Materials Transactions, Vol. 49, No. 5 (2008) pp.

1081 to 1083
Special Issue on Platform Science and Technology for Advanced Magnesium Alloys, IV
#2008 The Japan Institute of Metals

Effect of CaO on Oxidation Resistance and Microstructure of Pure Mg

Seong-Ho Ha* , Jin-Kyu Lee and Shae K. Kim
Korea Institute of Industrial Technology, 7-47 Songdo-dong, Incheon 406-840, Korea

This study was carried out to examine reduction mechanism of CaO and to investigate behaviors of CaO in pure Mg in terms of
microstructure, oxidation resistance and phase formation. Pure Mg was used instead of Mg alloys to minimize the effects of other elements. With
respect to CaO content, microstructures of the alloys showed prominent grain refinement. Mg2 Ca phase was formed even in 0.07CaO added pure
Mg by reduction, while Mg2 Ca phase was formed over 1.35Ca added pure Mg. With respect to CaO content, the hardness of CaO added pure Mg
was increased by grain refinement. From oxidation test by TGA, the oxidation behavior of CaO added Mg was similar to that of Ca added Mg.
From AES result, there was the thin oxide layer mixed with MgO and CaO in CaO added Mg. [doi:10.2320/matertrans.MC200786]

(Received October 10, 2007; Accepted January 30, 2008; Published March 19, 2008)
Keywords: magnesium, calcium oxide, Mg2 Ca phase, reduction

1. Introduction by CaO addition. Major concern of the study is to establish

In recent years, vehicle weight reduction has been issue in
automotive industry due to necessity of fuel economy and
global warming. The demand for Mg alloys has increased due
to their low density and good specific mechanical proper-
ties.1) Mg is the lightest of all structural metals and is thus
very attractive for applications in transportation. The total
vehicle weight can be reduced by using Mg alloy compo-
nents, such as seat frame, instrument panel, steering column
and transmission housing. Each 10% saving in dead weight of
a motor vehicle corresponds to an increase in fuel economy
of 5.5%. The problem of Mg and Mg alloys during the
process is that they are oxidized at high temperature and
especially ignited during melting and casting processes in an
ambient atmosphere. Melt protective gases such as SF6 have
been used in order to prevent the ignition of molten Mg
during melting and casting. However, the use of these gases
is limited because of theirs high cost and significant impact
on global warming potential. SF6 is one of the most potent
greenhouse gases currently known and is commonly used
during melting and casting of Mg alloys to prevent the
ignition of the melt. Its global warming potential is about
23,900 times worse than carbon dioxide (100-year time
horizon), and for this reason it has come under intense
environmental scrutiny. The Kyoto Protocol calls for a
reduction in SF6 use. Environmental and economical con-
cerns have caused the Mg industry to look for possibilities to
replace or reduce the outlets of SF6 in the casting process. If
gas concentration increases during melting, it has also bad
influence on corrosion resistance of steel based equipments.2)
In order to reduce the use of melt protective gases, many
researchers have performed studies to improve the flame
resistance of Mg alloys through alloying. It is well known
that Ca are used to improve the flame and oxidation
resistance of Mg alloys.3) You et al. ascribed the decrease
of the oxidation resistance of Mg-Ca alloys to the formation
of MgO/CaO protective film at elevated temperature.4)
On the other hand, Kim et al. studied environment-friendly
manufacturing process of Mg alloys without protective gas
*Corresponding author, E-mail: badguy99@kitech.re.kr
simultaneously (1) cleaner production routes with reduced or
without SF6 gas, (2) cost reduction by alloying low cost
element as well as/reducing or eliminating high cost SF6 ,
(3) original process abilities, esp. fluidity and hot tearing
susceptibility, (4) original mechanical properties and (5)
safety during applications by raising ignition temperature
in pure Mg and Mg alloy.5–7)
This paper discussed reduction mechanism of CaO and
behaviors of CaO in pure Mg in terms of microstructure,
oxidation resistance and phase formation. Pure Mg was used
instead of Mg alloys to minimize the effects of other
elements.
2. Experimental Procedures
Pure Mg was melted in a steel crucible in an electric
resistance furnace at 1003 K under SF6 and CO2 gases. The
desired fraction of CaO was added into molten pure Mg. The
CaO added melts were cast into steel mold preheated at
473 K. In order to observe the change of the microstructures
by CaO addition, the cast billets were polished and observed
using an Olympus PME3 microscopy. The variation of
hardness was investigated by Rockwell hardness tester (15T
scale). The obtained values of hardness were based on the
average of 7 tests. Oxidation test was performed with TGA
(Thermogravimetric analysis) to investigate the effect of CaO
on oxidation resistance of pure Mg. TGA was performed at
773 K for 5 hrs. After TGA, the surface was analyzed by AES
(Auger Electron Spectroscope). Mapping Mg, Ca and O by
EPMA was performed to investigate the existence of CaO
in pure Mg. Phase analysis was performed using XRD with
the scan speed of 0.08/s.
3. Results and Discussion
Figure 1 shows the microstructures of CaO added pure
Mg. As shown in the picture, the morphology of 0.07CaO
added Mg was composed of
- Mg grains. With respect to
CaO content, the grains were refined. This tendency was
more prominent in 0.28CaO added Mg. It was due to the
formation of Mg2 Ca in 0.28CaO added Mg, although Ca
1082 S.-H. Ha, J.-K. Lee and S. K. Kim

