1. Introduction
Fusion power, with its clean, safe, sustainable, and virtually limitless fuel supply, is regarded as the ideal solution for ultimate energy source of the world
[1],
[2],
[3],
[4]. In recent years, controlled nuclear fusion, represented by magnetic confinement (such as tokamaks and stellarators) and inertial confinement, has achieved milestone progress: the Experimental Advanced Superconducting Tokamak (EAST) successfully achieved steady-state operation for over 1000 seconds with plasma temperature up to 100 million K
[5],
[6], the tungsten-W Environment in Steady-state Tokamak (WEST) reached a key technological milestone by maintaining hydrogen plasma for 1337 seconds at 50 million K
[7],
[8], while the Inertial Confinement Fusion (ICF) program achieved a historic breakthrough in energy gain (
Q = 4.13)
[9],
[10]. These achievements signal that controlled nuclear fusion is accelerating its transition from scientific experimental reactors to DEMOnstration reactors and eventually to commercial fusion power plants. Major global economies have set the timeline for commercial nuclear fusion grid-connected power generation between 2040 and 2050
[11].
However, a major challenge standing in the way of this ambitious goal is the extreme environments faced by plasma-facing materials (PFMs)
[12],
[13],
[14]. PFMs generally defines the armor materials of the blanket, limiter, and divertor, that directly facing the plasma in fusion vessel
[15],
[16],
[17],
[18]: the first wall (
Fig. 1a and b), which serves as the lining of the blanket; the limiter (
Fig. 1c), a physical structure used to define the last closed flux surface (LCFS) that has largely been superseded by divertors in modern tokamaks like ITER, where it is used only during plasma start-up; and the divertor (
Fig. 1d), a sophisticated exhaust system designed to extract heat, impurities, and helium ash by diverting magnetic field lines to a remote target chamber. PFMs are exposed to extreme fusion scenarios combined with high heat flux, plasma, and high-energy neutrons
[12],
[13],
[19],
[20],
[21],
[22],
[23],
[24]. The first walls would face a steady-state heat flux of 2 MW/m
2, while the heat flux in the divertor target regions reaches as high as 10-20 MW/m
2. Complex plasma-surface interactions (PSI) would cause severe physical sputtering, chemical corrosion, and fuel retention. Irradiation by 14.1 MeV neutrons of the deuterium-tritium fusion reaction would cause transmutation of the elements and displacement damage approximately 200 dpa in a single year, and transmutation reactions. During transient events such as edge localized modes (ELMs), vertical displacement events (VDEs), plasma disruptions, runaways, the surface temperature of the material can surge to over 2000 K
[22],
[25],
[26],
[27],
[28],
[29],
[30],
[31]. Under these extreme coupled conditions, ensuring the structural integrity and the stability of the materials is the core challenge in PFMs research and development
[32],
[33].
Ideal PFMs should combine high thermal conductivity, high mechanical strength, and fracture toughness, excellent thermal shock and thermal fatigue resistance, as well as low activation and long-term radiation stability
[32]. Unfortunately, no single material has been able to fully meet all these requirements
[21]. Taking tokamak as an example, the PFMs currently under extensive research include low-Z elements such as carbon and beryllium
[35],
[36], which offer superior plasma compatibility, as well as high-Z elements such as molybdenum and tungsten
[37],
[38], which exhibit excellent high-temperature stability. The selection, replacement, and upgrading of these PFMs have played a pivotal role in the development of fusion devices and the successful operation of fusion devices. The EAST serves as a case in point
[39]: during the initial design and construction phases, stainless steel was selected as the sole plasma-facing material. Subsequently, the PFMs (upper/lower divertors and first wall) were upgraded to full-carbon materials in 2008, and the first wall was upgraded to a Mo alloy (Ti-Zr-Mo alloy) in 2012. In 2014 and 2021, the upper divertor and lower divertor were successively upgraded to full tungsten divertors. The upgrades to the PFMs in EAST have provided strong support for achieving long-pulse and steady-state operation
[5],
[6], leading to continuous major scientific and technological breakthroughs.
Tungsten is currently the standard reference material for PFMs
[37],
[40]. In 2023, ITER progressed into new baseline and decided to replace beryllium with tungsten as the first wall armor material
[4],
[22]. However, the limitations of tungsten are also obvious
[40]. Whether tungsten will be the final choice for PFMs in commercial fusion power plants remains to be seen
[4], and PFMs still have a long way to go before they can support successful commercial fusion. Researchers in the fusion community have made significant progress in controlled nuclear fusion and are currently at a critical juncture in the transition from experimental reactors to the DEMO reactor
[4],
[5],
[7],
[8],
[10],
[24],
[39]. It is therefore particularly important to review the history of PFM development, systematically summarize the current state of research on PFMs in service at various state-of-the-art tokamak devices, introduce promising PFM candidates, and present the results of evaluating these candidate materials in real fusion environments, which is the leitmotif of this paper.
