β-Cyclodextrin–chitosan based carriers enhance the solubility and dissolution of baicalein through inclusion complex formation

Abstract

Baicalein (BA) is a poorly water-soluble flavonoid whose limited aqueous solubility restricts its oral bioavailability and pharmaceutical performance. This study aimed to enhance BA solubility and dissolution behavior through the development of binary (BA–β-cyclodextrin) and ternary (BA–β-cyclodextrin–chitosan) inclusion systems. Solid-state characterization was performed using DSC, FTIR and XRD analysis, while thermodynamic evaluation was conducted through Higuchi–Connors phase-solubility analysis. Molecular docking and MM-GBSA calculations were employed to support host–guest interaction mechanisms. Equilibrium solubility studies revealed that pure BA exhibited an intrinsic solubility of 7.50 µg/mL, whereas the optimized 1:2 BA:β-CD binary complex achieved 65.99 µg/mL, corresponding to approximately 8.8-fold enhancement. The phase-solubility diagram displayed an AL-type profile with a calculated apparent stability constant (K_s) of 5.3 × 10² M⁻¹, confirming predominant 1:1 inclusion complex formation. Dissolution studies further demonstrated significantly improved cumulative drug release for binary systems and sustained-release behaviour for ternary complexes. These findings provide integrated thermodynamic, physicochemical, and computational evidence supporting cyclodextrin-assisted solubilization of baicalein and highlight the potential of combined inclusion–polymer systems for improving the pharmaceutical performance of poorly soluble flavonoids.

Introduction

Baicalein (BA; 5,6,7-trihydroxyflavone) (Fig. 1A) is the aglycone part of baicalin (7-D-Glucuronic acid-5,6-dihydroxyflavone) (Fig. 1B), a naturally occurring flavonoid isolated primarily from Scutellaria baicalensis and Oroxylum indicum^1,2. BA has emerged as a promising candidate owing to its broad spectrum of anticancer activities through modulation of several key oncogenic signalling pathways, including PI3K/Akt, MAPK, and NF-κB cascades, thereby regulating apoptosis, cell cycle progression, proliferation, and angiogenesis³. These effects have been documented across multiple cancer types, particularly lung, colon, and breast cancers^4–6. In addition, BA plays a role in suppressing tumour initiation and progression through its antioxidant and anti-inflammatory activities⁷.

Fig. 1.

Chemical structures of (A) Baicalein (5,6,7-trihydroxyflavone), (B) Baicalin (7-D-Glucuronic acid-5,6-dihydroxyflavone), (C) β-cyclodextrin, and (D) chitosan.

Despite BA’s potent pharmacological profile, its clinical translation remains limited by unfavourable pharmacokinetics. BA exhibits low aqueous solubility¹⁵, a low dissolution rate¹⁶, low membrane permeability¹⁷, extensive first-pass metabolism (primarily glucuronidation and sulfation), and rapid systemic elimination¹⁸. Consequently, BA shows poor oral bioavailability, reported to be approximately 13% to 23%¹⁹, thereby limiting its therapeutic activity and necessitating the development of enhanced delivery strategies.

To overcome such limitations, pharmaceutical research has increasingly focused on improving the pharmacokinetic profile through nanotechnology-enabled and polymer-based delivery systems^8,9. Among the most promising approaches are β-cyclodextrin (β-CD) inclusion complexation (Fig. 1C) and chitosan-based polymeric delivery (Fig. 1D), which offer synergistic, biocompatible solutions^4–6.

β-Cyclodextrins are cyclic oligosaccharide with a hydrophilic outer surface and a relatively small hydrophobic internal cavity (Fig. 1). β-Cyclodextrin (β-CD) was selected as the primary inclusion agent due to its hydrophobic cavity structure, which enables encapsulation of poorly soluble drug molecules through host–guest interactions⁵. β-CD has been widely used in pharmaceutical formulations to enhance aqueous solubility, dissolution rate, and stability of hydrophobic compounds. Its regulatory acceptance and well-established safety profile further support its suitability for oral delivery applications. Several studies have shown that β-CD can enhance aqueous solubility and dissolution, reduce crystallinity, and protect flavonoids from oxidative degradation^10,11. BA can form an inclusion complex by accommodating within the β-CD cavity, creating a stabilising hydrophobic environment. This complexation may improve dissolution kinetics and solubility, which may lead to enhanced BA’s cytoprotective and antioxidant effects^12,13.

