Chad Parish

This paper provides an overview of advanced scanning transmission electron microscopy (STEM) techniques used for characterization of irradiated BCC Fe-based alloys. Advanced STEM methods provide the high-resolution imaging and chemical analysis necessary to understand the irradiation response of BCC Fe-based alloys. The use of STEM with energy dispersive x-ray spectroscopy (EDX) for measurement of radiation-induced segregation (RIS) is described, with an illustrated example of RIS in proton- and self-ion irradiated T91. Aberration-corrected STEM-EDX for nanocluster/nanoparticle imaging and chemical analysis is also discussed, and examples are provided from ion-irradiated oxide dispersion strengthened (ODS) alloys. Finally, STEM techniques for void, cavity, and dislocation loop imaging are described, with examples from various BCC Fe-based alloys.

A compact electron cyclotron resonance plasma source has been utilized at Sandia National Laboratory to expose heated W samples (1270 K) to 50–75 eV He ions at fluxes on the order of 1019 m-2 s-1 and fluences on the order of 1024 m-2. Scanning electron microscopy (SEM) analysis of the surface has indicated bubbles up to 150 nm in diameter that exhibit signs of bursting near the surface. Comparisons have been made between W samples prepared from warm-rolled W sheet stock and ITER-Grade W rod stock. Focused ion beam (FIB) cross sectioning has been used with SEM and transmission electron microscopy (TEM) to identify large sub surface bubbles (100 nm diameter) at depths up to one micron as well as a dense layer of smaller bubbles (<10 nm diameter) within the first 100 nm of the surface, similar to bubble layers observed on higher flux experiments. SEM-Electron Backscatter Diffraction (EBSD) analysis has identified a unique surface morphology feature associated with the exposed ITER-Grade W as well as features similar to previous EBSD studies of rolled W stock. Thermal desorption spectroscopy (TDS) analysis has identified that pre-existing He bubbles found in the Sandia He-ion exposed samples do alter the D trapping and desorbing behavior in W. The findings from these preliminary characterization studies are presented and discussed in context with results from similar plasma exposure stages at other facilities around the world.

Nanostructured ferritic alloys (NFAs) are new generation materials for use in high temperature energy systems, such as nuclear fission or fusion reactors. However, joining these materials is a concern, as their unique microstructure is destroyed by traditional liquid-state welding methods. The microstructural evolution of a friction stir welded 14YWT NFA was investigated by atom probe tomography, before and after a post-weld heat treatment (PWHT) at 1123K. The particle size, number density, elemental composition, and morphology of the titanium-yttrium-oxygen-enriched nanoclusters (NCs) in the stir and thermally-affected zones were studied and compared with the base metal. No statistical difference in the size of the NCs was observed in any of these conditions. After the PWHT, increases in the number density and the oxygen enrichment in the NCs were observed. Therefore, these new results provide additional supporting evidence that friction stir welding appears to be a viable joining technique for NFAs, as the microstructural parameters of the NCs are not strongly affected, in contrast to traditional welding techniques.

In a magnetic fusion energy (MFE) device, the plasma-facing materials (PFMs) will be subjected to tremendous fluxes of ions, heat, and neutrons. The response of PFMs to the fusion environment is still not well defined. Tungsten metal is the present candidate of choice for PFM applications such as the divertor in ITER. However, tungsten's microstructure will evolve in service, possibly to include recrystallization. How tungsten's response to plasma exposure evolves with changes in microstructure is presently unknown. In this work, we have exposed hot-worked and recrystallized tungsten to an 80 eV helium ion beam at a temperature of 900 °C to fluences of 2 × 1023 or 20 × 1023 He/m2. This resulted in a faceted surface structure at the lower fluence or short but well-developed nanofuzz structure at the higher fluence. There was little difference in the hot-rolled or recrystallized material's near-surface (=50 nm) bubbles at either fluence. At higher fluence and deeper depth, the bubble populations of the hot-rolled and recrystallized were different, the recrystallized being larger and deeper. This may explain previous high-fluence results showing pronounced differences in recrystallized material. The deeper penetration in recrystallized material also implies that grain boundaries are traps, rather than high-diffusivity paths.

