Surface treatment of PFMs, while extremely effec­tive for the current day short-pulse tokamaks (pulses typically less than a few seconds), are of limited value for the next-generation (quasi-steady state) machines because of the significant surface erosion expected. However, forming graphite or CFC homogenously with various erosion-mitigating elements is possible. The mechanisms for this mitigation are twofold: (1) geometric shielding by low erosion yield par­ticulate, and (2) changes in local chemical reactivity due to the presence of doping atoms. Promising results have been obtained by doping of graphite with boron, which resides substitutionally in the graphite lattice, trapping migrating interstitials and altering the electronic structure of the material. Boron doping has been shown60 to both reduce the erosion due to oxygen, and to significantly reduce the sputtering yield due to methane formation. How­ever, other factors, such as the drastic reduction in thermal conductivity that is unavoidable in boro — nized graphite, need to be factored into the overall picture. Boron is discussed in some detail here, mostly because it had received great early attention. However, newer dopant combinations61 have served to suppress erosion, and they have also not had such a negative impact on thermal conductivity, and are therefore considered superior for PFC application.

As discussed by Garcia-Rosales,62 until the mid-1990s, boron was the primary ‘dopant’ of interest in fusion PFM graphites for chemical erosion mitiga­tion. As seen in Figure 23,63 the inclusion of up to 15% boron in graphite can result in significant (an order ofmagnitude in the peak regions) reduction in the erosion yield. Studies to date indicate that for effective suppression a minimum of 3% boron in graphite is required. The mechanism behind this suppression may include the reduced chemical activ­ity of the boronized material, as demonstrated by the increased oxidation resistance64 or the suppressed diffusion caused by the interstitial trapping at the boron sites. From the mid-1990s onward, many other metallic element additions to graphite were studied for

their possible beneficial effect on erosion mitigation. Specifically, elements such as silicon, titanium, tung­sten, and vanadium have been studied with varying levels of success.62 These elements are somewhat less effective in erosion mitigation than boron, though a factor of two reduction is to be expected.65-67 In a recent review by Balden,61 considerably higher reduc­tions in erosion are noted. More recently, emphasis has been on the use of multielement doping strategies.61

Because the removal of graphite is significant both in terms of gross material loss (possible consumption of the entire wall for power devices) and enhanced tritium retention for the resulting carbon dust, the effect of any additive to graphite in terms of physical properties or impact on plasma performance when eroded (Section 4.18.2.1) needs to be considered. With the exception of boron, additive elements dis­cussed in the previous paragraph will all have a negative impact on plasma performance in compari­son with carbon atoms, and therefore the balance of reduced mass loss compared to enhanced parasitic radiative plasma loss (eqn [2]) must be considered. As for physical properties, at levels well below the threshold at which they are effective for erosion suppression (~3%), they are direct substitutional elements in the graphite lattice, effecting significant reduction in thermal conductivity (due to their mass-defect phonon scattering.) In contrast, titanium doping, as evidenced in materials such as the Russian RGTi material, serves to enhance the graphitization

process, resulting in very well-crystallized materials of high (though somewhat anisotropic) thermal con­ductivity. A comparison of several element additions to graphite and their effects on the properties of graphite has been carried out by Paz68 and discussed by others.47,69 It is seen (Figure 2468) that all the metallic inclusions studied, with the exception of silicon,70 had, at these graphitization temperatures, the effect of enhancing the effective length (perfec­tion) of the basal plane of the graphite crystals, which is directly linked to enhanced thermal conduc­tivity. In comparison the basal crystal lengths of the Poco nuclear graphite see Table 2 and the Russian RGTi,71 which is processed at a similar graphitiza — tion temperature but with an applied electric current are shown in Table 2. The right hand side of Figure 24 shows the effect of varying the amount of titanium on the crystallite size, indicating that there is an increase in crystallite size with an increase of up to a few atomic percent of titanium.

In addition to the thermal process for chemical erosion described in the previous section, a second route to erosion, which is limited to the regions very near the surface, is also of importance.62 Specifically, for ions of <100eV, the formation of sp3 complexes occur at the surface of the graphite with very low binding energy (<2 eV compared to the 7.4 eV bind­ing energy for carbon in bulk graphite). Because of this very low binding energy the complex can be easily physically sputtered from the surface. Doping of graphites is also somewhat effective in reducing this surface erosion, attributed to the build-up of higher mass ‘dopant’ elements as the carbon atoms
are preferentially sputtered from the surface, effec­tively armoring the surface.62

For machines that will run in steady state such as ITER, moisture and oxygen evolving from the sur­face may not be a significant issue. However, oxygen is the most damaging impurity to current tokamaks through its presence in the molecular form, or as water vapor, and its tendency to be strongly adsorbed by carbon PFMs. Consequently, this impurity has a large impact on the plasma performance and erosion. The release of oxygen from irradiated carbonaceous films has been reviewed by Haasz72,73 and others. It has been clearly demonstrated that the carbon flux away from the first wall is directly related to the evolving oxygen. Typically, the oxygen enters the plasma from the PFMs in the form of CO or CO2. Figure 25 shows the strong temperature depen­dence of the erosion yield of a variety of graphites and codeposited (near amorphous redeposited car­bon) materials.74 It is noted that the data of Figure 25 are measurements of erosion yield by thermal oxida­tion in an O2 environment. Without special PFM surface treatment, such as plasma glow discharge and bake-out of the surface material, these fluxes dominate the surface erosion. For this reason, exten­sive research has been conducted into modifica­tion of graphite surfaces with impressive success in enhanced plasma performance.75 These improve­ments are due not so much to suppressed carbon erosion as to the decrease in the amount of oxygen released from the graphite. Toward this end, doped graphites have been modified to incorporate thermally and physically sputter-resistant carbides by doping

