Oxidation of Metals

, Volume 86, Issue 3–4, pp 193–203 | Cite as

Internal Oxidation of Ternary Alloys Forming a High Oxygen Conductive Oxide

  • R. Mertel
  • C. H. Konrad
  • M. Terock
  • R. Völkl
  • U. Glatzel
Original Paper

Abstract

Five ternary alloys consisting of a noble base metal (Ni, Co, Fe, Cu) and two reactive metals (Zr + Y, Ce + Gd) being able to form a high oxygen ion conductive oxide were internally oxidized under low oxygen partial pressures. All alloys developed either a continuous yttria-stabilized zirconia phase or a continuous gadolinia-doped ceria phase behind the front of internal oxidation. A Ni–Ce–Gd alloy showed extraordinarily high internal oxidation rates of up to 120 µm2/s at 900 °C. High internal oxidation rates in these ternary alloys were not limited to low concentrations of the reactive metals. The type of the internal oxide phase was found to be more important for the internal oxidation kinetics than the noble base metal.

Keywords

Internal oxidation Diffusion YSZ GDC 

Introduction

Wagner [1] described the internal oxidation of a less noble metal in solid solution with a nobler base metal. With increasing content of the less noble metal, the rate of internal oxidation is supposed to diminish and the formation of an oxide scale is favored. If the less noble element is initially present in an intermetallic phase, Gesmundo and Gleeson [2] as well as Anžel et al. [3] predict the so-called in situ internal oxidation of this intermetallic phase and subsequent coarsening of internal oxide particles with increasing content of the less noble alloying metal.

Kloss et al. [4] observed in situ internal oxidation of a binary Ni–Zr alloy at 1000 °C under low oxygen partial pressure. They found that even minor additions of Y accelerate internal oxidation considerably. The theory of Gesmundo et al. [2] could not explain the behavior of these ternary Ni–Zr–Y alloys. Konrad et al. [5] later detected that internal oxidation rates of Ni–Zr–Y alloys are not dependent on the total Zr and Y contents but more so on the Zr/Y ratio. The internal oxidation kinetics of Ni–Zr–Y alloys varied with the Zr/Y ratio very much in the same way as the oxygen ion diffusivities of bulk yttria-stabilized zirconia (YSZ) ceramics vary with the zirconia/yttria ratio. In both materials, fastest kinetics are observed at the ration of about 9/1. Furthermore, the absolute oxygen diffusivities in bulk YSZ and internally oxidized Ni–Zr–Y alloys turned out to be very close. Another interesting observation was that the oxide phase behind the internal oxidation front in Ni–Zr–Y alloys forms a continuous and interpenetrating network. Konrad et al. [5] concluded that oxygen diffusion through the evolving oxide phases themselves is controlling internal oxidation in the Ni–Zr–Y system.

Fuhrmann et al. [6] too investigated binary Ni–Zr and ternary Ni–Zr–Y systems. Both systems showed a transition from internal to external oxidation at unusually high contents of the less noble elements. Binary Ni–Zr alloys exhibited a transition from internal to external oxidation at around 35 at% Zr, whereas in ternary Ni–Zr–Y alloys this transition occurred at even higher contents of Zr + Y of about 55 at%. The transition in ternary Ni–Zr–Y came along with a major decrease of oxidation rates due to the formation of a dense yttria scale.

The question arises whether the Ni–Zr–Y system is unique showing exceptionally fast internal oxidation or whether other ternary alloys being able to form a high oxygen conductive oxide like YSZ or gadolinia-doped ceria (GDC) [7, 8] behave similar. Therefore, ternary alloys M–Zr–Y and M–Ce–Gd were studied here.

Experimental Procedures

Five ternary alloys with near-eutectic compositions were arc-melted under argon from pure elements (>99.9 %) to ingots of about 100 g (Table 1). In M–Zr–Y alloys, the Y content reflects the optimal Zr/Y atomic ratio for fastest internal oxidation according to Konrad et al. [5]. In NiCe8Gd1, the Ce/Gd atomic ratio reflects the ratio for highest oxygen conductivity in GDC oxides [7]. The oxide volume fractions after complete internal oxidation of the reactive elements is expected to vary between 26 and 30 %. The compositions were verified by energy-dispersive X-ray (EDX, Thermo Noran System Seven) analysis and micro-X-ray fluorescence (µ-XRF, EDAX Orbis PC SDD) analysis of the as-cast alloys.
Table 1

