Self-propagating high-temperature synthesis of nonstoichiometric wüstite

Article abstract of DOI:10.1016/j.jallcom.2011.12.057

This paper describes the self-propagating high-temperature synthesis (SHS) of nonstoichiometric Fe_xO (x = 0.833-1), with particular focus on the effects of nonstoichiometric Fe content and diluent addition on the phase of the SHS product. In th

Full text of DOI:10.1016/j.jallcom.2011.12.057

Journal of Alloys and Compounds 520 (2012) 59–64
Contents lists available at SciVerse ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jallcom
Self-propagating high-temperature synthesis of nonstoichiometric wüstite
a
b
b,∗
Maki Hiramoto , Noriyuki Okinaka , Tomohiro Akiyama
a
Graduate School of Engineering, Hokkaido University, Sapporo 060-8628, Japan
Center for Advanced Research of Energy and Materials, Faculty of Engineering, Hokkaido University, Sapporo 060-8628, Japan
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 16 November 2011
Received in revised form
This paper describes the self-propagating high-temperature synthesis (SHS) of nonstoichiometric Fe O
x
(
x = 0.833–1), with particular focus on the effects of nonstoichiometric Fe content and diluent addition
on the phase of the SHS product. In the SHS process, the raw materials Fe, NaClO4 (oxidizer), and NaCl
diluent) were thoroughly mixed in the desired ratio by ball milling, and the lower surfaces of the disk-
1
3 December 2011
(
Accepted 14 December 2011
shaped green compacts were subsequently electrically ignited to produce FexO through the propagation
of the sustainable exothermic reaction. X-ray diffraction analysis showed that the SHS products com-
prised double phases of FexO and Fe3O4. The peaks of products with 0.947 ≤ x ≤ 1.00 shifted to lower
angles in comparison to those of the product with x = 0.833 attributed to the lattice parameter distortion
of the crystal structure because of the Fe defects. In the presence of the NaCl diluent, the raw materi-
als were converted to high-purity FexO powders during the SHS process. Without the NaCl diluent, the
lattice parameter of SHS Fe0.947O corresponded to the theoretical lattice parameter. Nonstoichiometric
Available online 11 January 2012
Keywords:
Self-propagating high-temperature
synthesis
Nonstoichiometric oxides
Wüstite
Point defects
compounds of Fe O (0.942 ≤ x ≤ 0.952) were obtained through SHS without additional external heating.
x
Crystal structure
©
2011 Elsevier B.V. All rights reserved.
1
. Introduction
partial pressure, whereas the nonstoichiometric oxides CoxO and
NixO have a limited range of possible nonstoichiometric states.
The large nonstoichiometric range of FexO results in complicated
electronic and ionic disorder. In addition, the defect and transport
properties of FexO have been the subject of extensive experimental
studies and theoretical considerations.
In addition to FexO, other iron oxide phases are also of tech-
nological importance and have found a number of applications
including use as recording materials, electrophotographic devel-
opment agents [5], in catalysis, and as gas sensors [6]. Intensive
investigations of the thermodynamic and structural properties as
well as the electrical properties of the FexO system have been exe-
cuted. These studies suggest that defect ordering within the FexO
phase may be complex [7]. The unusual electronic properties of
FexO including the fact that it is a semiconductor whose carrier type
changes from p to n type around x = 0.92 [8] make FexO interesting
in its own right.
Nonstoichiometric transition metal oxides are convention-
ally produced based on equilibrium. However, the conventional
method suffers from the limitations of being time- and energy-
consuming, because the raw materials must be kept at a high
temperature for an extended period in order to reach the equilib-
rium compositions of the products [3].
