Preparation of Novel Electrospun Organic–Inorganic Hybrid Nanofibers Based on Polycaprolactone, Glycine Betaine, and Calcium Carbonate (P/G/CaCO 3 ‑EPnF)

: Biomedicine requires materials able to respond to specific needs without affecting the or‑ ganism. Organic–inorganic fibrillar polymeric matrices possess unique properties that may fulfill these needs. In the present study, different topology‑controlled poly( ε ‑caprolactone)‑based fibrillar matrices containing glycine betaine at varying concentrations (0.5, 1, and 2% w/v ) were prepared via electrospinning. The matrices were used as substrates in calcium carbonate crystallization as‑ says with gas diffusion to obtain a single organic–inorganic hybrid material. The resulting matrices and crystalline material were characterized using spectroscopic, microscopic, and thermogravimet‑ ric analyses. The incorporation of glycine betaine into a poly( ε ‑caprolactone) mesh modified the diameter of the fibers, without affecting the thermal behavior of the matrices. However, the chemi‑ cal and morphological characteristics of the matrices did influence in vitro inorganic mineralization. The thermogravimetric analysis of the matrices, performed after the mineralization tests, demon‑ strated the existence of a new organic–inorganic hybrid material with unique properties, which is discussed in the present study.


Introduction
Calcium carbonate (CaCO 3 ) is the most abundant calcium salt on Earth and can be found in hydrated and anhydrous forms.Three known polymorphs exist for the anhydrous form: vaterite, aragonite, and calcite [1].These polymorphs are present in marine sediments and rocks, in addition to forming part of living organisms (e.g., the exoskeleton of crustaceans, shells, and corals) through a natural process called biomineralization [2,3].CaCO 3 has been used in various fields of application.Its medical applications are of special interest, where its use has been described in bone implant and tissue engineering, drug delivery, etc., due to its biocompatibility and non-toxicity [1,4,5].
Organic-inorganic hybrid materials are a promising candidates for use in biomedicine [4].In biomineralized hybrid materials, both the organic and inorganic components play an important role.The lesser organic component drives the mineralization process, and the greater inorganic component underlies the formation of the crystal, which closely interacts with the organic matrix.This matrix can, in turn, stabilize metastable structures, which occurs in corals that stabilize aragonite [6].This is possible due to the presence of functional groups in the organic matrix produced by specialized cells that recognize and stabilize specific crystalline phases and that act as nucleation centers [3,7,8].Electrospinning (ES) is a technique widely used in different fields, and it allows for the fabrication of fibrillar meshes composed of the nano/microfibers of polymeric materials with a high surface area [9][10][11].Electrospinning consists of applying a high voltage to a polymer solution coming out of a metal needle; the drop of solution deforms and acquires a conical shape called a "Taylor cone".When the applied potential exceeds the surface tension, fibers are formed and ejected to a collector; at the same time, the solvent evaporates, and a nano/microfiber mesh is formed [10,12,13].An important aspect of ES is that it allows for the orientation of the functional groups of the polymeric material to be used during the fiber-manufacturing process.If a positive voltage is applied, the interface between the air and the polymer jet will be positively charged, which triggers the dipoles present in the solution to be attracted to the surface [10,14,15].The final traits of the polymetric fibers depend on the environmental conditions (temperature and humidity), the characteristics of the material (viscosity, conductivity, molecular weight, etc.), and the parameters used (pressure, voltage, and needle-collector distance) [16].
In biomedicine, materials that can be applied in specific ways are needed, and a key challenge in responding to this need is the development of new hybrid nanostructure materials [17].One interesting proposal is the preparation of organic-inorganic hybrid materials.Such preparations result in hybrid electrospun materials that present properties different to those of the originating materials.The optical, mechanical, and thermal properties, among others, of fibrillar polymeric meshes can be improved via the incorporation of inorganic materials [4,18,19].
The present study used a previously electrospun poly (ε-caprolactone) (P), a synthetic, biodegradable, and biocompatible aliphatic polyester with excellent mechanical and biomedical properties [3,20,21].To propose a novel hybrid material, this study incorporated glycine betaine (N,N,N-trimethylglycine betaine) (G), a zwitterion osmolyte found in animal organisms, into the fibrillar polymeric matrices.Glycine betaine possesses a carboxylate group that can be negatively charged and a positively charged tetramethylammonium group [22,23].Glycine betaine has been used to improve the characteristics of polymeric membranes by increasing wettability and the antifouling property, among others [24], demonstrating its ability to decrease the adverse effects of medical devices by way of improving hemocompatibility when incorporated into polymeric matrices [22].Additionally, G has nutraceutical characteristics and bioactive properties [25], and it has even been shown to improve the hemocompatibility of electrospun films [22].
Although each component of the fibrillar organic matrix influences the biomineralization process of CaCO 3 via the presence of functional groups [3,26], up to now, only the fabrication of P matrices covered with vaterite for biomedical use has been reported [27].However, this study is the first report on (i) the preparation of a novel electrospun polymer nanofiber (EPnF) mesh based on P and G (P/G-EPnF) and (ii) the use thereof as a template for the in vitro crystallization of CaCO 3 through gas diffusion (GD) as a potential organic-inorganic hybrid nanofiber (P/G/CaCO 3 -EPnF) material for biomedical applications by generating a new hybrid material created from three components widely described and used in the biomedical area.