75 76 77 78 79

with titanium,‘ boron,‘ ’ 7 beryllium,‘ and silicon.’ Comprehensive reviews can be found for the chemi­cal erosion of graphite46,47,55,56,80 doped graphite by hydrogen,62 and also an article on the surface treat­ment of a graphite wall by Winter.75

At the present time, the ITER first wall and shielding blanket is still undergoing a major redesign to over­come some of the main design shortcomings that were identified in the context of design review
conducted in 2007; for example, the thermal load requirements were updated, in light of experimental experience.16 Most important in this respect was the recognition that the upper X-point region would see much higher loads during burn than 0.5 MW m~2; long transients (approximately up to 5-10 s) of plasma contact with the wall would have to be with­stood. In addition, NB shine-through at low densities would necessitate high heat flux first-wall protection, and a new requirement has been introduced to pro­vide remote maintainability of the first-wall panel to be done in situ and independently of the shield mod­ule (which would also have to be maintainable). The rationale for the ongoing effort is described by Lowry et a/.156 Proposed design modifications are being developed while trying to avoid and minimize changes to other components which are on the criti­cal fabrication path, especially the vacuum vessel, which is under fabrication.

The main features of the proposed design are the following: (1) to abandon the port-limiters and to exploit the first wall for plasma startup by relying on more benign plasma start-up scenarios, including an early X-point formation; (2) to use suitably shaped plasma-facing surfaces to hide edges such that there is no illumination of component surfaces by

Support pins

Halo current path
through the tile

Halo current from plasma

Figure 18 Inner wall guard limiter tile (exploded view, top, and prototype, bottom). The five castellated Be slices have interslice and outer slice internal toroidal edges ski-slope shadowed. The slices are held on an inconel carrier by pins which allow bowing under thermal load. The RH bolts are designed to be shadowed by the next installed tile. Reproduced with permission from Riccardo, V. J. Nucl. Mater. 2009, 390-391, 895-899.

normal or near-normal field lines, emanating in the near SOL; (3) provide power load capability of 4-5 MW m~2, in order to be able to use the first wall as a limiter for startup and termination; and (4) with­standing transients is still subject to discussion. In particular, it must be noted that the vulnerability to damage induced by thermal transients is recognized and linked to the feasibility and efficiency of all processes required for full remote maintenance of the first-wall panels, which is yet to be demonstrated.

In practical terms, the approach adopted is to pro­vide a shadowed poloidal band in the center of the first-wall panel, the two sides being shaped in a form typical for limiters both to provide the shadowing of the band and to ensure that the toroidally facing edge of the first-wall panel is shadowed. Because of the port regions on the low-field side, which contain a variety of structures with varying power handling capabilities, and because the toroidal field ripple is variable with the toroidal field, it is not possible to exploit the entire
wall surface in this location. For this reason, the first — wall panels on the low-field side have the poloidal bands between the ports advanced with respect to those in line with the ports (see Figure 19). The amount of set back required at the edges of the first wall is determined by the penetration angle ofthe field lines and the power scrape-offlength, with the optimi­zation taking into account the differing power handling capability of the front face and the edge of the first wall.

Considerations discussed here are limited to some problems associated with the design of the beryllium tiles and prediction of PWI effects during operation in ITER.

An important design driver for the first wall in the past was the specification of the thermal load during off-normal transient events.3 In particular, the thick­ness of the beryllium tiles had to be such as to prevent overheating of the joints and possible damage of the coolant pipes (see Section 4.19.6.2.2). Also, the thickness of the tile determines the temperature

Be wall

CFC strikepoints W elsewhere

(a)

Figure 19 (a) View of the low field side first-wall surface showing how the panels in line with the port openings are recessed with respect to those between. It also shows the shielded central section of the panels allowing for access to the mechanical and hydraulic connections. Reproduced from Hawryluk, R. J.; etal. Nucl. Fusion 2009, 49, 15, 065012; with permission from IAEA. (b) Allocation of armor materials. Reproduced from Hawryluk, R. J.; etal. Nucl. Fusion 2009, 49, 15, 065012; Federici, G.; Loarte, A.; Strohmayer, G. Plasma Phys. Contr. Fusion 2003, 45, 1523-1547, with permission from IOPP.

gradient and the thermal stress under a prescribed thermal load during steady-state. Limits on the tile temperature during operation arise as a result of many processes including melting, excessive vapori­zation, thermal fatigue, reduced mechanical integrity, and chemical reactions during accidental exposure of armor or structure to air or steam. The last one of the above processes is important as explosion of hydro­gen liberated from the steam-Be reaction is a major concern. In the past, a tile thickness of 10 mm was adopted. This corresponded to a Be maximum tem­perature limit of ^650-750 °C, roughly the level at which the relevant Be material properties (including mechanical, embrittlement, thermal fatigue, and swelling effects) start to degrade considerably.

Because of the differences in the product of the elastic modulus and the coefficient of thermal expan­sion (E) between beryllium and copper or copper alloys (EBe/EcucrZr = 2.4), large thermal stresses are set up around the bond between the beryllium tile and copper allot heat-sink.

