arXiv subjects

Elizaveta Semenova

Publications and source records attributed to Elizaveta Semenova.

At least 37 records · Page 2Linked to original sources

arXiv2024

A nanolaser with extreme dielectric confinement

The interaction between light and matter can be enhanced by spatially concentrating the light field to boost the photon energy density and increasing the photon dwell time to prolong energy transfer between light and matter. Traditionally, strong spatial light localization has been achieved using plasmonics, which, despite its effectiveness, entails ohmic losses. Recent advances in nanostructured dielectrics offer an avenue for achieving strong light confinement without metallic losses. However, previous studies primarily focused on minimizing the optical mode volume without adequately addressing light-matter interactions. Here, we develop a nanolaser that simultaneously localizes the electromagnetic field and excited carriers within the same region of a dielectric nanobridge. This extreme dielectric confinement of both light and matter achieves a mode volume below the diffraction limit and a subwavelength carrier volume without the introduction of lateral quantum confinement, enabling continuous-wave lasing at room-temperature. Moreover, we observe a strong correlation between the mode field and carrier distribution, and unexpectedly, the enhanced mode field localization automatically leads to more pronounced carrier localization, promoting self-alignment of light and matter, which significantly reduces the laser threshold. We quantify the intensified light-matter interaction with a newly proposed interaction volume, which generalizes the concept of mode volume to a broad class of active media. Our work lays the ground for developing ultra-efficient optoelectronic devices by greatly enhancing light-matter interactions through advanced material nanostructuring.

physics.optics↗

arXiv2024

InAs/InP quantum dot based C-Band all-fiber plug-and-play triggered single-photon source integrated using micro-transfer printing

Fiber-based long-haul quantum communication would greatly benefit from a robust and deterministically integrated source of quantum state. Here, we report the design, fabrication, and optical characterization of InAs/InP quantum dots in the InP H1 point defect 2D photonic crystal cavity, integrated with the standard single-mode fiber using a micro-transfer printing technique. The device was placed in a compact cryocooler maintaining a cryogenic temperature of 15 K with single-photon emission characterized by $g^{(2)}(0)=0.14(14)$ and reliable and stable emission (intensity fluctuations given by a standard deviation $σ= 0.13$), so that an all-fiber based connection between two laboratory nodes through an open area was established and utilized for testing the quantum channel. In this way, we demonstrate a plug-and-play all-fiber single-photon source operating in the third telecom window, where standard telecommunication fiber networks can be used as a low-loss medium.

physics.optics↗

arXiv2024

Xenon plasma-focused ion beam milling as a method to deterministically fabricate bright and high-purity single-photon sources operating at C-band

Electron beam lithography is a standard method for fabricating photonic nanostructures around semiconductor quantum dots (QDs), which are crucial for efficient single and indistinguishable photon sources in quantum information processing. However, this technique lacks direct 3D control over the nanostructure shape, complicating the design and enlarging the 2D footprint to suppress in-plane photon leakage while directing photons into the collecting lens aperture. Here, we present an alternative approach to employ Xenon plasma-focused ion beam (Xe-PFIB) technology as a reliable method for the 3D shaping of photonic nanostructures containing low-density self-assembled InAs/InP quantum dots emitting in the C-band range of the 3rd telecommunication window. We explore both deterministic and non-deterministic fabrication approaches, resulting in mesas naturally shaped as truncated cones. As a demonstration, we fabricate mesas using a heterogeneously integrated structure with a QDs membrane atop an aluminum mirror and silicon substrate. Finite-difference time-domain (FDTD) simulations show that the angled sidewalls significantly increase photon extraction efficiency, achieving η = 0.89 for NA = 0.65. We demonstrate experimentally a high purity of pulsed single-photon emission (~99%) and a high extraction efficiency of η = 0.24, with the latter surpassing the highest reported values obtained using electron beam lithography in the C-band.

physics.optics↗

arXiv2024

Federated Learning for Non-factorizable Models using Deep Generative Prior Approximations