2. Plasma-facing materials: the-state-of-the-art
Modern designs for commercial fusion power plants are primarily based on the deuterium-tritium (DT) fusion reaction, which yields an α-particle and a 14.1 MeV neutron, as shown in
Fig. 2. Alongside the high-energy neutron irradiation and high heat flux, PFMs should also withstand PSI, which have been recognized as a pivotal challenge in transitioning controlled fusion into a viable energy source since the beginning of magnetic confinement fusion
[1],
[41],
[42],
[43],
[44]. The PSI influence tokamak operations through several critical mechanisms: plasma-driven erosion limits the operational lifespan of PFMs and introduces impurities into the core, while deposition alters surface morphology and can lead to the long-term sequestration of tritium inventories, raising significant safety and fuel-cycle concerns
[14],
[45], as schematically illustrated in
Fig. 3.
2.1. Carbon & graphite materials
Early tokamak devices typically employed stainless steel for vacuum vessels and utilized either stainless steel or refractory metals, such as molybdenum and tungsten, as limiter materials
[24],
[35]. Taking the Princeton Large Torus (PLT) as a primary example, initial experiments revealed that sputtering from the high-Z tungsten limiter led to severe contamination of the plasma core
[35],
[41]. These partially stripped tungsten ions caused intense line radiation, cooling the plasma core and degrading confinement. To mitigate this, PLT replaced the tungsten limiter with nuclear-grade graphite in 1978
[35]. The low-Z carbon ensures a good plasma compatibility, and its exceptional high-temperature performance, characterized by superior thermal shock resistance, fatigue endurance, and the tendency to sublime rather than melt, has been proved highly successful
[36]. This transition pioneered the use of carbon-based PFMs, establishing them as the preferred choice for limiters, divertors, and first-wall components for decades
[35],
[36],
[46]. Notable examples include the Joint European Torus (JET) device, which progressively covered its Inconel inner wall with carbon fiber composite (CFC) tiles around 1994
[24], and the EAST, which upgraded to a full-carbon first wall and divertor in 2008
[39].
However, as fusion research progressed toward longer pulses and higher power, significant drawbacks of carbon-based PFMs emerged
[36],
[47]. Key issues include neutron irradiation damage, which induces dimensional instability and degrades thermal conductivity, leading to excessive surface temperatures. Furthermore, the combination of physical sputtering, chemical erosion, and radiation-enhanced sublimation results in substantial mass loss and the subsequent re-deposition of carbon. This re-deposition process traps tritium, resulting in a high tritium inventory that poses both a critical safety risk and a significant economic barrier to the fuel cycle. Consequently, these challenges have led to the gradual phasing out of carbon and graphite as candidate PFMs for the next generation of fusion devices
[4],
[36]. As of now, DIII-D and ADITYA Upgrade are the only few tokamaks that still using graphite and carbon fiber composites as first wall materials
[46],
[48],
[49],
[50].
2.2. Beryllium
The inherent limitations of carbon led to the consideration of beryllium as a PFM
[35]. As early as the mid-to-late 1980s, a few tokamak devices (such as UNITOR, ISX-B, and JET) selected beryllium as the wall material
[35]. JET even replaced its CFC first wall material with beryllium in 2009, continuing to use it until its decommissioning in 2024
[16],
[24]. The practical application of beryllium as a PFM in JET has provided valuable insights for other tokamak devices. ITER identified beryllium as the preferred candidate material for first walls as early as the initial design phase in the 1990s
[38],
[42]. Beryllium features high thermal conductivity, minimal contamination of the core plasma, and low radiation power loss; at the same time, beryllium has demonstrated experimentally verified exceptional oxygen gettering ability, which can significantly improve plasma purity and reduce the effective charge of the plasma
[22],
[35]. Compared to carbon, beryllium exhibits lower tritium retention
[23],
[51],
[52],
[53],
[54]. However, as the understanding of beryllium’s comprehensive properties deepens, its use in extreme fusion environments is constrained by multiple physical and engineering factors
[24],
[55],
[56],
[57]: it has a low physical sputtering threshold and high sputtering yield; the toxicity of beryllium imposes stringent requirements on its post-processing; beryllium has a low melting point and poor high-temperature stability, and it undergoes significant volumetric swelling, neutron-induced embrittlement, and degradation of mechanical properties under high-temperature neutron irradiation. Although its thermal stability can be improved to some extent through alloying or the preparation of intermetallic compounds (beryllides), the inherent brittleness of such materials remains difficult to overcome
[35]. ITER has progressed into new baseline since 2023 and decided to fully replace the first wall material from beryllium to tungsten
[22],
[23].
2.3. Tungsten
In the 1980s, only a few fusion devices utilized high-Z materials such as tungsten and molybdenum for limiters
[35]. Following the major upgrade of PLT device in 1978, which involved the complete replacement of the original tungsten-based limiters with carbon-based materials, the application of tungsten-based materials in the fusion field entered a prolonged period of stagnation
[35]. It was not until the early 1990s that tungsten regained widespread attention within the fusion community due to its excellent erosion resistance
[38]. After decades of development, far more is known in operating tokamaks with high-Z metals, and tungsten has become the reference standard for the most promising PFM in the fusion field
[37],
[58],
[59],
[60]. In its new baseline, ITER will assemble a full-tungsten first wall, including bulk tungsten and tungsten heavy alloy with tungsten-coated steel as back-up option for some locations
[4]. Tungsten possesses excellent comprehensive properties, including a high melting point, low saturated vapor pressure, high thermal conductivity, high sputtering threshold, high thermal stress resistance, low swelling and low tritium retention rate
[37],
[61]. However, the disadvantages of tungsten are also obvious
[30],
[31],
[40],
[62],
[63],
[64],
[65],
[66],
[67], including poor machinability, low recrystallization temperature (~1500 K), high ductile-to-brittle transition temperature (DBTT) (~600 K, which increases after irradiation), low oxidation resistance, the surface morphology change (erosion, fuzz formation, and cracking) due to PSI and the corresponding properties deterioration and increased fuel retention, and the potential loss of melt layer during transient events (i.e., ELM, VDE and disruption).
