Hot stage

The equipment employed to analyze the samples was an Olympus BX51 polarized optical microscope with a Linkam hot-stage coupled to control the temperature and the heating and cooling rates. To control the thermal process, liquid nitrogen was employed in the Linkam hot-stage. The micrographs were taken by an Olympus SC50 camera linked to the microscope. The samples were previously dissolved in DMF, and the solutions with a concentration of around 4% were drop-casted on a glass substrate and dried at room temperature before the measurements. The growth rate of the spherulites observed was calculated from the slope of the spherulite radius versus time plots, which were always found to be linear.

Pairs of otherwise identically-prepared samples were used with only the substrate being different (Si or AF). The film thickness profile of each sample was determined by measuring the surface topography of the film using AFM, and averaging the 3D topography along the direction of translational quasi-invariance of the sample to obtain a 2D thickness profile. The pairs of samples were then placed side-by-side for simultaneous annealing in either the home-built vacuum oven or on a hot stage (Linkam, UK). For a given pair of samples, the annealing temperature was held constant (between 120 and 160 °C), and chosen such that the PS was in its liquid melt state inducing the capillary-driven leveling of the thickness profiles. After a chosen duration of annealing t, the samples were rapidly quenched to room temperature, deep into the glassy state of PS, where the leveling process was temporarily halted. The broadening of the thickness profiles were measured using AFM. The samples were then further annealed, quenched, and measured again using AFM. The process of alternate annealing and AFM measurements was repeated until the measured thickness profiles became self-similar (Fig. 1b).

Second harmonic generation is an optical process that can occur upon exposure of a crystalline solid to a high power laser beam by which the sample re-emits a wave of half the wavelength (double frequency) of the incoming wave. The latter process is possible only if the sample crystallizes in a non-centrosymmetric crystal structure. This is true for both DMU forms II (space group P2₁2₁2) and I (space group Fdd2). Two DMU–water mixtures were analyzed by TR-SHG. Temperature was monitored using a Linkam hot stage which was systematically purged with N₂ before each experiment, so that the analyzed sample was in an inert atmosphere during the measurement. The solutions were first cooled to −50 °C at a cooling rate of 10 °C/min, then kept at −50 °C for two hours and finally heated at 5 °C/min. TR-SHG measurements were performed every minute during the heating process with a measurement time of 3 s each.