Federated learning (FL) allows for collaborative model training across decentralized clients while preserving privacy by avoiding data sharing. However, current FL methods assume conditional independence between client models, limiting the use of priors that capture dependence, such as Gaussian processes (GPs). We introduce the Structured Independence via deep Generative Model Approximation (SIGMA) prior which enables FL for non-factorizable models across clients, expanding the applicability of FL to fields such as spatial statistics, epidemiology, environmental science, and other domains where modeling dependencies is crucial. The SIGMA prior is a pre-trained deep generative model that approximates the desired prior and induces a specified conditional independence structure in the latent variables, creating an approximate model suitable for FL settings. We demonstrate the SIGMA prior's effectiveness on synthetic data and showcase its utility in a real-world example of FL for spatial data, using a conditional autoregressive prior to model spatial dependence across Australia. Our work enables new FL applications in domains where modeling dependent data is essential for accurate predictions and decision-making.

stat.ML↗

arXiv2024

High-throughput quantum photonic devices emitting indistinguishable photons in the telecom C-band

Single indistinguishable photons at telecom C-band wavelengths are essential for quantum networks and the future quantum internet. However, high-throughput technology for single-photon generation at 1550 nm remained a missing building block to overcome present limitations in quantum communication and information technologies. Here, we demonstrate the high-throughput fabrication of quantum-photonic integrated devices operating at C-band wavelengths based on epitaxial semiconductor quantum dots. Our technique enables the deterministic integration of single pre-selected quantum emitters into microcavities based on circular Bragg gratings. Respective devices feature the triggered generation of single photons with ultra-high purity and record-high photon indistinguishability. Further improvements in yield and coherence properties will pave the way for implementing single-photon non-linear devices and advanced quantum networks at telecom wavelengths.

quant-ph↗

arXiv2024

On-Demand Generation of Indistinguishable Photons in the Telecom C-Band using Quantum Dot Devices

Semiconductor quantum dots (QDs) enable the generation of single and entangled photons, useful for various applications in photonic quantum technologies. Specifically for quantum communication via fiber-optical networks, operation in the telecom C-band centered around 1550$\,$nm is ideal. The direct generation of QD-photons in this spectral range and with high quantum-optical quality, however, remained challenging. Here, we demonstrate the coherent on-demand generation of indistinguishable photons in the telecom C-band from single QD devices consisting of InAs/InP QD-mesa structures heterogeneously integrated with a metallic reflector on a silicon wafer. Using pulsed two-photon resonant excitation of the biexciton-exciton radiative cascade, we observe Rabi rotations up to pulse areas of $4π$ and a high single-photon purity in terms of $g^{(2)}(0)=0.005(1)$ and $0.015(1)$ for exciton and biexciton photons, respectively. Applying two independent experimental methods, based on fitting Rabi rotations in the emission intensity and performing photon cross-correlation measurements, we consistently obtain preparation fidelities at the $π$-pulse exceeding 80$\%$. Finally, performing Hong-Ou-Mandel-type two-photon interference experiments we obtain a photon-indistinguishability of the full photon wave packet of up to $35(3)\%$, representing a significant advancement in the photon-indistinguishability of single photons emitted directly in the telecom C-band.

quant-ph↗

arXiv2023

Heterogeneous integration of single InAs/InP quantum dots with the SOI chip using direct bonding

Quantum information processing with photons in small-footprint and highly integrated silicon-based photonic chips requires incorporating non-classical light sources. In this respect, self-assembled III-V semiconductor quantum dots (QDs) are an attractive solution, however, they must be combined with the silicon platform. Here, by utilizing the large-area direct bonding technique, we demonstrate the hybridization of InP and SOI chips, which allows for coupling single photons to the SOI chip interior, offering cost-effective scalability in setting up a multi-source environment for quantum photonic chips. We fabricate devices consisting of self-assembled InAs QDs embedded in the tapered InP waveguide (WG) positioned over the SOI-defined Si WG. Focusing on devices generating light in the telecom C-band compatible with the low-loss optical fiber networks, we demonstrate the light coupling between InP and SOI platforms by observing photons outcoupled at the InP-made circular Bragg grating outcoupler fabricated at the end of an 80 $μ$m-long Si WG, and at the cleaved edge of the Si WG. Finally, for a device with suppressed multi-photon generation events exhibiting 80% single photon generation purity, we measure the photon number outcoupled at the cleaved facet of the Si WG. We estimate the directional on-chip photon coupling between the source and the Si WG to 5.1%.

