Solubility-Enhanced α‑Mangostin Bilayer Films: Synergizing Mucoadhesion and Protection for Effective Aphthous Ulcer Treatment

Abstract

Aphthous ulcers are prevalent and painful oral mucosal lesions that substantially compromise patient quality of life, requiring effective localized treatment. α-Mangostin, a bioactive compound from mangosteen pericarp, exhibits potent anti-inflammatory properties but suffers from poor aqueous solubility. This study developed a double-layer mucoadhesive film incorporating an α-mangostin-based solid dispersion (MPESD) to enhance solubility and provide dual-functionality: protection and therapy. Mangosteen peel extract was formulated into solid dispersions using PVP K30, significantly enhancing solubility from undetectable levels to 8.12 ± 0.04 mg/mL (1:4 ratio). The films, comprising an ethylcellulose backing layer and a gelatin-HPMC mucoadhesive layer, were fabricated via solvent casting. Mechanical testing confirmed suitable puncture strength (1.63 N/mm²) and flexibility (20% elongation) for oral application. The inclusion of HPMC improved surface wettability and prolonged in vitro residence time to >8 h, while the optimized film achieved 96% drug release within 2 h. These findings demonstrate that the MPESD-loaded double-layer film presents a promising formulation strategy with the potential for managing aphthous ulcers.

Introduction

Aphthous ulcers, commonly known as canker sores, represent the most prevalent oral mucosal lesions in the general population, with a prevalence reaching 25% and a recurrence rate as high as 50% within three months. , Although these lesions typically heal spontaneously within 7–10 days without scarring, they cause considerable pain and discomfort, significantly impacting patients’ quality of life by affecting eating and swallowing functions. , The etiology of aphthous ulcers remains unclear; however, these lesions characteristically develop on soft oral tissues such as the buccal mucosa, tongue, gingiva, and inner lip, presenting as inflammatory ulcers.

Current treatment approaches are symptomatic and depend on disease severity and recurrence frequency. Topical mouth patches containing various active ingredientsincluding corticosteroids, anesthetic agents, nonsteroidal anti-inflammatory drugs (NSAIDs), and herbal extractsare commonly employed. However, conventional patches have significant limitations, including easy detachment from the application site. Film-based formulations overcome these drawbacks by providing prolonged residence time, reducing exposure to irritants, and facilitating faster healing and pain reduction while decreasing recurrence rates, thereby improving therapeutic efficacy. Since aphthous ulcer patches must remain in the oral cavity, their development requires the use of mucoadhesive polymersthose that are hydrophilic, capable of swelling, and able to form bonds with oral mucosal molecules. − Examples include hydroxypropyl methylcellulose (HPMC), polyvinylpyrrolidone (PVP), poly­(vinyl alcohol) (PVA), and chitosan. Given the inflammatory nature of aphthous ulcers, incorporating anti-inflammatory and antioxidant agents into film formulations can further enhance therapeutic outcomes.

Garcinia mangostana Linn. (mangosteen), widely distributed throughout Southeast Asia including Malaysia, Indonesia, and Thailand, contains xanthonesnotably α-mangostinin its dried pericarp. α-Mangostin is the predominant and pharmacologically significant compound in mangosteen, demonstrating multiple biological activities including anti-inflammatory, antioxidant, antiallergic, antibacterial, antifungal, and anticancer properties. Winarni et al. reported the mangosteen peel extract (MPE) improved the open wound healing process in diabetic mice by antioxidant properties. Furthermore, α-mangostin has been incorporated into polymer-based wound dressings, where it demonstrated rapid bactericidal effects against wound infections caused by Staphylococcus aureus, highlighting its potential as a therapeutic agent in infection-associated wound management. However, as a hydrophobic compound classified as Biopharmaceutics Classification System Class 2, α-mangostin exhibits extremely low aqueous solubility, limiting its incorporation into pharmaceutical formulations. Solid dispersion technology is a highly promising strategy used to enhance the therapeutic potential of herbal extracts by improving their solubility. This technique improves aqueous solubility of poorly soluble compounds by dispersing it as a solid within an inert matrix, typically composed of hydrophilic polymers in either crystalline or amorphous form. Literature review reveals multiple polymers used to prepare α-mangostin solid dispersions, including alginate, PVP, chitosan, polymethacrylates (Eudragit), and HPMC, either individually or in combination. ,−

Based on the aforementioned rationale, this research aims to develop a double-layer film formulation for aphthous ulcer treatment incorporating α-mangostin within polymers that enhance its solubility while possessing mucoadhesive properties. This approach is intended to maximize contact α-mangostin with oral mucosa to optimize therapeutic efficacy and promote ulcer healing. Additionally, physicochemical properties of the resulting filmsincluding drug loading percentage, mucoadhesion duration and strength, and drug release profilewill be characterized. The developed film may potentially be extended for treating oral ulcers resulting from various etiologies.

