Advertisement

High-temperature oxidation behaviour of low-entropy alloy to medium- and high-entropy alloys

  • Nana Kwabena Adomako
  • Jeoung Han Kim
  • Yong Taek Hyun
Article

Abstract

The high-temperature oxidation behaviour of CoCrNi, CoCrNiMn, and CoCrNiMnFe equimolar alloys was investigated. All three alloys have a single-phase face-centred cubic structure. Thermogravimetric analyses (TGA) were conducted at temperatures ranging from 800 to 1000 °C for 24 h in dry air. The kinetic curves of the oxidation were measured by TGA, and the microstructure and chemical element distribution in different regions of the specimens were analysed. The oxidation kinetics of the three alloys followed the two-stage parabolic rate law, with rate constants generally increasing with increasing temperature. CoCrNi displayed the highest resistance to oxidation, followed by CoCrNiMnFe and CoCrNiMn exhibiting the least resistance to oxidation. The addition of Mn to CoCrNi increased the oxidation rate. The oxidation resistance of CoCrNiMn was enhanced by the addition of Fe. Less Mn Content and the formation of more Cr2O3 were responsible for the reduction in the oxidation rates of CoCrNiMnFe. The calculated activation energies of CoCrNiMn and CoCrNiMnFe at 800, 850 and 900 °C were 108 and 137 kJ mol−1, respectively, and are comparable to that of Mn diffusion in Mn oxides. The diffusion of Mn through the oxides at 800–900 °C is considered to be the rate-limiting process. The intense diffusion of Cr at 1000 °C contributed to the formation of CrMn1.5O4 spinel with Mn in the outer layer of CoCrNiMn and Cr2O3 in the outer layer of CoCrNiMn.

Keywords

High-entropy alloys Oxidation Thermogravimetric analyses Diffusion 

Introduction

High-entropy alloys (HEAs) are new solid-solution alloys that contain more than five elements in roughly equal proportions. HEAs increase the maximum entropy value of the entire solid solution and provide more thermodynamic freedom in the equilibrium state, thus allowing the alloying design to achieve the material’s stability for practical applications [1]. Because of their highly unique properties such as good corrosion resistance [2], good fatigue resistance [3], high fracture toughness, a wide range of Young’s moduli, and high yielding strengths [4], they are now the focus of significant attention in materials science and engineering. In particular, CoCrNiMnFe HEAs, first reported by Cantor et al. [5], have been demonstrated to have excellent strain hardening capability, low yield stress, and exceptional ductility at cryogenic temperatures [6].

Although the application of HEAs to high-temperature environments has not been extensively studied, the unusual sluggish diffusion behaviour of HEAs is expected to provide better creep resistance as compared to commercial Fe–Cr-based stainless steels [7]. Therefore, HEAs could be promising materials for high-temperature applications, such as for the heat-exchange tubes of power-generating facilities. To guarantee the reliability of HEAs over long-term exposure at high temperatures, investigating their high-temperature oxidation is necessary. However, the oxidation behaviour of HEAs has not been extensively studied. Kai et al. [1] studied the oxidation behaviour of FeCoNi, FeCoNiCr, and FeCoNiCrCu equimolar alloys over the temperature range of 800–1000 °C in dry air. In their study, the ternary and quaternary alloys were single-phase, whereas the quinary alloy was two-phase. The oxidation behaviour and the characteristics of the scales that formed during the oxidation test were very different depending on the alloy composition. The oxidation resistance of the alloys generally followed the order of FeCoNiCr > FeCoNiCrCu > FeCoNi. Daborwa et al. [8] studied the influence of Cu content on the high-temperature oxidation behaviour of AlCoCrCuxFeNi high-entropy alloys (x = 0; 0.5; 1). In their report, the oxidation rate of the HEAs decreased with an increase in Cu content and the values of their parabolic constants were on the level similar to those observed in Ni–Al intermetallic alloys. Also, Huang and Yeh [9] reported the atmospheric oxidation of both AlSiTiCrFeNiMo0.5 and AlSiTiCrFeCoNiMo0.5 HEAs at 900–1100 °C. The alloys to exhibited a good oxidation resistance at T ≤ 1000 °C and the presence of Cr in the alloys was attributed to oxidation resistance. However, the detailed oxidation mechanism and scale constitution were not given.

Co–Cr–Ni-based equiatomic alloys, such as CoCrNiFe, CoCrNi, CoCrNiMn, CoCrNiMnFe, have been reported in the literature to possess a stable single-phase FCC crystal structure [10]. In addition, one of their unique properties is their ability to exhibit a high ductility and strength when temperature decreases down to 77 K [11]. In particular, the three-component CoCrNi alloy exhibits higher work hardenability and excellent ductility compared with the four-component or five-component single-phase FCC HEAs [12]. Comparisons of the oxidation behaviour of these alloys seem to be very interesting but have not been systematically studied yet.

In the present work, the high-temperature oxidation behaviour of CoCrNi, CoCrNiMn, and CoCrNiMnFe alloys were studied in dry air and the roles of Mn and Fe additions on the oxidation kinetics of the CoCrNi ternary alloy were examined.

