arXiv subjects

Weilu Gao

Publications and source records attributed to Weilu Gao.

35 records · Page 2Linked to original sources

arXiv2022

Perfectly Perform Machine Learning Task with Imperfect Optical Hardware Accelerator

Optical architectures have been emerging as an energy-efficient and high-throughput hardware platform to accelerate computationally intensive general matrix-matrix multiplications (GEMMs) in modern machine learning (ML) algorithms. However, the inevitable imperfection and non-uniformity in large-scale optoelectronic devices prevent the scalable deployment of optical architectures, particularly those with innovative nano-devices. Here, we report an optical ML hardware to accelerate GEMM operations based on cascaded spatial light modulators and present a calibration procedure that enables accurate calculations despite the non-uniformity and imperfection in devices and system. We further characterize the hardware calculation accuracy under different configurations of electrical-optical interfaces. Finally, we deploy the developed optical hardware and calibration procedure to perform a ML task of predicting the intersubband plasmon frequency in single-wall carbon nanotubes. The obtained prediction accuracy from the optical hardware agrees well with that obtained using a general purpose electronic graphic process unit.

cs.ET↗

arXiv2022

Physics-aware Complex-valued Adversarial Machine Learning in Reconfigurable Diffractive All-optical Neural Network

Diffractive optical neural networks have shown promising advantages over electronic circuits for accelerating modern machine learning (ML) algorithms. However, it is challenging to achieve fully programmable all-optical implementation and rapid hardware deployment. Furthermore, understanding the threat of adversarial ML in such system becomes crucial for real-world applications, which remains unexplored. Here, we demonstrate a large-scale, cost-effective, complex-valued, and reconfigurable diffractive all-optical neural networks system in the visible range based on cascaded transmissive twisted nematic liquid crystal spatial light modulators. With the assist of categorical reparameterization, we create a physics-aware training framework for the fast and accurate deployment of computer-trained models onto optical hardware. Furthermore, we theoretically analyze and experimentally demonstrate physics-aware adversarial attacks onto the system, which are generated from a complex-valued gradient-based algorithm. The detailed adversarial robustness comparison with conventional multiple layer perceptrons and convolutional neural networks features a distinct statistical adversarial property in diffractive optical neural networks. Our full stack of software and hardware provides new opportunities of employing diffractive optics in a variety of ML tasks and enabling the research on optical adversarial ML.

cs.ET↗

arXiv2022

Device-system Co-design of Photonic Neuromorphic Processor using Reinforcement Learning

The incorporation of high-performance optoelectronic devices into photonic neuromorphic processors can substantially accelerate computationally intensive operations in machine learning (ML) algorithms. However, the conventional device design wisdom is disconnected with system optimization. We report a device-system co-design methodology to optimize a free-space optical general matrix multiplication (GEMM) hardware accelerator by engineering a spatially reconfigurable array made from chalcogenide phase change materials. With a highly-parallelized hardware emulator constructed based on experimental information, we demonstrate the design of unit device by optimizing GEMM calculation accuracy via reinforcement learning, including deep Q-learning neural network, Bayesian optimization, and their cascaded approach, which show a clear correlation between system performance metrics and physical device specifications. Furthermore, we employ physics-aware training approaches to deploy optimized hardware to the tasks of image classification, materials discovery, and a closed-loop design of optical ML accelerators. The demonstrated framework offers insights into the co-design of optoelectronic devices and systems with reduced human-supervision and domain-knowledge barriers.

cs.ET↗

arXiv2022

Evidence for Phonon-Assisted Intertube Electronic Transport in an Armchair Carbon Nanotube Film

The electrical conductivity of a macroscopic assembly of nanomaterials is determined through a complex interplay of electronic transport within and between constituent nano-objects. Phonons play dual roles in this situation: their increased populations tend to reduce the conductivity via electron scattering while they can boost the conductivity by assisting electrons to propagate through the potential-energy landscape. Here, we identify a phonon-assisted coherent electron transport process between neighboring nanotubes in temperature-dependent conductivity measurements on a macroscopic film of armchair single-wall carbon nanotubes. Through atomistic modeling of electronic states and calculations of both electronic and phonon-assisted junction conductances, we conclude that phonon-assisted conductance is the dominant mechanism for the observed high-temperature transport. The unambiguous manifestation of coherent intertube dynamics proves a single-chirality armchair nanotube film to be a unique macroscopic solid-state ensemble of nano-objects promising for the development of room-temperature coherent electronic devices.

