Corrosion Evolution of T91 Steel in Static Lead-Bismuth Eutectic Under an Oxidising Environment

This study investigates the corrosion mechanisms of T91 steel in static lead-bismuth eutectic under high-temperature oxidizing conditions, revealing that grain boundary attack driven by chromium and oxygen diffusion can be mitigated by a stable oxide scale, while unexpectedly forming a unique iron-enriched body-centered cubic surface layer.

The Gist

Imagine a nuclear power plant of the future, one that runs on super-hot liquid metal instead of water. This liquid is a mixture of lead and bismuth (LBE), acting like a super-efficient, molten "soup" that carries heat away from the reactor core. To hold this scorching soup, engineers use special steel pipes made of a material called T91.

However, there's a problem: just like how iron rusts when left in the rain, this steel can "melt away" or corrode when dipped in the hot liquid metal. The scientists in this paper wanted to figure out exactly how this happens when the liquid metal is exposed to oxygen (like air), especially at very high temperatures (700°C).

Here is the story of their discovery, broken down with some everyday analogies:

1. The Setup: A Hot, Oxygenated Bath

Think of the T91 steel as a high-performance athlete. The liquid metal is a hot bath. In the past, scientists studied what happens when this athlete takes a bath in "reducing" water (water without oxygen). But in this study, they added oxygen to the mix, simulating a more realistic, slightly "rusty" environment. They left the steel in this bath for different amounts of time: a short dip (70 hours), a medium soak (245 hours), and a long marathon (506 hours).

2. The Three Ways the Steel Got Hurt

The researchers found that the steel didn't just rot evenly. It attacked in three distinct patterns, like a virus spreading through a city:

• The "Highway" Attack (Intergranular Corrosion): Imagine the steel is made of tiny bricks (grains) held together by mortar (grain boundaries). The corrosion didn't eat the bricks; it ate the mortar first. The liquid metal and oxygen snuck into the cracks between the grains, traveling like cars on a highway. This happened quickly at first.
• The "Flood" Attack (Wider Area Corrosion): As time went on, the damage got worse. The "highways" became so wide that the water flooded the actual bricks. The corrosion spread from the cracks into the middle of the grains, eating away large chunks of the steel.
• The "Safe Zone" (Unaffected Regions): Surprisingly, some parts of the steel remained perfectly healthy, even after 500 hours. These spots had formed a perfect, invisible shield that stopped the liquid metal from touching the steel at all.

3. The Big Surprise: The "Ghost" Layer

The most shocking discovery was what formed on the outside of the steel.
Usually, when steel rusts, you expect to see a layer of iron oxide (rust). But the scientists found something weird: a layer that looked like pure iron, but with a specific crystal structure (BCC) that matched the steel underneath, not a typical rust.

The Analogy: Imagine you are peeling an orange. Usually, you expect to find the white pith (rust) under the skin. But here, the scientists found that the "skin" had turned into a layer of pure, solid fruit flesh that looked exactly like the inside, but was actually a new, strange layer formed by the steel losing its oxygen to the liquid metal. It was like the steel "shed" its rust to reveal a fresh, iron-rich coat.

4. The "Melting" of the Steel's Skeleton

Inside the steel, the structure is like a tightly woven basket (martensite). When the oxygen and liquid metal attacked, they stole a key ingredient called Chromium from the steel to build a protective wall (oxide) on the surface.

• The Result: Because the steel lost its Chromium, the "basket" structure became unstable. At the high heat of 700°C, the tight basket unraveled and turned into a loose pile of round balls (ferrite).
• The Metaphor: Think of it like a house of cards. If you pull out the central support (Chromium), the whole structure collapses and settles into a flat, stable pile. The steel didn't just rust; it physically changed its shape and personality.

5. The Cracks and The "Leak"

As the protective oxide layer grew thicker, it started to crack, much like a thick layer of mud drying in the sun.

• The Analogy: Imagine a thick coat of paint on a wall. As the wall expands and contracts with heat, the paint cracks. Once the paint cracks, the liquid metal (the "soup") leaks through the cracks, dives deep into the steel, and starts dissolving it from the inside out. This is why the corrosion eventually turned from "highway cracks" into a "flood."

6. The Lesson: The Perfect Shield

The most important takeaway is about the "Safe Zones." The steel only survived in the areas where a dense, continuous shield of Chromium and Silicon oxides formed.

• The Metaphor: Think of this shield as a raincoat. If the raincoat has holes or is patchy, the rain (liquid metal) soaks through and ruins the clothes (steel). But if you have a perfect, unbroken raincoat, the steel stays dry and strong, even in the storm.

Summary

This paper tells us that at super-high temperatures, T91 steel doesn't just rust; it undergoes a complex transformation. It tries to build a shield, but if that shield cracks, the liquid metal invades, steals the steel's strength (Chromium), and forces the steel to change its internal structure.

The Bottom Line: To make nuclear reactors safe and long-lasting, we need to ensure the steel can always wear a perfect, unbroken "raincoat" of oxide. If we can control the chemistry to keep that coat intact, the steel will survive the hot liquid metal bath for decades.

Technical Summary

1. Problem Statement

Liquid Lead-Bismuth Eutectic (LBE) is a promising coolant for Generation IV fast nuclear reactors due to its favorable neutronic and thermal properties. However, the corrosion of structural materials, specifically Ferritic/Martensitic (F/M) steels like T91, remains a critical bottleneck.

• The Gap: While corrosion mechanisms are well-studied at lower temperatures (<600°C) and in reducing environments, there is a lack of understanding regarding T91 behavior at high temperatures (>700°C) under oxidizing conditions.
• The Challenge: At 700°C, the formation of a stable, protective passivation layer (typically required to prevent corrosion) is thermodynamically difficult. Previous studies suggested that without a stable oxide, LBE penetrates the steel, causing severe degradation. This study aims to elucidate the corrosion mechanisms, microstructural evolution, and the role of oxide stability under these specific high-temperature, oxidizing conditions.