The as-cast compositions of ternary alloys

Alloy

Noble base metal

at%

Reactive metal

Zr or Ce

at%

Minor reactive metal

Y or Gd

at%

Oxide content after internal oxidation

vol%

CoZr10Y1

89.1

9.5

1.4

29.8

CuZr9Y1

90.2

8.6

1.2

26.1

FeZr10Y1

88.8

9.8

1.4

29.1

NiZr9Y1

89.9

8.8

1.3

28.1

NiCe8Gd1

91.1

8.0

0.9

26.1

Samples 4 × 4 × 5 mm3 for oxidation tests were cut out of the ingots, polished with SiC paper grit 1200, and cleaned in ethanol. In order to inhibit the oxidation of the noble base metal, the oxidation tests were carried out under low oxygen partial pressures listed in Table 2. The oxygen partial pressures were given by the dissociation of water vapor \({\text{H}}_{2} {\text{O}}_{\left( g \right)} { \leftrightarrows }{\text{H}}_{2\left( g \right)} + 1/2{\text{O}}_{2\left( g \right)}\) at the temperature of the internal oxidation heat treatment according to Nanko et al. [9]. The water vapor pressure was set by 10 l/min Ar/2 % H2 carrier gas flowing through a water bubbler at 22 °C. The oxidation times were 1–3 h at 700 and 800 °C and 1 and 2 h at 900 °C.
Table 2

Oxygen partial pressures

T in  °C

700

800

900

p(O2) in Pa

4.1 × 10−16

1.2 × 10−13

1.3 × 10−11

Cross sections of the oxidized samples were embedded in phenolic resin with carbon filler, ground with SiC paper (until grit 1200), and polished with diamond paste (6, 3, 1 µm). Microstructures were investigated using light microscopy and scanning electron microscopy (SEM, 1540 EsB Cross-Beam, Zeiss) equipped with an energy-dispersive X-ray spectroscopy (EDX). The depths d of the internal oxidation fronts were measured on SEM micrographs. For every data point, five micrographs from different specimen areas were taken. The error of the mean penetration depth was ±25 µm. X-ray diffraction (XRD) measurements were carried out on bulk specimens with a D8 advance diffractometer (Bruker AXS, Germany) set up in Bragg–Brentano geometry with a copper anode (λCu Kα1 = 0.15418 nm), a Ge monochromator, and a 0.3° divergence slit in the primary beam. An axial Soller slit with 2.5° opening angle was inserted into the secondary beam in front of a 1D detector. Possible phases were identified with the aid of the PDF-4+ 2015 powder diffraction database from The International Centre for Diffraction Data.

Results

In the as-cast state, all alloys are either two or three phase (Fig. 1; Table 3.). In all alloys, peaks in XRD diffractograms could be associated to a solid solution phase with the noble base metal and an intermetallic phase formed by the noble base metal and the reactive metal as predicted by the binary alloy phase diagrams. Intermetallic phases, Co17Y2 and Ni17Y2, containing the noble base metal and the minor reactive metal are detected in CoZr10Y1 and NiZr9Y1, respectively. Rietveld pattern refinement for quantitative classification of phase fractions was not successful on the bulk samples.
Fig. 1

Backscattered electron SEM micrographs of the as-cast microstructures

Table 3

Phases present in the as-cast state according to XRD measurements and/or EDX analysis

Alloy

CoZr10Y1

CuZr9Y1

FeZr10Y1

NiZr9Y1

NiCe8Gd1

Phases

Co

Cu

Fe

Ni

Ni

Co11Zr2

Cu9Zr2

Fe2Zr

Ni5Zr

Ni5Ce

Co17Y2

 

Fe85Zr9Y6

Ni17Y2

 

Peaks in XRD diffractograms of FeZr10Y1 could be associated to (Fe) solid solution and Fe2Zr. However, for some peaks present in the diffractograms no corresponding phase could be found in the PDF-4+ 2015 powder diffraction database. On backscattered electron SEM micrographs (Fig. 1), a third phase is identified. EDX revealed a composition of approximately 85 at% Fe, 9 at% Y, and 6 at% Zr. Ternary Cu–Zr–Y and Ni–Ce–Gd phase diagrams could not be found in literature; however, EDX analysis (Fig. 4) indicates that Y and Gd are dissolved preferentially in the intermetallic phases.