Experimental

Three different alloys with nominal compositions of 33.3Co–33.3Cr–33.3Ni [low-entropy alloy (LEA)], 25Co–25Cr–25Ni–25Mn [medium-entropy alloy (MEA)], and 20Co–20Cr–20Ni–20Mn–20Fe (high-entropy alloy HEA) in at.% were produced by plasma arc melting. The alloys were then homogenised at 1000 °C for 24 h and cooled in the furnace. After, they were cold rolled to 50% reduction, followed by annealing at 800 °C for 2 h.

For thermogravimetric analyses of the high-temperature oxidation process, specimens with dimensions of 4 mm × 8 mm × 20 mm were carefully cut from the rolled plate. The oxidation test specimens were polished to a 0.3 mm finish. Prior to oxidation, the specimens were ultrasonically cleaned in acetone and methanol and dried. Oxidation experiments were performed using a SETARAM SETSYS Evolution thermogravimetric analyser. In order to determine the starting temperature, a heating oxidation test was carried out from room temperature to the target temperature at 5 °C min−1 heating rate in dry air. It was confirmed that noticeable oxidation behaviour started above 700 °C for all samples. Next, an isothermal oxidation test was carried out. First, a specimen was heated to test temperatures ranging from 800 to 1000 °C at a rate of 30 °C min−1 in protective Ar gas. When the temperature reached the test temperature, the samples were oxidised in dry air flowing at 16 mL min−1 for 24 h. The mass change of the samples was measured as a function of exposure time by an electronic balance technique with an accuracy of 0.01 mg. After the test, the oxidised samples were cooled down to room temperature at a cooling rate of 85 °C min−1.

The phase components of the alloys before and after oxidation were examined by a Rigaku D/Max-2500VL/PC X-ray diffractometer (XRD) with Cu Kα radiation at a tube voltage of 40 kV and a current of 100 mA. The characterisation of the oxidised specimens was conducted using a scanning electron microscope (SEM, JEOL JSM-5800) equipped with an energy-dispersive spectroscope (EDS). Aqua regia was used to etch the surfaces of the samples for optical microscopy.

Results

Microstructure

Figure 1 shows the microstructure of the alloys after homogenisation, cold rolling, and subsequent annealing. The figure reveals CoCrNi having an equiaxed grain structure with extensive annealing twinning, whereas CoCrNiMn has coarse grains. For CoCrNiMnFe, most of the grains are elongated. XRD patterns of the alloys are shown in Fig. 2. All three alloys have an fcc single-phase structure, with a lattice constant of 0.3567, 0.3601, and 0.3599 nm for CoCrNi, CoCrNiMn and CoCrNiMnFe, respectively. The highest lattice constant of CoCrNiMn is thought to be due to the high atomic fraction of Mn, which has a high lattice constant of 0.8913 nm. Although the three alloys have the same fcc structure, the 2θ values for each alloy are different depending on the compositional change. In particular, by adding Mn and/or Fe to CoCrNi, a slight shift of approximately 0.42° of the 2θ values to the left were observed.
Fig. 1

Optical microscopic images of a CoCrNi, b CoCrNiMn, and c CoCrNiMnFe

Fig. 2

XRD patterns of the alloys

High-temperature oxidation behaviour

The mass gain versus the exposure time curves of the three alloys over the temperature range 800–1000 °C for 24 h are shown in Fig. 3. According to this graph, the oxidation rate increased with increasing temperature for all test results. The oxidation curves of the three alloys followed the two-stage parabolic rate law, i.e. a fast initial stage followed by a steady-state stage.
Fig. 3

Mass-gain versus the exposure time curves for a CoCrNi, b CoCrNiMn, and c CoCrNiMnFe at 800, 850, 900, and 1000 °C in dry air for 24 h

CoCrNi exhibited the lowest oxidation rate among the alloys. It showed a negligible mass gain at 800 °C, which slightly increased at higher temperatures. The oxidation rates of CoCrNiMn and CoCrNiMnFe were noticeable, with CoCrNiMn alloy exhibiting the highest oxidation rate. It is well noted that, the addition of Mn had a major influence on the oxidation rates of the alloys. After 24 h of oxidation of the alloys, no serious spallation was observed. Generally, spallation of the oxide layer is mainly due to compressive stress in the oxide layer from the mismatch between the thermal expansion coefficients of the matrix and the oxide. In the present case, the oxide layers were not thick enough to produce the degree of compressive stress required for spallation.

