Abbreviations

bcc Body-centered cubic CLAM China low activation martensitic steel CVD Chemical vapor deposition fcc Face-centered cubic HFR High flux reactor HIP Hot isostatically pressed ITER International Thermonuclear Experimental Reactor

PCA Prime candidate alloy

PRF Permeation reduction factor

RAFM Reduced activation ferritic/martensitic steel

4.16.1 Introduction

As fusion energy research progresses over the next several decades, and ignition and energy production

are attempted, the fuel for fusion reactors will be a combination of deuterium and tritium. From a safety point of view, these are not the ideal materials. The reaction of deuterium with tritium produces a-particles and 14.1 MeV neutrons. These neutrons are used not only to breed the tritium fuel, but also interact with other materials, making some of them radioactive. Although the decay of tritium produces only a low — energy p-radiation, it is difficult to contain tritium. Additionally, being an isotope of hydrogen, tritium can become part of the hydrocarbons that compose our bodies.

From the tritium point of view, the fusion facility can be divided into three components: the inner vessel area where the plasma is formed, the blanket where tritium production occurs, and the tritium exhaust and reprocessing system. There is the potential for tritium release in all the three sections of the facility. The tritium cycle for a fusion reactor begins in the blanket region. It is here that the tritium is produced by the interaction of neutrons with lithium. Specifically, the reaction is given symbolically as 6Li(n, a)3H. A neutron that has been thermalized, or lowered in energy by interaction with surrounding materials, is absorbed by 6Li to produce both an a-particle (helium nucleus) and a triton. Elemental lithium contains ^7.5% 6Li. As a breeder material in a fusion plant, lithium is enriched in the 6Li isotope to various degrees, depend­ing on the particular blanket design. The 7Li isotope also produces a small amount of tritium via the 7Li (n, a)3H + n reaction. The cross-section for this endo­thermic reaction is much smaller than that for the 6Li reaction. Upon release from the lithium breeder, the tritium is separated from other elements and other hydrogen isotopes. It is then injected as a gas or frozen pellet into the torus, where it becomes part of the plasma. A fraction of the tritium fuses with deuterium as part of the fusion process, or it is swept out of the chamber by the pumping system. If tritium is removed from the torus by the pumping system and sent to the reprocessing system, it is again filtered to separate other elements and other isotopes of hydrogen. All through the different steps, there is the potential for permeation of the tritium through the materials containing it and for its release to the environment. The probability of this occurring depends on the location in the tritium cycle. This chapter describes hydrogen permeability through two categories of materials that will be used in fusion reactors: candidate plasma-facing and structural materials.

The plasma-facing materials in future fusion devices will be heated by high-energy neutrons, by direct interaction of the plasma particles, and by electromagnetic energy released from the plasma. These plasma-facing materials must be cooled. It is primarily through the cooling tubes passing through the plasma-facing materials that tritium losses can occur in the primary vacuum vessel. The three mate­rials typically used for plasma-facing applications are carbon, tungsten, and beryllium. In this report, we describe the behavior of these materials as plasma­facing materials and how tritium can be lost to the cooling system.

The term ‘structural material’ is used here to describe materials that serve as the vacuum boundary in the main chamber, as the containment boundary for the blanket region, and as the piping for cooling and vacuum lines. These materials can be ferritic and austenitic steels, vanadium alloys, and zirconium alloys, as well as aluminum alloys in some locations, or potentially ceramics. We give a complete list of the different types of structural materials and review their tritium permeation characteristics.

Materials with a low permeability for tritium are being considered as barriers to prevent the loss of tritium from fusion plants. There are a few metals with relatively small values of permeability, but as a whole, metals themselves are not good barriers to the transport of tritium. Ceramics, on the other hand, are typically very good barriers if they are not porous. In most cases, the low permeation is due to extremely low solubility of hydrogen isotopes in ceramic mate­rials. Bulk ceramics, such as silicon carbide, may one day be used as tritium permeation barriers, but most of the current barrier development is for coatings of oxides, nitrides, or carbides of the metals themselves. We show in this review that many such oxides and nitrides may exhibit extremely good permeation behavior in the laboratory, but their performance as a barrier is significantly compromised when used in a radiation environment. We review the permeation parameters of materials being considered for barriers.