The difficulty to successfully join low thermal expansion armor materials such as beryllium and tung­sten to high thermal expansion heat sink materials has been a major problem and has been discussed in

Section 4.19.5. Thermomechanical modeling has shown the desirability of using very small tiles of brush like structure for PFC armor because of the reduction ofthe stress at the armor-heat sink interface. The proper selection of the size of beryllium tile is an important issue which impacts all aspects of compo­nent manufacturing such as increased cost of machin­ing, nondestructive examination features, reliability and repair of unbonded tiles, etc.

In general, the fatigue life issue is difficult to quantify because of a number of factors. The thermal stresses depend on the temperature profile and the degree of constraint in the tiles. Tile castellations must be introduced to further relieve the constraints, and these have been sized following an extensive program of coupled thermal and mechanical analyses using finite elements codes such as ANSYS185 and ABAQUS.186

More conventionally deposited coatings of Er2O3 or Y2O3 were developed such as arc source plasma deposition,25 electron beam-physical vapor deposition (EB-PVD),2 and radiofrequency (RF) sputtering.26 Because of the variation in the quality of the coating and the test conditions, reported stability in liquid lithium was different in different tests. Er2O3 produced by EB-PVD was heavily damaged in Li already at 500 °C,27 and Er2O3 and Y2O3 coatings produced by RF sputtering were also exfoliated at 500 ° C.28 How­ever, Er2O3 fabricated by arc source plasma deposition showed promising results, as shown in Figure 8. Depo­sition on a higher temperature substrate produced a highly crystalline Er2O3 coating, which was shown to be stable in Li for 1000 h at 700 “C.1 The stability in Li is known to be enhanced by improving the purity and crystallinity of the coating. An oxide layer at the coating/substrate interface may cause extensive exfoli­ation because Li introduced through cracks would preferentially attack the oxidized interface.

4.21.3.3.3 Other coating technologies

The efforts in using the physical coating processes explained so far are essential for establishing the
coating concepts. However, further efforts are also necessary to enhance the engineering feasibility of coating on large and complex surfaces. For this purpose, the sol-gel method (in other words, metal-organic deposition, MOD) and metal-organic chemical vapor deposition (MOCVD) have been explored.

Er2O3 coating was formed on stainless steel by the sol-gel method. Crystallinity of the coating depended on annealing atmosphere and tempera — ture.29 MOCVD was applied to coating Er2O3 on V-alloys and other materials. Successful coating on the inner surface of a tube was demonstrated.3

Table 1 provides a non-exhaustive summary of ceramic breeder blanket designs, and key parameters, compiled from the references in this chapter. Most of the concepts have a ferritic-martensitic steel as the structural material. Typical values for the expected blanket neutron wall load are about 3 MW m~ , and its lifetime is mostly considered to be limited by about 150 dpa for the structural material. For a DEMO reactor, which is not intended to be a power plant with high availability, typical lithium burnups are 11% (Li4SiO4) or 17% (Li2TiO3), and fast neutron damage in the reduced activation ferritic/ martensitic (RAFM) steels is about 70 dpa.47,48 A typical lifetime for a power reactor blanket is esti­mated to be of the order of 5 years, implying about ten replacements during a 60-year reactor life. The blanket structure should, therefore, be designed with low-activation materials, enabling it to be recycled typically in a period up to 100 years. Such activation requirements have led to Li4SiO4 and Li2TiO3 as the preferred ceramic breeder systems for the European HCPB concept. A case study for Li4SiO4 has been elaborated by Fischer and Tsige-Tamirat.49

Beryllium is a low-Z material with good thermal char­acteristics, described in Chapter 2.11, Neutron Reflector Materials (Be, Hydrides) and Chapter 4.19, Beryllium as a Plasma-Facing Material for Near-Term Fusion Devices. Additionally, it is a good getter for oxygen impurities in the plasma. The low-Z minimizes the radiation losses from the plasma, and the oxygen removal keeps the plasma clean. For these reasons, beryllium has been used in the JET fusion reactor and will be the first wall material for the International Thermonuclear Experimental Reactor (ITER). Beryllium has interesting hydrogen retention behavior. Beryllium may also be used as a neutron multiplier in the blanket area of future fusion devices to increase the tritium breeding ratio.

Abramov et a/.72 used two grades of beryllium in their permeation-diffusion experiments. These were high-purity (99%) and extra grade (99.8%). Adding to the validity of their experimental result was the fact that the authors used multilayer permeation theory analysis to take permeation through the outer oxide layer into consideration. For a lower
purity material (98%), Tazhibaeva et a/.73 also used the multilayer permeation analysis to determine dif — fusivity. Jones and Gibson74 studied tritium diffusiv — ity and solubility for arc-cast beryllium in the temperature range of 673-1173 K. Beryllium was exposed to tritium gas for various temperatures, durations, and pressures during isothermal anneals. After removing the samples to another experimental system, the samples were heated to various tempera­tures. For the initial heating, the tritium release would rise, but soon fall to zero. Elevating the tem­perature would reestablish the tritium release, but again the release would fall. While this behavior is not typical of diffusion controlled release, the data were analyzed to extract an effective diffusivity. The different reported diffusivities are shown in Figure 8. It can be seen that the diffusivity reported by Abramov et a/.72 is considerably larger than those of Tazhibaeva et a/.73 and Jones and Gibson.74 It is apparent that the purity of the beryllium played a strong role in deter­mining the effective diffusivity. Oxygen, the primary impurity in beryllium provides a strong trap for hydrogen. Thompson and Macaulay-Newcombe75,76 examined the diffusion of deuterium in single-crystal and polycrystalline beryllium. The effective diffusivity in the single-crystal material was lower than that for the polycrystalline material. The polycrystalline results agreed quite well with the results reported by Abramov et a/.72 They suggested that the lower diffusiv — ity seen for the single-crystal samples was the true diffusivity for beryllium, and that the polycrystalline results represented diffusion along the grain boundaries.

