arXiv subjects

Kai Fu

Publications and source records attributed to Kai Fu.

At least 19 recordsLinked to original sources

arXiv2026

Thermal Stability and Phase Transformation of Conductive $\alpha$-$(\mathrm{Al}_{x}\mathrm{Ga}_{1-x})_{2}\mathrm{O}_{3}/\mathrm{Ga}_{2}\mathrm{O}_{3}$ Heterostructures on Sapphire Substrates

Thermal stability and phase transformation of conductive $\alpha$-$(\mathrm{Al}_{0.16}\mathrm{Ga}_{0.84})_{2}\mathrm{O}_{3}/\mathrm{Ga}_{2}\mathrm{O}_{3}$ heterostructures on sapphire substrates were investigated using in situ high-temperature X-ray diffraction (HT-XRD), scanning electron microscopy (SEM), and atomic force microscopy (AFM). Conductive $\alpha$-$(\mathrm{Al}_{0.16}\mathrm{Ga}_{0.84})_{2}\mathrm{O}_{3}/\mathrm{Ga}_{2}\mathrm{O}_{3}$ heterostructures with fluorine (F) doping were grown by mist chemical vapor deposition on sapphire substrates, achieving a Hall mobility of $28~\mathrm{cm^{2}\,V^{-1}\,s^{-1}}$ and an electron concentration of $1.4\times10^{20}~\mathrm{cm^{-3}}$. The heterostructures exhibited thermal stability up to approximately $550$--$

cond-mat.mtrl-sci↗

arXiv2026

Crystallinity Evolution of MOCVD-Grown $\beta$-Ga$_2$O$_3$ Films Probed by In Situ HT-XRD under Different Reactor Heights

The crystallinity of $\beta$-Ga$_2$O$_3$ thin films grown by metal-organic chemical vapor deposition (MOCVD) is strongly influenced by reactor design and the resulting growth environment. In this work, we investigate the role of reactor height on the crystallinity evolution of MOCVD-grown $\beta$-Ga$_2$O$_3$ films by directly comparing long- and short-chamber showerhead configurations. Structural evolution was probed by in situ high-temperature X-ray diffraction (HT-XRD) as the MOCVD-grown films were heated from 25~$^\circ$C to 1100~$^\circ$C. Temperature-dependent XRD reveals a consistent redshift of the $\beta$-Ga$_2$O$_3$~($-201$) reflection after HT-XRD heating and subsequent cooling to room temperature for both reactor geometries, indicating a similar thermally driven strain response. Quantitative rocking-curve analysis shows a non-monotonic temperature dependence of the ($-201$) full width at half maximum (FWHM), with minimum values of approximately 2.03$^\circ$ and 2.72$^\circ$ for the short- and long-chamber films, respectively, reflecting differences in mosaic alignment established during growth. Atomic force microscopy further shows that short-chamber-grown films exhibit smoother surfaces, with root-mean-square roughness values of approximately 7.7~nm before and 7.3~nm after HT-XRD heating, compared to 19.3~nm and 12.3~nm, respectively, for long-chamber-grown films. Overall, these results indicate that reactor height influences the initial crystalline and morphological templates of $\beta$-Ga$_2$O$_3$ films and modulates their elevated-temperature structural response, providing practical insights for optimizing MOCVD reactor design for high-quality $\beta$-Ga$_2$O$_3$ growth.

cond-mat.mtrl-sci↗

arXiv2026

Dosimetry for Proton Therapy Using a \beta-Ga$_2$O$_3$ Metal-Semiconductor-Metal Detector with Low-Noise Amplification