physics.optics↗

arXiv2023

PriorCVAE: scalable MCMC parameter inference with Bayesian deep generative modelling

Recent advances have shown that GP priors, or their finite realisations, can be encoded using deep generative models such as variational autoencoders (VAEs). These learned generators can serve as drop-in replacements for the original priors during MCMC inference. While this approach enables efficient inference, it loses information about the hyperparameters of the original models, and consequently makes inference over hyperparameters impossible and the learned priors indistinct. To overcome this limitation, we condition the VAE on stochastic process hyperparameters. This allows the joint encoding of hyperparameters with GP realizations and their subsequent estimation during inference. Further, we demonstrate that our proposed method, PriorCVAE, is agnostic to the nature of the models which it approximates, and can be used, for instance, to encode solutions of ODEs. It provides a practical tool for approximate inference and shows potential in real-life spatial and spatiotemporal applications.

stat.ML↗

arXiv2023

Deep learning and MCMC with aggVAE for shifting administrative boundaries: mapping malaria prevalence in Kenya

Model-based disease mapping remains a fundamental policy-informing tool in the fields of public health and disease surveillance. Hierarchical Bayesian models have emerged as the state-of-the-art approach for disease mapping since they are able to both capture structure in the data and robustly characterise uncertainty. When working with areal data, e.g.~aggregates at the administrative unit level such as district or province, current models rely on the adjacency structure of areal units to account for spatial correlations and perform shrinkage. The goal of disease surveillance systems is to track disease outcomes over time. This task is especially challenging in crisis situations which often lead to redrawn administrative boundaries, meaning that data collected before and after the crisis are no longer directly comparable. Moreover, the adjacency-based approach ignores the continuous nature of spatial processes and cannot solve the change-of-support problem, i.e.~when estimates are required to be produced at different administrative levels or levels of aggregation. We present a novel, practical, and easy to implement solution to solve these problems relying on a methodology combining deep generative modelling and fully Bayesian inference: we build on the recently proposed PriorVAE method able to encode spatial priors over small areas with variational autoencoders by encoding aggregates over administrative units. We map malaria prevalence in Kenya, a country in which administrative boundaries changed in 2010.

cs.LG↗

arXiv2023

Fine-tunable near-critical Stranski-Krastanov growth of InAs/InP quantum dots

Emerging applications of self-assembled semiconductor quantum dot (QD)-based nonclassical light sources emitting in the telecom C-band (1530 to 1565 nm) present challenges in terms of controlled synthesis of their low-density ensembles, critical for device processing with an isolated QD. This work shows how to control the surface density and size of InAs/InP quantum dots over a wide range by tailoring the conditions of Stranski-Krastanow growth. We demonstrate that in the near-critical growth regime, the density of quantum dots can be tuned between $10^7$ and $10^{10} cm^{-2}$. Furthermore, employing both experimental and modeling approaches, we show that the size (and therefore the emission wavelength) of InAs nanoislands on InP can be controlled independently from their surface density. Finally, we demonstrate that our growth method gives low-density ensembles resulting in well-isolated QD-originated emission lines in the telecom C-band.

cond-mat.mes-hall↗

arXiv2022

Electrically-driven Photonic Crystal Lasers with Ultra-low Threshold

Light sources with ultra-low energy consumption and high performance are required to realize optical interconnects for on-chip communication. Photonic crystal (PhC) nanocavity lasers are one of the most promising candidates to fill this role. In this work, we demonstrate an electrically-driven PhC nanolaser with an ultra-low threshold current of 10.2 μA emitting at 1540 nm and operated at room temperature. The lasers are InP-based bonded on Si and comprise a buried heterostructure active region and lateral p-i-n junction. The static characteristics and the thermal properties of the lasers have been characterized. The effect of disorder and p-doping absorption on the Q-factor of passive cavities was studied. We also investigate the leakage current due to the lateral p-i-n geometry by comparing the optical and electrical pumping schemes.

