Liquid metals power advanced nuclear energy systems

Abstract

The development of advanced nuclear energy systems, known for their cleanliness and sustainability, is a key strategy for achieving a low-carbon energy transition. Liquid metal (LM)-powered advanced nuclear energy systems demonstrate sustainability and environmental friendliness, as well as being irreplaceable in specific areas. This paper charts a comprehensive scene of applications, challenges, and prospects of LMs in advanced nuclear energy (fusion and fission). First, next-generation fission reactors that use LM coolants, such as sodium or lead, are currently under design and construction. However, the coupling mechanisms of multiphase and multiphysics interactions remain unresolved due to various challenges, including corrosion and lead-water interactions. Second, the exploration of new LM-cooled reactors should emphasize sustainable development while ensuring basic performance. Lastly, the unique properties of LMs, including efficient energy transport and tritium breeding, position them as crucial materials in fusion system design. However, surface characteristics and the magnetohydrodynamic (MHD) effect remain major technical challenges. LMs have already left their mark in nuclear energy and are expected to be an effective solution to overcoming the energy crisis.

Graphical abstract

Public summary

• •This paper reviews the applications, challenges, and prospects of liquid metals in advanced nuclear energy systems.
• •Key challenges arise from interactions between liquid metals and other substances.
• •Liquid metals advance nuclear energy toward cleaner and more sustainable development.

Introduction

Under the framework of the Paris Agreement, many countries have established corresponding “dual-carbon targets.”¹ In this context, the imperative of low-carbon transformation and sustainable energy development has become undeniable. Nuclear energy, as a clean energy source, plays a crucial role in the utilization of low-carbon energy systems..^2,3,4 In December 2023, 22 countries jointly issued the Declaration to Triple Nuclear Energy at the United Nations Framework Convention on Climate Change. The European Council has also formally included nuclear energy in the Net Zero Emissions Industry Act, signaling that nuclear energy is playing an essential role in the mitigation of global climate issues.

As shown in Figure 1, the integration of liquid metals (LMs) with advanced nuclear energy systems has gained increasing attention from scholars, with the number of related publications rising steadily over the past decade. Advanced nuclear energy systems, particularly those based on LM-cooled fourth-generation fast neutron reactors, present a promising avenue. These reactors can convert non-fissile isotopes, such as U-238 or Th-232, into fissile Pu-239 or U-233, thereby enabling nuclear fuel breeding and significantly enhancing the utilization rate of natural uranium.⁵ Additionally, fast reactors can transmute spent fuel, including actinides such as Am-241 and Cm-244, as well as long-lived fission products like Se-79, significantly reducing their radioactive half-lives and transforming waste into a valuable resource.^6,7,8 For example, Am-241 can be used for measurement and medical diagnosis.⁹ Consequently, advanced nuclear energy systems utilizing LMs can guide nuclear energy development toward a more sustainable future. Additionally, fusion energy represents a truly clean and sustainable energy source. Fusion reactors, which derive their fuel from seawater, have the potential to fully resolve humanity’s energy challenges. Energy is produced through the fusion reaction of deuterium and tritium, and the byproduct is helium, which is environmentally benign.^10,11,12

Figure 1.

Annual publication trends (2014–2024) on LM applications in advanced nuclear systems including the generation IV fission and fusion reactors

Green renewable energies such as wind and solar power are gradually becoming the backbone of human daily energy consumption. However, these energies are intermittent and unstable, and nuclear energy can make up for these shortcomings to jointly contribute to energy decarbonization and power peak regulation. In addition, LM-cooled reactors are easy to miniaturize and modularize.¹³ These compact nuclear energy systems have irreplaceable roles in specific applications. For instance, in the areas of harsh-environment energy supply, deep-sea exploration, and space base operations, they possess unique advantages that other energy sources cannot match.^14,15,16

To summarize, the advantages of nuclear energy include its low-carbon nature, sustainability, stability, and flexibility, as well as its irreplaceability in specific areas. LMs offer hope for developing advanced nuclear energy systems that meet these needs. The successful application of LMs in generation IV reactors has made them a popular research target, and they will undoubtedly contribute to the development of advanced nuclear energy systems and enabling efficient utilization of low-carbon energy.

LMs amalgamate properties from various cutting-edge materials, particularly excelling in their thermal transport capabilities. These features provide them with a unique advantage in facilitating the transfer and conversion of energy.^17,18 The significant energy output inherent in nuclear energy, whether through fission or fusion, distinguishes it from other energy sources. Consequently, this huge demand for energy transfer means LMs can fully exert their role in the sphere of nuclear energy.^19,20,21

LMs are typically characterized as metals with a melting point below 300°C, although there is not an exact definition. In a broad sense, LMs can be categorized into three primary groups: alkali metals, heavy metals, and metals with low melting points. Table 1 offers a comparative overview of the thermophysical properties of typical LMs.

Table 1.

Thermophysical properties of LMs at 0.1 MPa, 600°C

Heat transfer fluid	Tmin (°C)	Tmax (°C)	Cp (kJkg−1 K−1)	k (Wm−1K−1)	$\rho$ (kgm−3)	$\mu$ (mPa s)
Alkali metals22,23
Na-K	−12	785	0.87	26.2	750	0.18
K	64	766	0.76	34.9	705	0.15
Na	98	883	1.25	46.0	808	0.21
Li	180	1,342	4.16	49.7	475	0.34
Heavy metals22
LBE	125	1,533	0.15	12.8	9,660	1.08
Bi	271	1,670	0.15	16.3	9,940	1.17
Pb	327	1,743	0.15	18.8	10,324	1.55
Fusible metals22,23
Ga	30	2,237	0.36	50.0	6,090	0.77
In	157	2,072	0.24	47.2	6,670	0.75
Sn	232	2,687	0.24	33.8	6,330	1.01

Alkali metals exhibit enticing properties, including specific heat capacity, viscosity, and other thermophysical attributes, making them a subject of extensive research in the domain of energy transportation.²⁴ However, all alkali metals have a fatal flaw: they are highly reactive and have a high probability of steam or hydrogen explosion when exposed to water.²⁵ Heavy metals such as lead (Pb) and bismuth (Bi) are also evergreen in nuclear energy. In contrast to alkali metals such as sodium (Na), lead-bismuth eutectic alloy (LBE) offers a broader range of operational temperatures, higher densities, harder neutron spectra, and superior chemical stability.²⁶ Consequently, lead-cooled fast reactors (LFRs) constitute one of the primary research reactor types in the context of generation IV and small reactors. The final category encompasses low-melting-point LMs, typified by gallium (Ga)-based and Bi-based alloys. These materials boast unique attributes—they remain liquid at room temperature, exhibit non-toxicity, and possess favorable interface properties. Consequently, they have garnered considerable attention in fields like robotics and biomedicine.^27,28 Thanks to the exceptional thermophysical properties of LMs, their application in nuclear energy has also garnered increasing attention.²⁹

This paper provides a comprehensive review of the applications, challenges, and future potential of LMs as exceptional heat transport materials in advanced nuclear energy systems. The applications of LMs in three domains encompass millennium energy: generation IV reactors, new LM-cooled conceptual reactors, and fusion engineering as the ultimate energy source for humanity. This paper is organized as follows.

