Plasmonic nanostructure design and characterization via Deep Learning

Deep Convolutional Mixture Density Network for Inverse Design of Layered Photonic Structures

Machine learning (ML) techniques, such as neural networks, have emerged as powerful tools for the inverse design of nanophotonic structures. However, this innovative approach suffers some limitations. A primary one is the nonuniqueness problem, which can prevent ML algorithms from properly converging because vastly different designs produce nearly identical spectra. Here, we introduce a mixture density network (MDN) approach, which models the design parameters as multimodal probability distributions instead of discrete values, allowing the algorithms to converge in cases of nonuniqueness without sacrificing degenerate solutions. We apply our MDN technique to inversely design two types of multilayer photonic structures consisting of thin films of oxides, which present a significant challenge for conventional ML algorithms due to a high degree of nonuniqueness in their optical properties. In the 10-layer case, the MDN can handle transmission spectra with high complexity and under varying illumination conditions. The 4-layer case tends to show a stronger multimodal character, with secondary modes indicating alternative solutions for a target spectrum. The shape of the distributions gives valuable information for postprocessing and about the uncertainty in the predictions, which is not available with deterministic networks. Our approach provides an effective solution to the inverse design of photonic structures and yields more optimal searches for the structures with high degeneracy and spectral complexity.

Simulation and Fabrication of Nanoscale Spirals Based on Dual-Scale Self-Assemblies

Archimedean spirals in nanometer scale have shown remarkable plasmonic responses derived from their linear and rotational asymmetry. Despite the unique optical properties of nanoscale spirals, their applications have been limited due to the difficulty in fabricating large-scale arrays with uniform and systematic control of the morphology. Here, we report simulation results of spiral morphologies, which are used to design a scalable fabrication process for nanoscale spirals and predict their plasmonic responses. First, self-consistent field theory (SCFT) simulations were performed to design optimal templates to guide self-assembly into spiral morphologies. Using the SCFT results, we developed a scalable fabrication process, which is based on the micron-scale assembly of microspheres combined with glancing angle deposition and nanoscale assembly of block copolymers, to induce the formation of uniform nanospirals with diverse size, handedness, orientation, and winding number. Finally, finite-difference time-domain simulation results show linear dichroism and electric field intensity enhancement effects of these nanospirals, which are highly dependent on the winding number of the spirals, indicating the importance of precise control of the structural parameters.

Deep learning modeling approach for metasurfaces with high degrees of freedom

Metasurfaces have shown promising potentials in shaping optical wavefronts while remaining compact compared to bulky geometric optics devices. The design of meta-atoms, the fundamental building blocks of metasurfaces, typically relies on trial and error to achieve target electromagnetic responses. This process includes the characterization of an enormous amount of meta-atom designs with varying physical and geometric parameters, which demands huge computational resources. In this paper, a deep learning-based metasurface/meta-atom modeling approach is introduced to significantly reduce the characterization time while maintaining accuracy. Based on a convolutional neural network (CNN) structure, the proposed deep learning network is able to model meta-atoms with nearly freeform 2D patterns and different lattice sizes, material refractive indices and thicknesses. Moreover, the presented approach features the capability of predicting a meta-atom's wide spectrum response in the timescale of milliseconds, attractive for applications necessitating fast on-demand design and optimization of a meta-atom/metasurface.

Deep red fluoride dots-in-nanoparticles for high color quality micro white light-emitting diodes

In this study, a novel nanostructure of fluoride red emitting phosphor is synthesized via soft templates. K₂SiF₆:Mn⁴⁺ nanocrystals in the range of 3-5 nm diameter are found inside the porous K₂SiF₆:Mn⁴⁺ nanoparticle hosts, forming unique dots-in-nanoparticles (d-NPs) structures with controlled optical properties. The porous K₂SiF₆:Mn⁴⁺ d-NPs exhibit a sharp and deep red emission with an excellent quantum yield of ∼95.9%, and ultra-high color purity with the corresponding x and y in the CIE chromaticity coordinates are 0.7102 and 0.2870, respectively. Moreover, this nanophosphor possesses good thermal stability in range of 300 K-500 K, under light excitation of 455 nm. The K₂SiF₆:Mn⁴⁺ d-NPs are covered onto a surface of 100×100 µm² blue-yellow In_xGa_1-xN nanowire light-emitting diode (LED) to make warm white LEDs (WLEDs). The fabricated WLEDs present an excellent color rendering index of ∼95.4 and a low correlated color temperature of ∼3649 K. Porous K₂SiF₆:Mn⁴⁺ d-NPs are suggested as a potential red component for high color quality micro WLED applications.