The parabolic rate law [Eq. (1)] was used to describe the oxidation kinetics of the three alloys.

$$\left( {\frac{\Delta W}{A}} \right)^{2} = k_{{\text{P}}} t$$
(1)
where \(\Delta W\) is the mass gain, A is the unit surface area of the specimen, k is the oxidation rate constant, and t is the oxidation time. The oxidation rate constants were obtained from the slope of a linear regression line fitted on the mass gain per surface area versus time plot. The oxidation rate constants and their correlated activation energy values, calculated from the linear regressions at different temperatures and compositions, are listed in Table 1. Detailed interpretation of the oxidation rate constants and their activation energy values is given in the discussion section. Assuming that the parabolic rate constant follows the Arrhenius behaviour, \(k_{\text{P}}\) should have a temperature dependence of:
$$k_{\text{p}} = k_{0} \exp \left( {\frac{{ - E_{\text{a}} }}{RT}} \right)$$
(2)
where \(k_{0}\) is a pre-exponential factor, R and T are the gas constant and temperature, respectively, and \(E_{\text{a}}\) is the apparent activation energy of oxidation. Figure 4 shows the graph of 1/T versus \(\ln k_{\text{p}}\) for the studied alloys. The activation energies of oxidation (\(E_{\text{a}}\)) were calculated to be 400 kJ mol−1 for CoCrNi, 153 kJ mol−1 for CoCrNiMn, and 203 kJ mol−1 for CoCrNiMnFe. The difference in the \(k_{\text{p}}\) and \(E_{\text{a}}\) values of the three entropy alloys was probably due to diffusion and the different nature of scales that were formed on the alloys. Also, the presence of Mn is thought to have reduced the activation energy.
Table 1

Steady-state oxidation rate constants and apparent activation energies of low-, medium-, and high-entropy alloys

Alloys

Oxidation rate constants/g2 cm−4 s−1

Activation energy/kJ mol−1

References

Temperature/°C

Temperature range

800

850

900

1000

800–1000 °C

800–900 °C

CoCrNi

4.05 × 10−15

3.876 × 10−14

2.52 × 10−13

4.78 × 10−12

400

433

Present Work

CoCrNiMn

2.45 × 10−11

4.967 × 10−11

6.83 × 10−11

3.81 × 10−10

153

108

 

CoCrNiMnFe

5.69 × 10−12

9.337 × 10−12

2.12 × 10−11

1.92 × 10−10

203

137

 

FeCoNi

2.31 × 10−10

1.02 × 10−10

3.60 × 10−9

159

[1]

FeCoCrNi

1.95 × 10−13

1.77 × 10−12

4.08 × 10−12

[1]

3.09 × 10−11

9.86 × 10−11

6.92 × 10−12

1.70 × 10−11

FeCoNiCrCu

1.30 × 10−12

6.45 × 10−12

3.84 × 10−10

317

[1]

CrMnFeCoNi

1.6 × 10−11

4.1 × 10−11

130

[25]

Mn

125

[26]

Fig. 4

Arrhenius plots of 1/T versus ln \(k_{\text{p}}\) for temperatures at a 800, 850, 900, and 1000 °C and b 800, 850, and 900 °C

X-ray diffraction analyses of the oxide layers

The XRD patterns of the programmed alloys after 24 h oxidation at 800, 900, and 1000 °C are shown in Fig. 5. For all oxidation temperatures in the CoCrNi alloy, fcc and Cr2O3 peaks were observed. The Cr2O3 peaks intensify with an increase in oxidation temperature. This shows the growth of an external oxide scale. The excellent oxidation resistance of CoCrNi can be explained by the presence of Cr2O3, which is known to be a good passivation layer against further oxidation [13]. At 1000 °C, an additional peak is observed near 30°, which matches the peak of NiCr2O4 spinel. However, the volume fraction of NiCr2O4 spinel seems to be very low. Similar peak was observed during the study of the oxidation behaviour of AlCoCrFeNi (Al12 and Al15) at 1050 °C [14]. Mn2O3 is revealed as the major XRD peak for the oxidation of CoCrNiMn at 800 °C. At 900 °C, the Mn2O3 peaks disappear and Mn3O4 (hausmannite) peaks are observed. At 1000 °C, CrMn1.5O4 spinel peaks are revealed and the intensity of the Mn3O4 peaks increases. The peaks of the substrate (fcc) were not detected by XRD for all temperatures due to the thickness of the scales. For the CoCrNiMnFe alloy, the intensities of the fcc peaks decreased with increasing oxidation temperature from 800 to 900 °C and disappeared at 1000 °C. After oxidation at 800 °C, the oxide scales contained mainly Mn2O3. When the temperature increased to 900 °C, Mn3O4 peaks formed according to the XRD patterns. Finally, after oxidation at 1000 °C, the oxide scales were mainly composed of Mn3O4 (hausmannite) and Cr2O3.
Fig. 5

XRD analyses of scales formed at 800, 900, and 1000 °C on a CoCrNi, b CoCrNiMn c CoCrNiMnFe after 24 h oxidation

Oxide layer observations using SEM and EDS

Figure 6 shows the SEM images of the cross-sectional area of the studied alloys after 24 h oxidation. The thicknesses of the oxide layers of CoCrNi at 800, 900, and 1000 °C are 1, 3, and 6 μm, respectively. This indicated a good resistance to oxidation among the other alloys. CoCrNiMn and CoCrNiMnFe showed relatively thick oxide layers in the range of 6–31 μm as the temperature increased. Also, the interface between the oxide layer and the substrate for these two samples were very rough at 1000 °C.
Fig. 6