Graphite is the moderator material of several Generation IV nuclear reactor concepts, as well as the British Advanced Gas-cooled Reactors (AGR). Porosity can heavily influence the material properties, me- chanical irradiation response, and neutron induced shrinkage or swelling of nuclear-grade graphite. Due to the sub-micron size of several types of pores found in graphite, only a high-resolution imaging tech- nique such as electron tomography are capable of visualizing these features in three dimensions. In this research, we used electron tomography to characterize as-received and neutron irradiated samples of IG-110 nuclear-grade graphite to show for the first time the 3D structure of both native and irradiation- induced nano-cracks. This technique also reveals unique characteristics of graphite such as the structure that surrounds pores and could be used to inform molecular dynamic simulations of irradiated graphite and experimental techniques such as gas-absorption. This research also shows the utility of this technique for the study of other nuclear porous carbon-based materials.

The equilibrium shape of 6H-SiC nanostructures and their surfaces were studied by analyzing nano-void (10 nm) shapes, which were introduced in monocrystalline 6H-SiC by hightemperature neutron irradiation, using transmission electron microscopy. The nano-voids were determined to be irregular icosahedrons truncated with six {1100}, twelve f1103}, one smaller top-basal, and one larger bottom-basal planes, which suggests that {1100} and f1103} are the next stable surface class after the basal planes. The relatively frequent absence of the {1100} surface in the nano-voids indicated that the (1103Þ surface type is energetically rather stable. These non-basal surfaces were found not to be atomically flat due to the creation of nanofacets with half unit-cell height in the c-axis. The {1100} and f1103} surfaces were classified as two and four face types according to their possible nanofacets and surface termination, respectively. We also discuss the surface energy difference between the (1103) and (1103) face types in relation to the energy balance within the equilibrium, but irregular, polyhedron, in which the (1103) surface had double the surface energy of the (1103Þ surface (3900 erg/cm2)

Ultrasonic additive manufacturing (UAM) is a solid-state additive manufacturing process that uses fundamental principles of ultrasonic welding and sequential layering of tapes to fabricate complex three-dimensional (3-D) components. One of the factors limiting the use of this technology is the poor tensile strength along the z-axis. Recent work has demonstrated the improvement of the z-axis properties after post-processing treatments. The abnormally high stability of the grains at the interface during post-weld heat treatments is, however, not yet well understood. In this work we use multiscale characterization to understand the stability of the grains during post-weld heat treatments. Aluminum alloy (6061) builds, fabricated using ultrasonic additive manufacturing, were post-weld heat treated at 180, 330 and 580 °C. The grains close to the tape interfaces are stable during post-weld heat treatments at high temperatures (i.e., 580 °C). This is in contrast to rapid grain growth that takes place in the bulk. Transmission electron microscopy and atom-probe tomography display a significant enrichment of oxygen and magnesium near the stable interfaces. Based on the detailed characterization, two mechanisms are proposed and evaluated: nonequilibrium nano-dispersed oxides impeding the grain growth due to grain boundary pinning, or grain boundary segregation of magnesium and oxygen reducing the grain boundary energy.

Tungsten is the primary candidate for plasma-facing materials (PFMs) in a magnetic fusion energy (MFE) devices such as ITER due to its high melting point, excellent erosion resistance, and low sputtering yield. However, tungsten will suffer from heat, neutron irradiation, large flux (1022-1024 He/m2s)- low energy (tens of eV to hundreds of eV) helium and hydrogen ion exposure, as well as microstructure evolution (such as recrystallization). We have exposed hot-rolled and recrystallized tungsten to 65 or 80 eV helium ions with a flux of 0.5 × 1022 or 5 × 1022 He/m2s to total fluence from 0.6 × 1024 to 4 × 1024 He/m2 at a temperature of ≈1100 K. The results show that recrystallized tungsten samples exhibit a roughened surface morphology in certain grains with orientations close to <001>, while hot-rolled tungsten samples still maintain the smooth surface under all helium-ion exposure conditions. All the samples exhibit relatively shallow helium bubble penetration (<100 nm) from cross-section TEM observation. Much deeper helium bubble penetrations were observed in the low flux exposed samples. Under the same flux conditions, the helium bubbles penetrate deeper at higher fluence. The recrystallized samples show deeper bubble distributions compared to hot-rolled samples. While bubble sizes do present distinct distributions between hot-rolled and recrystallized samples with rough or smooth surface, the differences become less pronounced under the higher fluence conditions.

