**2. Materials and Methods**

A Zr-based Vit 105 MG (Zr52.5Cu17.9Ni14.6Al10Ti5) plate with a thickness of 2 mm was produced by suction casting into a Cu-mold. The as-cast samples were cut into thin samples by electrical discharge machining. On each of the samples, the as-cut surface was mechanically polished while the as-cast surface was kept intact for indentation tests. The final sample had an area of ~15 mm<sup>2</sup> and a thickness of 1 mm.

In contrast to previous studies [11], where hydrogen was added to the alloy by processing the melt in a H2-containing gas, hydrogen charging of samples was performed after quenching using a cathodic charging technique in a 0.5 mol/L sulfuric acid solution. The samples were charged at a current density ~10 mA/cm<sup>2</sup> (lower than the critical current density for the formation of hydride [25]) for various charging times in order to control the dissolved hydrogen content (*c*H). Hydrogen contents up to *c*H > 0.5 H/M (hydrogen-to-metal ratio) were investigated, with most studies focusing on *c*H ≤ 0.29 H/M where no specimen cracking was observed. The structure of the samples before and after hydrogen charging was investigated with X-ray di ffraction (Cu-K α line).

Nanoindentation tests were conducted at room temperature using a Nano Indenter G200 (formerly MTS; now KLA-Tencor, Milpitas, CA, USA). Indentation with a Berkovich indenter in the standard XP indentation head was performed on the samples to measure hardness *H* and Young's modulus *E* using the continuous sti ffness measurement technique [26]. Reduced Young's modulus *Er* is calculated using the slope of the unloading curve and precalibrated contact area of the indenter. The modulus of the sample *Es* can be derived from the equation:

$$11/E\_I = \begin{pmatrix} 1 \ -\nu\_i^2 \end{pmatrix} / E\_i + \begin{pmatrix} 1 \ -\nu\_s^2 \end{pmatrix} / E\_{\text{s}\prime} \tag{1}$$

where Young's modulus of the tip *Ei* = 1140 GPa. Poisson's ratios of the diamond tip and the sample are ν2 *i* = 0.07 and ν2 *s* = 0.37, respectively. The indentation depth reaches 2 μm in order to reduce the effect of surface roughness.

A spherical diamond tip with a diameter of *R* = 650 nm was used in the DCM-1 indenter head. The tip radius was calibrated on a fused silica reference sample. Arrays of 12 × 12 indents which covered an area of 33 × 33 μm<sup>2</sup> were measured with a constant loading rate of 0.1 mN/s to a maximum load of 1 mN corresponding to a displacement of ~50 nm. Displacement bursts, which are commonly referred to as "pop-ins", were visible in the load-displacement curves and automatically identified from displacement-rate data using a MATLAB script (R2013b, MathWorks, Natick, MA, USA). See details in [27]. In order to measure *c*H, the average hydrogen content of the sample, melt extraction was carried out on the samples after indentation tests using a Hydrogen analyzer (G8 Galileo, Bruker, Billerica, MA, USA). The hydrogen analyzer was calibrated by a commercial standard hydrogen containing

sample (501-529, Leco, St Joseph, MI, USA). All the indentation tests and melt extraction analysis were performed within 72 h after the samples were charged with hydrogen.
