Search

Abstract All-optical nonlinear activation functions (NAFs) are crucial for enabling rapid optical neural networks (ONNs). As linear matrix computation advances in integrated ONNs, on-chip all-optical NAFs face challenges such as limited integration, high latency, substantial power consumption, and a high activation threshold. In this work, we develop an integrated nonlinear optical activator based on the butt-coupling integration of two-dimensional (2D) MoTe2 and optical waveguides (OWGs). The activator exhibits an ultra-broadband response from visible to near-infrared wavelength, a low activation threshold of 0.94 μW, a small device size (~50 µm2), an ultra-fast response rate (2.08 THz), and high-density integration. The excellent nonlinear effects and broadband response of 2D materials have been utilized to create all-optical NAFs. These activators were applied to simulate MNIST handwritten digit recognition, achieving an accuracy of 97.6%. The results underscore the potential application of this approach in ONNs. Moreover, the classification of more intricate CIFAR-10 images demonstrated a generalizable accuracy of 94.6%. The present nonlinear activator promises a general platform for three-dimensional (3D) ultra-broadband ONNs with dense integration and low activation thresholds by integrating a variety of strong nonlinear optical (NLO) materials (e.g., 2D materials) and OWGs in glass.

The number of layers of 2D materials is of great significance for regulating the performance of nanoelectronic devices and optoelectronic devices, where the optimal thickness of the target sample should be determined before further physical research or device manufacturing steps. At present, a variety of different optical technologies have been proposed to determine the thickness of samples by using the relationship between the number of layers and optical properties, including optical contrast, optical imaging, Raman spectra, photoluminescence spectra, nonlinear spectra, near-field optical imaging, ellipsometry spectra, and hyperspectral imaging. In the past decade, the rapidly growing number of 2D materials and their heterostructures has exceeded the capacity of traditional experimental and computational methods. In recent years, machine learning (ML) is emerging as a powerful tool to support such traditional methods, which brings new opportunities to tap the potential of optical technology in a more perspicacious way. The application of optical technology has greatly facilitated the acquisition of optical information of materials, while ML algorithms provide a fast, high-throughput, and intelligent way to complete back-end data processing and inference. A profound integration of conventional optical technologies and ML algorithms significantly helps 2D materials progress in basic studies and practical applications and further promotes industrial manufacturing. [ABSTRACT FROM AUTHOR]