SEM micrographs of cross sections of the alloys after 24 h oxidation

Figure 7 shows the energy-dispersive X-ray spectroscopy (EDS) elemental maps of the cross section of the CoCrNi alloy at 800, 900, and 1000 °C. Table 2 shows the chemical compositions of the points identified in Fig. 7, as measured by EDS. The oxide scales consisted of a continuous single layer of Cr2O3, which were consistent with the XRD results. The thickness of the scales formed at 800 °C was negligible, which increased significantly with increasing temperature. EDS point analyses confirmed the oxide layer to be Cr2O3.
Fig. 7

SEM-EDS micrographs of the cross sections of CoCrNi after oxidation at a 800 °C, b 900 °C, and c 1000 °C for 24 h

Table 2

Chemical compositions of the points (at.%) identified in Fig. 7

Elements

800 °C

900 °C

1000 °C

Spectrum

Spectrum

Spectrum

1

2

3

1

2

3

1

2

3

Cr

25.3

27.9

31.0

20.0

29.90

33.5

20.7

23.2

22.6

Co

27.0

34.7

34.6

0.6

36.0

34.2

0

0.1

0.9

O

27.0

2.9

0

78.3

0

0

79.3

76.5

76

Ni

20.6

34.6

34.5

0

34.1

32.4

0

0.2

0.8

The EDS mapping images of the oxide layer of CoCrNiMn alloy at 800, 900 and 1000 °C are shown in Fig. 8. Table 3 shows the chemical compositions of the points identified in Fig. 8, as measured by EDS. The oxide scales were composed of Mn and Cr oxide layers for all temperatures, with Co and Ni barely detected. At 800 °C, the oxide scales consisted of an outer layer of Mn2O3 and a thin inner layer of Cr2O3. The inner layer has a Cr content of approximately 7.5 at.%. As the temperature increased to 900 °C, the outer layer is identified as Mn3O4. This oxide has a tetragonal crystallographic structure that is stable in the temperature range of 850–1150 °C, whereas the cubic Mn2O3 phase is stable up to 850 °C [15]. The thickness of the inner Cr2O3 layer increased with EDS point analysis, showing an increase in the Cr content to 16.9 at.%. However, this layer is not visible in the XRD patterns because of the thickness of Mn3O4 outer layer. It should be noted that the EDS measurements were taken on the sample cross sections, whereas XRD analysis was performed on the surfaces of the oxides. A discontinuous internal oxide layer consisting of Mn3O4 is seen beneath the Cr2O3 layer. After oxidation at 1000 °C for 24 h, the scales consisted of a thick continuous outer layer of Mn3O4 and CrMn1.5O4 spinel. The inner layer is composed of discontinuous Mn3O4.
Fig. 8

SEM-EDS micrographs of the cross sections of CoCrNiMn after oxidation at a 800 °C, b 900 °C, and c 1000 °C for 24 h

Table 3

Chemical compositions of the points (at.%) identified in Figs. 8

Elements

800 °C

900 °C

1000 °C

Spectrum

Spectrum

Spectrum

1

2

3

1

2

3

1

2

3

Cr

1.3

7.5

21.0

0.2

16.7

16.8

13.4

7.3

9.7

Co

0.1

0.2

19.1

0.1

0.2

4.3

0.8

2.9

0

O

71.3

70.9

35.9

68.1

66.3

59.3

71.7

60.9

67.2

Ni

0

0.2

16.9

0

0.2

3.4

0.3

0.9

0.4

Mn

27.4

21.2

7.2

31.6

16.4

16.2

13.8

28

22.4

The EDS analyses of the oxide scales of CoCrNiMnFe alloy for all test temperatures are represented in Fig. 9. It shows the presence of Mn and Cr in all temperatures and in some cases, small amounts of Fe, Ni, and Co. Table 4 shows the chemical compositions of the points identified in Fig. 9, as measured by EDS quantitative analyses. At 800 °C, the oxides scales were similar to the oxides in CoCrNiMn at 800 °C. The oxide scales consisted of two oxide layers: a continuous Mn2O3 outer layer and a Cr2O3 inner layer. The Cr2O3 layer was thicker than that of the CoCrNiMn. At 900 °C, the oxide scales were composed of an outer layer of Mn3O4 and an inner layer of Cr2O3. The dispersion of Mn caused several depleted zones in the base material. At 1000 °C, the oxides consisted of a thick layer of both Cr2O3 and Mn3O4. Both Mn and Cr were heavily dispersed and concentrated throughout the oxide layer.
Fig. 9

SEM-EDS micrographs of the cross sections of CoCrNiMnFe after oxidation at a 800 °C, b 900 °C, and c 1000 °C, for 24 h

Table 4

Chemical compositions of the points (at.%) identified in Fig. 9

Elements

800 °C

900 °C

1000 °C

Spectrum

Spectrum

Spectrum

1

2

3

1

2

3

1

2

3

Cr

2.8

14.4

14.9

4.0

13.0

19.0

12.1

10.3

14.0

Co

0

11.0

3.4

0.2

0.4

24.7

1.1

1.2

3.5

O

76.2

49

57.2

74.3

69.4

0

78.2

72.3

69.4

Ni

0.3

8.7

2.6

0.1

0.2

20.2

0.8

0.7

3.1

Mn

20.8

4.0

18.1

20.8

16.4

9.8

6.9

13.1

6.1

Fe

0

12.7

3.9

0.6

0.6

26.4

1.0

2.5

3.9

Discussion

The main aim of this research was to investigate the oxidation behaviour of CoCrNi, CoCrNiMn, and CoCrNiMnFe alloy, and in particular, to examine the roles of Mn and Fe addition on the oxidation kinetics. For all three alloys, the oxidation kinetics followed the parabolic rate law. This was a clear indication that solid-state diffusion was the rate-limiting step for the reactions. The oxidation rate constant generally increased with increasing temperature because of the rapid migration of elements through the scales.