Fig. 4a shows the fuzz formation in tungsten when exposed to helium plasma.
Fig. 4b~d show the melting, cracking and recrystallization in tungsten caused by high heat flux.
Fig. 4e~g show the synergistic effects in heat and plasma on the surface morphology change of tungsten. The order of exposure to heat or plasma first, or heat and plasma simultaneously determines the ultimate surface morphology of tungsten
[68]. Under neutron irradiation, tungsten undergoes neutron embrittlement, as well as irradiation-induced transmutation, which leads to deterioration in its mechanical properties
[69],
[70]. Additionally, there is a mismatch in thermal expansion coefficients between tungsten and the heat sink materials, such as copper and stainless steel
[37]. The shortcomings of tungsten-based materials can be mitigated to some extent through the design of tokamak devices
[24],
[25],
[71],
[72],
[73],
[74],
[75],
[76],
[77]. The switch to tungsten first wall in ITER new baseline loses the oxygen gettering action of beryllium and its ability to purify the plasma in the early current ramp-up phase
[4],
[22]. An associated diborane glow discharge boronization system is introduced to mitigate such risk
[71]. While the PFMs community is working to propose solutions for some of these issues
[58],
[67],
[78],
[79]. Nevertheless, it remains undermined whether tungsten-based PFMs will be able to withstand the operating conditions in a fusion power plant.
2.4. Molybdenum
Molybdenum is a high-Z PFM that has been extensively studied alongside tungsten and has already been used in the divertors and first walls of various fusion devices. Although molybdenum (melting point 2895 K) has a lower melting point than tungsten (melting point 3687 K), the two metals have comparable thermal conductivities
[81]. Furthermore, the thermal conductivity of molybdenum is less affected by thermal decay in high-temperature environments, and it exhibits lower transmutation rates and hydrogen isotope retention rates compared with tungsten
[82]. However, pure molybdenum has a high DBTT ranging from room temperature to 425 K, a low recrystallization temperature (approximately 1300 K), poor oxidation resistance, and relatively poor high-temperature mechanical properties
[83]. These disadvantages can be improved to some extent through material modification strategies such as alloying or oxide dispersion strengthening (ODS)
[83],
[84]. For example, the classic titanium-zirconium-molybdenum (TZM) alloy successfully reduces the DBTT to 223-273 K, raises the recrystallization temperature to roughly 1600 K, and enhances its oxidation resistance, creep resistance, and high-temperature strength
[83]. TZM alloys have accumulated extensive operational experience in several tokamaks; for instance, the EAST tokamak device has applied TZM tiles as the primary PFMs since 2012
[39]. However, as a typical high-Z material, molybdenum still faces severe challenges like those of tungsten, including the degradation of surface morphology and macroscopic properties caused by intense PSI, as well as the need to improve its resistance to neutron irradiation
[35]. The post treatment of molybdenum should be taken into consideration in real DT nuclear fusion operations since molybdenum is a high-activation element
[85]. Furthermore, like tungsten, the erosion of molybdenum in steady-state conditions is primarily caused by low- and medium-Z impurities in the plasma
[24],
[66]. Experimental campaigns of long pulse and high parameters in EAST have further demonstrated that melting and boiling of TZM could be induced by the transient heat flux during plasma disruptions
[31]. Moreover, the behavior of different components in the TZM varies within the fusion core, which would affect the service performance of the material.
Over the half-century history of magnetic confinement fusion, carbon, beryllium, molybdenum, and tungsten have been extensively investigated and used as PFMs. Far more has been learnt from tens of thousands of experimental campaigns in numbers of tokamaks worldwide
[23],
[24],
[39],
[86]. In the journey towards DEMO and commercial fusion, carbon and beryllium have gradually been excluded from the shortlist for PFMs, mainly due to the high erosion rates and high tritium retention, as shown in
Fig. 5. In a sense, molybdenum can be considered a viable alternative to tungsten in current fusion devices. Tungsten has become the reference standard for current and future PFM. Numerous existing or under-construction tokamaks use or plan to use tungsten as a critical PFM, associated with the implementing new system design and modifying existing systems (e.g., wall conditioning system, plasma boundary control system, disruption mitigation system) to mitigate the extreme environments faced by tungsten armor
[4]. However, one should keep in mind that whether tungsten will be the ultimate choice for PFM in future commercial fusion reactors remains highly uncertain
[4],
[23],
[76].