Biologically interfaced nanoplasmonic sensors

Understanding biointerfacial processes is crucial in various fields across fundamental and applied biology, but performing quantitative studies via conventional characterization techniques remains challenging due to instrumentation as well as analytical complexities and limitations. In order to accelerate translational research and address current challenges in healthcare and medicine, there is an outstanding need to develop surface-sensitive technologies with advanced measurement capabilities. Along this line, nanoplasmonic sensing has emerged as a powerful tool to quantitatively study biointerfacial processes owing to its high spatial resolution at the nanoscale. Consequently, the development of robust biological interfacing strategies becomes imperative to maximize its characterization potential. This review will highlight and discuss the critical role of biological interfacing within the context of constructing nanoplasmonic sensing platforms for biointerfacial science applications. Apart from paving the way for the development of highly surface-sensitive characterization tools that will spur fundamental biological interaction studies and improve the overall understanding of biological processes, the basic principles behind biointerfacing strategies presented in this review are also applicable to other fields that involve an interface between an inorganic material and a biological system.

Deep Q-network to produce polarization-independent perfect solar absorbers: a statistical report

Using reinforcement learning, a deep Q-network was used to design polarization-independent, perfect solar absorbers. The deep Q-network selected the geometrical properties and materials of a symmetric three-layer metamaterial made up of circular rods on top of two films. The combination of all the possible permutations gives around 500 billion possible designs. In around 30,000 steps, the deep Q-network was able to produce 1250 structures that have an integrated absorption of higher than 90% in the visible region, with a maximum of 97.6% and an integrated absorption of less than 10% in the 8-13 µm wavelength region, with a minimum of 1.37%. A statistical analysis of the distribution of materials and geometrical parameters that make up the solar absorbers is presented.

Inverse design of an integrated-nanophotonics optical neural network

Artificial neural networks have dramatically improved the performance of many machine-learning applications such as image recognition and natural language processing. However, the electronic hardware implementations of the above-mentioned tasks are facing performance ceiling because Moore's Law is slowing down. In this article, we propose an optical neural network architecture based on optical scattering units to implement deep learning tasks with fast speed, low power consumption and small footprint. The optical scattering units allow light to scatter back and forward within a small region and can be optimized through an inverse design method. The optical scattering units can implement high-precision stochastic matrix multiplication with mean squared error <10^-4 and a mere 4 × 4 μm² footprint. Furthermore, an optical neural network framework based on optical scattering units is constructed by introducing "Kernel Matrix", which can achieve 97.1% accuracy on the classic image classification dataset MNIST.

Development of use-specific high-performance cyber-nanomaterial optical detectors by effective choice of machine learning algorithms

Due to their inherent variabilities, nanomaterials-based sensors are challenging to translate into real-world applications, where reliability and reproducibility are key. Machine learning can be a powerful approach for obtaining reliable inferences from data generated by such sensors. Here, we show that the best choice of ML algorithm in a cyber-nanomaterial detector is largely determined by the specific use-considerations, including accuracy, computational cost, speed, and resilience against drifts and long-term ageing effects. When sufficient data and computing resources are provided, the highest sensing accuracy can be achieved by the k-nearest neighbors (kNNs) and Bayesian inference algorithms, however, these algorithms can be computationally expensive for real-time applications. In contrast, artificial neural networks (ANNs) are computationally expensive to train (off-line), but they provide the fastest result under testing conditions (on-line) while remaining reasonably accurate. When access to data is limited, support vector machines (SVMs) can perform well even with small training sample sizes, while other algorithms show considerable reduction in accuracy if data is scarce, hence, setting a lower limit on the size of required training data. We also show by tracking and modeling the long-term drifts of the detector performance over a one year time-frame, it is possible to dramatically improve the predictive accuracy without any re-calibration. Our research shows for the first time that if the ML algorithm is chosen specific to the use-case, low-cost solution-processed cyber-nanomaterial detectors can be practically implemented under diverse operational requirements, despite their inherent variabilities.

Teaching optics to a machine learning network

In this Letter, we demonstrate how harmonic oscillator equations can be integrated in a neural network to improve the spectral response prediction for an optical system. We use the optical properties of a one-dimensional nanoslit array for a practical implementation of the study. This method allows to build more generalizable relations between the input parameters of the array and its optical properties, showing a 20-fold improvement for parameters outside the range used for the training. We also show how this model generates the output spectrum from phenomenological relationships between the input parameters and the output spectrum, indicating how it grasps the physical mechanisms of the optical response of the structure.