Temperature, 1000/ T (K-1)

If hydrogen isotopes migrate along the grain bound­aries, it is logical that the rate of migration would be affected by oxygen segregated to those boundaries.

The very limited results for hydrogen isotope solubility in beryllium are shown in Figure 9. In the earlier described experiments by Jones and Gibson,74 the solubility was seen to be effectively independent of temperature in the temperature range 550-1250 K. For sintered, distilled a-beryllium, Shapovalov and Dukel’skii77 reported similar values of solubility for the temperature range 673-1473 K. In experiments using 98.5% and 99.8% pure beryl­lium samples, Swansiger78 used gaseous uptake of tritium to determine the solubility. The amount of tritium uptake did not increase with increasing sample size. The solubility for the two purity materi­als was also seen to be the same. For temperatures below 650 K, the apparent solubility increased; this strange effect was attributed to trapping. It is interesting to note the fact that the apparent
solubility over temperatures at which the three research groups‘ ’ ’ performed their experiments varied by less than one order of magnitude even though the activation energies varied by 96 kJ mol-1. It should be questioned whether the reported values really represent the solubility of hydrogen isotopes in bulk beryllium.

For all plasma-facing materials, there is concern that implantation of energetic deuterium and tritium could lead to excessive retention and permeation. Implantation of hydrogen isotopes into a material with a low recombination-rate constant can lead to a majority of the hydrogen being pushed into the bulk of the material. In the limiting case of slow recombi­nation, 50% of the hydrogen exits the front face and 50% exits the rear face. Langley79 implanted 25 keV deuterium into 99.1% pure hot isostatic pressed beryllium. The retention was seen to be 100% until the particle fluence reached 2 x 1022 D m — . The retention flattened to a limit of ^2.8 x 1022 D m-2. Wampler80 recorded similar results for his implanta­tion of 0.5 and 1.5 keV deuterium into 99.6% pure beryllium samples. Saturation occurred at 0.31 D/Be in the implant zone. Yoshida et a/.81 used 99% pure beryllium in his implantation experiments with 8 keV deuterons. Transmission electron microscopy revealed bubble formation at all temperatures between room temperature and 873 K. The bubbles were not removed even by annealing at temperatures up to 973 K. Plasma exposure was used by Causey et a/.82 and by Doerner et a/.83 in low-energy, high-fluence deuterium exposures to beryllium. In both studies, the fractional retention was extremely low and decreased with increasing temperature. Open poros­ity in the implant zone was listed as the likely cause of the low retention. Chernikov et a/.84 and Alimov et a/.85 showed bubbles and microchannels to be responsible for the behavior of implanted hydrogen in beryllium. At 300 K, very small bubbles with a high volume density are formed even at low fluences. As the fluence is increased, the bubbles agglomerate into larger bubbles and then form microchannels that eventually intersect with the surface. For irradiation at 500-700 K, small facetted bubbles and large oblate, gas-filled cavities are formed. This microstructure was seen to extend well beyond the implant zone. Alimov et a/.85 postulated that the hydrogen retention in the porous region was due to binding to the beryl­lium oxide that forms on the pore surfaces.

Beryllium is known as a neutron multiplier because of the reaction 9Be + n! 8Be + 2n. Another neutron reaction for beryllium is 9Be + n! 4He + 6He, followed by 6He decaying to 6Li. 6Li has a very large cross-section to absorb a thermal neutron and produce a helium atom and a tritium atom. Baldwin and Billone86 calculated the amount of tritium that could be produced in a large fusion device of the future. In an experiment, they exposed beryllium to a neutron fluence of 5 x 1026 n m-2 with 6% of the neutrons having energy > 1 MeV. The resulting tri­tium level was determined to be 2530 appm. Scaling up to a fusion reactor with 50 Mg of beryllium exposed to 3 MWy m-2 results in the production of

5.5 kg of tritium. This is a sizeable quantity of tritium. The relevant question is whether this tritium would be released during normal operation of the fusion plant. Baldwin and Billone86 examined exactly that question in their experiment. The samples containing the 2530 appm of tritium were heated in stepped anneals to determine the release rate of tritium from beryllium materials with different densities. The annealing began with a very long anneal at 773 K, and the temperature was increased in incre­ments of 100 K. For each temperature, there was a nondiffusional burst of release followed by a rapid decrease in the release rate. The release behavior for the different materials was similar, but the fractional release was greater for the less dense materials. Andreev et a/.87 irradiated hot-pressed beryllium at 373 K. After neutron irradiation, thermal desorption spectroscopy was used with a heating rate of10 K s-1. Release began to occur at -~773 K. The temperature at which maximum release occurred depended on the neutron fluence. The sample irradiated to a fluence of 3 x 1025nm-2 had a peak release at 1080 K, while the sample irradiated to the higher fluence of 1 x 1026nm-2 exhibited a peak release at a lower temperature of 1030 K. The authors examined the microstructure of the samples after the release anneals. If the anneal was stopped at 973 K, pores with a diameter of 2-16 pm were formed. If the anneal was taken to 1373 K, the pore diameters increased to 25-30 pm.