Materials and Methods

Materials

Dark-purple mature mangosteen fruits (G. mangostana Linn.) were purchased from a local market in Phitsanulok, Thailand. α-Mangostin standard (analytical reagent, Lot No. 00013095-122) was obtained from ChromaDex, Inc. (Irvine, CA, USA). Gelatin type A with bloom strength 300 (Sigma-Aldrich, Steinheim, Germany) was used as the mucoadhesive layer material. Ethylcellulose (100 cP, analytical reagent, Lot No. SB09013T10) was supplied by Colorcon (Singapore) for a protective layer. Polyvinylpyrrolidone K30 (PVP K30) was obtained from BASF (Shanghai, China). Hydroxypropyl methylcellulose (HPMC; AnyCoat-C AN5, 2910 substitution type, viscosity 5 mPa·s in 2% solution) was supplied by Lotte Fine Chemical (Seoul, Korea). Glycerin and castor oil were procured from Vanrut Co., Ltd. (Bangkok, Thailand). Fresh porcine intestinal tissue was purchased from a local butcher on the morning of in vitro residence time experiment. All other reagents, including sodium chloride, potassium chloride, dipotassium hydrogen phosphate, disodium hydrogen phosphate anhydrous, and potassium dihydrogen phosphate, were of analytical grade and obtained from standard commercial suppliers.

Preparation of Mangosteen Peel Extracts

The peels of dark-purple mature mangosteen fruits were separated from the arils, cleansed, chopped, and hot-air-dried at 50 °C for 1 week, then milled to a particle size of approximately 2 mm. The milled material (800 g) was macerated in 95% ethanol (4 L) under gentle agitation (20 rpm) using an orbital shaker at room temperature for 72 h. The resulting mixture was filtered sequentially through a coarse sieve, cotton wool, and Whatman No. 1 filter paper. The filtrate was concentrated under reduced pressure using a rotary vacuum evaporator (BUCHI, Interface T-300, Switzerland) at 50 °C for 18–19 h to yield the crude mangosteen peel extract (MPE). The α-mangostin content was subsequently determined spectrophotometrically using a UV–vis spectrophotometer (Evolution 60, Thermo Fisher Scientific, USA) at 317 nm, following dissolution of the MPE in methanol and quantification against a standard curve of α-mangostin in methanol (2–10 μg/mL). Results were expressed as mean values ± standard deviation (n = 3).

Preparation of Mangosteen Peel Extract Solid Dispersion

Solid dispersions of the mangosteen peel extract (MPESD) were prepared using a melting-solvent technique. , Crude extract and polyvinylpyrrolidone K30 (PVP K30) were weighed according to predetermined weight ratios (1:3, 1:4, 1:5, and 1:7 w/w). The crude extract and PVP K30 were dissolved in ethanol and heated at approximately 60 °C until a homogeneous mixture was obtained. The mixture was rapidly cooled in an ice bath and subsequently dried in a hot air oven at 60 °C. The dried solid dispersion was pulverized using a glass mortar, passed through a sieve to ensure uniform particle size, and stored in opaque, sealed containers protected from light. To minimize the risk of age-induced recrystallization, the further studies were conducted within 7 days of preparation.

Drug Solubility

The solubility of α-mangostin from the MPESD formulations was evaluated using an equilibrium solubility method. MEPSD powder, equivalent to 10 mg of α-mangostin, was suspended in 1 mL of deionized water and continuously stirred at 250 rpm for 8 h at ambient temperature to ensure equilibrium was reached. Subsequently, the samples were centrifuged at 14,000 rpm for 20 min, the clear supernatant was collected and diluted with methanol. Then, the α-mangostin content was quantified spectrophotometrically at 317 nm. All solubility studies were performed in triplicate.