Intensity-modulated proton therapy (IMPT) employs proton radiation rather than conventional X-rays to treat cancerous tumors. This approach offers significant advantages by minimizing the radiation exposure of surrounding healthy tissue, leading to improved patient outcomes and reduced side effects compared to traditional X-ray therapy. To ensure patient safety, each treatment plan must be experimentally validated before clinical implementation. However, current dosimetry devices face limitations in performing angled beam measurements and obtaining multi-depth assessments, both of which are essential for verifying IMPT treatment plans. In this study, the performance of a \beta-Ga$_2$O$_3$-based metal-semiconductor-metal (MSM) detector with a low-noise amplifier is studied and evaluated under various proton radiation doses and energy levels delivered by a MEVION S250i proton accelerator. The detector performance is also compared with that of an ionization chamber. The \beta-Ga$_2$O$_3$ detector exhibits a linear response with proton dose for single-spot irradiations, and its response to varying proton energies closely matches both the ion chamber data and simulated dose distributions. These findings highlight the potential of \beta-Ga$_2$O$_3$-based detectors as robust dosimetry devices for IMPT applications.

physics.med-ph↗

arXiv2026

A Vision-Language-Action Model with Visual Prompt for OFF-Road Autonomous Driving

Efficient trajectory planning in off-road terrains presents a formidable challenge for autonomous vehicles, often necessitating complex multi-step pipelines. However, traditional approaches exhibit limited adaptability in dynamic environments. To address these limitations, this paper proposes OFF-EMMA, a novel end-to-end multimodal framework designed to overcome the deficiencies of insufficient spatial perception and unstable reasoning in visual-language-action (VLA) models for off-road autonomous driving scenarios. The framework explicitly annotates input images through the design of a visual prompt block and introduces a chain-of-thought with self-consistency (COT-SC) reasoning strategy to enhance the accuracy and robustness of trajectory planning. The visual prompt block utilizes semantic segmentation masks as visual prompts, enhancing the spatial understanding ability of pre-trained visual-language models for complex terrains. The COT- SC strategy effectively mitigates the error impact of outliers on planning performance through a multi-path reasoning mechanism. Experimental results on the RELLIS-3D off-road dataset demonstrate that OFF-EMMA significantly outperforms existing methods, reducing the average L2 error of the Qwen backbone model by 13.3% and decreasing the failure rate from 16.52% to 6.56%.

cs.RO↗

arXiv2025

In situ XRD Study of Strain Evolution in AlGaN/GaN HEMT at High Temperatures up to 1000 {\deg}C

The thermal stability and structural evolution of a GaN high-electron-mobility transistor (HEMT) heterostructure grown on a Si (111) substrate were investigated using in situ high-temperature X-ray diffraction (HT-XRD), reciprocal space mapping (RSM), Raman spectroscopy, and rocking-curve (RC) analysis at varying temperatures. The heterostructure, consisting of a p-GaN cap, an AlGaN barrier, and a GaN channel supported by two AlGaN/AlGaN superlattice (SL) buffer layers, maintained clear and periodic satellite peaks up to a temperature of 1000 deg C, confirming excellent structural integrity. Symmetric and asymmetric RSM results reveal that both the Si and GaN diffraction peaks shift to lower angles with increasing temperature, consistent with thermal expansion, and show no significant broadening or relaxation throughout the heating process. The c-lattice constant follows the theoretical expansion predicted by the multi-frequency Einstein model, whereas the a-lattice expansion is slower due to in-plane strain constraints imposed by the underlying Si substrate and buffer layers. Rapid lattice contraction during the fast-cooling stage induces a residual compressive strain of approximately 0.3 percent in the GaN channel after cooling. Raman spectra further confirm this strain state through a blue shift of approximately 1.5 cm-1 of the GaN E2 (high) phonon mode, corresponding to an in-plane strain of about 0.2 percent. Rocking-curve analysis reveals an increase in both screw and edge dislocation densities by 28 percent and 12 percent, respectively. These results collectively demonstrate that the GaN HEMT heterostructure exhibits robust crystalline stability up to 1000 deg C, with only minor strain redistribution and limited dislocation activity, providing experimental evidence for GaN device applications under high-temperature conditions.