This report begins with a review of the processes that control the uptake and transport of hydrogen isotopes through materials. The parameters used to define these processes include diffusivity, solu­bility, permeability, trapping characteristics, and recombination-rate coefficients. We examine the transport of hydrogen isotopes in plasma-facing mate­rials, discuss the conditions that exist in the main torus, and look at the ways in which tritium can be lost there. Next, we consider the tritium transport properties of structural materials, followed by the transport properties in barrier materials, including oxides, nitrides, and carbides of structural metals, as well as low-permeation metals. The application of tritium barriers is discussed in some detail: both the theoretical performance of barriers and their observed performance in radiation environments, as well as an example of tritium permeation in the blanket of a fusion reactor. We conclude by summar­izing the tritium permeation properties of all the materials, providing the necessary parameters to help designers of fusion reactors to predict tritium losses during operation.

The base temperature influences the thermal shock behavior in two ways. First, a higher base tempera­ture influences the damage, cracking, and melting threshold. All of them are essential because they limit the operational conditions and when exceeded cause enhanced material degradation. Therefore, life­time estimates based on RT data will yield unrealistic conclusions.

Second, crack formation strongly depends on the plastic deformation at high temperatures and even more on the stress developed during cool down. To understand the influence of a higher base temperature, one has to be aware of the typical shape of the yield and tensile strength curve for W or a W alloy.105,157 While the decrease in strength is rather high at low temperatures, the curve flattens at high temperatures despite a drop in strength when exceed­ing the recrystallization temperature. As a result, the high temperature plastic deformation induced by the combination of a heated surface and ‘cool’ base material can be significantly reduced by a small increase in base temperature. Combining this effect with the increased ductility of W at the given base temperature, brittle crack formation can be avoided when heating

the material above a certain threshold.82’157’164’165 This temperature threshold is related to the DBTT but is not necessarily identical to it.

4.17.4.1.2 Repetition rate

In addition to the parameters mentioned above, the damage, cracking, and melting thresholds are deter­mined by the number of load repetitions, because of continuing material degradation such as hardening and recrystallization. This is of particular interest for short transient events with a high repetition rate in magnetic (ELMs) and inertial fusion devices. Up to now the simulation of submillisecond events (ELMs, IFE) has been performed only up to a rela­tively low number of cycles; large numbers of pulses (e. g., >106 ELM pulses during the life-time of the ITER divertor) are not feasible in the majority of the above-mentioned test facilities.

Tritium retention and transport is a critical phenom­enon for graphite in fusion systems in general, and it is the subject of a chapter by Causey, San Marchi, and Karnesky in this series. In the previous section of this

Mean ambient thermal conductivity (W m-1 K-1)

chapter, the interaction of the plasma particle flux with the surface of graphite was discussed. However, the fate of the implanted particles, most impor­tantly deuterium and tritium, following their impact with the graphite surface is also an important issue and is seen by some as the major impediment to the use of graphite as a PFM.85 Quantification ofthe problem and determination ofpossible mitigat­ing steps is complicated by experimental data, which can vary by orders of magnitude,86-92 as reviewed by Wilson.93

The primary concern over retention of fuel in the PFC is the inventory of hydrogen adsorbed into the graphite and the subsequent release of near-surface hydrogen (due to physical or chemical sputtering, etc.) as plasma discharge begins. The hydrogen sput­tered from the wall oversupplies the plasma edge with fuel, causing instabilities and making plasma control problematic. Tritium inventory concerns are generally safety-related but can have significant economic consequences because of the high cost of tritium. Tritium release to the environment in an accident situation had limited the allowed inventory in TFTR, and was a significant consideration for the sighting of the ITER. It has been estimated84 that as much as ~1.5 kg of tritium would reside in the graph­ite PFM of ITER, corresponding to an additional fuel cost of 1.5-3 million dollars.