Preparation of P/G-EPnF
Poly (ε-caprolactone) 16% (w/v) was prepared in acetic acid/ethyl acetate/ultra-pure water 3:2:1 (v/v) at 40 • C under magnetic stirring for 6 h.To construct the P/G-EPnF meshes, 40% (w/v) G was dissolved in water under magnetic stirring for 10 min at 75 • C.This solution was added to a 3:2:1 (v/v/v) acetic acid/ethyl acetate/Milli-Q water mixture at concentrations of 0.5%, 1%, and 2 % (w/v), with the water replaced as needed to obtain each concentration.Subsequently, the P solution was added, and the resulting solution was kept at 40 • C under magnetic stirring for 6 h.After preparation, the solutions were deposited in 5 mL Luer-lock syringes (Nipro ® , Osaka, Japan).The ES process was carried out in an ES LE-10 instrument (Fluidnatek ® , Singapore).The proper working parameters for preparing the fibers are summarized in Table 1.
Table 1.ES parameters for the preparation of A-P/G-EPnF and R-P/G-EPnF meshes.The abbreviation of RPM corresponds to revolutions per minute.
The topographic aspects of the P/G-EPnF meshes used as scaffolds in the CaCO 3 mineralization study were checked.For this, fiber meshes with random (R-P/G-EPnF) or aligned (A-P/G-EPnF) distributions were obtained using a flat or a rotary (10 cm diameter) collector, respectively (Figure 1).Both collectors were covered with aluminum foil.
Crystals 2023, 13, x FOR PEER REVIEW 3 of 20 consumed in this research was Milli-Q grade water obtained using a BIOBASE (SCSJ-30V) instrument.

Preparation of P/G-EPnF
Poly (ε-caprolactone) 16% (w/v) was prepared in acetic acid/ethyl acetate/ultra-pure water 3:2:1 (v/v) at 40 °C under magnetic stirring for 6 h.To construct the P/G-EPnF meshes, 40% (w/v) G was dissolved in water under magnetic stirring for 10 min at 75 °C.This solution was added to a 3:2:1 (v/v/v) acetic acid/ethyl acetate/Milli-Q water mixture at concentrations of 0.5%, 1%, and 2 % (w/v), with the water replaced as needed to obtain each concentration.Subsequently, the P solution was added, and the resulting solution was kept at 40 °C under magnetic stirring for 6 h.After preparation, the solutions were deposited in 5 mL Luer-lock syringes (Nipro ® ).The ES process was carried out in an ES LE-10 instrument (Fluidnatek ® ).The proper working parameters for preparing the fibers are summarized in Table 1.
The topographic aspects of the P/G-EPnF meshes used as scaffolds in the CaCO3 mineralization study were checked.For this, fiber meshes with random (R-P/G-EPnF) or aligned (A-P/G-EPnF) distributions were obtained using a flat or a rotary (10 cm diameter) collector, respectively (Figure 1).Both collectors were covered with aluminum foil.The abbreviation of RPM corresponds to revolutions per minute.

Crystallization of CaCO 3
The GD technique was used to prepare CaCO 3 crystals in vitro on P/G-EPnF meshes.The GD method was performed as we described in our previous works [28,29].Briefly, a 200 mM CaCl 2 •2H 2 O solution was prepared in 200 mM TRIS at pH 9.These solutions were deposited in 1 mL crystallization vials containing cut P/G-EPnF meshes (8 × 8 mm).The vials were placed in a crystallization cell containing a 25 mM (NH 4 )HCO 3 solution used to create a CO 2 source (Figure 2).All GD assays were performed in an oven (LabTech ® , Sorisole ，Italy) at 25 • C for 24 h.After the crystallization time, the resulting meshes were washed with ethanol at increasing concentrations (50-100%), thereby obtaining hybrid matrices of P/G-EPnF with CaCO 3 crystals (P/G/CaCO 3 -EPnF).

Crystallization of CaCO3
The GD technique was used to prepare CaCO3 crystals in vitro on P/G-EPnF meshes.The GD method was performed as we described in our previous works [28,29].Briefly, a 200 mM CaCl2 • 2H2O solution was prepared in 200 mM TRIS at pH 9.These solutions were deposited in 1 mL crystallization vials containing cut P/G-EPnF meshes (8 × 8 mm).The vials were placed in a crystallization cell containing a 25 mM (NH4)HCO3 solution used to create a CO2 source (Figure 2).All GD assays were performed in an oven (Lab-Tech ® ) at 25 °C for 24 h.After the crystallization time, the resulting meshes were washed with ethanol at increasing concentrations (50-100%), thereby obtaining hybrid matrices of P/G-EPnF with CaCO3 crystals (P/G/CaCO3-EPnF).The G and P/G-EpnF meshes were evaluated with a Nicolet IS 50 instrument (Thermo Fisher Scientific, Waltham, MA, USA).Spectra were obtained in the range of 4000 to 500 cm −1 with a resolution of 4 cm −1 , and the final spectrum corresponds to an average of 16 scans.

Contact Angle (CA)
The CA of the P/G-EPnF meshes was measured in a Dataphysics OCA 15EC machine (DataPhysics Instruments GmbH).A dosing volume of 12 µ L ultrapure water was used.For this, 1 × 10 cm P/G-EPnF meshes were cut and glued to slides.Five replicates were performed for each assay.For aligned meshes, CA measurements were performed in both directions, i.e., frontal to the fibers and lateral to the fibers.

Scanning Electron Microscopy (SEM)
P/G-EPnF and P/G/CaCO3-EPnF were coated with a 10 nm layer of gold and analyzed using SEM with an Interspect-F50 instrument operating at 20 kV.The SEM images were analyzed using ImageJ software (NIH, http://rsb.info.nih.gov/ij/,accessed on 12 October 2022).A total of 3 representative images of each sample were used, and 100 fibers were measured (n = 300).