cond-mat.mes-hall↗

arXiv2021

Wafer-scale, full-coverage, acoustic self-limiting assembly of particles on flexible substrates

Self-limiting assembly of particles represents the state-of-the-art controllability in nanomanufacturing processes where the assembly stops at a designated stage1,2, providing a desirable platform for applications requiring delicate thickness control3-5. Most successes in self-limiting assembly are limited to self-assembled monolayers (SAM) of small molecules on inorganic, chemically homogeneous rigid substrates (e.g., Au and SiO2) through surface-interaction mechanisms6,7. Similar mechanisms, however, cannot achieve uniform assembly of particles on flexible polymer substrates8,9. The complex configurations and conformations of polymer chains create a surface with non-uniform distributions of chemical groups and phases. In addition, most assembly mechanisms require good solvent wettability, where many desirable but hard-to-wet particles and polymer substrates are excluded. Here, we demonstrate an acoustic self-limiting assembly of particles (ASAP) method to achieve wafer-scale, full-coverage, close-packed monolayers of hydrophobic particles on hydrophobic polymer substrates in aqueous solutions. We envision many applications in functional coatings and showcase its feasibility in structural coloration.

physics.atm-clus↗

arXiv2021

Physics-Guided and Physics-Explainable Recurrent Neural Network for Time Dynamics in Optical Resonances

Understanding the time evolution of physical systems is crucial to revealing fundamental characteristics that are hidden in frequency domain. In optical science, high-quality resonance cavities and enhanced interactions with matters are at the heart of modern quantum technologies. However, capturing their time dynamics in real-world scenarios suffers from long data acquisition and low analysis accuracy due to slow convergence and limited time window. Here, we report a physics-guided and physics-explainable recurrent neural network to precisely forecast the time-domain response of resonance features with the shortest acquired input sequence being 7\% of full length, and to infer corresponding resonance frequencies. The model is trained in a two-step multi-fidelity framework for high-accuracy forecast, where the first step is based on a large amount of low-fidelity physical-model-generated synthetic data and second step involves a small set of high-fidelity application-oriented observational data. Through both simulations and experiments, we demonstrate that the model is universally applicable to a wide range of resonances, including dielectric metasurfaces, graphene plasmonics, and ultrastrongly coupled Landau polaritons, where our model accurately captures small signal features and learns essential physical quantities. The demonstrated machine learning algorithm offers a new way to accelerate the exploration of physical phenomena and the design of devices under resonance-enhanced light-matter interaction.

physics.optics↗

arXiv2021

Band Structure Dependent Electronic Localization in Macroscopic Films of Single-Chirality Single-Wall Carbon Nanotubes

Significant understanding has been achieved over the last few decades regarding chirality-dependent properties of single-wall carbon nanotubes (SWCNTs), primarily through single-tube studies. However, macroscopic manifestations of chirality dependence have been limited, especially in electronic transport, despite the fact that such distinct behaviors are needed for many applications of SWCNT-based devices. In addition, developing reliable transport theory is challenging since a description of localization phenomena in an assembly of nanoobjects requires precise knowledge of disorder on multiple spatial scales, particularly if the ensemble is heterogeneous. Here, we report an observation of pronounced chirality-dependent electronic localization in temperature and magnetic field dependent conductivity measurements on macroscopic films of single-chirality SWCNTs. The samples included large-gap semiconducting (6,5) and (10,3) films, narrow-gap semiconducting (7,4) and (8,5) films, and armchair metallic (6,6) films. Experimental data and theoretical calculations revealed Mott variable-range-hopping dominated transport in all samples, while localization lengths fall into three distinct categories depending on their band gaps. Armchair films have the largest localization length. Our detailed analyses on electronic transport properties of single-chirality SWCNT films provide significant new insight into electronic transport in ensembles of nanoobjects, offering foundations for designing and deploying macroscopic SWCNT solid-state devices.

cond-mat.mes-hall↗

arXiv2021

Real-time Multi-Task Diffractive Deep Neural Networks via Hardware-Software Co-design