A source of trapped hydrogen, not discussed in detail here, which may dominate the tritium inventory in ITER-like machines, is the ‘codeposited layer.’94 This layer is formed by the simultaneous deposition of carbon, which is eroded from the first wall, and hydrogen. Thick layers of carbon redepos­ited to low erosion areas are common, and have been seen in all large tokamaks utilizing graphite PFMs. As this layer grows, the hydrogen contained therein cannot be liberated by surface sputtering and becomes permanently trapped. This problem is unique to graphite and will require continual sur­face conditioning to minimize the total inventory of trapped species. It represents a vast sink for tritium and therefore must be managed in some way. Below is a discussion of the retention of tritium in bulk graphite.

The physical process involved in the retention of hydrogen, as it corresponds to graphite PFMs, is fairly well understood. The energetic hydrogen iso­topes are implanted to depths of less than a micron in the PFM surface. Once implanted, the hydrogen ions are either trapped, reemitted, or diffuse through the bulk. At temperatures less than 100 °C,95,96
the majority of ions are trapped near the end of their range. These trapped ions are not in solution in the graphite, but are held97 in the highly defected struc­ture. The amount of hydrogen isotope that can be accommodated is largely dependent on the implan­tation temperature,96,98 the trap types and densities (defects), and, to a lesser extent, by the implantation depth.9 , One model for bulk hydrogen trapping presented by Atsumi101 is shown in Figure 28. In his work, two distinct traps have been identified. The lower energy trap (2.6 eV) is associated with edge planes of the graphite crystals, with total trapping therefore depending on the effective size and accessi­bility of the crystals to diffusing hydrogen species. The second, higher energy trap (4.4 eV), is associated with dangling bonds, albeit at the edge of an interstitial loop. As the formation of small interstitial loops is one of the primary effects of neutron irradiation, the formation of these deeper traps is directly affected by neutron fluence. The total retained isotopic H can reach as much as 0.4—0.5 H/C in the implanted layer

95,100,102

at room temperature.

As the amount of implanted hydrogen increases toward its saturation value, a larger fraction of ions are released from the graphite surface. At intermedi­ate and high temperatures (>250 °C), diffusion of hydrogen in the graphite lattice occurs. This diffu­sion is most likely along internal surfaces such as micropores and microcracks, while transgranular dif­fusion has been seen above 750 °C.103,104 This bulk diffusion, along with the associated trapping of hydrogen at defect sites, has been studied widely with quite variable results. This variation can be seen in Figure 29, where the temperature depen­dence of the hydrogen diffusion coefficient for sev­eral carbon and graphite materials is shown.

It is expected that the diffusion of hydrogen through graphite would be highly dependent on the graphite microstructure, which may explain the wide range of the data of Figure 29. In any event, the transport of hydrogen through the bulk graphite and associated solubility limits can significantly increase the hydrogen inventory for fusion devices. The effect of the perfection of graphitic structure on the solu­bility of hydrogen is shown by Atsumi’s data105 in Figure 30. The data in Figure 30 indicate that the more defect-free, highly graphitized materials have a lower solubility limit. Further evidence for the role of structural perfection comes from the observation that materials that have been disordered by neutron irradiation have significantly higher solu­bility for hydrogen.105,106

The effect of atomic displacements on the hydro­gen retention of graphite was first shown by Wampler using 6 MeV ion beams.107 Wampler used four types of intermediate and high quality graphites, which were irradiated with a high-energy carbon beam at room temperature, and then exposed to deuterium gas. Wampler’s results indicated that the residual deuterium concentration increased by more than a factor of 30-600 appm for displacement doses appropriate to ITER. However, for reasons that are not entirely clear yet, neutron-irradiated high-quality CFCs retain significantly less tritium than expected

Absorption (Rate-determining step) (>90%)

Desorption О О Detrapping

Figure 28 Schematic of the processing of hydrogen trapping and diffusion in graphite. Reproduced from Atsumi, H. J. Nucl. Mater. 2003, 313-316, 543-547.

from the earlier work. This was reported by Atsumi105 and clearly shown in the work of Causey108 (Figure 31). Causey irradiated high thermal conduc­tivity MFC-1 unidirectional composite and FMI-222

