Plasmonics最新文献

This study reports the fabrication and performance evaluation of advanced multilayer hybrid plasmonic SERS sensors based on tri-material core–shell nanostructures deposited on macroporous silicon (McPSi) substrates. Two distinct nanoparticle architectures AgNPs core/AuNPs shell/AgNPs shell and AuNPs core/AgNPs shell/AuNPs shell—were successfully synthesized and integrated onto the McPSi surface, forming highly efficient SERS-active platforms. Owing to the synergistic plasmonic coupling between the multilayer metallic shells and the high surface area of the porous silicon substrate, both configurations exhibited outstanding sensing performance. The AgNPs core/AuNPs shell/AgNPs shell and AuNPs core/AgNPs shell/AuNPs shell sensors achieved ultralow detection limits of 1 × 10⁻¹² M and 2 × 10⁻¹³ M, respectively. Moreover, remarkably high Raman enhancement factors of 1 × 10⁸ and 5 × 10⁷ were obtained, indicating strong electromagnetic field localization and dense plasmonic hot-spot formation. These results demonstrate the superior signal amplification capability and exceptional sensitivity of the proposed hybrid sensors, highlighting their strong potential for ultra-trace detection of environmental contaminants and advanced biosensing applications.

A 4-parameter gas sensor that can simultaneously detect methane (CH₄), relative humidity (RH), hydrogen (H₂), and hydrogen sulfide (H₂S) is designed and analyzed. It uses a 4D-type coreless optical fiber as the substrate, gold (Au) as the plasmonic medium, and cryptophane-A/ultraviolet-cured fluorosiloxane complex (Crypt-A/UVCFS), polyvinyl alcohol (PVA), Pd-WO_3, and zinc oxide (ZnO) as the sensing materials. Simulation results show that in the ranges of methane concentrations (c(CH₄)) of 0%-2%, RH of 30%-70%, hydrogen concentration (c(H₂)) of 0%-2%, and hydrogen sulfide (c(H₂S)) concentration of 0–80 ppm, the average CH₄ sensitivity, RH sensitivity, H₂ sensitivity, and H₂S sensitivity are − 76 nm/%, -2.2 nm/%RH, -153 nm/%, and 15 nm/ppm, respectively.

A novel dual U-notched surface plasmon resonance (SPR)-based photonic crystal fiber (PCF) sensor in the mid-infrared (MIR) regime has been investigated, exhibiting an application for refractive index (RI) prediction through the machine learning (ML) technique. The structural design parameters were selected using an L9 Taguchi orthogonal-array parameter screening analysis, in which two signal-to-noise ratio (SNR) responses, corresponding to peak confinement loss and full width at half maximum (FWHM), were analyzed through their delta values, and the resulting configuration has been verified by confirmation simulation. The sensor achieves maximum and average wavelength sensitivities of 30,000 nm/RIU and 19,010 nm/RIU, respectively, with an amplitude sensitivity of 968.20 RIU^− 1, a resolution of 3.33 × 10^− 6 RIU, and maximum and average figures of merit (FOM) of 1034.06 RIU^− 1 and 569.87 RIU^− 1, respectively, over the assessed RI range. To enable accurate RI prediction without repeated full-wave computation, the ML models have been trained on numerical data generated by finite-element simulation in COMSOL Multiphysics, using the resonance wavelength and peak confinement loss jointly as inputs. Six regression methods, namely linear regression, second- and third-degree polynomial regression, support vector regression (RBF), Gaussian process regression (RBF) and XGBoost, have been evaluated by using the metrics, such as mean absolute error (MAE), root mean square error (RMSE), R², and mean absolute percentage error (MAPE) under the nested leave-one-out cross-validation, with an independent hold-out set. The GPR performed best, attaining an MAE of 1.16 × 10^− 4 RIU, an RMSE of 1.61 × 10^− 4 RIU, R² = 0.99956, and a MAPE of 0.0087%. The proposed sensor detects analytes with RIs between 1.326 and 1.353 over the MIR wavelength range of 2940–3460 nm and is applicable to biological analysis, chemical detection, and environmental monitoring within that range.