Preparation of Double-Layer Film Loaded with MPESD

Double-layer mucoadhesive films, consisting of a backing layer and a drug-loaded mucoadhesive layer, were prepared using the solvent casting method. The backing layer was prepared initially, dried completely, and subsequently served as the substrate for the mucoadhesive layer.

The backing layer solution was prepared by mixing ethyl cellulose (15% w/w) with absolute ethanol (77.5% w/w) and castor oil (7.5% w/w, equivalent to 50% of the ethyl cellulose content) until a homogeneous mixture was obtained. The solution was subjected to sonication for 20 min to ensure uniform dispersion and eliminate air bubbles. The mixture was then cast onto a Teflon sheet mold and dried in a hot air oven at 60 °C for 4 h until completely solvent-free.

After the backing layer was formed, the mucoadhesive layer was prepared by dissolving gelatin (4 g) in hot deionized water (20 mL), followed by the addition of HPMC (0, 0.04, and 0.12 g), glycerin (1.6 g), and MPESD powder (1.125 g). The mixture was stirred thoroughly until a homogeneous dispersion was obtained, then cast onto the dried backing layer and subsequently dried at 45 °C for 2 h to obtain double-layer mucoadhesive films.

Scanning Electron Microscopy

The surface morphology and cross-sectional architecture of the double-layer mucoadhesive films were evaluated using scanning electron microscopy (SEM). Film samples (1 cm × 1 cm) were mounted onto aluminum stubs and sputter-coated with a gold layer to ensure electrical conductivity. Micrographs were obtained using a scanning electron microscope (Model 1455VP, LEO Co., Ltd., England) at selected magnifications to characterize the integrity of the bilayer structure and surface topography.

Evaluation of Mechanical Properties

The mechanical integrity of the double-layer films was assessed by determining puncture strength and elongation at break. Film samples (1.5 cm × 1.5 cm) were analyzed using a texture analyzer (TA.XT, Stable Micro Systems Co., Ltd., England) equipped with a 30 kg load cell and a spherical probe (P/5S). The test parameters were set as follows: pretest contact force of 5 g and test speed of 0.30 mm/s. The probe was driven downward until film rupture occurred, and the maximum force required to break the film was recorded. Puncture strength and elongation at break (%) were calculated using eqs and , respectively. All measurements were performed in quintuplicate (n = 5), and results were expressed as mean ± standard deviation (SD).

$$\mathrm{P}\mathrm{u}\mathrm{n}\mathrm{c}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{g}\mathrm{t}\mathrm{h}=\frac{\mathrm{B}\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{k}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{f}\mathrm{o}\mathrm{r}\mathrm{c}\mathrm{e}(\mathrm{N})}{\mathrm{A}\mathrm{c}\mathrm{s}({\mathrm{m}\mathrm{m}}^2)}$$ 1

where Acs is cross-sectional area of the edge of the dry film located in the path of cylindrical opening of the film holder (Acs = 2 rd, where r is the radius of the hole, d is the thickness of the film).

$$\%\mathrm{E}\mathrm{l}\mathrm{o}\mathrm{n}\mathrm{g}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}=\frac{\sqrt{{\mathrm{r}}^2+{\mathrm{D}}^2}-\mathrm{r}}{\mathrm{r}}\mathrm{X}100$$ 2

where r is the radius of the film exposed in the cylindrical hole of the film holder and D is the displacement of the probe from point of contact to the point of film puncture.

In Vitro Mucoadhesive Residence Time

Porcine intestinal mucosa tissue was employed as a model for oral mucosa. Fresh tissues were rinsed with phosphate-buffered saline (PBS), cut into sections (2.5 cm × 1.5 cm), and fixed onto glass slides using double-sided tape, as illustrated in Figure . Then, the film was applied to the tissue-coated slide, which was then vertically positioned against the inner wall of a beaker containing 200 mL of artificial saliva maintained at 37 ± 1 °C. The assembly was placed in a water bath shaker operated at 15 ± 1 strokes per min. The time required for complete detachment of the film from the mucosal tissue was recorded. Observations were made at 30 min intervals over an 8 h period. The residence time for each film was recorded as the duration from initial application until complete detachment. The detachment time of the 5 films was recorded as the in vitro residence time. Artificial saliva was freshly prepared consisting of sodium chloride (0.084%), potassium chloride (0.12%), dipotassium hydrogen phosphate (0.026%), and distilled water (99.77%), with the pH adjusted to 7.0 using lactic acid.