Deep neural networks (DNNs) have substantial computational requirements, which greatly limit their performance in resource-constrained environments. Recently, there are increasing efforts on optical neural networks and optical computing based DNNs hardware, which bring significant advantages for deep learning systems in terms of their power efficiency, parallelism and computational speed. Among them, free-space diffractive deep neural networks (D$^2$NNs) based on the light diffraction, feature millions of neurons in each layer interconnected with neurons in neighboring layers. However, due to the challenge of implementing reconfigurability, deploying different DNNs algorithms requires re-building and duplicating the physical diffractive systems, which significantly degrades the hardware efficiency in practical application scenarios. Thus, this work proposes a novel hardware-software co-design method that enables robust and noise-resilient Multi-task Learning in D$^2$NNs. Our experimental results demonstrate significant improvements in versatility and hardware efficiency, and also demonstrate the robustness of proposed multi-task D$^2$NN architecture under wide noise ranges of all system components. In addition, we propose a domain-specific regularization algorithm for training the proposed multi-task architecture, which can be used to flexibly adjust the desired performance for each task.

cs.LG↗

arXiv2020

Artificial Intelligence Accelerators based on Graphene Optoelectronic Devices

Optical and optoelectronic approaches of performing matrix-vector multiplication (MVM) operations have shown the great promise of accelerating machine learning (ML) algorithms with unprecedented performance. The incorporation of nanomaterials into the system can further improve the performance thanks to their extraordinary properties, but the non-uniformity and variation of nanostructures in the macroscopic scale pose severe limitations for large-scale hardware deployment. Here, we report a new optoelectronic architecture consisting of spatial light modulators and photodetector arrays made from graphene to perform MVM. The ultrahigh carrier mobility of graphene, nearly-zero-power-consumption electro-optic control, and extreme parallelism suggest ultrahigh data throughput and ultralow-power consumption. Moreover, we develop a methodology of performing accurate calculations with imperfect components, laying the foundation for scalable systems. Finally, we perform a few representative ML algorithms, including singular value decomposition, support vector machine, and deep neural networks, to show the versatility and generality of our platform.

cs.ET↗

arXiv2020

Terahertz Excitonics in Carbon Nanotubes: Exciton Autoionization and Multiplication

Excitons play major roles in optical processes in modern semiconductors, such as single-wall carbon nanotubes (SWCNTs), transition metal dichalcogenides, and 2D perovskite quantum wells. They possess extremely large binding energies (>100~meV), dominating absorption and emission spectra even at high temperatures. The large binding energies imply that they are stable, that is, hard to ionize, rendering them seemingly unsuited for optoelectronic devices that require mobile charge carriers, especially terahertz emitters and solar cells. Here, we have conducted terahertz emission and photocurrent studies on films of aligned single-chirality semiconducting SWCNTs and find that excitons autoionize, i.e., spontaneously dissociate into electrons and holes. This process naturally occurs ultrafast (<1~ps) while conserving energy and momentum. The created carriers can then be accelerated to emit a burst of terahertz radiation when a dc bias is applied, with promising efficiency in comparison to standard GaAs-based emitters. Furthermore, at high bias, the accelerated carriers acquire high enough kinetic energy to create secondary excitons through impact exciton generation, again in a fully energy and momentum conserving fashion. This exciton multiplication process leads to a nonlinear photocurrent increase as a function of bias. Our theoretical simulations based on nonequilibrium Boltzmann transport equations, taking into account all possible scattering pathways and a realistic band structure, reproduce all our experimental data semi-quantitatively. These results not only elucidate the momentum-dependent ultrafast dynamics of excitons and carriers in SWCNTs but also suggest promising routes toward terahertz excitonics despite the orders-of-magnitude mismatch between the exciton binding energies and the terahertz photon energies.

cond-mat.mtrl-sci↗

arXiv2019

Groove-Assisted Global Spontaneous Alignment of Carbon Nanotubes in Vacuum Filtration