In order to obviate the drawbacks associated with the
conventional production of FexO, we applied self-propagating high-
temperature synthesis (SHS) as an alternative production method
[9]. This method harnesses the thermal energy generated by an
exothermic chemical reaction between metal powders and an
It is well known that nonstoichiometric compounds have unique
magnetic, electrical, thermal, optical, and mechanical properties
owing to the effect of lattice vibrations [1,2]. Nonstoichiomet-
ric transition metal oxides such as TiOx, VOx, FexO, MnxO, CoxO,
and NixO crystallize in the simple rock salt structure with vary-
ing degrees of intrinsic defects, typically present on the cationic
sublattice. The aggregation of these intrinsic defects confers semi-
conductor properties on these nonstoichiometric oxides even
without the addition of rare metal dopants.
Wüstite (FexO) is one of the phases of iron oxide and is a
well-known nonstoichiometric compound. The intriguing phys-
ical and chemical properties of FexO have garnered unflagging
interest for more than 90 years on the basis of both practical
and cognitive considerations. For instance, FexO is a major con-
stituent of the oxide-scale formed on iron and iron-based alloys,
and plays an important role in the reduction processes of iron ores.
◦
FexO is stable only at temperatures above 570 C at atmospheric
pressure; below 570 C, FexO disproportionates into Fe O and Fe
◦
3
4
[
3]. FexO can also exist as a metastable phase at room tempera-
ture by quenching from high temperatures where FexO exists in
the equilibrium state [4]. FexO exhibits relatively large deviation
from stoichiometry depending on the temperature and the oxygen
∗ Corresponding author. Tel.: +81 11 706 6842; fax: +81 11 726 0731.
E-mail address: takiyama@eng.hokudai.ac.jp (T. Akiyama).
0
925-8388/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2011.12.057

6
0
M. Hiramoto et al. / Journal of Alloys and Compounds 520 (2012) 59–64
Table 1
Properties of raw materials used for synthesizing FexO (x = 0.833–1) by SHS.
Raw materials Source
Particle diameter (␮m) Purity (mass%)
Fe
NaClO4
NaCl
Kojundo Chemical 3–5
Aldrich >10
Kojundo Chemical >10
99.9
98
99.9
oxidant comprised of various raw materials to generate the desired
product. The principle of SHS is that once one end of the starting
mixture is ignited, an exothermic sustainable reaction initiates and
causes a combustion wave to propagate through the sample, and
the desired product is finally generated without additional energy
[
10,11]. In addition to saving energy, SHS offers several benefits
including minimization of the operating time, simplification of pro-
cedure and equipment, and facile control of the composition of
the product by methodical variation of the molar ratio of the raw
materials [12,13].
To the best of our knowledge, there are no published reports on
SHS of nonstoichiometric FexO, despite its scientific significance.
Therefore, the purpose of this study was to synthesize nonstoichio-
metric FexO (0.833 ≤ x ≤ 1.00) using the SHS method, focusing on
the effects of the nonstoichiometric Fe content and diluent addi-
tion on the phase of the SHS product. The findings of this study are
expected to pave the way for a new FexO production route with the
benefits of energy saving and minimized production time.
Fig. 2. Schematic diagram of the experimental apparatus for self-propagating high-
temperature synthesis.
2.
Experimental
The properties and sources of the raw materials used for the synthesis of FexO
(
0.833 ≤ x ≤ 1.00) are listed in Table 1. NaClO4 was selected as an oxidizing reagent
for metallic iron [10]. Note that NaCl was added to the raw materials as a diluent to
provide the effect of decreasing the adiabatic flame temperature.
Fig. 1 shows the process flow sheet for the synthesis of FexO via the SHS method.
