Design, Characterization, and Enhanced Performance of Electrospun Chitosan-Based Nanocomposites Reinforced with Halloysite Nanotubes and Cerium Oxide Nanoparticles for Wound Healing Applications

The electrospun 3D scaffolds exhibit a high surface area, excellent water and gas permeability, and a structure that closely mimics ECM of body tissues, making them highly promising for wound healing applications [64]. After preparation, the CS-PEO-HNT and CS-PEO-HNT-CeONP electrospun mats were heated at 80 °C for 3 h to remove unbound acetic acid and convert CS into its water-insoluble form. SEM revealed the uniform formation of nanofibers in both mats. The composite solution containing HNTs produced smooth, defect-free nanofibers (Figure 1a), with an average fiber diameter of approximately 151 ± 63 nm (Figure 1c).

In contrast, the morphology of composite nanofibers containing both HNTs and CeONPs showed less homogeneity. Individual nanofibers exhibited a rough surface, likely due to the aggregation of different nanoparticles on the fiber surface (Figure 1b). The average diameter of these fibers was larger, measuring 233 ± 86 nm (Figure 1d).

EDX was performed on CS-PEO-HNT-CeONP to confirm the presence of CeONPs (Ce) and HNTs (Si) within the nanofibers (Figure 2). Elemental distribution maps indicated a uniform dispersion of HNTs and CeONPs throughout the electrospun material.

CS interacts with the surface of HNTs, and this interaction can be enhanced through heating, as previously noted [45]. The incorporation of HNTs resulted in a reduction in the swelling degree of the CS-PEO-HNT electrospun mats compared to the CS-PEO material (Table 1). Conversely, the addition of CeONPs led to an increase in the swelling degree of the CS-PEO-CeONP composite in both water and 0.9% NaCl solution (Table 1), attributed to the rearrangement of intermolecular bonds within the CeONP system. Similar swelling properties were previously observed in CS-PEO-CeONP composites (Table 1, [36]). Additionally, we have reported similar results for CS-based films containing CeONPs, which exhibited increased hydrophilicity compared to pure CS films [57].

The WAXS of the CS-PEO-HNT-CeONP electrospun composite was quite challenging due to the low mass content of HNTs and CeONPs and the high porosity of the sample. The halloysite peaks overlapped with those of CS and PEO and appeared as a shoulder at 2θ of 12°. Weak reflections of CeONPs were observed at 2θ values of 28.7°, 33.0°, 47.5°, and 57.0°, corresponding to the (110), (200), (220), and (311) crystal planes of CeO₂ (fluorite cubic crystal structure), respectively: ICDD PDF Map #34-394, data from National Institute of Standards and Technology, Gaithersburg, MD, USA.

More intense reflections characteristic of CeONPs were observed in a CS-PEO-HNT-CeONP film of similar composition (Figure 3). This suggests that the morphology of the porous nonwoven composite influences the X-ray diffraction analysis. We have previously observed similar effects when comparing the diffractograms of multilayer composites in film and nonwoven forms [65].

The mechanical properties of the materials are summarized in Table 2, and their stress–strain curves are shown in Figure 4. The deformation behavior of all samples was typical of electrospun materials, with no necking observed beyond the yield point. The final strain values were in the range of 5–6%, which is suitable for biomedical applications.

The incorporation of nanofillers into the CS polymer matrix resulted in materials with enhanced mechanical properties. The observed near-doubling of the Young's modulus in the dual-filler composite (CS-PEO-HNT-CeONP) compared to the pure CS-PEO mat supports our hypothesis of a synergistic reinforcing effect, where the HNTs provide the primary structural support and the CeONPs potentially contribute to additional interfacial bonding. The high strength of the HNT-containing material is attributed to the reinforcement of the polymer matrix by rigid nanotubes and the interaction between positively charged CS and negatively charged HNTs (Table 2). The introduction of CeONPs also increased the stiffness of the material, albeit to a lesser extent, due to the formation of additional bonds between polymer macromolecules and nanoparticle surfaces [36,57]. However, the formation of nanocomposites led to a slight reduction in the final deformation of the samples (Table 2), a common phenomenon in polymer-inorganic nanocomposites [66]. The most significant reduction in ε_b was observed in the electrospun mat containing both HNTs and CeONPs.