Prediction Network of Metamaterial with Split Ring Resonator Based on Deep Learning

The introduction of "metamaterials" has had a profound impact on several fields, including electromagnetics. Designing a metamaterial's structure on demand, however, is still an extremely time-consuming process. As an efficient machine learning method, deep learning has been widely used for data classification and regression in recent years and in fact shown good generalization performance. We have built a deep neural network for on-demand design. With the required reflectance as input, the parameters of the structure are automatically calculated and then output to achieve the purpose of designing on demand. Our network has achieved low mean square errors (MSE), with MSE of 0.005 on both the training and test sets. The results indicate that using deep learning to train the data, the trained model can more accurately guide the design of the structure, thereby speeding up the design process. Compared with the traditional design process, using deep learning to guide the design of metamaterials can achieve faster, more accurate, and more convenient purposes.

Deep learning: a new tool for photonic nanostructure design

Early results have shown the potential of Deep Learning (DL) to disrupt the fields of optical inverse-design, particularly, the inverse design of nanostructures. In the last three years, the complexity of the optical nanostructure being designed and the sophistication of the employed DL methodology have steadily increased. This topical review comprehensively surveys DL based design examples from the nanophotonics literature. Notwithstanding the early success of this approach, its limitations, range of validity and its place among established design techniques remain to be assessed. The review also provides a perspective on the limitations of this approach and emerging research directions. It is hoped that this topical review may help readers to identify unaddressed problems, to choose an initial setup for a specific problem, and, to identify means to improve the performance of existing DL based workflows.

Inverse design of plasmonic metasurfaces by convolutional neural network

Artificial neural networks have shown effectiveness in the inverse design of nanophotonic structures; however, the numerical accuracy and algorithm efficiency are not analyzed adequately in previous reports. In this Letter, we demonstrate the convolutional neural network as an inverse design tool to achieve high numerical accuracy in plasmonic metasurfaces. A comparison of the convolutional neural networks and the fully connected neural networks show that convolutional neural networks have higher generalization capabilities. We share practical guidelines for optimizing the neural network and analyzed the hierarchy of accuracy in the multi-parameter inverse design of plasmonic metasurfaces. A high inverse design accuracy of $\pm 8\;{\rm nm}$±8nm for the critical geometrical parameters is demonstrated.

Parametric optimization of visible wavelength gold lattice geometries for improved plasmon-enhanced fluorescence spectroscopy

We report the exploitation of spectroplasmonics for innovations in optical transducer development, specifically in the well-established application of labeled fluorescent analytes known as fluorescence spectroscopy. Presented herein are comprehensive analyses of nanoscale plasmonic lattice feature geometries using finite-difference time-domain software to determine the largest surface electric ($E$E) field enhancement resulting from localized surface plasmon resonance for reducing the limit of detection of plasmon-enhanced fluorescence. This parametric optimization of the critical dimensions of the plasmon resonance of noble metal nanostructures will enable improved excitation and emission enhancement of fluorophores used in visible wavelength fluorescence spectroscopy.

Topological encoding method for data-driven photonics inverse design

Data-driven approaches have been proposed as effective strategies for the inverse design and optimization of photonic structures in recent years. In order to assist data-driven methods for the design of topology of photonic devices, we propose a topological encoding method that transforms photonic structures represented by binary images to a continuous sparse representation. This sparse representation can be utilized for dimensionality reduction and dataset generation, enabling effective analysis and optimization of photonic topologies with data-driven approaches. As a proof of principle, we leverage our encoding method for the design of two dimensional non-paraxial diffractive optical elements with various diffraction intensity distributions. We proved that our encoding method is able to assist machine-learning-based inverse design approaches for accurate and global optimization.

Compounding Meta-Atoms into Metamolecules with Hybrid Artificial Intelligence Techniques

Molecules composed of atoms exhibit properties not inherent to their constituent atoms. Similarly, metamolecules consisting of multiple meta-atoms possess emerging features that the meta-atoms themselves do not possess. Metasurfaces composed of metamolecules with spatially variant building blocks, such as gradient metasurfaces, are drawing substantial attention due to their unconventional controllability of the amplitude, phase, and frequency of light. However, the intricate mechanisms and the large degrees of freedom of the multielement systems impede an effective strategy for the design and optimization of metamolecules. Here, a hybrid artificial-intelligence-based framework consolidating compositional pattern-producing networks and cooperative coevolution to resolve the inverse design of metamolecules in metasurfaces is proposed. The framework breaks the design of the metamolecules into separate designs of meta-atoms, and independently solves the smaller design tasks of the meta-atoms through deep learning and evolutionary algorithms. The proposed framework is leveraged to design metallic metamolecules for arbitrary manipulation of the polarization and wavefront of light. Moreover, the efficacy and reliability of the design strategy are confirmed through experimental validations. This framework reveals a promising candidate approach to expedite the design of large-scale metasurfaces in a labor-saving, systematic manner.

