The effect of aluminothermic reaction on the progress of carbothermic reaction in simultaneous mechanochemical reduction of CuO and ZnO

Article abstract of DOI:10.1016/j.jallcom.2009.04.066

In this research, the effect of aluminothermic reaction on the progress of carbothermic reaction in simultaneous mechanochemical reduction of CuO and ZnO has been studied. A mixture of CuO and ZnO powders along with different amounts of aluminum and graph

Full text of DOI:10.1016/j.jallcom.2009.04.066

Journal of Alloys and Compounds 482 (2009) 522–525
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jallcom
The effect of aluminothermic reaction on the progress of carbothermic
reaction in simultaneous mechanochemical reduction of CuO and ZnO
M.S. Marashi∗, J. Vahdati Khaki
Department of Materials Science and Metallurgical Engineering, Engineering Faculty, Ferdowsi University of Mashhad, Mashhad, Iran
a r t i c l e i n f o
a b s t r a c t
Article history:
In this research, the effect of aluminothermic reaction on the progress of carbothermic reaction in simulta-
neous mechanochemical reduction of CuO and ZnO has been studied. A mixture of CuO and ZnO powders
along with different amounts of aluminum and graphite was milled in a planetary ball-milling machine.
The adiabatic temperature and overall reaction enthalpy were varied by changing the aluminum/carbon
ratio. The results showed that by increasing the amount of aluminum, the exothermic aluminothermic
reaction provided the required heat for the endothermic carbothermic reactions. Consequently the oxide
powders were entirely reduced during ball-milling process and ␣-brass–Al2O3 nanostructured compos-
ites were obtained. X-ray diffraction (XRD) and transmission electron microscopy (TEM) were used to
characterize the final products. By utilizing the Scherrer’s method, the mean grain size of the products
was calculated to be in the nanometer range, further confirmed by TEM observations.
Received 27 October 2008
Received in revised form 12 April 2009
Accepted 16 April 2009
Available online 23 April 2009
Keywords:
Mechanochemical processing
Nanostructured composite
X-ray diffraction
© 2009 Elsevier B.V. All rights reserved.
1. Introduction
reduction of CuO by C. But according to the author’s investiga-
tion, the effect of metallothermic reaction on carbothermic reaction
Reduction of metal oxides is usually carried out at high
temperatures in appropriate conditions with respect to equilib-
rium thermodynamics. However, because of the high activity of
fresh surfaces, mechanochemical reduction is possible [1]. During
mechanical alloying repeated welding and fracturing of powder
particles increases the area of contact between the reactant pow-
der particles due to a reduction in particle size and allows fresh
surfaces to come into contact repeatedly; this allows the reaction
to proceed without the necessity for diffusion through the product
layer [2]. The redox reactions in powder mixtures of oxide and pure
metal during reactive milling allow both the refining of metals and
the direct production of alloys from their respective oxides, thus
attracting much attention [3].
In the past couple of decades, reduction of many metallic
oxides by solid metallic reducing agents (metallothermic reduc-
tion) through mechanical alloying has been extensively studied.
Solid state reduction of CuO by Ca, Al, Mg, Ti, Mn, Si, Fe, Ni [2–8] and
ZnO by Al [9], Ca [3], Ti [10] and Mg [11] have been investigated up
to now. Schaffer and McCormic also applied the reaction between
CuO, ZnO and Ca to synthesize ␤ꢀ-brass [3].
during mechanical alloying has not been studied yet.
In this work, we have used CuO/ZnO powders as starting mate-
rials and Al/carbon as reducing agents and studied the effect of
aluminothermic reaction on carbothermic reaction during simul-
taneous reduction of mentioned oxides and formation of ␣-brass.
2. Experimental work
CuO (96% purity, Merck 102761), ZnO (99% purity, Fluka 96484), Al (99.5% purity,
Mashhad Powder Metallurgy Co.) and C (99.5% purity, Mashhad Powder Metallurgy
Co.) powders were mixed to give the following desired stoichiometric composition:
0.3ZnO + 0.7CuO + (1 − 3x)C + 2xAl → 0.3Zn + 0.7Cu + (1 − 3x)CO + xAl2O3 (1)
where x is a variable. Changing the value of x will influence the weight percent
of the primary carbon and aluminum, which consequently affects the adiabatic
temperature of the reaction.
Adiabatic temperature can be calculated from the following equation [13]:
ꢁ
ꢃ
ꢁ
ꢃ
ꢂ
i
ꢂ
i
ad
T
T
m
ꢀ
ꢀ
ꢀ
ꢀH2◦98
=
Cpi s dT]
+
(ꢀHmi ) +
Cpi l dT]
(2)
i
T
m
298
where ꢀH2◦98 is the standard enthalpy of reaction, Tmi is the melting point of product
i, Cpi s is the heat capacity of product i in the solid state, ꢀHmi is the enthalpy of fusion
In addition to metallothermic reduction, carbothermic reduc-
tion can take place during mechanical alloying. Yang et al. [12] and
Tokumitsu [1] have reported the possibility of mechanochemical
of product i and Cpi l is the heat capacity of product i in the liquid state. If vaporizing
of product i occurs, heat capacity of product i in the vapor state (Cpi v) and enthalpy
of vaporization of product i (ꢀHvi ) should be included in Eq. (2). The change in the
adiabatic temperature vs. x, by using Eq. (2) is plotted in Fig. 1. Four different starting
compositions were chosen for the experiments. x, Tad, and the content of primary Al
and C for each studied composition are shown in Table 1.
∗
Corresponding author. Tel.: +98 9155201436.
E-mail address: ma marashi 81@yahoo.com (M.S. Marashi).
Mechanical alloying (MA) was performed under argon atmosphere in a planetary
ball-mill at room temperature with a rotation speed of 260 rpm. The milling media
0925-8388/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2009.04.066

