An Investigation into the Nature of the Oxide Layer Formed on Kovar (Fe–29Ni–17Co) Wires Following Oxidation in Air at 700 and 800 °C
Oxid Met
An Investigation into the Nature of the Oxide Layer Formed on Kovar (Fe-29Ni-17Co) Wires Following Oxidation in Air at 700 and 800 C
Paul M. Yates 0 1 2
Christopher F. Mallinson 0 1 2
Phillip M. Mallinson 0 1 2
Mark J. Whiting 0 1 2
Julie A. Yeomans 0 1 2
0 Materials Science , AWE, Aldermaston, Reading RG7 4PR , UK
1 Department of Mechanical Engineering Sciences, The University of Surrey , Guildford, Surrey GU2 7XH , UK
2 Mark J. Whiting
This work provides new insight and evidence that challenges and extends the accepted view of the oxidation behaviour of Kovar (ASTM-15). Specimens of 2 mm diameter Kovar wire were oxidised in air at 700 or 800 C for 10 min. The resulting oxide layers were analysed by X-ray diffraction, X-ray photoelectron spectroscopy, scanning electron microscopy with energy-dispersive X-ray spectroscopy, scanning transmission electron microscopy and Raman spectroscopy. Oxide layers of approximately 2 and 4 lm thickness were formed at 700 and 800 C, respectively. These were found to contain iron, cobalt and traces of nickel. The combination of analysis techniques revealed that the oxide contains Fe2O3 in addition to (Fe, Co, Ni)3O4, a spinel oxide, in contrast to the combinations of Fe3O4, Fe2O3 and FeO that are typically reported. The oxide layer was found to be complex, consisting of multiple layers with different compositions, which is overlooked in the existing literature.
Raman spectroscopy; Spinels; X-ray photoelectron spectroscopy; (XPS); Scanning electron microscopy (SEM); Glass-to-metal seal
Introduction
The alloy Kovar, ASMT-15 (Fe–29 wt% Ni–17 wt% Co), is widely used in
glassto-metal sealing because of its electrical properties and in particular its thermal
expansion, which closely matches frequently used borosilicate glasses [
1
]. It has
been common practice, for more than 60 years, to pre-oxidise Kovar prior to sealing
&
to a glass so as to improve wetting and joining [
1, 2
]. However, the literature
regarding the oxidation mechanisms and products is limited in its depth and is often
contradictory.
Previous studies have used a range of oxidising atmospheres, such as air [
3, 4
],
air in closed tubes [4], dry air [
3, 5–7
], purified air [8], steam and carbon dioxide [
9
]
and nitrogen with water and hydrogen added [
10
]. The work presented here focuses
solely on air as the oxidising atmosphere.
A range of oxides are reported in the literature, with some authors reporting
various oxide layers composed only of iron oxides, for example Fe3O4 [
9, 11
],
Fe3O4, Fe2O3 and FeO [5] or FeO [
3
], whilst other authors report an oxide layer
containing iron, cobalt and nickel in a spinel structure [
8
]. The composition of the
oxide layer formed on Kovar is often assumed to be independent of oxidation
temperature [
11, 12
], and a wide range of temperatures have been studied in the
literature. There is, however, some evidence that the oxidation temperature affects
the type of oxide formed [
6–8
].
Characterisation of the oxide is usually performed by X-ray diffraction (XRD)
which has limited capabilities to differentiate between Fe3O4, a spinel, and other
similar crystal structures that have also been proposed to be present in the oxide
layer, such as CoFe2O4 which is also a spinel.
The presence of cobalt in the oxide may affect the bonding of Kovar to glass
because of the differences in the solubility of cobalt as compared to iron in a glass or
altered dissolution of iron from the CoFe2O4 spinel as opposed to Fe2O3. The
presence, or absence, of FeO in the oxide layer and potential for dissolution of FeO
into glass, may affect seal quality significantly because FeO is known to be a
network modifier in glasses. Crystallisation characteristics of the glass may also be
modified by dissolution of cations from the metal, which could be detrimental to the
seal quality. This study aims to explore a specific oxidation treatment relevant to a
common glass-to-metal seal and to characterise the oxide layer produced. The
production of the particular seals is discussed further elsewhere [
13
].
Experimental Procedures
Since the purpose of this study was to understand the oxide on Kovar (ASTM-15),
which was then going to be used in a glass-to-metal seal, the samples chosen for
investigation were 2-mm-diameter wire (purchased from Testbourne Ltd), which
was cut into straight 30 mm lengths, in the as-drawn state. In addition to these,
Kovar (ASTM-15) plates were also analysed. These were purchased from Future
Alloys as 50 9 50 9 2 mm sheets, which were polished using standard
metallographic techniques to a 1-lm diamond finish. All samples were cleaned using
acetone and isopropyl alcohol prior to oxidation.
For oxidation, the samples were placed in preheated muffle furnaces for 10 min,
then removed and placed on an alumina block to air-cool. The wire samples were
stood vertically so as to avoid damage to, or contamination of, the oxide. Some
samples were pr (...truncated)