*3.2. Characterization of Fish Gelatin Electrospun Mats*

The OM and SEM images of the electrospun mats obtained from FG and FG+NaOH solutions are reported in Figure 2. Interestingly, the OM images (Figure 2a,d) show that bead-free and regular microfibers were obtained for both FG and FG+NaOH solutions. In contrast to previous findings [31], our results demonstrate that electrospinning of FG from a citric acid/water solution without the addition of acetic acid to form microfibers was possible. However, although both the solutions turned out to be electrospinnable, the fibrous morphology of mats obtained from FG solution was not preserved over time, since the mats turned into films, with a barely recognizable fibrous structure, in few hours, as demonstrated by the SEM micrographs (Figure 2b).

**Figure 2.** OM (**a**,**d**) and SEM images (**b**,**c**,**e**,**f**) of electrospun mats from FG (**a**,**b**,**c**) and FG+NaOH (**d**,**e**,**f**) solutions: mats as spun (**a**,**b**,**d**,**e**) and after the thermal treatment (**c**,**f**).

Addition of NaOH to FG solutions increased the pH values from 1.8 to 3.7 and improved the fiber stability over time. The obtained fibrous mat was able to preserve its morphology better than fibers obtained from FG solution (compare Figure 2b,e), although fiber fusion at their contact points could be noticed.

In agreement with previous literature findings [21,30], the thermal treatment performed on these mats immediately after their fabrication came out to get a beneficial effect on the resulting morphology, since the fibrous morphology of the mats is better maintained with respect to the as spun mats (Figure 2c,f). It is pointed out that, even if the fiber diameter was hardly measurable due to the many fusion points among fibers, after the thermal treatment a fiber diameter of 2.19 ± 0.07 μm and 4.42 ± 0.05 μm was evaluated for FG mats and FG+NaOH mats, respectively.

To assess the extent of the crosslinking reaction between gelatin and citric acid, the amount of ε-amino groups of gelatin reacted with citric acid was calculated. In line with the morphological results, only a small amount of ε-amino groups was crosslinked with citric acid in the mats obtained from FG solution, leading to a crosslinking extent of 12%. On increasing the pH up to 3.7 in the FG+NaOH solution, crosslinking extents of 25% and 38% were achieved for FG+NaOH mat as spun and thermally treated, respectively. These results are in agreement with previous findings [35], and highlight that the

crosslinking reactions can take place also at room temperature in the water solution of FG with citric acid, even if an increase of temperature up to 80 ◦C is needed to speed up such crosslinking reaction.

On the basis of the above-described results, the following considerations can be drawn. For the mat obtained from FG solution, the obtained morphology is well explained by the low crosslinking extent of the fibers and it is mainly ascribed to the low pH of the solution, since at pH 1.8 the crosslinking reactions can difficultly take place. Moreover, the widely reported mechanism driving the crosslinking of proteins or molecules containing amino groups in the presence of citric acid lies in the formation of reactive citric anhydride from citric acid and in the nucleophilic substitution occurring between the carboxyl groups of the anhydride and the amino groups of the considered protein or molecule [21,34,35]. Since type-A fish gelatin is employed in this work and its isoelectric point is in the pH range of 6.0–9.5 [36], the protonation of the amine groups takes place in the strong acidic conditions of FG solutions (pH 1.8), thus limiting the crosslinking reaction. An increase of pH up to 3.7, even if lower than fish gelatin isoelectric point, would favor the deprotonation of –NH3 <sup>+</sup> groups in –NH2, and the above described nucleophilic substitution was more likely to occur with the formation of amide groups. The pH value of 3.7 was, thus, selected in order to achieve the best compromise between the number of amine groups available for crosslinking and the known crosslinking mechanism of citric acid, which has been reported to occur at the highest extent at pH 3.5 [21].

The influence of the electrospinning process on structural properties of gelatin was investigated on fish gelatin powder and on the obtained mats through wide-angle WAXD analysis. In that way, the relative triple-helix content of fish gelatin materials was analyzed in detail (Figure 3). It is well known that the collagen WAXD pattern includes two broad diffraction bands. The first one, centered at about 8◦, related to the triple helix diameter, while the second one at around 21◦ is related to the distance between amino acidic residues in the helix. These reflections are typically observed also in the pattern of partially renaturated gelatin powder and gelatin films [37].

**Figure 3.** WAXD patterns of FG powder, FG, FG+NaOH, and thermal treated FG+NaOH electrospun mats.

In agreement with these data, results reported in Figure 3 show that fish gelatin powder exhibits the two reflections centered at about 8◦ and 21◦, as previously reported [38,39]. However, reflection at 8◦ disappears after the electrospinning process of both FG and FG+NaOH solutions, as a consequence of acidic pH. This result can be explained considering that, as previously observed for gelatin solubilized in acetic acid solutions [8], citric acid prevents the gelatin's partial renaturation which takes place during gelling from aqueous solution and a random coil conformation is favored, decreasing the number of single left-hand helix chains and residual triple-helix conformations. Furthermore, citric acid also influences the diffraction reflection located at about 21◦, whose intensity decreases as a consequence of the decrease of the single left-hand helix chain content. Addition of NaOH to FG solution, with a consequent increase of pH from 1.8 to 3.7, does not change the diffraction pattern significantly, whereas the thermal treatment performed on FG+NaOH mat further decreases the broad band at about 21◦.