First, millennium energy: LM-cooled fast reactor details the role and challenges of LMs in generation IV advanced fission nuclear energy systems currently in the design and construction phase. New LM conceptual reactors introduces new LM-cooled fission nuclear reactors that are currently in the exploratory phase. The efficient energy transport of LMs may become an effective way to develop the ultimate energy of humankind—fusion engineering. Ultimate energy: fusion engineering examines the development and application of LMs in fusion energy systems, focusing on blanket and core components, including plasma-facing elements. Finally, prospects and conclusions presents conclusions and future perspectives.

Millennium energy: LM-cooled fast reactor

Researchers have been committed to applying LMs with excellent thermophysical properties to fission reactors. The Generation IV International Forum (GIF-IV) selected Na-based and Pb-based metals as candidate coolant materials for fourth-generation nuclear reactors.³⁰ Figure 2 outlines the global history of LM-cooled reactor construction and design^31,32, while Table 2 summarizes the design parameters of typical LM-cooled reactors.

Figure 2.

History of LM-cooled reactors constructed worldwide

Table 2.

The design parameters of typical LM-cooled reactors

Nations	Acronym	Coolants	Thermal power (MW)	Outlet temperature (°C)
China31,33,34	CLEAR-I	LBE	700	390
	CEFR	Na	65	530
Russia31	BREST-300-OD	Pb	700	535
	SVBR-100	LBE	280	477
Europe31	ALFRED	Pb	300	520
France34	Super Phenix	Na	2,990	545
United States31,34	WLFR	Pb	950	535

Nuclear energy systems impose significantly higher safety requirements compared to other clean energy sources. Even when these systems operate stably and efficiently, all potential accidents must be fully controlled, ensuring consequences remain within acceptable limits. The development of LM-cooled nuclear energy systems also follows the overall technical route from verification and optimization of basic performance to research on challenges and mitigation measures under safety issues and finally to engineering applications. Safety is an eternal topic and research hotspot for nuclear energy. Performance verification and enhancement, along with risk mitigation, are two inseparable and continuously integrated parts of LM-cooled reactor design. This section first examines research progress and trends in heat transfer of LM-cooled fast reactors (LMFRs) and then discusses key safety challenges and potential solutions.

Heat energy transfer of LMs in reactors

It is essential to identify an efficient heat transfer medium capable of operating at higher temperatures than water to achieve the higher economics and safety required by generation IV nuclear energy systems. On the one hand, the coolant outlet temperature directly influences the economics of a nuclear power plant. On the other hand, to ensure the integrity of the fuel element, it is crucial to expel the generated heat as swiftly as possible. The excellent thermophysical properties of LMs are considered to be highly promising coolant candidates.

LMs exhibit distinctive flow heat transfer characteristics due to their low Prandtl number. This means that the LMs are significantly different from water in the heat transfer. Early investigations into the flow and heat transfer characteristics of LMs primarily concentrated on Na and NaK eutectic alloys. NaK eutectic alloys, with a melting point of 12°C and a boiling point of 785°C at atmospheric pressure, have a significantly lower melting point than pure sodium. Moreover, NaK eutectic alloy has been demonstrated to exhibit a higher Nusselt number compared to Na,³⁵ which implies greater heat transfer capability and energy output. Therefore, NaK eutectic alloy is more suitable for various space reactors, including SNAP-10A, AFSPS, BUK, and TOPAZ-I, among others.^36,37,38 The disadvantages of NaK include its low boiling point, the production of radioactive Na-24, and its chemical reactivity, which is similar to all alkali metals.

Initially, researchers conducted numerous studies primarily focusing on the fundamental heat transfer capabilities of LMs as coolants. The Oak Ridge National Laboratory (ORNL) in the United States constructed a large-scale sodium flow experimental setup called FFM.³⁹ FFM-2A generated a substantial amount of validation data widely used for studying the flow and heat transfer characteristics of sodium in 19-wire-wrapped rod bundles. Operational and experimental data from PHENIX, a pool-type sodium-cooled fast reactor (SFR) in France, are also frequently employed as benchmarks.⁴⁰ On this foundation, Mohamed⁴¹ developed heat transfer correlations specifically for LBE, leveraging the existing heat transfer data of Na and NaK bundle flow. In order to precisely ascertain the flow heat transfer as well as the natural circulation capacity of Pb-based metals, bundle flow experiments with different geometries have been carried out at KIT,⁴² KTH,⁴³ University of Pisa,⁴⁴ and USTC.⁴⁵ The flow heat transfer characteristics of LMs in conventional situations have been fully investigated. Subsequent work focuses on validating and refining the applicability of the model for diverse flow channels and different metal fluids.⁴⁶

As shown in Figure 3A, it should be noted that oxidative corrosion leads to the formation of corrosion products within the LM, which then circulate with the coolant.⁵³ These impurities have a notable impact on the thermophysical properties of LMs.⁵⁴ Researchers have conducted preliminary investigations into the behavior of LMs and their impurities under multiphysics field conditions. For instance, Gladinez et al.^55,56 examined the formation of metal oxides and their interaction with flow heat transfer. Wang et al.⁵⁷ developed a multiphysics-coupled model to analyze the generation of impurities due to corrosion within LMs and their impact on fuel assembly performance. Ji et al.⁵⁸ proposed an integrated neutron-thermal-hydraulic material approach to assess the effects of oxidative corrosion products on fuel assemblies in lead-bismuth reactors. However, the subsequent migration and diffusion behavior of the corrosion products have not been addressed. Existing research efforts are fragmented and lack a comprehensive, systematic study that integrates the formation, dissolution, migration, deposition, and flow heat transfer of corrosion particles.

Figure 3.

LM corrosion and performance optimization

(A) Corrosion product impurities dissolve into LBE.⁴⁷

(B) Manifold microchannel for ultra-high heat flux cooling.⁴⁸

(C) Wire-wrapped fuel design in reactors.