Deep Learning Meets Nanophotonics: A Generalized Accurate Predictor for Near Fields and Far Fields of Arbitrary 3D Nanostructures

Deep artificial neural networks are powerful tools with many possible applications in nanophotonics. Here, we demonstrate how a deep neural network can be used as a fast, general purpose predictor of the full near-field and far-field response of plasmonic and dielectric nanostructures. A trained neural network is shown to infer the internal fields of arbitrary three-dimensional nanostructures many orders of magnitude faster compared to conventional numerical simulations. Secondary physical quantities are derived from the deep learning predictions and faithfully reproduce a wide variety of physical effects without requiring specific training. We discuss the strengths and limitations of the neural network approach using a number of model studies of single particles and their near-field interactions. Our approach paves the way for fast, yet universal, methods for design and analysis of nanophotonic systems.

Design of task-specific optical systems using broadband diffractive neural networks

Deep learning has been transformative in many fields, motivating the emergence of various optical computing architectures. Diffractive optical network is a recently introduced optical computing framework that merges wave optics with deep-learning methods to design optical neural networks. Diffraction-based all-optical object recognition systems, designed through this framework and fabricated by 3D printing, have been reported to recognize hand-written digits and fashion products, demonstrating all-optical inference and generalization to sub-classes of data. These previous diffractive approaches employed monochromatic coherent light as the illumination source. Here, we report a broadband diffractive optical neural network design that simultaneously processes a continuum of wavelengths generated by a temporally incoherent broadband source to all-optically perform a specific task learned using deep learning. We experimentally validated the success of this broadband diffractive neural network architecture by designing, fabricating and testing seven different multi-layer, diffractive optical systems that transform the optical wavefront generated by a broadband THz pulse to realize (1) a series of tuneable, single-passband and dual-passband spectral filters and (2) spatially controlled wavelength de-multiplexing. Merging the native or engineered dispersion of various material systems with a deep-learning-based design strategy, broadband diffractive neural networks help us engineer the light-matter interaction in 3D, diverging from intuitive and analytical design methods to create task-specific optical components that can all-optically perform deterministic tasks or statistical inference for optical machine learning.

A Bidirectional Deep Neural Network for Accurate Silicon Color Design

Silicon nanostructure color has achieved unprecedented high printing resolution and larger color gamut than sRGB. The exact color is determined by localized magnetic and electric dipole resonance of nanostructures, which are sensitive to their geometric changes. Usually, the design of specific colors and iterative optimization of geometric parameters are computationally costly, and obtaining millions of different structural colors is challenging. Here, a deep neural network is trained, which can accurately predict the color generated by random silicon nanostructures in the forward modeling process and solve the nonuniqueness problem in the inverse design process that can accurately output the device geometries for at least one million different colors. The key results suggest deep learning is a powerful tool to minimize the computation cost and maximize the design efficiency for nanophotonics, which can guide silicon color manufacturing with high accuracy.

Self-Learning Perfect Optical Chirality via a Deep Neural Network

Optical chirality occurs when materials interact differently with light in a specific circular polarization state. Chiroptical phenomena inspire wide interdisciplinary investigations, which require advanced designs to reach strong chirality for practical applications. The development of artificial intelligence provides a new vision for the manipulation of light-matter interaction beyond the theoretical interpretation. Here, we report a self-consistent framework named the Bayesian optimization and convolutional neural network that combines Bayesian optimization and deep convolutional neural network algorithms to calculate and optimize optical properties of metallic nanostructures. Both electric-field distributions at the near field and reflection spectra at the far field are calculated and self-learned to suggest better structure designs and provide possible explanations for the origin of the optimized properties, which enables wide applications for future nanostructure analysis and design.

The inverse design of structural color using machine learning

Efficiently identifying optical structures with desired functionalities, referred to as inverse design, can dramatically accelerate the invention of new photonic devices, and this is especially useful in the design of large scale integrated photonic chips. Structural color with high-resolution, high-saturation, and low-loss holds great promise in image display, data storage and information security. However, the inverse design of structural color remains an open challenge, and this impedes practical application. Here, we propose an inverse design strategy for structural color using machine learning (ML) technologies. The supervised learning (SL) models are trained with the geometries and colors of dielectric arrays to capture accurate geometry-color relationships, and these are then applied to a reinforcement learning (RL) algorithm in order to find the optical structural geometries for the desired color. Our work succeeds in finding simple and accurate models to describe geometry-color relationships, which significantly improves the efficiency of the design. This strategy provides a systematic method to directly encode generic functionality into a set of structures and geometries, paving the way for the inverse design of functional photonic devices.

