Synergistic effects of W–Ta alloys exposed to various ELM-like fusion events

Recently tungsten alloys have shown capabilities in suppressing morphological changes when exposed to harsh fusion conditions. Morphological changes due to He+ and D+ ions at high temperatures (1000–2000 K) can lead to easier erosion and sputtering of tungten (W) particles towards quenching the fusion plasma. This study investigated W–3Ta alloy as potential plasma facing materials (PFMs) exposed to different edge-localized mode (ELM)-like energies and ion loadings. Performance of W–3Ta has been tested as a function of ELM-like energies (0.6–1.5 GW m−2) and varying He+ and D+ ion concentrations (100% He+, 50% He+:50% D+,10% He+:90% D+ (similar ELM-like reactor condition), and 100% D+ ions). Early-stage fuzz formation and surface pores were observed during 100% He+ ion irradiation and mixed ion loadings but not observed during 100% D+ ion loadings. Pores decreased in quantity and increased in size with increasing transient heat loading. Grain boundaries acted as trapping sites of helium and deuterium atoms during dual beam ion irradiations, as observed through increased pore size along the grain boundary. Qualitative erosion estimates were measured using witness plates. Increased erosion was observed with increasing D+ ion concentration, which is attributed to D supersaturation in the near surface region affecting thermal and mechanical properties. More erosion was observed for W–3Ta samples compared to pure W, evident through large (>10 µm) cracks on the surface, and an increase in W metal x-ray photoelectron spectroscopy peaks under similar loading conditions. The slight enhancement in erosion is attributed to the slightly decreased thermal properties, ∼10% thermal diffusivity drop, of W–3Ta at high temperatures. The damage threshold was impacted by ion loading, where distinct cracking was observed for 100% He+ at a lower heat loading. This matrix study using reactor-like loading supports the need for extensive studies on future W alloy PFMs before selecting one as an alternative.

Original content from this work may be used under the terms of the Creative Commons Attribution 4.0 licence.Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.

Introduction
Tungsten (W) has long been studied as the main plasma facing material (PFM) for the ITER divertor due to its high thermal conductivity, high melting temperature [1], and reduced tritium retention [2][3][4].Serious issues may occur during long duration high particle flux interactions with these PFMs and/or divertor materials of fusion tokamak reactors due to drastic surface morphology changes.Helium (He) ions form nanotendril growths on the surface via He + ion implantation at temperatures between 1000-2000 K on W, with this phenomenon during reactor transient operations [5][6][7].The mechanisms for nanotedril formation, parameter specifics, and the resulting microstructural effects have been studied extensively over the last two decades [8,9].Deuterium (D) irradiation plays a very different role and D + ions can coalesce leading to blisters at surface temperatures below 700 K [10][11][12][13].During simultaneous He + and D + , as well as D + only, irradiations at high temperatures the tungsten surface did not show any signs of blistering due to D + ion rapid desorption while the He + ion produced the typical porous fuzz structures [14].He + ions cannot be avoided with current plasma concentrations, as it is one of the byproducts of D + and Tritium (T + ) fusion reactions.In recent years the study of plasma disturbances known as edge localized modes (ELMs) have been studied extensively with signs of massive erosion if not mitigated [2,15].The concentration of W within the fusion plasma is of concern, since concentrations as low as 10 −5 could quench the reaction [2].Previous work simulating fusion environments have either consisted of He + and D + plasmas formed via premixed ratios of gases [16,17] or have focused on plasma concentrations that would have more He than D in a tokamak device [18,19].Some studies have focused on lower concentrations of He + compared to D + ions [20], but very few have combined these ion loadings with transient heat loading to simulate an ELMlike event.Transient heat loading has been simulated using millisecond lasers (this study) [21,22], electron beams [23], and quasistationary plasma accelerators [24].A few recent studies at near reactor conditions using mixed gases and one ion beam have been completed by another group on ITERgrade tungsten [25,26] No study to the best of our knowledge has investigated near reactor ion loadings via dual beam ion irradiations simultaneously with transient heat loading at various ELM energies in a near complete nuclear fusion matrix environment.
As previously mentioned the damaging mechanisms of each, D + and He + irradiations, occur in different temperature regimes.Virtually all D is desorbed from the bulk of W below 1100 K with decreased retention with increasing temperature [27,28].In this study the irradiation temperature was ∼100 K higher than the maximum total desorption temperature of D and thus the D is assumed to be fully desorbed during the irradiation process.While for He the threshold temperature for desorption is ∼1000 K and majority of He is desorbed below 2000 K, which aligns well with the tungsten fuzz window [29,30].During mixed ion beam irradiations, the D + ions may escape the surface through bubbles formed by the He + ions at temperatures below 800 K as evident by previous thermal desorption spectroscopy experiments on W [31,32].The pores created by He + irradiations also act as sinks and immobilize D + near the surface in which less D was retained in the bulk [33].Both helium and deuterium have decreased mobility at grain boundaries and impurities, seen experimentally and theoretically [34][35][36].He mobility near the grain boundary and preferential growth has been observed experimentally to greatly be affected by the incoming ion energy and base sample temperature [36].When more D + concentration is present, the D + ions tend to saturate the surface competing against the He + ions and disrupt bubble formation and nucleation to induce signs of early-stage fuzz formation.This phenomenon leads to suppressed fuzz formation on the surface due to the D + ion super-saturation and has been seen in many previous studies [19,20].When ion irradiations and transient heat loadings were combined the cracking networks led to deeper mobility of ions into the material [21,37].When pure W was exposed to simultaneous ion loading (D and He) and ELM-like heat loading the sample eroded more material via splashing and dust formation [5,19,38].More erosion is most likely due to the D gas building up quickly sub surface and with a roughened surface more of the heat load is absorbed and can not easily be dissipated.
With the emerging interest in DEMO reactors as the next generation fusion pilot plants, there has been interest in W alloys and advanced alloys to combat morphological changes due to low energy ion irradiation.Alternative elements, mostly high-z refractory metals, have been investigated in previous studies [39][40][41] with promising hindrance in sputtering and erosion, but lack the thermal properties W has. In general, one next logical approach is to investigate these promising alternatives as alloying materials with W.Many alloys have shown to reduce fuzz growth by allowing faster desorption of the He + ions or lower entrapment in general.Most W alloys, such as those alloyed with rhenium, have shown to improve ductility and decrease the risk of crack formations due to thermal shock [42].W-Ta is one of the alloys being investigated due to its enhanced resistance to thermal shock [43].W-Ta alloys have also experienced less dense fuzz under similar irradiation parameters to pure W [44].In this study W-3Ta is investigated because of its ability to suppress fuzz growth and cracking under ELM-like heat loadings, while having a slightly favorable thermal conductivity over W-5Ta.
One of the major focuses of this study is to investigate if W-Ta alloys are a good candidate for future fusion reactors in a near complete nuclear fusion environment and matrix study including various simultaneous ion loadings paired with transient and steady state heat loads.In this study, dual ion beams from two sources equidistant from the heater and sample surface with applied transient heat loading from a pulsed ms laser simultaneously interacted with the W-Ta alloy.Extensive Field-Emission Scanning Electron Microscope imaging, and focused ion beam (FIB) was used to compare both surface and subsurface morphology changes in addition to detailed xray photoelectron spectroscopy (XPS) for elemental analysis.Optical profilometry was used to measure broad scheme surface melting and cracking effects while witness plates assisted in measuring surface erosion.