Chitosan was incorporated as a secondary polymeric component due to its biocompatibility, mucoadhesive properties, and ability to modulate drug release kinetics. The combination of cyclodextrin inclusion with polymeric stabilisation offers potential advantages in balancing solubility enhancement with controlled-release behaviour¹⁴. Together, these platforms represent a complementary strategy to address BA’s solubility problem.

In this study, the aim is to develop and characterise a β-CD-based inclusion complex of BA followed by a chitosan-based polymeric coating to overcome BA’s inherent solubility and dissolution limitations and enhance formulation performance. Molecular docking analysis was additionally performed to probe the host–guest inclusion of BA within the β-CD cavity and to provide mechanistic support for the experimental formulation strategy. Although β-cyclodextrin-based systems for baicalein delivery have been previously reported, the present study provides a comprehensive and integrative investigation that extends beyond conventional formulation approaches. In contrast to earlier studies that primarily focused on preparation and basic characterization, this work systematically combines solid-state analysis (DSC, FTIR, X-ray), quantitative equilibrium solubility assessment, full Higuchi–Connors phase-solubility analysis with determination of the apparent stability constant (K_s), dissolution performance evaluation, and molecular modeling (docking and MM-GBSA free energy analysis) within a unified experimental framework. Moreover, both binary (BA–β-CD) and ternary (BA–β-CD–chitosan) systems were comparatively developed and mechanistically correlated, allowing direct evaluation of polymer-mediated modulation of inclusion behaviour and release kinetics. This integrated thermodynamic–computational–performance-based approach provides deeper mechanistic insight into baicalein complexation and release behavior, thereby offering a more comprehensive understanding of cyclodextrin-assisted solubilization strategies for poorly soluble flavonoids.

Therefore, the objective of this study was to develop and systematically characterize binary (BA–β-CD) and ternary (BA–β-CD–chitosan) inclusion complexes to enhance aqueous solubility and dissolution performance of baicalein. A comprehensive approach integrating solid-state characterization, quantitative solubility analysis, Higuchi–Connors phase-solubility evaluation, dissolution profiling, and molecular docking was employed to provide mechanistic and thermodynamic validation of complex formation.

Materials and methods

Material

BA (≥ 98% purity), β-cyclodextrin (β-CD, pharmaceutical grade), and chitosan (medium molecular weight, degree of deacetylation ≥ 75%) were purchased from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany). Ethanol (analytical grade), formic acid acid, and sodium hydroxide were also obtained from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany). Purified water was used throughout the study.

Method

Molecular docking and binding energy calculations

Molecular modeling was used to provide qualitative structural support for host–guest inclusion and to rationalize experimental trends; it is not intended as a quantitative predictor of polymeric complex stability or in vivo performance¹⁵. Given the macromolecular nature of chitosan, a representative oligomeric fragment was modelled to approximate local interaction behaviour. This approach enables qualitative evaluation of binding orientation and interaction energy while acknowledging that full polymeric dynamics cannot be completely captured by static docking simulations.

Ligand preparation

The chemical structure of the ligands was prepared using the LigPrep module (Schrödinger LLC, USA). Ionization states and possible tautomers were generated using Epik at a target pH of 6.0 ± 1.0, to match the experimental conditions. The ligand structures were minimised using the OPLS4 force field before docking.

Binary docking (baicalein–β-cyclodextrins)

For binary docking studies, β-CD was used as the host. Receptor structures were prepared using the Structure Preparation workflow. Final minimization was performed using MacroModel with the OPLS4 force field¹⁶. Molecular docking of BA into β-CD was performed using Glide in standard precision (SP) mode. A receptor grid was generated by centering the grid box on the β-CD cavity, with dimensions sufficient to accommodate the ligand (12 Å). After docking, the top-ranked pose was selected for analysis based on GlideScore and visual inspection of key interactions and geometric stability.