Bayesian Machine Learning in Metamaterial Design: Fragile Becomes Supercompressible

Designing future-proof materials goes beyond a quest for the best. The next generation of materials needs to be adaptive, multipurpose, and tunable. This is not possible by following the traditional experimentally guided trial-and-error process, as this limits the search for untapped regions of the solution space. Here, a computational data-driven approach is followed for exploring a new metamaterial concept and adapting it to different target properties, choice of base materials, length scales, and manufacturing processes. Guided by Bayesian machine learning, two designs are fabricated at different length scales that transform brittle polymers into lightweight, recoverable, and supercompressible metamaterials. The macroscale design is tuned for maximum compressibility, achieving strains beyond 94% and recoverable strengths around 0.1 kPa, while the microscale design reaches recoverable strengths beyond 100 kPa and strains around 80%. The data-driven code is available to facilitate future design and analysis of metamaterials and structures (https://github.com/mabessa/F3DAS).

Intelligent metasurface imager and recognizer

There is an increasing need to remotely monitor people in daily life using radio-frequency probe signals. However, conventional systems can hardly be deployed in real-world settings since they typically require objects to either deliberately cooperate or carry a wireless active device or identification tag. To accomplish complicated successive tasks using a single device in real time, we propose the simultaneous use of a smart metasurface imager and recognizer, empowered by a network of artificial neural networks (ANNs) for adaptively controlling data flow. Here, three ANNs are employed in an integrated hierarchy, transforming measured microwave data into images of the whole human body, classifying specifically designated spots (hand and chest) within the whole image, and recognizing human hand signs instantly at a Wi-Fi frequency of 2.4 GHz. Instantaneous in situ full-scene imaging and adaptive recognition of hand signs and vital signs of multiple non-cooperative people were experimentally demonstrated. We also show that the proposed intelligent metasurface system works well even when it is passively excited by stray Wi-Fi signals that ubiquitously exist in our daily lives. The reported strategy could open up a new avenue for future smart cities, smart homes, human-device interaction interfaces, health monitoring, and safety screening free of visual privacy issues.

Controlling the degrees of freedom in metasurface designs for multi-functional optical devices

This review focuses on the control over the degrees of freedom (DOF) in metasurfaces, which include the input DOF (the polarization, wavelength and incident angle of the input light and some dynamic controls), parameter DOF (the complex geometric design of metasurfaces) and output DOF (the phase, polarization and amplitude of the output light). This framework could clearly show us the development process of metasurfaces, from single-functional to multi-functional ones. Advantages of the multi-functional metasurfaces are discussed in the context of various applications, including 3D holography, broadband achromatic metalenses and multi-dimensional encoded information. By combining all the input and output DOF together, we can realize ideal optical meta-devices with deep subwavelength thickness and striking functions beyond the reach of traditional optical components. Moreover, new research directions may emerge when merging different DOF in metasurfaces with other important concepts, such as parity-time symmetry and topology, so that we can have the complete control of light waves in a prescribed manner.

Deep learning for accelerated all-dielectric metasurface design

Deep learning has risen to the forefront of many fields in recent years, overcoming challenges previously considered intractable with conventional means. Materials discovery and optimization is one such field, but significant challenges remain, including the requirement of large labeled datasets and one-to-many mapping that arises in solving the inverse problem. Here we demonstrate modeling of complex all-dielectric metasurface systems with deep neural networks, using both the metasurface geometry and knowledge of the underlying physics as inputs. Our deep learning network is highly accurate, achieving an average mean square error of only 1.16 × 10^-3 and is over five orders of magnitude faster than conventional electromagnetic simulation software. We further develop a novel method to solve the inverse modeling problem, termed fast forward dictionary search (FFDS), which offers tremendous controls to the designer and only requires an accurate forward neural network model. These techniques significantly increase the viability of more complex all-dielectric metasurface designs and provide opportunities for the future of tailored light matter interactions.

