The work identifies the conditions for thermodynamically favored spontaneous metal corrosion using potential-activity diagrams tailored for high-temperature molten fluorides. These diagrams provide insights into the thermodynamic phase stability of both solid and dissolved metal species, such as Cr, Cr(II), Cr(III), Ni, Ni(II), Fe, and Fe(II), along with their potential primary oxidizers, including Eu(III), O2, and HF, over a broad range of theoretical F− anion activities. The work further examines the practical implications, prospects, and challenges associated with the construction of these diagrams. The key objective of this project is to pinpoint crucial thermodynamic variables that substantially affect metal corrosion electrochemistry in the context of molten salt nuclear reactor applications.
In the past decade, there has been a notable increase in the publication of corrosion research focused on candidate metals and alloys for Gen-IV molten salt reactor applications.1 This trend highlights the growing interest in understanding and addressing the corrosion challenges inherent to this technology. Numerous studies have intentionally introduced oxidizers, including HF, CrF3, and EuF3, into the salt mixture to explore the impact of oxidizing impurities on the corrosion process, the driving force for corrosion, the rates of corrosion, and the underlying microstructural damage seen in candidate metals and alloys.2-3 However, these studies are often conducted phenomenologically4 and lack conceptualization of spontaneous corrosion processes explained by foundational thermodynamic and kinetics theories of corrosion.5-7 A step toward improving the thermodynamic foundations involves the construction and partial verification of potential-activity diagrams, similar to Pourbaix diagrams in aqueous systems. These diagrams can display the stability regions of species such as Cr, Cr(II), and Cr(III) in molten fluorides, as well as indicate the conditions necessary for the reduction of oxidizers and the occurrence of spontaneous corrosion.8 However, earlier works were constrained to the discussion of phase stability of Cr species, presuming them to be solvated to F− at 600°C. There is a gap in predicting and comparing oxidizer candidates in molten fluorides that can meet the conditions for spontaneous corrosion for a given metal. Specifically, which types of oxidizers, at what concentrations, and for which metals (e.g. Cr, Fe, and Ni) remain unexplored in many previous studies. Hence, the objective is to expand our original approach to define the phase stability of Ni, Fe, and other potential oxidizing elements, such as HF, O2, Eu(III), and Cr(III) when present in low finite concentrations. When considered together these help to define the conditions for spontaneous corrosion.
Table 2 reports some pertinent half-cell redox reactions, the standard electrode potential, , and the activity modified or Nernst equation considered in this work. Further discussions are given below about the application of potential-activity diagrams for analysis of the propensity for spontaneous corrosion in molten LiF-NaF-KF salts (or FLiNaK) with various impurities as the assumed predominant cathodic reactions. For thermodynamic reference, it is assumed that the of F− at 600°C is 0 and the partial pressure of all gaseous species (F2) to be 1 atm. It was noted that metal fluoride species are represented using the Roman numeral corresponding to the positively charged cation, e.g., Cr(II) as CrF2 and Eu(III) as EuF3.
RESULTS AND DISCUSSIONS
The lower potential boundary is set by the oxidation of metal salt cation components which will be assumed oxidized under all potentials. There are two different representations in the literature for determining the Nernst potential of salt metal cation reductions: one considers the metal fluoride compound (e.g., KF) as a singular entity, leveraging the more developed thermodynamic data for these species.10 However, this approach assumes a fluoride-dependent reaction, which remains a subject of debate and requires further experimental validation. The other assumes a fully ionized form (e.g., K+).6
At the vertical dashed line (representing F− activity of FLiNaK), a substantial difference in potential of roughly 500 mV between KF/K and K+/K is observed (Figure 1[a]). This implies that the K+/K reaction is thermodynamically more inclined to occur. However, this may not be the case for salts with a low F– activity. Given this comparison, K+/K will be considered as the lower potential boundary of FLiNaK salts in the subsequent discussion.
Figure 1(b) shows half-cell redox reactions involving gas generation and consumption if present. For instances, F− is reduced to form F2, HF to form H2, and O2– to form O2. Shaded regions represent stability fields dominated by the evolution of specific gas species. Concentrations of HF and O2– between 10 ppm and 1,000 ppm are considered. These are likely cathodic reactions when coupled with metal oxidation in molten salts. Other cathodic oxidations include the reduction of Cr(III) to Cr(II) or Ni(II) to Ni if NiF2 is present in the molten fluorides. For spontaneous metal dissolution, the potential should fall within these regions, typically aligned with H2 gas evolution. Metals with half-cell electrode potentials in the gray region will not corrode spontaneously unless an additional oxidizer like Eu(III) is present. The role of O2 as an oxidizer has not been investigated in detail due to uncertainties surrounding its solubility in FLiNaK at 600°C and the possibility for it to be oxidized by F− to form O2–.14
The reduction of higher valence cations, such as Eu(III) and Cr(III), which also act as oxidizers, is hypothetically capable of oxidizing all three elements Ni-, Cr-, or Fe-based on the imposed assumption. Additionally, for Cr, fluoride complexes such as CrF3− for Cr(II) and CrF63– in the case of Cr(III) may also form (and change the electrode potential) as suggested by both experimental and computational studies.8,17 The effect of fluoride complex formation on these Cr-F phase stability diagrams was investigated in our previous work.8
Our discussion herein of this work highlights some of the further research required to address various underlying assumptions. Factors such as activity coefficients, ratios between higher and lower valence states, fluoride complex formation, and the possibility of saturation all require thorough investigation. Moreover, substrate effects can change thermodynamic driving forces. Crystal orientation and nanocurvature can also affect the standard electrode potential.18-19 However, this work provides a foundation for understanding the spontaneity of metal corrosion in terms of phase stability diagrams. The authors’ latest work demonstrates the practicality of these diagrams in understanding experimental polarization curves.6 Beyond these insights, the work also establishes a thermodynamic framework that facilitates deeper comprehension of the contributory factors to corrosion, such as radiation.
This study utilizes potential-activity diagrams tailored for molten fluorides, designed to report in graphical form the driving forces behind various half-cell redox reactions forming a spontaneous corrosion cell involving one or more anodic and cathodic half-cell reactions. Such diagrams enable a comprehensive understanding of phase stability for different metal species and their potential predominant oxidizers during corrosion (e.g., Eu(III), O2, and HF) assuming a hypothetical range of fluoride activities. However, the study recognizes certain assumptions embedded within this framework and those require further experimental validation. Moreover, the diagram represents an equilibrium condition and does not inform corrosion kinetics. This framework could potentially predict conditions for spontaneous corrosion and serves as a tool for devising strategies for corrosion control and mitigation in these challenging environments, contributing to materials sustainability in fourth generation molten salt reactor technologies.
This work was supported as part of FUTURE (Fundamental Understanding of Transport Under Reactor Extremes), an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences. H.C. acknowledges the National Science Foundation Graduate Research Fellowship Program under Grant No. #1842490. Any opinions, findings, conclusions, or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the U.S. Department of Energy nor the National Science Foundation.