TGA results are shown in Figure 5a. All materials contained approximately 7% moisture, which evaporated between 120–130 °C. Thermal degradation of the composites began at 150–160 °C. Similar to other polysaccharide-based materials, these electrospun composites underwent thermo-oxidative degradation in two stages, as evident in the differential mass loss curves (Figure 5b). The first stage occurred between 170–300 °C, during which the samples lost about 50–55% of their initial weight. This stage involved multiple overlapping processes, with maximum intensity at 210–225 °C and 250–265 °C, reflecting the degradation characteristics of both polymers. No significant differences were observed between the thermal behavior of the CS-PEO mat and the HNT-containing nanocomposite (Table 3). However, the introduction of CeONPs shifted the degradation to lower temperatures, with t₅ and t₁₀ values decreasing by 8–10 °C (Table 3).

The second stage of intense degradation occurred above 360–370 °C. In this range, the CS-PEO-HNT-CeONP mat exhibited a more pronounced weight loss, with 12% of its initial weight lost between 365–385 °C. This was accompanied by a sharp exothermic peak in the DTA curve (Figure 5c), which was 90 °C lower and more intense compared to the other samples. This effect is attributed to the catalytic role of cerium oxide in the degradation of organic materials [67]. However, this catalytic effect is not critical for the biomedical application of the CS-PEO-HNT-CeONP composite, as it occurs during the deep degradation stage when the material is no longer functional. The reduced thermal stability of the composite fibers containing both HNTs and CeONPs aligns with mechanical and morphological results, suggesting that nanoparticle interactions increase the heterogeneity of the nanofibers.

The cytocompatibility of the fabricated electrospun mats was evaluated using mesenchymal stem cells (MSCs) as a model system, with particular emphasis on cellular adhesion and proliferation dynamics. MSCs were selected as a representative cell type for assessing potential cytotoxic effects relevant to various somatic cell populations. Quantitative analysis revealed a statistically significant reduction (p < 0.05) in MSC counts on CS-PEO-HNT-CeONP substrates compared to CS-PEO-HNT and to CS-PEO control (Table 4). Notably, MSC cultures on CS-PEO-HNT-CeONP matrices exhibited a distinct morphological response characterized by spontaneous spheroid formation, suggesting altered cell–substrate interactions.

Cell morphology also differed between the samples. On CS-PEO-HNT, adhesive cells were abundant, with most exhibiting a typical elongated shape and longitudinal striations. However, some cells appeared spherical with protrusions. The elongated cells were smaller in area compared to those on coverslips, and cell colonies were not distinctly visible. In contrast, CS-PEO-HNT-CeONP displayed many single cells with a nearly rounded shape, with only a few elongated cells present. Numerous small spheroid colonies, measuring up to 66 ± 11 μm in diameter, were observed on the sample surface. Some spheroids exhibited signs of cell migration along the colony periphery (Table 4, Figure 6).

In summary, the developed biohybrid scaffolds exhibited satisfactory cytocompatibility profiles in vitro when evaluated using rat MSCs as a model system. Quantitative assessment revealed that the incorporation of dual nanofillers (CS-PEO-HNT-CeONP) resulted in moderately reduced cellular compatibility compared to the HNT-only formulation (CS-PEO-HNT), as demonstrated by decreased cell adhesion efficiency and proliferative capacity (p < 0.05). Importantly, both scaffold variants maintained non-cytotoxic characteristics throughout the evaluation period, suggesting their fundamental biocompatibility despite the observed differences in cellular responses.

Subcutaneous implantation of the developed materials was performed to evaluate their in vivo biocompatibility and biodegradation. The results indicated uncomplicated wound healing in the absence of chronic pain, inflammation, or pronounced scar formation. The appearance and structure of the implanted CS-PEO-HNT-CeONP mat in a formalin-fixed specimen are shown in Figure 7a.