Ever since the discovery of carbon nanotubes (CNTs), it has long been a challenging goal to create macroscopically ordered assemblies, or crystals, of CNTs that preserve the one-dimensional quantum properties of individual CNTs on a macroscopic scale. Recently, a simple and well-controlled method was reported for producing wafer-scale crystalline films of highly aligned and densely packed CNTs through spontaneous global alignment that occurs during vacuum filtration [\textit{Nat.\ Nanotechnol}.\ \textbf{11}, 633 (2016)]. However, a full understanding of the mechanism of such global alignment has not been achieved. Here, we report results of a series of systematic experiments that demonstrate that the CNT alignment direction can be controlled by the surface morphology of the filter membrane used in the vacuum filtration process. More specifically, we found that the direction of parallel grooves pre-existing on the surface of the filter membrane dictates the direction of the resulting CNT alignment. Furthermore, we intentionally imprinted periodically spaced parallel grooves on a filter membranes using a diffraction grating, which successfully defined the direction of the global alignment of CNTs in a precise and reproducible manner.

physics.app-ph↗

arXiv2019

One-directional thermal transport in densely aligned single-wall carbon nanotube films

Individual carbon nanotubes (CNTs) possess extremely high thermal conductivities. However, the thermal conductivities and their anisotropy of macroscopic assemblies of CNTs have so far remained small. Here, we report results of directional thermal transport measurements on a nearly-perfectly aligned CNT film fabricated via controlled vacuum filtration. We found the thermal conductivity to be 43 +- 2.2 W m^-1 K^-1 with a record-high thermal anisotropy of 500. From the temperature dependence of the thermal conductivity and its agreement with the atomistic phonon transport calculation, we conclude that the effect of intertube thermal resistance on heat conduction in the alignment direction is negligible because of the large contact area between CNTs. These observations thus represent ideal unidirectional thermal transport, i.e., the thermal conductivity of the film is determined solely by that of individual CNTs.

physics.app-ph↗

arXiv2019

Terahertz Faraday and Kerr rotation spectroscopy of Bi$_{1-x}$Sb$_x$ films in high magnetic fields up to 30 Tesla

We report results of terahertz Faraday and Kerr rotation spectroscopy measurements on thin films of $\text{Bi}_{1-x}\text{Sb}_{x}$, an alloy system that exhibits a semimetal-to-topological-insulator transition as the Sb composition $x$ increases. By using a single-shot time-domain terahertz spectroscopy setup combined with a table-top pulsed mini-coil magnet, we conducted measurements in magnetic fields up to 30~T, observing distinctly different behaviors between semimetallic ($x < 0.07$) and topological insulator ($x > 0.07$) samples. Faraday and Kerr rotation spectra for the semimetallic films showed a pronounced dip that blue-shifted with the magnetic field, whereas spectra for the topological insulator films were positive and featureless, increasing in amplitude with increasing magnetic field and eventually saturating at high fields ($>$20~T). Ellipticity spectra for the semimetallic films showed resonances, whereas the topological insulator films showed no detectable ellipticity. To explain these observations, we developed a theoretical model based on realistic band parameters and the Kubo formula for calculating the optical conductivity of Landau-quantized charge carriers. Our calculations quantitatively reproduced all experimental features, establishing that the Faraday and Kerr signals in the semimetallic films predominantly arise from bulk hole cyclotron resonances while the signals in the topological insulator films represent combined effects of surface carriers originating from multiple electron and hole pockets. These results demonstrate that the use of high magnetic fields in terahertz magnetopolarimetry, combined with detailed electronic structure and conductivity calculations, allows us to unambiguously identify and quantitatively determine unique contributions from different species of carriers of topological and nontopological nature in Bi$_{1-x}$Sb$_x$.

cond-mat.mes-hall↗

arXiv2019

Macroscopically Aligned Carbon Nanotubes as a Refractory Platform for Hyperbolic Thermal Emitters

Refractory nanophotonics, or nanophotonics at high temperatures, can revolutionize many applications, including data storage and waste heat recovery. In particular, nanophotonic devices made from hyperbolic materials are promising due to their nearly infinite photonic density of states (PDOS). However, it is challenging to achieve a prominent PDOS in existing refractory hyperbolic materials, especially in a broad spectral range. Here, we demonstrate that macroscopic films and architectures of aligned carbon nanotubes work as excellent refractory hyperbolic materials. We found that aligned carbon nanotubes are thermally stable up to $1600^{\circ}$C and exhibit extreme anisotropy - metallic in one direction and insulating in the other two directions. Such extreme anisotropy makes this system a hyperbolic material with an exceptionally large PDOS over a broadband spectrum range (longer than $4.3μ$m) in the midinfrared, exhibiting strong resonances in deeply sub-wavelength-sized cavities. We observed polarized, spectrally selective thermal emission from aligned carbon nanotube films as well as indefinite cavities of aligned carbon nanotubes with volume as small as $\simλ^3/700$ operating at $700^{\circ}$C . These experiments suggest that aligned carbon nanotubes possess naturally large PDOS that leads to thermal photon densities enhanced by over two orders of magnitude, making them a promising refractory nanophotonics platform.

