The inverse-Cherenkov dielectric laser accelerator (ICR-DLA) holds great promise as a compact, on-chip accelerator for a wide range of future applications. However, two significant challenges—bunch dispersion/deflection and phase slippage—have hindered its development in the subrelativistic regime. In this paper, we propose an approach that addresses both issues simultaneously by utilizing a single laser pulse to illuminate a staged dielectric prism. Our method leverages the phase slippage experienced by subrelativistic electrons during high-gradient acceleration, allowing these electrons to encounter alternating focusing and defocusing forces throughout the acceleration process. This technique enables stable, long-range bunch transport within a miniaturized acceleration channel. We show that cascading focusing and acceleration of subrelativistic electrons can be achieved in a multistage acceleration structure, laying a foundation to bridge the gap between subrelativistic and relativistic regimes, which is crucial for the realization of a practical on-chip particle accelerator.

Integrating a high-power microwave source with an accelerating structure into a single device is a key strategy for developing compact, even portable, high-energy particle accelerators, which hold significant promise for scientific, medical, and industrial applications. To achieve this objective, we propose integrating an X-band coaxial transit-time oscillator (CTTO) microwave source with an X-band accelerating structure, thereby drastically minimizing the total volume of particle accelerators. Leveraging the CTTO’s advantages, low operating voltage, stable frequency, high power, and high efficiency, along with the high acceleration gradient of X-band accelerators, our design demonstrates that a CTTO operating at 50 kV (generating 7.8 MW of radio-frequency power at 9.3 GHz) can accelerate an electron beam from 50 keV to 10.96 MeV over a 320 mm length.

On-chip particle accelerators, especially dielectric laser accelerators utilizing the inverse-Cherenkov effect (ICR-DLAs), show great potential for next-generation compact applications. However, single-sided laser-driven ICR-DLAs suffer from two key drawbacks: (1) the achievable acceleration gradient is notably weaker than the incident laser field due to reflection losses; (2) asymmetric field distributions in the bunch channel cause particle deflection, reducing acceleration efficiency. To address these issues, we propose a novel surface plasmon polaritons (SPP)-based ICR-DLA design, which integrates dual metal films into a silicon prism. This structure achieves two key improvements: (1) excitation of SPP significantly increases the acceleration gradient; (2) the laser-induced surface wave is redistributed to form a symmetric field profile in the bunch channel, thereby minimizing deflection. The effectiveness of this design is validated through theoretical analysis and particle tracking simulations. This SPP-based approach marks a significant advance toward on-chip accelerator systems with high acceleration gradients and easily controllable laser-driven profiles.

Pursuing on-chip particle accelerators has long been a focal point for researchers, and the inverse Smith-Purcell dielectric laser accelerator holds great promise as a viable option. However, achieving the cascade acceleration and focusing of subrelativistic electrons on a chip remains one of the most formidable challenges. In this paper, we propose a dynamic phase method to address this issue. This method takes advantage of the phase slippage between the electron and the acceleration wave, enabling transverse alternate focusing, defocusing, and continuous acceleration. Simulation results show that a 29.50 keV electron beam with a 20 eV spread can achieve an acceleration gradient of 39 MV/m with a captured efficiency of 95% within a 120-mu m distance in a 200-nm-wide channel. Compared with the alternating phase method, this approach streamlines the fabrication process by removing the need for incremental adjustments of the phase velocity at each accelerating stage to match the electron energy. We anticipate that applying the dynamic phase-driven approach will significantly propel the development of on-chip accelerators.

The miniaturization of high-power microwave (HPM) sources and the expansion of their multiband operation capabilities play a crucial role in practical applications. This study presents a high-efficiency dual-beam dual-band HPM source that combines a coaxial transit-time oscillator (TTO) with a relativistic backward wave oscillator (RBWO). Unlike conventional compact inner and outer HPM structures, our design features an outer TTO structure for generating Ka-band high-frequency microwaves. The Ku-band low-frequency RBWO structure is nested inside, which significantly reduces the device’s size and weight. Both electron beams are guided by a common magnetic field, achieving significant device compactness while maintaining high-power output. Numerical simulations show that at 450 kV diode voltage and 0.65 T axial magnetic field, the Ka-band TTO generates 1.1 GW output power at 39.4 GHz with 38% conversion efficiency. The Ku-band RBWO simultaneously delivers 522 MW at 16.2 GHz with 36% efficiency.

