Plasma Spraying of W Coatings for Nuclear Fusion Applications: Advancements and Challenges

Abstract

The selection of a suitable plasma-facing material (PFM) that must protect the divertor, cooling systems, and structural components is an important challenge in the design of advanced fusion reactors and requires careful consideration. Material degradation due to melting and evaporation may lead to plasma contamination, which must be strictly avoided. Among the candidate materials, tungsten (W) is the most promising because of its thermo-mechanical and physical properties, which allow it to endure repeated exposure to extremely harsh conditions within the reactor. The plasma spraying (PS) technique is gaining increasing interest for the deposition of tungsten (W) coatings to protect heat sink materials, due to its relatively low cost, high deposition rates, and capability to coat complex-shaped surfaces and fix damaged coatings in situ. This review aims to provide a systematic assessment of tungsten (W) coatings produced by PS techniques, evaluating their suitability as PFMs. It discusses W-based materials, plasma spraying technologies, the role of the interface in joining W coating and metallic substrates such as copper alloys and steels, and the main issues related to coating surface erosion under steady-state and transient heat loads associated with advanced fusion reactor operation modes and off-normal events. Quantitative data available in the literature, such as porosity, oxygen content, thermal conductivity of the coatings, residual stresses accumulated in the coating–substrate interface, surface temperature, and material loss following heat load events, were summarized and compared to bulk W ones. The results demonstrate that, following optimization of the fabrication process, PS-W coatings exhibit excellent performance. In addition, previously mentioned advantages of PS technology make PS-W coatings an attractive alternative for PFM fabrication.

1. Introduction

Nuclear fusion is a potential solution to meet the growing global demand for sustainable energy. Fusion reactors could provide clean and abundant energy; however, materials involved are subjected to extreme operating conditions. Plasma-facing components (PFCs), including shielding systems and heat transfer units, are exposed to thermal loads and particle flux escaping from the plasma, leading to absorption, chemical bonding, surface deterioration due to physical or chemical sputtering, evaporation or arcing, and material re-deposition. These operating conditions lead to high temperatures on the surface of PFCs with consequent melting and evaporation. Additionally, thermal stresses due to plasma heat flux together with those induced by the manufacturing process may lead to the formation of cracks. Alongside their standard operating conditions, PFCs are expected to experience abnormal events such as disruptions, i.e., sudden plasma current termination; vertical displacement events (VDEs) characterized by uncontrolled plasma shifts; and various forms of edge-localized modes (ELMs), including controlled plasma “cleaning” from impurities. To illustrate the extreme conditions that plasma-facing components (PFCs) must withstand, it is useful to analyze some data. For example, a type I ELM, the most hazardous type for plasma-facing material degradation, can potentially deposit an energy quantity of approximately 1 GW m⁻² within a timeframe shorter than 1 ms. VDEs are capable of transferring significant quantities of plasma energy to specific sections of the reactor’s walls within a pulse timeframe of 100–300 ms. Energy densities from mitigated ELMs using pellet injection can reach up to 1 MJ/m² at frequencies of 50 Hz [1]. The electro-mechanical loads during an ITER plasma disruption, which involves thermal energy less than 350 MJ and poloidal magnetic energy less than 1300 MJ, result from two processes: current induction by fast-changing magnetic fields from the plasma current decay and movement, as well as a halo current, with a path partially in the plasma edge and flowing through the plasma-facing components [1,2,3,4,5,6,7,8].

Therefore, PFCs must meet high safety standards as well as low manufacturing costs, making research on the topic of great importance. Research campaigns were carried out with the goal of identifying materials that can withstand abnormal events and implementing more efficient plasma control methods. These studies focused on limiting PFC erosion to a minimum thickness that does not decrease under 3 mm for the divertor and extending the material’s lifespan. In demonstration-scale fusion reactors with high neutron flux, neutron exposure can impact the thermo-mechanical characteristics and result in altered material dimensions. The choice of PFMs should consider their compatibility with plasma (high or low Z) and substrate material, resistance to high temperatures, high level of resistance to erosion, excellent resistance to high-energy fluxes, high thermal conductivity and capability to withstand severe thermal shock, minimal radiation alterations, and low tritium retention. For example, very high Z materials offer excellent thermal properties but suffer from excessive impurity radiation under high plasma energy; on the other hand, very low Z materials can be employed in a higher range of plasma conditions but are highly susceptible to erosion. The only metallic materials able to satisfy these requests are refractory metals.

Among the refractory metals, tungsten (W) and molybdenum (Mo) have been identified as the most promising materials for PFM applications due to their exceptional properties, such as high thermal resistance, low sputtering under ion bombardment, and capability to withstand high heat fluxes. As compared to W, Mo possesses superior physical characteristics, including density, melting point, boiling point, and thermal conductivity [9,10,11,12,13]. Moreover, Mo has a lower ductile-to-brittle transition temperature (DBTT) and superior thermal shock resistance. A key benefit of Mo over W is its reaction to prolonged exposure to 14 MeV neutrons in a fusion power plant setting. The erosion and re-deposition of pure Mo and W exhibit similar behavior [14], with Mo retaining a smaller proportion of hydrogen isotopes than W [15]. Montanari et al. [16] discovered that in scenarios analogous to ITER disruption events, the ablation volume lost by Mo is approximately ten times greater than that of W, primarily due to its lower latent heat of fusion and vaporization. Therefore, the superior ductility and lower DBTT of Mo are offset by its higher erosion rate, making Mo less suitable than W for application in the divertor armors of future fusion reactors.

Alongside the clear benefits of W, there are significant disadvantages that make its application in PFC design difficult: W has poor machinability at room temperature and high DBTT [7,17,18,19,20,21]; therefore, conventional bulk manufacturing processes are not suitable. Nevertheless, it is possible to use W as the armor of a heat sink, for example, in the form of tiles or coatings. In these cases, manufacturing methods such as brazing can be adopted to join sintered W tiles to the heat sinks. The costs related to the application of these two technologies, brazing and sintering, are expensive, the process is time consuming, and they are difficult to apply on complex components. Therefore, a number of coating technologies have been investigated by many research studies, with a particular emphasis placed on vacuum or atmosphere plasma spraying (VPS and APS) [22], physical vapor deposition (PVD) [23], and chemical vapor deposition (CVD) [24,25,26]. Moreover, in a recent work [27], a novel approach combining magnetron sputtering and ion implantation was employed to develop a thin W coating on Inconel 718, both with and without a (~3 µm) Mo interlayer, for COMPASS-U tokamak.

