknowledge the financial support of the State
Committee of Science of the Republic of Armenia (Project 354).
’REFERENCES
(1) Merzhanov, A. G.; Yukhvid, V. I.; Borovinskaya, I. P. Self-
Propagating High-Temperature Synthesis of Cast Refractory Inorganic
Compounds. Dokl. Akad. Nauk USSR 1980,255, 120.
(2) Wang, L. L.; Munir, Z. A.; Maximov, Y. M. Thermite Reactions:
Their Utilization in the Synthesis and Processing of Materials. J. Mater.
Sci. 1993,28, 3693.
(3) Mukasyan, A.; Lau, C.; Varma, A. Influence of Gravity on
Combustion Synthesis of Advanced Materials. AIAA J. 2005,43, 225.
(4) Odawara, O.; Mori, K. Thermite Reaction in a Short-Time
Microgravity Environment. J. Mater. Synth. Process. 1993,1, 203.
(5) Mei, J.; Halldearn, R. D.; Xiao, P. Mechanisms of the Aluminium
Iron Oxide Thermite Reaction. Scr. Mater. 1999,41, 541.
(6) Dur~aes, L.; Costa, B. F. O.; Santos, R.; Correia, A.; Campos, J.;
Portugal, A. Fe
2
O
3
/Aluminum Thermite Reaction Intermediate and
Final Products Characterization. Mater. Sci. Eng. A 2007,465, 199.
(7) Bae, J. H.; Kim, D. K.; Jeong, T. H.; Kim, H. J. Crystallization of
Amorphous Si Thin Films by the Reaction of MoO
3
/Al Nanoengi-
neered Thermite. Thin Solid Films 2010,518, 6205.
(8) Sundaram, V.; Logan, K. V.; Speyer, R. F. Reaction Path in the
Magnesium Thermite Reaction to Synthesize Titanium Diboride. J. Mater.
Res. 1997,12, 2657.
(9) Yeh, C. L.; Li, R. F. Formation of TiB
2
Al
2
O
3
and NbB
2
Al
2
O
3
Composites by Combustion Synthesis Involving Thermite Reactions.
Chem. Eng. J. 2009,147, 405.
(10) Zhu, H. X.; Abbaschian, R. In-Situ Processing of NiAlAlumina
Composites by Thermite Reaction. Mater. Sci. Eng. A 2000,282,1.
(11) Xi, W.-J.; Li, N.; Zhang, T.; Zhu, W.-L.; Guo, H.-Z. Thermite
Reaction Synthesis of Nano-Sized NiAl Reinforced FeNiCrTiC
Composite Coating. J. Alloys Compd. 2010,504, 414.
(12) Cervantes, O. G.; Kuntz, J. D.; Gash, A. E.; Munir, Z. A.
Activation Energy of TantalumTungsten Oxide Thermite Reactions.
Combust. Flame 2011,158, 117.
(13) Manukyan, Kh.V.;Aydinyan, S. V.;Kirakosyan, Kh.G.; Kharatyan,
S. L.; Blugan, G.; M€uller, U.; Kuebler, J. Molten Salt-Assisted Combustion
Synthesis and Characterization of MoSi
2
and MoSi
2
Si
3
N
4
Composite
Powders. Chem. Eng. J. 2008,143, 331.
(14) Mukasyan, A. S., Martirosyan, K., Eds. Combustion of Heterogeneous
Systems: Fundamentals and Applications for Material Synthesis; Transworld
Research Network: Kerala, India, 2007.
(15) Won, C.; Nersisyan, H.; Won, H.; Lee, J. Refractory Metal
Nanopowders: Synthesis and Characterization. Curr. Opin. Solid State
Mater. Sci. 2010,14, 53.
(16) Nersisyan, H.; Lee, J.; Won, C. Combustion of TiO
2
Mg and
TiO
2
MgC Systems in the Presence of NaCl to Synthesize Nano-
crystalline Ti and TiC Powders. Mater. Res. Bull. 2003,38, 1135.
(17) Nersisyan, H.; Won, H.; Won, C.; Lee, J. Study of the Com-
bustion Synthesis Process of Nanostructured WC and WCCo. Mater.
Chem. Phys. 2005,94, 153.
(18) Zelikman, A.; Gorovits, N. Reduction of Molybdenum Oxide
by Hydrogen in the Presence of Salts. J. Gen. Chem. USSR 1954,24, 1879.
(19) Volkovicha, V.; Griffiths, T.; Thied, R.; Lewin, B. Behavior of
Molybdenum in Pyrochemical Reprocessing: A Spectroscopic Study of
the Chlorination of Molybdenum and Its Oxides in Chloride Melts. J. Nucl.
Mater. 2003,323, 93.
(20) Mobin, M. High Temperature Interaction of Metal Oxides and
Carbides with Ionic Salts. Sci. Eng. Compos. Mater. 1999,8, 257.
(21) Mobin, M.; Malik, A.; Ahmad, S. J High temperature interac-
tions of metal oxides with NaCl. Less Common Metals 1990,160,1.
(22) Erdogan, M.; Krakaya, I. Electrochemical Reduction of Tung-
sten Compounds to Produce Tungsten Powder. Metall. Mater. Trans. B
2010,41, 798.
(23) Shiryaev, A. Thermodynamics of SHS Processes: An Advanced
Approach. Int. J. SHS 1995,4, 351.
(24) Merzhanov, A. G.; Khaikin, B. I. Theory of combustion waves in
homogeneous media. Prog. Energy Combust. Sci. 1988,14 (1), 1–98.
(25) Patil, R.; Uplane, M.; Patil, P. Electrosynthesis of Electrochromic
Molybdenum Oxide Thin Films with Rod-Like Features. Int. J. Electrochem.
Sci. 2008,3, 259.
(26) Dubrovin, A. S.; Kuznetsov, V. L.; Ezikov, V. I.; Chirkov, N. A.;
Rusakov, L. N. Russ. Metall. (Engl. Transl.) 1968,5, 56.) 56.