(D) The ability of Ni patterns to enhance Ag wetting.⁴⁹

(E) Modifying the LM interface with phosphonic acid chemistry.⁵⁰

(F) Controlling the LM wettability with an fs laser.⁵¹

(G–I) LM-based nanofluid (G) preparation, (H) morphology, and (I) thermal conductivity as a function of CuGa mass ratio.⁵²

The increased viscosity and density of LMs lead to greater resistance to flow.⁵⁹ Consequently, much of the recent research has focused on optimizing heat transfer performance.^60,61,62 As shown in Figures 3B and 3C, the heat transfer of LMs can be enhanced by optimizing the flow channel design.⁴⁸ However, this approach has limited applicability in nuclear reactors. Instead, LM-cooled nuclear reactors frequently adopt a wire-wrapped fuel design to bolster both heat transfer and temperature uniformity.⁶³

LMs possess a significant characteristic known as high surface tension, which offers two ideas for enhancing heat transfer in these materials (Figures 3D–3I). Firstly, the high surface tension of LMs tends to make them less wettable, which hinders efficient heat transfer. Various approaches can enhance wettability, such as modification of the substrate material,^64,65 LM modification,^66,67 and adjustment of the external environment.^51,68 However, this is often contrary to the requirements of material compatibility. For instance, the stable oxide layer in LFRs, meant to mitigate corrosion, can negatively impact coolant wettability.⁶⁹ The second is the concept of LM-based nanofluids.^70,71,72 LMs are ideal base liquids because of their high thermal conductivity and high surface tension, which allows for the hosting of a larger volume fraction of nanoparticles. Khan et al. showed that the heat transfer coefficient for LM-based nanofluids exceeds that of conventional water-based nanofluids by a factor of over 3.41.⁷³ This scenario offers an attractive opportunity to improve the economic and safety aspects of existing LMFRs. Moreover, it contributes to the conceptualization of potential LMFR designs, such as Ga-cooled fast reactors. It is important to note that enhanced heat transfer often comes with an increase in pressure drop; this trade-off must be taken into account in the design phase. Additionally, the introduction of particulate matter raises the risk of corrosion deposition, and its impact on the performance of reactor fuel assemblies warrants further investigation. Furthermore, oxide particles and other impurities can significantly reduce the heat transfer performance of LMs.⁷⁴

To sum up, LMs serve as the medium for heat transfer between the heat source and heat trap in LMFRs. Extensive research has been conducted on heat transfer and flow models of LMs. The next urgent challenge to address is the impact of multiphase particles introduced into the reactor due to corrosion or the requirement for enhanced heat transfer. This issue involves a highly complex and multidisciplinary intersection.

Multiphase and multiphysics coupling behavior under security issues

In the nuclear energy industry, advanced reactor economics and safety are vital considerations. As a generation Ⅳ advanced nuclear energy system, the safety of LMFRs has not been thoroughly validated, slowing down the progress of commercial application.

Compatibility between LM coolant and structural materials is the main challenge for the development of metal fast reactors. As shown in Figure 4, in case of loss of control, accidents such as the rupture of heat transfer tubes due to corrosion may occur, accompanied by vigorous reactions between the LM and water or oxygen. It may also lead to metal solidification due to uncontrolled temperature fluctuations. The key to securing reactor safety is to understand the multiphase behavior and multiphysics coupling mechanism that contribute to accidents. These complex phenomena often pose challenges to multiple aspects of the 3C (control, cool, and contain) principles simultaneously, necessitating effective preventive and mitigative measures.

Figure 4.

Main challenges of LM-cooled reactors

(A) Materials compatibility.

(B) The interaction with water.⁷⁵

(C) Uncontrolled phase transition: boiling⁷⁶ and solidification.⁷⁷

The compatibility of LMs with structural materials

The unique corrosion properties of LMs, combined with the higher temperatures and stronger neutron irradiation in LMFRs, elevate material compatibility to a paramount concern. The deterioration of material properties stems from various factors, including LM corrosion (LMC),⁷⁸ LM embrittlement (LME),⁷⁹ and irradiation damage.⁸⁰ Corrosion is a unique and major material damage problem in LM fast reactors, so we focus on that.

The corrosion mechanisms and protective measures have been extensively studied.^81,82,83 Low-carbon nitrogen-controlled nuclear-grade stainless steel was recommended as the main material of SFRs.⁸⁴ It has been observed that, within the designated operational temperature range of SFRs (400°C–500°C), proper chemical control ensures that the mechanical properties of austenitic and ferritic steels remain unaffected by Na.⁸⁵ Consequently, the corrosion problem in SFRs is deemed a non-urgent concern. In contrast, corrosion in LFRs presents a significantly more severe challenge. It is primarily attributed to the higher solubility of steel in lead and LBE compared to that in sodium. Wetting is proposed as a key mechanism governing solubility and corrosion.⁸⁶

Referencing Figure 4A, two primary strategies exist to mitigate corrosion: first, improvement of the corrosion resistance of structural materials. The main methods to achieve this are material modification,^87,88 substrate optimization,⁷⁸ and using insoluble metals such as Nb and Mo.⁸⁹ Material modification typically takes the form of coatings and is often performed concurrently with substrate optimization. Common types include Al-containing coatings,⁹⁰ ceramic coatings,⁹¹ oxide coatings,⁹² and high-entropy alloy coatings.⁹³ It should be noted that high-temperature coating deposition techniques can alter substrate properties, necessitating further refinements in the coating process. Another issue is the durability of the coating. Studies have indicated that Nb dissolves in static Pb, while Mo exhibits more favorable behavior than Nb.⁹⁴ Similarly, Cairang et al.⁹⁵ also found that Mo has better antioxidant properties in oxygen-saturated LBE at 600°C, but the relationship of this mechanism to temperature and durability remains unresolved.

The other method is reducing the corrosive effects of Pb and LBE. The primary strategy is using gas corrosion inhibitors.⁹⁶ Recently, oxygen has gained recognition as an exceptionally effective non-metallic inhibitor. The control of the active oxygen concentration to facilitate the formation of a protective Fe₃O₄ oxide layer has emerged as a key research area.^97,98,99 However, the durability of Fe₃O₄ and controlling the uniform oxygen concentration are the main challenges. The specifics of the oxygen control technology will be further elaborated in the next subsection. The above methods require further consideration of corrosion resistance under complex multi-physical environments (irradiation, high temperature, flow heat transfer, and chemical).

