The increasing computing power of AI presents a major challenge for high-power chip solution and heat dissipation. Boron nitride nanosheet-based thermal interface materials (BNNS-based TIMs) exhibit excellent electrical insulation property, ensuring the secure and stable operation of chips. However, the efficiency of micro/nano interfacial thermal transport is limited, impeding further enhancements in the thermal conductivity (TC) of BNNS-based TIMs. Here, a strategy of surface functionalization is reported to unlock the trade-off between the intrinsic and interfacial thermal transport of BNNS within TIMs. These results suggest that the surface functionalization maintains the intrinsic high TC of BNNS while significantly increasing binding energy between micro/nano interfaces in BNNS-based TIMs, effectively reducing interfacial thermal resistance of BNNS joint interfaces and interfaces between BNNSs and the matrix by 50% and 26.1%, respectively. The BNNS-based TIMs exhibit excellent TC (≈21-25W/(m·K)) and ultralow Young's modulus, which can promote the development of flexible high-performance chip cooling technology in the AI industry.
Read full abstract