UC Berkeley

Temperature and thermal conductivity are key parameters in applications ranging from microelectronics to energy conversion. As we strive to advance the state of the art, feature sizes continue to shrink, operating temperatures are pushed even higher, and in general devices must function in increasingly extreme regimes. Consequently, many of the traditional methods used to measure temperature or thermal conductivity are no longer applicable. This continued paradigm shift thus necessitates the development of new thermal metrology.

In this dissertation, I develop far-field optical thermometry techniques to measure temperature and thermal conductivity in challenging environments – namely, at temperatures up to 3,000 K and in systems with sub-50 nm features. All of these techniques have the benefit of far-field, non-contact operation, which allows for measurements in scenarios where many conventional methods fail for both practical and fundamental reasons. I begin by describing a Planck spectral fitting method developed to measure the peak temperature of reduced graphene oxide (RGO)-based films that are Joule heated to temperatures as high as 3,000 K. I also describe a “conduction shoulder” fitting method that combines image processing with a nonlinear, multimode heat transfer model in order to experimentally determine the high-temperature thermal conductivity of these same RGO films. Next, I introduce a far-field optical nanothermometry technique based on the temperature-dependent luminescence of individual NaYF4:Yb3+,Er3+ nanoparticles. This technique allows for single-point temperature measurements with spatial resolution of 50 nm or better. I then report apparent self-heating effects at high excitation intensities that indicate an apparent temperature rise > 50 K at an excitation intensity of 2.5 x 105 W cm-2 if interpreted as thermal; however, I demonstrate that these effects are in fact non-thermal in nature. Instead, rate equation modeling shows that these effects are due to both radiative and non-radiative relaxation from high energy excited states in Er3+ that are heavily populated at high excitation intensities. Finally, I demonstrate that individual nanoparticle thermometers can be used to measure nanoscale hotspots that result from both optical heating via a tightly focused laser beam and non- uniform Joule heating caused by a geometric effect known as current crowding.