Experimental setup
W-3Ta (3 wt.% Ta) bulk material which was created by sintering <10 µm powder particles at 1773 K was obtained from American Elements.Four samples of pure-W (99.95% purity) were prepared from a single cold rolled and annealed sheet of tungsten supplied by Alfa Aesar.All the samples were rotary cut to 10 mm × 10 mm squares 2 mm thick and polished to a mirror like finish using SiC metallographic polishing equipment up to 1200-grit.Wt. % was verified using XPS post 1 keV Ar sputter cleaning confirming the wt.% within 0.3% [44].The average grain size pre-irradiation for the W-Ta alloy and pure-W was ∼3-4 µm, from previous EBSD measurements [20].The samples were exposed simultaneously to fusion like environments including; steady state heat flux via a 1-inch button heater (Heatwave Labs Inc. 101 491), transient heat loading via an Nd:YAG ms laser (Miyachi Unitek LW25A) equipped with fiber optic transport and collimating lenses, and dual ion beam loading (He + and D + ) via two end hall ion beams (Kaufman & Robinson EH-400).Detailed schematic of the ultra-high flux irradiation (UHFI-II) chamber can be seen in figure 1 in which all the experiments were carried out close to a complete nuclear fusion environment.All experiments were conducted continuously for 1 h with steady state heat loading, transient heat loading, and ion beam irradiations.During all experiments, samples were kept at a constant steady state temperature of 1223 K via a type K thermocouple and PID temperature controller.For precise temperature measurement verification, a single wavelength pyrometer was focused on the surface of each sample.Transient heat fluxes of 1 ms duration pulses at 1 Hz were imposed onto the W-3Ta samples, resulting in a total of 3600 simulated transient events.Different ELM-like heat loadings consist of 0.6, 0.9, 1.2, and 1.5 GW m −2 , noting the lower as an approximate melting threshold and upper limit bounded by maximum ITER heat flux in the divertor.Heat flux calculations were based on a 2 mm diameter spot from burn paper and power meter data.The data was taken inside the ultra-high vacuum (UHV) chamber, just before the sample surface, before starting each and every experiment.The larger diameter beam profile reduces hot spots and allows larger areas to characterize the transient heating effects.All ion loadings were completed at an energy of 100 eV while the concentrations varied to encapsulate possible loadings within the fusion reactor.Ion fluences were measured using improved Faraday cup measurements incorporating the use of high temperature shielded BNC cable connected to an ammeter outside of the chamber by using a BNC feedthrough.Measurements were taken at 22.5 • from each ion source as they are separated by 45 • with the laser irradiation normal to the surface.Fluences for He + and D + ion loadings were 1.08 × 10 23 ions m −2 while dual ion beam irradiations of 50:50 and 10:90 (D + : He + ) had total ion fluences of 2.16 × 10 23 ions m −2 for maintaining beam stability.A complete detailed list of the 17 irradiations, including those with no transient heat loading, are listed in table 1 along with sample IDs used throughout these studies.Sample IDs colored in purple represent tests completed under identical conditions for pure-W samples for later comparison.During all irradiations, a Mo witness plate (at 45 • angle to the sample surface on the heater) was used to collect eroded material from the sample.Witness plates were analyzed via ex situ XPS in CMUXE IMPACT-II.During the XPS measurements, the photoelectrons were excited by an xray source of Mg-Kα (energy = 1253.6eV) radiation source (SPECS XRC-1000), and the emitted photoelectrons were analyzed in an Omicron Argus hemispherical electron analyzer (EAC2000 SPHERA 547) using a round aperture of 0.63 mm (for imaging-XPS) and a second aperture at 6.3 mm (for XPS).All XPS spectra were analyzed with commercial Casa-XPS software using Shirley background subtraction [45,46].Other post-irradiation characterization consisted of SEM micrographs and FIB/SEM cross-sections using a FEI Quanta 3D FEG equipped with a liquid Ga + ion beam.Finally, white light profilometry equipment (Zygo ZeScope) was used to analyze surface roughness due to the transient heat loading.All post-analysis were used to characterize the morphology changes on the sample surface and subsurface due to ion exposures.The commercial ImageJ software was used for post-analysis of SEM micrographs to enhance the quality and to also calculate pore area and density [47].

