Impact and behavior of Sn during the Ni/GeSn solid-state reaction.

Ni-based intermetallics are promising materials for forming efficient contacts in GeSn-based Si photonic devices. However, the role that Sn might have during the Ni/GeSn solid-state reaction (SSR) is not fully understood. A comprehensive analysis focused on Sn segregation during the Ni/GeSn SSR was carried out. In situ X-ray diffraction and cross-section transmission electron microscopy measurements coupled with energy-dispersive X-ray spectrometry and electron energy-loss spectroscopy atomic mappings were performed to follow the phase sequence, Sn distribution and segregation. The results showed that, during the SSR, Sn was incorporated into the intermetallic phases. Sn segregation happened first around the grain boundaries (GBs) and then towards the surface. Sn accumulation around GBs hampered atom diffusion, delaying the growth of the Ni(GeSn) phase. Higher thermal budgets will thus be mandatory for formation of contacts in high-Sn-content photonic devices, which could be detrimental for thermal stability.

GeSn heterostructure micro-disk laser operating at 230 K.

We demonstrate lasing up to 230 K in a GeSn heterostructure micro-disk cavity. The GeSn 16.0% optically active layer was grown on a step-graded GeSn buffer, limiting the density of misfit dislocations. The lasing wavelengths shifted from 2720 to 2890 nm at 15 K up to 3200 nm at 230 K. Compared to results reported elsewhere, we attribute the increase in maximal lasing temperature to two factors: a stronger optical confinement by a thicker active layer and a better carrier confinement provided by a GeSn 13.8% / GeSn 16.0% / GeSn 13.8% double heterostructure.

Self-adapting denoising, alignment and reconstruction in electron tomography in materials science.

An automatic procedure for electron tomography is presented. This procedure is adapted for specimens that can be fashioned into a needle-shaped sample and has been evaluated on inorganic samples. It consists of self-adapting denoising, automatic and accurate alignment including detection and correction of tilt axis, and 3D reconstruction. We propose the exploitation of a large amount of information of an electron tomography acquisition to achieve robust and automatic mixed Poisson-Gaussian noise parameter estimation and denoising using undecimated wavelet transforms. The alignment is made by mixing three techniques, namely (i) cross-correlations between neighboring projections, (ii) common line algorithm to get a precise shift correction in the direction of the tilt axis and (iii) intermediate reconstructions to precisely determine the tilt axis and shift correction in the direction perpendicular to that axis. Mixing alignment techniques turns out to be very efficient and fast. Significant improvements are highlighted in both simulations and real data reconstructions of porous silicon in high angle annular dark field mode and agglomerated silver nanoparticles in incoherent bright field mode. 3D reconstructions obtained with minimal user-intervention present fewer artefacts and less noise, which permits easier and more reliable segmentation and quantitative analysis. After careful sample preparation and data acquisition, the denoising procedure, alignment and reconstruction can be achieved within an hour for a 3D volume of about a hundred million voxels, which is a step toward a more routine use of electron tomography.