As the downscaling of electronic devices pushes dimensions of its components towards the atomic limits, new measurement tools need to be developed to address new challenges. In particular, nanoscale heat transfer is a key mechanism which is known to limit the performance of nanoscale sized transistors in the processor chips and light emitting diodes and to define the performance of thermoelectric materials. In most cases the thermal properties of materials and devices defining their performance are to be investigated through all their three-dimensional (3D) structure (e.g. interconnects and transistor layers in the processor chips) and grain interfaces (e.g. in thermoelectric compounds) with the interfaces and layer thickness’ down to few nanometres. Furthermore, thermal properties at nanoscale lengths present many challenges to traditional techniques, e.g. thermal conductivity of films thinner than 100 nm is affected by surface scattering and mean-free path of the heat carriers (phonons and electrons) and often cannot be directly decoupled from interfacial resistance between the film and the substrate. In this report, we develop a novel approach addressing these challenges.
Here we report combination of an Ar-ion cross-sectional tool ideally suitable for the scanning microscopy (SPM) measurements - beam-exit cross-sectional polishing (BEXP) and a heated probe of scanning thermal microscopy (SThM) to measure thermal transport of buried layers of a wide range of materials. BEXP uses Ar-ion beams impinging on a sample at shallow angle (<10o). The cross-sectioned surface obtained has low-angle geometry and sub-nm surface roughness making it easily suitable for studies via standard SPM methods. We then use SThM to raster a thermosensitive probe on a surface quantifying probe-sample thermal resistance. By analysing the SThM signal of the wedge-shaped section of the probed material, we are able to extract the intrinsic thermal conductivity of the nanoscale thin layer material separating it from the probe-sample and sample-substrate thermal resistances by combining analytical and finite element modelling of the heat transport in the sample.
To validate our approach, we apply this method on different commonly used materials from semiconductors to insulators such as silicon dioxide, spin-on-glass and spin-on-polymers. Then we investigate multi-layered samples for optoelectronic applications such as MBE grown SiGe alloys and map the thermal conductance across the sample volume. Our results demonstrate its applicability for direct measurements of otherwise hard to obtain quantities for previously unknown materials. The ease of use of our method renders it suitable for a broad range of samples and opens new paths for fundamental and applied research.