CoCrNi displayed the highest resistance to oxidation among the three alloys for all temperatures. Among the elements in the low-entropy alloy, Cr had the highest diffusivity. The diffusivity of Cr, Ni, and Co in Ni matrix calculated at 1450 °C are 9.55, 4.21 and 6.05 × 10−13 m2 s−1, respectively [7]. Also, the diffusivity of Cr, Ni, and Co in Co matrix at 1490 °C are 2.69, 1.98, and 1.63 × 10−13 m2 s−1, respectively [7]. Therefore, during oxidation, there was the fast diffusion of Cr onto the surface of the alloy. The diffused Cr reacted with oxygen to form an outer nonporous layer of chromia (Cr2O3). This oxide layer has the tendency of preventing or reducing the inward diffusion of oxygen and/or the outward diffusion of cations (Ni and Co) [16]. Therefore, more Cr self-diffuses from the substrate and through the oxide layer as the temperature increased. Cr2O3 is stable in air up to 1100 °C [17] and hence can provide a long-term protection at higher temperatures.

Mn is a very important element in fcc single-phase HEAs because it can reduce stacking fault energies from 30–35 to 20–25 mJ m−2 [18]. Thus, the possibility of twinning at room temperature increases and a lower rate of dislocation annihilation is expected because of the difficulty in cross-slipping [19]. However, because of the high reactivity of Mn with oxygen, the addition of Mn to CoCrNi seemed to deteriorate the high-temperature oxidation resistance of the medium-entropy alloy (CoCrNiMn) and high-entropy alloy (CoCrNiMnFe). During the high-temperature oxidation, oxides of Mn and Cr were formed rather than those of Fe, Co, and Ni. This is because, the Gibbs free energy of Mn and Cr oxides is more negative than those of the other elements [20]. The Gibbs free energy of Cr2O3, which is − 1053 kJ mol−1, is lower than that of Mn2O3, which is − 881 kJ mol−1 [21]. However, the formation of Mn2O3 overwhelmed the formation of Cr2O3.

Tsai et al. [7] examined the diffusivity of Cr, Mn, Fe, Co, and Ni in CoCrFeMnNi HEAs at 1330 °C. In their results, Mn exhibited the highest diffusivity and the lowest activation energy of diffusion. The diffusivities of the various elements in CoCrFeMnNi are: 1.69 × 10−13 m2 s−1 for Cr, 2.12 × 10−13 m2 s−1 Mn, 1.30 × 10−13 m2 s−1 for Fe, 0.98 × 10−13 m2 s−1 for Co, and 0.95 × 10−13 m2 s−1 for Ni. Unfortunately, the diffusivities of the elements in the medium-entropy alloy were not reported. However, it is reasonably assumed that the high diffusivity of Mn may have accelerated the formation of Mn-rich oxide layer. Cr has the second fastest diffusivity among the elements. If the diffusivity of Cr had far exceeded that of Mn, then a passive Cr2O3 layer might have formed on the top surface, resulting in improved oxidation resistance of the alloy. However, in reality, Cr2O3 was rather formed below Mn2O3. The total Cr contents in the oxide layers of CoCrNiMnFe were found to be greater than in CoCrNiMn, even though CoCrNiMn had a high Cr content in the substrate than CoCrNiMnFe. This is believed to be due to the high diffusivity of Cr in Fe, which is 4.19 × 10−12 m2 s−1 [7]. The diffusivity of Mn in Fe, Co, and Ni is 2.17, 5.87, and 24.2 × 10−12 m2 s−1, respectively [7]. The addition of Fe to the CoCrNiMn system is believed to have slightly induced faster outward diffusion of Cr in the CoCrNiMnFe alloy. In Table 1, we can see that the k p value of CoCrNi increased by two orders of magnitude at 800 °C when Fe was added to form FeCoCrNi alloy, as reported by Kai et al. [1]. At 800 °C, both alloys exhibited the parabolic rate law during oxidation, and Cr2O3 was the main oxide that formed on surfaces of the alloys after 24 h. However, the Cr2O3 layer that formed on CoCrNi, as seen in Fig. 7, was very thin and could hardly be detected by EDS as compared to the Cr2O3 layer that formed on FeCoCrNi, which was reported to be dense. Therefore, it can be deduced that the presence of Fe in FeCoCrNi induced faster outward diffusion of Cr.

