Fig. 1. Direct and indirect bandgaps of Ge as a function of 〈100〉 uniaxial tensile strain.

Fig. 2. (a) Scanning electron micrograph (SEM) of a suspended strained Ge wire attached to suspended Ge pads. (b) Zoomed-in SEM of the region indicated in (a). (c) COMSOL simulation of the strain distribution in a suspended Ge wire attached to suspended Ge pads.

Fig. 3. (a) Tilted scanning electron micrograph (SEM) of a substrate-adhered strained Ge wire and pads. (b) Zoomed-in tilted SEM of the edge of a pad, showing deflection due to stiction. (c) Zoomed-in tilted SEM of the substrate-adhered Ge wire. (d) Zoomed-in titled SEM showing a substrate-adhered Ge wire and pads next to an overhang of suspended Ge.

Fig. 4. Dependence of the Raman peak redshift on the excitation laser power for both suspended and substrate-adhered Ge wires with ∼4% strain. Both curves are the difference between the wire’s Raman peak location and the Raman peak location of relaxed bulk Ge.

Fig. 5. Raman spectra for suspended Ge wires with various pad dimensions, showing uniaxial strains from 0.79% to 5.71%.

Fig. 6. Photoluminescence (PL) from suspended Ge wires with various strains.

Fig. 7. Theoretically calculated fraction of electrons in the direct valley as a function of 〈100〉 uniaxial tensile strain, shown at several temperatures, plotted on a (a) linear scale and (b) logarithmic scale.

Fig. 8. Theoretical internal quantum efficiency of a Ge LED as a function of strain for various defect-limited minority carrier lifetimes at room temperature (300 K), assuming 1019  cm−3 n-type doping and 1017  cm−3 carrier injection.

Fig. 9. Theoretically calculated threshold current density for a Ge laser as a function of strain.