This manuscript presents a Kretschmann-configured long-range surface plasmon resonance (LRSPR) sensor that exploits the in-plane optical anisotropy of rhenium disulfide (ReS₂) for biomolecular refractive-index sensing. The 2S2G/Cytop/Ag/LiNbO₃/ReS₂ architecture operates at 633 nm, employs Ag for plasmon excitation, and uses LiNbO₃ as a protective passivation layer. Transfer-matrix calculations and COMSOL field simulations were used to examine the monolayer, bilayer, and trilayer ReS₂ at orientationsparallel (at θ = 0°) and perpendicular (at θ = 90°) to the Re–Re chains. Layer number and orientation substantially govern the resonance characteristics. At θ = 0°, trilayer ReS₂ yields the lowest reflectance (0.03647 a.u.), whereas at θ = 90°, it provides the highest imaging sensitivity (23,034 RIU⁻¹), detection accuracy (100/°), angular figure of merit (1200 RIU^-1), and penetration depth (644.25 nm). The resulting narrow resonance and deeper evanescent field markedly outperform the reference conventional SPR configuration. These findings establish an orientation-controlled ReS₂-assisted LRSPR sensor as a promising sensor for high-resolution, label-free biosensing.

Plasmonic nanolasers enable subwavelength optical confinement far below the diffraction limit, but practical implementation is constrained by Ohmic dissipation in noble metals. In this work, we present a computational investigation of anisotropic gold nanostars (AuNSs) coupled to colloidal CdSe/ZnS core-shell quantum dots operating at near-infrared wavelengths (780 nm). Integrating three-dimensional finite-difference time-domain (3D-FDTD) electrodynamics with boundary element method (MNPBEM) calculations, numerical results indicate that for the design geometry identified within the explored parameter space ((R_{text {core}} = 20,text {nm}, L_{text {tip}} = 30,text {nm}, r_{text {tip}} = 2.0,text {nm}, N_{text {tips}} = 8)), charge localization at sharp tip apexes yields a calculated Purcell factor of (F_P = 845 pm 110) (expanded uncertainty with coverage factor (k=2), approximately 95.45% coverage probability) and a normalized mode volume of (V_{text {eff}} = 1.76 times 10^{-7}(lambda /n)^3) ((V_{text {eff}} approx 35.2,text {nm}^{3})), enabling a cold-cavity modal threshold gain of (g_{text {th,modal}} approx 9.59 times 10^3text { cm}^{-1}) ((1.043 times 10^4text { cm}^{-1}) for (n=1.45)) and a calculated (sim 15)-fold threshold pump rate reduction relative to uncoupled emitters. Numerical consistency across independent solvers, spatial mesh convergence, and surrogate Monte Carlo sampling supports the robustness of the computational design. The reported findings provide computational design guidelines for low-threshold near-infrared plasmonic nanolasers within the explored parameter space.

A dual-interface photonic crystal fiber (PCF) sensor based on surface plasmon resonance (SPR) is proposed for the simultaneous detection of refractive index (RI) and temperature. In the dual-interface configuration, two independent sensing channels are formed on opposite sides of the PCF, and gold (Au) films are deposited on the polished surfaces to excite the SPR effect. Numerical results show that the resonance wavelength exhibits distinct spectral shifts in response to variations in both the analyte RI and temperature. For RI sensing, the proposed sensor achieves a maximum wavelength sensitivity of 8300 (mathrm{nm}/^circmathrm C) over an RI range of 1.34–1.42. For temperature sensing, a maximum sensitivity of 10.8 (nm^circtext{C}) is obtained within the temperature range of 10–80 (^{circ}text{C}). The RI and temperature detection resolutions are estimated to be (:text{1.20}times{text{10}}^{text{-5}}) (:text{RIU}) and (:text{9.26}times{text{10}}^{text{-3}}{^circ}text{C}), respectively. In addition, the dual-interface configuration facilitates metal-film deposition and improves the fabrication feasibility of the sensor. These results indicate that the proposed PCF-SPR sensor has promising potential for simultaneous RI and temperature monitoring because of its favorable sensitivity, relatively wide detection ranges, and potential fabrication feasibility.