Among the different techniques, plasma spraying is the most used due to its versatility; however, joining W and metallic material substrates, such as CuCrZr alloy [28,29,30,31] or martensitic steels [32,33], is a hard task due to the great mismatch of the coefficients of thermal expansion (CTE), leading to high residual stresses at the interface during cooling to ambient temperature. The arising stresses may induce the formation and propagation of cracks and the final detachment of the coating from the substrate either in the deposition process or during the operative life of the component. The use of an appropriate interlayer helps to solve this issue.

Despite the high number of reviews on plasma-facing materials (PFMs), a systematic and focused analysis of plasma-sprayed tungsten (W) coatings is still lacking. This review aims to fill that gap by providing a comprehensive overview of the current progress, challenges, and potential of plasma-sprayed W coatings on metallic substrates for fusion applications. Particular attention is given to issues related to coating–substrate compatibility, interface behavior, and the role of interlayers in mitigating residual stresses. This review is structured into four main sections: (i) W-based materials, (ii) plasma spraying technologies, (iii) the role of the interface in joining W coating and metallic substrates, and (iv) coating surface erosion under steady-state and transient heat loads. The following section (Section 2) presents a general introduction on the methods for improving the properties of W-based materials, in particular by tailoring the composition, followed by a special focus on plasma spraying technologies, VPS and APS in particular (Section 3). Section 4 provides details on the effects of various types of interlayers used to reduce residual stresses and mitigate differences in the thermal expansion coefficient between the coating and the substrate, thereby enhancing coating adhesion and stability. Finally, Section 5 explores the performance of substrate/interlayer/coating under different thermal load conditions. Particular attention has been paid to the research of novel deposition and post-process treatments.

The scope of this work was restricted to the thermal loading effects of plasma, without taking into consideration the neutron radiation damage. The latter is a complex and deeply profound subject, and it lends itself to a separate study.

2. W-Based Materials

Different approaches have been used to improve the properties and overcome the limits of pure W. W-based materials (Figure 1) can been produced through various techniques: composites, nanocrystalline materials, W-X (where X denotes elements such as Ta, Re, Mo, V, Ti, etc.) alloys manufactured by powder metallurgy, sinterization of W with ductile fibers, dispersion of ceramic particles of transition metals within W, and addition of rare-earth (RE) oxide particles to W [34,35].

The addition of oxide particles like La₂O₃, Y₂O₃, and TiO into the W matrix has been found to enhance several properties, including toughness [36,37,38,39,40,41], thermal shock resistance, creep resistance, recrystallization temperature, and grain growth at high temperature.

It is generally believed that oxygen (O) impurity is one of the main factors causing imperfect bonding and voids at the inter-lamellae of APS coatings and its introduction is unavoidable during the APS process [42,43], even using supersonic guns. O impurities usually exist in the form of oxides (WO₃) and segregate at grain boundaries with harmful effects on the thermo-mechanical properties of tungsten coatings [43]; however, their detrimental effects can be mitigated by altering the oxide types and incorporating RE elements. Thermal treatments, such as the two-step annealing treatment (793 °C for 1 h + 910 °C for 2 h at a pressure of 0.1 Pa) proposed by Jianjun et al. [44], can promote the reduction of tungsten oxides to W, with beneficial effects on both their content and distribution from grain boundaries to grains.

A similar effect could also be obtained by introducing active elements such as rare-earth elements [45,46]. In fact, the addition of RE hydrides, specifically LaH₂ powder, to the W powders used in APS coatings has been discovered to significantly decrease the oxide content. This is attributed to the formation of WO₂ and La₂O₃ oxides that grow at the expense of WO₃ and O₂ [46]. Hou et al. [47] investigated NbC-doped W coatings deposited by using the supersonic APS method. The NbC-doped coating exhibited improved characteristics such as reduced porosity, increased lamellae thickness, enhanced thermal conductivity, and superior hardness and crack resistance.

Several recent studies have been intensively focused on nanosized grains of W/W tungsten-based materials [48,49,50]. The superior mechanical properties of W-based nanocrystalline materials are primarily attributed to two factors: grain boundary hardening and dispersion hardening. The fundamental principle behind grain boundary hardening is that grain boundaries act as obstacles, thereby hindering the movement of dislocations. As a result, additional external force is necessary to overcome these obstacles, ultimately contributing to increased hardness. Furthermore, secondary-phase particles act as anchoring points, effectively obstructing the movement of dislocations. These materials exhibit exceptional mechanical properties and an unexpected resistance to irradiation damage, which may overcome the limitations of pure W and facilitate their industrial application. However, due to their high melting point and low recrystallization temperature, nanocrystalline W/W composites are challenging to fabricate.

Despite their promising properties, W-based materials are more susceptible to material loss through ablation, evaporation, or melting during transient thermal events such as ELMs or disruptions. This increased vulnerability, compared to pure W, can lead to significant plasma contamination, highlighting the need for further optimization in both composition and processing methods to balance performance with plasma-facing material stability.

3. Plasma Spraying Technologies

This section provides a comparison of the main plasma spraying (PS) technologies and discusses their potential for application in the fabrication of plasma-facing materials (PFMs).

Plasma spraying appears to be particularly appealing, given its simplicity, low expenses, and ability to cover large surfaces, and VPS is commonly used to produce coatings with low porosity and minimal impurities.

In earlier studies [51,52,53] of VPS-W coatings on F82H reduced activation steel, it was observed that the coating consisted of two distinct zones: a columnar area with a homogeneous orientation and an inhomogeneous area with randomly oriented grains, pores, and unmelted particles. The non-uniform regions decrease the coating thermal conductivity and diffusivity. Tokunaga et al. [53] suggested that excluding unmelted and semi-melted particles would contribute to the attainment of a completely regular columnar structure. The optimization of spraying temperature, gas flow parameters, and powder size enables the deposition of W coating that possesses withstanding characteristics [52,53,54,55]. In particular, Wang et al. [56] suggested a method of enhancing the VPS process for applying VPS W coatings on F82H substrates. Achieving optimal processing conditions, they improved the mechanical properties of the coating by increasing the mass density and ensuring greater microstructural uniformity of the W coating and interface, without intermetallic precipitates.

VPS requires specialized equipment, including a vacuum chamber and precise control over parameters such as pressure and temperature, which makes it more expensive compared to other spraying techniques like atmospheric plasma spraying (APS). Additionally, the inability to perform on-site work, along with the high costs and complexity of the process, makes the implementation of VPS particularly challenging. In order to greatly reduce the costs, shorten the fabrication time, and improve the coating durability, APS has also been used to deposit W. The microstructures of W samples manufactured by VPS and APS display similarities, including the laminar structure and columnar crystals within the laminar layers (e.g., see Figure 2); nonetheless, the APS coatings have poorer thermo-mechanical properties due to a higher number of oxides and residual porosity [52,53,54,55,56,57,58].