2. Methodology

• Material: T91 steel (Fe-9Cr-1Mo), supplied in a quenched and tempered condition (tempered martensite structure).
• Environment: Static Lead-Bismuth Eutectic (44.5 wt.% Pb, 55.5 wt.% Bi) at 700°C.
• Oxygen Control: The oxygen concentration was maintained between the equilibrium potentials of Pb/Bi oxides and Fe oxides to create a controlled oxidizing atmosphere (achieved by injecting water vapor).
• Exposure Times: Samples were immersed for 70, 245, and 506 hours to observe temporal evolution.
• Characterization:
  • SEM-EDX: Used for elemental mapping and line scans. Low accelerating voltages (5 kV) were employed to enhance spatial resolution and detect light elements like Oxygen.
  • EBSD (Electron Backscatter Diffraction): Used to analyze crystallographic phases, grain orientation, and misorientation to detect phase transformations (e.g., martensite to ferrite).
  • FIB-SEM: Used for higher-resolution verification of specific features.

3. Key Contributions & Novel Findings

The study challenges several established paradigms regarding T91 corrosion in LBE:

1. Discovery of an Fe-Enriched BCC Surface Layer: Contrary to previous literature which reports outer oxide layers as $Fe_3O_4$ (FCC) or $Fe_2O_3$ (Hexagonal), this study identified a Body-Centered Cubic (BCC) iron-enriched layer on the surface. This layer is likely ferritic steel rather than an oxide, formed via the reduction of initial iron oxides by inward-diffusing oxygen reacting with Cr/Si.
2. Mechanism of Martensite Decomposition: The authors demonstrate that localized Chromium (Cr) depletion near grain boundaries lowers the recrystallization temperature of the steel. At 700°C, this triggers the transformation of martensite laths into equiaxed ferrite grains, a process driven by the diffusion of Cr to the surface/oxide interface.
3. Evolution from Intergranular to Area Corrosion: The paper maps a clear progression: corrosion initiates along grain boundaries (intergranular), transitions to wider area corrosion as oxides crack and spall, and finally leads to deep LBE penetration.
4. Role of Oxide Stability: The study identifies that the stability and coherence of the Cr/Si-rich oxide scale are the deciding factors between passivation (no corrosion) and catastrophic failure (LBE penetration).

4. Detailed Results

A. Corrosion Morphologies

Three distinct patterns were observed, varying by exposure time and local microstructure:

1. Intergranular Internal Corrosion:
   • 70h: Oxidation follows fine martensite lath boundaries.
   • 245h/506h: Oxidation extends to larger prior austenite grain boundaries (PAGBs).
   • Composition: Grain boundaries are enriched with Cr and Si oxides (spinel-type).
2. Wider Area Corrosion:
   • Occurs when the protective oxide layer fails (cracks/spalls).
   • LBE penetrates deep into the grain interiors, dissolving Fe, Cr, and Si.
   • Cracking: Thermal expansion mismatch between the substrate and oxide scale leads to cracks, which act as channels for LBE ingress.
3. Unaffected Regions:
   • Areas where a continuous, dense, Cr/Si-enriched oxide layer formed successfully.
   • These regions resisted LBE wetting and penetration, despite the high temperature.

B. Microstructural Evolution

• Phase Transformation: In corroded regions, the original martensitic lath structure transformed into equiaxed ferrite grains. EBSD analysis showed a significant reduction in grain misorientation in these zones, confirming recovery and recrystallization.
• Cr Depletion: A distinct Cr-depleted zone was observed adjacent to the oxidized grain boundaries. This depletion is caused by Cr diffusing outward to form stable oxides.
• Surface Layer: The outermost layer was found to be Fe-enriched (>80 wt.%) with a BCC structure. This contradicts the standard assumption of an outer iron-oxide layer, suggesting a complex reduction mechanism where initial iron oxides lose oxygen to form internal Cr/Si oxides.

C. Mechanism of Progression

1. Initiation: Oxygen diffuses inward, reacting with Cr and Si to form internal oxides along grain boundaries.
2. Depletion & Transformation: Cr diffuses to the surface/oxide interface, depleting the underlying matrix. This lowers the recrystallization temperature, causing martensite to decompose into ferrite.
3. Failure: As internal oxides thicken, volume mismatch creates stress, leading to oxide cracking.
4. Penetration: Once cracked, LBE penetrates through the cracks, dissolving alloying elements and accelerating degradation from intergranular attack to bulk area corrosion.

5. Significance and Implications

• Material Design: The findings highlight that maintaining a continuous, adherent Cr/Si-rich oxide scale is the single most critical factor for corrosion resistance in LBE at 700°C.
• Operational Limits: The study suggests that simply controlling oxygen levels to form any oxide is insufficient; the oxide must be mechanically stable to prevent spallation.
• Predictive Modeling: The identification of the BCC Fe-enriched layer and the specific mechanism of martensite-to-ferrite transformation via Cr depletion provides new parameters for modeling corrosion kinetics in high-temperature liquid metal systems.
• Future Strategies: To improve T91 performance, future alloy designs should focus on optimizing Cr and Si content to ensure rapid formation of a stable oxide, and surface treatments to minimize oxide cracking.

In conclusion, this paper provides a comprehensive mechanistic understanding of T91 corrosion in high-temperature LBE, revealing that corrosion is a coupled process of diffusion, phase transformation, and mechanical failure of the oxide scale, rather than simple dissolution.