2, Thales Research and Technology, Palaiseau, , France
Heat management has always been a key factor in the development of novel technologies – from cooling rocket engine nozzles to heat dissipation in today’s computer chips. To help these developments, it is tempting to use the highly heat conductive one- and two-dimensional (1D and 2D) materials, such as carbon nanotubes (CNT), and, more recently, graphene and hexagonal boron nitride material composed of atomically thin Van-der-Waals bonded layers. However, the integration of these materials in the real-life applications is rather challenging and requires novel approaches, in particular, to merge a nanoscale nature of these materials with their application in a real-world macroscale device – in, so-called, thermal interface materials (TIM). Novel measurements approaches are needed to explore the TIM physical properties such as heat conductivity of their components and thermal resistance of the structural interfaces in structures that are often buried within the sample volume.
In our work, we aimed to resolve these by using Scanning Thermal Microscopy (SThM) enhanced with dedicated sample preparation techniques, namely a unique cross-sectional tool (Beam-Exit Cross-section Polishing, BEXP), used to prepare a nanoscale flat tapered section of a material with a shallow angle (<10°). For TIM consisting of CNT brush attached to a Si substrate, BEXP creates a “nanoforest” of vertically aligned carbon nanotubes whose lengths increase from less than 10 nm to more than 1 mm. SThM then measures the thermal resistance of this resulting shaved “nanoforest” by scanning a nanoscale sharp tip integrated with a heater and a temperature sensor as a function of the CNT length. These measurements complemented with an analytical modelling allowed to extract thermal conductivities of in-plane and out-of-plane of CNT “nanoforest”, as a result gaining significant insights into the nature of the TIM sample thermal anisotropy. By using this approach to fully structured CNT TIM’s in combination with focused-ion beam, we were able to map the thermal conductance of the individual materials of the structure. Finally, by applying a temperature gradient across the TIM we measured the temperature drop from the hot to cold side.
In summary, this work demonstrates the usefulness of nanoscale probe microscopy for TIM engineering, allowing to estimate anisotropy of thermal resistance of CNT bundles and interfacial thermal conductance, providing working solutions for the research and development of new generation of heat management components.