cond-mat.mtrl-sci↗

arXiv2025

In situ Study of p-NiO Film Quality at High Temperatures up to 1100 deg C

NiO is a promising p-type material for photovoltaics and power electronics, but its temperature limits remain unclear. Using in situ high-temperature X-ray diffraction (HT-XRD) from 30 to 1100 C, we track the structural evolution of NiO thin films in air. The film crystallizes from an amorphous phase to cubic NiO between 300 and 400 C, where the emergence and growth of the (111) diffraction peak correlate with an increase in electrical resistivity. Further increases in temperature lead to improved crystallinity and higher resistivity. At 1100 C, the formation of Ni2O3 is observed, resulting in a highly resistive film. This study establishes a clear correlation between phase evolution, crystallinity, and resistive behavior in NiO thin films.

cond-mat.mtrl-sci↗

arXiv2025

Control of Extraordinary Optical Transmission in Resonant Terahertz Gratings via Lateral Depletion in an AlGaN-GaN Heterostructure

Periodic metallic gratings on substrates can support a range of electromagnetic modes, such as leaky waveguide, guided-resonant, and Fabry-Perot (FP) cavity modes, which can strongly modulate optical transmission under resonant excitation. Here, we investigate how this coupling can be dynamically manipulated through charge-density control in a laterally patterned AlGaN/GaN heterostructure. The structure comprises metallic stripes separated by regions containing a two-dimensional electron gas (2DEG), forming a periodically modulated interface whose electromagnetic response is governed by the charge density between the stripes. In the unbiased state, the conductive 2DEG screens the incident terahertz field and suppresses excitation of guided modes. When the 2DEG is depleted, the change in boundary conditions allows efficient coupling into substrate resonances, producing a strong modulation at particular frequencies where extraordinary optical transmission (EOT) through the structure takes place. The results highlight the sensitive dependence of guided-mode-resonance (GMR) mediated EOT on inter-stripe charge distribution and demonstrate a direct interplay between carrier dynamics and resonant electromagnetic phenomena in the terahertz regime.

cond-mat.mtrl-sci↗

arXiv2025

Klear-AgentForge: Forging Agentic Intelligence through Posttraining Scaling

Despite the proliferation of powerful agentic models, the lack of critical post-training details hinders the development of strong counterparts in the open-source community. In this study, we present a comprehensive and fully open-source pipeline for training a high-performance agentic model for interacting with external tools and environments, named Klear-Qwen3-AgentForge, starting from the Qwen3-8B base model. We design effective supervised fine-tuning (SFT) with synthetic data followed by multi-turn reinforcement learning (RL) to unlock the potential for multiple diverse agentic tasks. We perform exclusive experiments on various agentic benchmarks in both tool use and coding domains. Klear-Qwen3-AgentForge-8B achieves state-of-the-art performance among LLMs of similar size and remains competitive with significantly larger models.

cs.AI↗

arXiv2025

High thermal conductivity of rutile-GeO$_2$ films grown by MOCVD: $52.9~\mathrm{W\,m^{-1}\,K^{-1}}$

Rutile germanium dioxide (r-GeO2) has recently emerged as a promising ultrawide-bandgap (UWBG) semiconductor owing to its wide bandgap (~4.4-5.1 eV), ambipolar doping potential, and high theoretical thermal conductivity. However, experimental data on the thermal conductivity of r-GeO2 epitaxial layers have not been reported, primarily due to challenges in phase control and surface roughness. Here, we report a high thermal conductivity of 52.9 +/- 6.6 W m^-1 K^-1 for high-quality (002) r-GeO2 films grown by metal-organic chemical vapor deposition (MOCVD) and characterized using time-domain thermoreflectance (TDTR). The phase control was achieved through a seed-driven stepwise crystallization (SDSC) approach, and the surface roughness was significantly reduced from 76 nm to 16 nm (locally as low as 1 A) via chemical mechanical polishing (CMP). These results highlight the promise of r-GeO2 as a UWBG oxide platform for power electronics applications.

