The project will use the FIB and TEM/STEM “hot” facilities at IMCL to prepare and examine exceptionally thin lamella for determination of the location and physical relationship of hydrogen and helium distributions in neutron-irradiated 316 stainless steel that served as a flux-thimble tube in a commercial PWR to ~76 dpa at ~320°C. These specimens are currently available at Texas A&M University in the form of mechanically and electro-polished quarter-ring tube segments supplied by Paula Freyer of Westinghouse and Frank Garner of Radiation Effects Consulting. The specimens were cut from a tube of 7.7 mm diameter with 1.2 mm thickness, with the quartering cuts yielding ~65° segments instead of 90° segments. After final polishing the specimen thickness is ~0.4 mm, with a gamma activity reading ~17 mR/hr at 6” and ~5 mR/hr at 12 inches, primarily arising from Co-60.
Microscopy measurements conducted in 2009 at PNNL on essentially identical specimens from the same reactor at ~70 dpa/315°C shows nano-bubbles 1-3 nm in diameter at a density of 1.6E23 m-3, resulting from measured helium and hydrogen concentrations of ~600 and ~2400 appm, respectively. Various recent atomistic modeling studies suggest that bubbles formed in various metals during neutron irradiation have a high-density helium core, surrounded by a hydrogen-enriched metal matrix interface or “halo” surrounding helium bubbles.
High-resolution electron energy loss spectroscopy (EELS) technique coupled with scanning transmission electron microscopy (STEM) will be employed as the major tool. There are two major challenges to overcome. First, multiple very-thin lamellae must be produced to allow clear images of many single, non-overlapping cavities in a minimum matrix, requiring significant effort in specimen preparation. Second, both the helium and hydrogen edge occurs within the bulk plasmon signal of the matrix, making background subtraction rather complex. However, recent soon-to-be published EELS studies conducted at Canadian Nuclear Laboratories on Inconel X750 from CANDU garter springs indicate that the halo effect appears to be real, but the neutron-flux-spectra and non-aqueous environment of the CANDU X750 produce a much higher, overlapping bubble density and a very low retained H/He ratio of ~0.1, leading to some remaining doubt that the hydrogen signal was not a plasmon-artifact. The 316 flux-thimble tube specimens examined in earlier studies had a much higher H/He ratio of ~4 and a lower density of more easily-separated bubbles at 70 dpa/315°C. Somewhat larger values expected at 76 dpa/320°C, allowing a much more conclusive determination of the halo presence and its characteristics compared to that derived in the X750 effort.