(a) (b)

(c) (d)

Fig. 3 XRD results of 1.34CaO added Mg.

60

55
Fig. 1 As-cast microstructures of (a) Mg-0.07CaO, (b) Mg-0.28CaO,

Hardness(HR15T)
(c) Mg-0.66CaO and (d) Mg-1.34CaO alloys.
50

45

(a) (b) 40

35
Pure Mg
30
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

wt%CaO

Fig. 4 Results of Rockwell hardness (HR15T) in CaO added Mg.

(c) (d)

Mg(l) þ CaO(s) ¼ Ca(s) þ MgO(s)

Fig. 2 EPMA mapping results of Mg, Ca and O for 1.34CaO added Mg:
(a) BSE image, (b) Mg, (c) Ca and (d) O.
solubility in Mg is 1.35% based on Mg-Ca binary phase
diagram.8) The previous study showed that CaO existed as
Al2 Ca intermetallic compound in Mg-Al base alloys by
reduction and the compound could contribute to grain
refinement of the alloy during solidification.
Figure 2 shows EPMA mapping result of (b) Mg, (c) Ca
and (d) O for 1.34CaO added pure Mg. Fig. 2(a) shows BSE
(back scattered electron) image of 1.34CaO added pure Mg.
Ca was distributed on the grain boundary. But O was not
detected all over the image. Figure 3 shows XRD result of
1.34CaO added pure Mg. 1.34CaO added pure Mg presented
Mg and Mg2 Ca peaks. Based on the results of EPMA and
XRD, CaO seemed to be reduced to Ca in pure Mg.
Thermodynamical consideration for reduction mechanism
of CaO in pure Mg was carried out referring to thermody-
namic table.9)
At 1000 K
Mg(l) þ 1/2O2 (g) ¼ MgO(s)

Gf ¼
492:858 kJ/mol ð1Þ
CaO(s) ¼ Ca(l) þ 1/2O2 (g)

Gf ¼ þ531:087 kJ/mol ð2Þ

Gf ¼ þ38:229 kJ/mol ð3Þ
Above equations show the possible chemical reactions
during stirring after adding CaO into pure Mg melt. The
oxidation reaction of Mg is given as eq. (1). Equation (2)
shows the reverse reaction against oxidation of Ca. Gibbs
free energy values (
Gf ) of the equations exhibited

492:858 and +531.087 kJ/mol, respectively. The reaction
equations of (1) and (2) were rearranged as (3). Because
Gibbs free energy of eq. (3) exhibits +38.229 kJ/mol, CaO
could not be thermodynamically reduced in Mg melt.
Although it could not be explained by thermodynamics, it
was confirmed that CaO reduced as Mg2 Ca in Mg through
EPMA and XRD analysis. CaO might be possible by (1)
surplus energy occurred during stirring and (2) additional
energy such as ignition by reaction between Mg melt and
moisture which exists in CaO powder as below
Mg(l) þ H2 O(g) ¼ MgO(s) þ H2 (g)

Gf ¼
300:145 kJ/mol ð4Þ
After that, it seems that the following phase was formed
by:
2Mg(l) þ Ca(s) ¼ Mg2 Ca(s)

Gf ¼
35:582 kJ/mol ð5Þ
Much work should be performed to understand the
reduction mechanism of CaO in pure Mg.
Figure 4 shows the dependence of hardness on the mass
fraction of CaO. The hardness of CaO added Mg was
increased with respect to CaO content by grain refinement.
 Effect of CaO on Oxidation Resistance and Microstructure of Pure Mg 1083