We irradiated four ferritic alloys with energetic Fe and He ions: one castable nanostructured alloy (CNA) containing Ti-W-Ta-carbides, and three nanostructured ferritic alloys (NFAs). The NFAs were: 9Cr containing Y-Ti-O nanoclusters, and two Fe-12Cr-5Al NFAs containing Y-Zr-O or Y-Hf-O clusters. All four were subjected to simultaneous dual-beam Fe + He ion implantation (650 °C, ~50 dpa, ~15 appm He/dpa), simulating fusion-reactor conditions. Examination using scanning/transmission electron microscopy (STEM) revealed high-number-density helium bubbles of ~8 nm, ~1021 m-3 (CNA), and of ~3 nm, 1023 m-3 (NFAs). STEM combined with multivariate statistical analysis data mining suggests that the precipitate-matrix interfaces in all alloys survived ~50 dpa at 650 °C and serve as effective helium trapping sites. All alloys appear viable structural material candidates for fusion or advanced fission energy systems. Among these developmental alloys the NFAs appear to sequester the helium into smaller bubbles and away from the grain boundaries more effectively than the early-generation CNA.

The family of layered carbides and nitrides known as MAX phase ceramics combine many attractive properties of both ceramics and metals due to their nanolaminate crystal structure and are promising potential candidates for application in future nuclear reactors. This investigation examines the effects of energetic heavy ion (5.8 MeV Ni) irradiations on polycrystalline samples of Ti3SiC2, Ti3AlC2, and Ti2AlC. The irradiation conditions consisted of midrange ion doses between 10 and 30 displacements per atom at temperatures of 400 and 700 °C, conditions relevant to application in future nuclear reactors and a relatively un-explored regime for this new class of materials. Following irradiation, a comprehensive analysis of radiation response properties was compiled using grazing incidence X-ray diffraction (XRD), nanoindentation, scanning electron microcopy (SEM), and transmission electron microscopy (TEM). In all cases, XRD and TEM analyses confirm the materials remain fully crystalline although the intense atomic collisions induce significant damage and disorder into the layered crystalline lattice. X-ray diffraction and nanoindentation show this damage is manifest in anisotropic swelling and hardening at all conditions and in all materials, with the aluminum based MAX phase exhibiting significantly more damage than their silicon counterpart. In all three materials there is little damage dependence on dose, suggesting saturation of radiation damage at levels below 10 displacements per atom, and significantly less retained damage at higher temperatures, suggesting radiation defect annealing. SEM surface analysis showed significant grain boundary cracking and loss of damage tolerance properties in the aluminum-based MAX phase irradiated at 400 °C, but not in the silicon counterpart. TEM analysis of select samples suggest that interstitials are highly mobile while vacancies are immobile and that all three materials are in the so-called point defect swelling regime between 400 and 700 °C. All results are consistent with previous work involving traditional and MAX phase ceramics. Results show the aluminum MAX phases are not fit for application near 400 °C and that the silicon MAX phase is more damage tolerant at 400–700 °C.