cond-mat.mtrl-sci↗

arXiv2025

Quantifying AI-to-Clinical Translation: The Algorithm-to-Outcome Concordance (AOC) Framework

Background: Despite high in-silico performance (AUC >0.80), 85% of AI cancer biomarkers fail clinical translation, exposing a critical algorithm-to-outcome gap. Methods: We introduce the Algorithm-to-Outcome Concordance (AOC) framework, integrating model accuracy (AUC), clinical correlation (Corr), and trial heterogeneity. We validated AOC across 6 neoantigen vaccine trials (2017-2025) and 3 independent melanoma immunotherapy cohorts (n=188 patients). Results: AOC ranged 0.18-0.79 across trials, with failed trials (ORR <15%) showing AOC <0.40. External validation revealed unstable algorithm-outcome correlation (C-index: 0.49-0.61, p>0.05), demonstrating the necessity of explicit concordance assessment. Conclusions: AOC provides a quantitative framework for pre-trial risk assessment and adaptive trial design. Prospective validation is underway in KEYNOTE-942 extension studies.

q-bio.QM↗

arXiv2025

$\beta$-Ga$_2$O$_3$--Based Radiation Detector for Proton Therapy

Intensity modulated proton therapy (IMPT) is an advanced cancer treatment modality that offers significant advantages over conventional X-ray therapies, particularly in its ability to minimize radiation dose beyond the tumor target. This reduction in unnecessary irradiation exposure significantly lowers the risk to surrounding healthy tissue and reduces side effects compared to conventional X-ray treatments. However, due to the high complexity of IMPT plans, each plan must be independently validated to ensure the safety and efficacy of the radiation exposure to the patient. While ion chambers are currently used for this purpose, their limitations-particularly in angled-beam measurements and multi-depth assessments-hinder their effectiveness. Silicon-based detectors, commonly used in X-ray therapy, are unsuitable for IMPT due to their rapid degradation under proton irradiation. In this study, a $\beta$-Ga$_2$O$_3$-based metal-semiconductor-metal (MSM) detector was evaluated and compared with a commercial ion chamber using a MEVION S250i proton accelerator. The $\beta$-Ga$_2$O$_3$ detector demonstrated reliable detection of single-pulse proton doses as low as 0.26 MU and exhibited a linear charge-to-dose relationship across a wide range of irradiation conditions. Furthermore, its measurement variability was comparable to that of the ion chamber, with improved sensitivity observed at higher bias voltages. These results highlight the strong potential of $\beta$-Ga$_2$O$_3$ as a radiation-hard detector material for accurate dose verification in IMPT.

physics.med-ph↗

arXiv2025

Study of Optical Properties of MOCVD-Grown Rutile GeO2 Films

Rutile germanium dioxide (r-GeO$_2$) is a promising ultra-wide bandgap (UWBG) semiconductor, offering a high theoretical Baliga figure of merit, potential for p-type doping, and favorable thermal and electrical properties. In this work, we present a comprehensive optical investigation of crystalline r-GeO$_2$ thin films grown on r-TiO$_2$ (001) substrates via metal-organic chemical vapor deposition (MOCVD). Cathodoluminescence (CL) spectroscopy reveals broad visible emissions with distinct peaks near 470~nm and 520~nm. CL mapping indicates enhanced emission intensity in regions with larger crystalline domains, highlighting the correlation between domain size and optical quality. X-ray photoelectron spectroscopy (XPS) confirms the presence of Ge$^{4+}$ oxidation state and provides a bandgap estimation of $\sim$4.75~eV based on valence band and secondary electron cutoff analysis. UV--Vis transmittance measurements show a sharp absorption edge near 250--260~nm, corresponding to an optical bandgap in the range of 4.81--5.0~eV. These findings offer valuable insights into the defect-related emission behavior and band-edge characteristics of r-GeO$_2$, reinforcing its potential for future applications in power electronics and deep-ultraviolet optoelectronic devices.