Figure 2. Microstructure of W deposited by APS (Ref. [58]).

Figure 2. Microstructure of W deposited by APS (Ref. [58]).

To address these limitations, several strategies have been proposed, including the optimization of deposition parameters, powder doping, and post-treatment processes. However, spraying tungsten remains a challenging task due to the numerous variables involved. Producing dense, defect-free coatings requires that powder particles injected into the plasma jet melt completely without overheating. This delicate balance can only be achieved through the precise control of several key parameters, such as gun power, powder particle size distribution, particle velocity, spraying distance, feeding distance, substrate temperature, and surrounding atmosphere [59,60]. When these conditions are carefully optimized, both APS and VPS have demonstrated the capability to produce thick, dense functionally graded (FG) layers with porosity levels as low as 1–2 vol% and free from cracks [61,62]. Moreover, powder doping has emerged as a promising strategy to reduce the detrimental effects of oxygen impurities, which negatively affect the thermo-mechanical performance of W coatings. For example, Heuer et al. [57] successfully used W + WC powders to prevent oxidation and achieve a pure W coating.

Wang et al. [63] employed post-processing treatments involving vacuum and H₂ annealing, thereby reducing oxygen content and porosity and improving the purity of the coating: the decomposition and reduction in W oxide lead to decreased porosity and increased density due grain growth in the form of columnar crystals. The APS coating demonstrated improved thermal conductivity, microhardness, and overall performance after annealing. Ajdari et al. [64] subjected the porous APS coating to a multi-step treatment, consisting in vacuum annealing, cold isostatic pressing, and H₂ annealing, obtaining the elimination of volatile phases from the surface and sub-surface regions. As demonstrated by Jianjun et al. [44], vacuum annealing leads to enhanced microstructure, increased density, significantly reduced oxygen content, and elevated thermal diffusivity. A different approach was proposed by Daram et al. [65], who successfully used friction stir processing to reduce porosity and induce grain refinement of APS coatings. The hardness of the coating was enhanced by a factor of 1.4, and the treatment also improved its resistance to extreme thermal and mechanical stresses common for the fusion reactors.

Laio et al. [66] studied the microstructural changes and performance of W coatings obtained through electron beam remelting treatments (EBRTs), focusing on both short-term and prolonged treatment durations. The as-sprayed W coating exhibited numerous pores and characteristic small lamellar microstructures with a straightforward body-centered cubic phase structure. Following the EBRT treatments, a compact, remelted layer with a columnar crystal structure was formed on the surface of the W coatings. Prolonged EBRT resulted in changes to the microstructure, leading to a reduction in both porosity and oxygen content within the coating. The wear resistance, corrosion resistance, and thermal diffusivity of the W coatings were substantially enhanced.

Wang et al. [67] used submicron W powder with a mean diameter of 600 nm to produce APS coatings, which were then subjected to a vacuum annealing process. The microstructural modifications allowed higher density, improved thermal conductivity (reaching nearly double the pre-treatment level), larger grain size, fewer pores and imperfections, increased density and microhardness, and a decrease in W oxide content to be obtained. APS coatings subjected to hot isostatic pressing (HIP) exhibit a substantial improvement in thermal conductivity [68]. However, the application of this post-process treatment is restricted to small components. Recently, Guo [69] observed that the combination of detonation spraying and HIP significantly improves the microstructure and mechanical properties of W coatings and the performance under transient high heat flux conditions, showing better resistance to the thermal shock and cracking.

Table 1. The effect of post-treatments on the coating properties.

	Vacuum Annealing [48,63,64,67]	H2 Annealing [63,64]	Cold Isostatic Pressing [64]	Friction Stir Processing [65]	Hot Isostatic Pressing [68,69]	Electron Beam Remelting Treatment [66]
Oxygen content reduction	✓	✓	✓	✓		✓
Porosity reduction	✓	✓	✓	✓	✓	✓
Microhardness increase	✓	✓		✓		✓
Thermal conductivity increase	✓	✓			✓

highlights the effects of different post-treatment methods in improving tungsten coating properties. Vacuum annealing, hydrogen annealing, and electron beam remelting are particularly effective in reducing porosity and oxygen content. Other methods, like hot isostatic pressing and friction stir processing, enhance both density and mechanical properties.

Table 2. Comparison of main characteristics of coatings fabricated via different PS techniques and post-processing treatments.

Technology	Parameters	Coating/ Substrate	Coating Thickness	Coating Porosity vol.%	Oxygen Content	Thermal Conductivity W/mK	Ref.
Vacuum Plasma Spraying (VPS)	Gas Ar/H2, 35/15 L/min Spraying power, 40 kW Spraying distance, 250 mm Powder carrier Ar, 2.0 L/min	W/Cu	1 mm	7.6	0.35 wt.%	58	[54]
	75 W/25 Cu and 75 W/25 Cu	W/Cu composite		5.8–6.0	0.65–0.76 wt.%	45–120	[55]
Atmospheric Plasma Spraying (APS)	Gas Ar/H2, 35/15 L/min Spraying power, 40 kW Spraying distance, 250 mm Powder carrier Ar, 2.0 L/min	W/Cu	1 mm	12.9	1.2 wt.%	32	[54]
	75 W/25 Cu and 75 W/25 Cu	W/Cu composite		9.7–11.5	1.31–1.43 wt.%	19–28	[55]
APS + Vacuum Annealing	Gas Ar/H2, 47.5–50/3.5–4.0 L/min Spraying power, 25–30 kW Spraying distance, 70 mm Powder carrier Ar, 5.0 L/min	W	1 mm	N.A.	0.49–1.01 wt.%	33–37	[63]
	Vacuum annealing: 2 h, 800 °C + 3 h, 910 °C				0.36–0.59 wt.%	61–78
APS + H2 Annealing	Gas Ar/H2, 47.5–50/3.5–4.0 L/min Spraying power, 25–30 kW Spraying distance, 70 mm Powder carrier Ar, 5.0 L/min	W	1 mm	N.A.	0.49–1.01 wt.%	33–37	[63]
	H2 annealing: 2 h, 500 °C + 2 h, 600 °C				0.32–0.63 wt.%	65–71
APS + Vacuum Annealing + Cold Isostatic Pressing + H2 Annealing	Spraying power, 27.5 kW Gas Ar/H2, 40/7.0 L/min Powder carrier Ar, 9.4 L/min Spray distance, 8 cm	W/Mo	1 mm	18.0	Reduced	N.A.	[64]
	Vacuum annealing: 2 h, 800 °C + 3 h, 910 °C
	Cold isostatic pressing (175 MPa)			7.0
	Annealing in wet hydrogen: 1.5 h, 1500 °C
APS + Friction Stir Processing	Amperage 350 A Spraying distance 100–140 mm Powder carrier Ar 157.0 L/min	W/W	1 mm		Less than 1 wt%	N.A.	[65]
	Friction stir processing: 12 mm diameter WC-Co plate; rotation speed, 500 rpm; push force, 0.16 MPa; holding time, 30 s			Reduced	N.A.
APS + Electron Beam Remelting Treatment	Gas Ar/H2, 56.5/3.6 L/min Spraying power, 22 kW Spraying distance, 80 mm Powder carrier Ar, 17.5 L/min	W/AISI 316 L	500 μm	3.9	0.6 at.%	N.A.	[66]
	Electron beam remelting treatment (500 ms and a heat flux of 43.63 MW/m2 and 6000 ms and a heat flux of 13.75 MW/m2)			0.2–1.6	0.04–0.28 at.%
APS + Hot Isostatic Pressing	Amperage, 500 A Spraying distance, 320–455 mm Powder carrier Ar, 12–36 L/min	W/AISI 410		1.0–9.8	9.3–22.2 vol.%	9.7	[68]
	HIP (parameters N.A.)			No significant improvement for W coating	N.A.	10.8

