Transient oxidation of Mo–Si–B alloys: Effect of the microstructure size scale

Abstract

The composition and phase constituency of Mo–Si–B alloys are known to be important parameters in determining the oxidation response. For three phase Mo + T₂ + Mo₃Si alloys with constant composition and phase constituency, it is observed that a refined microstructure scale provides superior oxidation resistance. The transient stage of oxidation is shortened and the recession of the alloy is decreased with microstructural refinement. In order to identify the phase interaction during the transient stage, oxidation of each of the three alloy phases, Mo, Mo₃Si (A15) and Mo₅SiB₂ (T₂) has been investigated separately. Quantification of the separate phase size distributions by image analysis was coupled with the individual phase oxidation response to evaluate the overall oxidation behavior and phase interaction effects. A kinetic model for oxidation of Mo–Si–B alloys is proposed that incorporates the key role of microstructure scale on the transient stage and provides guidance for microstructure design.

Introduction

The high temperature behavior of Mo–Si–B alloys has been widely studied in the last decade due to the attractive alloy properties of high strength, creep resistance and phase stability. The alloys derive their oxidation protection from the continuous outer layer of borosilica glass that develops upon high temperature exposure. The oxidation mechanism of three phase Mo–Si–B alloys composed of Mo solid solution [Mo(ss)], Mo₃Si (A15) and Mo₅SiB₂ (T₂) phases has been described by Parthasarathy et al. [1]. Two distinct stages have been identified during isothermal oxidation in dry air above 800 °C. Initially, there are concurrent reactions between the molybdenum and oxygen to form molybdenum trioxide, which starts to sublime above 475 °C [2], [3], and oxidation of the boron and silicon from the A15 and T₂ phases to form a silica and a borosilica product, respectively. The competition between these reactions corresponds to the transient stage. Above 750–800 °C the borosilica reaction product, for some compositions, becomes continuous and slows down the oxidation and recession of the substrate, which corresponds to the second stage, the steady-state. Between 650 and 750 °C the borosilica is not continuous and oxidation will lead to a complete pesting, or transformation into powder, of the samples.

Detailed observations on the early stage of the oxidation of a Mo–3Si–1B (wt.%) alloy were reported by Helmick et al. [4] for various temperature and gas flow conditions. In some cases they described the formation of channels and bubbles of gas in the borosilica layer. They observed the formation of these channels mainly at the grain boundaries of the alloy. The healing of these channels was suggested to correspond to the transition between the transient and steady-state stages of the oxidation.

The addition of B₂O₃ to SiO₂ glass decreases the viscosity. The experimental results of Yan et al. [5] on the viscosity of the glass as a function of temperature and B₂O₃ content indicate a sharp drop of about 3–4 orders of magnitude for a 5 wt.% addition at 1000 °C and a more gradual decline at higher addition levels. The direct consequence of reduced viscosity is an increased permeability of oxygen through the borosilica layer protecting the Mo–Si–B alloy. Consequently, the rate of mass variation during the steady-state stage increases with increasing boron to silicon ratio in the substrate [6].

The thermal cycling oxidation behavior of different Mo–Si–B alloy compositions was studied by Supatarawanich et al. [7] on samples with different microstructure size scales. It was reported that a larger amount of T₂ phase improved the oxidation resistance and that a composition close to eutectic (i.e. with a finer microstructure) was the most oxidation resistant. However, it is not clear from the results if the variation in oxidation behavior is mostly dependent on the alloy composition or the microstructure scale.

The aim of this work is to understand the role of microstructure scale on the oxidation behavior of Mo–Si–B alloys [in the three phase field Mo(ss), A15 and T₂] under controlled conditions, without the influence of composition variations. The objective is also to measure the influence of the microstructure size scale on both the transient and steady-state stages of oxidation as a basis for a microstructure-based kinetic model for the oxidation behavior. In order to understand how the three phases interact with each other during the oxidation of Mo–Si–B alloys it is first necessary to know how they react with oxygen independently of each other.

Molybdenum exhibits a complex oxidation behavior that is sensitive to process variables. Depending on the temperature, oxygen partial pressure and gas flow the products and the kinetics of oxidation will be different [2], [3], [8]. In air, below 475 °C, MoO₃(s) forms the oxide layer. MoO₃ starts to sublime (vapor pressure of oxide > 10⁻⁴ Pa) at temperatures as low as 475 °C mainly by the formation of (MoO₃)₃, while above 700 °C sublimation of the oxide becomes the major process controlling oxidation [3]. Between 475 and 700 °C oxidation is a competition between formation of an oxide scale and volatilization of the oxide. For this reason, molybdenum shows a rapid mass loss during oxidation at high temperatures (above 700 °C). Due to continuous sublimation of the oxide molybdenum does not exhibit a protective behavior. The oxidation is characterized by linear recession rates, which depend on the temperature, partial pressure of oxygen and gas flow rate.

The oxidation of pure Mo₃Si has not been studied extensively. However, Ochiai [9], [10] compared the oxidation resistance of pure Mo₃Si with Cr or Al alloyed Mo₃Si in air at 900 °C. The pure Mo₃Si sample lost mass in a linear manner without reaching a steady-state. The addition of Cr or Al to the Mo₃Si intermetallic was observed to improve the oxidation resistance. In order to develop a more complete perspective on the oxidation of Mo₃Si, new results are presented on the oxidation behavior at 1100 °C in air.

Oxidation of the T₂ phase was studied by Yoshimi et al. [11] between 700 and 1400 °C. The results of their work demonstrated that above 1000 °C oxidation develops in two steps, as for multiphase Mo–Si–B alloys, with transient and steady-state stages. They observed that after 24 h oxidation two distinct layers covered the alloy substrate. The outermost surface was identified as a borosilica layer containing some molybdenum solid solution inclusions. After oxidation the protective borosilica layer on the surface of the sample was found to have a lower boron to silicon ratio than in the substrate. While the overall mass loss is dominated by the volatilization of MoO₃, boron depletion in the glass indicates that B₂O₃ also volatilizes during oxidation, contributing to the mass loss. The second layer consists of a matrix of Mo(ss) with dispersed amorphous silica.

While there is some knowledge of the oxidation of the phase pure samples, previous work was conducted under different conditions so that direct comparisons are difficult. In the current study oxidation of the three phases was examined under the same conditions as for the Mo–Si–B alloys in order to identify their role during the transient and steady-state stages of oxidation. With this foundation, any changes in the alloy oxidation kinetics can be correlated with the variation in microstructure size scale.