The interactions between LMs and oxygen/water

When LMs undergo oxidation, solid oxides with much lower density but higher melting points than the LMs are often produced. The accumulation of these solid oxides can cause flow plugging and heat transfer degradation in the core or steam generator. In particular, excessive suspended Na₂O (with a melting point of approximately 1,275°C) can cause plugging in sodium metal flow regions.¹⁰⁰ Large quantities of PbO (with a melting point of approximately 885°C) were also responsible for causing core damage in a Soviet nuclear submarine.¹⁰¹

The primary methods employed for oxygen control in liquid Pb and LBE comprise gas-phase oxygen control,¹⁰² solid-phase oxygen control,¹⁰³ and electrochemical oxygen control techniques.¹⁰⁴ While gas-phase oxygen control techniques have seen considerable advancement, large mass transfer devices and the risk of leakage make gas methods unsuitable for large nuclear reactors. Researchers are increasingly focusing on the solid-phase oxygen control technique (PbO dissolution) due to higher mass transfer rates.^{103,105,106,107} Besides, the solid-phase approach circumvents issues such as solid oxide precipitation resulting from excessive oxidation and potential leakage of radioactivity. The PbO dissolution rate is positively correlated with the temperature and flow rate of LBE.¹⁰⁸ Oxygen diffusion and transport mainly receive the combined influence of convection and diffusion. Zhu et al.¹⁰⁹ found that the oxygen mass transfer coefficient peaks at 475°C were attributed to the faster increase in diffusive mass transfer compared to convective mass transfer at elevated temperatures. It is important to note that the development of oxygen control systems in reactors requires synergistic optimization of PbO dissolution and diffusion with coolant flow heat transfer and corrosion deposition. Electrochemical oxygen pumping is an emerging and clean technology that avoids the problems of leakage of radioactive gases and replenishment of solid oxides after depletion, but it is still in the exploratory stage.^110,111 Additionally, a fundamental challenge in oxygen control is the limited understanding of the non-uniform distribution of oxygen concentration, which warrants further investigation into oxygen transport properties in non-isothermal circuits.

Another concern is LM-water interactions. When the steam generator heat transfer tubes rupture (steam generator tube rupture [SGTR]) due to corrosion or vibration and other reasons, water will be in direct contact with the coolant in the form of a high-pressure jet. There will be a rapid release of a substantial amount of energy, resulting in violent fluctuations in pressure and temperature (Figure 4B illustrates the interactions of LMs with water).

In SFRs, SGTR results in a violent and highly exothermic chemical reaction, which generates various chemical byproducts, such as H₂, Na₂O, and NaOH. This reaction has the potential to lead to explosions and sodium fire.¹¹² Carnevali et al.¹¹³ observed explosion waves in the inert atmosphere, thereby disproving the earlier hypothesis of hydrogen burning in air. They proposed that the heating of sodium played a crucial role in causing the explosion. Observed sodium vapor fumes before each explosion provide further support for this explanation.^114,115 Furthermore, the pre-mixing phase of sodium and water remains poorly understood. Several studies have proposed that the vapor film between the two substances plays a critical role, but there is no consensus on its mechanism.^116,117 In LFRs, Pb and LBE are chemically inert to water, do not undergo exothermic reactions with water, and do not produce hydrogen. However, due to intense phase transition behavior, the consequences of SGTR still need to be studied in terms of system dynamic response, steam transport, pressurization, and overpressure release.¹¹⁸ Additionally, the interactions between Pb/LBE and water will promote the formation of oxides, potentially posing additional risks, which have not yet received much attention.

Violent pressure fluctuations and thermal fatigue effects due to rapid temperature changes seriously threaten the structural integrity of reactors.¹¹⁹ To investigate the interactions between Pb/LBE and water, several test platforms have been developed recently. For example, researchers of ENEA in Italy carried out a series of experiments called LIFUS5 to study the pressure evolution characteristics and verify the reliability of the SIMMER codes.¹²⁰ In addition, the Innovative Nuclear System Laboratory (INSL) at Shanghai Jiao Tong University designed and built a large-scale LBE-water interaction and jet test platform (LIJI).⁷⁵ A prominent feature of this facility is its incorporation of inner-cycle lead fume purification technology and a fully automated remote intelligent detection system. Remote operation further isolated the heavy metal hazards that occurred during the experiment. The LIJI system provides a standard for the safe protection of heavy metal experiments. Similar facilities also include LEST from Xi’an Jiaotong University¹²¹ and LEWIN from Westinghouse Company.¹²² However, none of the above studies explored the consequences of the accidents and how they were mitigated, which is due to the inability to directly observe the complex multiphase interactions within LMs in experiments. Zhang et al.¹²³ designed a sliced test section of double-sided glass on the LIJI platform and conducted small-scale visualization experiments. This was the first local visualization experiment to study the interactions between LBE and water.

The key point of SGTR is to clarify the fragmentation properties of the bubbles within the LM and whether the bubbles can enter the core. The challenge is the opacity of the LM, which prevents direct observation of LBE-water interactions and bubble migration behavior. As shown in Figure 5, several methods have been initially explored to observe the two-phase flow behavior inside LMs. These include indirect approaches, invasive measurement techniques (e.g., optical¹²⁷ and electrical¹²⁸ methods), and non-invasive imaging technologies such as radiography¹²⁴ and ultrasound.¹²⁹ In comparison, the main application of the radiological technique is imaging, but the resolution of rays is low due to excessive attenuation. Meanwhile, ultrasound technology is still being explored. There are two main types of ultrasonic applications (velocimetry and imaging). Wang et al.¹³⁰ measured the rise velocity of bubbles within gallium-based LMs based on ultrasonic velocimetry (UDV). Subsequently, Richter et al.¹³¹ developed the ultrasonic transit time technique (UTTT) to simultaneously obtain the velocity and trajectory of bubbles. The above measurement methods provide good ideas for the observation of the two-phase flow of LMs, but none of them can simultaneously satisfy the exquisite measurement of bubble size and velocity. Therefore, there is an urgent need to develop advanced measurement techniques for the two-phase flow of LMs. The INSL at Shanghai Jiao Tong University is developing advanced LM ultrasonic and fiber optic measurement technology. These AI-based reconstruction algorithms show promise for enabling real-time observation of gas-liquid (bubble) and liquid-solid (solidification) interfaces in modified living organisms.

Figure 5.

Measurement methods for two-phase flow of LMs (A) Measurements of two-phase flow of LMs. (B) High absorption characteristics of LBE metal for ray: X-ray,¹²⁴ γ ray,¹²⁵ and neurography.¹²⁶

Finally, it is worth noting that LMs react with oxygen or water to produce impurities, so reactors should develop online scavenging technologies for removing impurities from the coolant to reduce the risk of shutdown and economic losses.