physics.optics↗

arXiv2022

PriorVAE: Encoding spatial priors with VAEs for small-area estimation

Gaussian processes (GPs), implemented through multivariate Gaussian distributions for a finite collection of data, are the most popular approach in small-area spatial statistical modelling. In this context they are used to encode correlation structures over space and can generalise well in interpolation tasks. Despite their flexibility, off-the-shelf GPs present serious computational challenges which limit their scalability and practical usefulness in applied settings. Here, we propose a novel, deep generative modelling approach to tackle this challenge, termed PriorVAE: for a particular spatial setting, we approximate a class of GP priors through prior sampling and subsequent fitting of a variational autoencoder (VAE). Given a trained VAE, the resultant decoder allows spatial inference to become incredibly efficient due to the low dimensional, independently distributed latent Gaussian space representation of the VAE. Once trained, inference using the VAE decoder replaces the GP within a Bayesian sampling framework. This approach provides tractable and easy-to-implement means of approximately encoding spatial priors and facilitates efficient statistical inference. We demonstrate the utility of our VAE two stage approach on Bayesian, small-area estimation tasks.

cs.LG↗

arXiv2022

Second-harmonic generation tuning by stretching arrays of GaAs nanowires

We present a wearable device with III-V nanowires in a flexible polymer, which is used for active mechanical tuning of the second-harmonic generation intensity. An array of vertical GaAs nanowires was grown with metalorganic vapour-phase epitaxy, then embedded in polydimethylsiloxane and detached from the rigid substrate with mechanical peel off. Experimental results show a tunability of the second-harmonic generation intensity by a factor of two for 30% stretching which matches the simulations including the distribution of sizes. We studied the impact of different parameters on the band dispersion and tunability of the second-harmonic generation, such as the pitch, the length, and the diameter. We predict at least three orders of magnitude active mechanical tuning of the nonlinear signal intensity for nanowire arrays. The flexibility of the array together with the resonant wavelength engineering make such structures perspective platforms for future bendable or stretchable nanophotonic devices as light sources or sensors.

physics.optics↗

arXiv2021

Bayesian workflow for disease transmission modeling in Stan

This tutorial shows how to build, fit, and criticize disease transmission models in Stan, and should be useful to researchers interested in modeling the SARS-CoV-2 pandemic and other infectious diseases in a Bayesian framework. Bayesian modeling provides a principled way to quantify uncertainty and incorporate both data and prior knowledge into the model estimates. Stan is an expressive probabilistic programming language that abstracts the inference and allows users to focus on the modeling. As a result, Stan code is readable and easily extensible, which makes the modeler's work more transparent. Furthermore, Stan's main inference engine, Hamiltonian Monte Carlo sampling, is amiable to diagnostics, which means the user can verify whether the obtained inference is reliable. In this tutorial, we demonstrate how to formulate, fit, and diagnose a compartmental transmission model in Stan, first with a simple Susceptible-Infected-Recovered (SIR) model, then with a more elaborate transmission model used during the SARS-CoV-2 pandemic. We also cover advanced topics which can further help practitioners fit sophisticated models; notably, how to use simulations to probe the model and priors, and computational techniques to scale-up models based on ordinary differential equations.

stat.CO↗

arXiv2021

Bright Quantum Dot Single-Photon Emitters at Telecom Bands Heterogeneously Integrated on Si

Whereas the Si photonic platform is highly attractive for scalable optical quantum information processing, it lacks practical solutions for efficient photon generation. Self-assembled semiconductor quantum dots (QDs) efficiently emitting photons in the telecom bands ($1460-1625$ nm) allow for heterogeneous integration with Si. In this work, we report on a novel, robust, and industry-compatible approach for achieving single-photon emission from InAs/InP QDs heterogeneously integrated with a Si substrate. As a proof of concept, we demonstrate a simple vertical emitting device, employing a metallic mirror beneath the QD emitter, and experimentally obtained photon extraction efficiencies of $\sim10\%$. Nevertheless, the figures of merit of our structures are comparable with values previously only achieved for QDs emitting at shorter wavelength or by applying technically demanding fabrication processes. Our architecture and the simple fabrication procedure allows for the demonstration of a single-photon generation with purity $\mathcal{P}>98\%$ at the liquid helium temperature and $\mathcal{P}=75\%$ at $80$ K.