Absence of transient heat.
Particle loading will occur during normal operations, typically at lower fluxes than transients, nevertheless these cases need to be initially studied separately.These cases would also be helpful in future devices when ELMs can successfully be mitigated.If disruptions are mitigated the surface fuzzy structures are likely to grow to an equilibrium in depth and size.In the case of ELMs there can be increased erosion of these surfaces due to the fuzzy structures having a significantly lower thermal conductivity.In this study samples with no applied heat loading (XXHe|XXD-0 in table 1) were first loaded with different compositions of He + and D + ion loadings to serve as a basis of comparison for samples exposed to ELM-like transient heat loadings.SEM micrographs were taken of each sample post 1 h irradiation at the compositions outlined and repeated for clarity in figure 2. Low magnification SEM micrographs are presented in figures 2(a)-(d) to show grain sizes more effectively while high magnification micrographs are presented in figures 2(e)-(h) to better analyze surface pores.As expected for 100% He + ion loading there is visible evidence of early-stage nanotendrils (fuzz) on the surface.This is seen as a shale-like structure on the surface, as well as the apparent micrograph contrast/brightness differences for objects protruding from the surface.These fuzz-like structures have been known to form in the temperature operating range from 1000-2000 K for W, in which these experiments were conducted at 1223 K [6].He + ion bubbles form below the surface, building up causing these nanotendril growth from the surface [48].
Samples that contain partial amounts of D + ions show significant hindrance of early-stage fuzz formation due to D + ion super-saturation on the surface [18,19].The most probable pore sizes decrease with increasing D + concentration, evident in the bottom of figure 2 and also seen in the pore analysis later in figure 5. We noticed an increase in surface pore population density for mixed plasmas where reducing the concentration ratio of He and D, i.e. higher D + than He + , (figure 5).This can be attributed to increase in D + ions hindering the mobility of He + , thus the helium conglomerates locally.The size of the pores caused by the He + also indicates a reduction in volume before bursting which is directly influenced by the concentration of He + ions implanted.At 100% D + ion concentration, no pores are found on the surface, this indicates that the D is saturated on the surface and desorption is almost instantaneous.Rapid desorption can be assumed because the operating temperature was 1223 K, well above the blistering region (273-700 K) [11,12], as well as, beyond the maximum desorption temperature of 1100 K observed in other W studies [27,30].Also, the He + ions are noted as the sole cause of pores in this study as again no pores are present in the D only irradiations.The appearance of grain boundaries are easily distinguishable in micrographs of samples exposed to D + ions in figures 2(b)-(d).The early stage fuzz formation on the He + only exposure 'hides' the grain boundaries so that they are not easily distinguishable.The average grain size was ∼500 nm for all ion loadings with D + which was calculated using the linear intercept method.The appearance of these grains and reduction in size were also observed in the case of W-5Ta samples in a previous study [20].This phenomenon is attributed to the high temperature annealing alongside the D + ions possibly lowering of the recrystallization temperature, known as irradiationenhanced recrystallization [49,50].