X-ray Diffraction (XRD)
The GB, P, CaCO3, P/G0-EPnF, and P/G/CaCO3-EPnF meshes were evaluated in a Siemens D-5000 machine with CuKα radiation.The Diffrac.Eva program (Bruker ® ) was used to compare the different diffractograms and to determine the polymorphs of the CaCO3 crystalline material.The crystallography open database (COD) was used to index  The G and P/G-EpnF meshes were evaluated with a Nicolet IS 50 instrument (Thermo Fisher Scientific, Waltham, MA, USA).Spectra were obtained in the range of 4000 to 500 cm −1 with a resolution of 4 cm −1 , and the final spectrum corresponds to an average of 16 scans.

Contact Angle (CA)
The CA of the P/G-EPnF meshes was measured in a Dataphysics OCA 15EC machine (DataPhysics Instruments GmbH).A dosing volume of 12 µL ultrapure water was used.For this, 1 × 10 cm P/G-EPnF meshes were cut and glued to slides.Five replicates were performed for each assay.For aligned meshes, CA measurements were performed in both directions, i.e., frontal to the fibers and lateral to the fibers.

Scanning Electron Microscopy (SEM)
P/G-EPnF and P/G/CaCO 3 -EPnF were coated with a 10 nm layer of gold and analyzed using SEM with an Interspect-F50 instrument operating at 20 kV.The SEM images were analyzed using ImageJ software (NIH, http://rsb.info.nih.gov/ij/,accessed on 12 October 2022).A total of 3 representative images of each sample were used, and 100 fibers were measured (n = 300).

X-ray Diffraction (XRD)
The GB, P, CaCO 3 , P/G0-EPnF, and P/G/CaCO 3 -EPnF meshes were evaluated in a Siemens D-5000 machine with CuKα radiation.The Diffrac.Eva program (Bruker ® , Billerica, MA, USA) was used to compare the different diffractograms and to determine the polymorphs of the CaCO 3 crystalline material.The crystallography open database (COD) was used to index the peaks.The crystal structure was analyzed using the Rietveld re-finement method, by using half-width at half-height (FWHM), calculated with Scherrer's formula with TOPAS 4.2 software, Bruker AXS.2.4.5.Thermogravimetric Analysis and Derivative Thermogravimetry (TGA/DTG) P/G-EPnF and P/G/CaCO 3 -EPnF were evaluated using TGA/DTG with 1.5-8.5 mg of the sample, which was performed with a constant nitrogen flow rate of 30 mL/min and a constant 10 • C/min temperature increase rate from room temperature to 600 • C. The TGA/DTG measurements were performed on a NETZSCH TG 209 F3 Tarsus ® instrument (Selb, Germany).

Statistical Analysis
The normality of the fiber diameter was evaluated using the Shapiro-Wilk test, and the homogeneity of variance was evaluated using Fligner-Killeen.According to the results, a parametric test using one-way ANOVA and Tukey's multiple comparisons tests or a nonparametric test using the Kruskal-Wallis test was performed, followed by Dunn's test.A p-value of ≤0.05 was considered statistically significant.

ATR/FT-IR Spectroscopy
Infrared spectroscopy allows for the identification of functional groups in samples.Identified for G was a sharp absorption band at 3290 cm −1 , associated with the -OH group [25], and the absorption bands at 890 cm −1 and 1630 cm −1 are associated with the v C-C and carboxylic groups, respectively (Figure 3) [23,30].The typical bands of P were also identified.The stretching of the C=O carbonyl group was recorded at 1720 cm −1 , while the absorption bands of C-H stretching were found at 2945 cm −1 and 2860 cm the peaks.The crystal structure was analyzed using the Rietveld refinement method, by using half-width at half-height (FWHM), calculated with Scherrer's formula with TOPAS 4.2 software, Bruker AXS.
2.4.5.Thermogravimetric Analysis and Derivative Thermogravimetry (TGA/DTG) P/G-EPnF and P/G/CaCO3-EPnF were evaluated using TGA/DTG with 1.5-8.5 mg of the sample, which was performed with a constant nitrogen flow rate of 30 mL/min and a constant 10 °C /min temperature increase rate from room temperature to 600 °C.The TGA/DTG measurements were performed on a NETZSCH TG 209 F3 Tarsus ® instrument.

Statistical Analysis
The normality of the fiber diameter was evaluated using the Shapiro-Wilk test, and the homogeneity of variance was evaluated using Fligner-Killeen.According to the results, a parametric test using one-way ANOVA and Tukey's multiple comparisons tests or a non-parametric test using the Kruskal-Wallis test was performed, followed by Dunn's test.A p-value of ≤ 0.05 was considered statistically significant.

ATR/FT-IR Spectroscopy
Infrared spectroscopy allows for the identification of functional groups in samples.Identified for G was a sharp absorption band at 3290 cm −1 , associated with the -OH group [25], and the absorption bands at 890 cm −1 and 1630 cm −1 are associated with the v C-C and carboxylic groups, respectively (Figure 3) [23,30].The typical bands of P were also identified.The stretching of the C=O carbonyl group was recorded at 1720 cm −1 , while the absorption bands of C-H stretching were found at 2945 cm −1 and 2860 cm −1 , respectively corresponding to the stretching signals of the asymmetric and symmetric CH2 methylene groups.Further observed were absorption bands corresponding to the asymmetric (1236 and 1295 cm −1 ) and symmetric (1172 and 1044 cm −1 ) stretching of the C-O-C and C-O bonds, respectively (Figure 3) [25,31].The characteristic absorption bands of P were identified in P/G-EPnF.Moreover, the absorption bands of G in the P/G1-EPnF and P/G2-EPnF samples are indicated by black arrows in insets 1 and 2 in Figure 3.The small bands at 1630 cm −1 (-COO) and 890 cm −1 (C-C) demonstrate the presence of G at a very low concentration in the P/G-EPnF meshes.