M.S. Marashi, J.V. Khaki / Journal of Alloys and Compounds 482 (2009) 522–525
523
Fig. 1. The change in the adiabatic temperature for different values of x.
was two types of bearing steel balls with diameters of 10 and 12 mm and the chosen
weight ratio of the balls to the powder (BPR) was 40:1. In each run, the total amount
of milled powder mixture was 7.5 g.
The reduction progress in the ball-milled powders was characterized by means
of X-ray diffraction (XRD) with Cu K␣ (ꢁ = 1.54060 Å). The microstructure and mor-
phology of the reacted powders was studied by transmission electron microscopy
(TEM) (LEO 912 AB). The sample chosen for TEM observation was prepared by mix-
ing the powders with ethanol in an ultrasound stirrer and mounting the suspended
powders on a carbon coated grid. The crystallite size of the powders was calculated
using the Scherrer method from the line broadening of the diffraction lines in the
XRD patterns [2]:
0.9ꢁ
B cos ꢂ
d =
(3)
where d is the average grain size (nm), ꢁ the Cu K␣ wavelength (nm), B the diffraction
peak width at half maximum intensity (rad) and ꢂ the Bragg diffraction angle.
3. Results and discussion
3.1. Reduction by aluminum and carbon
Fig. 2. XRD patterns of sample 1 ball-milled for different time periods.
The first step was to ensure that the copper and zinc oxides
could be simultaneously reduced during the mechanochemical pro-
cess; in this regard, sample 1 was ball-milled for different time
periods. The adiabatic temperature in this sample was calculated
to be 2300 K, and according to Merzhanov’s criteria (Tad > 1800 K),
occurrence of self-propagating high temperature synthesis (SHS)
is possible [14]. Therefore, there is a high probability that the heat
produced by the metallothermic reaction is sufficient to activate
the carbothermic reduction reaction. Fig. 2 shows the XRD results
obtained from this sample.
that the reaction occurs as an SHS and the aluminothermic reaction
is responsible for activating the carbothermic reaction.
The changes in the (1 1 1) ␣-brass peak caused by increasing the
ball-milling time are shown in Fig. 3. By increasing the ball-milling
time a slight broadening could be seen in the (1 1 1) peak, its inten-
As can be seen in Fig. 2, after 2 h of ball-milling, no new peaks
were seen in the XRD pattern i.e. no detectable chemical reaction
has occurred. But after 3 h of ball-milling, the peaks corresponding
to the starting materials have disappeared and new peaks corre-
sponding to ␣-brass and ␣-Al2O3 can be seen. This fact indicates
Table 1
Tad and the amount of primary aluminum and carbon for different values of x in four
studied samples.
Sample no.
x
Tad (K)
Primary Al
Primary C
weight percent
weight percent
1
2
3
4
0.215
0.188
0.155
0.08
2300
1800
1300
298
12.09%
10.62%
8.81%
4.61%
4.44%
5.57%
6.77%
9.74%
Fig. 3. Changes of (1 1 1) ␣-brass peak vs. ball-milling time for sample 1.