Synergistic effects (He
The addition of transient heat loading in the form of ms laser pulses at 1 Hz to simulate large ELM-like events affected the surface morphology significantly.Micrographs of the whole laser irradiated area are seen in figure 3 for 0.6 and 1.2 GW m −2 heat loadings.Surface cracking was observed in all samples except those of mixed ion loading at 0.6 GW m −2 which under these experimental conditions appeared to not exceed the damage threshold.This indicates that the transient heat loading of 0.6 GW m −2 did not surpass the melting threshold for W-3Ta samples; this is similar to other studies of W and W-Ta alloys [51][52][53].Larger cracks were observed on the laser only exposure (figure 3(g)) indicating that the ion irradiation reduced the overall stresses in the material.Samples exposed to the same flux, but different ratios of He + and D + ions, 50He|50D and 10He|90D, are imaged in figures 3(c), (e) and (d), (f ) at two different heat loadings of 0.6 and 1.2 GW m −2 respectively.Although no cracks were seen at 0.6 GW m −2 samples, massive crack networks were seen on samples exposed to 1.2 GW m −2 as a result of doubling the heat loading.Crack networks appear to be larger in the sample exposed to a higher D + concentration which may be a result of reduced thermal conductivity at the surface where the deuterium is rapidly being desorbed.Larger surface roughness caused by the larger cracks were also observed via optical profilometry (figure 6).These results are contrary to previous investigations by Gago et al that showed an increase in deuterium, at the same total He + +D + flux, caused less severe cracking [25].Though noted in the article, the results are heavily dependent on the ion flux and heat loading applied where differences in cracking observed were not clear between different samples.Here, we show the cracking appears to be more severe with mixed plasma irradiation that has imore deuterium than helium, at the same ion flux.
The small crack features found on the sample exposed to He + ions only, with the lowest power heat loading, 0.6 GW m −2 are imaged in figure 3(a) with an inset to better show the small cracks.The cracks are most likely caused by the He + ions reducing surface reflectivity during fuzz growth leading to more laser absorption and thus heat load on the where the left is for a laser only exposure (g) and the right is for deuterium ion loading with the applied heat loading (h).Insets shown on micrographs (a), (g), and (h) show higher magnification images from their marked respective regions.
surface [22,40,54].The fuzz structures have a lower thermal conductivity leading to easier ablation alongside the changes in optical reflectivity [5].These phenomena are captured while measuring the surface pores (figure 5).When no transient heat loading was present there was an increase in overall pore area for increasing He + ion concentrations evident by the peak of the normal distribution; this matches well with the early results by Cipiti and Kulcinski [14].High magnification micrographs, taken at the center of the laser irradiated area, for all exposures can be seen in figure 4. Figures 4(a)-(d) represents He + only exposures at increasing heat loads from left-to-right.At 0.6 GW m −2 the early-stage fuzz formation is still observed, but has the appearance of being annealed back into the surface during the irradiation.Though there is some annealing present, there is an appearance of an equilibrium being reached for the early-stage fuzz formation, similarly seen in simulations [55].Note, at this equilibrium stage, the surface reflectivity is still reduced which has allowed for small surface cracks to form on the surface of the sample at 0.6 GW m −2 heat loading.At higher heat loadings the early-stage fuzz formations cease to appear and the helium is likely diffusing out of the sample creating a smoother yet still porous surface.The pores appear to grow in size with increasing heat loading indicating an increase in helium mobility in the sample.These effects are also seen in mixed plasma exposures and discussed more in the following paragraph where we first analyze the pore differences between different ion exposures.
A full pore analysis can be seen in figure 5.The most probable area of pores for each loading, in the absence of transient heat loading, was ∼400 nm 2 for He + only ion exposure, ∼325 nm 2 for 50He|50D exposure, ∼200 nm 2 for 10He|90D exposure, and no pores were seen for D + only exposure.The quantity of pores changed significantly over the different ion exposures.He + only exposures suffered from the most pores and highest area with no applied heat loading.Samples exposed to 50He|50D mixed plasmas had the fewest number of pores, but the pores were on average larger than that of the 10He|90D sample which is a result of reduced helium concentration and mobility over the deuterium.With increasing heat loads all samples exposed to the various ion loadings had pores that increased in size and decreased in quantity with the exception of 50He|50D loading where fewer pores were observed in the non-loaded case than the 0.6 GW m −2 case.This may be a result of the sampling area observed.For all 10He|90D samples the pores were smaller in average size and larger in quantity of that compared to 50He|50D up to 1.5 GW m −2 where the pores were very similar over all ion beam loadings.One reason for the pores being similar is a result of extreme melting and reintegration of the surface material back into the bulk during the 1 h irradiation.Melting was more clear in SEM micrographs of He + only loadings at higher heat loading due to melting and solidification seen as folds in figure 3(d).Such effects can be understood from Marangoni effect [56,57],when a negative surface tension is observed which pools the melting horizontally in the fold like pattern.With the SEM micrographs in figure 4 and the pore diagrams in figures 5, 10He|90D ratio (fusion reactor like environment) may be more favorable due to signs of delayed severe melting at low heat loads.Grain boundaries did not play a large role in trapping helium in this study, evident by the He + only experiments in figures 4(a)-(d) not having larger pores at the grain boundary interfaces.This is similar to previous studies where no heat load was applied, but under similar base temperatures  and He + ion energy [36].The micrographs in figures 4(e)-(l), mixed plasma experiments, also indicate the largest pores appear on the grain boundaries which decrease the mobility of atoms and can act as traps [34,35,58].The applied heat loading also aided in the mobility through the bulk.The addition of D was found to play a significant role on the overall mobility of helium in this study.
Optical profilometry results further confirm the above conclusions for cracking appearance (figure 6).The He + only and 10He|90D loading measurements are shown for all loading conditions used during these studies, starting from no loading to 1.5 GW m −2 where surface roughness increases with heat loading.With no heat loading the samples were close to that of pristine, polished and unirradiated, before heat loading (0.02 µm), because surface features were on nanometer scale, while resolution of the optical profilometer was micrometer scale to cover the image of the entire laser irradiated area.Note, He + only ion loading has distinguishable cracking at 0.6 GW m −2 due to the increase in surface roughness and was clearly visible in figure 3(a).No cracking was visible on the surface of the 50He|50D or 10He|90D at 0.6 GW m −2 which was concluded with a similar roughness to that of no loading.At larger heat loading the 10He|90D sample surface roughness surpasses the He + only loading indicating that larger cracks formed raising the surface with possibly more melting and erosion.Optical profilometry on 10He|90D samples with ELM-like heat loading of 1.2 GW m −2 and both mixed plasmas with heat loading of 1.5 GW m −2 were rougher than all He + only ion irradiations.This can be explained by the fact that the large-scale crack networks as aforementioned are the leading cause to the broad surface roughness.From these findings we can also conclude that an increase in the D + ion loading led to larger crack networks that would easily allow for deep diffusion of D [59].This phenomena is likely playing a large role in increased erosion for increased D + ion concentration ratio in the plasma.As previously mentioned for He + only exposures the early-stage fuzz formations on the surface allow for increased laser power absorption leading to cracking at lower heat loadings.Erosion is investigated more thoroughly in the following sections with indications that when more D + (in concentration) is present there is significantly more erosion.