This study presents the design and numerical investigation of a metasurface absorber/transmission for terahertz (THz) wave applications. A resonance-engineering technique is employed to generate multiple closely spaced resonances and merge them to achieve ultra-wideband absorption. The metasurface unit cell is realized using thermally switchable vanadium-dioxide (VO₂) as both the resonator and ground plane, while silicon-dioxide (SiO₂) is utilized as the dielectric cavity and impedance-matching layer. In the conducting phase of VO₂, the proposed structure exhibits ultra-wideband absorption more than 90% ranging from 4.04 THz to 19.47 THz, corresponding to an absolute bandwidth of 15.43 THz and a fractional bandwidth of 131.2% among the highest reported for VO₂-based THz absorbers to the best of current knowledge. The temperature-controlled phase transition of VO₂ enables a reconfigurable transmission response, where more than 80% transmission is achieved over the 2–13.31 THz band, yielding a fractional bandwidth of approximately 147.7% in the insulating phase. The absorber performance is further analysed under oblique incidence to assess its practical applicability, demonstrating stable absorption for incident angles up to approximately 60° for both transverse electric and transverse magnetic polarizations. A transmission line based equivalent circuit model is used to further validates its simulated results, demonstrating good agreement between the theoretical and simulated responses. Detailed fabrication steps, working mechanism using electric, magnetic and power-flow graph of the proposed structure are presented, and a comprehensive parametric analysis is carried out to evaluate fabrication tolerance. These features make the metasurface highly promising for emerging THz applications, spectroscopic imaging, non-invasive biosensing, stealth technology, electromagnetic shielding, and frequency-selective filtering.

We propose a Ridge-loaded Nanocavity Array (RLNA) for ratiometric refractive-index sensing. A metallic nanoridge embedded within each nanoslit modifies the transmission spectrum to produce two distinct resonances exhibiting anti-correlated intensity modulation in response to changes in the analyte refractive-index, thereby enabling dual-peak intensity-ratio interrogation. Finite element method simulations show an average ratiometric sensitivity of (350~textrm{RIU}^{-1}) over the simulated range of (n=1.30)–1.36. The proposed all-metal architecture realizes dual-peak operation without requiring additional functional materials or complex multilayer structures. The transmission-mode intensity-ratio readout provides a self-referenced sensing metric that suppresses common-mode intensity fluctuations, providing a potentially scalable all-metal platform suitable for plasmonic refractive-index sensing.

Spoof surface plasmons (SSPs) are special bounded surface waves on the corrugated metallic waveguides in Terahertz (THz) and/or microwave bands. SSPs modes have been proposed for various potential applications due to their advantages such as the tight near-field localization and enhanced light-matter interactions. However, previous studies on their high-efficiency excitations are only on a single type plasmonic waveguide and thus limited to single-band THz operation, which limit their applications for multi-band and multi-functional plasmonic devices and systems. To address this challenge, the dual-band excitations of THz SSPs on a composite metallic grating by integrating different rectangular groove depth are proposed and investigated. The composite metallic grating can simultaneously support dual-band tightly bounded SSPs modes, which the cutoff frequency of each mode is determined by the individual groove depth independently. The dual-band SSP waveguide is then designed in THz band and its propagation characteristics are also verified by both dispersion lines and good transmission performances of unit cell. Besides, high-efficiency near-field dual-band excitations based on wave vector matching mechanism via the charged electron beam are also investigated and analyzed via particle-in-cell simulations. The output field intensity, SSP power profiles and spectrum of dual-band SSPs at two distinct frequencies of around 0.76THz and 0.96THz are demonstrated. Compared to previous single-band excitations on single-periodic metallic grating, the ability to guide dual-band SSPs and the simultaneously high efficiency electronic excitations on composite meta-structures are promising to the development of novel dual-band multi-functional plasmonic devices and systems in THz photonics.

Traditional design methods for nanophotonic devices suffer from low efficiency and high computational cost. Here, we propose a deep-learning-based framework integrating a variational autoencoder (VAE) with FDTD simulations. Using only 4,000 data items, the VAE learns the mapping from structural parameters to optical responses. It generates device structures from target spectra within seconds, which is over three orders of magnitude faster than conventional methods. The inversely designed cross-dual-ring metasurface absorbers achieve 99% absorption at 10.2 μm and 96% at 13.4 μm, with average peak wavelength errors of 20 nm and 30 nm, and intensity errors below 3%. All designs are validated by independent FDTD simulations. The VAE’s continuous latent space enables one‑to‑many mapping, yielding multiple structurally distinct solutions for a single target spectrum and thus enhancing fabrication tolerance. Converging within only 30 epochs, this approach offers a practical and scalable solution for rapid photonic device design.