*4.1. Comparison with Data on He bubble Nucleation in Pd and PdT0.6*

Although very little microscopy work has been done on fcc metals implanted with He at cryogenic temperatures, He bubble formation has been studied in aged Pd tritide, PdT0.6. Tritium decay does not induce displacement damage in the lattice, so the vacancy concentration is limited to thermal vacancies, although H can stabilize vacancies, and formation of β-phase hydride can cause microstructural changes [18]. Bubbles are at equilibrium when the He pressure inside the bubble equals the surface tension of the host material, *p* = 2γ/*r*, where *p* is the He pressure, γ is the surface energy of the host material, and *r* is the bubble radius [19]. Bubbles formed at room temperature in PdT0.6 are typically highly over-pressurized. He bubbles reach 1–2 nm in diameter in the first 8 months (1.34 at.% He) of storage. Bubble density estimates vary widely due to film thickness measurement error, overlapping bubbles, and the possible presence of bubble sizes near or below the resolution limit of the microscope, but values of 1017–1019 bubbles/cm<sup>3</sup> are reported for aging times of 2–8 months [7–9]. If a thickness of 50 nm is assumed for the samples implanted in this work, the density varied from 1017–1018 bubbles/cm3 with implantation dose from 2 <sup>×</sup> 1016 to 9 <sup>×</sup> 1016 ions/cm2. Bubble diameter did not vary with implantation dose in this work and is similar to that observed in aged PdT0.6. Nuclear Magnetic Resonance (NMR) has been used to measure the phase and pressure of He bubbles in aged PdT0.6. After annealing samples aged for 1 year, solid He diffusion was observed below −73 ◦C, and melting was observed from −73 ◦C to 7 ◦C, with corresponding bubble pressures of 6–11 GPa, and an average density of 120 He/nm<sup>3</sup> within a bubble [20]. Aging for 8 years, followed by multiple deuterium exchanges to remove the tritium, resulted in an average density of 90 He/nm3 and the presumed coexistence of liquid and solid phases from −230 ◦C to −133 ◦C [21].

#### *4.2. Comparison with Theory of He Bubble Nucleation and Growth at Cryogenic Temperatures*

Several mechanisms could result in a lack of bubble growth with implantation time. Low He diffusivity at cryogenic temperature will contribute to a lower nucleation and growth rate. Low diffusivity may promote a higher bubble nucleation rate, but a lower bubble growth rate due to He becoming trapped very near its implantation site. Bubble diameter remained constant at 1.2 nm after the initial observation of He bubbles at a fluence of 2.35 <sup>×</sup> 1016 ions/cm<sup>2</sup> (6 at.%), but bubble density increased with implantation dose. Since the He implantation profile, shown in Figure 1b, results in lower concentrations near the surfaces, nucleation will take longer in these regions than at the center of the foil where the He concentration is highest. TEM imaging captures all the material within the thickness of the sample, which could result in a visible increase in bubble density as nucleation occurs first in the higher concentration, and subsequently in the lower concentration regions. This may contribute to an experimentally measured increase in areal bubble density.

At cryogenic temperatures, the low thermal vacancy concentration prohibits cavity growth by absorption of thermal vacancies, which cause rapid cavity growth at elevated temperatures. If the mobility of irradiation- or hydrogen-induced vacancies is low, bubble growth can only occur from dislocation loop punching. Assuming a Pd vacancy migration value of 0.63 eV [22], Pd vacancies are not mobile over the timescale of these experiments at −100 ◦C. Wolfer [23] has succinctly summarized He bubble growth in metals as a function of homologous temperature, T*h*, and pressure. In the case of an isolated single bubble at temperatures below T*<sup>h</sup>* = 0.25, theory indicates that the bubble pressure is high enough for loop punching only at μ/5, where μ is the shear modulus of the material [24]. Below this pressure at T*<sup>h</sup>* = 0.25, bubble growth is not expected to occur. In Pd, μ = 42 GPa, making the pressure required for loop punching from an isolated single bubble 8.4 GPa. The NMR measurements discussed above [20] found bubble pressures of 6–11 GPa in the −173 ◦C to 77 ◦C range in 1 year old PdT0.62, very close to the threshold for loop punching.