displays the main plasma spraying techniques (VPS and APS) and post-processing treatments and their effect on the key coating’s quality characteristics (especially porosity, oxygen content, and thermal conductivity). The table below illustrates that VPS and APS combined with the post-treatments allow good coating quality to be achieved.

4. The Role of the Interface in Joining W Coating and Metallic Substrates

Since transient heat loads in fusion reactors could cause temperature peaks exceeding the safe temperature range of heat sink materials [70,71,72,73,74], the application of the PFMs to protect them is of great importance. Nevertheless, it is not an easy task. In fact, owing to the relevant difference in CTE and the elastic modulus between W and metallic substrates (Cu alloys or martensitic steels), it is necessary to deposit an interlayer to improve the bonding and accommodate the arising stresses which could lead to fracturing and detachment of the coating. To ensure that the coating–interlayer–substrate system performs well, several key features must be attained: low levels of impurities (notably oxides) and porosity, strong coating adhesion, and the prevention of crack propagation. Among the factors that contribute to enhancing the bonding, the deposition temperature and particle velocity play a key role. It is widely acknowledged that particle velocity can lead to higher-density coatings, but it also negatively affects inter-lamellar bonding [75]. The different coating adhesion mechanisms are typically categorized into three distinct groups: mechanical, physical, and chemical/metallurgical [76]. Even when mutually insoluble materials like W and Cu are involved, local melting facilitates closer contact between the two, ultimately enhancing the bond between them [77]. Because of these factors, higher deposition temperatures normally result in enhanced adhesion, even if they may also trigger or intensify the oxidation of metals [75,78,79].

There are two distinct approaches for selecting the interlayer material: (i) a ductile material exhibiting intermediate properties between the coating and substrate or (ii) a functionally graded material (FGM).

In the 1990s, the feasibility of 5 mm-thick coatings on CuCrZr and their capability to withstand heat fluxes up to 5 MW m⁻² in cooling conditions relevant for the ITER reactor were demonstrated [80]. Figure 3 shows the section of a pipe fitting made of CuCrZr alloy coated by a 5 mm-thick W coating deposited through APS: the 800 µm-thick interlayer consists of a mixture of Ni, Al-Si, and Ni-Al obtained by first depositing a layer of pure Ni and then a stratification of layers consisting of graded mixtures of Al-12% Si and Ni-20% Al [81]. The interlayer accumulates stress, protecting the W coating: the total strain of the coating determined by X-ray diffraction is ~5 × 10⁻⁴, reaching a value of ~10⁻² at the interface [33].

The ductile materials used as interlayers include Ti, along with various systems featuring Ni, such as Ni-Al-Si and Ni-Al-Cr. Ni-based systems exhibit exceptional bonding capabilities [30,82,83]. Despite the excellent thermo-mechanical performance of these systems, the presence of Ni represents a serious drawback in nuclear fusion reactors such as ITER and DEMO due to its high activation under neutron irradiation.

Niu et al. [84] investigated the VPS deposition of tungsten coatings onto copper alloy substrate directly and with gradient Cu + W and Ti + W interlayers. The presence of Ti_xW_y and TiH₂ phases was detected in the joints with a Ti interlayer. The interaction between Ti and H₂ plasma forming gas at high temperature is responsible for the formation of the brittle TiH₂ phase [85,86] with consequent weakening of the coating. By using the VPS technique and carefully avoiding the interaction between Ti and H₂ [87], Cho et al. [88] were able to deposit W onto graphite with a Ti interlayer with sufficient adhesion strength after thermal exposure tests.

Recently, different interlayers, made of the same elements of coating and substrate, were developed and investigated. These interlayers are the FGMs fabricated by spraying layers with amounts of the two metals, gradually changing from the substrate to the pure W coating.

The formation of intermetallic precipitates significantly affects the W/steel FG interlayers. During the fabrication of this material, two hard and brittle intermetallic phases, Fe₂W and Fe₇W₆, are observed to be formed using various methods such as VPS, APS, and other techniques [74,89]. Deposition parameters should be carefully controlled to avoid intermetallic phases.

In addition, Heuer et al. [60] observed some intriguing phenomena. The thermal conductivity of the W-steel FGM was not correlated with the quantity of oxides present. The oxides at the splat interfaces in steel coatings are more pronounced due to their greater thickness, compared to those in W coatings, yet the occurrence of steel splats (encompassing Fe-Cr-oxides) does not inherently suggest poorer adhesion than W splats.

The same authors [60] also demonstrated that shorter spraying distances result in reduced interaction between molten droplets and oxygen during the flight phase of spraying. Moreover, the substrate’s pre-heating temperature is a crucial factor influencing the oxidation process, and the appropriate deposition temperature should be selected for each material combination. For example, the performance of the coating with the FGM interlayer is significantly influenced by residual stresses that result mostly from the high temperature alterations that occur during the spraying process. Residual stresses can be classified into two types: the restricted shrinkage of molten droplets as they solidify (quenching stresses) and the difference in thermal expansion between the substrate and the deposited coating as they cool from the deposition temperature to ambient temperature (thermal stresses). Typically, the stresses resulting from quenching are of a tensile nature. On the other hand, thermal stresses can either be tensile or compressive, depending on the difference in the coefficients of thermal expansion between the coating and the substrate. The PS W coating can result in large residual stresses due to the large mismatch in the coefficient of thermal expansion (CTE) between W and the substrate material. Measuring residual stresses caused by manufacturing processes is crucial for evaluating the mechanical reliability of components produced by plasma spraying.