The CS-PEO-HNT mat in the implant pocket formed crumpled clumps up to 460 μm thick. Degradation began by the end of the first month, primarily through surface layer delamination. By the third month, the ratio of disintegrated to preserved mat was approximately 1:3 to 1:4. Single cells were visible within the mat, and a thin-walled connective tissue capsule without aseptic inflammation began to form.

In contrast, CS-PEO-HNT-CeONP remained intact in the implant pocket without clumping, as shown in macroscopic images of formalin-fixed samples 30 days post-surgery (Figure 7). Histological analysis of the implantation site at various time intervals revealed that the dual-filler CS-PEO-HNT-CeONP mat degraded at a slower rate compared to CS-PEO-HNT. By the end of the first month, the mat remained dense, with no peripheral delamination. A key feature of its biocompatibility was the early formation of a connective tissue capsule. By the second month, the scaffold exhibited signs of degradation, accompanied by the formation of a well-defined connective tissue capsule with evident vascularization and a reduction in aseptic inflammation when compared to CS-PEO-HNT. By the third month, the mat thickness was 500–700 µm, with surface degradation occurring at approximately ¼ of its thickness. Inflammatory infiltration was largely absent, persisting only in areas of higher material disintegration. A thin, well-organized connective tissue capsule composed of collagen fibers was evident, with no signs of infiltration. The histological patterns of the implantation site at different time intervals are shown in Figure 8.

In summary, histological analysis confirmed the superior in vivo performance of the dual-filler composite. The CS-PEO-HNT-CeONP scaffold exhibited a slower degradation rate, superior structural integrity without clumping, and a significantly attenuated inflammatory response—as qualitatively assessed by histology—compared to the HNT-only formulation. Critically, these findings align with our initial mechanistic hypothesis: the structural reinforcement provided by HNTs, combined with the anti-inflammatory and proposed matrix-stabilizing effect of CeONPs, creates a more controlled and biocompatible degradation profile. This synergy promotes the early formation of a well-organized connective tissue capsule. While these qualitative results are highly promising, future studies incorporating quantitative analyses—such as immunohistochemical cell counts (e.g., for CD68+ macrophages) or cytokine profiling (e.g., TNF-α, IL-1β)—would provide robust objective evidence to further solidify these claims. Nonetheless, the present results underscore the potential of the CS-PEO-HNT-CeONP material as an effective scaffold for managing extensive skin defects, as it provides durable coverage while establishing an optimal regenerative microenvironment.

The wound-healing properties of the developed materials were assessed using a wound model that heals by primary intention (per primam intentionem). On the 21st day post-surgery, veterinary examinations of all animal groups revealed no spontaneous mortality or clinical signs of ill health, such as behavioral stress symptoms, abnormalities in coat or visible mucous membranes, or local complications like purulent infections, tissue necrosis, or contact allergic dermatitis. Over the 21-day observation period, neither the CS-PEO-HNT nor the CS-PEO-HNT-CeONP scaffolds underwent significant resorption. All samples remained sutured in place, albeit with varying degrees of adherence (Figure 9). By the end of the experimental period, wounds in the CS-PEO and CS-PEO-HNT groups exhibited complete scaffold detachment, which was facilitated by the presence of a newly regenerated epidermal layer. In contrast, wounds treated with CS-PEO-HNT-CeONP showed incomplete epithelial regeneration, and the scaffold consequently maintained persistent adhesion to the underlying granulation tissue.

Histological analysis data are presented in Figure 10. Skin samples from the intact control animals (Figure 10B-a) clearly depict the well-defined collagen fibers characteristic of the reticular dermis. In the sham-operated control group, scab elements with minimal leukocyte infiltration and cellular debris were observed at 21 days post-surgery. The granulation tissue surface was covered by a continuous layer of stratified squamous epithelium, which was thicker compared to intact areas (Figure 10A-b). Histological sections of the skin flap revealed well-formed granulation tissue replacing the wound defect (Figure 10B-b), with extensive vascularization characterized by branching loops of engorged capillaries and vertical vessels. At this stage, signs of epidermal keratinization and papillary layer formation were evident. The granulation tissue consisted of abundant fibroblasts and thin collagen fibers arranged in a reticular pattern, with no signs of inflammation and a moderate presence of macrophages.

