Electrically pumped topological laser with valley edge modes.

Quantum cascade lasers are compact, electrically pumped light sources in the technologically important mid-infrared and terahertz region of the electromagnetic spectrum1,2. Recently, the concept of topology3 has been expanded from condensed matter physics into photonics4, giving rise to a new type of lasing5-8 using topologically protected photonic modes that can efficiently bypass corners and defects4. Previous demonstrations of topological lasers have required an external laser source for optical pumping and have operated in the conventional optical frequency regime5-8. Here we demonstrate an electrically pumped terahertz quantum cascade laser based on topologically protected valley edge states9-11. Unlike topological lasers that rely on large-scale features to impart topological protection, our compact design makes use of the valley degree of freedom in photonic crystals10,11, analogous to two-dimensional gapped valleytronic materials12. Lasing with regularly spaced emission peaks occurs in a sharp-cornered triangular cavity, even if perturbations are introduced into the underlying structure, owing to the existence of topologically protected valley edge states that circulate around the cavity without experiencing localization. We probe the properties of the topological lasing modes by adding different outcouplers to the topological cavity. The laser based on valley edge states may open routes to the practical use of topological protection in electrically driven laser sources.

Long wavelength mid-infrared multi-gases spectroscopy using tunable single-mode slot waveguide quantum cascade laser.

Single-mode tunable quantum cascade lasers (QCLs) are promising for high-resolution and highly sensitive trace gases sensing across the mid-infrared (MIR) region. We report on the development of a tunable single-mode slot waveguide QCL array in the long wavelength part of the MIR regime (>12 µm). This laser array exhibits a tuning range of around 12 cm-1, from 735.3 to 747.3 cm-1. Using this developed single-mode tunable QCL, we demonstrate individual gas sensing, yielding the detection limit of 940 ppb and 470 ppb for acetylene and o-xylene, respectively. To verify the potential of the developed QCL array in multi-species gas detection, laser absorption measurements of two mixed gases of acetylene and o-xylene were conducted, showing the absorption features of the corresponding gases agree well with the theoretical predictions.

Impact of Cavity Geometry on Microlaser Dynamics.

We experimentally investigate spatiotemporal lasing dynamics in semiconductor microcavities with various geometries, featuring integrable or chaotic ray dynamics. The classical ray dynamics directly impacts the lasing dynamics, which is primarily determined by the local directionality of long-lived ray trajectories. The directionality of optical propagation dictates the characteristic length scales of intensity variations, which play a pivotal role in nonlinear light-matter interactions. While wavelength-scale intensity variations tend to stabilize lasing dynamics, modulation on much longer scales causes spatial filamentation and irregular pulsation. Our results will pave the way to control the lasing dynamics by engineering the cavity geometry and ray dynamical properties.

Efficient Avalanche Photodiodes with a WSe2/MoS2 Heterostructure via Two-Photon Absorption.

Two-dimensional (2D) materials-based photodetectors in the infrared range hold the key to enabling a wide range of optoelectronics applications including infrared imaging and optical communications. While there exist 2D materials with a narrow bandgap sensitive to infrared photons, a two-photon absorption (TPA) process can also enable infrared photodetection in well-established 2D materials with large bandgaps such as WSe2 and MoS2. However, most of the TPA photodetectors suffer from low responsivity, preventing this method from being widely adopted for infrared photodetection. Herein, we experimentally demonstrate 2D materials-based TPA avalanche photodiodes achieving an ultrahigh responsivity. The WSe2/MoS2 heterostructure absorbs infrared photons with an energy smaller than the material bandgaps via a low-efficiency TPA process. The significant avalanche effect with a gain of ∼1300 improves the responsivity, resulting in the record-high responsivity of 88 µA/W. We believe that this work paves the way toward building practical and high-efficiency 2D materials-based infrared photodetectors.

Widely tunable single-mode slot waveguide quantum cascade laser array.

We report designs and experimental demonstrations of a widely tunable single-mode quantum cascade laser array based on slot waveguide structures in the mid-infrared region. The laser array device realized a continuous tuning range of 71 cm-1 from 9.66 µm to 10.37 µm at 300 K only using the current tuning without any external heatsink temperature adjustments, in good agreement with the design. Stable single-mode operations free of undesired mode-hops have been obtained over the whole tuning range. Another slot waveguide QCL array with a 41 cm-1 continuous tuning range around 7.3 µm has also been realized with the same design principle, demonstrating the universal applicability of the array design. The broadly continuous tuning with simple processing makes the array device a suitable candidate for mid-infrared sensing and spectroscopy application.

Beam combining of a broadly and continuously tunable quantum cascade laser.

