We initially developed an efficient solver to study photodetectors composed of multiple semiconductor layers with varying thicknesses and doping concentrations. Subsequently, we employed it as the forward solver for three different numerical optimization methods aimed at designing Si-Ge photodetectors with larger bandwidth, higher quantum efficiency, and lower phase noise. Our work offers new insights into the design of high-performance photodetectors—a challenging task due to computation time, design constraints, and the complexity of estimating sensitivity to design parameters.
We present a study on the accuracy of three neural network architectures, namely fully-connected neural networks, recurrent neural networks, and attention-based neural networks, in predicting the coupling response of broadband microresonator frequency combs. These frequency combs are crucial for technologies like optical atomic clocks. Optimizing their spectral features, especially the dispersion in coupling to an access waveguide, can be computationally demanding due to the large number of parameters and wide spectral bandwidths involved. To address this challenge, we employ machine learning algorithms to estimate the coupling response at wavelengths not present in the input training data. Our findings demonstrate that when trained with data sets encompassing the upper and lower limits of each design feature, attention mechanisms achieve over 90% accuracy in predicting the coupling rate for spectral ranges six times wider than those used in training. This significantly reduces the computational burden for numerical optimization in ring resonator design, potentially leading to a six-fold reduction in compute time. Moreover, devices with strong correlations between design features and performance metrics may experience even greater acceleration.
We fabricate and characterize mono- and few- layers of MoS2 and WSe2 on glass and SiO2/Si substrates. PbS quantum dots and/or Au nanoparticles are deposited on the fabricated thin metal dichalcogenide films by controlled drop casting and electron beam evaporation techniques. The reflection spectra of the fabricated structures are measured with a spatially resolved reflectometry setup. Both experimental and numerical results show that surface functionalization with metal nanoparticles can enhance atomically thin transition metal dichalcogenides’ absorption and scattering capabilities, however semiconducting quantum dots do not create such effect.
In order to protect optoelectronic and mechanical properties of atomically thin layered materials (ATLMs) fabricated over SiO2/Si substrates, a secondary oxide or nitride layer can be capped over. However, such protective capping might decrease ATLMs’ visibility dramatically. Similar to the early studies conducted for graphene, we numerically determine optimum thicknesses both for capping and underlying oxide layers for strongest visibility of monolayer MoS2, MoSe2, WS2, and WSe2 in different regions of visible spectrum. We find that the capping layer should not be thicker than 60 nm. Furthermore the optimum capping layer thickness value can be calculated as a function of underlying oxide thickness, and vice versa.
Graphene’s controllable optical conductivity and mechanically strong structure make it a suitable material to de- sign tunable localized surface plasmon resonance (LSPR) sensors. In this work, we theoretically and numerically demonstrate that the resonance wavelength of an LSPR sensor can be tuned to any value within a reasonably wide range of wavelengths by changing the voltage applied to graphene layer. Theoretical results reveal a higher sensitivity with respect to regular LSPR sensors.
Conference Committee Involvement (1)
Applications of Machine Learning 2024
20 August 2024 | San Diego, California, United States
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.