Space-qualifying silicon photonic modulators and circuits.

Reducing the form factor while retaining the radiation hardness and performance matrix is the goal of avionics. While a compromise between a transistor's size and its radiation hardness has reached consensus in microelectronics, the size-performance balance for their optical counterparts has not been quested but eventually will limit the spaceborne photonic instruments' capacity to weight ratio. Here, we performed space experiments of photonic integrated circuits (PICs), revealing the critical roles of energetic charged particles. The year-long cosmic radiation exposure does not change carrier mobility but reduces free carrier lifetime, resulting in unchanged electro-optic modulation efficiency and well-expanded optoelectronic bandwidth. The diversity and statistics of the tested PIC modulator indicate the minimal requirement of shielding for PIC transmitters with small footprint modulators and complexed routing waveguides toward lightweight space terminals for terabits communications and intersatellite ranging.

Integrated photonic metasystem for image classifications at telecommunication wavelength.

Miniaturized image classifiers are potential for revolutionizing their applications in optical communication, autonomous vehicles, and healthcare. With subwavelength structure enabled directional diffraction and dispersion engineering, the light propagation through multi-layer metasurfaces achieves wavelength-selective image recognitions on a silicon photonic platform at telecommunication wavelength. The metasystems implement high-throughput vector-by-matrix multiplications, enabled by near 103 nanoscale phase shifters as weight elements within 0.135 mm2 footprints. The diffraction manifested computing capability incorporates the fabrication and measurement related phase fluctuations, and thus the pre-trained metasystem can handle uncertainties in inputs without post-tuning. Here we demonstrate three functional metasystems: a 15-pixel spatial pattern classifier that reaches near 90% accuracy with femtosecond inputs, a multi-channel wavelength demultiplexer, and a hyperspectral image classifier. The diffractive metasystem provides an alternative machine learning architecture for photonic integrated circuits, with densely integrated phase shifters, spatially multiplexed throughput, and data processing capabilities.

Device architectures for low voltage and ultrafast graphene integrated phase modulators.

The atomic layer thin geometry and semi-metallic band diagram of graphene can be utilized for significantly improving the performance matrix of integrated photonic devices. Its semiconductor-like behavior of Fermi-level tunability allows graphene to serve as an active layer for electro-optic modulation. As a low loss metal layer, graphene can be placed much closer to active layer for low voltage operation. In this work, we investigate hybrid device architectures utilizing semiconductor and metallic properties of the graphene for ultrafast and energy efficient electro-optic phase modulators on semiconductor and dielectric platforms. (1) Directly contacted graphene-silicon heterojunctions. Without oxide layer, the carrier density of graphene can be modulated by the directly contact to silicon layer, while silicon intrinsic region stays mostly depleted. With doped silicon as electrodes, carrier can be quickly injected and depleted from the active region in graphene. The ultrafast carrier transit time and small RC constant promise ultrafast modulation speed (3dB bandwidth of 67 GHz) with an estimated Vπ·L of 1.19 V·mm. (2) Graphene integrated lithium niobite modulator. As a transparent electrode, graphene can be placed close to integrated lithium niobate waveguide for improving coupling coefficient between optical mode profile and electric field with minimal additional loss (4.6 dB/cm). Numerical simulation indicates 2.5× improvement of electro-optic field overlap coefficient, with estimated V π of 0.2 V.