3. Promising candidates
The selection of PFMs is highly challenging, with compatibility with high-temperature fusion plasmas being the primary requirement
[12],
[13],
[14],
[23],
[33]: materials with low atomic numbers or high sputtering resistance are preferred; if materials with high atomic numbers, such as tungsten, are used, the concentration of impurity introduction must be strictly controlled. Meanwhile, high thermal conductivity, high strength, matching thermal expansion coefficients, and low neutron activation are key performance indicators
[13],
[50]. By combining peer-reviewed data in the Pauling File database (melting points, thermodynamic and mechanical parameters, etc.) with data from DFT calculations (surface binding energy, interstitial hydrogen, and isotope formation energy), Fedrigucci et al. screened 21 of the most promising PFM candidate materials
[81]. Their model successfully identified traditional PFMs such as tungsten, molybdenum, and carbon, confirming its reliability. These screening results provide an important reference for PFMs' selection. In this section, a brief overview of the screened materials will be presented. It should be noted that this shortlist of 21 PFM candidates should be treated with caution; materials not included in the shortlist should not be simply excluded from consideration as PFM candidates, such as TiC
[48],
[87]. When determining whether a material is suitable for PFM, it is still necessary to combine theoretical predictions with experimental evaluation.
3.1. Transition metals and their alloys
In addition to the metals already in service, such as tungsten and molybdenum, their 21 PFM candidates also include metals such as copper, rhenium, tantalum, and chromium, as well as their alloys. Although copper has extremely high thermal conductivity
[81],
[88],
[89], it has significant limitations as a PFM. Copper has low hardness and a low softening temperature, making it difficult to maintain mechanical stability under extreme thermal loads
[81],
[90]. Copper exhibits intensive erosion under hydrogen plasma flux, which limits its in-service lifetime
[91]. As such, pure copper is not suitable for directly use as a potential PFM. However, Cu-Cr-Zr alloys have become a key heat sink material in blanket structures, playing a central role in efficiently dissipating heat
[92],
[93],
[94]. In addition, there are a few studies focusing on Cu-W and Cu-Li alloys as prospective materials for divertors and first-wall components
[92],
[95].
As a rare metal, rhenium primarily serves as an alloying element in tungsten-based PFMs
[79],
[96]. Rhenium significantly enhances the thermomechanical properties of tungsten, the so-called Rhenium Effect
[79],
[96], including improved ductility, high-temperature strength, plasticity, and corrosion resistance. In terms of microstructure, rhenium effectively stabilizes the grain structure, raises the recrystallization temperature, and improves weldability, while significantly reducing the degree of recrystallization embrittlement of the material under high-temperature or irradiation conditions
[79],
[96]. However, the extremely high cost and scarcity of rhenium severely limit its large-scale application. Based on economic considerations, platinum-group metals are also typically excluded from the selection of PFMs
[81].
Tantalum was previously considered a potential PFM
[97]. The energy threshold at which tantalum forms a fibrous fuzz structure under helium ion flux is higher than that of tungsten
[97]. However, the drawbacks of pure tantalum are equally significant: its thermal conductivity is inferior to that of tungsten, and it exhibits exothermic hydrogen absorption and high tritium retention, which limit its application as a potential PFM
[81]. Forming alloys with tungsten, typically with a composition of Ta
0.5W
0.5, not only lowers the DBTT of tungsten to room temperature but also possesses improved high-temperature thermomechanical properties
[98],
[99],
[100],
[101].
Chromium possesses a sequence of properties
[102],
[103],
[104], such as high melting point of 2180 K, excellent corrosion resistance, and superior low-activation behavior, making it a good candidate for structural components in fusion reactor vessel. The use of pure chromium as a PFM is limited by its high DBTT and low recrystallization temperature
[104],
[105],
[106]. Its ductility at room temperature is highly susceptible to significant effects from impurity elements such as oxygen and nitrogen, which increases the difficulty of fabrication and welding. Alloying can effectively address the shortcomings of pure chromium. For example, the Cr-10%W alloy exhibits significantly improved yield strength and work hardening capacity at high temperatures
[104] and is proposed to be a promising candidate material for monoblock structures in PFCs
[106].
Medium-entropy alloys (MEAs) and high-entropy alloys (HEAs) are emerging as transformative PFMs due to their unique multi-principal element design, which shifts the paradigm away from traditional alloys
[67],
[107],
[108],
[109],
[110],
[111]. Currently, research focuses on their exceptional irradiation tolerance, where the inherent lattice distortion effect creates a complex potential energy landscape that promotes the spontaneous recombination of Frenkel pairs, effectively suppressing the nucleation of dislocation loops and voids that lead to hardening
[112],
[113],
[114],
[115]. HEAs have shown significant potential in resisting radiation-induced surface morphology changes under high-flux plasma exposure
[116],
[117],
[118]. The prospect of HEAs lies in their high degree of tunability through materials modelling, allowing for the concurrent optimization of high-temperature mechanical strength and low-activation characteristics
[119],
[120]. However, the chemical complexity of HEAs inherently leads to preferential sputtering due to disparities in atomic masses and binding energies, resulting in a dynamic surface evolution that necessitates the careful selection of constituent elements to ensure plasma compatibility and structural integrity
[118],
[121],
[122],
[123]. Moreover, simultaneously achieving high fracture toughness and high thermal conductivity remains intrinsically challenging. Fracture toughness generally benefits from severe lattice distortion, chemical complexity, and heterogeneous local environments. However, these same characteristics also increase electron and phonon scattering, thereby reducing thermal conductivity, which is indispensable for efficiently dissipating the extreme steady-state heat fluxes encountered by PFMs
[108],
[124],
[125],
[126]. Furthermore, the outstanding radiation tolerance of HEAs/MEAs originates from their compositional complexity and lattice distortion, which create a rugged potential energy landscape that facilitates point-defect recombination and suppresses irradiation-induced defect clustering. Nevertheless, these same characteristics inevitably increase electron and phonon scattering, leading to reduced thermal diffusivity and consequently compromising heat dissipation capability. In addition, prolonged plasma exposure may induce preferential sputtering, elemental redistribution, and local phase evolution, further affecting both thermal transport and microstructural stability
[120],
[127],
[128],
[129]. Therefore, the trade-off between thermal management and radiation stability arises because the same structural features that improve irradiation resistance also impede efficient heat transport. Future development of HEAs/MEAs should therefore focus on optimizing compositional complexity and microstructural architecture through advanced alloy design and processing strategies to balance thermal conductivity, fracture toughness, and irradiation tolerance. If these competing requirements can be reconciled, HEAs and MEAs are expected to provide a promising material foundation for the next generation of magnetic confinement fusion reactors
[109],
[127],
[128],
[129],
[130].