Due to the toxicity of beryllium, there have been relatively small numbers of experiments performed on the behavior of hydrogen isotopes in beryllium. The apparent diffusion coefficient of hydrogen in beryllium is strongly affected by purity levels. The values determined for the solubility of hydrogen in beryllium all fall within one order of magnitude even though the apparent activation energy differs by 96 kJ mol — . Implantation of hydrogen into beryllium results in the formation of bubbles and eventually open channels or porosity. Connection of the porosity to the surface facilitates the release of hydrogen from the beryllium as the particle fluence is increased. The tendency to form bubbles would suggest that the solubility of hydrogen in beryllium is extremely small. It is possible that the values determined for the solubility of hydrogen in beryl­lium actually represent the amount of hydrogen absorbed on the external surface and on the grain boundaries. The measured diffusivity may represent migration along the grain boundaries. More experi­ments, and experiments with single crystals, are needed to answer these questions. For beryllium used for long times in future fusion devices, tritium produced by neutron reactions on the beryllium is likely to dominate tritium retention in beryllium. Tritium inventory from eroded beryllium codepos­ited with tritium may play a strong role in tritium inventory, but that effect is not covered in this review.

Data in the literature on mechanical properties of neutron-irradiated tungsten are very limited.234’238’239 However, in combination with experimental results obtained for other refractory metals, it has been shown that in metals with a bcc lattice structure’ irradiation hardening causes a steep increase in yield stress and a decrease in ductility.110 Conse­quently, the material fails by brittle cleavage frac­ture as soon as the yield stress exceeds the cleavage strength. Therefore, the increase of the DBTT depends on the neutron fluence, the neutron spec­trum (will be addressed by the International Fusion Materials Irradiation Facility, IFMIF), and the irradiation temperature. The radiation hardening in bcc alloys at low temperatures (<0.3 Tm) occurs even for doses as low as ~0.15—0.6 dpa (irradiation of plasma facing materials for ITER and DEMO, PARIDE campaigns217), which corresponds to the expected ITER conditions. Therefore, operation of tungsten at temperatures >1000 °C would be preferred as full or at least partial recovery of defect-induced material degradation is achieved by annealing at 1200 °C.234 This implies that the near­surface part of a W component will retain its ductility, which has a beneficial effect on the crack resistance at the plasma loaded surface. However, such temperatures are not feasible at the interface to the heat sink where tungsten will be in contact with copper (ITER) or steel (DEMO), which are limited to significantly lower operational tempera­tures. Hence, better understanding of the irradiation effects on tungsten at temperatures between 700 and 1000 °C is needed, particularly related to reactor

application in DEMO.109,110,240

In addition to the influencing factors on the DBTT mentioned above, that is, neutron fluence, neutron spectrum, and irradiation temperature, the material’s composition also plays an important role. While the addition of Re has a beneficial effect on the material’s ductility in the nonirradiated state, under neutron irradiation it results in more rapid and severe embrit­tlement than it is observed for pure W2 9 Similarly, less mechanical strength and an increased loss of ductility compared to pure W is found for particle — strengthened W alloys (e. g., W—1% La2O3) when tested up to 700 °C. The only exception among all explored tungsten alloys might be mechanically alloyed W-TiC (see Section 4.17.3.3) that showed no irradiation hardening as measured by Vickers hard­ness at 600 °C.87

Finally, the mechanical properties are influenced by neutron-induced He-generation and the transmu­tation of tungsten. While He generation in W is, com­pared to CFC and Be, very small (~0.7 appm He per dpa) and its influence on the mechanical properties of W negligible,73,83,224 the transmutation of W into Re and subsequently Os significantly alters the mate­rial structure and its properties. The generation of significant amounts of ternary a and subsequently а-phases results in extreme material embrittlement and will cause shrinkage. In combination with ther­mally induced strains, this might produce high tensile stresses causing the extremely brittle material to extensively crack and perhaps even crumble to powder.36

Physical sputtering results from the elastic transfer of energy from incoming projectiles to atoms on the surface of the target material. Target atoms can be sputtered when the energy they receive from the collisional cascade of the projectile exceeds the bind­ing energy of the atom to the surface. The physical sputtering rate is usually referred to as the sputtering yield, Y, which is defined as the ratio of the number of atoms lost from a surface to the number of incident energetic particles striking the surface. The lower the binding energy of surface atoms, the larger the physical sputtering yield. As physical sputtering can be approximated using a series of binary collisions within the surface, it is relatively easy to estimate

the physical sputtering yield of given projectile-target scenarios. Monte-Carlo based simulation codes (such as transport of ions in matter (TRIM))53 have been used to generate extensive databases of sputtering yields based on incident particle angle, energy, and mass, for a variety of targets54 including beryllium.

Measurement ofthe physical sputtering yield from a beryllium surface is complicated by the natural affinity of beryllium for oxygen. A beryllium surface will quickly form a thin, stable, passivating oxide surface layer when exposed to atmosphere. In ion beam devices, it is possible to clean any oxides from the beryllium surface before a measurement and with careful control of the residual gas pressure, make the measurements before the oxide reforms on the surface and alters the measurement.55 It has also been shown that it is possible to deplete the beryllium surface of oxide by heating the sample to tempera­tures exceeding 500 °C, where the beryllium below the oxide can diffuse through the oxide to the surface,56 thereby allowing measurements on a clean beryllium surface. The comparison between the calcu­lated sputtering yield and measurements made using mass-selected, monoenergetic ion-beams devices impinging on clean beryllium surfaces is fairly good.57

Measurements of sputtering yields in plasma devices, however, are complicated by several factors. In plasma devices, the incident ions usually have a temperature distribution and may contain different charge state ions. Each different charge state ion will be accelerated to a different energy by the elec­trostatic sheath in the vicinity of the surface. When hydrogenic plasma interacts with a surface, one must also account for a distribution of molecular ions striking the surface. In the case of a deuterium plasma, for example, the distribution of molecular ions (D+, Dj, D3) must be taken into account as the incident molecule disassociates on impact with the surface and a Dj ion becomes equivalent to the bombardment of two deuterium particles with one-half the incident energy of the original Dj ion. Figure 3 shows the change to the calculated sputter­ing yield when one includes the effects of molecular ions in a plasma, compared to the calculated sputter­ing yield from pure D+ ion bombardment.

