The aforementioned polymers were coated in a home-built hot-walled viscous flow reactor. Nitrogen (99.999%) was used as the carrier gas and the chamber pressure was ~ 3.5 Torr. Al2O3 was deposited using trimethyl aluminum (TMA, 98%, STREM Chemicals) and deionized water as precursors. Two different recipes were developed and used to coat the SAR-302503 samples. The first ALD recipe used relatively short precursor and reactant dose times separated by a nitrogen purge. The exposure times for TMA/N2/H2O/N2 were 0.1/45/0.1/30 s, respectively. The second method used similar dose times, but a “hold” step was added after the precursor dose. During the hold, the reactor chamber was isolated from the pump, allowing more time for the precursors to interact with the samples. The process timing followed TMA(hold)/N2/H2O(hold)/N2 = 0.5(60)/180/0.5(60)/180 s. The purge times are longer for the hold experiments to ensure complete precursor removal before the co-reactant dose. However, other experiments in our flow-tube reactors indicate that shorter purge times, similar to those for the shorter exposure times would also be sufficient. The depositions for each recipe were run at 50 and 100 °C. During coating, samples were positioned in the reactor to allow all surfaces to be exposed to process gases. As experimental controls, several polymer samples were exposed to the same thermal cycle, but did not receive any precursor exposure. Before starting the ALD run, samples were held in the reactor at the deposition temperature for at least 5 min under dry N2 flow to permit water desorption from the surface and near surface of the polymer. Experiments done with shorter thermal exposure times showed similar results. A silicon wafer was placed in the reactor with the polymers for each run to monitor Al2O3 growth. The different ALD treatments performed in this study are listed in Table 1.