Liquid Crystal Testing

Wang, Yuanyuan, et al. Applied Soft Computing 116 (2022): 108326.

This study presents a novel approach for detecting validly deformed particles in liquid crystal modules (LCMs) circuits, a critical task for ensuring the quality and functionality of these modules. The presence of conductive particles in LCMs can disrupt the circuit, making particle detection essential. However, challenges such as particle aggregation, varying particle sizes, overlapping deformations, and uneven illumination complicate accurate detection.

To address these challenges, the paper proposes the Multi-Scale Deep Adversarial Network (MSDA-Net), a robust solution for real-time detection of valid particles. The method utilizes a compact, multi-branch Generator for efficient, accurate detection of valid particles. This Generator employs a Coarse-to-Fine multi-scale feature extraction, which improves detection speed without compromising accuracy. Additionally, the Logical Focusing Attention (LFA) module refines the center location of the particles, enhancing detection precision.

The MSDA-Net also integrates a Multi-branch Deep Adversarial Strategy to distinguish valid from invalid particles based on their visual characteristics, leading to improved accuracy and reduced false positives. Extensive testing on real particle datasets shows that MSDA-Net outperforms state-of-the-art detection methods, demonstrating not only high detection accuracy but also the ability to process in real-time.

These results underscore the potential of MSDA-Net for improving the reliability and efficiency of particle detection in liquid crystal module circuits, offering a promising solution for the electronics manufacturing industry.

Li, Weijun, et al. Science Bulletin 66.21 (2021): 2199-2206.

This study explores the application of halide perovskite single crystals (HPSCs) in liquid crystal-based X-ray detection, demonstrating a new solvent-volatilization-limited-growth (SVG) strategy to control crystal growth and achieve high-quality HPSCs. Traditional crystal growth methods often lead to inconsistent rates and quality, but the SVG method stabilizes the process, resulting in single crystals with low trap density (2.8 x 10⁸ cm⁻³) and a high charge carrier mobility-lifetime product (μτ = 0.021 cm²/V). This exceptional crystal quality is crucial for optoelectronic applications, particularly in sensitive X-ray detectors.

The study further demonstrates that surface defects in the HPSCs are passivated using oxygen-containing molecules, reducing leakage currents by a factor of two. This improvement enhances the sensitivity of the X-ray detectors, which achieved a remarkable sensitivity of 1274 µC/(Gyair cm²) and a detection limit of 0.56 μGyair/s under 120 keV hard X-ray exposure.

The findings underscore the potential of SVG-grown HPSCs for applications beyond X-ray detection, including alloy composition analysis and metal flaw detection, signaling broad industrial implications. Additionally, the study provides insights into crystal growth dynamics that could advance the fabrication of high-quality polycrystalline films, thus facilitating the development of next-generation liquid crystal-based devices and systems.