The trajectory of the incoming ions can also be altered by the presence of electrostatic and magnetic sheaths. Plasmas also contain varying amounts of impurity ions, originating either from PWIs in other locations of the device, or ionization of residual back­ground gas present in the device and these impurity ions, or simply neutral gas atoms, may interact with
the surface. Finally, the incident flux from the plasma is usually so large that the surface being investigated, and its morphology, becomes altered by the incident flux and a new surface exhibiting unique character­istics may result.

Nevertheless, the physical sputtering yield from beryllium surfaces exposed to plasma ion bombard­ment has been measured in several devices. Unfortu­nately, there is little consensus on the correct value of the physical sputtering yield. In JET, the largest confinement device to ever employ beryllium as a PFC sputtering yield measurements range from values far exceeding47 to values less58 than one would expect from the predictions of TRIM. In the Plasma Interaction with Surface Components Experimental Station B (PISCES-B) device, systematic experiments to measure the physical sputtering yield routinely show values less59-61 than those expected from TRIM. This difference is shown in Figure 3, where the energy dependence of the calculated yield is com­pared to experimental measurements.

Another primary difference between the condi­tions in an ion beam device and those encountered in a plasma device has to do with the neutral density near the surface being investigated. In an ion beam experiment, the background pressure is kept very low

Figure 3 Calculated sputtering yields from pure D+ bombardment at normal incidence compared to that calculated fora (0.25,0.47,0.28) mix of D+, Dj, and Dj; also shown is the measured yield from such a plasma.

so that the surface being probed maintains its clean properties. On the other hand, the incident flux in a plasma device is usually several orders of magnitude larger than in an ion beam device, ensuring that the surface remains clean because of the large incident flux. However, this plasma-facing surface undergoes not only energetic ion bombardment, but also bom­bardment by neutral atoms and molecules.

The neutral density in plasma generators is typi­cally on the order of 1020 m~3 (a few millitorr) which is necessary for breakdown of the plasma. The esti­mated neutral atom flux is approximately equal to the incident ion flux to the surface61 and it is often not possible to alter significantly this flux ratio. In the case of a beryllium surface which can form a hydride (see Section 4.19.3.1.3), the presence of adsorbed deuterium on the surface could affect the measured sputtering yield by decreasing the beryllium concen­tration at the surface and altering the binding energy of surface beryllium atoms.

Some evidence of this effect may be discerned in data from JET measurements of the beryllium sputtering yield. Two sets of sputtering yield mea­surements have been reported from JET; one from beryllium divertor plate measurements and the other from beryllium limiter measurements. In the divertor region, one expects a neutral density similar to that encountered in plasma generators (1020 m~3 or more) and the measured sputtering yield is lower than that predicted by TRIM calculations.58 When sputtering measurements are made on the limiter, where the neutral density is typically lower, the sputtering yield agrees with, or exceeds, the calculated value.47 Of course, other issues such as impurity layers on the divertor plate and angle ofincidence questions tend to confuse the results. However, the data sets from JET are consistent with the impact of neutral absorption on the beryllium plasma-facing surface.

Effects associated with plasma operation will need to be taken into account when predicting sputtering yields from different areas of confinement devices. In addition to the low-energy neutral atom flux and higher-energy charge exchange neutral flux, the impact of small impurity concentrations in the inci­dent plasma flux will also have a large impact on the expected sputtering yield. Some of the implications of the formation of a mixed-material surface are discussed in the next section and in Section 4.19.3.3.

4.20.2.1 Pure Copper

Copper is widely used where high electrical or ther­mal conductivity is required. Pure copper is defined as having a minimum copper content of 99.3%. Copper with oxygen content below 10 ppm is called ‘oxygen — free.’ ‘Oxygen-free, high conductivity’ (OFHC) grade copper has room temperature electrical conductivities equal to or greater than 100% International Annealed Copper Standard (IACS), where 100% IACS = 17.241 nO m at 20 °C.3 Copper grades with the ASTM/SAE unified number system (UNS) designation C10100, C10200, C10400, C10500, and C10700 are classified as OFHC copper. Grades C10400, C10500, and C10700 have significant silver content, which creates activa­tion hazards. Only C10100 and C10200 are considered for fusion systems.

The use of unalloyed copper is often limited by its low strength. Copper can be strengthened by various processes, for example, cold working, grain refine­ment, solid solution hardening, precipitation hard­ening, dispersion strengthening, etc. While these approaches can significantly increase the strength, they can also lead to a pronounced reduction in con­ductivity. The challenge is to design a material with the best combination of strength and conductivity.

Cold work can significantly increase the strength of pure copper and has a relatively moderate effect on conductivity.4 However, cold-worked copper can be softened at relatively low temperatures (^200 °C) because of its low recrystallization temperature.5 A recent study has shown that ultrahigh-strength and high-conductivity copper can be produced by introducing a high density of nanoscale twin bound — aries.6 The tensile strength of the nano-grained cop­per can be increased by a factor of 10 compared to conventional coarse-grained copper, while retaining a comparable conductivity. The potential of high — strength, high-conductivity bulk nano-grained cop­per in nuclear energy systems, however, has not been widely explored.