Ternary docking (baicalein–β-cyclodextrins–chitosan)

For ternary docking, a short oligomer of chitosan (6-monomer unit) was prepared to represent the polymeric receptor in the ternary system. The BA-β-CD generated inclusion complexes were assembled, and the inclusion geometry was retained (frozen); the complex was then treated as a single ligand and docked onto the chitosan oligomer. A receptor grid was generated on the chitosan structure to capture surface-level interactions. Docking was performed using Glide SP mode with the input ligand conformation preserved to maintain the inclusion geometry. The best docking poses were selected based on GlideScore and visual inspection of key interactions and geometric stability.

MM-GBSA binding free energy calculations

Prime MM-GBSA calculations were performed on the selected complexes (binary and ternary) to further validate the docking results and estimate binding free energies. The MM-GBSA method was applied using the VSGB implicit solvation model and OPLS4 force field. The binding free energy (ΔG_bind) was calculated based on the equation below:

1

For the binary system, BA was defined as the ligand and β-CD as the receptor. For the ternary system, BA-β-CD was defined as the ligand and chitosan as the receptor. MM-GBSA results were used to evaluate relative stability in the binary and ternary systems. Although molecular docking and MM-GBSA calculations provide valuable insight into binding orientation and interaction energetics, it is important to recognise their inherent limitations. Docking simulations represent static approximations of molecular interactions and do not fully capture dynamic solution behaviour, polymer flexibility, solvent effects, or physiological complexity. Furthermore, binding energy predictions do not directly translate into in-vivo pharmacokinetic outcomes such as absorption, distribution, metabolism, or bioavailability. Therefore, the computational findings should be interpreted as supportive mechanistic evidence that complements experimental characterisation rather than as definitive predictors of biological performance.

Complex formation

Preparation of BA–β-CD inclusion complexes: BA–β-CD inclusion complexes were prepared at BA:β-CD molar ratios of 1:1, 1:2 and 1:3 using the co-evaporation method. Briefly, β-CD was dissolved into purified water, while BA was dissolved into ethanol¹⁶. The BA solution was added dropwise to the β-CD solution under continuous stirring at 40 °C and maintained for 2 h to allow complexation. The organic solvent was removed under reduced pressure, and the resulting dispersion was freeze-dried or vacuum-dried to obtain a solid inclusion complex, which was gently pulverized and stored in a desiccator until further analysis¹⁷.

Preparation of BA-β-CD-chitosan ternary complexes: Ternary complexes were prepared by coating pre-formed BA-β-CD inclusion complexes with chitosan. A chitosan solution containing 0.1–0.2% (w/v) chitosan at pH 5.5 was prepared by dissolving chitosan in 1% (v/v) aqueous acetic acid. The BA-β-CD complexes were slowly added to the chitosan solution under agitation and further stirred for 4 h. The resulting ternary complexes were freeze-dried to obtain BA-β-CD-chitosan powders for all three β-CD ratios. The preparation of binary and ternary complexes is depicted in Fig. 2¹⁸.

Fig. 2.

Schematic illustration of the multistep preparation process for binary (A) and ternary complexes (B).

The selected BA:β-CD molar ratios (1:1, 1:2, and 1:3) were chosen based on commonly reported stoichiometric inclusion patterns in cyclodextrin systems and to systematically investigate the influence of increasing host concentration on complex formation efficiency and dissolution behaviour¹⁹. The 1:1 ratio represents the typical stoichiometry for hydrophobic guest inclusion within the β-CD cavity, while higher β-CD ratios (1:2 and 1:3) were evaluated to determine whether excess host availability could further enhance solubility and dissolution performance⁵.

Fourier transform infrared (FTIR) spectroscopy

FTIR analysis was performed to investigate potential intermolecular interactions and confirm the formation of complexes. FTIR spectra of pure BA, β-CD, pure chitosan, and the binary (BA-β-CD) and ternary (BA-β-CD-chitosan) complex systems were recorded using FTIR spectrophotometer (Shimadzu Corporation, Kyoto, Japan). For FTIR analysis, samples were ground to a fine powder and analysed using the KBr pellet method. Each spectrum was scanned in the range of 4000–400 cm⁻¹ with a resolution of 4 cm⁻¹ and 32 scans. The resulting FTIR spectra were interpreted by comparing characteristic bands and identifying peak shifts, intensity changes, broadening, or disappearance of bands, which may indicate complex formation and intermolecular interactions²⁰.