First, Fe, NaClO4, and NaCl powders were mixed in the desired ratio to obtain a total
mass of 20 g. The mixture was then ball-milled in a 140 mm diameter alumina pot
containing 25 alumina balls of 10 mm diameter. The mill was operated at 60 rpm
for 3 h under atmospheric conditions. After the milling process, 10 g of the sample
was pressed into a disk of 20 mm in diameter at 20 MPa for 10 min using a uniaxial
single-acting press. Fig. 2 shows the schematic diagram of the experimental appa-
ratus used for the SHS. The apparatus includes a reactor, a control unit, and a gas
control system [12]. The green compact of the sample was placed into a graphite
crucible with dimensions of 40 mm × 25 mm × 120 mm, and a disposable carbon
foil (5 mm × 200 mm × 0.1 mm) used as an igniter was also placed in contact with
the lower surface of the sample. Prior to commencing the reaction, the reactor was
evacuated using a rotary pump and then filled with argon gas of 99.999% purity at
atmospheric pressure. The sample was ignited by an electrically heated carbon foil
for 5 s. After completion of the ignition process, the exhaust valve of the reactor was
kept open for 20 min to allow the sample to cool down completely. The cooled sam-
ple was subjected to ultrasonic cleaning to remove NaCl and was dried overnight at
◦
110 C. The product was separated to retrieve the fraction with sizes less than 25 ␮m
and the morphology was observed using scanning electron microscopy (SEM). The
remaining product was ground in an agate mortar to obtain particles with sizes
less than 25 ␮m. The phase of the product obtained was identified by powder X-ray
diffraction (XRD) using Cu K␣ radiation (ꢀ = 1.5418
A˚ ).
Fig. 1. Schematic illustration of SHS procedure.

M. Hiramoto et al. / Journal of Alloys and Compounds 520 (2012) 59–64
61
Table 2
Molar specific heat of product materials Fe0.947O and NaCl.
Specific heat, Cp (J mol 1
−
K
−1
)
Cp = A + B × 10−3 T + C × 105
T
−2
+ D × 10−6 T2
Temperature range (K)
Product
A
B
C
D
Fe0.947O
NaCl (g)
48.794
8.372
0
0.383
−2.891
0
−1.715
0
0
298.15–1650
1650–
298.15–3600
65.839
37.56
0.217
3
. Results and discussion
having an irregular shape (Fig. 3(3)). These particles are shown at
higher magnification in the respective segments of Fig. 3 for clar-
ity. The SEM observations suggest that the spherical micron-sized
products crystallized from melted droplets because the Tad was suf-
ficiently higher than the melting point of the products. As shown
in Fig. 3(2b), the size of the product was smaller than that of raw
materials. This may be due to the presence of the added NaCl. It is
recognized that nanosized materials can be generated by an SHS
method known as alkali metal molten salt assisted combustion
The reaction involved in the SHS of Fe0.947O is given as follows:
0.947Fe + 0.25NaClO = Fe0.947O + 0.25NaCl(g) + 212.7 kJ (1)
4
The SHS process is only effective for strongly exothermic chemical
reactions. An important parameter in this regard is the adiabatic
flame temperature, Tad, which was calculated using the following
equation:
[
14]. The reduced dimensions of the products generated herein may
ꢁ
ꢃ
ꢀ
T
stem from the fact that these products were produced in the molten
NaCl. The submicron products were thus obtained after removal of
NaCl in water. The irregularly shaped SHS Fe0.947O particles were
produced by melting and sintering.
The SHS reaction for the formation of FexO is expressed by the
following equation:
ad
ꢂ
◦
niCip dT = −ꢁHr
(2)
2
98.15
i
Table 2 indicates the molar specific heat of the products in Eq. (1)
for evaluating the Tad from the heat balance of SHS due to the iron
oxidation, as given by Eq. (2). The data of Table 2 were referred to the
HSC 5.11 thermodynamic software. Tad evaluated for the synthesis
xFe + 0.25NaClO → FexO + 0.25NaCl(g), (x = 0.833–1.00) (3)
4
of Fe0.947O was 2042 K, which is above the melting point of Fe0.947
1651 K).
Fig. 3(1) shows the overall SEM image of the SHS-generated
O
(
Fig. 4(a) shows the powder XRD patterns of the SHS FexO
(x = 0.833–1.00) samples. All the FexO samples corresponded to the
Fe0.947O crystals with sizes less than 25 ␮m. Within this figure, SHS
Fe0.947O particles with spherical morphology of various dimensions
can be observed (Fig. 3(2a, 2b)) as well as SHS Fe0.947O products
FexO and Fe O4 phase. The value of x exerted no significant effect
on the phase of the products. The defect clusters of nonstoichio-
metric transition metal oxides consist of cation vacancies, cation
3
Fig. 3. SEM images of (1) SHS Fe0.947O crystals with sizes less than 25 (m and Fe0.947O with (2a, 2b) spherical shape of different sizes, and (3) irregular shape.