4.0	on n
IG-430U
| |Trap 2 (crystallite edge surface) I I Trap 1 (interstitial cluster loop)
1.9 X 1024 5.4 X 1024 0.23dpa 0.65dpa
Figure 32 Loading of defects in irradiated graphite with hydrogen. Reproduced from Atsumi, H.; Muhaimin, A.; Tanabe, T.; Shikama, T. J. Nucl. Mater. 2009, 386-388, 379-382.

magnitude less than that expected from the earlier work on GraohNOL-N3M.109

In some more recent work by Atsumi,110 different neutron-irradiated graphites were irradiated and the amount of hydrogen that could be entrained in the crystal was measured. Specifically, Atsumi110 irra­diated three isomolded grades of graphite (IG-110, IG-430U, and IG-880U) in the JMTR reactor below 200 °C to various fluences and then baked the sam­ples in an atmosphere of hydrogen. Figure 32 shows the relative abundance of hydrogen that can be loaded into the crystallite edge and interstitial cluster loop — type defects (Type 1 and Type 2 defects of Figure 28) of the IG-430U graphite. Clearly, both defects are produced during irradiation and are accessible to postirradiation loading of hydrogen. In the same work, Atsumi carried out a series of preloading anneal­ing of samples, which suggested that the edge-type defects would be preferentially annihilated. In this context, it is important to note that all work on hydrogen or tritium retention in irradiated graphite has followed the approach of irradiating the material at a relatively low temperature and then loading and unloading and measuring the released hydrogen from the sample at a comparatively higher tem­perature. This may be of significance in that, as inferred in the work of Atsumi and others,105,106,108 the relative crystalline perfection (amount of intrin­sic defects) is strongly related to hydrogen retention. As discussed in Section 4.18.3 and Chapter 4.10, Radiation Effects in Graphite, irradiation at low temperature may result in a significantly different microstructure (an abundance of simple interstitial and vacancy clusters) as compared to the more fusion
relevant irradiation temperatures (formation of more perfect interstitial discs and collapsing of vacancy complexes). Moreover, the postirradiation annealing of a low-temperature-irradiated microstructure will likely not produce representative microstructures of irradiation at more relevant higher temperatures. For this reason, data generated to date should be consid­ered as a guide for the trends likely to occur rather than as quantitative information on the actual tritium retention that will occur in fusion devices. Moreover, they are likely overly conservative.

The erosion mechanisms that affect the erosion/ damage of the first wall in ITER are (1) sputtering erosion by D-T ions and charge-exchange neutrals

during normal operation; and (2) evaporation and loss of melt layers during off-normal transient events such as thermal quench disruptions, ELMs, VDEs, and runaway electrons impact. There are additional localized erosion phenomena such as arcing, over­heating with evaporation, and, possibly, loss of melt layer on exposed edges, but it is very difficult to make predictions of these effects for ITER. Special design attention has been given to avoid the misalignments of PFCs and avoid thermal overloading with possible localized damage.

4.19.6.2.2.1 Erosion of Be wall during normal operation

Calculations have been done to compute erosion of the first wall (due to fuel charge-exchange neutrals and ions, and impurity ions).205,206 It was found that about 20-40 g of Be per 400 s discharge are eroded from the wall with a beryllium peak erosion rate of the order of 0.1 nms~ . These predictions are con­firmed by extrapolation of experimental data from JET.1 0 This erosion rate would be acceptable from a component lifetime standpoint, especially during the low duty-factor operation of ITER. However, the total amount of eroded material may be significant. This material will most likely go to the divertor, and this will affect the composition of the divertor sur­face; therefore, it will affect the divertor performance
and contribute to tritium codeposition and dust inventories. Modeling of the influx of the eroded beryllium on the divertor is in progress to extrapolate from present machines and, in particular, to account for effects arising from material mixing including codeposition as expected in ITER. Several studies have been recently published on this subject (see, e. g., Kirschner et a/.207,208).

In recent years, FCIs made of ceramic materials such as silicon carbide composite (SiC/SiC) have been proposed for both electrical and thermal insulation between the liquid breeder and the channel walls. Although the electrical resistivity of SiC/SiC is lower than that of the candidate insulator coating materials, the use of an insert that is much thicker than the coating allows for sufficient reduction of induced electrical currents.