Differential scanning calorimetry (DSC)

Thermal behaviour and physical state changes of BA, β-CD, chitosan, and their corresponding binary and ternary complexes were evaluated using differential scanning calorimetry (DSC). Measurements were carried out using a Mettler Toledo DSC instrument (Mettler-Toledo International Inc., Greifensee, Switzerland). Accurately weighed samples (approximately 3–5 mg) were placed in standard aluminium pans and hermetically sealed, with an empty aluminium pan used as a reference. The samples were heated over a temperature range of 25–450 °C at a heating rate of 10 °C/min under a nitrogen purge atmosphere. Thermograms were analysed to identify changes in melting endotherms, peak shifts, or disappearance of characteristic peaks, which were considered evidence of successful complex formation and changes in crystallinity²⁰.

Powder X-ray diffraction (PXRD) analysis

Powder X-ray diffraction (PXRD) analysis was performed to investigate the crystalline characteristics and solid-state transformation of baicalein (BA), physical mixtures, binary inclusion complexes (A, B, C), and ternary systems (1, 2, 3).

Diffraction patterns were recorded using an X-ray diffractometer equipped with Cu Kα radiation (λ = 1.5406 Å), operated at 40 kV and 30 mA. Samples were gently ground and uniformly spread on a sample holder to obtain a smooth surface. Data were collected over a 2θ range of 5°–70° with a step size of 0.02° and a scanning rate of 2°/min.

All measurements were conducted at room temperature. Diffractograms were analyzed to compare peak positions, intensities, and crystallinity changes among pure BA, physical mixtures, and prepared complexes. Reduction in peak intensity, peak broadening, or disappearance of characteristic reflections was interpreted as evidence of reduced crystallinity or possible inclusion complex formation.

HPLC method

The quantification of baicalein was performed using a validated HPLC method as previously described by Malkawi et al. (submitted for publication)²¹. Chromatographic separation was achieved using a Diamonsil C18 column (150 mm × 4.6 mm, 5 μm) maintained at 40 °C. The mobile phase consisted of 1% formic acid in water and acetonitrile (60:40, v/v), pumped at an isocratic flow rate of 1.0 mL/min. The injection volume was set at 35 µL, and UV detection was performed at 271 nm.

The method was validated in accordance with ICH Q2(R1) guidelines. Linearity was established over a concentration range of 1–50 µg/mL with a correlation coefficient (R^2) of 0.9993. The method demonstrated high accuracy, with recovery rates ranging from 98.72% to 103.81% and a relative standard deviation (RSD) below 4%. Precision was confirmed with intra-day RSD values between 0.53% and 3.29% and inter-day RSD values between 0.23% and 1.60%. Specificity results confirmed no interference from formulation excipients such as chitosan and β-cyclodextrin.

In vitro dissolution study

In vitro dissolution studies were conducted to evaluate the release behaviour of pure BA, BA–β-CD) inclusion complexes, and BA-β-CD-chitosan ternary complexes. Dissolution testing was performed using a USP Apparatus II (paddle method). An accurately weighed amount of each formulation, equivalent to the same BA dose, was placed in 900 mL of dissolution medium maintained at 37 ± 0.5 °C and stirred at a paddle rotation speed of 100 rpm. The dissolution medium consisted of phosphate buffer pH 6.8, selected to simulate intestinal conditions and ensure sink conditions^22,23. Phosphate buffer pH 6.8 was selected as the dissolution medium to simulate intestinal conditions, where baicalein is primarily absorbed following oral administration. Since baicalein is a weakly acidic flavonoid with limited solubility in aqueous media, evaluation under near-neutral intestinal pH provides relevant insight into its dissolution-limited absorption behaviour. The use of 900 mL medium volume in USP Apparatus II ensured maintenance of sink conditions throughout the experiment and minimised saturation effects.