During exposure to low oxygen partial pressures at 700–900 °C (Table 2), the reactive and the minor reactive metals of the intermetallic phases selectively oxidize internally. The internal oxidation fronts in NiZr9Y1 and FeZr10Y1 (Fig. 2) are comparatively flat, whereas NiCe8Gd1 (Fig. 3), CuZr9Y1 (Fig. 4), and CoZr10Y1 (Fig. 5) show irregularly shaped internal oxidation fronts seemingly correlated to the as-cast microstructure. Internal oxidation in all five alloys is apparently from the in situ type [2].
Fig. 2

Microstructure of FeZr10Y1 at the internal oxidation front after exposure to 700 °C

Fig. 3

Microstructure of NiCe8Gd1 at the internal oxidation front after exposure to 800 °C

Fig. 4

Microstructure and elemental distributions of CuZr9Y1 at the internal oxidation front after exposure to 800 °C

Fig. 5

Microstructure and elemental distributions of CoZr10Y1 at the internal oxidation front after exposure to 800 °C

In internally oxidized CoZr10Y1, CuZr9Y1, FeZr10Y1, and NiZr9Y1, monoclinic, tetragonal, and cubic YSZ [10] were detected by XRD (see Table 4). With the exception of CuZr9Y1, a Y2O3 phase was verified too in these alloys. Oxygen partial pressures of less than 10−17 Pa and 10−12 Pa are needed in order to prevent the oxidation of Fe at 700 and 900 °C, respectively [11]. Hence, a porous iron oxide scale is observed (Fig. 2) after exposing FeZr10Y1 between 700 and 900 °C to an atmosphere with the oxygen partial pressures of ~4 × 10−16 and 1 × 10−11 Pa, respectively. Peaks of the iron oxides FeO, Fe2O3, and Fe3O4 are found in XRD diffractograms.
Table 4

Phases in the alloys after internal oxidation heat treatment at 900 °C

Alloy

CoZr10Y1

CuZr9Y1

FeZr10Y1

NiZr9Y1

NiCe8Gd1

Phases detected by XRD

Co

Cu

Fe

Ni

Ni

mon-ZrO2

mon-ZrO2

FeO

mon-ZrO2

CeO2

cub-ZrO2

cub-ZrO2

Fe2O3

cub-ZrO2

Gd2O3

tet-ZrO2

tet-ZrO2

Fe3O4

tet-ZrO2

 

Y2O3

 

mon-ZrO2

Y2O3

 
  

cub-ZrO2

  
  

tet-ZrO2

  
  

Y2O3

  

mon-ZrO2 monoclinic (P21/c) solid solution mixture of ZrO2 and Y2O3, tet-ZrO2 tetragonal ((Fm3m) solid solution mixture of ZrO2 and Y2O3, cub-ZrO2 cubic (P42/nmc) solid solution mixture of ZrO2 and Y2O3

Molybdic acid preferably etches the metallic phase of the internally oxidized ternary M–Zr–Y and M–Ce–Gd alloys. On the etched cross sections, tridimensional structures of continuous oxide phases are observed (Fig. 6). Structure and size distributions are bimodal, and larger structures reflect former intermetallic phases transformed into a fine mixture of both an oxide phase and a solid solution phase of the noble base metal by in situ internal oxidation. The larger structures in Fig. 6 are subdivided into fine connected networks of oxide and porosity where the molybdic acid has dissolved the metallic phase.
Fig. 6

Secondary electron SEM micrographs of oxide phases in alloys CoZr10Y1, NiCeGd1, and FeZr10Y1 after internal oxidation at 800 °C. Cross sections of oxidized samples were etched selectively with molybdic acid leaving continuous oxide networks at the specimen surface

Plotting internal oxidation depths versus the square root of time, the experimental values generally fall on straight regression lines indicating parabolic internal oxidation kinetics (Figs. 7, 8, 9). In addition to results from this study, internal oxidation depths of binary alloys from literature (Stott et al. [12], Nagorka et al. [13], Konrad [14]) are shown in Fig. 8 for 800 °C and in Fig. 9 for 900 °C, respectively.
Fig. 7