Different methods can be used to evaluate residual stresses, some applications of standard (XRD, nanoindentation and Incremental hole drilling) and novel (in situ coating properties sensor) methods are presented in the following paragraphs, and the results they provide are summarized in Table 3.

Residual stresses of a W coating deposited through PS in Ar-H₂ atmosphere on a substrate of CuCrZr with an FGM W-Cu interlayer were evaluated via XRD and nanoindentation tests [81,90,91], and their values are comparable to thermal stresses; consequently, it appears that a significant portion of quenching stresses is being relieved through yielding. The interlayer is composed of W and Cu splats, displaying a complex and irregular shape. XRD analysis showed that the preferred grain orientation of W and Cu varies depending on their locations within the interlayer. The residual stresses found in the interlayer exhibit varying signs, with compressive stresses present in W and tensile stresses present in Cu. At the midpoint of the interlayer, they achieved the maximum value, where the proportions of the two metals deposited were approximately equal. The residual stresses within the interlayer, measured in W, are significantly lower than the 275 ± 50 MPa threshold for crack formation, as reported in reference [92], but are still not insignificant. Under external loading conditions, these factors may attain a critical stress threshold.

The grain morphology, grain size, and dislocation density of W and Cu do not change across the interlayer while the texture of both metals shows variations depending on the position in the interlayer [93]. The growing hardness seen in the W zones of the interlayer as it moves from the substrate towards the coating is due to the development of the [100] texture component with localized plasticity [94]. Near the substrate and within the interlayer, the strongest Cu texture components are [100] and [110], which correspond to the lowest mechanical strength [95].

Matějíček et al. in [96] employed an in situ coating properties (ICPs) sensor, which was supplemented by four-point bending, hardness testing, and microstructural examinations. Measurements indicated that metallic coatings, including Cu, W, and W-Cu composites, exhibited residual stress levels in the hundreds of MPa without substantial cracking. Both Cu coatings exhibited a consistent coating modulus across the entire strain range that was tested. Hardness exhibited significantly less variability. The residual stress profiles in W/Cu functionally graded materials were substantially influenced by the gradation profile.

Values of residual stresses (Rs) for various systems are displayed in Table 3. To facilitate a clearer comparison of these values, Rs values were approximately calculated using Hook’s law, based on the strain values provided in the referenced works. The Young’s modulus value required for residual stress calculation is provided. It should be mentioned that the Young’s modulus of PS coatings is usually significantly lower than the bulk values, which means that the residual stress values calculated using the theoretical bulk values are overestimated and the real values would be lower. The mock-ups manufactured with the optimal PS parameters and the highest residual stress values within the interlayer were selected.

Table 3. Residual stress (Rs) values for different plasma sprayed coatings.

Coating/Interlayer/Substrate	Young’s Modulus, GPa	Rs Substrate MPa	Rs Interlayer MPa	Rs Coating MPa	Measurement Method	Ref.
APS W (≈1.4 mm)/ FG W/Cu (≈1.2 mm)/ CuCrZr	110 (Cu bulk) 400 (W bulk)	60–86	103 (Cu) −273 (W)	−149	XRD	[90]
APS W (≈1.4 mm)/ FG W/Cu (≈1.2 mm)/ CuCrZr	86 (Cu in [96]) 270 (W in [96])	N.A.	23 (Cu) −115 (W)	N.A.	Nanoindentation	[91]
APS W (≈4 mm)/ Ni-Si-Al (≈800 µm)/ CuCrZr	110 (Cu bulk) 190 (Ni bulk) 68 (Al bulk) 400 (W bulk)	106	884 (Al) 285 (Ni)	180	XRD	[81]
APS W (≈3 mm)/ Al-12%Si/W (500 µm)/ AISI 316	200 (AISI 316) 68 (Al bulk) 400 (W bulk)	1060	29 (Al)	358	XRD	[81]
APS W (≈3 mm), Al-12%Si/W (350 µm)/ AISI 420	200 (AISI 420) 68 (Al bulk) 400 (W bulk)	690	N.A.	−358	XRD	[81]
APS W (≈1 mm)/ FG W/Cu/ Cu	-	150 (prediction)	300	180 (prediction)	In situ coating properties sensor	[96]
VPS W (350–900 µm)/ FG W/steel/ EUROFER97	-	N.A.	−500	−360	Incremental hole drilling	[22]

Table 3. Residual stress (Rs) values for different plasma sprayed coatings.

Coating/Interlayer/Substrate	Young’s Modulus, GPa	Rs Substrate MPa	Rs Interlayer MPa	Rs Coating MPa	Measurement Method	Ref.
APS W (≈1.4 mm)/ FG W/Cu (≈1.2 mm)/ CuCrZr	110 (Cu bulk) 400 (W bulk)	60–86	103 (Cu) −273 (W)	−149	XRD	[90]
APS W (≈1.4 mm)/ FG W/Cu (≈1.2 mm)/ CuCrZr	86 (Cu in [96]) 270 (W in [96])	N.A.	23 (Cu) −115 (W)	N.A.	Nanoindentation	[91]
APS W (≈4 mm)/ Ni-Si-Al (≈800 µm)/ CuCrZr	110 (Cu bulk) 190 (Ni bulk) 68 (Al bulk) 400 (W bulk)	106	884 (Al) 285 (Ni)	180	XRD	[81]
APS W (≈3 mm)/ Al-12%Si/W (500 µm)/ AISI 316	200 (AISI 316) 68 (Al bulk) 400 (W bulk)	1060	29 (Al)	358	XRD	[81]
APS W (≈3 mm), Al-12%Si/W (350 µm)/ AISI 420	200 (AISI 420) 68 (Al bulk) 400 (W bulk)	690	N.A.	−358	XRD	[81]
APS W (≈1 mm)/ FG W/Cu/ Cu	-	150 (prediction)	300	180 (prediction)	In situ coating properties sensor	[96]
VPS W (350–900 µm)/ FG W/steel/ EUROFER97	-	N.A.	−500	−360	Incremental hole drilling	[22]

The data in the table illustrate the impact of various interlayer materials on residual stress levels, thereby highlighting the benefits of using FG interlayers. The different substrate materials could also affect the joint stability. It is noticeable that in the system of W plasma sprayed on the AISI 316 from [81], the maximum residual stress was accumulated in the substrate, and the interlayer did not play its role as a “buffer” between the substrate and the coating, leading to coating exfoliation from the substrate. Instead, W coating on AISI 420 showed good adhesion.