Uncontrolled phase transition of LMs

In order to maintain the long-term stable operation of a reactor, it is crucial to ensure that the coolant remains in a molten state of a controllable and safe temperature range. One of the main advantages of advanced nuclear energy systems is that they can rely on passive heat transfer for long-term core cooling, even under extreme conditions. As a result, it is difficult to ensure reactor safety in the event of temperature destabilization, such as coolant solidification, boiling, or thermal stratification. Thermal stratification not only leads to thermal fatigue of structural components but also affects or delays the system’s transition to natural circulation, which is unacceptable.

In SFRs, the solidification of Na is rarely taken into consideration due to its low melting point of only 98°C. However, it is important to acknowledge the risk of sodium boiling since its boiling point (785°C) does not sufficiently eliminate this concern.¹³² Unlike water, sodium boiling occurs predominantly in an annular flow pattern but only at low flow rates.⁷⁶ The refined modeling of sodium boiling faces challenges primarily due to limitations in measurement techniques, leading to the absence of a fully developed theory to date.¹³³ Currently, the systematic program remains biased in the calculation of the two-phase pressure drop after boiling, which makes it difficult to converge in the transient state.¹³⁴ Consequently, the primary focus of research on sodium boiling modeling has been on improving numerical stability across various transient conditions.^135,136

In LFRs, the primary risks associated with the solidification of Pb and LBE encompass potential structural damage arising from mechanical stress due to thermal expansion,¹³⁷ as well as impaired heat transfer capability due to the blockage of coolant flow channels.¹³⁸ Recently, the Westinghouse designed the LEFREEZ facility to bolster lead-based reactor development.¹³⁹ Besides, KTH is planning natural-cycle-based experiments using the TALL-3D facility to investigate the solidification mechanism of LBE.¹⁴⁰ The SEASAME project led by the Center for Advanced Research and Development (CRS4) in Sardinia, Italy, in the framework of Horizon 2020, has also designated LBE solidification as one of its main topics.¹⁴¹ However, the public availability of data from these experiments remains relatively limited. As shown in Figure 4C, Ito et al.⁷⁷ employed pulsed neutron imaging to visualize the transient behavior of one-dimensional solid-liquid interfaces. However, there is currently no explanation for the trends and mechanisms governing the solidification front on a three-dimensional scale. Furthermore, the solidification mechanism of lead-bismuth alloys has been relatively understudied and lacks comprehensive, intuitive kinetic, thermodynamic, and phenomenological models. Additionally, the extent to which localized temperature increases resulting from solidification can promote remelting remains unclear. The crucial aspect of LM solidification lies in analyzing the thermodynamic conditions that prevent solidification, thereby providing design guidelines for the thermal insulation boundary of the reactor. Furthermore, solidification can lead to the core overheating, especially if it results in piping blockage. To mitigate this potentially serious accident, developing directional solidification methods is a promising approach. By ensuring that the solidification front primarily occurs along the axial direction of the pipeline, we can minimize the risk of pipeline blockage.

In order to enable the large-scale commercial operation of LMFRs, the validation of the safety objectives becomes imperative. The challenging environment within these reactors, along with the unique characteristics of LM, make the compatibility of the LM with materials the primary concern. Additionally, there are several other safety issues specific to LMFRs that require investigation. These include boiling and solidification, as well as the interactions of the coolant with oxygen and water. Reactors involve multiphase-multiphysics coupling, particularly under accident conditions, and thus, these issues must be addressed from a multidisciplinary perspective. For instance, corrosion involves particle-LBE multiphase flow and chemical reactions, while SGTR is accompanied by multiphase, multicomponent interactions and phase transitions among water, vapor, and LBE. Furthermore, the measurement of LMs poses a significant challenge. Lastly, it is imperative to propose prevention and mitigation measures for specific accidents.

New LM conceptual reactors

In addition to the LM coolants selected by GIF, other LMs with excellent thermal or neutron properties are also being explored for application in fission reactors. This section will provide a discussion on the roles played and challenges faced by LMs in conceptual reactors.

Compared to Na, Pb, and LBE, Ga has a significantly lower melting point (29.8°C), allowing it to remain in a liquid state at room temperature. In addition, Ga-based metals have a wider operating temperature range and higher thermal conductivity, so they are also considered as an emergency coolant in the event of a severe reactor accident.¹⁴² The excellent thermal properties of Ga-based metals suggest that reactors utilizing Ga-based metals as a coolant have lower operational requirements and enhanced safety.¹⁴³ Ga exhibits chemical inertness toward both air and water,¹⁴⁴ thereby offering prominent advantages over Na in terms of chemical properties and safety.

The potential application of Ga-based metals in reactors has attracted the attention of researchers.^145,146,147 Sawada et al.¹⁴⁸ conducted an investigation with a gallium-cooled liquid metallic-fueled fast reactor into the impact of various liquid plutonium (Pu)-based metal alloy fuels on the core characteristics of a gallium-cooled fast reactor. Despite Ga’s significant absorption cross-section for fast neutrons, they achieved a favorable neutron physics performance through a well-designed reactor core. Lee et al.¹⁴⁹ studied the feasibility of using molten Ga with suspended nanoparticles as a coolant for a reactor, and they found that nanofluids reduce neutron absorption cross-sections. In addition to its primary role as a coolant, Seo et al.¹⁵⁰ and Zamora¹⁵¹ also proposed the utilization of Ga for mitigating reactor core accidents to reduce the overall risk of the entire reactor system.

The main obstacle to using Ga as a coolant is its relatively large neutron absorption cross-section, approximately four times Na, resulting in a lower neutron economy. Proper alloying and nanofluid can reduce the neutron absorption cross-section, and only a relatively small amount of high-quality coolant is required in the reactor, so the use of Ga cannot be excluded solely based on its cross-section. The cost of Ga is a significant factor limiting its large-scale application in reactors. Ga is never found in its pure state. It often needs to be recovered as a byproduct of mineral processing since mining minerals solely for Ga extraction is not economically viable.¹⁵² The efficient extraction of Ga resources is a problem that must be solved when applying Ga to reactor processes.