Xu et al. [22] investigated the roles of Mn in the high-temperature oxidation resistance of steels. The authors revealed that high Mn content stimulated the formation of CrMn1.5O4 coarse spinel by reacting with Cr. Their results are in agreement with the results obtained in the present study, where the outer oxide layer of CoCrNiMn at 1000 °C consisted of CrMn1.5O4 spinel and Mn3O4. Most of the diffused Mn formed Mn3O4 with O2 and the rest, forming CrMn1.5O4 with dispersed Cr in the layer. The formation of these oxides resulted in the severe depletion of Mn in the oxide–substrate interface, thus making the concentration of Mn beneath the scales insufficient to form a continuous oxide layer but rather, forms a discontinuous internal oxides of Mn3O4. For the CoCrNiMnFe alloy, the oxides that formed at 1000 °C were Cr2O3 and Mn3O4. These oxides were continuous in the outer layer and discontinuous in the inner layer. The Cr2O3 prevented or slowed down further oxidation or growth of Mn3O4. Holcomb et al. [23] studied the high-temperature oxidation behaviour of CoCrFeNi and CoCrFeNiMn HEAs. The authors reported the excellent oxidation resistance of CoCrFeNi than CoCrFeMnNi alloy. Mn was found to be more harmful, increasing the oxidation rate constant values than Cr was helpful reducing it. It was therefore concluded that, for CoCrFeMnNi HEAs to be used in high-temperature environments, less Mn content, whiles maintaining Cr levels, should be pursued. Hence, we can say that the reason for the higher oxidation resistance of CoCrNiMnFe than CoCrNiMn alloy is believed to be due to (1) less Mn content and (2) the formation of a more stable Cr2O3 phase.

From Table 1, the activation energy of oxidation of the alloys, calculated at temperatures of 800, 850, and 900 °C are 433 kJ mol−1 for CoCrNi, 108 kJ mol−1 for CoCrNiMn, and 137 kJ mol−1 for CoCrNiMnFe. Laplanche et al. [24] investigated the oxidation behaviour of CoCrNiMnFe at the temperature range of 500–900 °C and they found the activation energy of oxidation to be 130 kJ mol−1. This value was compared to the activation energy of oxidation of pure Mn, which is 122 kJ mol−1 at the temperature range of 400–900 °C [25], and the activation energy of diffusion of Mn in Mn oxides, which is 125 kJ mol−1 [26]. Their activation energy was very well comparable to both the activation energy of oxidation and diffusion of Mn. It was therefore strongly concluded that, the diffusion of Mn through the oxides was the rate-limiting process for the reaction. In our work, the activation energy values of CoCrNiMn and CoCrNiMnFe, calculated at 800, 850 and 900 °C, are well comparable to the activation energy of diffusion of Mn cations in Mn oxides and the activation energy of oxidation of pure Mn. Therefore, the upward diffusion of Mn through the oxides is considered to be the rate-limiting process for the oxidation reaction at the temperatures of 800, 850, and 900 °C. However, the activation energy values for the overall oxidation test (800–1000 °C) for CoCrNiMn and CoCrNiMnFe are 153 and 203 kJ mol−1, respectively. These values are greater than those calculated from 800 to 900 °C. The reason can be attributed to the heavy dispersion of Cr through the oxide layers of both alloys at 1000 °C, as seen from the EDS mapping images in Figs. 9 and 10. EDS point analysis data, as seen in Tables 3 and 4, show a high Cr content in the outer layer of both alloys at 1000 °C as compared to that at 800 and 900 °C. The Cr content in CoCrNiMn in the presence of more dispersed Mn formed CrMn1.5O4 spinel intermixed with already-formed Mn3O4. In the case of CoCrNiMnFe, the dispersed Cr content was higher than the dispersed Mn. This resulted in Cr forming Cr2O3, which intermixed with Mn3O4.
Fig. 10

Schematic diagram of the oxide scales formation of CoCrNi, CoCrNiMn, and CoCrNiMnFe

Based on the experimental results, a schematic diagram representing the oxidation mechanism of CoCrNi, CoCrNiMn, and CoCrNiMnFe are displayed in Fig. 10. The mechanism of oxidation is mainly based on the rate of diffusion of metal cations to the metal surface to form metal oxides with diffused oxygen and the diffusional growth of both Mn2O3 and Cr2O3 scales. The oxidation mechanism of Al–Co–Cr–Ni–(Fe or Si) HEAs as proposed by Butler et al. [27] ultimately depended on the diffusion of oxygen and the diffusion and concentration of Al and/or Cr within the alloy. In their result, alloy with high enough Cr concentrations but relatively low Al concentration formed Cr2O3 scale as the outer scale and Al2O3 as the internal subscale while alloys with high enough Al concentration but relatively low Cr concentrations formed Al2O3 as the major oxides scales and exhibited the lowest mass gains. In both system, the oxidation process were also determined by the diffusional growth of the external Cr2O3 or Al2O3 scale by Cr and/or Al.