1.

Schematic illustration of the in vitro residence time test.

Wettability Test

The surface wettability of the double-layer films was assessed by measuring the contact angle between purified water and the surface of the mucoadhesive layer. Measurements were performed using a drop shape analyzer (Model DSA25E, Scientific Promotion Co., Ltd., Hamburg, Germany) employing the sessile drop technique. The experimental parameters were set as follows: test liquid, purified water; temperature, 25 °C; droplet volume, 10 μL; application rate, 10 μL/s; and needle diameter, 0.5 mm. Each measurement was repeated five times per sample and contact angle values were recorded and expressed as mean ± standard deviation.

In Vitro Drug Release and Kinetic Drug Release Analysis

The in vitro release profile of α-mangostin from the double-layer films was evaluated using the Franz diffusion cell method. The membrane-free Franz cell setup was designed to simulate direct contact of the film with an aphthous ulcer (open wound) in the oral cavity. The film was placed directly on the receptor medium without a semipermeable membrane in order to evaluate the α-mangostin release from the system, rather than permeation across a buccal barrier. The receptor chamber was filled with 7 mL of artificial saliva (pH 7.0) and maintained at 37 ± 1 °C under magnetic stirring. Aliquots of 0.5 mL were withdrawn at predetermined time intervals (0.25, 0.5, 1, 2, 4, and 6 h) and immediately replaced with an equal volume of fresh artificial saliva maintained at the same temperature to ensure sink conditions. The withdrawn samples were analyzed using UV–visible spectrophotometry at a wavelength of 317 nm to determine the α-mangostin concentration. A standard curve of α-mangostin was prepared under identical conditions to facilitate quantification. The cumulative amount of α-mangostin released was calculated and plotted as a function of time.

The release kinetics of α-mangostin were evaluated by fitting the release data to four kinetic models: zero-order, first-order, Higuchi, and Korsmeyer–Peppas models. The analysis was performed using the DDsolver add-in (version 1) in Microsoft Excel 365, and the corresponding model equations are presented below.

Zero-order model

$$\frac{M_{\mathrm{t}}}{M_{\infty }}=k_0\mathrm{t}$$ 3

First-order model

$$\ln \left(\frac{M_{\mathrm{t}}}{M_{\infty }}\right)=k_1\mathrm{t}$$ 4

Higuchi model

$$\frac{M_{\mathrm{t}}}{M_{\infty }}=k_{\mathrm{H}}\mathrm{t}$$ 5

Korsmeyer–Peppas

$$\frac{M_{\mathrm{t}}}{M_{\infty }}=k_{\mathrm{K}\mathrm{P}}{t}^n$$ 6

where M _t and M _∞ represent the cumulative amount of drug released at time t and the total amount of drug contained in the film, respectively. The parameters k ₀, k ₁, k _H, and k _KP denote the kinetic constants for the zero-order, first-order, Higuchi, and Korsmeyer–Peppas models, respectively. The diffusional exponent n indicates the drug-release mechanism. For film systems, n ≤ 0.45corresponds to Fickian diffusion, 0.45 < n < 0.89 indicates anomalous (non-Fickian diffusion) transport, and n ≥ 0.89 represents Case II transport. The best-fitting model was determined based on the coefficient of determination (R ²).

Statistical Analysis

All the quantification results are presented as mean ± SD from triplicate experiments. The data were analyzed by one-way analysis of variance (ANOVA). Differences were statistically significant when the p value was ≤ 0.05.

Results and Discussion

Optimization of Mangosteen Peel Extract Solid Dispersion

Mature mangosteen peel was selected as the primary α-mangostin source based on its rich bioactive content. Ethanolic extraction followed by solvent evaporation yielded a mangosteen peel extract (MPE) of 12.1%. UV–visible spectrophotometric analysis at 317 nm confirmed a notably high α-mangostin content of 38.63 ± 0.35% w/w, which surpasses the 29–30% w/w previously reported by Naing et al. for ethanol extracts of Thai mangosteen peel.