An FCI is attractive for application to dual­coolant Li-Pb blanket concepts in which heat is removed by both a high flow rate of He and Li-Pb.41,42 This concept, however, may not be applied to self-cooled liquid metal blankets because these blankets need to avoid thermal insulation between the coolant and the first wall or the blanket structural components and coolants.

The most effective structural condition assessment programs are those that focus on the components most important to safety and at risk due to environ­mental stressor effects. Aging assessment methodolo­gies have been developed to provide a logical basis for identifying the critical concrete structural ele­ments and degradation factors that can potentially impact the performance of these structures.83

An evaluation of the impact on plant risk due to structural aging can also be used in the selection of structural components for evaluation.84 Probabilistic risk assessments conducted to date indicate that the structural systems generally play a passive role in miti­gating design basis (or larger) internal initiating events: a notable exception being the pressure-retaining func­tion of the containment following a degraded core incident involving failure of the reactor pressure vessel. The structural components play an essential role in mitigating extreme events initiated by earthquake, wind, and other extreme influences, and their failure probabilities due to external events can be higher. Moreover, failure of major structural components may impact the operation of a number of mechanical and electrical systems and lead to so-called common cause failures. Thus, deterioration of structural com­ponents and systems due to aging and other aggressive environmental influences may be more serious in terms of overall plant risk than might be evident from a cursory examination of their role in accident mitiga­tion. The significance of structural aging and deterio­ration to plant risk can be evaluated by considering the impact that they have on risk associated with external initiating events, especially earthquakes. It is in miti­gating the effects of strong ground motion due to earthquakes that structural systems play a particularly significant role. The apparent impact of structural aging can be investigated using a margins analysis to assess suitability for continued service. Sensitivity analysis can help to identify the structures of impor­tance that should warrant particular attention.

A third approach involves the combination of finite-element analysis and nondestructive testing methods for evaluation of aging and degradation of concrete containments.85 The CONMOD Project objective was to find a practical means to determine the condition of a containment structure as well as how this condition can be expected to change with time under the influence of various loading condi­tions, including aging. Applications of the approach developed include (1) identification of the critical parts of a structure for nondestructive evaluation including critical parameters, (2) updated structural analyses using input from nondestructive evaluations, (3) prediction of nondestructive responses for a known condition at a given time using finite-element method modeling techniques, and (4) prediction of the nondestructive evaluation responses using finite — element modeling techniques based on a known con­dition and how this will change because of aging processes. One of the conclusions of this study was that development ofnew containment designs should focus on establishing rules, designs, and novel ideas on how to significantly improve the accessibility of the concrete structures for diagnostic investigations.

Table 2 summarizes lithium compounds considered as breeding material and lists some key properties.

Lithium oxide was favored in early blanket concepts, in particular because of its very high lithium density and good thermal conductivity. Its biggest disadvantage is its strong sensitivity to moisture and the evidence it shows of significant swelling under irradiation.

The silicates studied most widely are Li2SiO3 and Li4SiO4. In practice, traces of Li2SiO5 or Li6Si2O7 can be found as well.

A variety of lithium zirconates have also been stud­ied, such as metazirconate Li2ZrO3 and Li8ZrO6 octa — zirconate by Roux and coworkers.26,33,35,52,53,62,63,156 Because of activation issues for Zr in the fusion blanket spectra, these compounds became less attractive. Zirconates have been shaped in pellets as well as pebbles, and their irradiation performance at high burnups was considered promising in terms of pellet integrity and tritium release characteristics.33

Lithium titanates were initially introduced at AECL by Kopasz and coworkers,64’65 and early tri­tium release experiments have shown promising results. Li2TiO3 has been studied by Japanese and EU parties. Hoshino et a/66-70,179-181,190 found that not only oxygen-deficient but also lithium oxide-deficient defects form on changing the atmo­sphere from hydrogen to oxygen. Thus, the doubly nonstoichiometric composition, Li2-xTiO3-j, has been confirmed. Further, it has been shown by ther­mal diffusivity measurement that 95Li2TiO3 has a higher thermal conductivity than 100Li2TiO3.66

Lithium aluminates have been studied mostly in the form of g-LiAlO2. Data accumulated have been summarized by Billone and Grayhack.71 Due to its low lithium density and modest tritium release per­formance, the material has gradually begun to receive less consideration.62

More recently, Zhu et a/.56 in China started inves­tigations of lithium tantalates such as Li3TaO4, which has a reasonably high lithium density.

