LICORICE

Optoelectronic devices based on III-V semiconductor are fabricated on expensive and relatively small III-V wafers, mostly made of critical raw materials. Silicon, in contrast, is an abundant material, available in the form of large wafers at the heart of the entire microelectronics industry. Its use as a substitute for the III-V substrates would be a major step towards the sustainable use of III and V materials, and the fabrication of fully integrated photonic chips. However, the epitaxial growth of III-V on Si results in a high density of dislocations, which is detrimental to device performance and reliability. In a previous project, we investigated filtering techniques to reduce their density before they thread through the device active region. While significant improvements have been made, dislocation densities are still too high. The LICORICE project proposes to explore a new paradigm based on the design of trapping interfaces within the device to minimize the dislocation length in contact with the active layers, and to prevent the formation of new dislocation segments during device operation. Layers of a few nanometers will be studied to optimize the interface trapping process in terms of strain, while retaining good electronic and optical properties. The challenge is indeed to place these trapping interfaces as close as possible to the active layers of the device without degrading its performance by introducing optical losses or electrical resistances. To this end, a consortium including a recognized epitaxy laboratory on the French side and a group expert in the study of crystal defects and interfaces by electron microscopy on the German side will study dedicated heterostructures in order to optimize the trapping interfaces in terms of thicknesses, composition, placement with respect to the active layers, etc. Finally, lasers will be fabricated and tested, to demonstrate performance and reliability comparable to that of devices grown on III-V substrates.