Sub-surface morphology effects
Subsurface imaging was conducted by Ga + FIB milling into the surface.He + and D + only and 10He|90D samples exposed to ELM-like heat loading of 1.2 GW m −2 were imaged alongside no loading case for comparison.All micrographs were imaged at 52 • off-normal (figure 7).Samples exposed to ion irradiation only appeared significantly different than those with heat loading.First, we observe that there are no pores present for the deuterium only case which again confirms that the gas is desorbed during the irradiation process at 1223 K.This was also observed in other studies on pure-W even at lower base temperatures [25] and again confirms that the He + ion irradiation causes the porous structures.The He + only exposure (figure 7(a)) gives an enhanced look at the early-stage fuzz formation and shows the affected depth of the helium to be ∼350 nm (angle corrected).The 10He|90D sample is similar to the normal micrograph with an enhanced look at the grain-like structures.The affected depth is comparable to the pore size where no subsurface pores are observable.With the applied heat loading there was a significant reduction in early-stage fuzz formation where it appears that any sign of it has vanished and is similar to that of pure W [18].In conjunction with the erosion results (discussed in the following section 3.3) the He + only ion loading with added heat loading suppresses/anneals the surface such that signs of earlystage fuzz growth do not appear [55].Furthermore, the pores are similar in size for heat loaded areas of He + ion irradiated samples where larger pores again are seen near the grain boundaries because the grain boundaries aid in absorption, as well as impede the mobility of bubbles.
Though the fluence was double for dual beam irradiations over that of single beam, the sub-surface micrographs and erosion studies strongly suggest for the role of D + in surface erosion either by sputtering [60,61] or by populating near the surface allowing easier erosion under the heat loading [19,62].