The poor aqueous solubility of crude mangosteen peel extract is largely governed by the physicochemical properties of its primary bioactive constituent, α-mangostin. As a highly lipophilic prenylated xanthone with a rigid planar structure and a high octanol–water partition coefficient (Log P 4.6), α-mangostin presents a significant thermodynamic barrier to aqueous dissolution, further compounded by strong intermolecular hydrogen bonding within its stable crystal lattice. , To address this limitation, MPE solid dispersions (MPESD) were successfully prepared using PVP as a hydrophilic carrier to enhance solubility and dissolution rate. The resulting MPESD formulations appeared as yellowish-brown dry powders.

The effect of the carrier ratio (MPE/PVP K30) on drug loading and aqueous solubility is summarized in Table . The ratio of MPE to PVP K30 significantly influenced both drug loading and solubility. Increasing the proportion of PVP K30 from 1:3 to 1:7 resulted in a decrease in % drug loading, ranging from 5.76% to 3.33%, which is attributable to the dilution effect of the carrier.

1. Effect of MPE/PVP K30 Ratios on % Drug Loading and Solubility of α-Mangostin Solid Dispersions (MPESD) Compared to Physical Mixture and MPE.

formulation/ratio	drug loading (%)	solubility of α-mangostin (mg/mL)
solid dispersions (SD): Mass ratio of MPE/PVP K30
1:3	5.76 ± 0.01	2.45 ± 0.11
1:4	4.40 ± 0.01	8.12 ± 0.04
1:5	4.36 ± 0.01	8.77 ± 0.21
1:7	3.33 ± 0.01	6.62 ± 0.33
comparator comparisons
physical mixture (PM) (1:4)	-	0.96 ± 0.00
MPE	-	unable to detect

As expected, solid dispersion markedly improved the aqueous solubility of α-mangostin compared to the MPE alone (where solubility was undetectable) and the physical mixture (0.96 mg/mL). Solubility increased with higher PVP K30 content up to 1:5, achieving a maximum of 8.77 ± 0.21 mg/mL, while the 1:4 ratio exhibited comparable solubility (8.12 ± 0.04 mg/mL). This finding is in agreement with prior work demonstrating that PVP K30 solid dispersions markedly improve α-mangostin solubility (∼0.2 μg/mL to 2.7 mg/mL) via amorphization, drug–polymer interactions, and nanomicelle formation. Vongsak et al. also prepared solid dispersions of mangostin-containing compounds via solvent evaporation. FT-IR and XRD confirmed amorphization and hydrogen bonding between mangostin and PVP K30. Notably, the ratio 1:7 resulted in a significant decline in solubility despite having the highest polymer content. This phenomenon suggests that an excess of PVP K30 may increase microenvironment viscosity or compete for hydration shells, thereby hindering rather than facilitating the dissolution of α-mangostin. , While this study demonstrated improved solubility, long-term physical stability of the amorphous solid dispersion was not assessed. Recrystallization remains a known risk in amorphous systems and could compromise dissolution performance, especially under elevated temperature and humidity. These findings should therefore be regarded as short-term, and future work should address comprehensive stability testing to confirm that the amorphous state and its associated dissolution benefit are maintained over time. In this study, the 1:5 ratio demonstrated the highest solubility. However, the 1:4 ratio was selected for incorporation into the film formulation. This decision was based on the negligible difference in solubility between the 1:4 and 1:5 ratios, coupled with the higher drug loading efficiency of the 1:4 formulation.

Physicochemical Properties of α-Mangostin Bilayer Films

Aphthous ulcers are among the most common oral mucosal lesions, frequently causing significant pain and discomfort that interfere with eating, speaking, and overall quality of life. , Effective management requires not only pain relief but also promotion of wound healing at the mucosal site. To address these needs, a double-layer film was designed to provide dual functionality, offering both protective and therapeutic effects.

Double-layer mucoadhesive films incorporating MPESD were successfully prepared using the solvent-casting method. The films consisted of an ethyl cellulose (EC) backing layer and a drug-loaded mucoadhesive layer, fabricated sequentially by casting the backing layer first and allowing it to dry completely before deposition of the mucoadhesive layer. The mucoadhesive layer was formulated from gelatin blended with HPMC at concentrations of 0%, 0.15%, and 0.45% and incorporated MPESD. Glycerin was incorporated as a plasticizer to reduce intermolecular hydrogen bonding within the gelatin matrix, thereby improving chain mobility and film flexibility. The addition of HPMC was expected to influence film mechanical performance, which could modulate mucoadhesion.