Sedano and coworkers57 report on the develop­ment of lithium plumbates such as Li8PbO6, revisit­ing earlier work of Hayashi et a/.72

Body-centered cubic (bcc)-structured V-Cr-Ti alloys (particularly composition ranges of around V-4Cr-4Ti and V-15Cr-5Ti) have low neutron cross-sections and the isotopes that do form with neutron capture have short half lives (51V has a half-life of <4 min). As noted in Chapter 4.12, Vanadium for Nuclear Systems, these characteris­tics, along with reasonable operating temperatures (limited by radiation hardening and helium bubble formation to «575-775 K129), make V-Cr-Ti an attractive material for first walls and blankets, but the tritium retention characteristics of vanadium alloys leave much to be desired. Vanadium has a large solubility for hydrogen and a very large diffu­sion coefficient for hydrogen. These two traits make the permeability of hydrogen in vanadium compara­ble to that of titanium and palladium.1 Vanadium absorbs hydrogen exothermically. Additions of chro­mium tend to increase this energy, while titanium additions tend to decrease it and, to first order, alloys with roughly equal and small amounts of chro­mium and titanium (such as V-4Cr-4Ti) are assumed to react similarly to hydrogen isotopes as pure vana — dium.129 Vanadium alloys form hydrides below «450 K,129 which is below the typical operating temperatures.

The diffusivity of hydrogen in vanadium is «10- m2 s-1 in the range of operating temperatures, a larger value than in most metals.129 There has been extensive experimental measurement of several V-Cr-Ti alloys using different hydrogen isotopes. These are summar­ized in Figure 14. Schaumann et at}31 and Cantelli eta/.132 independently measured the diffusivity of both protium and deuterium in pure vanadium after charg­ing them with gas at «775 Kusing the Gorsky anelas — tic relaxation effect.131 Both groups found that the prefactor did not depend strongly on the isotope, whereas the activation energy did. This is contrary to the common naive expectation, where the activation energies would be identical and the prefactors would differ by a factor of the square root of the mass. The two groups also each reported a deviation from expo­nential behavior at lower temperatures, which could be attributed to some combination of surface effects, trapping, and a V-H phase transition at «200 K.1 However, deviation from Arrhenius behavior is com­mon in bcc metals133 and has been reported in a number of studies of vanadium. The transition tem­perature from exponential behavior varies widely with the technique used to measure diffusivity and the group that measures it,131,132,134-139 and has been as high as 813 K when measuring uncharged specimens using the absorption technique.134 This supports the notion that much of the deviation reported in the literature may be due to surface recombination limita­tion at lower temperatures. Compounding the recom­bination limitation of the vanadium base metal is the fact that surface oxides (particularly TiO) form and limit recombination more.140,141 At lower tempera­tures, there was an increased deviation in the measured Dh/Dd from /2, which has also been observed by others100,138 and in the diffusivity of titanium.142 The heavier hydrogen isotopes do not diffuse much slower than predicted until temperatures below ~-373 K, thus the physics associated with deviations from the predic­tions of classical rate theory cannot be exploited for the use in fusion applications.

