Achieving high-power performance at long wavelengths has been a longstanding goal in the field of nitride LEDs, however it poses significant challenges. These include declining efficiency with increasing wavelength, undesired hue shifts with increasing current, and the difficulty of maintaining material quality while incorporating high indium compositions in quantum wells. In this work, I employ predictive modeling techniques to shed light on these hurdles, informing the design of devices that effectively overcome these obstacles. In particular, I will demonstrate how the efficiency-droop and hue-shift problems in the green spectral range are caused by an increase in the operating carrier density rather than material degradation. Then, I will discuss how progress towards longer red wavelengths is hindered by the difficulty in incorporating higher indium concentrations in quantum wells. To alleviate these issues, I explore the high-dimensional configurational landscape of quaternary III-nitride emitters using statistical-learning techniques, and identify promising designs that emit in the red spectral range.

Towards a multiscale model of nanostructured optoelectronic devices, we have proposed a nonequilibrium Green’s function approach, with intraband scattering self-energies computed in the self-consistent Born approximation, and interband self-energies (Shockley-Read-Hall and optical transitions) included in terms of semiclassical generation-recombination rates. This model provides a quantum-kinetic description of tunneling, miniband transport, hopping, and carrier extraction in terms of quasi-Fermi levels and electric fields, germane to the drift-diffusion framework. Another innovative aspect of the model is the numerical evaluation of functional derivatives, which led us to the development of small-signal AC NEGF model of dissipative carrier transport in semiconductor nanostructures.

In this presentation, we propose an effective scattering-potential approach for treating interface-roughness scattering of moving electrons in a superlattice structure. Based on obtained effective scattering potentials, we further derive a generalized Boltzmann transport equation by including a self-consistent internal scattering force. In addition, we solve this equation exactly beyond the relaxation-time approximation, and meanwhile, analyze the dependence of conduction current on interface-roughness parameters at various temperatures and DC electric fiield strengths. Finally, we reveal a microscopic mechanism associated with non-ohmic transport behavior by analyzing features in steady-state non-equilibrium electron occupation function and its dependence on interface roughness parameters.

Using nano-cathodoluminescence performed in scanning transmission electron microscope (STEM-CL), we have investigated a photonic-bandgap-crystal (PBC) laser structure at T = 17 K. In cross-sectional STEM images the full device structure is clearly resolved. The most dominant luminescence originates from the 3-fold MQW of the active region. The MQW shows a distinct peak wavelength change in growth direction indicating different structural and/or chemical properties of the individual quantum wells. In detail, a clear shift from 427 nm to 438 nm from the first to the top QW is observed, respectively.

Scaling quantum computing while maintaining quantum coherence at cryogenic temperatures is still a challenging issue. It emphasizes the need for an optical link between the control processor in the cryogenic environments and memory units kept in room temperature to mitigate thermal noise-induced decoherence. This study examines the utilization of VCSELs as an optical link in cryogenic environments. The study explores microcavity-gain resonance conditions and their temperature dependency, develops electrical models considering limited thermionic emission, and analyzes internal thermal profiles during low-temperature operation. The research includes characterizing fabricated devices and addressing key factors, such as p-doped DBR, that limit energy-efficient performance.

Accurately determining numerical values for key model parameters for any semiconductor devices is extremely important for analyzing the device characteristics and model-based device design optimization. However, their experimental determination can be very difficult since measurement results involve interaction of many parameters and isolating the influence of a single parameter is often not possible. One of the ways to solve this issue is deep learning. We achieve accurate determination of key laser diode model parameters such as internal loss, Auger coefficient, and free-carrier absorption coefficient of a fabricated ridge-waveguide 850 nm GaAs/AlGaAs laser diode(LD) applying the trained deep neural network (DNN). We use a LD TCAD simulator, PICS3D, for producing training and testing data. The accuracy of our approach is confirmed by comparing the simulation result with the actual measurement result for the LD L-I characteristics using extracted model parameters by DNN.

Continuous-wave (cw) operation of integrated nanoscale lasers is a key ingredient for on-chip optical processing schemes in Si photonic circuits. Here, we demonstrate cw-lasing from individual InAs nanowires at mid-infrared wavelengths (2.4-2.7 µm) without any external cavity for mode confinement. Using finite difference time domain (FDTD) modelling of the threshold gain, optimal single Fabry-Perot nanowire laser geometries with diameter > 800 nm and lengths of 10-30 µm are realized by site-selective growth methods. Corresponding nanowires exhibit cw-lasing with thresholds around 10-30 kW/cm2 at lasing emission up to 70K.