In our experimental result, oxides of Mn and/or Cr were mostly formed due to their high diffusivity values. Increasing the temperature caused more cations to diffuse from the substrate through the oxide scales. This leads to the growth of the metal oxides layers. For CoCrNi alloy, Cr diffused rapidly to the outer surface and formed Cr2O3 with oxygen. An increase in temperature to 900 and 1000 °C caused more Cr to diffuse through Cr2O3 layer, which increased the thickness of the scales. Cr-depleted zones are seen just below the oxide–substrate interface. In the case of CoCrNiMn and CoCrNiMnFe, Mn diffused rapidly to the surface to form Mn2O3 with O2. Cr, with second highest diffusivity, also diffused to the surface and formed Cr2O3 beneath the Mn2O3 layer. As the temperature increased to 900 °C, Mn3O4 formed as the outer layer and the thickness of the inner Cr2O3 layer increased due to the continual diffusion of metal cations through the scales. This additional transport of Mn and Cr from the substrate to the oxide scales resulted in the depletion of Mn and Cr close to the oxide–substrate interface. Severe depletion of Mn occurred in CoCrNiMn as compared to CoCrNiMnFe such that the Mn concentration in the alloy near the oxide–substrate interface was insufficient to form a continuous oxide layer, thereby kinetically favouring the formation of internal discontinuous Mn3O4. At 1000 °C, the rate of diffusion of both Cr and Mn increased such that they were heavily dispersed throughout the scales of both alloys. The scales formed on CoCrNiMn consisted of an outer layer of CrMn1.5O4 and Mn3O4 and an inner layer of discontinuous Mn3O4, whereas the scales formed on CoCrNiMnFe consisted of Cr2O3 and Mn3O4.

Conclusions

The high-temperature oxidation behaviour of LEA, MEA, and HEA was studied at 800, 850, 900, and 1000 °C in a dry air environment. The oxidation kinetics and the influence of the addition of Mn and Fe were investigated. The findings are summarised below.
  1. 1.

    The oxidation behaviour of the three alloys followed the parabolic rate law over the temperature range of 800–1000 °C, with CoCrNi exhibiting the highest resistance to oxidation, followed by CoCrNiMnFe and CoCrNiMn recording the least resistance to oxidation.

     
  2. 2.

    The addition of Mn to CoCrNi increased the oxidation rate because of the high diffusivity of Mn. The oxidation resistance of CoCrNiMn was enhanced by the addition of Fe. Cr has a high diffusivity in Fe. Less Mn content and the formation of more stable Cr2O3 were responsible for the reduction in the oxidation rates of CoCrNiMnFe.

     
  3. 3.

    The activation energy of CoCrNiMn and CoCrNiMnFe at temperatures, ranging from 800 to 900 °C, was 108 and 137 kJ mol−1, respectively. These values are comparable to that of Mn diffusion in Mn oxides. Hence, the diffusion of Mn through the oxides from 800 to 900 °C is considered to be the rate-limiting process.

     
  4. 4.

    The diffusion of Cr increased at 1000 °C, thus forming CrMn1.5O4 spinel with Mn in the outer layer of CoCrNiMn and Cr2O3 in the outer layer of CoCrNiMnFe. The intense diffusion of Cr at 1000 °C is believed to have increased the activation energy of both CoCrNiMn and CoCrNiMnFe at 1000 °C.

     

Notes

Funding

This research was also supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (2015R1C1A1A02036622).