Comprehensive erosion study
Witness plates positioned at 45 • to the W-3Ta sample surfaces during all ion exposure experiments were taken from the irradiation chamber and transferred to the UHV XPS facility.Mo was used because it is not likely to sputter at the ion energies investigated, has a high melting temperature, is extremely ductile at room temperature, and has only minor peak overlaps.Survey scans were taken for all the cases and in addition to obvious Mo, C, and O XPS peaks, W and Ta peaks were observed (figure 8). Figure 8 clearly indicates peaks associated with ambient conditions, the Mo witness plate, and the W-3Ta sample by analyzing the W-3Ta sample itself and the most erosion captured on the Mo witness sample, the 10He|90D 1.5 GW m −2 heat loading.Something that is observed is the large size comparison between W and Ta in the two samples.The W peaks on the Mo witness are much lower while the Ta peaks are approximately the same which may indicate that more Ta is eroded than just 3 Wt.% ratio compared to W. It is extremely difficult to quantify the results due to peak overlaps (with Mo), thus the data is analyzed qualitatively.High resolution XPS spectra over the W4f and Ta4f and Mo4p regions were obtained to compare the presence of W and Ta for all experiments.We identified the XPS peaks of WO 3 4 f 5/2 (∼37.7 eV), WO 3 4 f 7/2 (∼36 eV), W4f 5/2 (∼34.5 eV), W4f 7/2  (∼31.8 eV), Ta 2 O 5 4 5/2 (28 eV), Ta 2 O 5 4 7/2 (25.2 eV), and the broad Mo4p peak (∼40-34.5 eV) [63].Since the W loss peak was shadowed by Mo4p and not useful for discussions, it is not presented here.All highly resolved XPS data presented here was peak normalized for better comparison of XPS spectra from different samples.The observed XPS peaks of W and Ta oxides are obvious due to ambient exposure of samples during the transfer process from the irradiation chamber to the XPS chamber.Any pure metal (primarily W) peaks observed were very likely eroded from the sample during irradiation in the vacuum environment and then preserved via a native oxide layer growth on the surface during sample transfer transport.XPS measurements show oxygen saturation, evident by similar peak magnitudes even with increased heat loading.This is not surprising as XPS is a surface analysis technique with sampling depths on the order of a few nanometers.
As previously discussed and shown via optical profilometry data (figure 6) and subsurface micrographs (figure 7) there was significant erosion with added D + ion concentration.This is also portrayed in XPS measurements where figure 9 shows qualitative results for different ion loadings with 1.2 GW m −2 transient heat loading.Although the single ion beam irradiations were at half the flux of dual beam ion irradiations it is clear that He + only ion loading produced very little erosion and the Mo4p peak and a small, not above background, W4f 7/2 peak is visible.With D + only loading there is clear W erosion in 3 out of the 4 observable W and WO 3 peaks.The presence of TaO 5 peaks are clearly visible on all loadings, but close to the background for He + only further suggesting D + to play a large role in erosion.In the top of figure 10, one can see that the magnitude of TaO 5 peaks in W-3Ta samples exposed to He + only exposure are similar to that of the spectrum gathered on pure-W sample (purple) and thus minimal on the order of background.This is not the case for the 10He|90D case where pure-W showed no signs of Ta peaks, as expected while W-3Ta had significant amounts of TaO 5 .Comparisons with W are explored more in the next section, but are mentioned here to better understand the W-3Ta results.The increased erosion with exposure to D + only ions was not surprising as this has been seen in many of our other studies on pure W and W-5Ta samples exposed to similar conditions [18,19].When samples were exposed to dual ion beam loading, the peaks were smoother suggesting more counts over background and thus more erosion.
The use of different transient heat loadings were also observed via XPS for all ion loadings (figure 10).It is clear that increased heat loading enhances erosion.Also, the appearance of oxygen saturation (formation of corresponding material oxides) with regard to XPS sampling depth is seen.This was most apparent in 10He|90D samples where all XPS spectrums were very similar even with drastic heat loading conditions.Samples exposed to heat loads of 0.76 GW m −2 have previously been completed on W-5Ta alloys where TaO 5 was not observed [18].Our results here indicate that the TaO 5 peaks become observable at lower heat loadings only for 10He|90D case and all other results are similar to previous studies on W-5Ta, due to smaller amounts of erosion and not sample saturated.Overall the erosion for 50He|50D is similar to that for pure W in our previous study, in the sense that saturation occurred to the sampling depth of XPS with increased heat loadings.

