2.4. Measurement Strategy

The AFM is relatively slow, compared to FSC. The AFM used in this work needs about 30 min to collect a single image with the needed resolution of 512 × 512 pixel per image. True in situ imaging is therefore limited to slow processes such as crystallization near the equilibrium melting temperature (Tm0) or near the glass transition (T_g). In the latter case, combination with the fast cooling capabilities of the FSC are advantageous since the maximum of the crystallization rate can easily be bypassed for most polymers. A typical temperature–time profile for such studies is shown in Figure 4a. It consists of melting the sample by heating it to above Tm0, quenching to the crystallization temperature (T_c), annealing for several hours and collecting the AFM images, cooling and final heating. The cooling steps, if possible, are performed at rates far above the critical cooling rate to avoid any crystallization or nucleation [2]. The final heating scan, in contrast, is performed at a rate which gives the best results regarding detection of the temperature and enthalpy of melting for the used sensor and the sample under investigation.

Program for studying isothermal crystallization by AFM-FSC combination. (a)—taking AFM images during slow crystallization at a temperature either close to Tm0 or, using quenching, close to T_g. (b)—time-resolved imaging, using ultra-fast cooling and heating ability of FSC. In both cases, the final FSC heating scan after crystallization is taken at 5000 K/s starting at ambient temperature.

Since crystallization must be slow to allow in situ AFM imaging, there is no useful calorimetric signal during the isotherm. The final heating scan is therefore used to determine the crystallinity from the enthalpy of fusion and the crystal stability from the melting temperature. If a rate of 100,000 K/s is not fast enough to prevent further crystallization on cooling below T_c, as for polyethylene [53], heating can start directly from the annealing temperature. In this case melting may partially occur in the transient stage at the beginning of the FSC scan when the programmed heating rate is not yet reached.

If crystallization is fast and in situ observation is not possible, the fast cooling capability of the FSC allows for a different experiment approach, as shown in Figure 4b. As before, the sample is first heated to above Tm0 and then quenched to T_c. After an appropriate annealing time, crystallization is interrupted by cooling the sample at overcritical rate to a temperature significantly below the glass transition. If this temperature is low enough for the studied polymer, then crystallization is fully stopped and there is sufficient time to collect AFM images with high resolution. After taking the AFM image, the sample is quickly heated back to T_c and crystallization is allowed to proceed for a certain time. Then the sample is again quenched and the next image is taken. This procedure is repeated until the sample is fully crystallized or any intermediate stage of interest is reached. More complex temperature–time profiles are possible, as shown in Section 3.3 below. The final heating scan always allows for quantifying the reached crystallinity. The advantage of this method is that it provides enough time for AFM imaging after only 1 ms annealing, which is not accessible directly, not even with a video type AFM.

For quantification of crystallinity of the sample, its mass (m) was determined from the measured heat capacity step height at the glass transition temperature of the fully amorphous sample and the known step height in the specific heat capacity [54]:

where ΔΦ is the measured heat flow step at the glass transition of the fully amorphous sample, ß is the heating rate and Δc_p is the step in specific heat capacity at the glass transition.

Crystallinity was determined as the ratio between the measured heat of fusion and the known heat of fusion of the 100% crystalline polymer:

where ΔH is the measured heat of fusion and ΔH^∞ is the heat of fusion of the fully crystalline material, taken from the ATHAS data bank [55].