Nanoindentation experiments performed in 5 and 18 μm thick vapor deposited polycrystalline lithium films at 31 °C reveal the mean pressure lithium can support is strongly dependent on length scale and strain rate. At the smallest length scales (indentation depths of 40 nm), the mean pressure lithium can support increases from ∼23 to 175 MPa as the indentation strain rate increases from 0.195 to 1.364 s−1. Furthermore, these pressures are ∼46–350 times higher than the nominal yield strength of bulk polycrystalline lithium. The length scale and strain rate dependent hardness is rationalized using slightly modified forms of the Nabarro–Herring and Harper–Dorn creep mechanisms. Load–displacement curves suggest a stress and length-scale dependent transition from diffusion to dislocation-mediated flow. Collectively, these experimental observations shed significant new light on the mechanical behavior of lithium at the length scale of defects existing at the lithium/solid electrolyte interface.

Nanoindentation experiments performed in high-purity vapor deposited lithium films at 31 °C reveal a strain rate and length scale dependence in the stress at which pop-in type events signal an abrupt transition from diffusion to dislocation-mediated flow. The stress level at which the transition to dislocation-mediated flow occurs varies with the strain rate and ranges from 88 to 208 times larger than the nominal yield strength of bulk, polycrystalline lithium. Variation in the indentation strain rate reveals the relationship between the stress required to initiate the transition and the length scale at which the transition occurs follows the power-law relation, hardness × depth1.17 = 1.545 N/m0.83, where the magnitude of the exponent and constant reflect the defect structure of the film. A rationalization of the transition is provided through direct comparisons between the measured cumulative distribution function (CDF) and the CDF hypothesized for the activation of a Frank–Read source.

Nanoindentation has been used to measure the elastic modulus of 5 and 18 μm thick high-purity vapor deposited polycrystalline lithium films at 31 °C. Over indentation depths ranging from 150 to 1100 nm, the modulus is found to vary with film thickness from 9.8 GPa ± 11.9% to 8.2 GPa ± 14.5%. These results are well within the range of lithium's orientation dependent elastic modulus, which spans approximately 3.1 to 21.4 GPa. The measured values may also indicate (111) and (100) texture for the 5 and 18 μm thick films, respectively. The potential effects of pileup and surface contamination are found to be negligible if any at all. Small but discernible changes in damping capability near the free surface may provide insight into the subsurface defect structure and the potential for localized heating. Numerous experimental challenges are addressed and key metrics are used to validate the measured elastic modulus.

A simple stochastic model is developed to determine the pop-in load and maximum shear stress at pop-in in nanoindentation experiments conducted with spherical indenters that accounts for recent experimental observations of a dependence of these parameters on the indenter radius. The model incorporates two separate mechanisms: pop-in due to nucleation of dislocations in dislocation-free regions and pop-in by activation of preexisting dislocations. Two different types of randomness are used to model the stochastic behavior, which include randomness in the spatial location of the dislocations beneath the indenter and randomness in the orientation of the dislocations, i.e., randomness in the stress needed to activate them. In addition to correctly predicting the experimentally observed average maximum shear stress at pop-in, the model also correctly describes the scatter in pop-in loads and how it varies with indenter radius. Monte Carlo simulations are used to validate the model and visualize the scatter expected for a limited number of tests.

In-situ tensile tests have been performed in a dual beam focused ion beam and scanning electron microscope on as-grown and prestrained single-crystal molybdenum-alloy (Mo-alloy) fibers. The fibers had approximately square cross sections with submicron edge lengths and gauge lengths in the range of 9–41 μm. In contrast to previously observed yield strengths near the theoretical strength of 10 GPa in compression tests of ∼1–3-μm long pillars made from similar Mo-alloy single crystals, a wide scatter of yield strengths between 1 and 10 GPa was observed in the as-grown fibers tested in tension. Deformation was dominated by inhomogeneous plastic events, sometimes including the formation of Lüders bands. In contrast, highly prestrained fibers exhibited stable plastic flow, significantly lower yield strengths of ∼1 GPa, and stress–strain behavior very similar to that in compression. A simple, statistical model incorporating the measured dislocation densities is developed to explain why the tension and compression results for the as-grown fibers are different.

Extended abstract of a paper presented at Microscopy and Microanalysis 2011 in Nashville, Tennessee, USA, August 7–August 11, 2011.