6
2
M. Hiramoto et al. / Journal of Alloys and Compounds 520 (2012) 59–64
Fig. 4. (a) XRD patterns of FexO (x = 0.833–1.00) synthesized by SHS method. The reaction involved in the SHS of FexO is given as follows: xFe + 0.25NaClO4 → FexO + 0.25NaCl
◦
(
g) (x = 0.833–1.00). (b) Enlargement of XRD patterns of SHS FexO in 2ꢂ region of 41.5–42.5 .
interstitials, and holes. The simplest cluster is the so-called 4:1 clus-
ter, which consists of 4 cation vacancies, 1 cation interstitial, and
a variable number of holes. The 16:5 cluster formed by the aggre-
gation of 4:1 clusters is an element of the inverse spinel structure
This methodical variation of the molar ratio of the raw materials
showed that in the case of xexpt = 0.947, the calculated composition
(xcalc) was coincident with the experimental composition.
The reaction involved in SHS of Fe0.947O with/without NaCl dilu-
ent is given as follows:
typical of Fe O [7]. Therefore, a product with large nonstoichio-
3
4
metric value tends to have a high Fe O content. However, the XRD
3
4
data indicate that there is no difference in the Fe O4 content with
(0.947–3.788z)Fe + (0.25–z)NaClO4 + zNaCl → (1–4z)Fe0.947O
3
decreasing x. This may be indicative of the fact that product gener-
ation is dependent on the adiabatic flame temperature and cooling
rate.
+
0.25NaCl, (z = 0 or 0.05)
(6)
Tad evaluated for the SHS reaction with the addition of the NaCl
diluent was 1805 K. The SHS reaction was not self-sustaining when
more than 5 mol% NaCl diluent was utilized. This suggested that
the addition of more than 5 mol% diluent caused extinction of the
reaction without sufficient generation of self-propagating heat. In
general, empirical evidence indicates that reactions are not self-
sustaining when Tad is less than 1800 K [15].
◦
◦
Fig. 4(b) shows the enlarged XRD patterns (41.5 < 2ꢂ < 42.5 )
of FexO (x = 0.833–1.00) samples synthesized by the SHS method.
The peaks of the products with 0.947 ≤ x ≤ 1.00 shifted to lower
angles in comparison to that of the product with x = 0.833. This was
attributed to the lattice parameter distortion of the crystal structure
because of the Fe defects.
Table 3 displays the results of the quantitative estimation, the
lattice parameter, and the calculated composition x of FexO in the
case where the experimental composition of Fe in the raw material
was varied. The experimental composition (xexpt) is the ratio of the
Fig. 5(a) shows the XRD patterns of the Fe0.947O sample pro-
duced by SHS with/without NaCl diluent. Only FexO and Fe O4
3
phases were detected without the addition of the diluent. In con-
trast, upon addition of the NaCl diluent, un-reacted Fe was also
Fe content in the raw materials. The lattice parameter, aexperimental
,
detected in addition to the FexO and Fe O4 phases. This suggested
3
was calculated from the following equation:
that the contact area of the reactants decreased because of the addi-
tion of diluent. It was also observed that the intensity of the Fe3O4
peak was decreased in product prepared with diluent with accom-
panying increase in the intensity of the FexO peak, in comparison
to the product prepared without diluent.