Oxidation depths d as a function of oxidation time t at 700 °C

Fig. 8

Oxidation depths d as a function of oxidation time t at 800 °C

Fig. 9

Oxidation depths d as a function of oxidation time t at 900 °C

NiCe8Gd1 internally oxidizes fastest from the here investigated ternary alloys. Extraordinarily high internal oxidation depths of 1030 µm after 3 h at 800 °C and of 915 µm after only 2 h at 900 °C were observed. The alloys FeZr10Y1, NiZr9Y, and CoZr10Y1, all forming an interpenetrating YSZ network behind the internal oxidation front, oxidize comparably fast. CuZr9Y1 also forming an interpenetrating YSZ network however oxidizes considerably faster than the latter three alloys.
$$d = \sqrt {k_{p} } \cdot \sqrt t$$
(1)
Parabolic oxidation rates kp deduced from the internal oxidation depths according to Eq. 1 are given in Table 5. NiCe8Gd1 shows the highest internal oxidation rates, kp, of the investigated alloys of ~43, ~89, and ~121 µm2/s at 700, 800, and 900 °C, respectively.
Table 5

Internal oxidation rates, kp, of ternary alloys at 700, 800, and 900 °C in µm2/s

 

700 °C

800 °C

900 °C

CoZr10Y1

0.8

3.4

4.8

CuZr9Y1

5.6

21.9

63.6

FeZr10Y1

1.6

3.0

8.4

NiZr9Y1

2.7

8.3

21.4

NiCe8Gd1

43.3

89.2

120.8

Discussion

All investigated ternary alloys undergo exceptionally fast internal oxidation. Stott et al. [12] internally oxidized Ni–Al and Ni–Cr alloys in Ni/NiO Rhines packs [15]. They generally found faster internal oxidation with decreasing content of the reactive metal in the binary alloys in accordance with the predictions of Wagner [1]. The fastest internal oxidation was measured in ~1 wt% Cr- or ~wt% Al-containing Ni-based alloys. However, internal oxidation depths at 800 and 900 °C are very much smaller compared to all the here investigated alloys even though the contents of reactive metals in these ternary alloys are much higher (see Figs. 8, 9). Konrad [14] too found much slower internal oxidation of binary Ni–Zr alloys at 800 °C under low oxygen partial pressure (see Fig. 8). According to Nagorka et al. [13], internal oxidation of Cu-1 at% Zr and Cu-1 at% Y wires at 800 °C in Rhines packs proceeded faster than in the above-mentioned Ni–Al and Ni–Cr binary alloys, however still much slower than in the ternary alloys of this study (see Fig. 8).

Kloss et al. [4] first assumed that oxygen diffusion through the internally formed oxide phase itself plays an important role in the internal oxidation of Ni–Zr–Y alloys. They suggested that high ionic conductivity in YSZ ceramics is responsible for fast internal oxidation in noble metal-based alloys with the additions of both Zr and Y. Later, Konrad et al. [5] could explain such incredible fast internal oxidation in Ni–Zr–Y alloys by means of a model derived under the assumptions that oxygen diffusion through the oxides itself is the rate-controlling mechanism and the oxide phase behind the internal oxidation front is continuous, i.e., provides continuous diffusion paths to the surface.

These assumptions are fulfilled in the here investigated ternary alloys. Microstructure investigations (Fig. 6) prove that all ternary alloys in this study form continuous oxide phases behind the internal oxidation fronts in the temperature range from 700 to 900 °C. In Zr + Y containing alloys, a continuous YSZ is formed by internal oxidation and in the Ce + Gd containing alloy continuous GDC is observed. The highest internal oxidation rates of the NiCe8Gd1 alloy may be explained straightforwardly by the higher oxygen ion conductivity of GDC compared to YSZ, the oxygen ion conductivity in GDC is roughly one order of magnitude higher than that in YSZ [8, 16, 17].

From the ternary alloys forming a continuous YSZ phase, CuZr9Y1 oxidizes fastest internally. The study of Nagorka et al. [13] confirms our observation. In [13], the binary Cu-1 at% Zr and Cu-1 at% Y alloys show faster internal oxidation compared to binary Ni–Al, Ni–Cr, and Ni–Zr alloys (see Fig. 8). Even though the literature values are not consistent, oxygen diffusivities turned out to be generally much higher in Cu than in Ni. According to Narula et al. [18], the diffusivity of oxygen in copper DOCu is ~10−5 cm2/s at 900 °C, while the diffusivity of oxygen in Ni DONi is ~10−9 cm2/s at 900 °C according to Park et al. [19], i.e., orders of magnitude lower. In the Cu-based alloys, oxygen diffusion through the noble base metal may therefore play a more pronounced role in the internal oxidation kinetics than in Ni-, Co-, and Fe-based ternary alloys.