In addition to Ga, bismuth-based alloys are also explored as potential coolants. Dahy et al.¹⁵³ investigated the effects of Sn and Cd in Bi-Pb-Sn-Cd alloy on its thermal and nuclear performance as a fast neutron reactor coolant. Regarding thermal properties, the melting point of Bi-Pb-Sn-Cd alloy is lower and decreases with increasing Sn content, thereby reducing the material corrosion caused by bismuth. In terms of neutron performance, although the presence of Cd increases the neutron absorption cross-section of Bi-Pb-Sn-Cd alloy in the thermal neutron energy range, its cross-section remains small in the fast neutron energy range, which does not significantly affect its application in fast neutron reactors. It is crucial to note that the irradiation of Bi metals produces the highly toxic substance Po-210, which is one of the challenges encountered in lead-bismuth-cooled fast reactors.¹⁵⁴ Currently, the exploration of new LM-cooled reactors by researchers is still in the stage of validating basic performance and has not yet shifted its focus to the environmental impact and energy sustainability implications of these reactors.^155,156,157

In summary, scientists have been exploring the potential of conceptual LMs as a coolant for fission reactors, focusing on maintaining a balance between neutron performance, thermal performance, cost, and safety. It is worth noting that the development of new LM-based advanced nuclear energy systems needs to focus on sustainable development, recycling of nuclear waste, and green energy systems.

Ultimate energy: Fusion engineering

The roles of LMs in fusion

Thermonuclear fusion engineering offers an endless supply of clean energy. Controlled nuclear fusion is also a vital component of China’s three-step strategic approach to nuclear energy development. China’s significant investment in fusion technology, both in Hefei and elsewhere, demonstrates that China is at the forefront of global nuclear fusion research.¹⁵⁸

As shown in Figures 6A and 6B, light elements like hydrogen are subjected to high temperatures and pressures, causing tritium and deuterium nuclei to undergo fusion reactions, thereby releasing vast amounts of energy. This process generates high-speed charged particles that interact with a blanket, heating it and facilitating tritium breeding. Subsequently, the heat transfer fluid carries the energy from the blanket to an external system, which is responsible for electricity generation. The helium ash produced during the fusion process is collected by the divertor. Additionally, it also accepts the heat and particle flow from the plasma. Hence, the tritium breeder blanket and divertor are the central energy conversion components in a fusion reactor.

Figure 6.

Applications of LMs in fusion reactor

(A) Fusion reactor structure and principle.

(B) The roles played by LMs in the fusion reactor.

(C) The operable temperature and vapor pressures of the candidate LMs.

One of the most pressing challenges facing fusion reactor power plants is addressing the heat load on plasma-facing components (PFCs), particularly the divertor. LMs are recognized as the most promising solution to this issue, thanks to their outstanding energy conversion performance and “self-healing” properties.¹⁵⁹ After extensive research, the Fusion Energy Sciences (FES) initiated the LMPFC program in 2019.¹⁶⁰ Furthermore, China, Europe, and other nations have also undertaken responsive research efforts.^161,162,163

For candidates of LMs in fusion reactors, the primary factors to consider include their melting and boiling points, abundance and cost in nature, vapor pressures, and chemical compatibility with other component materials. The operable temperatures and vapor pressures of the candidate LMs are given in Figure 6C. Li, Sn, Ga, and their alloys have been identified as potential candidates for LMs in fusion applications.^{164,165,166,167,168} The temperature of the flowing liquid Li limiter is typically maintained between 200°C and 350°C to ensure lithium liquefaction while minimizing its evaporation.¹⁶⁹ It is worth noting that, unlike Li and its compounds, all other LMs necessitate the presence of beryllium, a neutron multiplier, to sustain tritium cycling within the fusion reactor. The advantage of using Li-Sn, as opposed to Li or Li-Pb, as a material for PFCs lies in its relatively higher operating temperature and resistivity.

PFCs

PFCs must endure direct exposure to the high heat flux and particle flow from the plasma. Conventional solid metals have demonstrated limitations in power handling, including issues like melting of leading edges, cracking, and performance degradation caused by plasma interactions and thermal expansion.¹⁷⁰ Any coating has a finite thickness and will, therefore, eventually become saturated with injected hydrogen-like substances and absorbing impurities.

LM-based divertors were initially introduced in the UWMAK-1 design.¹⁷¹ Subsequently, several LMPFC concepts have been explored in the United States APEX project.¹⁷² These concepts involve using LMs like Li or Li-Sn as the working fluid and employing electromagnetic and inertial forces to confine the liquid against the rear wall while controlling the hydrodynamic flow pattern. The realization of LMPFC applications necessitates addressing the following technical challenges: the wetting, splashing, and other behaviors of LMs due to the magnetohydrodynamic (MHD) effect, corrosion of the LMs, and tritium removal from Li metal.¹⁷³ Various forms of liquid lithium walls have been developed, including static free surfaces, capillary porous systems (CPSs),¹⁷⁴ and dynamic wall systems like the free-falling LiMIT system,¹⁷⁵ along with the forced-flow FLiLi system.

The primary function of the lithium wall is to solve the heat transfer problem. In the ongoing United States LMPFC development program, researchers have conducted numerical simulations and analyses to explore the feasibility of an open surface Li divertor. Smolentsev et al.¹⁷⁶ demonstrated that the velocity distribution is more advantageous for removing high surface heat flux from the plasma under the influence of the MHD effect. However, the Li flow in the divertor can become unstable due to the development of surface waves. In order to manage a heat flux of 10 molecular weight (MW)/m² (heat flux sustained by the International Thermonuclear Experimental Reactor divertor), the heat transfer coefficient of the system usually needs to exceed 4,000 W/m²/K.¹⁷⁷ The enormous demand for heat treatment necessitates adequate wetting of all surfaces of the lithium walls.

Due to its higher surface tension, liquid lithium can be readily pumped to the plasma’s edge, leveraging the capillary effect of a mesh. The concept of CPSs has been implemented as poleward limiters in several medium-sized plasma confinement experiments, notably those conducted at T-11M, FTU, and TJ-II. A significant milestone was reached in 2005 when the static LMPFC was first tested in the tokamak’s first wall.¹⁷⁸ Subsequently, dynamic liquid lithium limiters, FLiLi and LiMIT, were developed and tested.¹⁷⁹ In 2016, FLiLi underwent an upgrade to enhance surface temperature control, which included improved surface erosion resistance and real-time gas cooling capabilities.¹⁶³ Static free liquid surface configurations are difficult to consider as candidates for LMPFC due to surface instability.¹⁸⁰ In contrast to static liquid walls, the most promising designs for future fusion reactors involve lithium walls that harness capillary action or incorporate dynamic flows. The CPS employs capillary forces to securely retain liquid lithium within its porous structure, thereby significantly mitigating instability and preventing droplet splashing.¹⁸¹ However, lithium deuteride reduces the self-replenishment flow rate of lithium, counteracting capillary action.¹⁸⁰ Furthermore, this configuration exhibits slower heat transfer rates, which may not suffice for managing high thermal loads.¹⁸² The other form of dynamic lithium wall utilizes electromagnetic and gravitational forces to retain the LM within the channel. However, there is concern about the temperature evolution leading to proper wetting of the dynamic lithium wall, as inadequate wetting can cause droplet splashing instability. Currently, the highest recorded lithium coverage on the surface of the restrictor stands at 97.32%.¹⁸³ The primary challenge in addressing dynamic lithium wall systems lies in understanding the behavior of the lithium free surface. In particular, open LM surface flows can experience dry out under high heat flux conditions, resembling the critical heat flux (CHF) phenomenon seen in pressurized water reactors. Thus, balancing efficient heat transfer with wetting stability is a pivotal issue for LMPFC, and surface engineering is a promising new approach to tackle this challenge.^184,185,186 In addition to wettability, the behavior of lithium wall surfaces, particularly the splashing of the LM, is a significant concern. In severe spattering cases, the supporting framework of the solid substrate can quickly make contact with the plasma, resulting in erosion, evaporation, sputtering, and even melting of these solid materials.