cond-mat.mes-hall↗

arXiv2021

Droplet epitaxy symmetric InAs/InP quantum dots for quantum emission in the third telecom window: morphology, optical and electronic properties

The rapidly developing quantum communication technology requires deterministic quantum emitters that can generate single photons and entangled photon pairs in the third telecom window, in order to be compatible with existing optical fiber networks and on-chip silicon photonic processors. InAs/InP quantum dots (QDs) are among the leading candidates for this purpose, due to their high emission efficiency in the required spectral range. However, fabricating versatile InAs/InP QD-based quantum emitters is challenging, especially as these QDs typically have asymmetric profiles in the growth plane, resulting in a substantial bright-exciton fine structure splitting (FSS). This hinders the generation of entangled photon pairs and thus, compromises the versatility of InAs/InP QDs. We overcome this by implementing droplet epitaxy (DE) synthesis of low surface density (2.8$\times$10$^8$ cm$^{-2}$) InAs QDs on an (001)-oriented InP substrate. The resulting QDs are located in etched pits, have concave bases, and most importantly, have symmetric in-plane profiles. We provide an analytical model to explain the kinetics of pit formation and QD base shape modification. Our theoretical calculations of electronic states reveal the properties of neutral and charged excitons and biexcitons confined in such QDs, which agree with the optical investigations of individual QDs. The optical response of QD ensembles suggests that FSS may indeed be negligible, as reflected in the vanishing degree of linear polarization. However, single QD spectrum gathered from an etched mesa shows moderate FSS of (50$\pm$5) $μ$eV that we link to destructive changes made in the QD environment during the post-growth processing. Finally, we show that the studied DE QDs provide a close-to-ideal single-photon emission purity of (92.5$\pm$7.5) $\%$ in the third telecom window.

cond-mat.mes-hall↗

arXiv2020

Monolithic integration of InP on Si by molten alloy driven selective area epitaxial growth

We report a new approach for monolithic integration of III-V materials into silicon, based on selective area growth and driven by a molten alloy in metal-organic vapor epitaxy. Our method includes elements of both selective area and droplet-mediated growths and combines the advantages of the two techniques. Using this approach, we obtain organized arrays of high crystalline quality InP insertions into (100) oriented Si substrates. Our detailed structural, morphological and optical studies reveal the conditions leading to defect formation. These conditions are then eliminated to optimize the process for obtaining dislocation-free InP nanostructures grown directly on Si and buried below the top surface. The PL signal from these structures exhibits a narrow peak at the InP bandgap energy. The fundamental aspects of the growth are studied by modeling the InP nucleation process. The model is fitted by our x-ray diffraction measurements and correlates well with the results of our transmission electron microscopy and optical investigations. Our method constitutes a new approach for the monolithic integration of active III-V material into Si platform and opens up new opportunities in active Si photonics.

physics.app-ph↗

arXiv2019

Tunable MEMS VCSEL on Silicon substrate

We present design, fabrication and characterization of a MEMS VCSEL which utilizes a silicon-on-insulator wafer for the microelectromechanical system and encapsulates the MEMS by direct InP wafer bonding, which improves the protection and control of the tuning element. This procedure enables a more robust fabrication, a larger free spectral range and facilitates bidirectional tuning of the MEMS element. The MEMS VCSEL device uses a high contrast grating mirror on a MEMS stage as the bottom mirror, a wafer bonded InP with quantum wells for amplification and a deposited dielectric DBR as the top mirror. A 40 nm tuning range and a mechanical resonance frequency in excess of 2 MHz are demonstrated.

physics.app-ph↗