CA Measurements
P is a hydrophobic material [16], as was confirmed by the recorded CA values above 110 • for R-P/G0-EPnF and A-P/G0-EPnF (Figure 4).A similar behavior was found for R-P/G0.5-EPnF,R-P/G1-EPnF, and R-P/G2-EPnF, the CA values of which were close to 100 • (Figure 4a); however, when analyzing the results, it was possible to observe that there were statistically significant differences between the R-P/G1-EPnF and R-P/G2-EPnF groups.For the aligned fibers, CA was measured frontally and laterally (Figure 4b) since a spherical drop cannot be obtained on anisotropic structures [32].The results (Figure 4c,d) agree with those in previous reports [33], specifically in relation to how the fiber orientation and the measurement method are determinants of CA.In aligned fibers, a similar behavior was seen, where G-concentration-dependence was observed (Figure 4c,d).The increase in the degree of wetting in the meshes containing higher concentrations of G in the P/G-EPnF systems is explained by the ability of G to increase hydrophilicity through its ability to retain water [24].In the case of aligned fibers, it was also possible to observe an effect attributed to the topography of the substrate.In A-P/G2-EPnF, CA values 20 • lower than those in the frontal and lateral measurements of the fibers in the other meshes were observed (Figure 4c,d).However, a higher CA value in A-P/G1-EPnF regarding A-P/G0.5-EPnF in the lateral analysis of the fibers could also be observed (Figure 4d).The present results indicate that CA is also influenced by fiber orientation.This occurs since the wetting degree of a material depends on the surface free energy, which is sensitive to the chemical traits of the material.Therefore, the 20 • variation in CA for A-P/G2-EPnF indicates that fiber orientation modulates the surface free energy [11].These results confirm that CA is dependent on both the fiber orientation and the chemical characteristics of the material.

CA Measurements
P is a hydrophobic material [16], as was confirmed by the recorded CA values above 110° for R-P/G0-EPnF and A-P/G0-EPnF (Figure 4).A similar behavior was found for R-P/G0.5-EPnF,R-P/G1-EPnF, and R-P/G2-EPnF, the CA values of which were close to 100° (Figure 4a); however, when analyzing the results, it was possible to observe that there were statistically significant differences between the R-P/G1-EPnF and R-P/G2-EPnF groups.For the aligned fibers, CA was measured frontally and laterally (Figure 4b) since a spherical drop cannot be obtained on anisotropic structures [32].The results (Figure 4c,  d) agree with those in previous reports [33], specifically in relation to how the fiber orientation and the measurement method are determinants of CA.In aligned fibers, a similar behavior was seen, where G-concentration-dependence was observed (Figure 4c, d).The increase in the degree of wetting in the meshes containing higher concentrations of G in the P/G-EPnF systems is explained by the ability of G to increase hydrophilicity through its ability to retain water [24].In the case of aligned fibers, it was also possible to observe an effect attributed to the topography of the substrate.In A-P/G2-EPnF, CA values 20° lower than those in the frontal and lateral measurements of the fibers in the other meshes were observed (Figure 4c, d).However, a higher CA value in A-P/G1-EPnF regarding A-P/G0.5-EPnF in the lateral analysis of the fibers could also be observed (Figure 4d).The present results indicate that CA is also influenced by fiber orientation.This occurs since the wetting degree of a material depends on the surface free energy, which is sensitive to the chemical traits of the material.Therefore, the 20° variation in CA for A-P/G2-EPnF indicates that fiber orientation modulates the surface free energy [11].These results confirm that CA is dependent on both the fiber orientation and the chemical characteristics of the material.