Precipitates of the Laves and B2 phases were engineered in a newly-designed advanced ferritic alloy. Under creep test at 650 °C with 120 MPa, the material showed a steady-state minimum creep rate of 1 × 10−4 h−1, about one order of magnitude lower than T91. Microstructural characterization of the ferritic alloy revealed primarily ductile and partially brittle fractures after the creep test. Coarse Laves phase (~ 1 µm) was observed associating with the brittle fracture, resulting in reduced creep ductility. However, fine Laves phase precipitates (~ 100 nm) helped the dimple-ductile fracture and strengthened the material through impeding the motion of dislocations and boundaries. Unlike the B2 precipitates remained coherent exerting the classic Orowan bypassing mechanism at the brittle location, some of the B2 precipitates at the ductile location became incoherent and can develop an attractive interaction with dislocations. This coherency change of B2 precipitates, together with the nucleation of ultrafine (~ 40 nm) Laves phase precipitates during the creep test, would compensate for the coarsening-induced loss of Orowan strengthening of coherent B2 precipitates.

We observed that ß-SiC, neutron irradiated to 9?dpa (displacements per atom) at ˜1440?°C, began transforming to a-SiC, with radiation-induced Frank dislocation loops serving as the apparent nucleation sites. 1440?°C is a far lower temperature than usual ß???a phase transformations in SiC. SiC is considered for applications in advanced nuclear systems, as well as for electronic or spintronic applications requiring ion irradiation processing. ß-SiC, preferred for nuclear applications, is metastable and undergoes a phase transformation at high temperatures (typically 2000?°C and above). Nuclear reactor concepts are not expected to reach the very high temperatures for thermal transformation. However, our results indicate incipient ß???a phase transformation, in the form of small (~5–10?nm) pockets of a-SiC forming in the ß matrix. In service transformation could degrade structural stability and fuel integrity for SiC-based materials operated in this regime. However, engineering this transformation deliberately using ion irradiation could enable new electronic applications.

This work reports the first mechanical properties of Ti3AlC2-Ti5Al2C3 materials neutron irradiated at ∼400, 630 and 700 °C at a fluence of 2 × 1025 n m−2 (E > 0.1 MeV) or a displacement dose of ∼2 dpa. After irradiation at ∼400 °C, anisotropic swelling and loss of 90% flexural strength was observed. After irradiation at ∼630–700 °C, properties were unchanged. Microcracking and kinking-delamination had occurred during irradiation at ∼630–700 °C. Further examination showed no cavities in Ti3AlC2 after irradiation at ∼630 °C, and MX and A lamellae were preserved. However, disturbance of (0004) reflections corresponding to M-A layers was observed, and the number density of line/planar defects was ∼1023 m−3 of size 5–10 nm. HAADF identified these defects as antisite TiAl atoms. Ti3AlC2-Ti5Al2C3 shows abrupt dynamic recovery of A-layers from ∼630 °C, but a higher temperature appears necessary for full recovery.

Three types of microstructures, i.e., tempered-martensite (TM), ferrite (F), and dual-phase (TM + F), were developed in a castable nanostructured alloy that favors a high density of nanoprecipitates compared with the precipitates in current reduced-activation ferritic-martensitic steels. The effect of the distinct microstructures on tensile properties, Charpy impact toughness, and thermal helium desorption behavior was investigated with the full TM structure as a reference. The results indicated that the F domain in the TM + F structure governed the strength and slightly impaired the impact toughness. The full F structure exhibited the highest strength without compromising ductility, but it noticeably diminished impact toughness. All microstructures had a dominant helium desorption peak at ∼1070 °C. The higher density of nanoprecipitates and complex boundaries and dislocations in the TM + F structure enhanced the secondary helium desorption peak and extended the shoulder peak, in contrast to the full TM structure with an enlarged desorption peak associated with the ferrite-to-austenite transformation at ∼810–850 °C and the full F structure with a dominant desorption peak related to bubble migration at ∼1070 °C. These results suggest that components fabricated from functionally graded microstructures could be engineered to exploit the advantages of different microstructures for demanding application requirements.