Conclusions

Ternary alloys containing two reactive metals being able to form a high oxygen conductive oxide are prone to exceptionally fast internal oxidation. Fast internal oxidation is even more expected if the internal oxides form continuous diffusion paths from the surface to the alloy interior. Internal oxidation of NiCe8Gd1 is much faster than that of CoZr10Y1, CuZr9Y1, FeZr10Y1, and NiZr9Y1. Internal oxidation kinetics of such ternary alloys are apparently more dependent on the oxygen conductive oxide that is formed through internal oxidation than on the noble base metal. Noble base metals with high oxygen diffusivities may further accelerate internal oxidation. Fast internal oxidation in these ternary alloys is not limited to low concentrations of the reactive alloying metals.

Notes

Acknowledgments

The financial support from the Deutsche Forschungsgemeinschaft (DFG) within the project GL 181/32-1 is gratefully acknowledged.

References

  1. 1.
    C. Wagner, Zeitschrift für Elektrochemie 63(7), 772 (1959) .Google Scholar
  2. 2.
    F. Gesmundo and B. Gleeson, Oxidation of Metals 44, 211 (1995).CrossRefGoogle Scholar
  3. 3.
    I. Anžel, A. C. Kneissl, L. Kosec, and A. Krizman, Zeitschrift für Metallkunde 88(8), 38 (1997).Google Scholar
  4. 4.
    B. Kloss, M. Wenderoth, U. Glatzel, and R. Völkl, Oxidation of Metals 61, 239 (2004).CrossRefGoogle Scholar
  5. 5.
    C. Konrad, L. Fuhrmann, R. Völkl, and U. Glatzel, Corrosion Science 63, 187 (2012).CrossRefGoogle Scholar
  6. 6.
    L. Fuhrmann, C. H. Konrad, R. Völkl, and U. Glatzel, Corrosion Science 94, 218 (2015).CrossRefGoogle Scholar
  7. 7.
    B. Steele, Solid State Ionics 129, 95 (2000).CrossRefGoogle Scholar
  8. 8.
    S. M. Haile, Acta Materialia 51, 5981 (2003).CrossRefGoogle Scholar
  9. 9.
    M. Nanko, M. Ozawa, and T. Maruyama, Journal of the Electrochemical Society 147(1), 283 (2000).CrossRefGoogle Scholar
  10. 10.
    C. J. Howard and R. J. Hill, Journal of Material Science 26, 127 (1991).CrossRefGoogle Scholar
  11. 11.
    H. J. T. Ellingham, Journal of the Society of Chemical Industry 63, 125 (1944).CrossRefGoogle Scholar
  12. 12.
    F. Stott, G. Wood, D. Whittle, B. Bastow, Y. Shida, and A. Martinezvillafane, Solid State Ionics 12, 365 (1984).CrossRefGoogle Scholar
  13. 13.
    M. S. Nagorka, C. G. Levi, and G. E. Lucas, Metallurgical and Materials Transaction A 26, 859 (1995).CrossRefGoogle Scholar
  14. 14.
    C. H. Konrad, R. Völkl, and U. Glatzel, Oxidation of Metals 77, 149 (2012).CrossRefGoogle Scholar
  15. 15.
    F. H. Rhines, Transaction of Metallurgical Society of AIME 137, 246 (1940).Google Scholar
  16. 16.
    S. P. S. Badwal and F. T. Ciacchi, Ionics 6, 1 (2000).CrossRefGoogle Scholar
  17. 17.
    H. Inaba, Solid State Ionics 122, 95 (1999).CrossRefGoogle Scholar
  18. 18.
    M. L. Narula, V. B. Tare, and W. L. Worrell, Metallurgical Transactions B 14, 673 (1983).CrossRefGoogle Scholar
  19. 19.
    J.-W. Park and C. J. Altstetter, Metallurgical Transactions A 18, 43 (1987).CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2016

Authors and Affiliations

  • R. Mertel
    • 1
  • C. H. Konrad
    • 1
  • M. Terock
    • 1
  • R. Völkl
    • 1
  • U. Glatzel
    • 1
  1. 1.Metals and AlloysUniversity BayreuthBayreuthGermany

Personalised recommendations