The electrochemical pulse experiments of Boes and Zuchner143 derived an activation energy that is twice as large using the electrochemical pulse method, which was also supported by absorption experiments by Eguchi and Morozumi.134 Both tech­niques are influenced by the surface, while Gorsky effect measurements and electrical resistivity mea­surements are only influenced by the bulk.1 The

absorption experiments also indicated that the hydro­gen diffusion coefficient in vanadium alloys is decreased by additions of chromium (as well as iron and niobium), but that it could be increased by large titanium additions, which is thought to be due to electronic contributions.134 Most other experiments have found that moderate amounts of titanium decrease the hydrogen diffusion coefficient much more than chromium is able to.133,139 Ti’s strong ability to trap hydrogen isotopes may explain the discrepancy in these measurements. Increasing tita­nium content decreases the Dh/Dd ratio, while increasing chromium content increases the ratio.139

While the solubility of hydrogen in vanadium is lower than that in either zirconium or titanium, it is still very large, being greater than the value in palladium and much greater than the value in the other structural metals considered here (Figure 15). The reported hydrogen lattice solubilities in vana­dium alloys are in reasonable agreement, regardless of composition.1 Alloying additions do, however, change trapping in the alloy.

Titanium has a higher heat of solution for hydrogen than vanadium and titanium138 additions

increase the lattice parameter of vanadium.133 How­ever, titanium is a much stronger trap than other elements that increase the lattice parameter as much or more (including niobium, molybdenum, and zirconium).133 Pine and Cotts139 assert that tita­nium solute atoms trap not only hydrogen isotopes at nearest-neighbor interstitial sites, but also hydrogen substitutionally. They demonstrated that the binding energy varied from 3 kJ mol-1 in V-3Ti to 9.84 kJ mol-1 in V—8Ti. The trapping energy for D is larger than that for hydrogen for both alloys. However, it should also be noted that there is considerable short — range ordering in V—Ti alloys with more than ^4 at. % Ti.133 This ordering means that trapping will not obey an Oriani-type behavior, in which trapping would be linearly dependent on the number of solute atoms, because the solid solution is not random. The elements chromium, iron, and copper all reduce the lattice parameter of vanadium and the diffusivity change in alloys containing these elements is also much lower than that in alloys with Ti.133 In fits to the apparent diffusivity in tritium diffusion experi­ments in ternary V-Cr-Ti alloys, Hashizume eta/.135

show that, in addition to single titanium atoms, the most likely secondary trap is not chromium. Instead, the secondary trap has much higher energy and a lower concentration when compared to the monomer titanium trap. They also speculated that this was due to solute dimers and larger clusters. Interstitial oxygen, carbon, and nitrogen are also com­mon in vanadium alloys. One or more hydrogen atoms bind with single carbon or nitrogen atoms readily, and oxygen atoms tend to trap at least two hydrogen atoms each.144

Other defects, such as dislocations, may still be effective traps at 773 K.145 As with other materials, vanadium can be damaged by radiation, and this will likely be the dominant trap in fusion reactors.146,147

The recombination coefficient for hydrogen is over five orders of magnitude slower in vanadium than in nitrogen in the range of operating tempera­tures, and is relatively insensitive to the surface concentration of sulfur.129 Because of this and the high diffusivity of tritium, release is recombination limited in vanadium alloys. Deuterium ion-driven permeation experiments148 of V-15Cr-5Ti have

estimated the recombination-rate coefficient to be

2.4 X 10-29 m4 s-1 (although this measurement is three orders of magnitude lower than measurements on more dilute alloys149 and two orders of magnitude higher than measured in pure vanadium1 0). It should be noted that most measured recombination rates are lower bounds due to surface oxides. In environ­ments in which this native oxide layer may be dam­aged (such as by radiation in a fusion reactor), the actual recombination rate may be higher.147

V-Cr-Ti alloys have hydrogen permeabilities that are at least two orders of magnitude more than nearly any other blanket material and form detrimental hydrides.129’130’151-156 The ongoing stud­ies of permeation barriers may allow mitigation of this significant disadvantage so that V’s positive traits in a high-energy neutron environment can still be utilized.

