• Advanced Photonics
  • Vol. 4, Issue 4, 046005 (2022)
Chengkuan Gao, Prabhav Gaur, Shimon Rubin*, and Yeshaiahu Fainman
Author Affiliations
  • University of California, San Diego, Department of Electrical and Computer Engineering, La Jolla, California, United States
  • show less
    DOI: 10.1117/1.AP.4.4.046005 Cite this Article
    Chengkuan Gao, Prabhav Gaur, Shimon Rubin, Yeshaiahu Fainman. Thin liquid film as an optical nonlinear-nonlocal medium and memory element in integrated optofluidic reservoir computer[J]. Advanced Photonics, 2022, 4(4): 046005 Copy Citation Text show less

    Abstract

    Understanding light–matter interaction lies at the core of our ability to harness physical effects and to translate them into new capabilities realized in modern integrated photonics platforms. Here, we present the design and characterization of optofluidic components in an integrated photonics platform and computationally predict a series of physical effects that rely on thermocapillary-driven interaction between waveguide modes and topography changes of optically thin liquid dielectric film. Our results indicate that this coupling introduces substantial self-induced phase change and transmittance change in a single channel waveguide, transmittance through the Bragg grating waveguide, and nonlocal interaction between adjacent waveguides. We then employ the self-induced effects together with the inherent built-in finite relaxation time of the liquid film, to demonstrate that the light-driven deformation can serve as a reservoir computer capable of performing digital and analog tasks, where the gas–liquid interface operates both as a nonlinear actuator and as an optical memory element.

    1 Introduction

    Light–matter interaction resides at the core of advancing our understanding of both light and matter properties. The emergence of robust-integrated photonic platforms over the last two decades, characterized by miniaturized cross-section and higher refractive index contrast, was leveraged to enhance the optical intensity, thus achieving nonlinear response of various matter degrees of freedom. More recently, nonlinear integrated photonics platforms1 demonstrated promising capabilities to conduct basic scientific research in technologically attractive applications, such as frequency conversion and third-harmonic generation,2 supercontinuum generation,3 emerging quantum photonics applications,4 and are considered as an attractive platform for future nonconventional computation architectures.5 In particular, efficiency of neuro systems to process computational tasks, which are challenging to traditional Turing–von-Neumann machines6 due to sequential “line-by-line” operation, inspired the development of machine learning-based neuromorphic computing (NC) and recurrent neural network (RNN) computational models7,8 as well as its subset called reservoir computing (RC).9,10 Similarly to RNN, the recurrent signal in RC constitutes a memory capable of participating in parallel computation; however, in contrast to RNN, where computationally complex algorithms are required to tune internal weights in the network, in RC, the dynamics occurs in a fixed recurrent network, and only external weights are digitally tuned, thus leading to a significant reduction of the training computation time and the size of the required memory. Consequently, various physical systems can, in principle, serve as powerful RC platforms,11 where the complex and nonlinear signal output produced by the physical system is collected and used for supervised learning to obtain the desired set of weight coefficients during the digital learning stage. Key properties of physical RC systems should satisfy sufficiently large reservoir dimensionality, allowing us to separate features of distinct states according to their dynamics, short-term (fading) memory, and a balance between sensitivity and separability. Among the different proposed physical mechanisms, such as wave propagation,12 optical-based systems13,14 are particularly attractive due to their inherent parallelism, speed-of-light data propagation and processing, and relatively low energy consumption, leading to both on-chip1517 and free-space18,19 realizations.

    Chengkuan Gao, Prabhav Gaur, Shimon Rubin, Yeshaiahu Fainman. Thin liquid film as an optical nonlinear-nonlocal medium and memory element in integrated optofluidic reservoir computer[J]. Advanced Photonics, 2022, 4(4): 046005
    Download Citation