Searcharxiv⌕ Search

arXiv subjects

Shilong Chen

Publications and source records attributed to Shilong Chen.

3 recordsLinked to original sources

Influences of $N_2$ flow rate on the superconducting properties and dimensionality of epitaxial TiN thin films

TiN thin films have received great attention in the field of superconducting quantum computing due to their low loss characteristics derived from stable chemical and thermal properties, as well as excellent electrical conductivity. In this study, TiN thin films with a thickness of 35 nm were epitaxially grown on MgO substrates using the DC reactive magnetron sputtering technique. The crystal structure, film thickness, and surface morphology of the thin films were characterized. The films exhibit a very flat surface with a root mean square roughness of only a few tenths of a nanometer. The influences of nitrogen ($N_2$) flow rate on the surface morphology, superconducting transition temperature ($T_c$), residual resistivity ratio (RRR), normal-state resistivity ($ρ_n$), upper critical field, and superconducting dimensionality were investigated systematically. The experimental results show that the TiN film deposited with a $N_2$ flow rate of 1.5 sccm exhibits optimal performance, including the highest $T_c$ (4.93 K), the largest RRR, the lowest $ρ_n$, and the lowest slope of the out-of-film upper critical field near Tc. Moreover, it is found that the background vacuum inside the growth chamber has a significant impact on the superconducting properties of TiN films, and the degradation of vacuum degree can lead to the decrease of $T_c$. The upper critical field reveals a significant anisotropic feature. Importantly, the in-film upper critical field data demonstrates that the optimization of sample quality leads to deviation from two-dimensional superconducting behavior, which provides a new perspective on the influencing factors of superconducting dimensionality beyond the thickness of films. The present results provide useful references for a comprehensive understanding of the relationship between growth parameters and the superconducting properties of thin films.

cond-mat.supr-con↗

Physical origin of very-high-energy gamma rays from the low-luminosity active galactic nucleus NGC 4278 and implications for neutrino observations

Relativistic jets in active galactic nuclei (AGNs) are known to accelerate particles to extreme energies, yet the physical origin of very-high-energy (VHE) emission from low-luminosity (LL) AGNs remains unclear. NGC 4278, a nearby LL AGN, has recently been identified as a VHE source following detections by LHAASO. In this work, we present a multiwavelength and multimessenger analysis to investigate the origin of this emission. Swift X-Ray Telescope monitoring reveals a quasi-quiescent state characterized by the low X-ray flux. Modeling the broadband spectral energy distribution with the leptohadronic code AMES, we find that a standard one-zone synchrotron self-Compton model underpredicts the VHE flux, unless a relatively high Doppler factor ($δ\gtrsim 6$) is invoked. Alternatively, an external inverse-Compton (EIC) scenario-scattering seed photons from a radiatively inefficient accretion flow-provides a good description of the broadband emission with modest jet power and Doppler factor. We further explore neutrino production within a leptohadronic framework. The EIC model in the quasi-quiescent state yields the largest predicted number of muon neutrinos, reaching $N_{ν_μ} \sim 0.001$ over 15 yr of IceCube observations (assuming that 0.1\% of the Eddington luminosity is converted into high-energy protons). Future multimessenger observations are essential to unveil the details of the high-energy processes of NGC 4278.

astro-ph.HE↗

SketchGraphNet: A Memory-Efficient Hybrid Graph Transformer for Large-Scale Sketch Corpora Recognition

This work investigates large-scale sketch recognition from a graph-native perspective, where free-hand sketches are directly modeled as structured graphs rather than raster images or stroke sequences. We propose SketchGraphNet, a hybrid graph neural architecture that integrates local message passing with a memory-efficient global attention mechanism, without relying on auxiliary positional or structural encodings. To support systematic evaluation, we construct SketchGraph, a large-scale benchmark comprising 3.44 million graph-structured sketches across 344 categories, with two variants (A and R) to reflect different noise conditions. Each sketch is represented as a spatiotemporal graph with normalized stroke-order attributes. On SketchGraph-A and SketchGraph-R, SketchGraphNet achieves Top-1 accuracies of 83.62% and 87.61%, respectively, under a unified training configuration. MemEffAttn further reduces peak GPU memory by over 40% and training time by more than 30% compared with Performer-based global attention, while maintaining comparable accuracy.

cs.CV↗