*2.5. Characterization*

The thicknesses of all nonwovens were measured and their surface densities were determined by weighing. The surfaces of the nonwovens were examined using scanning electron microscopy (SEM) with energy dispersive spectroscopy (EDS) JEOL 6010LA (JEOL, Japan) at an accelerating voltage of 10 kV. Before the examination, the surfaces were vacuum sputtered with a 10 nm gold layer using a coater (Quorum EMS150R ES, UK). Diameter distributions of the fibers were determined from SEM micrographs.

Water contact angles (WCA) of 5 μl distilled water droplets with the nonwovens were determined at RT, using a goniometer 100-00-230 NRL Rame Hart (USA) with the Image Drop Analysis program. In each case, the WCA measurements were carried out five times and average values were calculated.

Thermal properties of the nonwovens were studied using differential scanning calorimetry (DSC 3 Mettler Toledo, Switzerland) during heating at 10 ◦C/min. Thermogravimetric analysis was performed using the TGA 550 from TA instruments (USA) in a nitrogen atmosphere at a heating rate of 20 ◦C/min.

Tensile drawing of nonwoven specimens was performed with the Linkam TST 350 Minitester (UK) at 25 ◦C. Three 10 mm wide strips of each material were drawn, with a distance between grips of 20 mm, at a rate of 2 mm/min (10%/min). The tests were carried out at least 48 h after preparation of the nonwovens.

Selected PLLA-based nonwovens were analyzed with Fourier-transform infrared spectroscopy (FTIR) using a Thermo Scientific Nicolet 6700 FT-IR instrument with an ATR (Attenuated Total Reflectance) (USA). FTIR spectra in the range of 500–4000 cm<sup>−</sup><sup>1</sup> were recorded with a 1 cm<sup>−</sup><sup>1</sup> resolution.

## **3. Results and Discussion**

The structure of the electrospun PLLA-based nonwovens, fiber morphology, and the fiber size distributions are shown in Figures 1 and 2, whereas parameters of the nonwovens are collected in Table 2. All nonwovens were composed of uniform and defectfree fibers, randomly distributed. It is worth noting that TGA thermograms, shown in Figure S3 in SI, evidence that the solvents evaporated during the electrospinning and post-electrospinning drying; weight loss at 100 ◦C was below 0.1% in each case. Moreover, the TGA thermograms in Figure S3 in SI and data collected in Table S1 in SI show that all the materials with LPSQ-COOMe exhibited better thermal stability in comparison to neat PLLA, as already reported [39].

**Figure 1.** SEM micrographs of nonwovens: PLLA (**a,b**), PL-CNT (**d,e**), PL-g-CNT (**g,h**) and fiber diameter distributions; PLLA (**c**), PL-CNT (**f**), PL-g-CNT (**i**). Water drops on nonwoven surfaces are shown in the insets.

**Figure 2.** SEM micrographs of nonwovens: PL-LPSQ5 (**a,b**), PL-LPSQ5-CNT (**d,e**), PL-LPSQ10 (**g,h**), PL-LPSQ10-CNT (**j,k**) and fiber diameter distributions: PL-LPSQ5 (**c**), PL-LPSQ5-CNT (**f**), PL-LPSQ10 (**i**), PL-LPSQ10-CNT (**l**). Water drops on nonwoven surfaces are shown in the insets.

**Table 2.** Characteristics of PLLA-based nonwovens: thickness, surface density, average fiber diameter, the most probable diameter (at the histogram maximum) and water contact angle (WCA).


Diameters of a majority of neat PLLA fibers were in the 1.5–3.5 μm range, with an average of 2.5 μm, whereas diameters of PL-CNT and PL-g-CNT were smaller, mostly below 2 μm, with an average of 1.7 and 1.8 μm, respectively. The decrease of the PL-CNT fiber diameters was caused by an increase in the electrical conductivity of the solution. As a result, the solution underwent a greater extension under the influence of the electrostatic field, which caused the reduction of fiber diameters [6,57]. The same applies to PL-g-CNT electrospinning process, although it is worth noting that PL-g-CNT concentration in the solution had to be decreased in comparison to that of PL-CNT because of increased viscosity due to grafting of polymer chains on MWCNT. The pores on the fiber surfaces resemble those observed in [44,58] due to the breath figure mechanism. As a result of the evaporative cooling of the solvent, water vapor condensed and droplets formed on the surface of the fiber jets, causing the formation of pores. Moreover, the pores were elongated in the directions parallel to the fiber axes, as a result of deformation during jet stretching.