To achieve the miniaturization and lightweight design of high-power microwave (HPM) systems, a compact permanent magnet TE11 diffraction-output relativistic magnetron is proposed and designed in this paper. Based on permanent magnet circuit theory, the volume requirement of the permanent magnets is significantly reduced by decreasing the inner radius of the magnetic circuit. Three-dimensional particle-in-cell simulation results show that, under an input voltage of 250 kV and an anode current of 3.4 kA, the device operates stably in the pi mode at 2.4 GHz, and successfully realizes the direct extraction of the TE11 mode. An average power of approximately 250 MW (with a peak power of 500 MW) is obtained at the output port, achieving a total efficiency of 29.47 percent. Although there is a slight trade-off in efficiency, this design dramatically reduces the volume and weight of the entire microwave source, effectively promoting the portability and engineering application of HPM systems.

The inverse Cherenkov dielectric laser accelerator holds great promise for compact on-chip acceleration, yet phase slippage and bunch dispersion in the subrelativistic regime remain critical bottlenecks. In this paper, we propose a three-stage cascaded terahertz prism accelerator that achieves continuous phase matching through a gradient incidence angle design, and simultaneously employs a dynamic phase bunching method. This allows electrons to experience alternating focusing and defocusing forces throughout acceleration, enabling stable long-range beam transport. Simulation results show acceleration from 100 keV to 1.84 MeV over 24 mm, with an average gradient of 71.4 MV/m and 100% transmission. This approach lays a foundation for bridging the subrelativistic to relativistic regimes in compact terahertz accelerator development.

On-chip inverse Cherenkov dielectric laser accelerators (ICR-DLAs) hold promise for next-gen compact applications. But single-sided laser-driven ICR-DLAs face two issues: (1) Weaker acceleration gradient than incident laser (reflection losses); (2) Asymmetric channel fields cause beam deflection, reducing efficiency. To address these, we propose a novel SPP-enhanced ICR-DLA with dual metal films in a silicon prism. It offers two gains: (1) SPP excitation boosts acceleration gradient; (2) Laser surface waves form symmetric channel fields, reducing deflection. Validated by theoretical analysis and particle simulations, this approach advances on-chip accelerators with high gradients and controllable beam profiles.

High power microwave (HPM) sources play a vital role in numerous applications, from directed energy weapons to high power radar systems. Among various HPM devices, the coaxial transit-time oscillator (c-TTO) holds great promise due to its high power capacity and bulk-wave characteristics. To further improve its performance in higher frequencies and output powers, this paper presents a novel solution to reduce the maximum operating field intensity, thereby enhancing the power capacity. Specifically, an internal waveguide within the inner conductor is mounted before the energy extraction cavities. By optimizing the configuration and position of the internal waveguide, the field intensity in the buncher is significantly reduced at the cost of adding the structure complexity and reducing the total output power. Through simulations, we demonstrate that the proposed c-TTO can generate power exceeding 1 GW at a frequency of 39.4 GHz. This research offers a promising pathway toward high power capacity Ka-band HPM sources.

The trend toward compact, multifrequency high-power microwave (HPM) systems is growing increasingly prominent. Here, we propose a compact dual-band nested relativistic backward wave oscillator (RBWO), comprising an outer coaxial slow wave structure (SWS) operating in the Ka-band and an inner hollow SWS operating in the Ku-band. The inner and outer ring-shaped electron beams propagate independently while sharing a common guiding magnetic field. This design reduces the size to less than half of previous schemes, significantly decreasing the system volume and weight. Particle-in-cell (PIC) simulations show that at a diode voltage of 500 kV and a guiding magnetic field of 0.84 T, the Ka-band (31.1GHz) and Ku-band (15.6GHz) microwave outputs reach 1.4 and 0.95 GW, with corresponding efficiencies of 27% and 38%, respectively.

The millimeter wave (such as Ka-band) relativistic backward wave oscillator (RBWO) operating at gigawatt power is contemporary pivotal due to the high vulnerability of vacuum breakdown in compact device size. This paper proposes an internal resonant output structure inside a coaxial RBWO, to reduce the maximum field intensity in the slow-wave structure. With optimization on size parameters in simulation, the maximum field intensity of the device is reduced by over half while the total output power is maintained. Simulation results show the proposed device can deliver 1 GW output at 39.85 GHz. This research presents a promising avenue toward achieving high-power millimeter-wave generators.