Figure 4 shows the ATR-FTIR spectra of the electrospun mats obtained from FG and FG+NaOH solutions before and after the thermal treatment, FG powder is also reported for the sake of comparison. The broad band above 3000 cm−1, observed in all spectra shown in Figure 4a, corresponds to the hydroxyl and amino groups [40]. Some changes can be observed in FTIR spectra, in particular, in those bands associated to the peptide bonds in gelatin: amide I (C=O stretching), amide II (N–H bending), and amide III (C–N stretching). As can be seen in Figure 4b), there was a shift of these bands to higher wavenumbers from FG mat to FG+NaOH mats, indicating that the citric acid incorporated into the formulation causes new interactions between the amino groups of gelatin and the carboxyl groups of citric acid [27,41], in accordance with the shift of the characteristic band related to the carboxyl group in citric acid from 1748 cm−<sup>1</sup> to 1715 cm−<sup>1</sup> [42].

**Figure 4.** ATR-FTIR spectra of FG powder, FG, FG+NaOH, and thermally treated FG+NaOH electrospun mats (**a**) from 4000 to 800 cm−<sup>1</sup> and (**b**) from 1850 to 1000 cm<sup>−</sup>1.

The band corresponding to amide I is related to the secondary structure of the protein backbone and it is generally used for the quantitative analysis of the secondary structures. Hydrogen bonding plays a significant role in stabilization of protein secondary structure. Indeed, inter-peptide hydrogen bonding stabilizes secondary structures (i.e., α-helix and β-sheet conformations), while peptide-water hydrogen bonding competes against peptide bond-peptide bond hydrogen bonding. Due to the central role of hydrogen bonding in protein folding, the analysis of this band is of great importance (Figure 5).

The assignment of absorption peaks in amide I band is as follows: two peaks from 1603 to 1616 cm−<sup>1</sup> support β-sheet conformation, the peak at 1634 cm−<sup>1</sup> corresponds to random coil conformation, the peak from 1641 to 1650 cm−<sup>1</sup> is associated to α-helix conformation, and the peak centered at 1670–1678 cm−<sup>1</sup> is assigned to the β-turn conformation of the hairpin-folded antiparallel β-sheet structure (Table 1).

As shown in Figure 5 and Table 1, electrospinning causes changes in the secondary structure of gelatin. While FG powder shows α-helix conformation, a random coil is the predominant conformation for electrospun mats, in accordance with the results found by XRD, which show amorphous structure with no remaining triple helix structure.

**Figure 5.** Curve fitting spectra of amide I band (black curve) for (**a**) FG powder, (**b**) FG, (**c**) FG+NaOH, and (**d**) thermal treated FG+NaOH electrospun mats (β-sheet conformation, red curve; random coil conformation, brown curve; α-helix conformation, blue curve; β-turn conformation, green curve).

**Table 1.** Resulting percentage of the curve fitting of amide I for FG powder, FG, FG+NaOH, and thermal treated FG+NaOH electrospun mats.


#### **4. Conclusions**

An environmentally friendly chemical strategy to successfully electrospin fish gelatin from a solution containing only citric acid in an aqueous solution was demonstrated. Citric acid was used as a benign acid to solubilize gelatin and allow the electrospinning process and, at the same time, as a crosslinking agent. The pH of the spinning solutions turned out to have a strong influence on the viscoelastic behavior of the solutions, as well as on the crosslinking extent that, in turn, influenced the fiber morphology stability. Rheological measurements provided evidence that in all solutions the solid-like behavior dominated over time, but the weak gel formed did not hinder electrospinnability of the solution. The solution at pH 3.7, showing higher values of G and G" moduli, was characterized by a higher extent of crosslinking with respect to the solution at pH 1.8. Although microfibers were obtained for both FG and FG+NaOH solutions, the increase of solution pH from 1.8 to 3.7 was necessary to maintain the fibrous morphology also in the mat. A subsequent thermal treatment at 80 ◦C of the electrospun mat turned out to significantly increase the morphological stability of the mat. The crosslinking degrees of the mats were in line with the morphological results (FG mat = 12%, FG+NaOH mat = 25%, thermally treated FG+NaOH mat = 38%). Gelatin denaturation after the electrospinning process was demonstrated by the absence of the diffraction band related to the triple helix diameter, as expected since gelatin solubilization in acidic solvents is known to prevent the partial renaturation of gelatin. ATR-FTIR characterization confirmed this result and demonstrated that the gelatin structure changed from α-helix to random coil conformation as a consequence of the electrospinning process. Although further studies are necessary to find the best solution composition to further optimize the final morphology as well as the crosslinking extent of the produced mats, this work is the first successful attempt to produce crosslinked electrospun fish gelatin fibers through the use of a citric acid/water solution.

**Author Contributions:** All authors contributed to the conceptualization and the methodology of this article; investigation, J.U., A.L., S.P.; data curation, J.U., A.L., S.P., M.L.F., P.G., K.d.l.C.; writing—original draft preparation, J.U., A.L.; writing—review and editing, M.L.F., S.P., K.d.l.C.; supervision, M.L.F., S.P., P.G., K.d.l.C.

**Funding:** This research was funded by the Spanish Ministry of Science, Innovation and Universities (RTI2018-097100-B-C22), the Basque Government (Department of Quality and Food Industry), the Provincial Council of Gipuzkoa (Department of Economic Development, the Rural Environment and Territorial Balance) and the Italian Ministry of University and Research (MIUR).

**Acknowledgments:** Jone Uranga thanks the Basque Government (PRE\_2015\_1\_0205) for her fellowship and for the mobility grant (EP\_2018\_1\_0050).

**Conflicts of Interest:** The authors declare no conflict of interest.