The resulting films contained 0.2 mg/cm² of α-mangostin and exhibited a uniform yellowish-brown appearance consistent with the characteristic color of the active compound, whereas blank films were transparent and colorless (Figure ). The selected drug loading of 0.2 mg/cm² was justified based on literature evidence showing that α-mangostin possesses anti-inflammatory, antibacterial, and wound-healing activity in topical formulations. ,, Structurally, the EC backing layer functions as a nonadhesive protective barrier that shields the oral lesion from external irritants such as food and beverages while preventing drug loss toward the oral cavity (Figure A). Moreover, it promoted unidirectional mangostin release toward the mucosa. In contrast, the mucoadhesive layer enables intimate and prolonged contact with the oral mucosal tissue (Figure B), facilitating localized delivery of α-mangostin. This localized delivery was designed to take advantage of α-mangostin’s anti-inflammatory and antioxidant properties, which may help relieve pain and support wound healing at the site of the lesion. Upon exposure to the oral environment, the mucoadhesive layer gradually hydrates and erodes, enabling controlled drug release, while the backing layer maintains structural integrity and directs drug diffusion toward the mucosal surface.

2.

Visual appearance of double-layer mucoadhesive films: MPESD-loaded film formulations containing HPMC at (A) 0%, (B) 0.15%, and (C) 0.45% concentrations, and (D) the blank formulation (No MPESD).

3.

Design of double-layer films loaded with MPESD and cross-sectional morphology (A) Schematic illustration of the bilayer film architecture comprising an ethyl cellulose protective backing layer and a MPESD-loaded mucoadhesive layer. (B) Representative cross-sectional SEM micrographs showing the intact bilayer structure, where both the protective backing layer and the mucoadhesive layer are clearly distinguishable with a well-defined interface.

The incorporation of MPESD into the mucoadhesive layer resulted in a smoother and more compact microstructure compared to the corresponding blank formulations across all HPMC concentrations, as evidenced by cross-sectional imaging (Figure ). This morphological improvement can be attributed to the presence of PVP K30 in the MPESD, which is known to function as a polymeric plasticizer when blended with gelatin. PVP contains both hydrophilic pyrrolidone and hydrophobic alkyl moieties, enabling it to form strong hydrogen bonds with the amino and carboxyl groups of the gelatin chains, thereby enhancing polymer miscibility and chain mobility within the matrix. , This intermolecular interaction disrupts the rigid crystalline arrangement of the standalone gelatin matrix, resulting in a more homogeneous and denser film structure with reduced internal stress. In contrast, formulations without MPESD displayed a rougher, more porous, and irregular morphology with visible cracking, which is consistent with the high internal stress and limited chain mobility that commonly arise during the drying process of unfilled polymer matrices.

4.

Cross-sectional SEM micrographs of double-layer mucoadhesive films containing varying concentrations of HPMC (0%, 0.15%, and 0.45%). Formulations are presented in the absence and presence of MPESD, illustrating the influence of both MPESD and HPMC concentration on the microstructural morphology of the mucoadhesive layer.

Mechanical Performance: Puncture Strength and Elongation at Break

Mechanical properties are critical indicators of a mucoadhesive film’s ability to withstand handling, application, and structural stresses within the dynamic oral environment. In this study, puncture strength and elongation at break were assessed to evaluate the robustness and flexibility of the double-layer films containing varying concentrations of HPMC. As presented in Table , film thickness ranged from approximately 421 to 452 μm across all formulations, with a slight decreasing trend observed at higher HPMC concentrations, though no significant differences were noted. This trend may be attributed to HPMC molecules occupying the interstitial spaces and free volume between the primary gelatin polymer chains, thereby reducing micropores and voids within the film matrix and enhancing overall structural compactness and uniformity. The incorporation of HPMC did not significantly alter the mechanical properties of the films, indicating that the flexibility of the film was preserved regardless of HPMC content. The puncture strength of approximately 1.6 N/mm² surpassed the commercial orodispersible film benchmark (>0.06 N/mm²), confirming adequate durability for intraoral application. Elongation at break values approaching 20% reflected sufficient flexibility without brittleness, consistent with previously reported HPMC-based buccal films (7–30%). These findings are consistent with previous studies on bilayered buccal films, where mechanical robustness is balanced with flexibility to ensure patient comfort and reliable retention at the application site.