Comparison to pure-W
The W-3Ta results shown in these studies are well substantiated and aligned with our previous investigations using pure-W exposed to similar ion beam parameters and ELM-like heat loading [19].In both studies the pore amount and an enhancement in size was observed with increasing ELM-like heat loading.We present further confirmation the role of D has in this study by increasing the concentration to 90% which follows the hypothesized trend in our previous studies.Both experiments show a clear dependence on ion loading concentrations to erosion, with mixed plasmas suffering from higher erosion.Enhanced erosion was obvious with increasing heat loading until what appeared to be a plateau that was reached at heat fluxes above ∼1.2GW m −2 .The XPS peaks seemed saturated at higher energies, which is likely caused by XPS sampling depth as previously mentioned.No clear transient temperature dependence to XPS peaks was present in pure W experiments, but was clearly visible for the results shown here.This could be due to longer transfer times in the previous experiment on pure W or improved laser parameters previously mentioned for these experiments.Subsurface imaging showed a decrease in local surface roughness and this phenomena was confirmed by optical profilometry studies on pure W and for W-3Ta for the lower heat loading tests.The surface roughness of W-3Ta samples were significantly lower than that of the pure W samples between 0.6 and 0.9 GW m −2 .
In this study four tests were completed on pure-W samples to aid in comparison because a larger laser spot size was used in this study.SEM micrographs for He + only and 10He|90D at the most severe heat loading of 1.5 GW m −2 are presented in figure 11.The ion only loading micrographs are inset in the Figure and labeled figures 11(a.3) and (b.3).Under the conditions in this study there is a significant decrease in early-stage fuzz formation for pure-W compared to the W-3Ta.It has been previously reported that the addition of Ta in W alloys delays the effect of low energy He + ion irradiations, but under these conditions this phenomenon was not observed [44].Mixed ion beam loading yielded similar results to that of the W-3Ta material, where the pores decreased in size and quantity with the addition of D + .This has been observed and is consistent with previous studies on pure-W where the D + ions are competing against the He + ions.
There is an obvious difference between the laser irradiated area for pure-W and W-3Ta alloys.Pure-W, even at the maximum heat loading studied here, showed no signs of having large crack networks like those seen even at low heat loading in W-3Ta.This indicates that the W-3Ta samples have decreased mechanical and thermal properties compared to pure-W which has been noted in other studies [43].Another observation is that the loading with D + had an increase in surface damage, evident by the SEM micrographs, similar to that found in the W-3Ta study, but this effect may be driven by the fact that the exposed flux was double and will need to be studied further for pure-W.High magnification micrographs show small cracks (∼10 µm) form on the surface during dual beam irradiations and heat loading, while the He + only irradiations with heat loading yielded the shale-like morphology seen in [19].Mo witness plates were used in a similar fashion to observe any material erosion during the pure-W tests, the results are shown in purple in figure 10.The He + only test on pure-W at 1.5 GW m −2 resulted in two large WO 3 XPS peaks while pure-W peaks were not observed at all.This is an indication that the W oxidized completely during sample transfer, only the WO 3 surface layer was eroded during the irradiation, or a combination of both.This phenomenon was also observed in previous experiments where an approximately 1 mm laser spot size was used [19].Larger WO 3 peaks than pure-W peaks were also observed for 10He|90D tests on pure-W at a heat loading of 1.5 GW m −2 , further suggesting less severe erosion for pure-W samples exposed to the same conditions.
The severe cracking observed in W-3Ta alloys, alongside XPS results, suggests more erosion of the W-3Ta samples because of its reduced thermal properties, ∼10% reduction in thermal diffusivity estimated for high temperatures [64].Recent experiments show varying results for the mechanical properties of W-Ta alloys depending on the manufacturing process.The samples in this study were prepared the same as our previous study [18], which showed that W-5Ta had a slightly reduced hardness over pure W. In another study the Charpy impact energy for W-3Ta is approximately the same for pure W at 1223 K with pure W being very slightly reduced [64].At this time it is not clear which material properties, if not both, are contributing to the enhanced erosion of W-Ta alloys as it depends heavily on the material manufacturing.Future experiments need to be conducted on pure W with improved laser parameters with in situ erosion measurements by using a quadrupole mass spectrometer for example, but as of now, there is an appearance of more erosion for W-3Ta tests.