Among the three types of lithium walls, the CPS and dynamic walls have greater application possibilities, but there are still some technical difficulties. The first one is the need to solve the free surface behavior properties to prevent droplet splashing and maintain sufficient wetting. Additionally, the lithium wall needs to handle the heat flux sufficiently and persistently to cope with the real plasma.

It is also worth noting that the PFC is directly exposed to the plasma, and in the event of a plasma collapse, the lithium wall will be directly burned out. As shown in Figure 7A, the development of machine learning provides an effective means for plasma prediction and control.^191,192 For example, Zheng et al.¹⁹³ successfully predicted plasma collapse based on a deep parameter-based transfer learning approach. Seo et al.¹⁸⁷ went a step further to present a deep reinforcement learning technique to avoid disruptive tearing instability in a tokamak. However, the coupling between the plasma and the liquid lithium wall has not yet been addressed,¹⁹⁴ which necessitates further coordinated control studies.

Figure 7.

Challenges and advances of LMs in fusion

(A) Prediction with artificial intelligence techniques in plasmas.¹⁸⁷

(B) Diagram of a tokamak.

(C) Temperature distribution.

(D) Contour diagrams of pressure distribution in ducts.¹⁸⁸

(E) Visualization of instantaneous velocity and currents upon entering the magnetic field.¹⁸⁹

(F) The coupling phenomenon between MHD effect and buoyancy.^190.

Blanket and energy conversion

Another significant application of LMs is as a blanket in tokamak. As depicted in Figure 7B, the blanket serves as one of the central energy conversion components in fusion engineering. The design of the breeder blanket poses a substantial challenge in fusion reactor engineering due to stringent performance requirements in terms of heat load and neutron flux. The primary functions of the blanket encompass plasma inclusion, tritium breeding, and energy acquisition. Tritium self-sustainability can be achieved by utilizing neutrons to react with lithium, given that the chemical activity of Li is higher than that of Pb and Sn. Furthermore, lead can act as a neutron multiplier in the nuclear reaction. Thus, PbLi becomes the primary candidate for the LM blanket.¹⁹⁵

The advantages of an LM blanket include high geometrical adaptability, exceptional heat-carrying capacity, efficient tritium cycling, and the ability for online replenishment of LMs. Furthermore, liquid lithium boasts robust heat transfer properties capable of meeting the substantial energy transport requirements of fusion reactors.¹⁹⁶ These merits were among the reasons LMs were chosen as the primary candidate for the APEX program.¹⁷²

Table 3 overviews the primary LM blanket cladding concepts from various regions worldwide. LM blankets typically employ coolants such as helium, water, and LMs. There are typically three cooling configurations for the cladding: self-cooling,²⁰¹ separate cooling (utilizing helium or water),²⁰⁶ and dual coolant (combining LMs and helium).²⁰⁷

Table 3.

Partial LM breeding blanket concept in different countries

Nations	Blanket	Tritium breeder	Coolant	Outlet temperature (°C)
Europe197,198,199,200	WCLL	PbLi	water	325
	HCLL	PbLi	He	500
	DCLL	PbLi	PbLi/He	550/427
	A-DC	PbLi	PbLi/He	700/480
	TAURO	PbLi	PbLi	950
United States201,202,203	ARIES-AT	PbLi	PbLi	1,050
	ARIES-ACT1	PbLi	He	1,030
	HT DCLL	PbLi	He	460
	ITER TBM	PbLi	He	360
	TSLL	PbLi	PbLi	550
China204	SLL	PbLi	He	450
	DLL	PbLi	PbLi/He	700
	HTL	PbLi	PbLi/He	900–1,000
Japan205	FFHR	Flibe	Flibe	600

Despite the numerous advantages of LM blankets over conventional solid blankets, several technical challenges still impede their widespread application, necessitating urgent solutions.

Firstly, LMs generate MHD effects in high-magnetic-field environments due to their conductive nature, resulting in induced currents. These currents, in turn, give rise to Lorentz forces.²⁰⁸ As shown in Figure 7D, the pressure drop due to MHD effects, including the influence of the Lorentz force, can be reduced through the use of flow channel inserts (FCIs).^188,209 Furthermore, FCIs have also been applied to the dual coolant lithium lead (DCLL) cladding in the context of the European Union for the Development of Fusion Energy (EURO) project.²¹⁰ However, the LMMHD effect gives rise to electromagnetic flux coupling,^211,212,213 three-dimensional MHD phenomena (Figure 7E),^214,215,216 and magnetic convection (Figure 7F).^217,218,219 An in-depth understanding of the complex flow behavior induced by the MHD effect is necessary for advancing LM breeding blanket designs.