cond-mat.mtrl-sci↗

arXiv2025

Phase Competition and Rutile Phase Stabilization of Growing GeO2 Films by MOCVD

Rutile germanium dioxide (r-GeO2) is an ultra-wide bandgap semiconductor with potential for ambipolar doping, making it a promising candidate for next-generation power electronics and optoelectronics. Growth of phase-pure r-GeO2 films by vapor phase techniques like metalorganic chemical vapor deposition (MOCVD) is challenging because of polymorphic competition from amorphous and quartz GeO2. Here, we introduce seed-driven stepwise crystallization (SDSC) as a segmented growth strategy for obtaining r-GeO2 films on r-TiO2 (001) substrate. SDSC divides the growth into repeated cycles of film deposition and cooling-heating ramps, which suppress the non-rutile phases. We discuss the underlying mechanisms of phase selection during SDSC growth. We demonstrate continuous, phase-pure, partially epitaxial r-GeO2 (001) films exhibiting x-ray rocking curves with a FWHM of 597 arcsec. SDSC-based growth provides a generalizable pathway for selective vapor-phase growth of metastable or unstable phases, offering new opportunities for phase-selective thin-film engineering.

cond-mat.mtrl-sci↗

arXiv2025

RLEP: Reinforcement Learning with Experience Replay for LLM Reasoning

Reinforcement learning (RL) for large language models is an energy-intensive endeavor: training can be unstable, and the policy may gradually drift away from its pretrained weights. We present \emph{RLEP}\, -- \,Reinforcement Learning with Experience rePlay\, -- \,a two-phase framework that first collects verified trajectories and then replays them during subsequent training. At every update step, the policy is optimized on mini-batches that blend newly generated rollouts with these replayed successes. By replaying high-quality examples, RLEP steers the model away from fruitless exploration, focuses learning on promising reasoning paths, and delivers both faster convergence and stronger final performance. On the Qwen2.5-Math-7B base model, RLEP reaches baseline peak accuracy with substantially fewer updates and ultimately surpasses it, improving accuracy on AIME-2024 from 38.2% to 39.9%, on AIME-2025 from 19.8% to 22.3%, and on AMC-2023 from 77.0% to 82.2%. Our code, datasets, and checkpoints are publicly available at https://github.com/Kwai-Klear/RLEP to facilitate reproducibility and further research.

cs.CL↗

arXiv2025

Phase Evolution and Substrate-Dependent Nucleation of Quartz GeO$_2$ Films Grown by MOCVD on r- and c-Plane Sapphires

Ultrawide-bandgap (UWBG) semiconductors, such as GeO$_2$, are gaining significant attention for their potential in high-performance applications, particularly in piezoelectric devices. Despite extensive research, a comprehensive understanding of the growth dynamics and phase evolution of GeO$_2$ films via metal-organic chemical vapor deposition (MOCVD) remains insufficient. In this study, we investigate the growth behavior and morphological evolution of GeO$_2$ thin films on r-plane and c-plane sapphire substrates for the MOCVD growth process. The temporal evolution of crystallization and the amorphous-to-quartz phase transition are systematically elucidated for the first time. As growth time increases, the spherulitic quartz patterns expand in size, and elevated growth temperatures are found to enhance the crystallization rate. Distinct morphological symmetries emerge depending on the substrate orientation: quadrangular patterns on r-plane sapphire and hexagonal patterns on c-plane sapphire. Atomic force microscopy reveals that these spherulitic domains exhibit pyramid-like surface topography, consistent with volumetric contraction during the amorphous-to-quartz phase transition. These findings offer new insights into the phase evolution and substrate-dependent crystallization behavior of GeO$_2$ films grown by MOCVD.