In addition to experimental campaigns, some investigators employed numerical computation to study the behavior of different interlayers and determined the most favorable parameters [97,98,99,100]. For instance, finite element modeling (FEM) was used to examine the case of VPS-W on Cu, taking into account the interlayers W/Cu, Ti, and NiCrAl [90,100]. The results of these simulations indicate that W/Cu performed optimally in terms of minimizing stress concentration at the interface without causing a substantial rise in surface temperature.

The authors in [101,102,103,104] found that FGM interlayers with at least 1 mm thickness are required to have a significant impact on the steel substrate, with greater thickness also reducing creep but increasing temperature in the W and the interface. Furthermore, Dose et al. [105] demonstrated that higher amounts of single layers in the FGM system result in a decrease in the desired maximum stress within the FGM.

Recently, FGM interlayers of W/EUROFER have been tested on large components to assess the feasibility of the VPS technique on real parts [106]. The mock-ups up to 500 × 250 mm² were successfully fabricated for the first time. The results are promising and show no heat accumulation, but further improvements are still necessary since the corners of the plates show different mechanical properties and therefore might be weak spots.

5. Coating Surface Erosion Under Steady-State and Transient Heat Loads

As previously discussed, PFMs have to withstand steady-state thermal loads up to 20 MW/m² and transient high energy loads such as disruptions (20–100 MJ/m²), edge-localized modes (ELMs—1–3 MJ/m²; duration 0.1–10 ms), and vertical displacement events (VDEs—~60 MJ/m²; duration 100–300 ms) which may cause melting, vaporization, surface cracking, and materials damage. Therefore, experimental tests by using various approaches and numerical calculations simulating the above-mentioned heat loads have been carried out, and the main results are illustrated in this section.

Tokunaga et al. [107] investigated the VPS-W/F82H coating through experiments involving steady and cyclic heat loading induced by an electron beam, and finite element analysis (FEA) of temperature patterns and thermal strain. No macroscopic fractures or exfoliation occurred under steady heat loading conditions with surface temperature in the range 700–750 °C. The thermal and adhesion properties at the interface of VPS-W-coated F82H joints have been found to be satisfactory. Conversely, exfoliation has taken place at the interlayer of the VPS-W coatings near the interface between the VPS-W and F82H due to cyclic heat loading at 1300 °C.

The same VPS-W/F82H system with a 1 mm-thick W coating was brazed onto an oxygen-free high-purity copper (OFHC) block that featured a cooling tube, with the aim of evaluating its suitability as a PFM [108]. The heat flux was modified from 0.38 to 3.4 MW/m² while the parameters of water flow (velocity, 18.0 m/s; pressure, 0.7 MPa; room temperature) were kept constant. It was observed that VPS-W/F82H/OFHC samples typically reached a lower surface temperature than APS-W/F82H/OFHC samples. The temperature difference, which reaches 435 °C at a heat flux of 3.4 MW/m², was attributed to different thermal conductivity between the two types of coatings. Surface modification, exfoliation, and cracking were not induced by thermal fatigue experiments conducted up to 200 cycles with a heat flux of 3.2 MW/m².

Other studies on the performance of APS W-coated components, specifically those with W-Cu water-cooled parts, subjected to brief thermal shocks evidenced partial melting of the W surface, and with thermal loads of 2.0 GW/m² the molten zone is deeper, and numerous cracks and precipitates were observed [109].

Single-pulse tests were performed by using an electron beam with a pulse duration of 5 ms on the APS pure W and APS W-Cu FGM coatings at various (0.22, 0.34, 0.56, and 0.9 GW/m²) power densities in the JUDITH facility [110]. SEM observations revealed micro-cracks in both coatings at the lowest power density of 0.22 GW/m². As power density increases, particle release and coating surface melting occur with more pronounced effects in pure W coating as compared to the APS W-Cu FGM one. The phenomenon is accompanied by cracks running parallel to the surface in the vicinity of the melted area, but no vertical cracks were detected. The authors attributed such behavior to the coating’s lamellar structure.

Physical simulations of transient heat loads similar to ITER disruption events were carried out by using a laser pulse (wavelength λ = 1064 nm, pulse duration τ = 15 ns, pulse energy Ep~8 J, size of the focal spot Φ = 200 µm, surface power density on the focal plane I = 1.7 × 10¹² W·cm⁻²) [58,111,112]. The parameters correspond to a plasma electronic temperature T_e ≈ 1.218 × 10⁶ K at the surface and allow the material degradation to be simulated during the plasma disruption process. Each pulse produces a crater surrounded by a ridge resulting from molten metal ejected and flowing outward; they have a surface with plates surrounded by jagged boundaries and cracks, due to thermal stress. As shown in the example in Figure 4, the shape and depth of the crater depend on the material (bulk W, plasma-sprayed W, and W-1%La₂O₃). A relevant role is played by thermal conductivity that is different in bulk W with equiaxed grains, plasma sprayed with a lamellar structure, and W-1%La₂O₃ with oxide particle distribution. A numerical model was developed by using Ansys code and APDL language, and data from calculations were used to determine the crater shape in both bulk and PS-W coatings and compared to experimental data [113].

Among the tests carried out to investigate the coatings’ behavior when subjected to irradiation, Hou et al. [114] examined the impact of the presence of TiC on the microstructure and radiation resistance of W-based coatings produced via supersonic atmospheric plasma spraying (SAPS). Experiments involving irradiation were performed by means of a transverse flow CO₂ laser with a power of 2.50 kW, a laser spot diameter of 2 mm, and an energy density of approximately 10^10–12 W/cm², resulting in a surface temperature in the order of 10⁵ K. The as-sprayed TiC-free and TiC-doped coatings primarily consisted in a lamellar structure. The TiC phase was mainly found in the lamellar gaps of the coating and exhibited a strip-like morphology. Comparative analysis revealed that the as-sprayed TiC-doped coating exhibited a compact microstructure, significantly reduced porosity, decreased O content, and enhanced thermal conductivity compared to the as-sprayed TiC-free coating. Finally, laser irradiation remelted areas, unremelted areas, and/or regions with mixed characteristics were observed. The TiC-free coating was more prone to developing cracks than the TiC-doped one. Within the remelted areas, the strip-like TiC particles modified into spherical ones and were predominantly found along the W grain boundaries contributing to grain refinement.

Kikuchi et al. [115] tried to replicate the predicted conditions of I ELMs for the ITER operation by employing a magnetized coaxial plasma gun. They then exposed F82H steel with a VPS W coating to 10 plasma pulses, each of approximately 0.3 and 0.9 MJ m⁻². No surface cracking or morphology modification was observed for the plasma exposure of 0.3 MJ m⁻², but they emerged at 0.9 MJ m⁻². Regardless, despite the surface melting, no evidence of the W coating detached from the F82H substrate was found. The heat diffusion length within the W coating during the plasma pulse was estimated to be between 100 and 200 μm, a distance that is smaller than the W thickness; therefore, negligible thermal stresses are produced at the interface between the W coating and the substrate.