SEM Analysis
The SEM analysis served to study mesh morphologies.In the micrographs obtained for the different EPnF meshes, it was possible to observe continuous fibers with some defects on their surfaces and the presence of flat fibers.This is attributed to the solvent system used, in which water was incorporated.Previous works have reported the difficulty of manufacturing P 16% fibers in this solvent (acetic acid/ethyl acetate/Milli-Q water) [34].In this work, it was possible to obtain fibers with nanometer and micrometer diameters (Figures 5 and 6); a similar result has been recently reported [21].When the rotary collector was used, fibers that were not completely aligned were obtained.
These findings indicate that the final fiber diameter depends on both the type of collector used and the chemical composition of the polymer solution used in the electrospinning process.This is because the final traits of the fibers are influenced by environmental factors (temperature and humidity), characteristics of the polymer solution (concentration, polymer, loading, viscosity, conductivity, etc.), and specific parameters of the ES process (voltage, distance, and pressure) [35].However, there is a fan effect that occurs in the rotating manifold at high speed, which causes the solvent to evaporate faster [10], and when G is added, some characteristics of the polymeric solution change.Therefore, new parameters for the ES process were required, presenting a similar situation to that observed in the CA measurements.
The SEM images of the CaCO 3 crystals obtained through GD (Figure 7) showed a homogenous distribution of the crystalline material in the control tests using a glass substrate (Figure 7a).Further observed were rhombohedral structures, corresponding to smoothsurfaced calcite crystals (29.6 ± 8.1 µm), and flower-like hexagonal structures (38 ± 4.8 µm in diameter) formed by stacked flakes (≈70 nm in height) (Figure 7b-d), which correspond to the polymorph vaterite.The surface of the flower-like structures presented a nanogranular texture (38.5 ± 14 nm) (Figure 7d).The crystal boundary ended at 120 • angles, indicating that these hexagonal plates correspond to the phase (001) monocrystals of vaterite and the hexagonal boundaries parallel to {110}, as described elsewhere [36].Although the morphologies of the crystals are quite varied, characteristic morphologies were described for the CaCO 3 polymorphs (namely, rhombohedral calcite, needle-like aragonite, and spherical vaterite) [5,[36][37][38], and these allowed us to approximate the observed polymorphs using SEM, which were later confirmed using XRD (please see Table 2).
An analysis of the P/G-EPnF and P/G/CaCO 3 -EPnF meshes revealed that, independent of the fiber orientation and composition of the P/G-EPnF meshes, the crystals were not distributed homogeneously over the matrix.Instead, agglomerates of crystals in different zones were observed, and they decreased the surface energy of the crystals that were smaller due to the presence of a greater number of nucleation centers (Figure 8a) [37].Furthermore, the CaCO 3 crystals closely interacted with proximally growing fibers (Figure 8b).This occurred because there was a geometric match between the functional groups of the organic matrix and the crystal lattice of a specific phase [26,38].Further found was that both the lattices and respective compositions influenced the process.This is explained by the fact that the environment where mineralization occurs determines the morphology and size of the crystals obtained [7,8].However, the CaCO 3 crystals had defects on the crys-talline surface in the presence of meshes (Figure 8c), as opposed to what occurred in the absence of meshes (Figure 7).The SEM images of the CaCO3 crystals obtained through GD (Figure 7) sho homogenous distribution of the crystalline material in the control tests using a glas strate (Figure 7a).Further observed were rhombohedral structures, correspond smooth-surfaced calcite crystals (29.6 ± 8.1 µ m), and flower-like hexagonal structu ± 4.8 µ m in diameter) formed by stacked flakes (≈70 nm in height) (Figure 7b-d), correspond to the polymorph vaterite.The surface of the flower-like structures pre a nanogranular texture (38.5 ± 14 nm) (Figure 7d).The crystal boundary ended a angles, indicating that these hexagonal plates correspond to the phase (001) monoc of vaterite and the hexagonal boundaries parallel to {110}, as described elsewher Although the morphologies of the crystals are quite varied, characteristic morpho were described for the CaCO3 polymorphs (namely, rhombohedral calcite, need aragonite, and spherical vaterite) [5,[36][37][38], and these allowed us to approximate t served polymorphs using SEM, which were later confirmed using XRD (please see 2).The SEM images of the CaCO3 crystals obtained through GD (Figure 7) showed a homogenous distribution of the crystalline material in the control tests using a glass substrate (Figure 7a).Further observed were rhombohedral structures, corresponding to smooth-surfaced calcite crystals (29.6 ± 8.1 µ m), and flower-like hexagonal structures (38 ± 4.8 µ m in diameter) formed by stacked flakes (≈70 nm in height) (Figure 7b-d), which correspond to the polymorph vaterite.The surface of the flower-like structures presented a nanogranular texture (38.5 ± 14 nm) (Figure 7d).The crystal boundary ended at 120° angles, indicating that these hexagonal plates correspond to the phase (001) monocrystals of vaterite and the hexagonal boundaries parallel to {110}, as described elsewhere [36].Although the morphologies of the crystals are quite varied, characteristic morphologies were described for the CaCO3 polymorphs (namely, rhombohedral calcite, needle-like aragonite, and spherical vaterite) [5,[36][37][38], and these allowed us to approximate the observed polymorphs using SEM, which were later confirmed using XRD (please see Table 2).Table 2. Crystal sizes of calcite (C) and vaterite (V) obtained in GD crystallization assays on the R-P/G/CaCO 3 -EPnF and A-P/G/CaCO 3 -EPnF meshes using SEM.determine the predominance of rhombohedral crystals.Meanwhile, the flower morphology of the vaterite crystals was lost in the P/G0/CaCO3-EPnF lattices, and only a few elliptical crystals were observed.However, in P/G0.5/CaCO3-EPnF,P/G1/CaCO3-EPnF, and P/G2/CaCO3-EPnF, the coexistence of vaterite crystals with flower and elliptical morphologies was observed (Figure 8d-f).Our findings obtained by carrying out SEM and XRD analyses of the crystals show that the topography of the meshes of R-P/G-EPnF and A-P/G-EPnF did not affect the growth of the CaCO3 crystals.In mineralization, the presence of other ions or the deletion of crystal lattice ions on the crystal surface generates a heterogeneous distribution of energy on the crystal surface, thereby impacting the rate of crystal growth [2,37,39,40].This was evidenced by alterations on the crystal surface (Figure 8c) and by comparing crystal size.The CaCO3 crystals obtained on P/G0/CaCO3-EPnF were six times smaller than the crystals observed on the glass substrates, while the crystals obtained on P/G0.5/CaCO3-EPnF,P/G1/CaCO3-EPnF, In the control tests (i.e., in the absence of the P/G-EPnF meshes), vaterite and calcite crystals were obtained, as subsequently confirmed using XRD.These results indicate that P/G-EPnF meshes favor the formation of rhombohedral crystals.Although these grow in agglomerates and fuse together, thus preventing counting, it was qualitatively possible to determine the predominance of rhombohedral crystals.Meanwhile, the flower morphology of the vaterite crystals was lost in the P/G0/CaCO 3 -EPnF lattices, and only a few elliptical crystals were observed.However, in P/G0.5/CaCO 3 -EPnF, P/G1/CaCO 3 -EPnF, and P/G2/CaCO 3 -EPnF, the coexistence of vaterite crystals with flower and elliptical morphologies was observed (Figure 8d-f).Our findings obtained by carrying out SEM and XRD analyses of the crystals show that the topography of the meshes of R-P/G-EPnF and A-P/G-EPnF did not affect the growth of the CaCO 3 crystals.