We demonstrate a new generation of composition-tuned, ternary GaAsSb nanowire lasers on silicon with emission wavelengths tuned to below the Si bandgap. By solving previous limitations in the growth of III-As-Sb based nanowires, resonator cavities with extended lengths > 7 µm and high Sb-content (~30%) are realized as a base for bulk-type or quantum-well based nanowire lasers. Bulk GaAsSb nanowire lasers with high radiative efficiency and low threshold are enabled by use of lattice-matched InAlGaAs surface passivation layers. Coaxial InGaAs multi-quantum well (MQW) active regions grown on GaAsSb nanowire templates open further scope of tailoring material gain and lasing wavelength.

Aeluma has developed breakthrough technology to manufacture high-performance compound semiconductor devices, such as photodetectors and lasers, on large-diameter substrates. This path to scaling and cost reduction could enable broad market adoption of these high-performance technologies. Aeluma’s offerings include photodetectors, photodetector arrays, and lasers for silicon photonics. Key to the technology is the ability to deposit compound semiconductor device structures on mismatched substrates including 12-inch Silicon, to subsequently manufacture devices with large-scale microelectronics foundries, and to integrate optical devices with electronics using either wafer-scale packaging or direct monolithic integration.

The role of optical feedback in a Talbot based cavity has been studied for many years in laser arrays. Yet, those studies mainly deal with improving (or stabilizing) the global, slow-timescale properties of the laser array such as far-field pattern and optical spectrum. In our work, we shed a new light on Talbot based cavity feedback by exploring the fast-timescale properties of each lasers in the array and by studying destabilization instead of stabilization. We experimentally show that a 1D-array of semiconductor lasers emitting around 445 nm exhibits sub-ns intensity fluctuations that can be partially synchronized between lasers.

Two monolithically integrated MLLDs operating at telecom wavelength are compared. A quantum dot (QD) MLLD with an RF of about 50.1 GHz and a quantum well (QW) MLLD with an RF of about 51.4 GHz. The RF tunability is characterized by sweeping the pump current of the laser, the temperature, and the reverse voltage of the saturable absorber (SA). The QW-MLLD has a tuning range of 31 MHz with an average repetition frequency (RF) linewidth of 53 kHz, while the QD-MLLD has a smaller tuning range of 26 MHz with a higher average RF linewidth of 172 kHz. Both MLLDs are used in a terahertz time-domain spectroscopy (THz-TDS) setup with asynchronous optical sampling (ASOPS) using two lasers with slightly different RFs. One laser serves as a pump laser and the other as a probe laser. The measurement speed can be set directly by the RF difference and is limited only by data acquisition and stability, allowing ultra-fast measurement speeds with a very compact system.

Quantum dot lasers are expected to have excellent radiation hardness since carriers are spatially localized within the quantum dots and unable to move freely in-plane and interact with radiation-induced defects. Indeed, early experimental observations of reduced threshold current increase in quantum dot lasers with respect to quantum well lasers grown on native substrates were reported for heavy ion and proton environments. We recently observed the robustness of InAs quantum dot lasers grown on silicon and irradiated in neutron environments thus demonstrating radiation hardness also extends to quantum dot lasers grown on non-native substrates. We will discuss our experimental work and modeling effort to understand how the electronic structure of the quantum dot system impacts radiation hardness.

Optical frequency combs have become essential components in a wide variety of technologies, with semiconductor laser diodes becoming increasingly relevant due to their seamless integration with photonic circuitry. We study the high temperature operation of a single-mode, passively mode-locked ridge waveguide laser based on InAs/InP quantum dashes designed for C-band operation. We present experimental evidence of stable optical frequency comb generation at temperatures up to 85°C, with peak widths of the repetition rate below 10 MHz. When deployed in high temperature environments, these devices can offer power savings of up to 80%.

A parametric study was conducted on coupled-cavity on-chip lasers to investigate the feasibility of reducing the lasing linewidth. The study showed that the coupled-cavity structure achieved up to 7 orders of magnitude linewidth reduction. Increasing the number of QW/QD layers (or QD density-per-layer) resulted in higher optical power and narrower linewidths. However, in the QW case, increasing the layers reduced efficiency and increased the input-power requirement for locking, while in the QD case, increasing the QD layers/density increased the efficiency and decreased the input-power requirement. The study recommends minimizing the number of QW layers and maximizing the number of QD layers at moderate and low current injection, respectively.