3.2. Transition metal compounds
The shortlist of 21 PFM candidates also includes several transition metal compounds, primarily high-melting-point borides, carbides, and nitrides
[81]. Most of these compounds can be classified as ultra-high temperature ceramics (UHTCs)
[131],
[132],
[133],
[134],
[135],
[136],
[137],
[138],
[139],
[140],
[141],
[142],
[143], and some are key materials for the fourth-generation fission reactors
[144],
[145],
[146]. Due to their high melting points, excellent high-temperature stability, high thermal conductivity, and good mechanical properties, these compounds have attracted widespread attention
[131],
[147]. However, systematic research on the performance of UHTCs in fusion environments remains limited, particularly regarding PSI, high-temperature neutron irradiation, and their synergistic effects
[14],
[148],
[149],
[150]. Furthermore, the inherent brittleness and poor thermal shock resistance of the UHTCs limit their application prospects
[14],
[131],
[151],
[152],
[153],
[154],
[155],
[156],
[157],
[158],
[159],
[160],
[161],
[162],
[163],
[164],
[165].
Four borides (HfB
2, ZrB
2, TiB
2, and VB
2) appeared on their shortlist; except for VB
2 (which has a melting point of 2723 K), the melting points of the other three borides all exceed 3273 K
[166]. Nasseri used Geant4 simulations to investigate the behavior of ZrB
2 and HfB
2 in interaction with neutrons
[167]. They found that HfB
2 is a better neutron absorber than ZrB
2 across broad neutron spectrum, while ZrB
2 yields more helium particles under thermal neutrons and is more susceptible to irradiation damage
[167]. Garrison et al. reported the first study of ZrB
2 behavior under 30 keV He ion irradiation at temperatures up to 1120 K
[168]. ZrB
2 demonstrated the ability to maintain structural integrity under high-temperature He ion irradiation
[168], indicating that ZrB
2 is a promising PFM candidate.
Fig. 6a~d show the morphology modification of the metal borides after deuterium plasma irradiation, indicating the erosion of the grain boundaries. Recent experiments involving deuterium plasma exposure have shown that TiB
2 and ZrB
2 exhibit boron preferential erosion
[14],
[169], leaving behind a metal enriched surface, as shown in
Fig. 6e and f. While this is beneficial to some extent for maintaining plasma purity
[169], changes in the composition and structure of the substrate can modify its properties. The impact of these changes on its properties and lifetime warrants attention and requires further investigation
[170]. Boron plays a key role in the irradiation effects of these materials, as
10B decays into He and Li atoms after neutron irradiation, triggering the formation of cavities or bubbles in the matrix and grain boundaries
[150],
[171]. Researchers at Oak Ridge National Laboratory have conducted studies on the neutron irradiation effects of
11B-enriched Ti
11B
2 [150] and Zr
11B
2 [148], respectively, revealing that using
11B-enriched Zr
11B
2 can effectively mitigate the negative effects of helium production caused by
10B transmutation. They found that fine-grained ceramics can improve radiation resistance of the diborides, as compared in Figs. 6 g and 6 h. However, irradiation-induced defects can significantly reduce the thermal conductivity of ZrB
2 [148]. The MB
2-WC (M = Ti, Zr, Hf) composites showed excellent stability in their thermal conductivity when exposed to the deuterium plasma at temperatures up to 1473 K
[170]. Under neutron irradiation, lattice swelling in hexagonal diborides exhibits slight anisotropy, leading to grain boundary microcracking, a phenomenon observed in both ZrB
2 and TiB
2 [148],
[150]. Regarding oxide impurities, the relatively large lattice parameter changes in Zr-O impurities after irradiation may intensify trans granular cracking formation
[148].