A-P/G1-EPnF
In mineralization, the presence of other ions or the deletion of crystal lattice ions on the crystal surface generates a heterogeneous distribution of energy on the crystal surface, thereby impacting the rate of crystal growth [2,37,39,40].This was evidenced by alterations on the crystal surface (Figure 8c) and by comparing crystal size.The CaCO 3 crystals obtained on P/G0/CaCO 3 -EPnF were six times smaller than the crystals observed on the glass substrates, while the crystals obtained on P/G0.5/CaCO 3 -EPnF, P/G1/CaCO 3 -EPnF, and P/G2/CaCO 3 -EPnF were three to four times smaller than those obtained on the control.These did, however, increase in size with increasing G concentrations in the mesh (Table 2).Previous experiments on the CaCO 3 crystals in electrospun P meshes show that fiber distribution influences nucleation and the calcite-to-vaterite ratio [28].
The presence of new functional groups on the substrate favors the formation of nucleation centers [3].This situation is similar to that which occurs in nanocellulose, where it acts as a scaffold for CaCO 3 growth [8], as determined by the presence of hydroxyl groups of nanocellulose that interact with Ca 2+ ions causing modifications in the speed and direction of growth, ultimately altering the final morphology of calcite crystals.These authors indicate that nanocellulose, due to electrostatic interactions and the steric effect, prevents the sustained growth of the crystals, generating a decrease in the crystalline size because the nanocellulose interacts with specific crystalline phases, thereby preventing or delaying the vertical growth of that phase, which consequently remains stable.Therefore, in this work, it is proposed that Ca 2+ is attracted by the carboxylate group of the G molecule, favoring the presence of new nucleation centers (Figure 9).delaying the vertical growth of that phase, which consequently remains stable.Therefore, in this work, it is proposed that Ca 2+ is attracted by the carboxylate group of the G molecule, favoring the presence of new nucleation centers (Figure 9).
However, it has been described that the TRIS buffer used in CaCO3 crystallization assays could influence crystalline growth by forming soluble complexes with Ca 2+ ions, decreasing their availability in the medium [41][42][43].Therefore, the presence of TRIS is essential to maintain the pH of a crystallization solution at pH 9. It is important to note that the concentration of TRIS was not varied in the different crystallization tests, which is why it does not represent a new variable in our results.Moreover, its concentration can be considered negligible in this assay.Table 2. Crystal sizes of calcite (C) and vaterite (V) obtained in GD crystallization assays on the R-P/G/CaCO3-EPnF and A-P/G/CaCO3-EPnF meshes using SEM.Crystal sizes are presented in micrometer (µm).Results are presented as mean ± standard deviation.However, it has been described that the TRIS buffer used in CaCO 3 crystallization assays could influence crystalline growth by forming soluble complexes with Ca 2+ ions, decreasing their availability in the medium [41][42][43].Therefore, the presence of TRIS is essential to maintain the pH of a crystallization solution at pH 9. It is important to note that the concentration of TRIS was not varied in the different crystallization tests, which is why it does not represent a new variable in our results.Moreover, its concentration can be considered negligible in this assay.
An analysis of the XRD diffractograms for the R-P/G-EPnF (Figure 11a-d) and A-P/G-EPnF (Figure 11e-h) meshes used as substrates in the GD crystallization tests revealed the typical crystalline peaks associated with P in all samples and the presence of calcite.Although hexagonal crystals of vaterite were observed, its abundance was lower than that of calcite, thus explaining the absence of crystallographic peaks for this or other polymorphs that could be present in the meshes.This was confirmed by a Rietveld analysis, where it was shown that the composition of the samples was 100% calcite, with the exception of A-P/G0.5-EPnF,which had a composition of 97.4% calcite and only 2.6% vaterite.Therefore, our experimental findings demonstrate the stabilization almost exclusively of calcite in the P/G-EPnF electrospun meshes, independent of their composition.However, the crystallite sizes of calcite and vaterite are modified depending on the substrate where the crystallization takes place.This behavior is not associated with the topography of the substrate, but there is a relationship in terms of the concentration of G in the meshes.In Table 3, we can see that, in both R-P/G-EPnF and A-P/G-EPnF, there is a decrease in crystallite size at concentrations of 0.5 and 2% G.
responded to calcite, with crystalline peaks at 2 29.4° (014), 39.4° (113), 43.1° (022), 47.4° (018), and 48.5° (116) (COD #2100992), and to vaterite, with 2 crystalline peaks at 27° (12-1) and 32.7° (202) (COD #9014701) (Figure 10a).Most of the crystals formed were calcite, which represents 67.1% of the sample, while vaterite represents 32.9% of the sample.In turn, numerous peaks were observed for the XDR of G, the most intense of which occurred at 2 angles of 15.3°, 22.5°, and 25.18° (PDF #13-0815) (Figure 10b).For the P/G0-EPnF mesh, the XRD of R-P/G0-EPnF identified the presence of most characteristic crystalline peaks at 2 21.3° and 23.6°, which correspond to the crystallographic planes (110) and (200), respectively [44] (Figure 10c).When analyzing the diffractogram of P/G2-EPnF, only the crystallographic peaks associated with P appeared (Figure 10d).This could be due to the interaction of G with P during the electrospinning process, which could cause the G crystalline structure loss [45].The random and aligned fibers of P/G-EPnF had the same XRD pattern, so the diffractograms of R-P/G0-EPnF and R-P/G2-EPnF are reported in Figure 10.An analysis of the XRD diffractograms for the R-P/G-EPnF (Figure 11a-d) and A-P/G-EPnF (Figure 11e-h) meshes used as substrates in the GD crystallization tests revealed the typical crystalline peaks associated with P in all samples and the presence of calcite.Although hexagonal crystals of vaterite were observed, its abundance was lower than that of calcite, thus explaining the absence of crystallographic peaks for this or other polymorphs that could be present in the meshes.This was confirmed by a Rietveld analysis, where it was shown that the composition of the samples was 100% calcite, with the exception of A-P/G0.5-EPnF,which had a composition of 97.4% calcite and only 2.6% vaterite.Therefore, our experimental findings demonstrate the stabilization almost exclusively of calcite in the P/G-EPnF electrospun meshes, independent of their composition.However, the crystallite sizes of calcite and vaterite are modified depending on the substrate where the crystallization takes place.This behavior is not associated with the topography of the substrate, but there is a relationship in terms of the concentration of G in the meshes.In Table 3, we can see that, in both R-P/G-EPnF and A-P/G-EPnF, there is a decrease in crystallite size at concentrations of 0.5 and 2% G. Additionally, various crystallographic peaks associated with calcite were detected in both P/G-EPnF systems (Figure 11 c-h); however, the 2 crystallographic peaks that appeared at 43.1° and 47.5° were not observed in R-P/G0-EPn or R-P/G0.5-EPnF(Figure 11a,  b).
This could be explained by the interaction of the organic matrix with certain crystalline phases preventing or delaying growth, as mentioned above.Nevertheless, this does not mean that the interaction prevents the growth of other crystalline phases associated with the same polymorph [8].The Rietveld refinement method was carried out using the Scherer method, through which it was possible to determine the crystallite sizes of the crystals obtained in the samples (Table 3), where the sizes of the crystallites for calcite (C) had a similar behavior with both mesh topographies.The Rietveld analysis showed the largest crystallite sizes for the control and both fibrillar systems R-P/G and A-P/G in the absence of G. Thus, this analysis Table 3. Crystallite sizes of calcite (C) and vaterite (V) obtained in GD crystallization assays on the R-P/G/CaCO 3 -EPnF and A-P/G/CaCO 3 -EPnF meshes determined using XRD.The length scales of the crystallite size are presented in nanometers (nm).
This could be explained by the interaction of the organic matrix with certain crystalline phases preventing or delaying growth, as mentioned above.Nevertheless, this does not mean that the interaction prevents the growth of other crystalline phases associated with the same polymorph [8].
The Rietveld refinement method was carried out using the Scherer method, through which it was possible to determine the crystallite sizes of the crystals obtained in the samples (Table 3), where the sizes of the crystallites for calcite (C) had a similar behavior with both mesh topographies.The Rietveld analysis showed the largest crystallite sizes for the control and both fibrillar systems R-P/G and A-P/G in the absence of G. Thus, this analysis showed that the concentration of G1 in both fibrillar mesh systems (P/G1-EPnF) presented the largest crystallite sizes, and the crystallite size of vaterite (V) in the A-P/G-EPnF system was 89 nm, identical to that in the control assay.However, the size of the crystallites did not correlate with the increase in the diameter of the fibers observed using SEM when increasing the concentration of G in both fibrillar systems.The crystalline cell parameters of each assay are reported in Table 4.

