Prior to the TDTR measurement, a thin layer of Al optical transducer (≈90 nm) was deposited (via magnetron sputtering) onto the sample. The water contents in the sample after Al coating were expected to be negligible, due to the long-time pumping in vacuum before deposition (90 °C baking for 2 h at 1× 10^-4 Torr, followed by pumping at <10^−7 Torr overnight). The Al coating prevents the water vapor from diffusing into the sample. To check these assumptions, AFM was used to verify the thicknesses of several hydrophilic samples (PANa, PALi, PVPA, PMCP, PDDA). A portion of the polymers was scratched off the substrate by a razor blade before Al deposition to create a step edge that facilitates the AFM measurement.
The height of the step edges matched with the thicknesses measured by ellipsometry at 90 °C to within the combined experimental uncertainties of <8 nm. In a TDTR measurement, the ratio of in-phase
(Vin) and out-of-phase (Vout) signals of the reflected probe beam recorded by an radio frequency (RF) lock-in amplifier with the reference frequency set at f is measured. Changing the modulation
frequency f between 1 and 9 MHz allowed the modification of the sensitivity of Vin/Vout with respect to thermal conductivity and heat capacity for thin films with thicknesses of ≈100 ∼ 200 nm. We typically
acquire data for Vin/Vout at three modulation frequencies, 1.1, 5.1, and 9.1 MHz, as a function of pump-probe delay time from −25 to 3600 ps. To extract the thermal conductivity and heat capacity from the TDTR data, we used a heat diffusion model to fit the three Vin/Vout curves simultaneously. The model consists of three layers: the Al transducer, the thin film for investigation, and the Si substrate. All
the geometric parameters and thermal properties of the Al transducer and Si substrate are measured separately or adopted from literature values; the only fitting variables are and C of
the thin film. The effects of the interfacial thermal conductance for the Al/polymer and polymer/Si interfaces are small since the Kapitza lengths of the interfaces are on the order of a
few nanometers. Therefore both of the interfacial conductances is set to a nominal value of 100 MW m^−2 K^−1 in the modeling. The TDTR measurements is validated by testing a reference
sample of PMMA. The extracted and C are 0.19 ± 0.02 W m^−1 K^−1 and 1.60 ± 0.15 J cm^−3 K^−1, respectively, which is consistent with prior studies.