The primary concern with carbides as potential PFM candidates is their high carbon atomic content, which poses risks of high corrosion rates and high tritium retention like those associated with graphite
[14],
[81]. Numbers of studies utilized carbides as dispersoids to toughen tungsten-based materials
[78],
[87],
[172],
[173],
[174],
[175],
[176],
[177],
[178]. The melting points of transition metal carbides are higher than those of their corresponding borides and nitrides, while their thermal conductivity is slightly lower. Generally speaking, the order of fracture toughness is nitrides > borides > carbides, while the order of flexural strength is borides > nitrides > carbides
[14]. Fedrigucci et al. identified HfC, TaC, W
2C, and WC as the four carbides on their shortlist of 21 PFM candidates
[81], while conventional nuclear materials such as TiC and ZrC were not included on the shortlist
[144],
[149],
[179], a point that is open to debate. In terms of the crystallography of these carbides, HfC exhibits a wide range of carbon-deficient stoichiometric ratios
[180], while TaC is a line compound in the phase diagram
[181]. WC is a line compound below 2657 K; between 2657 and 3058 K, it exhibits non-stoichiometric ratios accompanied by polymorphic transformations
[182]. W
2C is a phase stable at high temperature above 1523 K and possesses several polymorphs
[182]. The composition, constituents, and microstructure of carbides depend on their preparation methods; imperfections such as non-stoichiometric ratios, free carbon, and oxygen impurities may be present, which can affect the thermal and mechanical properties of the material to varying degrees
[48],
[144],
[181],
[183]. When carbides are exposed to plasma, preferential sputtering occurs, leaving a carbon-deficient surface with modified thermal properties
[14]. This may lead to considerable thermal gradients between the surface and the substrate
[13],
[183], a factor that warrants careful attention. Helium ion implantation in transition metal carbides induces the formation of high-density point defects and helium vacancy clusters, which in turn cause lattice swelling, surface blistering, and helium embrittlement
[184],
[185]. The thermal conductivity of carbides decreases significantly after neutron irradiation
[149],
[179],
[186]; however, their swelling caused by neutron irradiation is less severe than that of borides
[14],
[149],
[150],
[187], which may be attributed to the higher melting point of carbides and thus the corresponding excellent structural stability. The thermal conductivity of metal-rich carbides such as W
2C
[95], Ta
2C
[188], and Zr
2C
[189] is significantly lower than that of the corresponding metal monocarbides. Furthermore, since W
2C undergoes a polymorphic transformation at high temperatures
[182], its suitability for direct use as a PFM requires further assessment.
Two transition metal nitrides, TiN and
θ-TaN, are included in their list, while nitrides with higher melting points and bond dissociation energies, such as ZrN and HfN, are not
[81]. When considering nitrides as fusion core materials, the first factor to keep in mind is the transmutation of
14N into
14C under neutron irradiation, as well as the potential high fuel retention issues that may result
[14],
[81],
[190]. Ab initio theoretical calculations indicate that TiN exhibits good resistance to hydrogen permeation
[191], a finding supported by deuterium permeation experiments
[192]. Zhang et al. proposed TiN as a tritium permeation barrier coating
[192]. However, nitrides face the risk of nitrogen loss at high temperatures, accompanied by changes in composition, crystal structure, and properties. Bull et al. conducted experimental validation of nitride stability and found that TiN showed the resistance to nitrogen loss at 1773 K, yet less stable than ZrN and
h-BN
[191]. The low melting point of Cr
xN would limit its potential application as hydrogen permeation barrier coating
[193].
θ-TaN has attracted attention primarily due to its exceptionally high theoretical thermal conductivity, which has been experimentally confirmed by various research teams
[194],
[195],
[196]. Reports indicate that the thermal conductivity of
θ-TaN single crystals at room temperature differs from 502 W/m·K
[196] to 1100 W/m·K
[194] by two independent groups. Furthermore, the preparation of
θ-TaN crystals is an extremely harsh process
[194],
[196],
[197] and its properties are largely unexplored.
3.3. Low-Z compounds
Their list also includes low-Z compounds composed of B, C, and N, namely BN, BC
2N, and C
3N
4, primarily due to their high thermal conductivity, high specific heat capacity, and high Debye temperature
[81]. C
3N
4 has found widespread application as a nano-functional material in fields such as catalysis
[198],
[199]. It has been reported that the decomposition temperature of C
3N
4 is below 1273 K
[200],
[201]. Research on BC
2N
[202],
[203],
[204] and C
3N
4 [205],
[206],
[207] have largely relied on theoretical calculations, and there remain many unknowns regarding the experimental determination of their thermal and mechanical properties
[208],
[209]. Another two typical low-Z compounds, boron carbide (B
4C) and silicon carbide (SiC) should also be taken into consideration. B
4C is limited by its low thermal conductivity, brittleness, poor thermal shock resistance, and potential risk of high tritium retention
[171]. SiC possesses advantages such as low activation, high-temperature resistance, radiation resistance, and low neutron absorption, and has been studied and used in the nuclear systems for decades
[48],
[210],
[211],
[212],
[213],
[214],
[215],
[216],
[217]. Recent H-mode plasma discharge experiments conducted in the DIII-D tokamak device
[48],
[210] have shown that CVD-grown SiC exhibits good sputtering resistance, low impurity emission, and morphological stability, making it a promising candidate for PFMs. However, the significant degradation in thermal conductivity of these low-Z compounds under neutron irradiation poses a major challenge that needs to be addressed
[171],
[213],
[214],
[218].