In the group treated with the CS-PEO mat, a thick scab heavily infiltrated with leukocytes and cellular debris was observed (Figure 9a). Epithelialization beneath the scab was complete, with the formation of a stratum corneum, although the epithelial layer was thinner than in intact areas and the control group (Figure 10A-c). The underlying layer exhibited dense, parallel bundles of collagen fibers without the formation of a three-dimensional mesh structure (Figure 10B-c). The cellular composition was predominantly fibroblasts, with few fibrocytes and macrophages. Vascularization was minimal, lacking loop-like vessels, and no inflammatory response was noted.

In the CS-PEO-HNT group, a substantial scab with mild leukocyte infiltration and cellular debris was present. The scab had completely detached from the skin surface (Figure 9b). The epithelial layer, though fully formed and covering the wound defect, was thinner than in intact areas (Figure 10A-d). Collagen fibers in the developing dermis were arranged in dense parallel bundles (Figure 10B-d), with fibroblasts being the predominant cell type and a moderate number of fibrocytes. A small number of macrophages were observed, but no inflammatory response was detected. Vascularization was moderate, without the formation of loop-like or vertical vessels.

In the CS-PEO-HNT-CeONP group, a thick scab with moderate leukocyte infiltration and cellular debris was observed. However, epithelial regeneration was incomplete, as a central area of the wound defect persisted, covered by the closely adhered scaffold (Figure 9c). At the periphery of the former wound, where epithelialization had occurred, the scaffold had detached. The tissue beneath the non-epithelialized central defect exhibited sparse, thin collagen fibers and signs of inflammation, including dilated thin-walled vessels (hyperemia), lymphocytes, and numerous active macrophages. Furthermore, in contrast to the other groups, the region beneath the newly formed epithelium contained fewer collagen fibers, which were arranged in a loose, reticulate pattern rather than the dense, parallel bundles seen elsewhere (Figure 10B-e). The cellular composition in this area included randomly distributed fibroblasts alongside a high density of active macrophages.

In summary, the comparative analysis revealed that the two composite scaffolds mediate distinct and mechanistically insightful wound healing patterns. The CS-PEO-HNT scaffold facilitated rapid epithelialization but resulted in a thinner epithelial layer and the formation of parallel collagen bundles—a histological signature of fibrotic healing and foreign body encapsulation.

In striking contrast, the CS-PEO-HNT-CeONP composite demonstrated a truly synergistic effect, orchestrating a fundamental trade-off between healing kinetics and tissue quality. While the initial epithelial migration was temporarily delayed, the ultimate outcome was the regeneration of a dermal architecture strikingly similar to native skin morphology. We propose that this occurs through a defined mechanistic framework: the HNTs provide sustained structural support, preventing scaffold collapse and enabling guided tissue formation, while the CeONPs attenuate the inflammatory response and modulate the oxidative microenvironment. Together, they create a pro-regenerative niche that promotes the development of an organized, reticular collagen structure instead of scar tissue.

This insight—that a delay in closure can be strategically exploited to achieve superior regeneration—opens a compelling avenue for future therapeutic design. It suggests the potential for a dual-phase treatment strategy, where an initial dressing promotes epithelialization, followed by the application of CS-PEO-HNT-CeONP to guide the remodeling phase and minimize scarring. Alternatively, a single, smart dressing could be engineered to temporally release bioactive factors, addressing both phases of healing simultaneously.

Consequently, these results robustly position the CS-PEO-HNT-CeONP composite not merely as a passive dressing, but as an active scaffold for advanced wound management, where the primary objectives are the quality of healing and scar reduction, moving beyond the paradigm of simply accelerating wound closure.

