*3.5. AFM Characterizations*

Figure 7 shows that all the studied sampled are characterized by the typical filamentary structure of the tissue fibers on the nanoscale.

Thanks to the high resolution achieved by the AFM technique, it is possible to notice that both the two sample pairs COL\_GPTES (Figure 7c), COH\_GPTES (Figure 7d) and COL\_G-PNPA (Figure 7e), COH\_G-PNPA (Figure 7f) are rougher than the untreated tissues COL\_UT (Figure 7a), COH\_UT (Figure 7b), indicating that the GPTES and PNPA intimately and homogeneously wrap the fibers. This result is in good agreemen<sup>t</sup> with the EDS Si mapping. In the lower right corner of Figure 7f, it is possible to observe the presence of a lack in the coating; the study of the line profiles in that area allows us to evaluate the thickness of the coating that ranges between 2.5 and 4 nm.

**Figure 7.** AFM micrograph of COL\_UT (**a**), COH\_UT (**b**), COL\_GPTES (**c**), COH\_GPTES (**d**), COL\_G-PNPA (**e**), and COH\_G-PNPA (**f**).

### *3.6. PNPA in Vitro Di*ff*usion Studies*

In vitro diffusion studies on the developed cotton-based textiles were carried out with the aim of evaluating their ability to release in a controlled manner the synthesized PEA derivative compared to a standard solution of the molecule (Scheme 3).

**Scheme 3.** Schematic representation on the PNPA controlled release from the functional sol–gel coated cotton by stimuli effect.

In the performed experiments, Strat-M® membranes were used as a synthetic alternative, which is predictive of the diffusion process occurring through human skin. The studies involved two different textiles, such as COL\_G-PNPA and COH\_G-PNPA, prepared employing the obtained G-PNPA functional sol. The obtained results are expressed as cumulative diffused amount (%) and the diffusion profiles for the tested items are reported in Figure 8.

**Figure 8.** In vitro diffusion profiles.

The two G-PNPA-coated cotton textiles, COL\_G-PNPA and COH\_G-PNPA, show similar results in the performed in vitro diffusion studies. In the case of the COL\_G-PNPA sample, indeed, the amount of released PNPA is equal to 5% within the first hour reaching the 48% and 66% in 6 and 24 h, respectively, while the COH\_G-PNPA sample exhibits a value of 7% after the first hour achieving the 53% and 73% at the time points of 6 and 24 h. On the other hand, the 14% of PNPA is released after the first hour from the control solution reaching the 69% and 89% at 6 and 24 h. The experimental data confirm the ability of both the developed textiles to release in a controlled way the synthesized PEA derivative. These results could be related to the rate-limiting steps in drug release reported in the literature [33,34]: the drug diffusion within the polymer matrix and the rate of polymer swelling. Depending on the presence of sweat, simulated with the buffer solution in contact with the PNPA-treated fabric, the weak interactions between the treated fabric and the antioxidant molecule slowly disappear until the drug is completely released. The presence of the sol–gel matrix increases the bonding interactions between the network and the drug, slowing down its release.