The addition of 5 and 10 wt.% of LPSQ-COOMe to PLLA resulted in thinner fibers, mostly with diameters ranging from 1 to 3 μm, and average diameters of 1.8 and 1.7 μm, respectively, as shown in Figure 2 and Table 2. The decreased average fibers diameters of PL-LPSQ5 and PL-LPSQ10 as compared to those of PLLA could result from a decrease of solution viscosity resulting from the presence of LPSQ-COOMe. Fibers of PL-LPSQ5-CNT and PL-LPSQ10-CNT were even thinner, mostly with diameters ranging from 0.5 to 1.5 μm and from 1 to 2 μm, with average values of 0.9 and 1.2 μm, respectively. The nonwoven thickness ranged from 0.68 to 0.84 mm. The thickest was the nonwoven of neat PLLA, whereas the thinnest were the nonwovens with MWCNT, composed of the markedly thinner fibers.

Figures 1 and 2 show also water droplets placed on the nonwoven surfaces, whereas the WCA values are listed in Table 2. All the obtained electrospun PLLA-based nonwovens were hydrophobic, with WCA values close to 155◦. It is known that hydrophobicity of nonwovens results from their surface roughness and air entrapped between fibers [59].

DSC heating thermograms of PLLA-based nonwovens are compared in Figure 3, whereas the calorimetric parameters are collected in Table 3. Each thermogram shows a glass transition, a cold crystallization exotherms, and a melting endotherm. In addition, on all thermograms, pre-melting exotherms, attributed to the transition from the disordered orthorhombic alpha' form to the ordered orthorhombic alpha form [60], are visible, with a maximum at 157–159 ◦C. The cold crystallization exotherms, especially those of the materials with LPSQ-COOMe, exhibit long high-temperature tails, overlapping the pre-melting recrystallization exotherms, hence the latter are less pronounced and their enthalpies listed in Table 3 are only approximate. However, in each case the sum of enthalpies of the exothermic effects was equal to the melting enthalpy, evidencing that the nonwovens were amorphous before heating in DSC. Tg of neat PLLA nonwoven was at 59 ◦C, whereas the cold crystallization peak (Tcc) was at 97 ◦C. Tgs of both PL-CNT and PL-g-CNT were nearly the same, at 58 ◦C. Tgs of the nonwovens of PLLA blends with 5 and 10 wt.% of LPSQ-COOMe were lower, at 55 and 52–53 ◦C, respectively, due to the plasticizing effect of the additive. However, taking into account the Tgs of PLLA, LPSQ-COOMe and PLLA blends with LPSQ-COOMe, it can be calculated, using the Fox equation, that in PLLA with 5 and 10 wt.% of LPSQ-COOMe about 3 and 5 wt.% of the additive, respectively, is dispersed on a molecular level. This is in accordance with ref. [38], where phase separation was observed in PLA blends with LPSQ-COOMe. Figure S4 in SI shows SEM EDS mapping of silicon in PL-LPSQ5 and PL-LPSQ10 that confirms the presence of LPSQ-COOMe and its dispersion in the PLLA matrix. Inclusions of the additive were not discernible, most probably being too small, as in ref. [38]. It is also observed that Tcc of composite materials decreased in comparison to that of neat PLLA nonwoven, to 92 ◦C for PL-CNT and even more, to 87 ◦C, for PL-g-CNT. Tccs of PL-LPSQ5 and PL-LPSQ10 nonwovens were even lower, at 86 ◦C and 82 ◦C, respectively. The presence of CNT further enhanced the decrease of Tcc, to 81 ◦C for both PL-LPSQ5-CNT and PL-LPSQ10-CNT. In turn, the melting peak temperatures (T m) of the materials were at 173–175 ◦C. The only exception was PL-g-CNT, whose T m of

179 ◦C was higher than that of PL-CNT, indicating the melting of thicker crystals, despite its low Tcc. CNT are known to nucleate crystallization of PLA [61], whereas LPSQ-COOMe increases the mobility of PLA chains and also enhances its cold crystallization [38]. It is worth noting that PLLA in PL-g-CNT was optically pure. Moreover, the additives can enhance the orientation of PLLA in the fibers during electrospinning, which also can promote cold-crystallization of the polymer.

**Figure 3.** DSC heating thermograms of PLLA-based nonwovens.

**Table 3.** Calorimetric parameters of PLLA-based nonwovens: Tg–the glass transition temperature, Tcc, Hcc–the cold crystallization peak temperature and enthalpy, respectively, Tm and Hm–the melting peak and enthalpy, respectively, the enthalpy of the pre-melting recrystallization is given in brackets.


Figure 4 shows the tensile behavior of the nonwovens, whereas the mechanical parameters are collected in Table 4. Fibers in the nonwovens were randomly distributed, without any preferred orientation. When the nonwovens were strained, the fibers tended to orient parallel to the stretching direction. Loosely connected structure without strong bonds between the fibers at their cross points facilitated the fiber alignment during drawing. However, some fibers broke at early stages of drawing, thus decreasing the tensile stress. In Figure 4 the engineering stress, calculated as a ratio of force to initial cross-section area of the nonwoven, is plotted vs. engineering strain. It is seen that with increasing strain the stress increases, passes through a maximum, and decreases, less or more sharply. The maximum stress value recorded during the drawing of neat PLLA nonwoven was 0.5 MPa. Similar values measured for electrospun neat PLA nonwovens were reported in [49,62]. The presence of 0.1 wt.% of CNT increased significantly the tensile strength to 0.8 and 0.95 MPa for PL-CNT and PL-g-CNT, respectively. The effect of 5 wt.% of LPSQ-COOMe on the PLLA nonwoven strength was weak, but PL-LPSQ10 exhibited a strength of 0.95 MPa.