We report a cost-efficient method to demonstrate the beam combining of five laser elements in an array of tunable slot waveguide quantum cascade lasers in the mid-infrared region at around 10 µm. An aspherical lens with five fine-tuned mini mirrors was employed to collimate the individual beams from the laser array. To verify the feasibility of this beam combining approach, the combined beams were coupled into a hollow-core fiber gas cell with a low numerical aperture (N.A.) of 0.03 and a coupling efficiency >= 0.82, for gas sensing of binary compound gases of ammonia and ethylene simultaneously.

High-harmonic generation in polycrystalline CdTe nano-films via macroscopic investigations.

Bright high harmonics generation (HHG) in CMOS-compatible nano-films can provide new opportunities for integrated coherent ultra-violet sources and attosecond photonic devices. Up to now, most HHG studies have been limited to single crystals. Polycrystalline materials, which consist of many grains separated by grain boundaries and normally have random crystallographic orientations, have rarely been explored for HHG. Understanding and predicting the HHG properties in polycrystalline nano-films are important owing to its merits of low cost and diversified properties, but challenging due to their complicated electronic structures. Here, we for the first time experimentally discover the correspondence between HHG in polycrystalline matters and macroscopic material parameters, to the best of our knowledge. Pumped by a mid-infrared femtosecond laser centered at 7.1 µm wavelength, bright and long-term stable harmonics extending to 25th orders (284 nm) are demonstrated in polycrystalline cadmium telluride (CdTe) nano-films. It is found that the HHG strengths in the transmission and the reflection behave differently as a function of the material thickness in the range from 6 nm to 4 µm, which is highly correlated to the measured macroscopic conductivity. The experimental findings agree well with the recent theoretical prediction [Phys. Rev. B103(15), 155426 (2021)10.1103/PhysRevB.103.155426]. This work provides a simple gauge to study and predict HHG in complicated polycrystalline and amorphous nano-systems, and paves the way for novel strong-field nanophotonics based on polycrystalline nano-films.

Triaxially strained suspended graphene for large-area pseudo-magnetic fields.

Strain-engineered graphene has garnered much attention recently owing to the possibilities of creating substantial energy gaps enabled by pseudo-magnetic fields (PMFs). While theoretical works proposed the possibility of creating large-area PMFs by straining monolayer graphene along three crystallographic directions, clear experimental demonstration of such promising devices remains elusive. Herein, we experimentally demonstrate a triaxially strained suspended graphene structure that has the potential to possess large-scale and quasi-uniform PMFs. Our structure employs uniquely designed metal electrodes that function both as stressors and metal contacts for current injection. Raman characterization and tight-binding simulations suggest the possibility of achieving PMFs over a micrometer-scale area. Current-voltage measurements confirm an efficient current injection into graphene, showing the potential of our devices for a new class of optoelectronic applications. We also theoretically propose a photonic crystal-based laser structure that obtains strongly localized optical fields overlapping with the spatial area under uniform PMFs, thus presenting a practical route toward the realization of graphene lasers.

Anomalous Single-Mode Lasing Induced by Nonlinearity and the Non-Hermitian Skin Effect.

Single-mode operation is a desirable but elusive property for lasers operating at high pump powers. Typically, single-mode lasing is attainable close to threshold, but increasing the pump power gives rise to multiple lasing peaks due to inter-modal gain competition. We propose a laser with the opposite behavior: multimode lasing occurs at low output powers, but pumping beyond a certain value produces a single lasing mode, with all other candidate modes experiencing negative effective gain. This phenomenon arises in a lattice of coupled optical resonators with non-fine-tuned asymmetric couplings, and is caused by an interaction between nonlinear gain saturation and the non-Hermitian skin effect. The single-mode lasing is observed in both frequency domain and time domain simulations. It is robust against on-site disorder, and scales up to large lattice sizes. This finding might be useful for implementing high-power laser arrays.

A Review of Machine Learning for Near-Infrared Spectroscopy.

The analysis of infrared spectroscopy of substances is a non-invasive measurement technique that can be used in analytics. Although the main objective of this study is to provide a review of machine learning (ML) algorithms that have been reported for analyzing near-infrared (NIR) spectroscopy from traditional machine learning methods to deep network architectures, we also provide different NIR measurement modes, instruments, signal preprocessing methods, etc. Firstly, four different measurement modes available in NIR are reviewed, different types of NIR instruments are compared, and a summary of NIR data analysis methods is provided. Secondly, the public NIR spectroscopy datasets are briefly discussed, with links provided. Thirdly, the widely used data preprocessing and feature selection algorithms that have been reported for NIR spectroscopy are presented. Then, the majority of the traditional machine learning methods and deep network architectures that are commonly employed are covered. Finally, we conclude that developing the integration of a variety of machine learning algorithms in an efficient and lightweight manner is a significant future research direction.

Tailorable infrared emission of microelectromechanical system-based thermal emitters with NiO films for gas sensing.