Plasmonic nanoparticle simulations and inverse design using machine learning

Collective oscillation of quasi-free electrons on the surface of metallic plasmonic nanoparticles (NPs) in the ultraviolet to near-infrared (NIR) region induces a strong electromagnetic enhancement around the NPs, which leads to numerous important applications. These interesting far- and near-field optical characteristics of the plasmonic NPs can be typically obtained from numerical simulations for theoretical guidance of NP design. However, traditional numerical simulations encounter irreconcilable conflicts between the accuracy and speed due to the high demand of computing power. In this work, we utilized the machine learning method, specifically the deep neural network (DNN), to establish mapping between the far-field spectra/near-field distribution and dimensional parameters of three types of plasmonic NPs including nanospheres, nanorods, and dimers. After the training process, both the forward prediction of far-field optical properties and the inverse prediction of on-demand dimensional parameters of NPs can be accomplished accurately and efficiently with the DNN. More importantly, we have achieved for the first time ultrafast and accurate prediction of two-dimensional on-resonance electromagnetic enhancement distributions around NPs by greatly reducing the amount of electromagnetic data via screening and resampling methods. These near-field predictions can be realized typically in less than 10-2 seconds on a laptop, which is 6 orders faster than typical numerical simulations implemented on a server. Therefore, we demonstrate that the DNN is an ultrafast, highly efficient, and computing resource-saving tool to investigate the far- and near-field optical properties of plasmonic NPs, especially for a number of important nano-optical applications such as surface-enhanced Raman spectroscopy, photocatalysis, solar cells, and metamaterials.

Emerging role of machine learning in light-matter interaction

Machine learning has provided a huge wave of innovation in multiple fields, including computer vision, medical diagnosis, life sciences, molecular design, and instrumental development. This perspective focuses on the implementation of machine learning in dealing with light-matter interaction, which governs those fields involving materials discovery, optical characterizations, and photonics technologies. We highlight the role of machine learning in accelerating technology development and boosting scientific innovation in the aforementioned aspects. We provide future directions for advanced computing techniques via multidisciplinary efforts that can help to transform optical materials into imaging probes, information carriers and photonics devices.

Free-Form Diffractive Metagrating Design Based on Generative Adversarial Networks

A key challenge in metasurface design is the development of algorithms that can effectively and efficiently produce high-performance devices. Design methods based on iterative optimization can push the performance limits of metasurfaces, but they require extensive computational resources that limit their implementation to small numbers of microscale devices. We show that generative neural networks can train from images of periodic, topology-optimized metagratings to produce high-efficiency, topologically complex devices operating over a broad range of deflection angles and wavelengths. Further iterative optimization of these designs yields devices with enhanced robustness and efficiencies, and these devices can be utilized as additional training data for network refinement. In this manner, generative networks can be trained, with a one-time computation cost, and used as a design tool to facilitate the production of near-optimal, topologically complex device designs. We envision that such data-driven design methodologies can apply to other physical sciences domains that require the design of functional elements operating across a wide parameter space.

Metasurface-Based Molecular Biosensing Aided by Artificial Intelligence

Molecular spectroscopy provides unique information on the internal structure of biological materials by detecting the characteristic vibrational signatures of their constituent chemical bonds at infrared frequencies. Nanophotonic antennas and metasurfaces have driven this concept towards few-molecule sensitivity by confining incident light into intense hot spots of the electromagnetic fields, providing strongly enhanced light-matter interaction. In this Minireview, recently developed molecular biosensing approaches based on the combination of dielectric metasurfaces and imaging detection are highlighted in comparison to traditional plasmonic geometries, and the unique potential of artificial intelligence techniques for nanophotonic sensor design and data analysis is emphasized. Because of their spectrometer-less operation principle, such imaging-based approaches hold great promise for miniaturized biosensors in practical point-of-care or field-deployable applications.

Training artificial neural network for optimization of nanostructured VO2-based smart window performance

In this work, we apply for the first time a machine learning approach to design and optimize VO₂ based nanostructured smart window performance. An artificial neural network was trained to find the relationship between VO₂ smart window structural parameters and performance metrics-luminous transmittance (T_lum) and solar modulation (ΔT_sol), calculated by first-principle electromagnetic simulations (FDTD method). Once training was accomplished, the combination of optimal T_lum and ΔT_sol was found by applying classical trust region algorithm on the trained network. The proposed method allows flexibility in definition of the optimization problem and provides clear uncertainty limits for future experimental realizations.

