arXiv subjects

Kai Fu

Publications and source records attributed to Kai Fu.

At least 19 recordsLinked to original sources

arXiv2026

Structural characterization of the rutile-GeO2/TiO2 (001) heterointerface

Rutile-GeO2 (r-GeO2) has emerged as a promising ultrawide-bandgap (UWBG) semiconductor for future high-power device applications because of its favorable combination of physical and electrical properties. This study has used high-resolution transmission electron microscopy (HR-TEM) to compare heterointerfaces between thin films of GeO2 grown by metal-organic chemical vapor deposition on closely-lattice-matched rutile TiO2 (r-TiO2) substrates at temperatures of 840° C and 925° C. Cross-sectional TEM observations reveal interfaces with two entirely different structures: an amorphous GeOx/r-TiO2 heterostructure with an abrupt, highly faceted and non-planar interface, or a fully epitaxial r-GeO2/r-TiO2 heterostructure characterized by a more wavy but often facetted interface. Structural analysis of the latter sample revealed a high density of strain-relieving interfacial dislocations, identified primarily as Lomer-type or 60°-type, that accommodated the ~4% lattice mismatch between the two materials, while chemical analysis showed very limited interfacial intermixing. Crystallographic analysis showed that the interface faceting was associated with rutile {101} facets, with characteristic facet angles of ~114°; these observations support thermally driven TiO2 (001) substrate faceting rather than an interface reconstruction resulting from GeO2 epitaxy. Overall, these findings provide useful insights into the mechanisms governing interface formation and strain relaxation in r-GeO2/r-TiO2 heterostructures and suggest that further refinement of growth conditions should enable improved epitaxial growth of r-GeO2 for future ultrawide-bandgap electronic device applications.

cond-mat.mtrl-sci↗

arXiv2026

10 kV $β$-Ga$_2$O$_3$ Sub-Bandgap Triggered Impurity Photoconductive Semiconductor Switches (i-PCSSs): Lessons for Optimal Design

$β$-Ga$_2$O$_3$ photoconductive semiconductor switches (PCSSs) are attractive for high-voltage operation, but above-bandgap triggering requires deep-ultraviolet excitation. Here, we investigate a 650-um-thick, 0.01% Fe-doped $β$-Ga$_2$O$_3$ impurity PCSS with a 5-mm lateral gap under continuous-wave (CW) 808-nm sub-bandgap excitation at biases up to 10 kV ( 20 kV cm^-1). The device exhibited clear sub-bandgap photoconductive switching. As the optical power density increased from 0.64 to 3.18 W cm^-2, the ON-state current increased progressively while the ON-state resistance decreased, consistent with enhanced defect-assisted photocarrier generation. A higher applied bias further strengthened the switching response through enhanced field-assisted carrier transport. The ON/OFF ratio increases with optical power density but exhibits a nonmonotonic bias dependence, reaching approximately 15-16 near 5-7 kV. At 10 kV, increased OFF-state conduction and non-ideal transient behavior reduce the switching contrast. These results demonstrate near-infrared photoconductive switching but also highlight critical trade-offs in photoconductive thresholds for impurity states.

cond-mat.mtrl-sci↗

arXiv2026

Point Processes and Gap Statistics for Translation Surfaces

We study the behavior of saddle connection directions on translation surfaces from the point process viewpoint. We define a direction point process for saddle connections of length at most $R>0$ on every translation surface. Given a general ergodic $\mathrm{SL}(2,\mathbb R)$-invariant measure $μ$ on a stratum of translation surfaces, we also define a slope point process associated with $μ$. By utilizing the dynamics and geometry of strata, we show that for almost every translation surface, the direction point processes, rescaled by $R^2$, converge in distribution to the slope point process as $R\to\infty$. Using Palm distribution theory, we prove the existence of limiting joint gap distributions and obtain new regularity results for gap distributions associated with the general $μ$. Finally, we study how these point processes and joint gap distributions behave as the genus $g$ tends to infinity. In the principal stratum, we show that the slope point processes, rescaled by $g^2$, converge in distribution to a Poisson point process. By passing to the corresponding Palm distributions, we show that the rescaled slope gaps converge jointly to independent exponential random variables, which is in stark contrast to the known non-exponential gap distributions in a fixed genus.

math.GT↗

arXiv2026

Bounds on saddle connections for flat spheres

We consider a flat metric with conical singularities on the sphere. Under the assumption that no partial sum of angle defects is equal to $2π$, we draw on the geometry of immersed disks to obtain an explicit upper bound on the number of saddle connections with at most $k$ self-intersections. Additionally, we establish an upper bound on their lengths for a surface with a normalized area. Finally, we apply these bounds to the counting of singular trajectories in irrational polygonal billiards.

math.GT↗

arXiv2026

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 $π$. 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 $(0, 2π)$. 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↗

arXiv2026

Uniform Length Estimates for Trajectories on Flat Cone Surfaces

This paper studies length estimates for trajectories on flat cone surfaces in terms of their self-intersection numbers. For an area-one flat cone surface, we obtain a lower bound for the length of a trajectory, with constants depending only on the flat metric. Our main focus is the case of convex flat cone spheres. We show that these constants can be chosen uniformly for such spheres with a positive curvature gap and a fixed number of singularities. Explicit values for these constants are also provided. Combined with a previously established upper bound, this yields uniform two-sided estimates for trajectory lengths on such flat cone spheres. As an application, we obtain uniform bounds for counting functions of trajectories on convex flat cone spheres and on convex polygonal billiards.

math.GT↗

arXiv2026

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

Thermal stability and phase transformation of conductive $α$-$(\mathrm{Al}_{0.16}\mathrm{Ga}_{0.84})_{2}\mathrm{O}_{3}/\mathrm{Ga}_{2}\mathrm{O}_{3}$ 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 $α$-$(\mathrm{Al}_{0.16}\mathrm{Ga}_{0.84})_{2}\mathrm{O}_{3}/\mathrm{Ga}_{2}\mathrm{O}_{3}$ 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 $β$-Ga$_2$O$_3$ Films Probed by In Situ HT-XRD under Different Reactor Heights

The crystallinity of $β$-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 $β$-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 $β$-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 $β$-Ga$_2$O$_3$ films and modulates their elevated-temperature structural response, providing practical insights for optimizing MOCVD reactor design for high-quality $β$-Ga$_2$O$_3$ growth.

cond-mat.mtrl-sci↗

arXiv2026

Dosimetry for Proton Therapy Using a β-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 β-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 β-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 β-Ga$_2$O$_3$-based detectors as robust dosimetry devices for IMPT applications.

physics.med-ph↗

arXiv2026

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↗

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↗

arXiv2026

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

In situ XRD Study of Strain Evolution in AlGaN/GaN HEMT at High Temperatures up to 1000 °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

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

$β$-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 $β$-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 $β$-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 $β$-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↗