When theory is applied to the case of a He bubble array, instead of an isolated bubble, the pressure required for loop punching increases to about 50% of the shear modulus [25]. Bubbles with radii less or equal to ~5*b*, where *b* is the Burgers vector of a prismatic dislocation loop, require a high enough He density to create bubble pressures sufficient for loop punching, independent of the bubble density. As bubbles grow by loop punching, the accumulation of loop debris in the regions between bubbles exerts an increasing and opposing force to the subsequent formation of loops, increasing the pressure required for loop punching dramatically. Thus, during the initial He accumulation period at cryogenic temperatures, one might expect slow growth by loop punching, but as the bubble density increases and the inter-bubble spacing decreases, the pressure required for loop punching likely becomes too high for additional growth to occur. This mechanism may have caused bubble growth to cease in this experiment, while bubble areal density continued to increase.

As the bubble pressure continues to increase beyond the window where dislocation loop punching is viable, inter-bubble fracturing may occur, possibly leading to blister formation and He release [23]. Blister formation has previously been observed after implanting bulk Pd with 300 keV He to 1 <sup>×</sup> <sup>10</sup><sup>18</sup> ions/cm<sup>2</sup> (70 at.% He at the peak) at <sup>−</sup><sup>180</sup> ◦C, a much higher He concentration that the total implanted dose in this work [1]. Theory suggests that, particularly below a bubble density of 3 <sup>×</sup> <sup>10</sup><sup>18</sup> bubbles/cm3, equilibration of the chemical potentials for gas atoms in the bubble and in interstitial solution could result in He diffusing freely throughout the solid without being trapped inside a bubble, causing rapid gas release once a critical concentration is reached [25].

### *4.3. Hydrogen-helium Interactions*

By performing in-situ He implantation in an ETEM, this work is uniquely suited to explore the interactions of H and He in a model fcc system. Point defects are introduced in the lattice during He implantation (see Figure 1b for the dpa profile). Hydrogen is known to interact with such defects in metals [26–28], which may have influenced the bubble nucleation and growth rates observed in this work. The binding energies of H to defects in Pd are lower than many other metals [29]. Experimentally determined binding energies of H to Pd defects are about: 0.15 eV (self-interstitial), 0.23 eV (vacancy), and 0.29 eV (He bubble) [27–29]. Hydrogen is expected to have been strongly bound to these defects in the present work, which was done at −100 ◦C (0.015 eV), and could increase the size of He clusters. In fcc metals, up to six H atoms can occupy a monovacancy [29], which could influence the trapping kinetics of He atoms to vacancies. Additionally, the presence of H in interstitial sites may influence the diffusion of interstitial He through the lattice, and therefore influence nucleation and growth kinetics. More simulation work is needed to verify potential effects of H on He bubble nucleation and growth.

#### **5. Conclusions**

Palladium metal was implanted with 10 keV 4He in-situ, at cryogenic temperature, in a H2 environment. No lattice expansion indicating β-phase hydride formation was observed. He bubbles 1.2 nm in diameter were observed to nucleate after 6 at.% He. Bubble size did not change with implantation time, but bubble density did increase. These initial experiments highlight the strength of the MIAMI-2 facility for in-situ TEM exploration of H2 interaction with He bubbles at various temperature extremes.

This preliminary work has highlighted the new combination of extreme environments (cryogenics, gas implantation, and reactive gas exposure) that can be explored during direct real-time observation within a TEM. Further work is needed to fully understand these initial observations. This future work would include comparison between in-situ He implantations in the presence and absence of H2 at the same temperature, for both ambient and low temperature, to deduce the effects of H2 and temperature on bubble formation, as well as development of methods to ensure hydride formation in the ETEM. This study points to a new multidimensional stressor approach to in-situ TEM experiments that permits greater understanding of the response to complex environments by materials.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/1996-1944/12/16/2618/s1, Figure S1: Under-focus image showing He bubbles during implantation, Figure S2: Over-focus image showing He bubbles during implantation.

**Author Contributions:** C.A.T. performed preliminary experiments at Sandia, aided in experimental planning, analyzed the data, and wrote the manuscript. J.A.H., G.G., E.A., S.B., and K.H. aided in experimental planning and performed the experiments. A.M. aided in data analysis and interpretation. D.B.R. aided in experimental planning and interpretation. J.D.S. prepared TEM samples and provided interpretation.

**Funding:** This work was performed, in part, at the Center for Integrated Nanotechnologies, an Office of Science User Facility operated for the U.S. Department of Energy (DOE) Office of Science. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International, Inc., for the U.S. DOE's National Nuclear Security Administration under contract DE-NA-0003525. The views expressed in the article do not necessarily represent the views of the U.S. DOE or the United States Government.

**Acknowledgments:** The authors would like to thank Warren York for his time in preparing TEM samples, and Norm Bartelt, Doug Medlin, and Trevor Clark for thoughtful discussion.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