2. Mechanical Properties of Double-Layer Films Containing Different Concentrations of HPMC.

HPMC	thickness (μm)	puncture strength (N/mm2)	elongation at break (%)
0%	452.13 ± 16.71	1.63 ± 0.03	19.15 ± 1.20
0.15%	444.07 ± 11.74	1.60 ± 0.08	19.32 ± 3.07
0.45%	421.13 ± 19.51	1.63 ± 0.10	20.83 ± 3.36

Mucoadhesive Performance

The in vitro residence time refers to the maximum duration that the double-layer films remain attached to the mucosal surface under simulated oral conditions. Table illustrates the effect of HPMC concentration on the residence time and contact angle of the double-layer films. The incorporation of HPMC significantly increased the in vitro residence time. The formulation without HPMC exhibited a short residence time of 0.88 ± 0.13 h, whereas formulations containing 0.15% and 0.45% HPMC demonstrated prolonged adhesion, remaining attached for the maximum test duration of 8 h. This enhancement suggests that the inclusion of HPMC strengthens the mucoadhesive properties of the gelatin-based layer.

3. In vitro Residence Time and Contact Angle of Double-Layer Films Containing Different Concentrations of HPMC.

HPMC	residence time (h)	contact angle (°)
0%	0.88 ± 0.13	89.47 ± 1.37
0.15%	>8	70.78 ± 0.85
0.45%	>8	63.33 ± 2.71

To confirm the effect of HPMC, contact angle measurements were performed to evaluate the wettability of the mucoadhesive layer, which is indicative of the affinity between the mucoadhesive polymer and the mucosal surface. The lower contact angle signifies greater hydrophilicity and affinity to the mucosal membrane. As shown in Table , the formulation without HPMC exhibited the highest contact angle (89.47°), indicating lower wettability and spreadability, which may result from the presence of more exposed hydrophobic domains of gelatin on the film surface.

In contrast, the addition of 0.15% and 0.45% HPMC reduced the contact angle to 70.78° and 63.33°, respectively. The increased hydrophilicity of the polymer matrix contributed to lower contact angles and improved wettability. This reduction correlates with increased residence time, confirming that HPMC plays a critical role in enhancing the mucoadhesive performance of the double-layer films. Furthermore, the enhanced wettability conferred by HPMC promoted effective spreading of the polymer over mucosal tissue, facilitating intimate contact and chain diffusion into the mucin network. , Additionally, higher HPMC concentrations provided a greater number of polymer chains available for mucin interpenetration, reinforcing stronger mucoadhesive bonding via hydrogen bonding and van der Waals forces, thereby extending film residence time in the oral cavity. Although this mucoadhesion residence-time test provided a useful comparative assessment of formulation residence time, it does not fully replicate physiological oral dynamics such as continuous salivary flow, tongue movement, and swallowing; therefore, the results should be interpreted as an in vitro screening indicator rather than a direct predictor of in vivo residence time.

It should be noted that this test offers a comparative assessment of residence time but does not fully replicate oral physiological conditions such as salivary flow, tongue movement, and swallowing. The results should therefore be interpreted as an in vitro screening indicator rather than a direct predictor of in vivo performance.

In Vitro Drug Release

The in vitro release profiles of α-mangostin from the double-layer films are presented in Figure . The study demonstrated a clear difference between the solid dispersion-based formulations and the unprocessed MPE. The MPE without solid dispersion exhibited negligible drug release, less than 1% throughout the 4 h study period. This confirms the hydrophobic nature of α-mangostin and validates the necessity of the solid dispersion technique to enhance aqueous solubility for effective therapeutic delivery.

5.

In vitro release profiles of α-mangostin from double-layer films. Film formulations include MPESD (1:4) containing (■) 0.45% HPMC, (□) 0.15% HPMC, and compared to (▲) MPE.