physics.optics↗

arXiv2018

Science and applications of wafer-scale crystalline carbon nanotube films prepared through controlled vacuum filtration

Carbon nanotubes (CNTs) make an ideal one-dimensional (1D) material platform for the exploration of exotic physical phenomena under extremely strong quantum confinement. The 1D character of electrons, phonons and excitons in individual CNTs features extraordinary electronic, thermal and optical properties. Since the first discovery, they have been continuing to attract interest in various disciplines, including chemistry, materials science, physics, and engineering. However, the macroscopic manifestation of such properties is still limited, despite significant efforts for decades. Recently, a controlled vacuum filtration method has been developed for the preparation of wafer-scale films of crystalline chirality-enriched CNTs, and such films immediately enable exciting new fundamental studies and applications. In this review, we will first discuss the controlled vacuum filtration technique, and then summarize recent discoveries in optical spectroscopy studies and optoelectronic device applications using films prepared by this technique.

cond-mat.mes-hall↗

arXiv2018

Direct Observation of Cross-Polarized Excitons in Aligned Single-Chirality Single-Wall Carbon Nanotubes

Optical properties of single-wall carbon nanotubes (SWCNTs) for light polarized parallel to the nanotube axis have been extensively studied, whereas their response to light polarized perpendicular to the nanotube axis has not been well explored. Here, by using a macroscopic film of highly aligned single-chirality (6,5) SWCNTs, we performed a systematic polarization-dependent optical absorption spectroscopy study. In addition to the commonly observed angular-momentum-conserving interband absorption of parallel-polarized light, which generates $E_{11}$ and $E_{22}$ excitons, we observed a small but unambiguous absorption peak whose intensity is maximum for perpendicular-polarized light. We attribute this feature to the lowest-energy cross-polarized interband absorption processes that change the angular momentum along the nanotube axis by $\pm 1$, generating $E_{12}$ and $E_{21}$ excitons. The energy difference between the $E_{12}$ and $E_{21}$ exciton peaks, expected from asymmetry between the conduction and valence bands, was smaller than the observed linewidth. Unlike previous observations of cross-polarized excitons in polarization-dependent photoluminescence and circular dichroism spectroscopy experiments, our direct observation using absorption spectroscopy allowed us to quantitatively analyze this resonance. Specifically, we determined the energy and oscillator strength of this resonance to be 1.54 and 0.05, respectively, compared with the values for the $E_{11}$ exciton peak. These values, in combination with comparison with theoretical calculations, in turn led to an assessment of the environmental effect on the strength of Coulomb interactions in this aligned single-chirality SWCNT film.

cond-mat.mes-hall↗

arXiv2016

Stability of High-Density Two-Dimensional Excitons against a Mott Transition in High Magnetic Fields Probed by Coherent Terahertz Spectroscopy

We have performed time-resolved terahertz absorption measurements on photoexcited electron-hole pairs in undoped GaAs quantum wells in magnetic fields. We probed both unbound- and bound-carrier responses via cyclotron resonance and intraexciton resonance, respectively. The stability of excitons, monitored as the pair density was systematically increased, was found to dramatically increase with increasing magnetic field. Specifically, the 1$s$-2$p_-$ intraexciton transition at 9 T persisted up to the highest density, whereas the 1$s$-2$p$ feature at 0 T was quickly replaced by a free-carrier Drude response. Interestingly, at 9 T, the 1$s$-2$p_-$ peak was replaced by free-hole cyclotron resonance at high temperatures, indicating that 2D magnetoexcitons do dissociate under thermal excitation, even though they are stable against a density-driven Mott transition.

cond-mat.quant-gas↗