Integration of photonic circuits on silicon offers a unique opportunity to address the scaling of inter- and intra-chip communications in an energy-efficient and cost-effective manner. Mode-division multiplexing (MDM) is deemed as one of the most promising technologies to increase aggregated data bandwidth and avoid a communication capacity crunch. In this invited talk, we review our latest advances on integrated silicon mode multiplexers, including new topologies based on subwavelength grating (SWG) metamaterials for extended broadband operation and higher-order mode support. Specifically, we report on an ultra-broadband multiplexer based on a phase shifter and a multimode interference (MMI) coupler both engineered with subwavelength metamaterials. Experimental measurements of a complete multiplexer-demultiplexer link show losses lower than 2 dB and crosstalk below -17 dB over a bandwidth of 245 nm (1427 – 1672 nm).

This work reports on the development of a novel polymer blend with thermally triggered shape memory and self-healing properties. Blends were prepared by mixing a self-healing ionomer (Surlyn 9520) and polycyclooctene (with and without crosslinking agent) in different ratios. The regions of thermal activation were determined by thermogravimetric analysis, differential scanning calorimetry and oscillatory shear rheology. Consecutively, the shape memory and self-healing behaviour were investigated by a torsion procedure and tensile testing respectively. It was found that ionomer/crosslinked polycyclooctene blends of 70/30 wt% lead to polymers showing partial macroscopic healing and repeatable shape memory characteristics. The new polymer system shows both dual and triple shape memory behaviour and a near to 100% stiffness recovery after healing of crosscuts at standard ionomer healing conditions. Furthermore, the relation between terminal relaxation and self-healing in blends is shown. This study introduces a triple shape memory polymer with self-healing properties by a blending strategy thereby clearing the path for more durable materials based on shape memory properties.