Another extreme loading test was performed by Kim et al. [116] by employing a hollow-electrode DC thermal plasma torch with a power output of 300 kW. The target was at a distance of 50 mm from the torch and was mounted on an actively cooled holder. The targets (cold-worked bulk W and PS-W) were exposed to 1.1–2.5 GJ/m² of energy density, corresponding to 1000–5000 occurrences of mitigated ELMs in ITER [117]. The total transferred energy was regulated by the duration of irradiation (180–420 s). PS-W exhibited a higher surface temperature compared to bulk W due to its lower thermal conductivity, a direct consequence of the lamellar structure. In bulk W, recrystallization was observed in a layer near the surface together with vertical cracks along grain boundaries. On the contrary, no vertical cracks occurred in PS-W: an annealing effect causes neighbor lamellas to merge, destroying inter-lamellar interfaces and preserving the material from the occurrence of vertical cracks. The changes are similar to the phenomena occurring in crystal grains [86]. On these grounds, it could be suggested that a multilayered W structure is advantageous, especially for long pulse operation in a future fusion reactor.

An issue of high technological relevance, namely the impact of the W powder size employed in the production of the APS coating, was investigated by Huang et al. [118]. Commercial W powders with median particle sizes of 0.6 µm and of 6 µm were deposited onto Cu substrates with sub-micrometric (SMW) and micrometric tungsten (MW) coatings, respectively. The SMW coating exhibits lower porosity and O content than the MW one. The experiments were performed at the high-intensity current pulsed electron beam facility, SOLO-M, in Tomsk, Russia. The samples were subjected to 25 pulses with a duration of 0.2 ms and power densities of 0.4, 0.6, and 0.8 GW/m². The SMW coating cools more rapidly than the MW one, the effects of superfast melt quenching are smaller, and the grains have nanocrystalline size [71]. A small portion of the APS-W coating surface was damaged by a shock wave resulting from HCPEB irradiation [119,120], and the mass loss of the SMW coating is greater than that of the MW coating because the SMW structure is less adherent.

Table 4 summarizes the results of the W coating irradiation tests with various energy sources used to simulate steady-state and transient heat load events. The heat load parameters, surface damage, and quantitative values such as coating and substrate surface temperature and material loss are reported.

Table 4. Results of W coating testing with electron beam and laser irradiation and pulsed plasma discharge to simulate steady-state and transient heat load events.

	Coating/Interlayer/Substrate	Parameters	Surface Damage	Measured Values	Ref.
Electron Beam Irradiation	VPS-W (0.6 mm)/-/RAFM steel	7.5 MW/m2 180 s	No modification	Tw = 700 °C Tsteel = 500 °C	[107]
		12 MW/m2 30 cycles: 60 s ON, 140 s OFF	Fine modification	Tw = 750 °C Tsteel = 350 °C
		40 MW/m2 30 cycles: 7 s ON, 230 s OFF	Cracks, exfoliation	Tw = 2100 °C Tsteel = 350 °C
	VPS-W (1 mm)/-/RAFM steel brazed to OFHC cooling tube	0.38–3.4 MW/m2 Ramp-up/plateau, ramp-down/rest 20/40/0 s	No modification	Tw = 350–1000 °C Tsteel = 50–250 °C TOFHC = 50–100 °C	[108]
	APS-W (1 mm)/-/RAFM steel brazed to OFHC cooling tube			Tw = 300–600 °C Tsteel = 50–350 °C TOFHC = 50–100 °C
	APS W (0.5 mm)/-/Cu	0.4–2.0 GW/m2 0.2 ms, 25 pulses.	Remelting, recrystallization, impurities, precipitates, cracks	T = 3500 °C	[109]
	APS W (1.5 mm)/-/Cu	3–8.5 MW/m2	No modification	T = 520–880 °C
	APS W (1 mm)/-/OFHC	0.22–0.9 GW/m2 Pulse 5 ms	Micro-cracks, particle release, surface melting	Materials loss 28 mg	[110]
	APS W (0.7 mm)/FG W-Cu (0.3 mm)/OFHC			Materials loss 18 mg
	APS W (micron and submicron powder size)/-/Cu	0.4–0.8 GW/m2 0.2 ms, 25 pulses	Remelting	T = 3500 °C Material loss 20–22 mg	[118]
	CVD W/-/Mo		Cracks, large internal stress	T = 1250 °C Material loss 1 mg
Laser irradiation	Bulk W	1.7 × 107 GW/m2 Pulse duration 15 ns Spot Ø 0.2 mm	Crater, ridge, plates with jagged boundaries and cracks, ablation sites from the pores	Material loss 0.37 × 105 µm3 Crater Ø 75 µm	[58,111,112]
	APS W/FG W-Cu/CuCrZr			Material loss 0.33 × 105 µm3 Crater Ø 80 µm
	W-1%La2O3 bulk			Material loss 3.06 × 105 µm3 Crater Ø 300 µm
	Bulk Mo			Material loss 4.00 × 105 µm3 Crater Ø 300 µm
	Supersonic APS Pure W	106–107 GW/m2 Traverse velocity 300 mm/min Spot Ø 2 mm	Remelting, cracks	N.A.	[114]
	Supersonic APS W doped with 1.5 wt% TiC		Remelting
Pulsed plasma discharge	W bulk (ITER grade)	1.5 MW/m2 0.2–0.6 ms, 10 pulses	Crack	N.A.	[115]
	W-2%wt Ta alloy		Crack
	VPS-W/-/RAF/M steel		Remelting and crack formation
	Cold-worked bulk W/Mo/graphite	4.3–5 MW/m2 pulse duration of 180–420 s	Recrystallization, bubbles, cracks	Recrystallization depth 1.7–3.0 mm	[116]
	APS W (0.3 mm)/Mo/graphite		Better resistance under long pulses

Table 4. Results of W coating testing with electron beam and laser irradiation and pulsed plasma discharge to simulate steady-state and transient heat load events.

