Thermal Property Estimation of Thin-Layered Structures by Means of Thermoreflectance Measurement and Network Identification by Deconvolution Algorithm

Abstract

Abstract Laser-induced forward transfer (LIFT) is a powerful tool for micro and nanoscale digital printing of metals for electronic packaging. In the metal LIFT process, the donor thin metal film is propelled to the receiving substrate and deposited on it. Morphology of the deposited metal varies with the thermodynamic responses of the donor thin film during and after the laser heating. Thus, the thermophysical properties of the multilayered donor sample are important to predict the LIFT process accurately. Here, we investigated thermophysical properties of a 100 nm-thick gold coated on 0.5 mm-thick sapphire and silicon substrates by means of the nanosecond time-domain thermoreflectance (ns-TDTR) analyzed by the network identification by deconvolution (NID) algorithm, which does not require numerical simulation or analytical solution. The NID algorithm enabled us to extract the thermal time constants of the sample from the nanosecond thermal decay of the sample surface. Furthermore, the cumulative and differential structure functions allowed us to investigate the heat flow path, giving the interfacial thermal resistance and the thermal conductivity of the substrate. After calibration of the NID algorithm using the thermal conductivity of the sapphire, the thermal conductivity of the silicon was determined to be 107–151 W/(m K), which is in good agreement with the widely accepted range of 110–148 W/(m K). Our study shows the feasibility of the structure function obtained from the single-shot TDTR experiments for thermal property estimation in laser processing and electronics packaging applications.