纳米片
氮化硼
材料科学
聚酰亚胺
复合材料
热导率
热的
氮化物
纳米技术
物理
图层(电子)
气象学
作者
Yiwei Wang,Guan Wang,Zhang Li,Bilal Iqbal Ayubi
标识
DOI:10.1021/acsanm.5c01083
摘要
Solid-state transformers (SSTs) are critical components in modern power systems, requiring insulation materials with high thermal conductivity and low dielectric loss to withstand prolonged high-frequency electrical stress and elevated temperatures. Although polyimide (PI) is widely used in SSTs, its low thermal conductivity and high dielectric loss limit its long-term performance. To address these challenges, we developed polydopamine-modified boron nitride nanosheets (BNNS-PDA) as fillers for PI composites. The PDA modification significantly improves the compatibility between BNNS and the PI matrix, resulting in composites with enhanced thermal conductivity (0.679 W/(m·K) at 5 wt %, 2.2 times that of pure PI) and reduced dielectric loss (0.00617 at 20 kHz, 37% lower than pure PI). These improvements lead to a 65% increase in high-frequency aging lifetime under 3 and 20 kHz, as well as a 62.5% reduction in surface temperature rise under 2 and 20 kHz. Molecular dynamics (MD) simulations reveal that PDA modification increases the interfacial interaction energy and hydrogen bond density between the BNNS and PI, enhancing interfacial stability. Density functional theory (DFT) calculations further visualize and quantify intermolecular hydrogen bonding interactions. When integrated with phase-field modeling, these enhanced interactions are reflected by elevated energy barrier parameters (α), effectively delaying electrical breakdown, and mitigating Joule heating under high-frequency conditions. However, excessively high α values accelerate the breakdown, highlighting the need to optimize doping levels for balanced performance. This study provides a multiscale understanding of the relationship between interfacial modifications and macroscopic performance, offering a practical strategy for designing high-performance insulation materials for high-frequency applications.
科研通智能强力驱动
Strongly Powered by AbleSci AI