3.4. Liquid plasma-facing materials
The concept of the liquid first walls in fusion reactors emerged as a radical solution to the insurmountable limitations of solid PFMs
[48],
[219],
[220],
[221]. Solid first walls like molybdenum and tungsten tiles suffer catastrophic degradation under extreme neutron irradiation and immense heat fluxes
[13],
[19],
[30],
[31]. Liquid first walls including liquid lithium, lead-lithium, or molten salts were proposed because they are intrinsically self-healing, theoretically immune to neutron embrittlement, and capable of continuously extracting extraordinarily high heat loads without structural failure
[222],
[223]. It remains largely confined to conceptual designs and small-scale loop experiments
[48]. Liquid metals moving across strong magnetic fields would generate severe magnetohydrodynamic drag
[223], which could cause flow instability and excessive pumping power requirements. Furthermore, high vapor pressure can lead to severe plasma contamination, quenching the fusion reaction, while the liquid medium aggressively corrodes underlying structural materials
[224]. DIII-D conducted the irradiation experiment on liquid lithium in a tokamak
[48]. When preheated to 623 K, uniform release of lithium vapor was achieved under high-confinement mode conditions, and lithium droplet spatter was effectively suppressed
[48]. While scientifically intriguing, liquid first walls currently represent a conceptual panacea that lacks the technological maturity to reliably sustain a burning plasma.
Table 1 summarizes the advantages and disadvantages of the potential PFMs for fusion energy. Despite decades of intensive research, no PFM has yet demonstrated an optimal combination of thermophysical performance, plasma compatibility, irradiation tolerance, and engineering feasibility required for future fusion power plants. While conventional materials such as tungsten remain the benchmark for near-term applications, emerging candidates offer new opportunities but still require systematic validation under fusion-relevant plasma and neutron environments. Bridging the gap between intrinsic material properties and reactor-scale performance will therefore remain a central challenge in PFMs research.
4. Qualification of the PFMs
To date, only few tokamaks (such as PTE, TFTR and JET) have conducted deuterium-tritium fusion experiments, and the results obtained have laid a solid engineering foundation and instilled confidence in the design of ITER and future DEMO
[4],
[24],
[53],
[76],
[77],
[225],
[226],
[227],
[228],
[229],
[230]. PFMs in controlled nuclear fusion devices are subjected to extreme and unique operating conditions: high-energy 14.1 MeV neutrons generated by deuterium-tritium fusion, plasma temperatures of hundreds of millions of degrees, and high heat fluxes
[13],
[19],
[231]. Such extreme environment far exceeds the empirical limits of existing fission reactors or aerospace materials. The assessment and evaluation of fusion reactor materials have gradually evolved into a systematic discipline with the focus on neutron irradiation damage, helium production rates, hydrogen isotope retention, thermomechanical fatigue, and PSI
[232],
[233],
[234],
[235]. The goal of evaluation is not to pursue a single high metric, but rather the material’s ability to operate stably over the long term under coupled multi-physical fields, including irradiation hardening and embrittlement, volumetric swelling and phase stability, thermal conductivity degradation, tritium retention and permeation, surface morphology evolution (fuzzing, cracking, and erosion), as well as chemical compatibility with the coolant and structural environment.
Evaluating PFMs is a very difficult and costly endeavor. Large number of facilities have been designed, built, and operated around the world to simulate one or, at most, two of the extreme conditions encountered in fusion scenarios
[13],
[28],
[68],
[236]. Laser or ion beams are used to simulate the shock of high heat flux and bombardment of intense plasma
[68],
[168],
[237],
[238]. Fission neutrons are used to simulate the transmutation and displacement damage by fusion neutrons
[233],
[234],
[235],
[239],
[240]. However, ion beams and fission neutrons differ significantly from fusion conditions in terms of the heat flux and neutron energy they can provide, as well as the resulting damage
[19],
[241],
[242], as summarized in
Fig. 7a. To address the challenge of scarce 14.1 MeV neutrons, several dedicated material qualification facilities have been proposed, such as IFMIF-DONES and IFMIF/EVEDA
[243],
[244],
[245].
Although each of these methods has its limitations, cross-validation with tokamak experimental data has gradually led to the development of reliable material databases and failure models
[4],
[22],
[23],
[24],
[39],
[76]. PSI experimental results have been accumulated over the long term in tokamak devices, as summarized in
Fig. 7b. JET device operated for the first time with all-tungsten and beryllium walls, systematically studying hydrogen isotope retention, impurity migration, and power load response under metal first-wall conditions
[24],
[52],
[76],
[227],
[246],
[247]. The results showed that, compared to carbon walls, metal walls significantly reduced tritium retention; however, they also revealed the sensitivity of radiation losses and confinement to tungsten impurities entering the plasma
[4]. This finding directly influenced the selection of first-wall materials and operational strategies for ITER and future reactor designs
[4],
[22],
[246]. EAST device, during long-pulse, high-power operations, provides data on the behavior of tungsten divertors under quasi-steady-state heat flux, verifying the thermal fatigue and recrystallization evolution patterns of tungsten under heat loads of the order of 10 MW·m
−2 [30],
[31],
[39],
[248],
[249]. WEST device, as a full-tungsten-wall tokamak specifically designed to study ITER divertor conditions, has demonstrated that under ITER-like heat flux and pulsed conditions, the evolution of tungsten surface morphology is closely related to hydrogen retention, providing critical data for tungsten lifetime prediction models
[4],
[250]. DIII-D device has conducted in-depth research on the damage mechanisms caused by transient thermal shocks during ELM and rupture events
[46],
[48],
[251],
[252],
[253]. It has demonstrated that under millisecond-scale high heat flux conditions, the material failure mechanism shifts from thermal fatigue to one dominated by melting and evaporation, which has directly advanced research on ELM suppression and divertor geometry design
[74],
[251],
[254].
