锡装饰的氮化硼纳米管增强液态金属界面热管理材料用于半导体封装

Liquid metal (LM) with notable thermal conductivity might be a promising candidate for the thermal management of central processing unit (CPU) and graphic processing unit (GPU) used for artificial intelligence (AI) since conventional polymer-based thermal interface materials (TIMs) are not sufficient to dissipate the heat produced by these advanced chips. However, it also has disadvantages, such as leakage, surface-spreading concerns, and processability difficulties due to its inherent high surface tension. The proposed solution to mitigate these issues might be metallic particle incorporation with LM, but the incorporation of those metal particles in general causes the intermetallic reactions, resulting in the deteriorated thermal performance and processability. Here, we suggest a facile and straightforward scheme using a composite of boron nitride nanotubes (BNNTs) with LM, while in situ decoration of Sn (tin) nanoparticles onto the surface of BNNT (BNNT-Sn) through a reduction reaction and electrostatic attraction improves BNNT dispersion and wettability in the LM matrix. Consequently, BNNT-Sn achieves superior phonon conduction pathways through the nanotubes by increasing the thermal conductivity as high as 30%, and more importantly, the wettability and processability properties are also enhanced by increasing the viscosity to 180 500~ ~750 mPa s from 2400 mPa s for neat LM with just 1.0 wt% of BNNT-Sn assimilation, with an improved rigidity evading leakage and adhesion as well. These improvements can be correlated with the intrinsic attributes of BNNT, such as an optimal aspect ratio and inherent thermal conductivity that contributed to the improvement of viscosity and thermal conduction.