Conclusions
This study focused on the effects of different ion loading of He + and D + on W-3Ta alloy exposed to ELM-like pulsed heating.When He + only ion loading interacted with the surface there were signs of early stage fuzz growth similar to that in previous studies.The addition of D + to the He + irradiation showed suppressed early-stage fuzz growth alongside the appearance of recrystallization.Surface pores decreased in size with dual plasma exposures due to the rapid desorption and saturation of D + .Samples exposed to 10He|90D had an observed increase in pore quantity over 50He|50D irradiations and is likely a result of decreased helium mobility.The most probable pore size decreased with increasing concentration of D + , where deuterium limits the mobility for 50He|50D, but also the decrease in concentration of He + in the 10He|90D directly influences the size of pores.No pores were observed for D + only exposure, most likely due to either rapid D + desorption or lower trapping sites due to the absence of He + .Subsurface imaging agrees well with the pore analysis taken using the surface normal micrographs.Surface pores grew in size and decreased in quantity with increase in transient heat loading.Larger pores were found on the grain boundaries of D + irradiated samples, agreeing to the fact that the grain boundaries are both a trap for helium and deuterium migration, but also a sink for all ion species to integrate into the bulk.For He + only exposures, while we noticed enhancement in the pore size, no appearance of grain boundary favoritism similar to previous studies and most likely due to low bulk mobility or the early-stage fuzz formations 'hiding' the grain boundaries.
The micrographs in conjunction with erosion studies suggest that the transient heat loading on the He + only exposure annealed the surface retarding early-fuzz growth and possibly saturating at a much lower depth.Cracking was observed earlier on the He + only exposure with 0.6 GW m −2 most likely due to the slightly reduced optical reflectivity which increased laser absorption.Such cracking was also observed during optical profilometry measurements.Comparison of all ion loadings at 1.2 GW m −2 clearly indicates enhanced erosion due to the addition of D + ion concentrations.The most severe erosion was witnessed during 10He|90D exposures, indicating that reactor conditions would be the most harmful for W-3Ta based PFMs.In this study we show enough erosion that the witness plates appear to saturate with respect to the sampling depth of XPS.Furthermore enough erosion occurred to lead to a measurable amount of TaO 5 erosion which did not appear in our previous studies.Though W-Ta alloys have shown great abilities in fuzz suppression, this study suggests that it is highly dependent on irradiation parameters where the evidence of early-fuzz formation was more prevalent in W-3Ta than pure-W samples.Large (>10 µm) surface cracking was only observed in W-3Ta alloys while pure-W had some small cracks for 10He|90D tests at 1.5 GW m −2 .In both materials there appeared to be significantly more damage with increasing D + concentrations.This was evident in W-3Ta samples where the flux was the same for both mixed plasma experiments and the cracks were larger for the 10He|90D cases.This may not be so evident with the pure-W samples because the flux was double for the mixed plasma experiments over the He + only cases, but nevertheless there was an appearance of more damage for the mixed plasma case.Further analysis of pure W and W-Ta alloys are needed, especially the need for in situ erosion detection, but at this time W-Ta alloys used in this study showed signs of enhanced erosion over pure-W.

Figure 1 .
Figure 1.Ultra high flux irradiation (UHFI-II) chamber at CMUXE used to simulate fusion ELM environment using two end hall ion sources for D + and He + ion irradiations, Nd:YAG for transient heat loading, and a button heater for steady state heat loading.Inset (lower right) shows the geometry of the heater surface used in each experiment.

Figure 3 .
Figure 3. Low magnification SEM micrographs of laser spot size where large cracks due to applied heat loading are seen.Various ion loadings are represented on the left and the two laser powers are represented at the top for figures (a)-(f ).The bottom two micrographs are taken at the same heat loadings of 1.2 GW m −2where the left is for a laser only exposure (g) and the right is for deuterium ion loading with the applied heat loading (h).Insets shown on micrographs (a), (g), and (h) show higher magnification images from their marked respective regions.

Figure 5 .
Figure 5. Surface pore area analyses for the majority of experiments compiled using ImageJ.Different ion loading concentrations of; He + only, 50% He + and 50% D + , 10% He + and 90% D + , from left to right.No heat loading at the bottom for each with increasing applied heat loads from 0.6, 0.9, 1.2, and 1.5 GW m −2 .Blue curves indicate a normal distribution fit.

Figure 6 .
Figure 6.Mean surface rougness for all W-3Ta samples exposed to various ion loadings as a result to different ELM-like transient heat loadings.Specific values for 100% He and 10% He exposures are shown in their respective colors for clarity.

Figure 7 .
Figure 7. Subsurface micrographs of W-3Ta samples produced by focused ion beam, images taken at 52 • angle to surface normal.Ion irradiation loading only on top and pulsed heat loading (1.2 GW m −2 ) included on the bottom with He + only (left), 10% He + and 90% D + (middle), D + only (right).

Figure 8 .
Figure 8. Survey scans taken for pre-exposed W-3Ta and for a Mo witness plate when W-3Ta was exposed to He:D 10:90 at a heat loading of 1.5 GW m −2 .

Figure 9 .
Figure 9. XPS of Mo witness plates for W-3Ta samples exposed to every ion loading and with ELM-like heat loading of 1.2 GW m −2 .

Figure 10 .
Figure 10.XPS of Mo witness plates for the majority of experiments conducted in this study and normalized.Different ion loading concentrations of; He + only, 50% He + and 50% D + , 10% He + and 90% D + , from left to right.All ELM-like heating with increasing applied heat loads from 0.6, 0.9, 1.2, and 1.5 GW m −2 .Purple lines indicate measurements on pure-W samples.

Figure 11 .
Figure 11.Heat loading of 1.5 GW m −2 for two different ion loadings (a) He + loading (b) He:D | 10:90 loading.(1) & (2) micrographs are different magnifications near the center of the laser heat loading, while (3) is a micrograph of the respective ion beam effects without the heat loading.

Table 1 .
Details of ion irradiations with transient heat loadings.Tests completed at a steady state temperature of 1223 K. Sample IDs colored in purple indicate pure-W samples irradiated at identical conditions alongside the W-3Ta samples.