524
M.S. Marashi, J.V. Khaki / Journal of Alloys and Compounds 482 (2009) 522–525
Fig. 4. Average grain size of ␣-brass vs. the milling duration for sample 1.
sity decreases and the peak is displaced to a higher Bragg angle. The
broadening and decrease in intensity of the peak could be related
to the decrease in ␣-brass grain size and increase in lattice micros-
train. Fig. 4 displays the average grain size of ␣-brass vs. ball-milling
time. The reduction in grain size and attainment of a steady-state
grain size are the result of interplay and balance between several
mechanisms occurring simultaneously during the milling process.
For instance, at first, formation of new defects particularly disloca-
tions might result in formation of high dislocation density regions in
the grains, pile up at the grain boundaries or irregular clusters in the
grains which results in reduction of crystalline size [15]. This mech-
anism is followed by work hardening and recovery mechanisms
[15]. Displacement of the peak to higher Bragg angles is related to
the decrease in the lattice parameter caused by the gradual depar-
ture of small amounts of zinc from the brass alloy [16]. The XRD
results also confirm this fact, such that after 40 h of ball-milling,
peaks corresponding to zinc could clearly be spotted (Fig. 1).
3.2. Effect of aluminothermic reduction on carbothermic
reduction
In order to study the effect of the aluminothermic reduction
reaction on the progression of the carbothermic reaction, samples
2, 3 and 4 were ball-milled for 20 h (under the same conditions
as sample 1) and the results were compared with sample 1. Fig. 5
shows the XRD patterns obtained from these samples.
␣-Brass peaks could be found in sample 2 which had a calculated
adiabatic temperature equal to 1800 K. Also, the peaks correspond-
ing to CuO have been eliminated and ZnO peaks have been intensely
reduced which indicate that the carbothermic reaction has been
activated in this sample. But due to the fact that the heat produced
by the aluminothermic reaction was low (because the adiabatic
temperature was low), the carbothermic reaction has not been com-
pleted. Peaks corresponding to ␣-Al2O3 could not be seen in the
XRD pattern in Fig. 5 for samples 1 and 2. This could be because of
Fig. 6. TEM images of sample 1 ball-milled for 20 h; (a) size of grains; (b) size of
particles.
the small size of Al2O3 particles and the fact that they are caught
in the dense copper matrix as reported by Ying and Zhang [17] and
Hwang and Lee [18].
The adiabatic temperature in sample 3 is calculated to be 1300 K;
therefore, SHS did not occur. Low intensity Cu peak in this sample,
suggesting that the aluminothermic reaction took place gradually.
A gradual aluminothermic reaction did not supply the required
temperature for occurrence of carbothermic reaction. Hence, even
after 20 h of ball-milling, CuO was partially reduced. The same
behavior was observed in sample 4. Because low amounts of alu-
minum were used in sample 4, the Cu peaks in the XRD pattern
had very low intensities. Although the carbothermic reaction did
Fig. 5. XRD patterns of samples ball-milled for 20 h.