Probabilistic Representation and Inverse Design of Metamaterials Based on a Deep Generative Model with Semi-Supervised Learning Strategy

The research of metamaterials has achieved enormous success in the manipulation of light in a prescribed manner using delicately designed subwavelength structures, so-called meta-atoms. Even though modern numerical methods allow for the accurate calculation of the optical response of complex structures, the inverse design of metamaterials, which aims to retrieve the optimal structure according to given requirements, is still a challenging task owing to the nonintuitive and nonunique relationship between physical structures and optical responses. To better unveil this implicit relationship and thus facilitate metamaterial designs, it is proposed to represent metamaterials and model the inverse design problem in a probabilistically generative manner, enabling to elegantly investigate the complex structure-performance relationship in an interpretable way, and solve the one-to-many mapping issue that is intractable in a deterministic model. Moreover, to alleviate the burden of numerical calculations when collecting data, a semisupervised learning strategy is developed that allows the model to utilize unlabeled data in addition to labeled data in an end-to-end training. On a data-driven basis, the proposed deep generative model can serve as a comprehensive and efficient tool that accelerates the design, characterization, and even new discovery in the research domain of metamaterials, and photonics in general.

Simultaneous Inverse Design of Materials and Structures via Deep Learning: Demonstration of Dipole Resonance Engineering Using Core-Shell Nanoparticles

Recent introduction of data-driven approaches based on deep-learning technology has revolutionized the field of nanophotonics by allowing efficient inverse design methods. In this paper, a simultaneous inverse design of materials and structure parameters of core-shell nanoparticles is achieved for the first time using deep learning of a neural network. A neural network to learn the correlation between the extinction spectra of electric and magnetic dipoles and core-shell nanoparticle designs, which include material information and shell thicknesses, is developed and trained. We demonstrate deep-learning-assisted inverse design of core-shell nanoparticles for (1) spectral tuning electric dipole resonances, (2) finding spectrally isolated pure magnetic dipole resonances, and (3) finding spectrally overlapped electric dipole and magnetic dipole resonances. Our finding paves the way for the rapid development of nanophotonics by allowing a practical utilization of deep-learning technology for nanophotonic inverse design.

Hybridization of singular plasmons via transformation optics

Surface plasmon resonances of metallic nanostructures offer great opportunities to guide and manipulate light on the nanoscale. In the design of novel plasmonic devices, a central topic is to clarify the intricate relationship between the resonance spectrum and the geometry of the nanostructure. Despite many advances, the design becomes quite challenging when the desired spectrum is highly complex. Here we develop a theoretical model for surface plasmons of interacting nanoparticles to reduce the complexity of the design process significantly. Our model is developed by combining plasmon hybridization theory with transformation optics, which yields an efficient way of simultaneously controlling both global and local features of the resonance spectrum. As an application, we propose a design of metasurface whose absorption spectrum can be controlled over a large class of complex patterns through only a few geometric parameters in an intuitive way. Our approach provides fundamental tools for the effective design of plasmonic metamaterials with on-demand functionality.

Plasmonic-Nanopore Biosensors for Superior Single-Molecule Detection

Plasmonic and nanopore sensors have separately received much attention for achieving single-molecule precision. A plasmonic "hotspot" confines and enhances optical excitation at the nanometer length scale sufficient to optically detect surface-analyte interactions. A nanopore biosensor actively funnels and threads analytes through a molecular-scale aperture, wherein they are interrogated by electrical or optical means. Recently, solid-state plasmonic and nanopore structures have been integrated within monolithic devices that address fundamental challenges in each of the individual sensing methods and offer complimentary improvements in overall single-molecule sensitivity, detection rates, dwell time and scalability. Here, the physical phenomena and sensing principles of plasmonic and nanopore sensing are summarized to highlight the novel complementarity in dovetailing these techniques for vastly improved single-molecule sensing. A literature review of recent plasmonic nanopore devices is then presented to delineate methods for solid-state fabrication of a range of hybrid device formats, evaluate the progress and challenges in the detection of unlabeled and labeled analyte, and assess the impact and utility of localized plasmonic heating. Finally, future directions and applications inspired by the present state of the art are discussed.

Plasmonic colours predicted by deep learning

Picosecond laser pulses have been used as a surface colouring technique for noble metals, where the colours result from plasmonic resonances in the metallic nanoparticles created and redeposited on the surface by ablation and deposition processes. This technology provides two datasets which we use to train artificial neural networks, data from the experiment itself (laser parameters vs. colours) and data from the corresponding numerical simulations (geometric parameters vs. colours). We apply deep learning to predict the colour in both cases. We also propose a method for the solution of the inverse problem – wherein the geometric parameters and the laser parameters are predicted from colour – using an iterative multivariable inverse design method.