Dual-band high-power microwave (HPM) sources, combining the distinct advantages of different frequencies (e.g., penetration depth, energy deposition efficiency, and target coupling), are highly valuable for military, industrial, and scientific applications. However, existing dual-band HPM systems typically require complex setups with two separate electron beams, posing significant challenges in beam control and thermal management. This paper proposes a compact, single-beam-driven dual-frequency coaxial transit-time oscillator that generates simultaneous Ku- and Ka-band outputs using cascaded extractors to efficiently extract energy from both the fundamental wave (14 GHz) and its second harmonic (28 GHz). Unlike conventional single-frequency devices, this design maintains a simple structure while delivering dual-band operation with enhanced power and efficiency. At 490 kV and 9.6 kA under a 0.83 T magnetic field in simulation, the device achieves 1.7 GW at 14 GHz (Ku-band) and 0.35 GW at 28 GHz (Ka-band), with a total output power of 2.05 GW and an efficiency of 44.6%. (c) 2025 Author(s). All article content, except where otherwise noted, is licensed under a Creative Commons Attribution-NonCommercial 4.0International (CC BY-NC) license (https://creativecommons.org/licenses/by-nc/4.0/).https://doi.org/10.1063/5.0294903

Output efficiency of a klystron is strongly determined by the beam bunching quality, and the size optimize of the extractor. In this work, an automated optimization framework for single-extractor S-band klystrons is developed using the original one-dimensional particle-field interaction code. The Particle Swarm Optimization algorithm maximizes output power by tuning cavity gap width, external quality factor, resonant frequency, and position, while energy-conservation and divergence-coefficient criteria ensure physically valid solutions. Simulations show that under small beam currents, optimized output efficiency (similar to 70%) consistent with 3D particle-in-cell (PIC) simulation results. For kA current operation, the framework provides reliable operational ranges, with PIC simulations yielding similar to 50% efficiency. The method efficiently guides cavity design, offering a robust and physically consistent tool for further multi-cavity optimization of klystrons.

Particle accelerators have played a pivotal role in scientific research, medical applications, and industrial processes. However, conventional radio-frequency accelerators face limitations due to their large size, high cost, and dependence on external microwave sources. This paper presents a novel linear accelerator design that integrates a high-power microwave source directly into the accelerator structure, eliminating the need for external drivers and enabling a more compact, cost-effective, and simplified system. The proposed design employs an X-band backward wave oscillator driven by a 50 keV hollow continuous electron beam. The generated microwave power is coupled into a central X-band standing-wave acceleration structure via a radial coupler. Simulations demonstrate that a witness beam passing through this structure can be accelerated from 50 keV to over 4 MeV across 10 acceleration cells. This work serves as a foundational study on integrated accelerator concepts, offering a pathway toward more efficient, scalable, and versatile particle acceleration systems.

The aim of this study is to develop a composite substrate for surface-enhanced Raman scattering (SERS) applications by utilizing a PMMA inverted anodic aluminum oxide (AAO) nanostructure array, with a modified MoS2 film as an interlayer between the gold nanofilm (Au film) and silver nanoparticles (AgNPs). The substrate not only possesses the electromagnetic enhancement mechanism (EM), but also introduces the chemical enhancement mechanism (CM) of MoS2 and its Raman internal standard characteristics. Utilizing the properties of hyperbolic metamaterials (HMM) to support high-wave-vector body plasmon polaritons (BPP), with AgNPs as the external coupling structure, the large-wave-vector scattering light conducted within the nanostructured array resonator excites the BPP in the HMM, expanding the energy into the gap of the AgNPs. This energy further couples with the local surface plasmon polaritons (LSP) on the surface, generating a high-density "hotspot" on the surface. Through a novel reverse fabrication method, we successfully fabricated a three-dimensional (3D) arrayed nano-composite architecture and used it as a flexible SERS sensor, which can be easily applied to largescale manufacturing. Finally, experimental verification demonstrated the excellent performance of this substrate. The limits of detection (LOD) for the probe molecules R6G and CV were found to be 6.93 x 10-12 M and 5.16 x 10-11 M, respectively. Additionally, in situ detection of CV on seafood surfaces resulted in a LOD of 3.78 x 10-6 M. Therefore, we believe that this flexible composite material provides a feasible solution for high-sensitivity molecular detection and has promising applications in the field of biosensing.

Traditional surface-enhanced Raman scattering (SERS) substrates typically rely on electromagnetic enhancement, which is predominantly generated by noble metal particles. While their practical limitations hinder widespread application. In contrast, semiconductor substrates have garnered increasing attention in the SERS field due to their capacity to address many of the shortcomings associated with noble metal particles. In this study, we investigated the SERS effect and enhancement mechanism of a novel transition metal carbide (ZrC), further enhancing its performance through morphological alterations and element doping. The detection limits for Rhodamine 6 G and Methylene blue using the prepared substrate were found to be remarkably low at 1 x 10-8 M and 1 x 10-6 M. To elucidate the enhancement mechanism, Density Functional Theory and Finite Element Analysis were employed. These simulations demonstrated that the observed enhancement arises from chemical mechanisms. The charge transfer process between the SERS substrate and the probe molecule resulted in significantly amplified Raman signals.