Alloying in copper can significantly improve mechanical strengths and raise the softening tempera­tures. However, additions of alloying elements also reduce electrical and thermal conductivity. Among the three alloying strengthening mechanisms, namely, solid solution hardening, precipitation hardening, and dispersion strengthening, solid solution hardening has the most detrimental effects on the conductivity4 and is the least favored mechanism to obtain high- conductivity, high-strength copper alloys.

Within the fusion program, another area of concern is related to the effects of radiation on the optical prop­erties of the dielectric materials to be used as trans­mission components (windows, lenses, and optical fibers) for the UV, visible, and near-IR wavelength diagnostic systems needed for control and safety, as well as maintenance (remote handling).21,26,140,154,155 Radiation-induced optical absorption (RIA) and light emission or RL impose severe limitations on the use of any optical material within an intense radiation field. For remote handling applications, the optical components will be expected to maintain their trans­mission properties under high levels of ionizing radi­ation (< 1 Gy s~ ) during hundreds of hours. For such applications, RIA imposes the main limitation, but can be tolerated. However, in the case of diagnostic applications, in addition to a higher level of ionizing radiation (tens to hundreds Gys~ ), the materials will also be subjected to atomic displacements >10~ dpas-1. It soon became clear that both RIA and RL would impose severe limitations on the main candidate materials (sapphire and silica). Of these two materials, sapphire is by far the most resistant to ionizing radiation. Although ionizing radiation can cause an increase in optical absorption because of trace impurities and vacancy defects present in the material, it is in general the displacement damage mechanism which induces absorption at first in the UV region as a result of oxygen vacancy-related defects.30,33,156-158 This fluence (dose) effect reduces the transmission in the UV region to essentially zero for doses above about 10~4dpa, and more slowly in the visible as the tails of the absorption bands begin to overlap into this region. Although sapphire shows more radiation resistance than SiO2 in terms of opti­cal absorption, the material was found to be unsuit­able for many diagnostic applications because of its intense RL, as will be seen below.

As with RIC, RL is ionizing flux (dose rate) dependent and hence will be a problem from the onset of operation of future fusion devices. Further­more, to assess RL clearly requires in situ measure­ments during irradiation. While many studies had been carried out on luminescence phenomena in SiO2 and sapphire, the problem was only addressed in a quantitative way because of fusion application requirements.159-164 Sapphire was quickly excluded from high-dose rate applications when it was shown that the photon emission for a typical diagnostic window dose rate would be comparable with the photon emission from the plasma.159 In contrast, certain grades of silica show virtually no RL in the UV-visible region, the emission being limited almost to the Cherenkov background. Quantitative lumines­cence data comparing UV grade sapphire and two types of silica, both of which show low RL, are given in Figure 13, indicating that suitable materials do exist in which the RL can be reduced to a minimum, although there are limited data on RL as a function of fluence.162-164 In particular, the KU1 and KS-4V quartz glass materials, provided by the Russian

Federation for the ITER diagnostics radiation testing program, have proved to be highly resistant to RL and RIA because of ionizing radiation and displace­ment damage, and are now reference materials.26,165-170 For ionizing radiation doses up to at least 100 MGy and for temperatures at or above about 100 ° C, very little absorption is induced in the KU1 material over the whole visible range; one must keep in mind however that with irradiation displacement dose the optical absorption related to oxygen vacancies in SiO2, as in all oxide materials, eventually renders them opaque in the UV and visible range.171-175

In an analogous way to the ECRH transmission windows, mention should be made of windows required for high-power laser transmission, that is, the LIDAR (light detection and ranging) system. This demanding diagnostic system being considered for ITER will require very high-quality transmission windows for the high-power laser pulses at about 500 and 1000 nm. It is estimated that transmission losses of the order of 5% may cause problems with the window integrity because of laser damage. However, such small decreases in the transmission corresponding to an optical density increase of only 0.02 are extremely difficult to measure by standard PIE of irradiated optical materials. Such measurements have to be per­formed in situ. In situ measurement is also required in order to determine possible radiation-enhanced absorption which can easily reach such small values. The possibility of radiation-enhanced dielectric breakdown due to the intense laser pulse and the
ionizing radiation has also to be considered. However, such a determination requires an elaborate in situ experiment. Work on laser-induced damage in KU1 and KS-4V has confirmed the limited influence of RIA and RIC on the damage threshold for high — power laser transmission.176 However, metallic depo­sition due to sputtering or evaporation can seriously reduce the damage threshold even for a few nanome­ter thickness, as may be seen in Figure 14. The effect is strongly material dependent, and furthermore self­cleaning with subthreshold laser pulses is not effective for all deposited materials.177,178

Although in general RL is considered to be a problem for diagnostic systems in future devices, it may be employed as a detector/converter for X-ray, UV, and particle emission from the plasma. The intense RL from Al2O3:Cr, for example, has been used for many years in ceramic fluorescent screens for accelerator beam alignment,179 and is now being developed with improved radiation resis­tance and rapid decay times for fusion applications, along with other alternative luminescent materials (Figure 15).1 0-1 2 Furthermore, RL is a potentially powerful tool capable of monitoring material modifi­cation during irradiation, but has been largely neglected within the fusion materials activities, in part because of the difficulty in interpreting the resulting emission spectra. However, the technique is now being successfully employed to study insulating materials such as aluminas and silicas, as well as breeding ceramics for fusion applications.183,184

