Quantitative study on thermoreflectance linear relation

Abstract

Standard thermoreflectance-based measurements have been routinely taken on thin metal transducer (Au or Al) deposited samples. This is based on the fundamental hypothesis that the reflectance change (ΔR/R) of the metal surface is directly and linearly related to the temperature change (ΔT), within a wide but finite temperature range (Trange). The quantitative study on Trange has been ignored for a long time, which would possibly cause severe measurement issues and impede the possible new applications that the thermoreflectance measurements are taken on new metals or even directly on non-metals. Here, we present an approach that combines multiple probe wavelengths’ nanosecond transient thermoreflectance technique with a transient temperature rise model to study the linear relation. This method enables fast and accurate determination of the Trange and the proportional coefficient (commonly called the thermoreflectance coefficient, Cth). We studied the commonly used metal transducers (Au and Al) and found that Au illuminated at 532 nm has a considerably larger Trange (from room T to at least 225 °C), with respect to Al illuminated at 785 nm (room T to 150 °C). The linear relationships of uncommon Ni and Ti metals are valid from room temperature to ∼115 °C, illuminated at 785 and 660 nm, respectively. Non-linearity was observed for Al, Ni, and Ti metals when the temperature was elevated above the quantified Trange. This method enables a facile and reliable platform to characterize thermoreflectance properties and better understand the mechanism of thermoreflectance linear relationship.