2
2
1
d2
h + k + l2
=
(4)
2
a
experimental
The following equation was used to calculate the composition x
of FexO from the lattice parameter, a [3]:
Fig. 5(b) shows the enlarged XRD patterns of the Fe0.947O sam-
ple produced by SHS with/without NaCl diluent in the 2ꢂ range
of 41.5–42.5 . This peak was coincident for the products obtained
a = 3.856 + 0.478x
(5)
◦
The results showed that the products obtained contained
with and without NaCl diluent.
approximately 70% FexO. The theoretical lattice parameters
Table 4 displays a comparison of the quantitative estimation,
the lattice parameter, and the calculated composition x of FexO for
the SHS products obtained with and without diluent. In the case
of diluent addition, the content of Fe3O4 impurity decreased, and
the purity of FexO was improved by more than 10%. This may be
accounted for by the decrease in the adiabatic flame temperature
and wave velocity generated by addition of the diluent, resulting
in a homogenous reaction with a narrow reaction zone [16]. There-
fore, the reactivity increases under these conditions and a higher
purity of FexO was obtained. However, the calculated composition
of the product prepared with the diluent shows a slight deviation
from the experimental composition of 0.947.
(
atheoretical) were obtained by substituting the experimental com-
position into Eq. (5). The aexperimental value tended to decrease as
the experimental composition of Fe in the raw material decreased.
Table 3
Quantitative estimation of composition and lattice parameter of SHS FexO product
from XRD patterns.
xexpt
FexO content
%)
Fe3O4 content
(%)
atheoretical
( A˚ )
aexperimental
( A˚ )
xcalc
(
0
0
0
0
1
.833
78 (± 4)
59 (± 2)
72 (± 2)
72 (± 3)
73 (± 2)
22.2 (± 1.3)
40.6 (± 1.8)
28.3 (± 1.4)
28.1 (± 1.6)
26.6 (± 1.1)
4.254
4.282
4.309
4.313
4.334
4.307
4.306
4.309
4.310
4.311
0.944
0.942
0.947
0.950
0.952
.891
.947
.957
.00
Fig. 6(a) shows the relationship between the experimental
and calculated composition of SHS FexO without NaCl diluent;
the calculated values are indicated next to the data points. The

M. Hiramoto et al. / Journal of Alloys and Compounds 520 (2012) 59–64
63
Fig. 5. (a) XRD patterns of the Fe0.947O sample produced by SHS with/without NaCl diluent, in which the synthesis reaction is expressed by the following equation:
◦
(
0.947–3.788z)Fe + (0.25–z)NaClO4 + zNaCl → (1–4z)Fe0.947O + 0.25NaCl (z = 0 or 0.05). (b) Enlarged areas of the spectra in (a) from 2ꢂ = 41.5–42.5 .
Table 4
Quantitative estimation of composition and lattice parameter of the SHS Fe0.947O with/without NaCl diluent from XRD patterns.
xexpt
Content of FexO (%)
Content of Fe3O4 (%)
Content of Fe (%)
atheoretical ( A˚ )
aexperimental ( A˚ )
xcalc
0
0
.947 (0 M NaCl)
.947 (0.05 M NaCl)
72 (± 2)
86 (± 3)
28.3 (± 1.4)
10.3 (± 1.0)
–
4.309
4.309
4.309
4.311
0.947
0.952
3.7 (± 0.6)
Fig. 6. (a) Relationship between experimental and calculated composition of SHS FexO without NaCl diluent. The calculated values are indicated next to the data points. The
experimental composition x is the ratio of the Fe content in the raw materials. The following equation was used to calculate the composition x of FexO from a, the lattice
parameter: a = 3.856 + 0.478x [3]. (b) Phase diagram for the Fe–O system [17]. Nonstoichiometric compounds of FexO (0.942 ≤ x ≤ 0.952) were obtained through SHS without
additional external heating.
composition of SHS FexO was limited to a narrow nonstoichiometric
range. Fig. 6(b) shows the phase diagram for the Fe–O system [17].