TGA/DTG Analysis
TGA/DTG assess the thermal stability of materials.In the present study, no differences were observed between R-P/G0-EPnF and A-P/G0-EPnF (Table 5 and Figure 12).It has been described that the decomposition temperature of P is 400 • C [46], while that for G is close to 300 • C [25,47,48].The thermal decomposition behaviors of R-P/G0-EPnF and A-P/G0-EPnF were 408 • C and 407 • C, respectively.In the meshes containing G, two decomposition processes were observed, one near 400 • C (associated with P) and another at 250 • C (attributable to the presence of G).The results for the TGAs of R-P/G-EPnF and A-P/G-EPnF after the crystallization tests are summarized in Table 6 and Figure 13.Only one thermal decomposition process of the organic-inorganic hybrid meshes was close to 400 °C, as it was associated with the decomposition temperature of P.However, when comparing the mass loss of the meshes in the absence of crystals, those containing CaCO3 showed a lower mass loss.This could be associated with the mineral component in the mesh having a decomposition temperature higher than 700 °C [8,49].
Importantly, an initial decomposition process near 250 °C was not observed in P/G/CaCO3-EPnF but was observed in P/G-EPnF.This can be explained by the interaction between Ca 2+ and G, which sustains incorporation into the crystalline matrix.In this work, it was hypothesized that the positively charged tetramethyl-ammonium group of G interacts with the carbonyl group of P through electrostatic interactions; this exposes the negatively charged carboxylate group of the G molecule, which has the ability to attract Ca 2+ , forming a nucleation center explaining the incorporation of P/G-EPnF meshes.This allows the organic part to be incorporated into the crystalline matrix [37,50,51], which explains These results show a pattern in the thermal decomposition process, with a slight increase in the decomposition temperatures of G and P depending on the concentration of G in the meshes.In another study, when G was incorporated into poly(vinyl alcohol) meshes, as the decomposition temperature of poly(vinyl alcohol) decreased, the thermal stability of G increased [25].This prior study demonstrates an interaction between poly(vinyl alcohol) and G that modifies thermal stability and that is dependent on the G concentration.Differences in behavior as compared to the present study, where the decomposition temperature of both materials increased, could be attributable to the existing specific interaction between materials.
The results for the TGAs of R-P/G-EPnF and A-P/G-EPnF after the crystallization tests are summarized in Table 6 and Figure 13.Only one thermal decomposition process of the organic-inorganic hybrid meshes was close to 400 • C, as it was associated with the decomposition temperature of P.However, when comparing the mass loss of the meshes in the absence of crystals, those containing CaCO 3 showed a lower mass loss.This could be associated with the mineral component in the mesh having a decomposition temperature higher than 700 • C [8,49].