Infrared gas sensors hold great promise in the internet of things and artificial intelligence. Making infrared light sources with miniaturized size, reliable and tunable emission is essential but remains challenging. Herein, we present the tailorability of radiant power and the emergence of new emission wavelength of microelectromechanical system (MEMS)-based thermal emitters with nickel oxide (NiO) films. The coating of NiO on emitters increases top surface emissivity and induces the appearance of new wavelengths between 15 and 19 µm, all of which have been justified by spectroscopic methods. Furthermore, a sensor array is assembled for simultaneous monitoring of concentrations of carbon dioxide (CO2), methane (CH4), humidity, and temperature. The platform shows selective and sensitive detection at room temperature toward CO2 and CH4 with detection limits of around 50 and 1750 ppm, respectively, and also shows fast response/recovery and good recyclability. The demonstrated emission tailorability of MEMS emitters and their usage in sensor array provide novel insights for designing and fabricating optical sensors with good performance, which is promising for mass production and commercialization.

W-type normal dispersion thulium-doped fiber-based high-energy all-fiber femtosecond laser at 1.7 µm.

We propose a parabolic W-type thulium-doped fiber for the 1.7 µm high-energy femtosecond pulsed laser. Despite its attractive normal dispersion, the fiber offers high gain in 1.7 µm region thanks to its distributed short-pass filtering effect. With a proper dispersion management in an all-fiber chirped pulse amplification (CPA) scheme, we demonstrate so far the highest pulse energy of 128.0 nJ in a stable pulse of 174 fs in the 1.7-1.8 µm region, which marks above an order of magnitude improvement in pulse energy while exhibiting the shortest pulse duration among fiber-based CPA works at 1.7 µm. Hence, we provide a pathway to an energy scalable and efficient femtosecond laser at 1.7 µm via a compact and elegant all-fiber solution.

Self-gating in semiconductor electrocatalysis.

The semiconductor-electrolyte interface dominates the behaviours of semiconductor electrocatalysis, which has been modelled as a Schottky-analogue junction according to classical electron transfer theories. However, this model cannot be used to explain the extremely high carrier accumulations in ultrathin semiconductor catalysis observed in our work. Inspired by the recently developed ion-controlled electronics, we revisit the semiconductor-electrolyte interface and unravel a universal self-gating phenomenon through microcell-based in situ electronic/electrochemical measurements to clarify the electronic-conduction modulation of semiconductors during the electrocatalytic reaction. We then demonstrate that the type of semiconductor catalyst strongly correlates with their electrocatalysis; that is, n-type semiconductor catalysts favour cathodic reactions such as the hydrogen evolution reaction, p-type ones prefer anodic reactions such as the oxygen evolution reaction and bipolar ones tend to perform both anodic and cathodic reactions. Our study provides new insight into the electronic origin of the semiconductor-electrolyte interface during electrocatalysis, paving the way for designing high-performance semiconductor catalysts.

20 W, 2 mJ, sub-ps, 258†nm all-solid-state deep-ultraviolet laser with up to 3 GW peak power.

We demonstrate an all-solid-state deep-ultraviolet (DUV) laser based on the frequency-quadrupling of a 1 µm, 1.2 ps, Yb: YAG Innoslab solid-state laser at a 10 kHz repetition rate, using LBO and BBO as second-harmonic generation and fourth-harmonic generation crystals, respectively. The DUV laser delivers 20 W, 2.0 mJ, 665 fs, 258 nm DUV pulses, with an overall conversion efficiency of ∼8.7% from 1 µm to DUV. The corresponding peak power of DUV pulses is up to 3 GW, which, to the best of our knowledge, is highest in reported kHz-rate all-solid-state DUV sources driven at 1 µm wavelength.

Far-field controllable excitation of phonon polariton via nanostructure engineering.

Hexagonal boron nitride (h-BN) as a natural mid-infrared (mid-IR) hyperbolic material which supports a strong excitation of phonon polariton (PhP) has enabled a new class of photonic devices with unprecedented functionalities. The hyperbolic property of h-BN has not only brought in new physical insights but also spurred potential applications. However, most of the current h-BN devices are designed repying on near-field excitation and manipulation of PhP. For fully realizing the potentials of h-BN, research on far-field controllable excitation and control of PhP is important for future integrated photonic devices. In this work, we exploit the designs of controllable far-field excitation of PhP in nanostructure-patterned h-BN thin film for deep subwavelength focusing (FWHM∼λ0/14.9) and interference patterns of 1D (FWHM∼λ0/52) and 2D standing waves (FWHM∼λ0/36.8) which find great potential for super-resolution imaging beyond diffraction limit. These polaritonic patterns could be easily tuned remotely by manipulating the polarization and phase of incident laser. This approach provides a novel platform for practical IR nanophotonic devices and potential applications in mid-IR bio-imaging and sensing.

All-fiber short-wavelength tunable mode-locked fiber laser using normal dispersion thulium-doped fiber.