	Coating/Interlayer/Substrate	Parameters	Surface Damage	Measured Values	Ref.
Electron Beam Irradiation	VPS-W (0.6 mm)/-/RAFM steel	7.5 MW/m2 180 s	No modification	Tw = 700 °C Tsteel = 500 °C	[107]
		12 MW/m2 30 cycles: 60 s ON, 140 s OFF	Fine modification	Tw = 750 °C Tsteel = 350 °C
		40 MW/m2 30 cycles: 7 s ON, 230 s OFF	Cracks, exfoliation	Tw = 2100 °C Tsteel = 350 °C
	VPS-W (1 mm)/-/RAFM steel brazed to OFHC cooling tube	0.38–3.4 MW/m2 Ramp-up/plateau, ramp-down/rest 20/40/0 s	No modification	Tw = 350–1000 °C Tsteel = 50–250 °C TOFHC = 50–100 °C	[108]
	APS-W (1 mm)/-/RAFM steel brazed to OFHC cooling tube			Tw = 300–600 °C Tsteel = 50–350 °C TOFHC = 50–100 °C
	APS W (0.5 mm)/-/Cu	0.4–2.0 GW/m2 0.2 ms, 25 pulses.	Remelting, recrystallization, impurities, precipitates, cracks	T = 3500 °C	[109]
	APS W (1.5 mm)/-/Cu	3–8.5 MW/m2	No modification	T = 520–880 °C
	APS W (1 mm)/-/OFHC	0.22–0.9 GW/m2 Pulse 5 ms	Micro-cracks, particle release, surface melting	Materials loss 28 mg	[110]
	APS W (0.7 mm)/FG W-Cu (0.3 mm)/OFHC			Materials loss 18 mg
	APS W (micron and submicron powder size)/-/Cu	0.4–0.8 GW/m2 0.2 ms, 25 pulses	Remelting	T = 3500 °C Material loss 20–22 mg	[118]
	CVD W/-/Mo		Cracks, large internal stress	T = 1250 °C Material loss 1 mg
Laser irradiation	Bulk W	1.7 × 107 GW/m2 Pulse duration 15 ns Spot Ø 0.2 mm	Crater, ridge, plates with jagged boundaries and cracks, ablation sites from the pores	Material loss 0.37 × 105 µm3 Crater Ø 75 µm	[58,111,112]
	APS W/FG W-Cu/CuCrZr			Material loss 0.33 × 105 µm3 Crater Ø 80 µm
	W-1%La2O3 bulk			Material loss 3.06 × 105 µm3 Crater Ø 300 µm
	Bulk Mo			Material loss 4.00 × 105 µm3 Crater Ø 300 µm
	Supersonic APS Pure W	106–107 GW/m2 Traverse velocity 300 mm/min Spot Ø 2 mm	Remelting, cracks	N.A.	[114]
	Supersonic APS W doped with 1.5 wt% TiC		Remelting
Pulsed plasma discharge	W bulk (ITER grade)	1.5 MW/m2 0.2–0.6 ms, 10 pulses	Crack	N.A.	[115]
	W-2%wt Ta alloy		Crack
	VPS-W/-/RAF/M steel		Remelting and crack formation
	Cold-worked bulk W/Mo/graphite	4.3–5 MW/m2 pulse duration of 180–420 s	Recrystallization, bubbles, cracks	Recrystallization depth 1.7–3.0 mm	[116]
	APS W (0.3 mm)/Mo/graphite		Better resistance under long pulses

Fusion reactor coatings face extreme conditions, requiring them to withstand steady-state heat fluxes of up to 20 MW/m² and transient events like disruptions, ELMs, and VDEs. Experimental testing using electron beams, laser pulses, and high-powered plasma guns simulates the operating conditions of fusion reactors. The effectiveness of coatings in resisting thermal shocks varies, with denser and treated coatings showing better performance. At high heat loads, recrystallization and surface modifications can occur, negatively impacting long-term stability.

Further investigation in future studies could reveal the potential for developing other types of interlayers to reduce high thermal stresses at the interface between the substrate and coating caused by different coefficients of thermal expansion. In particular, a V interlayer between W and steel showed promising results: V has a coefficient of thermal expansion in between W and steel and forms a solid solution with both. Nevertheless, the manufacturing process should be improved to reduce or, better, prevent the non-ideal contact between the coating and substrate that act as thermal barriers [121]. The same study demonstrated a curious result that direct W-steel joints show the best resistance under heat loads, and a deeper exploration of this phenomenon is crucial for performance optimization. At high heat loads, recrystallization and surface modifications can occur, negatively affecting long-term stability. Experiments conducted on the WEST tokamak [122] have provided some insight, though limited details are available at the moment regarding the plasma-sprayed coated PFCs. Nevertheless, post-mortem characterization is going to give some more information in the years to follow. Simulation experiments for VDE events with PS W, similar to those performed for bulk and W-La₂O₃ samples as referenced in [123], would also be extremely valuable for gaining a better understanding of the material’s behavior.

6. Conclusions and Future Directions

This work examined the achievements and problems related to the deposition of W and W-based materials on metallic substrates for nuclear fusion applications through plasma spraying technologies. From the revised literature, the following main conclusions can be drawn:

(i)

The adhesion of plasma-sprayed W on metallic substrates can be successfully attained through the deposition of suitable interlayers, which could be either ductile alloys or FGMs made of W and substrate metal.

(ii)

Since ductile alloys often are made of elements which activate under neutron irradiation, the deposition of FGMs seems to be the most promising approach.

(iii)

Among plasma spraying methods, APS combined with post-processing by means of techniques such as vacuum annealing, friction stir processing, or electron beam remelting represents a valid alternative to VPS with lower costs and similar performances in terms of coating density, hardness, and thermal properties.

(iv)

Since erosion is the main cause of component damage and plasma contamination, the present results indicate that bulk W and PS-W are the best choice as PFMs because they exhibit lower erosion if compared to W-1%La₂O₃ and Mo. The PS W coatings exhibit excellent thermo-mechanical resistance under constant and cyclic heat loads and are even superior to that of bulk W under long pulses. The coating density, however, remains a crucial factor in determining the coating quality for this application.

Based on the information gained from the literature, the authors would like to suggest some relevant topics for future research in the field of W-based plasma-sprayed coatings:

(i)

Testing alternative interlayers using the APS technique, particularly the V interlayer, which has shown promising results. The same studies have revealed that the most effective results were achieved with samples featuring direct W-EUROFER joints. A more in-depth examination of this unexpected finding would be of significant interest.

(ii)

Recently, scaled-up mock-ups were successfully produced by VPS. The results are promising and show the potential of technology transferring to the industrial scale. The tests of scaled up fabrication via APS could be very interesting, taking into account the cost efficiency of this technology.

(iii)

Experiments were conducted on the WEST tokamak, with limited details available regarding the plasma-sprayed coated PFC, suggesting the need for more comprehensive research in this area.

(iv)

Simulations of VDE events with PS W would be highly beneficial for understanding the behavior of PS W-based coatings.