Crab CS (Bioprogress, Shchelkovo, Russia) with a characteristic viscosity ([η]) of 2.56 dL/g and a viscosity-average molecular weight (Mη) of 6.0 × 10⁴ was used in this study. The characteristic viscosity was determined in a solvent consisting of 0.33 M CH₃COOH and 0.3 M NaCl using an Ubbelohde capillary viscometer at 20 °C. The molecular weight (Mη) was calculated using the Mark–Houwink equation: [η] = 3.41 × 10⁻³Mη^1.02 [68]. The degree of deacetylation, determined by ¹H NMR spectroscopy, was found to be 82%. PEO with a molecular weight of 9.0 × 10⁵ was purchased from Sigma-Aldrich (St. Louis, MO, USA), and HNTs were obtained from NaturalNano Inc. (Rochester, NY, USA). Citrate-stabilized CeONPs were synthesized following a previously reported procedure [69], yielding ultrasmall particles with a hydrodynamic diameter of approximately 2.3 nm (dynamic light scattering) and a core size of 1.5–2.5 nm (scanning electron microscopy, SEM), which form stable aqueous dispersions.

To ensure homogeneous nanofiller distribution, both HNTs and CeONPs were subjected to pre-dispersion protocols. The CS-stabilized CeONP dispersion was prepared according to established methodology [57]. Briefly, a 1% aqueous dispersion of citrate-stabilized CeONPs was mixed with a 0.1% CS solution in 1% acetic acid using an IL 100-6 ultrasonic homogenizer (Ultrasonic Techniques-INLAB, St. Petersburg, Russia) for 10 min, maintaining a mass ratio of 1:1 between citrate-stabilized CeONPs and CS.

For HNT dispersion, the predetermined mass of nanotubes was initially hydrated in distilled water for 24 h, followed by 10 min of sonication using an IL 100-6 ultrasonic disperser (Ultrasonic Techniques-INLAB, St. Petersburg, Russia). This pre-dispersion was subsequently combined with a 0.2% CS solution in 1% acetic acid under continuous sonication.

The polymer solution was prepared using a CS concentration of 3%, previously established as optimal for achieving appropriate viscosity for stable jet formation in our non-capillary electrospinning configuration. CS powder was initially dispersed in deionized water under vigorous stirring for several hours, followed by addition of glacial acetic acid with continuous stirring until complete dissolution was achieved. A 3% aqueous PEO solution was then incorporated to yield the final spinning solution (CS-PEO solution) containing 3% CS and 0.3% PEO in 70% v/v acetic acid.

The HNT-containing spinning solution (CS-PEO-HNT solution) was formulated by introducing the pre-dispersed HNT suspension into the CS-PEO solution under vigorous stirring. The final composite solution (CS-PEO-HNT-CeONP solution) was prepared by sequential addition of pre-dispersed HNT and CeONP suspensions (5% HNTs and 8% CeONPs by weight of CS) to the CS-PEO solution under identical mixing conditions.

Electrospinning was performed using a non-capillary Nanospider NS Lab 500 apparatus (Elmarco, Liberec, Czech Republic) with an applied voltage of 50–65 kV to ensure stable jet formation. The spinning electrode rotation speed was maintained at 10 rpm with a fixed inter-electrode distance of 24 cm. Fibers were collected on a paper substrate, producing two distinct nanofiber mat variants: CS-PEO-HNT and CS-PEO-HNT-CeONP. Following electrospinning, all samples were conditioned at room temperature for 5 days followed by thermal treatment at 80 °C for 3 h. All electrospinning parameters, including polymer concentrations, solvent composition, applied voltage, inter-electrode distance, and collector rotation speed, were kept identical to those used in our previous study [36] to ensure a valid comparison of the resulting materials.

The swelling capacity of the electrospun mats was evaluated gravimetrically by measuring their weight increase after immersion in both distilled water and physiological solution (0.9% NaCl). Precisely weighed air-dry samples (W_dry) were immersed in the respective solutions at room temperature (20 ± 2 °C) until swelling equilibrium was reached. The swollen mats were then carefully removed, excess surface liquid was blotted away using filter paper, and they were immediately weighed again (W_wet). The swelling degree was calculated according to the formula: Swelling (g/g) = [(W_wet − W_dry)/W_dry].