5. Summary and perspective
The environment for plasma-facing materials (PFMs) in fusion reactors is extremely harsh, characterized by high heat flux density, high-temperature plasma bombardment, and 14.1 MeV fast neutron irradiation. These extreme conditions, involving strong coupling and multiple physical fields, impose comprehensive demands on materials that far exceed those of conventional engineering systems. At the material level, these operating conditions ultimately translate into strict constraints on key properties such as melting point, heat capacity, thermal conductivity, thermal shock resistance, sputtering/corrosion resistance, irradiation stability, and low hydrogen isotope retention. A review of the development history of typical PFMs such as carbon, beryllium, molybdenum, and tungsten reveals that the selection and replacement of PFMs profoundly reflect the process of humanity’s deepening understanding of PSI, plasma control, etc. Carbon and beryllium were once widely adopted due to their good plasma compatibility but were gradually phased out due to severe erosion and fuel retention issues. Molybdenum possesses excellent thermal properties, but as a high-activation element, it is not conducive to long-term operation and post-processing in real DT fusion reaction. Currently, tungsten has become the mainstream choice for PFMs, and extensive PSI experience has been gained in tokamak devices such as JET, EAST, and WEST. However, tungsten suffers from issues such as neutron irradiation embrittlement, helium-induced damage, recrystallization, and thermal shock failure, all of which limit steady-state operation of long-pulse, high-parameter plasmas and the long-term safety of tokamak devices. The selection of PFMs for future DEMO and commercial fusion remains unresolved, and PFMs continue to be a significant challenge on the path to commercial fusion.
Developing higher-performance alloys and composites based on tungsten-matrix materials and optimizing component designs are among the appropriate approaches for improving the overall performance of tungsten-matrix PFMs in the short term. For example, strategies such as solid solution strengthening, dispersion strengthening, and fiber reinforcement can be employed to improve the toughness and radiation damage tolerance of tungsten-based materials. Ideally, these alloying elements, dispersoids, and reinforcements should themselves be suitable nuclear materials to prevent them from becoming weak points in nuclear applications. In addition to tungsten-based systems, other promising PFM candidates, represented by UHTCs, have gradually attracted attention over the past decade. UHTCs demonstrate high melting points, good thermal stability, and excellent mechanical properties; however, their inherent brittleness and poor thermal shock resistance remain issues to be addressed. UHTCs have typically been developed in fields such as aerospace; however, under the unique coupled thermal-particle-neutron irradiation scenarios of fusion, systematic research on their plasma compatibility, the effects of high-temperature high-dose neutron irradiation, and hydrogen isotope retention and transport behavior remains insufficient. Furthermore, challenges remain to be addressed regarding the synthesis of high-quality raw materials for UHTCs (including isotope enriched UHTCs), as well as the shaping, machining and joining technologies for complex components.
The current materials research and qualification system still struggles to comprehensively simulate the heat-plasma-neutron combination in a fusion reactor core. Looking ahead, the development of PFMs should be advanced in a coordinated manner across the following areas: clarifying material requirements and translating fusion operating scenarios into specific material performance metrics; accelerating the screening, synthesis, and testing of candidate materials through multiscale simulations and artificial intelligence—for example, ITER’s open-sourcing of its physical modeling and analysis tools, IMAS, will contribute to the global development of fusion energy; leveraging existing simulation and testing platforms to establish a materials database that reflects the synergistic effects of thermal shock, plasma erosion, and neutron irradiation; and mitigating the extreme conditions faced by PFMs at the system level by integrating plasma control, in-vessel component design optimization, improvements to supporting systems, etc.
The advancement of PFMs requires a shift away from focusing solely on improving the performance of individual materials toward system-level optimization: material selection, component design, plasma control strategies, wall conditioning, and cooling, etc., constitutes a complex multidisciplinary engineering challenge. Future material development should be more closely integrated with the reactor core design process, and materials scientists should engage deeply in the design, construction, and operation of fusion reactors from the “material-component-system” framework.
Finally, the history of PFM research also reflects the logic of the evolution of scientific understanding: from the initial use of high-Z metals in early tokamaks, which caused sputtering contamination of the plasma core, to the selection of low-Z materials such as carbon and beryllium; then, due to issues with these low-Z materials, i.e., high corrosion rates, fuel retention, and toxicity, they were gradually phased out; followed by the reassessment and adoption of the high-Z tungsten system as the current standard PFM; and finally, the current exploration of more complex material systems. This process represents a dialectical negation of negation in the understanding of fusion materials. The success of future fusion energy depends not only on breakthroughs in material performance itself, but also on interdisciplinary collaboration and the deep integration of systematic experimental validation with theoretical simulations. These efforts are an indispensable part of achieving long-term steady-state operation and ultimately commercial fusion reactors.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This work was financially supported by the National Natural Science Foundation of China (Nos. 52293372, 52032003, 52032001, 52332003, 50632070).
Declaration of Competing Interest
The authors have no competing interests to declare that are relevant to the content of this article.
Availability of data and materials
Not applicable.