A simple and efficient strategy is developed to effectively reduce the emission spectral width of through-space charge-transfer (TSCT) electroluminescent (EL) emitters by incorporating dual multiple resonance (dual-MR) groups into a single molecule and utilizing a spatially hindered connection approach. Based on this design principle, three proof-of-concept EL emitters, m-CzFBN, o-NOFBN, and o-CzFBN is successfully developed, featuring a rigid molecular backbone and TSCT characteristics. Unlike conventional TSCT molecules, which exhibit broad emission spectra and large full-width at half-maximum (FWHM) values, the newly designed molecules demonstrate significantly narrow full-width at half-maximum (FWHM) values. Notably, m-CzFBN displays an emission bandwidth as narrow as 15 nm in solution, representing the narrowest spectral width reported to date for any TSCT emitter. Moreover, all three emitters exhibit high photoluminescence quantum yields (PLQYs) exceeding 90%, which directly enhances their EL performance. A non-sensitized organic light-emitting diode (OLED) fabricated using m-CzFBN achieved an external quantum efficiency (EQE) of up to 34.0%, ranking among the highest efficiencies reported for narrow-band TSCT emitters. These findings provide a promising molecular design strategy for developing narrowband EL emitters suitable for ultra-high-definition (UHD) display applications.

We propose a scheme for a Figure-9 mode-locked fiber laser that utilizes liquid crystal variable retarders (LCs) with tunable splitting ratios and non-reciprocal phase shifts. Simulation results demonstrate that the parameter space can be effectively explored through full-wave scanning of the two LCs, while fast axes of both LCs are oriented at an angle of -1/4π to the horizontal plane. Under quasi-symmetric conditions, experimental results confirm the scheme's capability to control the mode-locking state. By varying the retardance of the LCs, we map the distribution of mode-locking states in the parameter space. Our findings reveal that, due to the difference in nonlinear transmission from the conventional scheme, this scheme has a similar distribution of mode-locking states but is more conservative. This innovative scheme offers a fast-response laser platform without moving parts, making it suitable for research in laser technology, intelligent mode locking, and ultrafast dynamics.

Random lasers have attracted much attention in recent years owing to their advantages of a simple fabrication process, low processing cost, and material flexibility for any lasing wavelengths. They provide a roadmap for the design of ultra-bright lighting, displays, etc. However, the threshold reduction in random nanolasers remains a challenge in practical applications. In this work, lower-threshold random laser action from monolayer molybdenum disulfide film-encapsulated Au nanoparticles (MoS2/Au NPs) is demonstrated. The observed laser action of the MoS2/Au NPs shows a lower threshold of about 0.564 µJ/mm2, which is about 46.2% lower than the threshold of random lasers based on Au NPs. We proposed that the charge transfer between MoS2 and the gain material is the main reason for the reduction in the random laser threshold. The finite-difference time-domain (FDTD) method was used to calculate the lasing action of these two nanostructures. When charge transfer is taken into account, the theoretically calculated threshold of the MoS2/Au NPs is reduced by 46.8% compared to Au NP samples, which is consistent with the experimental results. This study provides a new mechanism to achieve low-threshold and high-quality random lasers, which has the potential to facilitate the application of random lasers and the development of high-performance optoelectronic devices.

Multi-modal Large Language Models (MLLMs) have demonstrated impressive instruction abilities across various open-ended tasks. However, previous methods primarily focus on enhancing multi-modal capabilities. In this work, we introduce a versatile multi-modal large language model, mPLUG-Owl2, which effectively leverages modality collaboration to improve performance in both text and multi-modal tasks. mPLUG-Owl2 utilizes a modularized network design, with the language decoder acting as a universal interface for managing different modalities. Specifically, mPLUG-Owl2 incorporates shared functional modules to facilitate modality collaboration and introduces a modality-adaptive module that preserves modality-specific features. Extensive experiments reveal that mPLUG-Owl2 is capable of generalizing both text tasks and multi-modal tasks and achieving state-of-the-art performances with a single generic model. Notably, mPLUG-Owl2 is the first MLLM model that demonstrates the modality collaboration phenomenon in both pure-text and multi-modal scenarios, setting a pioneering path in the development of future multi-modal foundation models.