We report an all-fiber high pulse energy ultrafast laser and amplifier operating at the short wavelength side of the thulium (Tm) emission band. An in-house W-type normal dispersion Tm-doped fiber (NDTDF) exhibits a bending-induced distributed short-pass filtering effect that efficiently suppresses the otherwise dominant long wavelength emission. By changing the bending diameter of the fiber, we demonstrated a tunable mode-locked Tm-doped fiber laser with a very wide tunable range of 152 nm spanning from 1740 nm to 1892 nm. Pulses at a central wavelength of 1755 nm were able to be amplified in an all-fiber configuration using the W-type NDTDF, without the use of any artificial short-pass filter or pulse stretcher. The all-fiber amplifier delivers 2.76 ps pulses with an energy of ∼32.7 nJ without pulse break-up, due to the normal dispersion nature of the gain fiber, which marks so far, the highest energy amongst fiber lasers in the 1700 nm-1800 nm region.

High-energy mid-infrared intrapulse difference-frequency generation with 5.3% conversion efficiency driven at 3 µm.

Intrapulse difference-frequency generation (IPDFG) is a relatively simple technique to produce few-cycle mid-infrared (MIR) radiations. The conversion efficiency of IPDFG could be potentially improved by using the long driving wavelength to reduce the quantum defect. In this paper, we report a high-energy MIR IPDFG source with a record-high conversion efficiency of up to 5.3%, driven by 3 µm, 35 fs, 10 kHz pulses. The IPDFG output has a 5 µJ pulse energy and 50 mW average power. It spans over a spectral range from 6 to 13.2 µm. A 68 fs of IPDFG pulse width is measured, corresponding to 2.1 cycles, centered at 9.7 µm. The high-energy, two-cycle IPDFG pulses are used to produce a 3-octave supercontinuum in a KRS-5 crystal, spanning from 2 to 16 µm, with a 2.4 µJ pulse energy and a 24 mW average power.

9 µm few-cycle optical parametric chirped-pulse amplifier based on LiGaS2.

We report a long-wavelength mid-infrared (mid-IR), few-cycle optical parametric chirped-pulse amplifier (OPCPA) based on LiGaS2 crystals, pumped by a 1 µm Yb:YAG laser, at a 10 kHz repetition rate. The mid-IR OPCPA system generates pulses centered at 9 µm, with 1 4 µJ pulse energy and 140 mW average power. A 142 fs pulse width, which corresponds to less than 5 optical cycles at 9 µm, is measured by an interferometric autocorrelator. This is, to the best of our knowledge, the first long-wavelength mid-IR OPCPA pumped at 1 µm wavelength. It paves the way for the energy and power scaling of the ultrafast long-wavelength mid-IR lasers by utilizing advanced high-energy, high-power 1 µm pump lasers.

Multimicrojoule GaSe-based midinfrared optical parametric amplifier with an ultrabroad idler spectrum covering 4.2-16 µm.

We report a multimicrojoule, ultrabroadband midinfrared optical parametric amplifier based on a GaSe nonlinear crystal pumped at ∼2 µm. The generated idler pulse has a flat spectrum spanning from 4.5 to 13.3 µm at -3 dB and 4.2 to 16 µm in the full spectral range, with a central wavelength of 8.8 µm. The proposed scheme supports a subcycle Fourier-transform-limited pulse width. A (2+1)-dimensional numerical simulation is employed to reproduce the obtained idler spectrum. To our best knowledge, this is the broadest -3 dB spectrum ever obtained by optical parametric amplifiers in this spectral region. The idler pulse energy is ∼3.4 µJ with a conversion efficiency of ∼2% from the ∼2 µm pump to the idler pulse.

Quantitative Analysis of Gas Phase IR Spectra Based on Extreme Learning Machine Regression Model.

Advanced chemometric analysis is required for rapid and reliable determination of physical and/or chemical components in complex gas mixtures. Based on infrared (IR) spectroscopic/sensing techniques, we propose an advanced regression model based on the extreme learning machine (ELM) algorithm for quantitative chemometric analysis. The proposed model makes two contributions to the field of advanced chemometrics. First, an ELM-based autoencoder (AE) was developed for reducing the dimensionality of spectral signals and learning important features for regression. Second, the fast regression ability of ELM architecture was directly used for constructing the regression model. In this contribution, nitrogen oxide mixtures (i.e., N2O/NO2/NO) found in vehicle exhaust were selected as a relevant example of a real-world gas mixture. Both simulated data and experimental data acquired using Fourier transform infrared spectroscopy (FTIR) were analyzed by the proposed chemometrics model. By comparing the numerical results with those obtained using conventional principle components regression (PCR) and partial least square regression (PLSR) models, the proposed model was verified to offer superior robustness and performance in quantitative IR spectral analysis.