However, the highest strength was achieved for PL-LPSQ5-CNT and PL-LPSQ10-CNT, 1.1 and 1.2 MPa, which exceeded more than two times the strength of neat PLLA nonwoven. It is worth noting that the maximum stress of PL-LPSQ10-CNT was achieved at an elongation of approx. 80%, significantly larger than in the case of the other materials.

**Figure 4.** Stress-strain dependencies of PLLA-based nonwovens.


**Table 4.** Mechanical parameters of PLLA-based nonwovens.

CNT-containing materials are known to exhibit outstanding mechanical properties, due to stress transfer from the weaker polymer to the stronger nanofiller, which allows a higher loading to be achieved, thus the improvement in strength [14]. However, in the nonwovens studied the weight ratio of MWCNT to PLLA was only 1:1000, therefore we attribute the increase in strength also to the enhancement of PLLA chain orientation during electrospinning. An increase in electrical conductivity of the solution due to the presence of CNT results in a greater extension under the influence of the electrostatic field, which can promote not only the fiber diameter reduction but also the orientation of polymer chains. It must be noted that CNT are highly anisotropic particles, which tend to be oriented during fiber jet stretching. It is also worth noting that the effect of LPSQ-COOMe on the tensile properties of the PLLA nonwovens was different than on PLA films, whereas at 5 wt.% content of the additive a large elongation at break, 230%, was achieved, accompanied by a minor, approx. 10%, decrease of the tensile strength [38]. LPSQ-COOMe double-strand structure hampers coiling, and orientation of its macromolecules during electrospinning can also promote the orientation of PLLA chains, especially that they are capable of supramolecular interactions with PLLA. Side ester groups of LPSQ-COOMe may participate in weak C-H··· O=C hydrogen bonds, analogously to those in stereocomplex structures formed between enantiomeric chains of PLLA and poly(D-lactide) [63], and postulated for hybrid stereocomplex-PLA/LPSQ-COOMe blends [39]. The enhanced PLLA orientation can contribute to the improved strength of the modified nonwovens.

Figure 5 shows FTIR spectra of the selected nonwovens positioned perpendicular to the beam in the 2000–800 cm<sup>−</sup><sup>1</sup> range with the inset showing the enlarged peak at 1267 cm<sup>−</sup>1.

**Figure 5.** FTIR spectra of PLLAbased nonwovens in 2000–800 cm<sup>−</sup><sup>1</sup> range.

FTIR spectra of the selected nonwovens, shown in Figure 5, are typical for amorphous PLLA. 10 wt.% content of LPSQ-COOMe in PLLA corresponds to approx. 3 mol% content of functional groups, therefore the LPSQ-COOMe bands did not show up in the spectra, especially that some of them, e.g., those related to C=O or CH3 groups, are overlapping with PLLA bands [39]. The band near 1750 cm<sup>−</sup><sup>1</sup> corresponds to C=O stretching, whereas the bands of the CH3 asymmetric and symmetric bending and the first overtone of CH bending are seen near 1450, 1380, and 1360 cm<sup>−</sup><sup>1</sup> [64]. The bands near 1180, 1130, 1080, 1040 and 870 cm<sup>−</sup><sup>1</sup> are attributed to the asymmetric COC stretching and the asymmetric CH3 rocking, the CH3 symmetric rocking, the COC symmetric stretching, the C-CH3 stretching, and the C-COO stretching, respectively [64]. The band near 1270 cm<sup>−</sup><sup>1</sup> is attributed to the CH bending and COC stretching. The spectra of the PLLA-based nonwovens shown in Figure 5 were nearly identical except for the 1266–1267 cm<sup>−</sup><sup>1</sup> band. The peaks were significantly higher for the modified PLLA nonwovens than for the neat PLLA nonwoven, as shown in the inset in Figure 5. This band is sensitive to the presence of gauche–gauche (gg) conformers in the PLLA chain, which are less energy-favorable than the gauche– trans (gt) conformers [65]. It was recently demonstrated that gg conformers can originate from the stretching of the PLLA macromolecule [66]. Thus, the increased intensity of 1266–1267 cm<sup>−</sup><sup>1</sup> band may evidence enhanced stretching and orientation of PLLA chains in the modified PLLA fibers, which contributed to the increased tensile strength of the modified nonwovens.
