Search

We experimentally demonstrate the difference between a ridge polariton laser, and a conventional edge-emitting ridge laser operating under electron-hole population inversion. The horizontal laser cavities are 20 -- 60 $\mu$m long GaN etched ridge structures with vertical Bragg reflectors. We investigate the laser threshold under optical pumping and assess quantitatively the effect of a varying optically-pumped length. The laser effect is achieved for an exciton reservoir length of just 15% of the cavity length, which would not be possible in a conventional ridge laser, with an inversion-less polaritonic gain about 10 times larger than in equivalent GaN lasers. The modelling of the cavity free spectral range demonstrates the polaritonic nature of the modes.

We investigate the demonstration and quantification of the strong coupling between excitons and guided photons in a GaN slab waveguide. The dispersions of waveguide polaritons are measured from T=6~K to 300~K through gratings. They are carefully analyzed within four models based on different assumptions, in order to assess the strong coupling regime. We prove that the guided photons and excitons are strongly coupled at all investigated temperatures, with a small $(11 \%)$ dependence on the temperature. However the values of the Rabi splitting strongly vary among the four models: the "coupled oscillator" model over-estimates the coupling; the analytical "Elliott-Tanguy" model precisely describes the dielectric susceptibility of GaN near the excitonic transition, leading to a Rabi splitting equal to $82 \pm 10 \ meV$ for TE0 modes; the experimental ellipsometry-based model leads to smaller values of $55 \pm 6 \ meV.$ We evidence that for waveguides including active layers with large oscillator strengths, as required for room-temperature polaritonic devices, a strong bending of the modes dispersion is not necessarily the signature of the strong-coupling, which requires for its reliable assessment a precise analysis of the material dielectric susceptibility., Comment: 145 pages, 8 figures

Here, we experimentally and theoretically clarify III-V/Si crystal growth processes. Atomically-resolved microscopy shows that mono-domain 3D islands are observed at the early stages of AlSb, AlN and GaP epitaxy on Si, independently of misfit. It is also shown that complete III-V/Si wetting cannot be achieved in most III-V/Si systems. Surface/interface contributions to the free energy variations are found to be prominent over strain relief processes. We finally propose a general and unified description of III-V/Si growth processes, including the description of antiphase boundaries formation., Comment: includes the manuscript file and the supplemental materials