Nanotendril “fuzz” will grow under He bombardment under tokamak-relevant conditions on tungsten plasma-facing materials in a magnetic fusion energy device. We have grown tungsten nanotendrils at low (50?eV) and high (12?keV) He bombardment energy, in the range 900–1000?°C, and characterized them using electron microscopy. Low energy tendrils are finer (~22?nm diameter) than high-energy tendrils (~176?nm diameter), and low-energy tendrils have a smoother surface than high-energy tendrils. Cavities were omnipresent and typically ~5–10?nm in size. Oxygen was present at tendril surfaces, but tendrils were all BCC tungsten metal. Electron diffraction measured tendril growth axes and grain boundary angle/axis pairs; no preferential growth axes or angle/axis pairs were observed, and low-energy fuzz grain boundaries tended to be high angle; high energy tendril grain boundaries were not observed. We speculate that the strong tendency to high-angle grain boundaries in the low-energy tendrils implies that as the tendrils twist or bend, strain must accumulate until nucleation of a grain boundary is favorable compared to further lattice rotation. The high-energy tendrils consisted of very large (>100?nm) grains compared to the tendril size, so the nature of the high energy irradiation must enable faster growth with less lattice rotation.

Tungsten is the primary candidate for plasma-facing materials in the critical regions of a magnetic fusion energy (MFE) device, such as a tokamak. However, tungsten (and many other metals) grow copious nanotendril "fuzz" when exposed to fusion-relevant He-bearing plasmas, and the impact of this fuzz on reactor operation is still unknown. Further, the mechanism of fuzz growth is also poorly understood at present. In order to experimentally probe the growth mechanisms, we developed a new sample preparation technique to examine the interface of individual nanotendril roots with the substrate (substrate/tendril interface) using electron microscopy and measured the grain boundary character distributions (GBCDs) of substrate/tendril interfaces, the GBs within individual tendrils (tendril/tendril interfaces), and GBs within the substrate (substrate/substrate interfaces). We find the substrate/tendril and tendril/tendril GBCDs are essentially indistinguishable, but very different from the substrate/substrate GBCDs, which implies that tendril growth periodically forms new GBs at the substrate/tendril interface, and these GBs are pushed upward to become tendril/tendril interfaces within the fuzz mat. These results will help guide future computational modelling studies to elucidate the details of the growth mechanisms.

Plasma-facing materials in the divertor of a magnetic fusion reactor have to tolerate steady state plasma heat fluxes in the range of 10 MW/m2 for 107 s, in addition to fusion neutron fluences, which can damage the plasma-facing materials to high displacements per atom (dpa) of 50 dpa. Materials solutions needed for the plasma-facing components are yet to be developed and tested. The material plasma exposure experiment (MPEX) is a newly proposed steady state linear plasma device designed to deliver the necessary plasma heat flux to a target for testing, including the capability to expose a priori neutron-damaged material samples to those plasmas. The requirements of the plasma source needed to deliver the required heat flux are being developed on the Proto-MPEX device which is a linear high-intensity radio-frequency (RF) plasma source that combines a high-density helicon plasma generator with electron- and ion-heating sections. The device is being used to study the physics of heating overdense plasmas in a linear configuration. The helicon plasma is operated at 13.56 MHz with RF power levels up to 120 kW. Microwaves at 28 GHz (30 kW) are coupled to the electrons in the overdense helicon plasma via electron Bernstein waves and ion cyclotron heating at 7–9 MHz (30 kW) is via a magnetic beach approach. High plasma densities >6 1019/m3 have been produced in deuterium, with electron temperatures that can range from 2 to >10 eV. Operation with on-axis magnetic field strengths between 0.6 and 1.4 T is typical. The plasma heat flux delivered to a target can be >10 MW/m2 , depending on the operating conditions. An initial plasma material interaction experiment with a thin tungsten target exposed to this high heat flux in a predominantly helium plasma showed helium bubble formation near the surface, with no indication of source impurity contamination on the target.