In contrast, films loaded with MPESD showed rapid and extensive drug release. Both HPMC-containing formulations (0.15% and 0.45%) achieved approximately 96% cumulative release within 2 h. The formulation containing 0.45% HPMC exhibited a faster release rate at early time points compared to the 0.15% HPMC formulation. For instance, at 0.5 h, the 0.45% HPMC formulation released approximately 60% of the drug, whereas the 0.15% formulation released approximately 49%. This accelerated initial release observed in the higher HPMC concentration can be attributed to the hydrophilic nature of HPMC. An increased concentration of HPMC in the mucoadhesive layer promotes greater swelling upon contact with the release medium. This enhanced swelling facilitates increased water penetration into the polymer matrix, , thereby accelerating the dissolution of the α-mangostin solid dispersion. This is consistent with Suharyani et al., who reported that a higher swelling index promotes water absorption into the polymer matrix, subsequently facilitating the release of the active substance from the expanding film network. This mechanism is further supported by the wettability data, where the 0.45% HPMC formulation demonstrated a significantly lower contact angle (63°) compared to the 0.15% formulation (70°). The reduced contact angle indicates higher surface hydrophilicity, which synergizes with the swelling effect to promote faster water uptake and drug diffusion.

To investigate the drug-release mechanism, α-mangostin release from MPESD (1:4) double-layer films containing different concentrations of HPMC (0.15% and 0.45%) in artificial saliva was analyzed using zero-order, Higuchi, and Korsmeyer–Peppas kinetic models. Among these models, the Korsmeyer–Peppas model showed the best overall fit for the release data, as indicated by the highest correlation coefficient (R ²) (Table ). Nevertheless, the zero-order and Higuchi models provided additional insight into the dominant release processes.

4. Kinetic Parameters for α-Mangostin Release From MPESD (1:4) Double-Layer Films Containing Two Different HPMC Concentrations, Based on the Zero-Order, First-Order, Higuchi, and Korsmeyer–Peppas Models .

formulations	zero order		first order		higuchi		Korsmeyer–Pappas
	k 0	R2	k 1	R 2	kH	R 2	kKP	R 2	n
MPESD (1:4) + 0.15% HPMC	83.32	0.9807	1.35	0.9780	71.37	0.9253	80.57	0.9878	0.84
MPESD (1:4) + 0.45% HPMC	91.09	0.6986	1.92	0.9563	82.20	0.9807	77.52	1.0000	0.36

^aThe bold values showed the best-fitted model of kinetic drug release.

The formulation containing 0.15% HPMC showed good agreement with the zero-order model (R ² = 0.9807), suggesting a nearly constant drug-release rate associated with rapid hydration and swelling of the hydrophilic polymer matrix. In contrast, the formulation containing 0.45% HPMC was better described by the Higuchi model (R ² = 0.9807), indicating diffusion-controlled drug release through the polymer matrix.

Further interpretation using the Korsmeyer–Peppas model revealed that polymer concentration significantly influences the internal structure of the polymer matrix and the resulting release mechanism. The film containing 0.15% HPMC exhibited an n value of 0.84, indicating anomalous (non-Fickian) transport (0.45 < n < 0.89), involving both diffusion and polymer relaxation, which is consistent with the observed zero-order release behavior. This phenomenon may be attributed to the relatively loose polymer network formed at lower HPMC concentrations. Conversely, increasing the HPMC concentration to 0.45% resulted in an n value of 0.36, corresponding to Fickian diffusion and supporting the Higuchi model, suggesting diffusion-controlled drug release through a denser polymer network formed by increased polymer chain entanglement and hydrogen bonding.

Conclusions

This study successfully developed and characterized a double-layer mucoadhesive film incorporating α-mangostin solid dispersion (MPESD) for the treatment of aphthous ulcers. The implementation of solid dispersion technology using PVP K30 effectively overcame the inherent hydrophobicity of α-mangostin, enhancing its aqueous solubility from undetectable levels to approximately 8.12 mg/mL in the optimized 1:4 MPE-to-polymer ratio. The bilayer design achieved distinct layer-specific functionality: the ethylcellulose backing layer provided a protective barrier against oral irritants, while the gelatin-based mucoadhesive layer ensured prolonged retention at the ulcer sites, combining enhanced solubility, sustained mucoadhesion, and rapid targeted drug release. Mechanical testing confirmed that all formulations possessed suitable puncture strength and flexibility for intraoral application. Incorporation of HPMC into the mucoadhesive layer markedly enhanced overall film performance, with HPMC content serving as a key parameter in modulating both residence time and drug release rate while preserving mechanical integrity. Furthermore, the films demonstrated a rapid drug release profile, achieving approximately 96% α-mangostin release within 2 h. These findings suggest that the developed α-mangostin-loaded double-layer film provides a highly promising localized delivery platform for the potential treatment of aphthous ulcers.

The authors declare no competing financial interest.