Fig. 5 Weight gain curves of pure Mg and various CaO and added Mg
during TGA at 773 K for 5 hrs.
(a)
Atomic Concentration, %
for sputtering time. In CaO added Mg, there was the thin
oxide layer mixed with MgO and CaO. The oxidation
behavior of CaO added Mg was similar to that of Ca added
Mg studied by B. S. You et al.3)
Generally, MgO oxide layer exhibits protective behavior
below about 673 K, while weight gain by oxidation is
accelerated with respect to temperature above 673 K because
of porous MgO surface oxide film at high temperature. It
could be explained though Pilling-Bedworth ratio, R.10)
Pilling-Bedworth ratio means the ratio of oxide volume
to metal volume. Al is greater than 1, which explains the
protective effect of the Al oxide. Because the Pilling-
Bedworth ratio for Mg is 0.81, MgO oxide film on surface
of molten Mg could not act as a protective layer to prevent
ignition and oxidation of molten Mg. Although it could not
be explained by Pilling-Bedworth theory, it has been reported
that ignition and oxidation of Ca added Mg was suppressed
100
O due to the dense film which consisted of porous CaO (0.78)
and MgO.11) Further study should be performed to examine

Mg
80
the ignition suppression mechanism by Ca addition.
60

4. Conclusion
40

20
It was confirmed that CaO was reduced to Ca through
phase analysis. Mg2 Ca phase was formed even in 0.07CaO
0
0 20 40 60 80 100 120
added pure Mg by reduction, while Mg2 Ca phase was formed
Sputtering Time, min over 1.35% in Ca added Mg. With respect to CaO content, the
hardness of CaO added pure Mg increased by grain refine-
(b) 100
O
ment. From oxidation test by TGA, the oxidation behavior of
Mg CaO added Mg was similar to that of Ca added Mg for the
Atomic Concentration, %

80 Ca
previous study. And also, from AES result, there was the thin
60
oxide layer mixed with MgO and CaO in CaO added Mg.

40
20
0 2) W. Ha and Y. J. Kim: J. Alloys Comp. 422 (2006) 208–213.
0 20 40 60
Sputtering Time, min
Fig. 6 AES depth profiles sputtered from the surface of (a) pure Mg and
(b) CaO added Mg oxidized during TGA at 773 K for 5 hrs.
The hardness value of pure Mg was, as a hidden line, about
34 HR15T. The increase of hardness was more prominent
over 0.28CaO addition.
Figure 5 shows the weight gain of pure Mg and CaO added
Mg as a function of oxidation time at 773 K. The results
indicate that pure Mg showed a parabolic law due to the loose
and porous structure of MgO. On the contrary, the oxidation
rate was reduced with respect to CaO content for CaO added
Mg at the same temperature. Figure 6 shows the AES depth
profile sputtered form the surface. In pure Mg, it was
confirmed that there was a thick MgO oxide layer on the
surface because the concentration of Mg and O were constant
REFERENCES
1) R. Schmid-Fetzer and J. Grobner: Adv. Eng. Mater. 3 (2001) 947–961.
80 100 120
3) X. Zeng, Q. Wang, Y. Lu, W. Ding, Y. Zhu, C. Zhai, C. Lu and X. Xu:
Mater. Sci. Eng. A 301 (2001) 154–161.
4) B. S. You, M. H. Kim, W. W. Park and I. S. Chung: J. Kor. Inst. Met. &
Mater. 39 (2001) 446–450.
5) S. K. Kim, J. K. Lee, H. Cho, H. H. Jo, W. Ha and Y. J. Kim: Proc.
Magnesium Technology 2005, ed. By Neale R. Neelameggham,
Howard I. Kaplan and Bob R. Powell, (The Minerals, Metals and
Materials Society 2005) pp. 285–289.
6) J. K. Lee, Y. Y. Yoon and S. K. Kim: Proc. Magnesium Technology
2006, ed. By Alan A. Luo, Neale R. Neelameggham and Randy S.
Beals, (The Minerals, Metals and Materials Society 2006) pp. 185–189.
7) S. K. Kim: Proc. Magnesium Technology 2006, ed. By Alan A. Luo,
Neale R. Neelameggham and Randy S. Beals, (The Minerals, Metals
and Materials Society 2006) pp. 517–521.
8) T. B. Massalski: Binary Alloy Phase Diagram, (American Society for
Metals, Ohio, 1986) pp. 618–621.
9) Malcolm W. Chase, Jr: NIST-JANAF Thermochemical Tables, (Jour-
mal of Physical and Chemical Reference Data, Monograph No. 9).
10) N. B. Pilling and R. E. Bedworth: J. Inst. Metals 29 (1923) 529–582.
11) B. S. You, W. W. Park and I. S. Chung: Scripta mater. 42 (2000) 1089–
1094.