Finally, in connection with optical transmission components, one should note the flexibility and sim­plification in diagnostic design that the use of optical fibers would allow. However, this is not straightfor­ward; although RIA and RL are problems for optical window and lens components, in the case of optical fibers the situation is far worse because of the length of the optical path. Furthermore, because of the

manufacturing techniques, fibers with characteristics as good as those observed for the KU1 quartz glass for example have not been obtained. This has prompted an extensive collaborative research program to find the most suitable types of radiation-resistant fiber. Several different optical fibers have been examined to assess RIA and light emission, the viability of high-temperature operation and annealing, jacketing
material, and the influence of hydrogen loading. In addition, parallel work is being carried out on the possibility of photobleaching using high-intensity lasers to recover transmission, ‘holey’ fibers for improved transmission and radiation resistance, and fibers with extended blue — UV transmission.26,185-190 Irradiations have been carried out to total doses above 10MGy and 1022nm~ , and temperatures from about 30 to 300 °C. The most promising fibers are the hydrogen loaded KU1 and KS-4V, where above 400 nm they show the lowest RIA, as may be seen in Figure 16.139 Although the KU1 is the slightly better material up to about 700 nm, the intrinsic OH band and its harmonics notably affect transmission above 800 nm, so for a fiber required to transmit in the visible and IR regions, the hydrogen loaded KS-4V may be a better choice. For silica materials up to about 10 MGy, the main radiation damage mechanisms involve electron and hole­trapping; hence, the wide differences observed in induced absorption of the fibers tested are due to variations in intrinsic trapping centers (defects and impurities). In general, these trapping centers are thermally unstable, hence the effective thermal annealing for irradiation at higher temperature, or postirradiation thermal annealing. For higher doses, displacement damage leading to extensive structural damage begins to dominate, but by this time the fibers are of little use for diagnostic applications. Limited work is underway to examine the possibility of in situ photobleaching of the radiation-induced damage using high-intensity UV lasers, the potential of so-called ‘holey’ fibers (fibers containing an array of vacuum, air, or liquid filled holes) to improve radiation resistance, as well as fibers to extend trans­mission into the blue — UV region.

4.14.1.3.1 History

Master Curve methodology is based on the observa­tion that the fracture toughness transition curve for any ferritic steel has the same shape, no matter the steel (see Figure 2). Thus, a single ‘Master Curve’ can be used for all ferritic steels; the curve is simply shifted along the temperature axis to match a mean fracture toughness value, which is established from measured fracture toughness data for the particular steel being evaluated. The primary material fracture

toughness parameter is the transition temperature, T0, which characterizes the Master Curve position, and is defined as the temperature where the median fracture toughness is 100 MPa Vm (Figure 2).

The advantages of Master Curve technology over past methods for estimating the fracture toughness of materials (particularly irradiated materials) are (1) it is based on direct measurements of the property of interest (e. g., fracture toughness); (2) it provides a direct method of establishing the transition curve for irradiated materials (instead of inferring a shift in an assumed baseline bounding curve using Charpy data); and (3) it can be used for materials even with a limited availability of archival materials.

The development of Master Curve methodology was started in the 1980s by Wallin and his coworkers by the introduction of a mathematical model to describe the probability of cleavage fracture initiation in a material containing a distribution of potential fracture initiators (flaws). The model was completed by including the temperature dependence of Kjc, which was estimated empirically from a dataset including various ferritic structural steels. The scat­ter definition, the size adjustment, and the definition of the temperature dependence are the basic ele­ments of the Master Curve methodology described in ASTM E 1921.6-8 The approach has been verified in several round-robin and research programs.9

The first version of the Master Curve standard comprised a procedure for analyzing only single­temperature test data; in later versions, the approach was extended to consider multitemperature test data. The multitemperature approach requires finding a maximum likelihood solution for the value of the transition temperature, T0, from data measured over a range of temperatures, rather than at a single tem­perature. The first version of ASTM E1921 was approved in 1997 and issued in 1998 (ASTM E 1921­97). The multitemperature approach was included in the second revision, after which several other revisions with some minor changes have been released. The present revision, ASTM E 1921-08, describes proce­dures for the experimental determination of the elas­tic-plastic fracture toughness, Kjc, estimation of the reference temperature, T0, and principles for the lower bound curve definition of fracture toughness (Figure 3). Further detailed descriptions on the meth­odology and applications are given in McCabe eta/.10 and Sattari-Far and Wallin11.

The model12 has also been validated numerically to more accurately describe the true fracture behavior and the stress-strain distribution of bcc steels on a

micromechanical scale.13 The advanced numerical assessment capabilities presently available for multi­scale modeling of materials have made it possible to validate the main elements of the model. Despite several further developments primarily related to material inhomogeneity (in the basic model material macroscopic homogeneity is assumed), the basic approach being applied today is essentially the same as that developed over 20 years ago. No other defi­ciencies or assumptions requiring readjustment have been identified. Further verification of the approach and the validity of the empirically determined tem­perature dependence have been conducted. Some aspects, such as the lower shelf definition (Kmin), are practically impossible to verify only using experimen­tal methods, so that numerical modeling studies have been very instructive.

The Master Curve methodology is currently being used in both structural integrity and lifetime assess­ments. Typical areas ofapplication are pressure vessels and piping, nuclear RPV surveillance programs, other energy production structures, off-shore structures, and various welded components and bimetallic joints.