Conclusions
A topography-controlled organic-inorganic P/G/CaCO3-EPnF hybrid material was obtained using electrospun R-P/G/-EPnF and A-P/G/-EPnF matrices and the application of ES and the CaCO3 GD crystallization technique.The incorporation of G (0.5, 1, and 2% w/v) into these meshes influenced the fiber diameter without altering the mechanical properties.A thermal gravimetric analysis indicated the presence of a single thermal decomposition process in P/G/CaCO3-EPnF, suggesting the incorporation of G into the crys- Importantly, an initial decomposition process near 250 • C was not observed in P/G/ CaCO 3 -EPnF but was observed in P/G-EPnF.This can be explained by the interaction between Ca 2+ and G, which sustains incorporation into the crystalline matrix.In this work, it was hypothesized that the positively charged tetramethyl-ammonium group of G interacts with the carbonyl group of P through electrostatic interactions; this exposes the negatively charged carboxylate group of the G molecule, which has the ability to attract Ca 2+ , forming a nucleation center explaining the incorporation of P/G-EPnF meshes.This allows the organic part to be incorporated into the crystalline matrix [37,50,51], which explains the close relationship between the fibers and the crystals that was observed using SEM.
As has been shown elsewhere [52], similar findings using organic-inorganic materials demonstrated that CaCO 3 acts as a crosslinking agent between the proteins within the organic component, thereby improving the mechanical properties of the prepared hybrid material.Additionally, similar results were also reported when forming a unique, distinctive organic-inorganic material, with new properties given by the interaction between CaCO 3 and the functional groups of the organic matrix [8].

Conclusions
A topography-controlled organic-inorganic P/G/CaCO 3 -EPnF hybrid material was obtained using electrospun R-P/G/-EPnF and A-P/G/-EPnF matrices and the application of ES and the CaCO 3 GD crystallization technique.The incorporation of G (0.5, 1, and 2% w/v) into these meshes influenced the fiber diameter without altering the mechanical properties.A thermal gravimetric analysis indicated the presence of a single thermal decomposition process in P/G/CaCO 3 -EPnF, suggesting the incorporation of G into the crystalline matrix.Such incorporation demonstrates the formation of a new organic-inorganic hybrid material that has significant potential for use in biomedicine due to the properties and characteristics of each material of the fibrillar system.Therefore, the P/G/CaCO 3 -EPnF hybrid material could potentiate the characteristics of each of the components that compose it.

Figure 1 .
Figure 1.Schematic representation of the electrospinning setup and the final distribution of the fibers in each type of collector.(a) Flat collector and (b) rotating collector.Figure 1.Schematic representation of the electrospinning setup and the final distribution of the fibers in each type of collector.(a) Flat collector and (b) rotating collector.

Figure 1 .
Figure 1.Schematic representation of the electrospinning setup and the final distribution of the fibers in each type of collector.(a) Flat collector and (b) rotating collector.Figure 1.Schematic representation of the electrospinning setup and the final distribution of the fibers in each type of collector.(a) Flat collector and (b) rotating collector.

Figure 2 .
Figure 2. Schematic representation and photograph of the GD crystallization method.(a) Schematic of the crystallization cell with crystallization vials inside, and (b) photograph (top view) of the crystallization cell with vials containing P/G EPnF meshes.

Figure 2 .
Figure 2. Schematic representation and photograph of the GD crystallization method.(a) Schematic of the crystallization cell with crystallization vials inside, and (b) photograph (top view) of the crystallization cell with vials containing P/G EPnF meshes.

2. 4 .
Characterization Techniques 2.4.1.Attenuated Total Reflectance-Fourier Transform Infrared Spectroscopy (ATR/FT-IR) −1 , respectively corresponding to the stretching signals of the asymmetric and symmetric CH 2 methylene groups.Further observed were absorption bands corresponding to the asymmetric (1236 and 1295 cm −1 ) and symmetric (1172 and 1044 cm −1 ) stretching of the C-O-C and C-O bonds, respectively (Figure 3) [25,31].The characteristic absorption bands of P were identified in P/G-EPnF.Moreover, the absorption bands of G in the P/G1-EPnF and P/G2-EPnF samples are indicated by black arrows in insets 1 and 2 in Figure 3.The small bands at 1630 cm −1 (-COO) and 890 cm −1 (C-C) demonstrate the presence of G at a very low concentration in the P/G-EPnF meshes.Crystals 2023, 13, x FOR PEER REVIEW 5 of 20

Figure 4 .
Figure 4.The static water CA.(a) R-P/G-EPnF, (b) schematic representing the different ways of measuring CA on an anisotropic surface, (c) A-P/G-EPnF obtained frontal to the fibers, and (d) A-P/G-EPnF obtained lateral to the fibers.Asterisk (*) indicates p < 0.05.

Figure 6 .
Figure 6.Average fiber diameter of the (a) R-P/G-EPnF meshes obtained with the flat collector and (b) A-P/G-EPnF meshes obtained with the rotary collector.Asterisk (*) indicates p < 0.05.

Figure 6 .
Figure 6.Average fiber diameter of the (a) R-P/G-EPnF meshes obtained with the flat collector and (b) A-P/G-EPnF meshes obtained with the rotary collector.Asterisk (*) indicates p < 0.05.

Figure 9 .
Figure 9. Schematic representation of the interaction between P, G, TRIS, and Ca 2+ .

Figure 9 .
Figure 9. Schematic representation of the interaction between P, G, TRIS, and Ca 2+ .

Table 4 .
Unit cell volume and quantitative analysis of crystal structure and phase of CaCO 3 .