Smart inverse design of graphene-based photonic metamaterials by an adaptive artificial neural network

The burgeoning research of graphene and other 2D materials enables many unprecedented metamaterials and metadevices for applications on nanophotonics. The design of on-demand graphene-based metamaterials often calls for the solution of a complex inverse problem within a small sampling space, which highly depends on the rich experiences from researchers of nanophotonics. Conventional optimization algorithms could be used for this inverse design, but they converge to local optimal solutions and take significant computational costs with increased nanostructure parameters. Here, we establish a deep learning method based on an adaptive batch-normalized neural network, aiming to implement smart and rapid inverse design for graphene-based metamaterials with on-demand optical responses. This method allows a quick converging speed with high precision and low computational consumption. As typical complex proof-of-concept examples, the optical metamaterials consisting of graphene/dielectric alternating multilayers are chosen to demonstrate the validity of our design paradigm. Our method demonstrates a high prediction accuracy of over 95% after very few training epochs. A universal programming package is developed to achieve the design goals of graphene-based metamaterials with low absorption and near unity absorption, respectively. Our work may find important design applications in the field of nanoscale photonics based on graphene and other 2D materials.

Nanoplasmonic Approaches for Sensitive Detection and Molecular Characterization of Extracellular Vesicles

All cells release a multitude of nanoscale extracellular vesicles (nEVs) into circulation, offering immense potential for new diagnostic strategies. Yet, clinical translation for nEVs remains a challenge due to their vast heterogeneity, our insufficient ability to isolate subpopulations, and the low frequency of disease-associated nEVs in biofluids. The growing field of nanoplasmonics is poised to address many of these challenges. Innovative materials engineering approaches based on exploiting nanoplasmonic phenomena, i.e., the unique interaction of light with nanoscale metallic materials, can achieve unrivaled sensitivity, offering real-time analysis and new modes of medical and biological imaging. We begin with an introduction into the basic structure and function of nEVs before critically reviewing recent studies utilizing nanoplasmonic platforms to detect and characterize nEVs. For the major techniques considered, surface plasmon resonance (SPR), localized SPR, and surface enhanced Raman spectroscopy (SERS), we introduce and summarize the background theory before reviewing the studies applied to nEVs. Along the way, we consider notable aspects, limitations, and considerations needed to apply plasmonic technologies to nEV detection and analysis.

Intelligent nanophotonics: merging photonics and artificial intelligence at the nanoscale

Nanophotonics has been an active research field over the past two decades, triggered by the rising interests in exploring new physics and technologies with light at the nanoscale. As the demands of performance and integration level keep increasing, the design and optimization of nanophotonic devices become computationally expensive and time-inefficient. Advanced computational methods and artificial intelligence, especially its subfield of machine learning, have led to revolutionary development in many applications, such as web searches, computer vision, and speech/image recognition. The complex models and algorithms help to exploit the enormous parameter space in a highly efficient way. In this review, we summarize the recent advances on the emerging field where nanophotonics and machine learning blend. We provide an overview of different computational methods, with the focus on deep learning, for the nanophotonic inverse design. The implementation of deep neural networks with photonic platforms is also discussed. This review aims at sketching an illustration of the nanophotonic design with machine learning and giving a perspective on the future tasks.

Pushing the limits of optical information storage using deep learning

Diffraction drastically limits the bit density in optical data storage. To increase the storage density, alternative strategies involving supplementary recording dimensions and robust readout schemes must be explored. Here, we propose to encode multiple bits of information in the geometry of subwavelength dielectric nanostructures. A crucial problem in high-density information storage concepts is the robustness of the information readout with respect to fabrication errors and experimental noise. Using a machine-learning-based approach in which the scattering spectra are analysed by an artificial neural network, we achieve quasi-error-free readout of sequences of up to 9 bits, encoded in top-down fabricated silicon nanostructures. We demonstrate that probing few wavelengths instead of the entire spectrum is sufficient for robust information retrieval and that the readout can be further simplified, exploiting the RGB values from microscopy images. Our work paves the way towards high-density optical information storage using planar silicon nanostructures, compatible with mass-production-ready complementary metal-oxide-semiconductor technology.

Rapid phase retrieval of ultrashort pulses from dispersion scan traces using deep neural networks

The knowledge of the temporal shape of femtosecond pulses is of major interest for all their applications. The reconstruction of the temporal shape of these pulses is an inverse problem for characterization techniques, which benefit from an inherent redundancy in the measurement. Conventionally, time-consuming optimization algorithms are used to solve the inverse problems. Here, we demonstrate the reconstruction of ultrashort pulses from dispersion scan traces employing a deep neural network. The network is trained with a multitude of artificial and noisy dispersion scan traces from randomly shaped pulses. The retrieval takes only 16 ms enabling video-rate reconstructions. This approach reveals a great tolerance against noisy conditions, delivering reliable retrievals from traces with signal-to-noise ratios down to 5.