References

  1. 1.
    Kai W, Jang WL, Huang RT, Lee CC, Hsieh HH, Du CF. Air oxidation of FeCoNi-base equi-molar alloys at 800–1000°C. Oxid Met. 2005;63:169–92.CrossRefGoogle Scholar
  2. 2.
    Chou YL, Wang YC, Yeh JW, Shih HC. Pitting corrosion of the high-entropy alloy Co1.5CrFeNi1.5Ti0.5Mo0.1 in chloride-containing sulphate solutions. Corros Sci. 2010;52:3481–91.CrossRefGoogle Scholar
  3. 3.
    Hemphill MA, Yuan T, Wang GY, Yeh JW, Tsai CW, Chuang A. Fatigue behavior of Al0.5CoCrCuFeNi high entropy alloys. Acta Mater. 2012;60:5723–34.CrossRefGoogle Scholar
  4. 4.
    Wang F, Zhang Y, Chen G, Davies HA. Tensile and compressive mechanical behavior of a CoCrCuFeNiAl0.5 high entropy alloy. Int J Mod Phys B. 2009;23:1254–9.CrossRefGoogle Scholar
  5. 5.
    Cantor B, Chang ITH, Knight P, Vincent AJB. Microstructural development in equiatomic multicomponent alloys. Mater Sci Eng A. 2004;375–377:213–8.CrossRefGoogle Scholar
  6. 6.
    Gludovatz B, Hohenwarter A, Catoor D, Chang EH, George EP, Ritchie RO. A fracture-resistant high-entropy alloy for cryogenic applications. Science. 2014;345:1153–8.CrossRefPubMedGoogle Scholar
  7. 7.
    Tsai KY, Tsai MH, Yeh JW. Sluggish diffusion in Co–Cr–Fe–Mn–Ni high-entropy alloys. Acta Mater. 2013;61:4887–97.CrossRefGoogle Scholar
  8. 8.
    Dąbrowa J, Cieslak G, Stygar M, Mroczka K, Berent K, Kulik T, Danielewski M. Influence of Cu content on high temperature oxidation behavior of AlCoCrCuxFeNi high entropy alloys (x = 0; 0.5; 1). Intermetallics. 2017;84:52–61.CrossRefGoogle Scholar
  9. 9.
    Huang BP, Yeh J. Multi principal element alloys with improved oxidation and wear resistance for thermal spray coating. Adv Eng Mater. 2004;6:74–8.CrossRefGoogle Scholar
  10. 10.
    Wu Z, Bei H, Otto F, Pharr GM, George EP. Recovery, recrystallization, grain growth and phase stability of a family of FCC-structured multi-component equiatomic solid solution alloys. Intermetallics. 2014;46:131–40.CrossRefGoogle Scholar
  11. 11.
    Wu Z, Bei H, Pharr GM, George EP. Temperature dependence of the mechanical properties of equiatomic solid solution alloys with face-centered cubic crystal structures. Acta Mater. 2014;81:428–41.CrossRefGoogle Scholar
  12. 12.
    Zhao YL, Yang T, Tong Y, Wang J, Luan JH, Jiao ZB, Chen D, Yang Y, Hu A, Liu CT, Kai JJ. Heterogeneous precipitation behavior and stacking-fault-mediated deformation in a CoCrNi-based medium-entropy alloy. Acta Mater. 2017;138:72–82.CrossRefGoogle Scholar
  13. 13.
    Sugimoto K, Seto M, Tanaka S. Corrosion resistance of artificial passivation films of Fe2O3–Cr2O3–NiO formed by metalorganic chemical vapor deposition. J Electrochem Soc. 1993;140:1586–92.CrossRefGoogle Scholar
  14. 14.
    Butler TM, Weaver ML. Oxidation behavior of arc melted AlCoCrFeNi multi-component high-entropy alloys. J Alloys Compd. 2016;674:229–44.CrossRefGoogle Scholar
  15. 15.
    Speidel DH, Muan A. The system manganese oxide-Cr2O3 in air. J Am Ceram Soc. 1963;46:577–8.CrossRefGoogle Scholar
  16. 16.
    Sabioni ACS, Huntz AM, Philibert J, Lesage B, Monty C. Relation between the oxidation growth rate of chromia scales and self-diffusion in Cr2O3. J Mater Sci. 1992;27:4782–90.CrossRefGoogle Scholar
  17. 17.
    Smeltzer WW, Young DJ. Oxidation properties of transition metals. Prog Solid State Chem. 1975;10:17–54.CrossRefGoogle Scholar
  18. 18.
    Zaddach AJ, Niu C, Koch CC, Irving DL. Mechanical properties and stacking fault energies of NiFeCrCoMn high-entropy alloy. JOM. 2013;65:1780–9.CrossRefGoogle Scholar
  19. 19.
    Joo SH, Kato H, Jang MJ, Moon J, Tsai CW, Yeh JW, Kim HS. Tensile deformation behavior and deformation twinning of an equimolar CoCrFeMnNi high-entropy alloy. Mater Sci Eng A. 2017;689:122–33.CrossRefGoogle Scholar
  20. 20.
    Barin I, Platzki G. Thermochemical data of pure substances. 3rd ed. New York: VCH; 1995.CrossRefGoogle Scholar
  21. 21.
    Dean JA, Lange NA. Lange’s handbook of chemistry. 15th ed. New York: McGraw-Hill; 1999.Google Scholar
  22. 22.
    Xu X, Zhang X, Sun X, Lu ZP. Roles of manganese in the high-temperature oxidation resistance of alumina-forming austenitic steels at above 800 °C. Oxid Met. 2012;78:349–62.CrossRefGoogle Scholar
  23. 23.
    Holcomb GR, Tylczak J, Carney C. Oxidation of CoCrFeMnNi high entropy alloys. JOM J Miner Met Mater Soc. 2015;67:2326–39.CrossRefGoogle Scholar
  24. 24.
    Laplanche G, Volkert UF, Eggeler G, George EP. Oxidation behavior of the CrMnFeCoNi high-entropy alloy. Oxid Met. 2016;85:629–45.CrossRefGoogle Scholar
  25. 25.
    Païdassi J, Echeverría A. The oxidation of manganese in air at elevated temperatures. Acta Metall. 1959;7:293–5.CrossRefGoogle Scholar
  26. 26.
    Peterson PL, Chen WK. Cation self-diffusion and the isotope effect in Mn. J Phys Chem Solids. 1982;43:29–38.CrossRefGoogle Scholar
  27. 27.
    Butler TM, Alfano JP, Martens RL, Weaver ML. High-temperature oxidation behavior of Al–Co–Cr–Ni–(Fe or Si) multicomponent high-entropy alloys. JOM. 2015;67:246–59.CrossRefGoogle Scholar

Copyright information

© Akadémiai Kiadó, Budapest, Hungary 2018

Authors and Affiliations

  • Nana Kwabena Adomako
    • 1
  • Jeoung Han Kim
    • 1
  • Yong Taek Hyun
    • 2
  1. 1.Department of Materials Science and EngineeringHanbat National UniversityYuseong-gu, DaejeonRepublic of Korea
  2. 2.Titanium DepartmentKorea Institute of Materials ScienceGyeongnamRepublic of Korea

Personalised recommendations