The morphology of the resulting mats was characterized by SEM using a SUPRA 55VP setup (Carl Zeiss, Oberkochen, Germany) and wide-angle X-ray scattering (WAXS) using a Bruker D8 DISCOVER X-ray diffractometer equipped with CuKα radiation (Bruker, Karlsruhe, Germany). SEM images were obtained using both secondary electron and backscattered electron detectors.

To visualize the distribution of CeO₂ within the samples, the electrospun mats were frozen in liquid nitrogen and sectioned. The samples were then mounted on conductive tape and coated with a thin layer of platinum. Energy-dispersive X-ray spectroscopy (EDX) was performed using an EDX-Max 80 mm² detector (Oxford Instruments, Abingdon, UK) to analyze the distribution of cerium. Elemental maps were collected over the entire visible range of the samples.

The mechanical properties of the electrospun mats were investigated using an AG-100kNX Plus (Shimadzu, Kyoto, Japan) in uniaxial extension mode. Strip-shaped samples (2 × 30 mm) were stretched at a rate of 20 mm/min at room temperature according to ASTM D638. Stress–strain curves were recorded during the tests, and parameters such as Young's modulus (E), yield stress (σ_y), ultimate stress (σ_b), and elongation at break (ε_b) were determined. The experiment was performed in triplicate.

Thermal properties were analyzed using a DTG-60 setup (Shimadzu, Kyoto, Japan) capable of simultaneous thermogravimetric analysis (TGA) and differential thermal analysis (DTA). Samples were heated in air to a constant residual weight (550–600 °C) at a rate of 5 °C/min. Thermal stability indices, including τ₅ and τ₁₀ (the temperatures at which 5% and 10% weight loss occurred, respectively), T_fin (the temperature at which thermal degradation was complete), and m_fin − m₀ (residual weight), were derived from the TGA curves.

The cytocompatibility of the biopolymer scaffolds was evaluated using multipotent MSCs isolated from the visceral adipose tissue of healthy male Wistar rats according to a previously described protocol [57]. MSCs were selected for this study as a well-established model system for evaluating the biocompatibility of biomaterials intended for wound healing applications. Their high proliferative capacity, sensitivity to cytotoxic agents, and critical role in natural tissue regeneration processes—including immunomodulation, angiogenesis, and ECM remodeling—make them a highly relevant and stringent cell type for assessing the potential of scaffolds to support cell growth and function in vitro. Furthermore, the response of MSCs to biomaterials is often indicative of the material's interaction with a broad range of somatic cells involved in wound repair.

The cell culture experiments received ethical approval from the Commission for the Control of Care and Use of Laboratory Animals at the Almazov National Medical Research Centre (Protocol No. 21-12PZ#V3, 13 July 2021). Cells were grown under standard conditions at a density of 50,000 cells/mL for 72 h, with test samples compared to control coverslips. The experiment was performed in triplicate. For fluorescence imaging, we stained the cells using rhodamine-conjugated phalloidin (Thermo Fisher Scientific, Waltham, MA, USA; 1:500) to label F-actin and 4,6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich, St. Louis, MO, USA; 1:40,000) to label nuclei. Images were obtained using a Zeiss Axiovert inverted fluorescence microscope (Carl Zeiss, Jena, Germany) equipped with a Canon camera (Canon Europa N.V., Amstelveen, The Netherlands). We quantified cell adhesion by assessing 10 random microscopic fields, while qualitative morphological analysis was performed at 10× and 40× magnifications. All statistical analyses were performed with the nonparametric Mann–Whitney U test in GraphPad Prism 8 software (San Diego, CA, USA).

The in vivo experiments were conducted in compliance with the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85-23, revised 1996) and the European Convention for the Protection of Vertebrate Animals Used for Experimental and Other Scientific Purposes. The specific experimental protocol was reviewed and approved by the Institutional Animal Care and Use Committee of Pavlov First St. Petersburg State Medical University (Protocol No. PZ_21-02#Zhuravskii S.G. V3, 18 September 2023). Experiments were performed according to ARRIVE guidelines (https://arriveguidelines.org/↗ (accessed on 24 October 2025)).