The shaded region shows the composition range obtained by SHS,
and the stoichiometric composition of wüstite is generally accepted
to be x = 0.947. The nonstoichiometric compounds of SHS FexO
were obtained within the composition range of x = 0.942–0.952,
which are of the ± 0.005 limits of the stoichiometric composition
of 0.947.
(1) All of the products obtained were double phases of FexO and
Fe O . The FexO peaks of products with 0.947 ≤ x ≤ 1.00 shifted
3
4
to lower angles in comparison to that of the product with
x = 0.833. This shift was caused by the presence of lattice defects.
(2) The addition of a NaCl diluent to the raw material mixture dur-
ing SHS facilitated the conversion of the raw materials to yield
high-purity FexO powders.
(3) The lattice parameter of Fe0.947O, synthesized by SHS without
NaCl diluent, corresponded to the theoretical lattice parameter.
Nonstoichiometric compounds of FexO (0.942 ≤ x ≤ 0.952) were
obtained through SHS without additional external heating.
4
. Conclusions
The results reveal that the SHS method can be used to produce
nonstoichiometric FexO with the benefits of energy saving and min-
imization of production time. The purity of FexO obtained by this
The effects of nonstoichiometric content and diluent addition on
the self-propagating high-temperature synthesis (SHS) of nonsto-
ichiometric FexO (x = 0.833–1.00) were experimentally evaluated.
The following results were obtained:
method may be further improved by magnetic separation of Fe O4
3
and Fe.

6
4
M. Hiramoto et al. / Journal of Alloys and Compounds 520 (2012) 59–64
References
[9] H.C. Yi, J.J. Moore, J. Mater. Sci. 25 (1990) 1159–1168.
10] H. Ishikawa, K. Oohira, T. Nakajima, T. Akiyama, J. Alloys Compd. 454 (2008)
84–388.
11] K. Taniguchi, N. Okinaka, T. Akiyama, J. Alloys Compd. 509 (2011) 4084–4088.
12] T. Hirano, H. Purwanto, T. Watanabe, T. Akiyama, J. Alloys Compd. 441 (2007)
[
3
[1] M. Jiménez-Melendo, A. Domínguez-Rodríguez, J. Castaing, Acta Metall. Mater.
[
[
4
3 (1995) 3589–3604.
[
[
[
[
2] R. Aragón, Phys. Rev. B 46 (1992) 5328.
2
63–266.
13] I.S. Ahmed, S.A. Shama, H.A. Dessouki, A.A. Ali, Mater. Chem. Phys. 125 (2011)
26–333.
14] S.T. Aruna, A.S. Mukasyan, Curr. Opin. Solid State Mater. Sci. 12 (2008)
4–50.
3] C.A. McCammon, L.-G. Liu, Phys. Chem. Miner. 10 (1984) 106–113.
4] Y. Masaki, T. Fujii, M. Hayashi, K. Nagata, ISIJ Int. 51 (2011) 203–207.
5] Y.B. Khollam, S.R. Dhage, H.S. Potdar, S.B. Deshpande, P.P. Bakare, S.D. Kulkarni,
S.K. Date, Mater. Lett. 56 (2002) 571–577.
[
[
3
4
[6] G. Neri, A. Bonavita, S. Galvagno, P. Siciliano, S. Capone, Sens. Actuators B: Chem.
[
[
[
15] Z.A. Munir, U. Anselmi-Tamburini, Mater. Sci. Rep. 3 (1989) 277–365.
16] H. Jie-Cai, X.-H. Zhang, J.V. Wood, Mater. Sci. Eng. A 280 (2000) 328–333.
17] B. Sundman, J. Phase Equilib. 12 (1991) 127–140.
8
2 (2002) 40–47.
[
[
7] S. Mrowec, A. Podgórecka, J. Mater. Sci. 22 (1987) 4181–4189.
8] D.P. Johnson, Solid State Commun. 7 (1969) 1785–1788.