The ion energy determines the initial penetration depth of the particle and the vacancy concentration varies as a function of the implantation fluence.247 The higher the ion energy, the lower the fluence and the temperature required to create material damage beyond vacancies and vacancy clusters. For example, very small blisters were observed for 8 keV ions at RT and a fluence of 4 x 1021 He+-m~2 and above.247 In contrast, for low energies (<30 eV), temperatures >1300 K and fluences of about 1026 He+ m~2 and more are necessary to form bubbles and surface holes.149,248 The reason for the lack of blistering at low temperature and low ion energy is assumed to be the trapping of He-ions at defects/vacancies in the very near-surface range. With increasing temperature, the defects and He-atoms debond and the He-atoms diffuse toward the bulk, agglomerate, and result in blistering.249 Similarly, with increasing energy, the penetration depth of the He-atoms increases from nm for eV-ions up to 1.7 pm for 1.3 MeV He-ions and the probability of blister formation correspondingly rises. In both cases, whether the process is driven by He-diffusion or high penetration depth, blister­ing and exfoliation are expected to occur when the amount of He locally reaches 4at.% and 20-40 at.%, respectively.250

4.17.4.4.1.1 Influence of temperature

Vacancy mobility is dependent on temperature and starts at 523-573 k.251,252 As the mobility of vacancies and the formation of thermal vacancies are driving forces for the formation of bubbles, holes, and blis­ters, an increase in temperature increases the size and decreases the density of material damage.149,253 How­ever, it is not only the temperature during ion irradi­ation but also the annealing temperature during experiments such as thermal desorption measure­ments, which can influence the damage characteristics.247 The formation of holes and porous structures observed after thermal treat — ments,254 particularly for temperatures above the material-dependent recrystallization temperature, is related to the movement of vacancies, accelerating the expansion and coalescence of He bubbles, their migration to the surface,253 and subsequently the release of He. The latter is also a function of temper­ature, showing several release peaks between 400 and 1600 K related to different trapping sites247 and determines the amount of He retained as a func­tion of the incident fluence.242,247,255 However,

helium retention may be mitigated by cyclic He-implantation and high temperature heating, for example, flash heating to 2000 °C, because He flows away before critical amounts accumulate and form complex He-vacancy clusters with higher binding

250

energy.

In addition to the temperature-dependent chemical sputtering of beryllium when exposed to deuterium plasma, another temperature-dependent loss term is present in beryllium exposed to plasma bombard­ment at elevated temperature. The classical picture of the temperature dependence of erosion from chemically inert surfaces exposed to energetic parti­cle bombardment is composed of the superposition of a constant physical sputtering yield with an expo­nentially varying thermal sublimation curve. The classical picture is contradicted, however, by experi­ments that show an exponential increase in erosion at lower temperature that cannot be explained by classical thermodynamic sublimation. First observed by Nelson75 for a variety of metal surfaces, similar results have been measured for Be,76,77 W78 and C79,80 surfaces. In the case of carbon, this mechanism has been called RES.

In the case of beryllium, two explanations have been proposed and both rely on the large flux of ions incident during plasma bombardment to modify the

plasma-facing material surface. In the first, the inci­dent plasma ions, in addition to creating sputtered atoms from the surface, also create a population of surface adatoms. An adatom is an atom from a lattice site on the surface that has gained sufficient energy to leave its lattice location, yet does not have sufficient energy to escape from the surface as a sputtered atom. The atom then occupies a site on top of the regular lattice sites. Because an adatom does not have the same number of adjacent atoms as those in the lattice, it is less strongly bound to the surface and can therefore sublime at a lower temperature than one associates with equilibrium thermodynamic sublima­tion. In the second explanation, incident plasma ions that have thermalized somewhere below the surface of the lattice exert a stress on the surface atoms of the target again resulting in a lower binding energy of the surface atoms to the bulk of the material.

Measurements show atoms are being lost from the surface at thermal energies,77 rather than the energy associated with sputtered particles (i. e., on the order of electron volts). This seems to verify the loss mech­anism that occurs because of the thermal release of an ensemble of particles with a lower surface binding energy than that of bulk atoms of that element. Addi­tional measurements at elevated temperature have documented the variation in Be atom surface loss rate with changes of the incident flux of energetic particles.81 The larger the incident flux, the lower the onset temperature for the enhanced erosion. The implication of this enhanced loss rate at ele­vated temperature is a reduction of the permissible operating temperature of any plasma-facing material, or alternatively that the lifetime of a component operating at extreme temperature may be less than that expected based on the predictions from classical surface loss terms.