[Paper Review] Giant segregation transition as origin of liquid metal embrittlement in the Fe-Zn system

Reza Darvishi Kamachali, Theophilus Wallis+6 more|arXiv (Cornell University)|Apr 26, 2023

Corrosion Behavior and InhibitionMaterials Science29 references3 citations

TL;DR

This study reveals a giant Zn segregation transition at Fe grain boundaries (GBs) in the Fe-Zn system, driven by a magnetic miscibility gap and low Zn cohesive energy, which drastically increases Zn segregation (up to 60 at.%) at low temperatures. Using CALPHAD-integrated density-based modeling and HAADF-STEM validation, it demonstrates that this transition weakens GBs and promotes liquid phase formation, explaining the origin of liquid metal embrittlement (LME) and enabling preventive alloy and process design strategies.

ABSTRACT

A giant Zn segregation transition is revealed using CALPHAD-integrated density-based modelling of segregation into Fe grain boundaries (GBs). The results show that above a threshold of only a few atomic percent Zn in the alloy, a substantial amount of up to 60 at.\% Zn can segregate to the GB. We found that the amount of segregation abruptly increases with decreasing temperature, while the Zn content in the alloy required for triggering the segregation transition decreases. Direct evidence of the Zn segregation transition is obtained using high-resolution scanning transmission electron microscopy. Base on the model, we trace the origin of the segregation transition back to the low cohesive energy of Zn and a miscibility gap in Fe-Zn GB, arising from the magnetic ordering effect, which is confirmed by ab-initio calculations. We also show that the massive Zn segregation resulting from the segregation transition greatly assists with liquid wetting and reduces the work of separation along the GB. The current predictions suggest that control over Zn segregation, by both alloy design and optimizing the galvanization and welding processes, may offer preventive strategies against liquid metal embrittlement.

Motivation & Objective

• To identify the underlying mechanism of liquid metal embrittlement (LME) in Zn-coated steels, which remains a major failure risk in automotive manufacturing.
• To investigate whether Zn segregation at Fe grain boundaries (GBs) precedes and promotes LME, acting as a precursor to wetting, precipitation, or crack initiation.
• To determine the thermodynamic and atomic-scale origins of massive Zn segregation at GBs, particularly its dependence on temperature and alloy composition.
• To evaluate the impact of Zn segregation on GB cohesion and liquid phase formation, linking it to LME susceptibility.
• To develop a predictive framework for preventing LME through alloy and processing design by suppressing the segregation transition.

Proposed method

• Employed CALPHAD-integrated density-based Gibbs free energy formalism to model GB thermodynamics, incorporating bulk properties from the TCFE11 database.
• Used the equation $ G(T, ho,X_{Zn}) = X_{Fe}G_{Fe}(T, ho) + X_{Zn}G_{Zn}(T, ho) + ho^{2} riangle H^{B}(T,X_{Zn}) - T riangle S^{B}(T,X_{Zn}) $ to describe GB free energy as a function of density $\rho$, temperature $T$, and Zn composition.
• Mapped the GB phase diagram by computing miscibility gaps and segregation isotherms across temperatures, identifying a sharp transition in Zn segregation.
• Validated predictions with high-resolution HAADF-STEM, observing Zn-rich interfacial patches confirming phase decomposition at GBs.
• Performed DFT calculations to quantify the work of separation of a $\Sigma$5 GB as a function of Zn coverage, with and without magnetic ordering.
• Evaluated the thermodynamic barrier $\Delta G = G^L - G^{GB}$ for liquid phase formation at segregated GBs, comparing pre- and post-segregation transition states.

Figure 1: Grain boundary (GB) phase diagram. (a) The GB miscibility gap (circles) computed utilizing the CALPHAD-integrated density-based method. Thermo-Calc TCFE11 database was used. The miscibility gap of the parent $\alpha$ (BCC) bulk phase is shown for comparison in the background. The superscri

Experimental results

Research questions

• RQ1 What drives the abrupt increase in Zn segregation at Fe grain boundaries, and how does it depend on temperature and alloy composition?
• RQ2 How does the magnetic ordering in Fe influence the formation of a miscibility gap in Fe-Zn grain boundaries?
• RQ3 To what extent does Zn segregation weaken the grain boundary, and how does this affect the work of separation and crack initiation?
• RQ4 Can the segregation transition significantly reduce the energy barrier for liquid phase formation at GBs, thereby promoting liquid metal embrittlement?
• RQ5 Can the segregation transition be suppressed through alloy design or process optimization to prevent LME in Zn-coated steels?

Key findings

• A giant Zn segregation transition occurs at Fe grain boundaries, with Zn concentration in the GB increasing abruptly from low to up to 60 at.% at temperatures below ~623 K.
• The transition is triggered by a magnetic miscibility gap in the GB, confirmed by ab-initio calculations, and driven by the low cohesive energy of Zn.
• The required Zn content in the bulk alloy to trigger segregation decreases with decreasing temperature, highlighting the critical role of cooling stages in processing.
• Zn segregation reduces the grain boundary work of separation by nearly 50% at 50% Zn coverage, significantly weakening the interface and promoting crack initiation.
• The segregation transition drastically lowers the free energy barrier $\Delta G$ for liquid phase formation at GBs, facilitating interfacial wetting and LME.
• Experimental HAADF-STEM imaging confirms the formation of Zn-rich interfacial patches, providing direct evidence of the segregation-induced phase decomposition.

Figure 2: GB energy from DFT calculations. The energy of a $\Sigma$ 5 [100](013) GB is computed for various levels of Zn coverage (C: coverage), with and without the magnetic ordering effect. The black curves connect the GB energies for the end-members with no and full coverage. In the paramagnetic