Sample spinning to mitigate polarization artifact and interstitial-vacancy imbalance in ion-beam irradiation

Abstract

Accelerator-based ion-beam irradiation has been widely used to mimic the effects of neutron radiation damage in nuclear reactors. However, ion radiation is most often monodisperse in the incoming ions’ momentum direction, leading to excessive polarization in defect distribution, while the scattering under neutron irradiation is often more isotropic and has less radiation-induced polarization. Mitigation of the excess-polarization as well as the damage non-uniformity artifact might be crucial for making the simulation of neutron radiation by ion-beam radiation more realistic. In this work, a general radiation polarization theory in treating radiation as external polar stimuli is established to understand the natural material responses in different contexts, and the possibility to correct the defect polarization artifact in ion-beam irradiation. Inspired by Magic Angle Spinning in Nuclear Magnetic Resonance, we present a precise sample spinning strategy to reduce the point-defect imbalance effect in ion-beam irradiation. It can be seen that with optimized surface inclination angle and the axis of sample rotation, the vacancy-interstitial population imbalance, as well as the damage profile non-uniformity in a designated region in the target are both reduced. It is estimated that sample spinning frequency on the order of kHz should be sufficient to scramble the ion momentum monodispersity for commonly taken ion fluxes and dose rates, which is experimentally feasible.

Materials for Sustainable Nuclear Energy: A European Strategic Research and Innovation Agenda for All Reactor Generations

“…Moreover, contrary to bulk-techniques for mechanical testing, no good practices are fully established for those applicable to charged-particle irradiated specimens, e.g., nanoindentation [171,198]. Finally, serious issues of transferability to neutrons exist, because charged particles are injected at higher dose rate, have different energy spectrum, generally produce damage gradients in a limited penetration thickness and often inject foreign species, including impurities [168,170,[199][200][201][202][203][204][205]. Thus, theoretical and modelling work to ensure transferability, and a clear definition of suitable protocols, though already started [206], is still needed in order for charged particles to become fully usable screening tools.…”

Machine learning in nuclear materials research

“…We expect data-based ML to be as commonplace in the 2020s as curve-fitting with Excel spreadsheets today (see an example with SRIM/IM3D data in Ref. 3 ). This style of research will become ever more broadly used by experimentalists.…”

“…The departures from thermodynamic equilibrium in nuclear materials, as characterized by the volumetric energy dissipation rate and defect evolution activities, can reach extremes not seen in most materials applications, and so do the associated timedependent complexities in nuclear-energy applications: chemical due to transmutation and corrosion, microstructural due to radiation damage and gradients in temperature and other thermodynamic quantities, even electronic-structure complexities due to the excited-states behavior such as radiolysis in coolant and excitons in nuclear detector materials. The extreme service environments also make experimental data acquisition challenging, and small-scale laboratory tests whose intent is to mimic the in-core conditions (such as using ion accelerators to mimic neutron tests 3 ) may in reality still depart significantly in the actual microstructural physics that are sensitive to the boundary conditions.…”

“…For example, recently, machine learned interatomic potentials based on ab initio energies have been developed for many nuclear related materials, including properties of FLiBe 41 and chloride 42 molten salts, radiation damage in W 43 , and He bubble effects in the He-Be-W system 44 . Based on binary-collision Monte Carlo simulations, a fundamental study of radiation polarization and interstitial-vacancy imbalance in ion-beam irradiation was published where an optimized sample spinning strategy was designed to minimize the deviation between neutron exposure and ion-beam exposure based on neural-network representation of the damage profiles 45 .…”

“…62 Accelerated radiation tests by ions instead of neutrons, by definition, would impart a much higher dose rate (dpa/s) vis-à-vis the material's inherent relaxation processes. Also, with accelerator-based ion-beam radiations, the beam's monodisperse ion momenta cause "excess polarization" artifacts in the vacancy-interstitial imbalance 45 . And the very small damaged region (usually microns) in ion-beam tests also causes "size effects" in mechanical properties 63 that makes the resulting mechanical properties quite different from those of real centimeter-sized samples exposed to neutrons with much more uniform damage.…”