Thirteen austenitic stainless steels, nickel-base alloys, and ferritic alloys were irradiated using 2 MeV protons at 360 °C to a damage level of 2.5 displacements per atom (dpa). Comprehensive microstructural characterization was performed for irradiation-induced features, including dislocation loops, voids, precipitates, and radiation induced segregation (RIS). Dislocation loops formed in all alloys except 14YWT, while voids were observed in alloys 316 L, 310, C22, and 14YWT. Irradiation-induced formation of γ′ precipitates was observed in alloys 316 L, 310, 800, and 690; the irradiation-enhanced, long-range ordered Ni2Cr phase (Pt2Mo-type) was observed in alloys 690, C22, 625, 625Plus, 625DA, and 725; and G-phase was observed in alloy T92. No irradiation-induced precipitates were observed in alloys X750, 718 or 14YWT. Precipitation of the γ′ phase can be understood through segregation and clustering of Si, Al, and Ti. Overall, austenitic stainless steels are generally susceptible to irradiation damage in the form of loops, voids, precipitates, and RIS. Ni-base alloys have this same type of dislocation loops and RIS behaviors but are more resistant to void swelling. Ferritic alloys showed better resistance to loop formation, void swelling and irradiation-induced precipitation. From the degree of irradiation-induced microstructural change, alloy T92 was identified as the most radiation resistant among these alloys.

Modern electron microscopy techniques were used to conduct a thorough study of an irradiated urania fuel pellet microstructure to attempt at an understanding of high burnup structure formation in this material. The fuel was irradiated at low power to high burnups in a light water reactor, proving ideal for this purpose. Examination of grain size and orientation with strict spatial selectivity across the fuel pellet radius allowed for capturing the progression of the restructuring process, from its onset to full completion. Based on this information, the polygonization mechanism was shown to be responsible for restructuring, involving formation of low-angle grain boundaries with their initiation occurring at the original high-angle grain boundaries of the as-fabricated pellet and at the gas bubble-matrix interfaces. The low-angle character of boundaries between the subdivided grains disappeared in the fully developed high burnup structure, likely due to creep deformation in the pellet.

Nano-scale Ni-Mn-Si-rich precipitates formed in a reactor pressure vessel steel under high neutron fluence have been characterized using highly complimentary atom probe tomography (APT) and scanning transmission electron microscopy with energy dispersive spectroscopy (STEM-EDS) combined with STEM-EDS modeling. Using these techniques in a synergistic manner to overcome the well-known trajectory aberrations in APT data, the average upper limit Fe concentration within the precipitates was found to be ~6 at.%. Using this knowledge, accurate compositions of the precipitates was determined and it was found that the spread of precipitate compositions was large, but mostly centered around the G2-and G-phases. The use of STEM-EDS also allowed for larger areas to be examined, and segregation of minor solutes was observed to occur on grain boundaries, along with Ni-Mn-Si-rich precipitates that were smaller in size than those in the matrix. Solute segregation at the grain boundaries is proposed to occur through a radiation induced segregation or radiation enhanced diffusion mechanism due to the presence of a denuded zone about the grain boundary. It is also proposed that the reduced precipitate size at the grain boundaries is due to the structure of the grain boundary. The lack of Ni-Mn-Si precipitates observed in larger Mo-rich precipitates is also discussed, and the absence of the minor solutes required to form the Ni-Mn-Si precipitates results in the lack of nucleation. This is in contrast to cementite phases in which Ni-Mn-Si precipitates have been observed to have formed. It was also determined through this work that the exclusion of all the Fe ions during atom probe analysis is a reasonable approximation.

New candidate materials for GenIV or fusion nuclear energy systems, e.g., nanostructured ferritic alloys, are distinguished from older-generation nuclear materials by much smaller feature sizes and complex local nanochemistry and crystallography. Established and perspective nuclear materials, e.g. reactor pressure vessel steels or plasma-facing tungsten, also form small nanoscale structures under in-reactor service. Here, we discuss recent advances in materials characterization – high-efficiency X-ray mapping combined with datamining; transmission Kikuchi diffraction; and atom probe tomography – that make it possible to quantitatively characterize these nanoscale structures in unprecedented detail, which enables advances in understanding and modelling of radiation service and degradation.