M.S. Marashi, J.V. Khaki / Journal of Alloys and Compounds 482 (2009) 522–525
525
not take place in samples 3 and 4, the peaks corresponding to car-
bon were eliminated in the XRD patterns. Elimination of carbon
peaks is associated with its intense deformation during ball-milling
[19,13]. When ball-milling systems containing several phases, the
phase with the softest crystal structure undergoes more deforma-
tion compared to phases with harder crystal structures. Therefore,
the carbon peaks gradually decreased in intensity and completely
disappeared after 20 h of ball-milling. In this state, carbon particles
are believed to have an amorphous structure [19–23].
In order to study the microstructure and particle size of the
synthesized brass, transmission electron microscopy (TEM) was
utilized. Fig. 6 shows bright field images of sample 3 which was
ball-milled for 20 h. As can be recognize from Fig. 6a, the mean
grain size is about 20 nm while the grain size obtained from the
XRD pattern (using the Scherrer method) is about 10 nm. The pos-
sible reason for this distinction can be physical differences in values
being determined [24]. The X-ray structural method actually allows
us to determine a size of coherently scattering domains connected
with internal grain areas having a weakly distorted crystal lattice,
whereas the TEM method measures a complete grain size which
includes near boundary strongly distorted [24]. This distinction in
grain size can also be related to ignoring the effect of stacking faults
on broadening/shift of peak positions. Brass alloys have low stack-
ing fault energy, so high density of stacking fault in (1 1 1) planes
during mechanical alloying is expected. These faults can cause
anomalous hkl-dependent peak bordering in the X-ray diffraction
patterns. Nevertheless, in calculating grain size by Scherrer’s equa-
tion, it was assumed that the peak bordering is only due to small
grain size [2].
performed. It was observed that by changing the aluminum/carbon
ratio and consequently changing the adiabatic temperature, the
reduction reactions could take place in two different modes: grad-
ually and instantaneously. In order to activate the carbothermic
reduction reaction and the aluminothermic reduction reaction
simultaneously, the adiabatic temperature must be over 1800 K and
the aluminothermic reaction must occur instantaneously. After 3 h
of ball-milling, the peaks corresponding to the starting materials
were eliminated and brass peaks appeared in the sample which
had a 2300 K adiabatic temperature. While, even after 20 h of ball-
milling, complete reduction did not occur in the other samples.
References
[1] K. Tokumitsu, J. Solid State Ionics 101–103 (1997) 25–31.
[2] C. Suryanarayana, J. Prog. Mater. Sci. 46 (2001) 1–184.
[3] G.B. Schaffer, P.G. McCormick, J. Appl. Phys. Lett. 55 (1989) 45.
[4] J.M. Wu, Z.Z. Li, J. Alloys Compd. 299 (2000) 9–16.
[5] S.J. Hwang, D. Wexler, A. Calka, J. Mater. Sci. 39 (2004) 4659–4662.
[6] T.D. Shen, K.Y. Wang, J.T. Wang, M.X. Quan, J. Mater. Sci. Eng. A151 (1992)
189–195.
[7] G.B. Schaffer, P.G. McCormick, J. Metall. Trans. A21 (1990) 2789–2794.
[8] G.B. Schaffer, P.G. McCormick, J. Scripta Metall. Mater. 23 (1989) 835–838.
[9] J.M. Wu, J. Mater. Lett. 48 (2001) 324–330.
[10] G.B. Schaffer, P.G. McCormick, J. Metall. Trans. A22 (1991) 3019–3024.
[11] H. Yang, P.G. McCormick, J. Solid State Chem. 107 (1993) 258–263.
[12] H. Yang, G. Nguyen, P.G. McCormick, J. Scripta Metall. Mater. 32 (1995) 681–684.
[13] T.D. Shen, K.Y. Wang, J.T. Wang, M.X. Quan, J. Mater. Sci. Eng. A 151 (1992)
189–195.
[14] L. Takacs, J. Prog. Mater. Sci. 47 (2002) 355–414.
[15] B. Gómez, E. Gordo, J.M. Torralba, J. Sci. Eng. A 430 (2006) 59–63.
[16] A.E. Bayer, J. Mater. Eng. Perform. 6 (1997) 149–152.
[17] D.Y. Ying, D.L. Zhang, J. Mater. Sci. Eng. A 286 (2000) 152–156.
[18] S.J. Hwang, J.H. Lee, J. Mater. Sci. Eng. A 405 (2005) 140–146.
[19] E. Gaffet, M. Harmelin, J. Less-Common Met. 157 (1990) 201–222.
[20] C.C. Koch, J. Non-Cryst. Solids 117–118 (1990) 670–678.
[21] T.D. Shen, C.C. Koch, T.L. McCormick, R.J. Nemanich, J.Y. Huang, J.G. Huang, J.
Mater. Res. 10 (1) (1995) 139–148.
The particles do not have a constant and similar morphology and
the size of the particles was within 0.2–0.3 ␮m range (Fig. 6b).
4. Conclusion
[22] A.W. Weeber, H. Bakker, J. Phys. B 153 (1988) 93–122.
[23] N.J. Welham, J. Miner. Process. 67 (2002) 187–198.
[24] R.Z. Valiev, R.K. Islamgaliev, I.V. Alexandrov, J. Prog. Mater. Sci. 45 (2000)
103–189.
Simultaneous mechanochemical reduction of copper and zinc
oxides in the presence of carbon and aluminum was successfully