cond-mat.mtrl-sci↗

arXiv2025

Siegel-Veech Measures of Convex Flat Cone Spheres

A classical theorem of Siegel gives the average number of lattice points in bounded subsets of $\mathbb{R}^n$. Motivated by this result, Veech introduced an analogue for translation surfaces, known as the Siegel-Veech formula, which describes the average number of saddle connections of bounded length on the moduli space of translation surfaces. However, no such formula is known for flat surfaces with cone angles that are irrational multiples of $\pi$. A convex flat cone sphere is a Riemann sphere equipped with a conformal flat metric with conical singularities, all of whose cone angles lie in the interval $(0, 2\pi)$. In this paper, we extend the Siegel-Veech formula to this setting. We define a generalized Siegel-Veech transform and prove that it belongs to $L^\infty$ on the moduli space. This leads to the definition of a Siegel-Veech measure on $\mathbb{R}_{>0}$, obtained by integrating the Siegel-Veech transform over the moduli space. This measure can be viewed as a generalization of the classical Siegel-Veech formula. We show that it is absolutely continuous and piecewise real analytic. Finally, we study the asymptotic behavior of this measure on small intervals $(0,\varepsilon)$ as $\varepsilon \to 0$, providing an analogue of Siegel-Veech constants for convex flat cone spheres.

math.GT↗

arXiv2025

Carbon-Nanotube/$\beta$-Ga$_2$O$_3$ Heterojunction PIN Diodes

$\beta$-Ga$_2$O$_3$ is gaining attention as a promising semiconductor for next-generation high-power, high-efficiency, and high-temperature electronic devices, thanks to its exceptional material properties. However, challenges such as the lack of viable p-type doping have hindered its full potential, particularly in the development of ambipolar devices. This work introduces a novel heterojunction diode (HD) that combines p-type carbon nanotubes (CNTs) with i/n-type $\beta$-Ga$_2$O$_3$ to overcome these limitations. For the first time, a CNT/$\beta$-Ga$_2$O$_3$ hetero-p-n-junction diode is fabricated. Compared to a traditional Schottky barrier diode (SBD) with the same $\beta$-Ga$_2$O$_3$ epilayer, the CNT/$\beta$-Ga$_2$O$_3$ HD demonstrates significant improvements, including a higher rectifying ratio ($1.2 \times 10^{11}$), a larger turn-on voltage (1.96 V), a drastically reduced leakage current at temperatures up to 300 {\deg}C, and a 26.7% increase in breakdown voltage. Notably, the CNT/$\beta$-Ga$_2$O$_3$ HD exhibits a low ideality factor of 1.02, signifying an ideal interface between the materials. These results underline the potential of CNT/$\beta$-Ga$_2$O$_3$ heterojunctions for electronic applications, offering a promising solution to current limitations in $\beta$-Ga$_2$O$_3$-based devices.

physics.app-ph↗

arXiv2025

Degradation of 2.4-kV $Ga_{2}O_{3}$ Schottky Barrier Diode at High Temperatures up to 500 {\deg}C

Ga2O3 Schottky barrier diodes featuring a field plate and a composite SiO2/SiNx dielectric layer beneath the field plate were fabricated, achieving a breakdown voltage of 2.4 kV at room temperature. Electrical performance and degradation were analyzed via I-V and C-V measurements from 25 {\deg}C to 500 {\deg}C, revealing temperature-dependent transport, interface stability, and device stability. Upon returning to room temperature, the diodes exhibited nearly unchanged forward characteristics, while the breakdown voltage declined significantly from 2.4 kV to 700 V. This behavior indicates a temperature-induced reduction in the barrier height. Detailed analysis revealed that variable range hopping (VRH) dominated the leakage mechanism at moderate temperatures, while thermal emission (TE) became increasingly significant at temperatures exceeding 400 {\deg}C.

cond-mat.mtrl-sci↗