Our work on the PLASMIC project exploits Template Assisted Selective Epitaxy (TASE) for the monolithic integration of III–V material on a silicon platform. This technology was invented at IBM [Ref] and exploited for various electronic applications [Refs]. The core of the technique is to use an oxide cavity to guide the growth and starting growth from a Si surface small enough to allow only one nucleation point.
TASE provides a number of important advantages over many other growth methodologies in that we can grow on any crystalline orientation, even amorphous Si [Ref]. The challenges consist of scaling this integration technique from electronic (sub-100 nm) dimensions up to the millimeter scale for developing photonic components.
We aim to integrate active III–V photonic material in close connection with Si and passives. For this purpose, we are using hybrid plasmonic modes to concentrate the optical mode to achieve sub-wavelength confinement. Therefore, we follow two main integration approaches developed in our group: virtual substrate (VS) and direct cavity (DC) growth.
Using the DC approach, we recently demonstrated RT GaAs laser monolithically integrated on Si.
Both approaches allow us to downscale device dimensions and, in combination with plasmonics, help us explore the ultimate downscaling limit of monolithically integrated light sources.
Local III–V integration for active photonic devices is essential to enable fully integrated photonic circuits with hundreds or thousands of active components. Reducing device size also reduces the capacitance to be switched and is required for large-scale photonic circuits to be competitive with electronics when it comes to performance–power tradeoffs [Ref].
The PLASMIC project is also exploring alternatives to noble metals for the plasmonic metals, inspired by [Ref]. In particular, we are evaluating metal nitrides, which might not provide better plasmonic performance, but their improved stability and thermal behavior might lead to innovative device designs.