Secondly, in the case of a water-cooled lead-lithium blanket, interactions between LM and water may occur in the event of a breach or accident. ENEA conducted a series of experimental investigations on lithium-based metal-water interactions to assess potential risks using the LIFUS5/Mod3 facility.^220,221,222 This part of the study is similar to the previously discussed lead-water interactions. Notably, the interaction between lithium and water can produce hydrogen, potentially leading to a steam and hydrogen explosion. Data from LIFUS5/Mod3 indicated that chemical reactions have long-term effects. LIFUS5/Mod3 is a separate test facility, and an integral test facility, LIFUS5/Mod4, is being built at ENEA Brasimone to study phenomena under internal LOCA (loss of coolant accident) conditions.²²³ The latest goals include understanding the impact of chemical reactions and integrating them into safety analysis software.²²⁴

In conclusion, LMs are regarded as the most promising candidate for PFC materials in fusion reactors, primarily due to their exceptional energy conversion performance and self-healing properties. Additionally, their capability to serve as neutron multipliers positions them as a strong contender for blanket materials. However, there remain certain limitations in the application of LMs. Firstly, the study of wetting behavior and heat transfer to the first-wall material in a fusion reactor is a key area of research. The MHD effect, which impacts flow and heat transfer, is also a significant focus in LM blanket studies. Secondly, understanding the surface behavior of LMs, including aspects like droplet splashing and the interaction between LMs and plasma, represents a central technical challenge. Additional concerns encompass safety, particularly in accident scenarios involving lithium-water or lithium-fire reactions, the compatibility of LMs with structural materials, and the issue of LME. Therefore, the following steps should focus on intensifying fundamental research into LM’s interfacial behavior, the MHD effect, and the establishment of accident mitigation mechanisms.

Prospects and conclusions

Nuclear energy highlights the immense potential of LMs for energy conversion, especially in the context of large-scale energy development. Similarly, LMs are guiding nuclear energy toward a cleaner and more sustainable path. Specifically, it comprises both fission and fusion systems.

LM-cooled fission reactors offer unique advantages in sustainability, economy, and environmental friendliness, particularly due to their ability to transmute spent fuel, thereby converting nuclear waste into a valuable resource.²²⁵ Furthermore, the high boiling point of LMs enhances reactor safety and simplifies the reactor structure, paving the way for reactor miniaturization. The challenge of energy intermittency poses a significant hurdle for a future energy mix that will be dominated by renewable and clean energy sources. LMFRs present a new and attractive application: the integration of high-temperature LMs with heat storage materials, which enables heat recovery and energy storage within reactor systems.²²⁶ Additionally, the flexibility of LMFR and its combination with energy storage will promote the development of wind, solar, nuclear, and storage integration, contributing to the green and low-carbon energy transition.^227,228,229

The conventional flow heat transfer in LMs has been well studied. Future research should prioritize optimizing heat transfer for specific geometries and materials by employing a multidisciplinary, multiphysics coupling approach to examine how multiphase particles affect performance. However, challenges remain in overcoming the multiphase flow characteristics arising from corrosion, as well as in advancing measurement techniques to accurately capture these phenomena. Furthermore, the mechanisms underlying multiphase-multiphysics coupling, which are often triggered by safety accidents such as LM-water interactions and LMC, represent the core challenges limiting the application of LMs and are urgent problems that need to be addressed.

For the safety issues regarding LM-cooled reactors, a comprehensive technical path should be adopted, encompassing precaution, analysis, and mitigation measures. The most critical issue requiring precaution is corrosion, which should be addressed through the collaborative optimization of new materials development, oxygen control, and coatings. It is necessary to achieve precise oxygen regulation from static to dynamic conditions for specific materials under multiphysics coupling conditions. Furthermore, in-depth research should focus on coating fabrication techniques, particularly their integration with novel materials.

Reactors are complex systems that involve multiphysics coupling, and safety issues must be analyzed from a multidisciplinary perspective encompassing nuclear, thermal, mechanical, and material aspects. Moreover, due to the high temperatures, high pressures, corrosiveness, and opacity of LMs, conventional measurement techniques are often ineffective. Therefore, advanced measurement technologies are needed to analyze the multiphase evolution behavior in LMs. Ultrasonic measurement is a promising technique, but future research needs to address the challenges associated with high temperatures and image reconstruction algorithms.

To mitigate safety issues, corresponding measures should be established to ensure the containment of radioactivity and minimize accident consequences under extreme conditions, including the development of passive mitigation technologies that reduce human intervention. First, online purification systems to remove impurities and oxidation corrosion products from LMs should be developed. Filters and bubble adsorption techniques for impurity removal are attractive methods, but it is crucial to prevent bubbles from entering the core to avoid void-induced reactivity effects. The main mitigation targets for lead-water interactions are bubble migration, LM solidification, and the expansion, compression, and blockage of structural components by LM. Subsequent efforts should focus on developing passive, simple, stable, and reliable devices or technologies that reduce bubble capture by LM and promote bubble detachment. Regarding solidification, it is necessary to understand where it occurs and how the solidification front develops from an intrinsic mechanism perspective, a process that also relies on advanced measurement technologies.

When exploring new LMs as potential coolants, it is essential to consider neutronics, thermal performance, cost, and safety. Ideally, new LM coolants should address existing challenges such as corrosion, which requires ongoing efforts and testing.

LMs also present attractive applications in fusion reactors, specifically in PFCs and blankets, which will help accelerate the commercialization of fusion reactors. The excellent heat transport capacity of LMs can be coupled with the huge energy release of fusion. Lithium, in particular, offers the potential for tritium production and possesses the ability to capture impurities, presenting a notable competitive advantage in PFC and blanket materials. Nevertheless, the surface characteristics, MHD effects, and safety concerns associated with LMs remain primary technical challenges and areas of extensive research.

The above challenges should be addressed following an overall technical roadmap from mechanistic scaling experiment research and code development to in-tokamak testing. Mechanistic experiments should first verify and assess the wettability between LMs and structural materials and then analyze the flow and heat transfer characteristics under MHD effects. It is recommended to first conduct MHD scaling experiments using NaK metal, as it can achieve a dimensionless Hartmann number comparable to that inside the reactor under conditions of lower temperatures, smaller volumes, and weaker magnetic fields. Subsequent large-scale tests can employ target metals such as PbLi. Given the strong coupling among MHD effects, magnetic field variations, and LM flow conditions, developing multiphase, multiphysics simulation codes is essential. These codes should design non-uniform three-dimensional magnetic fields, thereby optimizing LM flow and heat transfer to avoid drying out. Integrating machine learning for real-time plasma prediction will enhance the accuracy of these codes for LM-facing component design. Finally, LM flow characteristics and surface evolution should be verified in a tokamak to gradually realize the application of LMs in fusion systems. Regarding safety, LM-related challenges in fusion (e.g., corrosion) mirror those in LM-cooled fission reactors. Shared solutions include advanced coatings, oxygen control, and material innovations.

Funding and acknowledgments

This work received support from the National Natural Science Foundation of China (nos. U20B2011 and 123B2086).

Author contributions

X.L. proposed the topic of the review and provided supervision. L.Z. drafted and revised the manuscript. C.D. and X.J. researched the data and proofread the figures and tables. All authors contributed to the manuscript and approved the final version.

Declaration of interests

The authors declare no competing interests.

Published Online: May 26, 2025