We present three different approaches to apply deep learning to inverse design for nanophotonic devices. The forward models use device parameters as inputs and device responses as outputs. This model works as a fast approximation method which can be integrated in the optimization loop, and can accelerate the optimization. The network is updated as we obtain more simulation data on the fly for better approximation. The inverse modeling uses a network trained with the device responses as inputs, and the device parameters as outputs. This way the network outputs the device structure given the target optical response. This network can also be updated as we obtain more data during the optimization and validation. The generative model we use is a variant of a conditional variational autoencoder, and the network learns the statistical characteristics of the device structure, and it generates a series of improved designs given the target device responses. By using these three models, we demonstrate how to design nanophotonic power splitters with multiple splitting ratios.
2D materials enable quantum well-like performance, while enjoying substrate independence. Together with their unique band-engineering potential, they pose an opportunity for exploring next generation devices. The rationale for heterogeneous integration is material function separation; that is to perform electrooptic switching in light-matter-enhanced or polaritonic material-mode combinations, while reserving the bosonic and weak-interacting character for photonics, ideally Si or SiN platforms for cost and loss competitiveness, respectively. Here we report on the first 2D material (TMD) integration into microring resonators (MRR), and demonstrated tunability to critical coupling regime. This system allows determining the TMD index via a semi-empirical approach, which is challenged by traditional ellipsometry due to the atom-thin TMD. We further discuss MRR-TMD electrooptic modulation contrasting spectrally ON versus OFF exciton tuning. We conclude by discussing optical nanocavity-TMD systems with applications in QED or LED emission, such as radiation and emission-channel engineering.
Precision and chip contamination-free placement of two-dimensional (2D) materials is expected to accelerate both the study of fundamental properties and novel device functionality. Current transfer methods of 2D materials onto an arbitrary substrate deploy wet chemistry and viscoelastic stamping. However, these methods produce a) significant cross contamination of the substrate due to the lack of spatial selectivity b) may not be compatible with chemically sensitive device structures, and c) are challenged with respect to spatial alignment. Here, we demonstrate a novel method of transferring 2D materials resembling the functionality known from printing; utilizing a combination of a sharp micro-stamper and viscoelastic polymer, we show precise placement of individual 2D materials resulting in vanishing cross contamination to the substrate. Our 2D printer-method results in an aerial cross contamination improvement of two to three orders of magnitude relative to state-of-the-art dry and direct transfer methods. Moreover, we find that the 2D material quality is preserved in this transfer method. Testing this 2D material printer on taped-out integrated Silicon photonic chips, we find that the micro-stamper stamping transfer does not physically harm the underneath Silicon nanophotonic structures such as waveguides or micro-ring resonators receiving the 2D material. Such accurate and substrate-benign transfer method for 2D materials could be industrialized for rapid device prototyping due to its high time-reduction, accuracy, and contamination-free process.
Here we report on an electrically-driven, CMOS compatible, ring resonator coupling modulation mechanism based on tuning of free carriers in Indium Tin Oxide (ITO). The modulator device consists of two ITO layers separated by oxide fabricated on the coupling region of silicon ring resonator. Micro ring resonators are extensively used in in integrated photonic applications and are highly sensitive to electro-optical effects. Majority of available ring based modulators are interactivity modulation where active region changes the phase or absorption of the stored optical mode in the cavity. Such devices are essentially limited by the photon lifetime in the cavity. In contrast, coupling modulation devices can change the cross-coupling coefficient of the resonator and take advantage of the non-quasi-static modulation regime. We demonstrate an electrically-driven, CMOS compatible, ring resonator coupling modulation mechanism based on tuning of free carriers in Indium Tin Oxide (ITO). The modulator device consists of two ITO layers separated by oxide fabricated on the coupling region of silicon ring resonator. We are investigating modulation performance of such CMOS compatible coupling modulation devices. We have demonstrated the first reservoir coupling ITO modulator by leveraging critical coupling effects on a SOI ring resonator.
Graphene, as the first identified two dimensional material, has shown great electro-optic response via Pauli-blocking for near IR frequencies and modulating functionality. However, this ability to modulate light is fundamentally challenged by its small optical cross-section leading to miniscule modal confinement factors in diffraction-limited photonics despite intrinsically high electro-optic absorption modulation (EAM) potential given by its strong index change. Yet the inherent polarization anisotropy in graphene and device tradeoffs lead to additional requirements with respect to electric field directions and modal confinement. The extinction ratio of graphene based EAM has, so far, been limited due to the small light matter interaction given the monolayer structure nature. Here we report an ultra-compact graphene based EAM by integrating graphene with a plasmonic slot waveguide. We show that the modal confinement and hence the modulation strength of a single-layer modulated graphene in this plasmonically confined mode is able to improve by more than 10x compared to diffraction-limited modes. Combined with the strong-index modulation of graphene the modulation strength could achieve more than 1dB/um, which is more than 2-orders of magnitude higher compared to Silicon platform graphene modulators. Furthermore, the modal confinement was found to be synergistic with performance optimization via enhanced light-matter-interactions. These results show that there is room for scaling 2D material EAMs with respect to modal engineering towards realizing synergistic designs leading to high-performance modulators.
The field of two-dimensional (2D) materials has the potential to enable unique applications across a wide range of the electromagnetic spectrum. While 2D-layered materials hold promise for next-generation photon-conversion intrinsic limitations and challenges exist that shall be overcome. Here we discuss the intrinsic limitations as well as application opportunities of this new class of materials, and is sponsored by the NSF program Designing Materials to Revolutionize and Engineer our Future (DMREF) program, which links to the President’s Materials Genome Initiative. We present general material-related details for photon conversion, and show that taking advantage of the mechanical flexibility of 2D materials by rolling MoS2/graphene/hexagonal boron nitride stack to a spiral solar cell allows for solar absorption up to 90%.