Metallic nanoparticles are of grown interest for scientific and technologic applications. In these systems, the interaction between electromagnetic fields and the free electrons forms a collective localized surface plasmon resonance (LSPR). The decay of this excited LSPR through Landau damping results in a nonthermal distribution of electrons and holes, so-called hot carriers. The generation and harnessing of hot electrons and holes in metallic nanoparticles is of significant interest for applications in photocatalysis, photovoltaics and sensing. Bimetallic nanostructures have shown outstanding properties for enhancing plasmonic hot carrier generation as they contain plasmonic metals which afford highly efficient conversion of electromagnetic radiation into hot carriers. In this work, we study the plasmonic hot-carrier generation in magnesium nanoparticles with metal atoms dopants (Ag, Au, Cu, Pd and Pt) and the dependence of the generation rates of these hot holes and electrons on the number of atoms dopants using first-principles time-dependent density functional theory (TD-DFT) approach.

Highly mismatched alloys (HMAs) – semiconductors where the alloying elements have highly different electronegativity – are used in solar cells and lasers due to their highly tunable band gaps. It was recently shown that doped HMAs also have widely tunable plasmon resonances, giving them plasmon resonances from the THz to the mid-IR [1]. We evaluate the potential of doped HMAs for mid-IR plasmonics by predicting their bulk plasmon frequency and associated losses. We characterize the quality of the plasmon resonance using a figure of merit |χ|^2/Im[χ], where χ is the electric susceptibility, which describes the potential of a material to scatter or absorb light [2]. We use a Green’s function method to develop a theory for the susceptibility of HMAs with CB anticrossing. We use a relaxation time approximation (RTA) to model material losses, with scattering times extracted from mobility measurements. From this model susceptibility, we predict HMAs with mid-IR plasmon resonances of comparable quality to other leading mid-IR plasmonic materials. [1] Allami and Krich, Phys Rev B (2021) [2] Miller et al., Optics Express (2016) [3] Mermin, Phys Rev B (1970)

We engineer an Aluminum (Al)-based plasmonic device coated with TiO₂ and SiO₂ layers for biosensing applications. First, the thicknesses of TiO₂, SiO₂, and Al layers are optimized under the angle interrogation scheme for a wavelength of 1550 nm. Over an optimized value of TiO₂, SiO₂, and Al film thickness, the variation trend in the performance parameters is studied for a range of thicknesses of 2D nanomaterials for the biofunctionalization of the sensing surface. Later, with the optimized intermediate layers, we present a comparative analysis of Al-based Kretschmann configuration with Graphene, MoS₂, MXene, and Fluorinated Graphene. It is observed that the TiO₂-SiO₂-Al-Fluorinated Graphene-based plasmonic device provides enhanced sensing parameters (sensitivity =120°/RIU, Figure of Merit = 430 RIU^-1).

This poster describes a new end-to-end virtual prototype solution we have developed for simulating the performances of the whole system of a CMOS Image Sensor Camera from the imaging lens system to the final image, through the optoelectronic sensor itself. Finite Difference Time Domain (FDTD) software is used to simulate how much light is absorbed by the CMOS sensor structure and the diffraction effects throughout the micro-lenses and pixels. 3D Charge Transport Solver is used to compute the probability to capture a photogenerated charge and get the quantum efficiency as a function of the incident angles, wavelengths, and pixel position. Finally, we combine light exposure onto the sensor from 3D environment raytracing software with quantum efficiency from photonics simulations to generate raw images and compute the final image.

High-quality Bi₂Se₃ thin-films laterally grown on a patterned sapphire substrate (PSS) have been developed using a catalyst-free vapor-phase transport deposition. Here, the structural properties of Bi₂Se₃ thin-films are greatly improved by covering the surface with Bi metal having a relatively low vaporization rate. Thus, the photoresponse of the fabricated IR conversion device using the Bi₂Se₃/PSS heterostrucutre exhibits excellent performance and high reliability with no degradation after continuous switching. Also, the Bi₂Se₃/PSS reveals a significantly higher near IR induced-photocurrent and a much faster photo-switching than the Bi₂Se₃ on a planar sapphire substrate.