ABSTRACT
Liquid metal (LM)-based thermal interface materials (TIMs) have the potential to dissipate high heat loads in modern electronics and often consist of LM microcapsules embedded in a polymer matrix. The shells of these microcapsules consist of a thin LM oxide that forms spontaneously. Unfortunately, these oxide shells degrade heat transfer between LM capsules. Thus, rupturing these oxide shells to release their LM and effectively bridge the microcapsules is critical for achieving the full potential of LM-based TIMs. While this process has been studied from an electrical perspective, such results do not fully translate to thermal applications because electrical transport requires only a single percolation path. In this work, we introduce a novel method to study the rupture mechanics of beds composed solely of LM capsules. Specifically, by measuring the electrical and thermal resistances of capsule beds during compression, we can distinguish between the pressure at which capsule rupture initiates and the pressure at which widespread capsule rupture occurs. These pressures significantly differ, and we find that the pressure for widespread rupture corresponds to a peak in thermal conductivity during compression; hence, this pressure is more relevant to LM thermal applications. Next, we quantify the rupture pressure dependence on LM capsule age, size distribution, and oxide shell chemical treatment. Our results show that large freshly prepared capsules yield higher thermal conductivities and rupture more easily. We also show that chemically treating the oxide shell further facilitates rupture and increases thermal conductivity. We achieve a thermal conductivity of 16 W m-1 K-1 at a pressure below 0.2 MPa for capsules treated with dodecanethiol and hydrochloric acid. Importantly, this pressure is within the acceptable range for TIM applications.
ABSTRACT
Room-temperature liquid metals (LMs) are attractive candidates for thermal interface materials (TIMs) because of their moderately high thermal conductivity and liquid nature, which allow them to conform well to mating surfaces with little thermal resistance. However, gallium-based LMs may be of concern due to the gallium-driven degradation of many metal microelectronic components. We present a three-component composite with LM, copper (Cu) microparticles, and a polymer matrix, as a cheaper, noncorrosive solution. The solid copper particles alloy with the gallium in the LM, in situ and at room temperature, immobilizing the LM and eliminating any corrosion issues of nearby components. Investigation of the structure-property-process relationship of the three-component composites reveals that the method and degree of additive blending dramatically alter the resulting thermal transport properties. In particular, microdispersion of any combination of the LM and Cu additives results in a large number of interfaces and a thermal conductivity below 2 W m-1 K-1. In contrast, a shorter blending procedure of premixed LM and Cu particle colloid into the polymer matrix yields a composite with polydispersed filler and effective intrinsic thermal conductivities of up to 17 W m-1 K-1 (effective thermal conductivity of up to 10 W m-1 K-1). The LM-Cu colloid alloying into CuGa2 provides a limited, but practical, time frame to cast the uncured composite into the desired shape, space, or void before the composite stiffens and cures with permanent characteristics.