At predetermined time intervals (0, 10, 20, 30, 40, 50, and 60 min), 3 mL samples were withdrawn and immediately replaced with an equal volume of fresh, pre-warmed dissolution medium to maintain a constant volume. The withdrawn samples were filtered through a 0.45 μm membrane filter, suitably diluted if necessary, and analysed for BA content using HPLC at the predetermined λmax of 271 nm. The cumulative percentage of BA released was calculated and plotted as a function of time. Dissolution profiles of the different formulations were compared to assess the effect of β-CD inclusion and chitosan incorporation on drug release behaviour, including enhancement and sustained-release characteristics^24,25. Statistical analysis was performed using one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test to compare dissolution performance between formulations at selected time points. Differences were considered statistically significant at p < 0.05. All experiments were conducted in triplicate, and data are presented as mean ± standard deviation (SD).

Solubility

The equilibrium solubility of pure baicalein (BA), BA–β-cyclodextrin (β-CD) binary inclusion complexes (Complexes A, B, and C), and BA–β-CD–chitosan ternary complexes (Complexes 1, 2, and 3) was determined using the shake-flask method. An excess amount of each formulation was accurately weighed and transferred into glass vials containing 1 mL of purified water. The suspensions were vortex-mixed and shaken at 37 ± 1 °C for 24 h to ensure equilibrium was achieved²⁶.

Excess conditions were confirmed by the presence of visible undissolved solid remaining at the bottom of the vials after 24 h equilibration, indicating that saturation had been reached. Moreover, the calculated dissolved amount (≤ 0.07 mg) was substantially lower than the initial mass added, further confirming that the system remained saturated throughout the experiment⁸.

Following equilibration, samples were centrifuged at 10,000 rpm for 10 min, and the supernatant was carefully collected and filtered through a 0.45 μm membrane filter to remove undissolved particles. The concentration of dissolved BA in the filtrate was quantified using the previously validated HPLC method. All experiments were conducted in triplicate, and data are presented as mean ± standard deviation (SD).

Phase-solubility study (Higuchi–Connors method)

The phase-solubility behavior of baicalein (BA) with β-cyclodextrin (β-CD) was performed according to the Higuchi–Connors method. β-CD aqueous solutions were prepared at increasing concentrations (0–12 mM) in purified water. An excess amount of pure BA was added to each vial to ensure saturation conditions. The suspensions were vortex-mixed and shaken at 37 ± 0.5 °C for 24 h to allow equilibrium to be reached.

After equilibration, samples were centrifuged at 10,000 rpm for 10 min, and the supernatants were filtered through a 0.45 μm membrane filter to remove undissolved drug particles. The concentration of dissolved BA was determined using the validated HPLC method described previously.

The BA concentration was calculated using the calibration Eq. (2):

2

where x represents BA concentration (µg/mL) and A is the chromatographic peak area (AUC).

Phase-solubility diagrams were created by plotting the equilibrium solubility of BA (µg/mL) against β-CD concentration (mM). The type of solubility profile (AL, AP, or AN) was verified based on the shape of the plot. For linear AL-type diagrams, the apparent stability constant (K_s) was calculated from the slope of the linear portion of the plot using the Higuchi–Connors Eq. (3):

3

where S₀ stand for the intrinsic solubility of BA in the absence of β-CD, and slope is the linear regression slope obtained from the phase-solubility plot. All experiments were performed in triplicate, and results were expressed as mean ± standard deviation (SD).

Drug content assay

Drug content of baicalein (BA) in the prepared binary BA–β-CD inclusion complexes and ternary BA–β-CD–chitosan complexes was determined using the validated HPLC method described in Sect. 2.2.6. An accurately weighed amount of each dried formulation was transferred to a volumetric flask, and BA was extracted using ethanol under sonication for 10–15 min to ensure complete drug extraction. The solution was centrifuged, and the supernatant was filtered through a 0.45 μm membrane filter before HPLC injection. BA concentration was calculated from the BA calibration curve. Drug content was expressed as percentage assay/label claim according to the following equation, adapted from the extraction-based assay approach reported by Devangan et al.²⁷.

where the amount of BA recovered represents the experimentally quantified BA amount obtained from the HPLC calibration curve, and the theoretical BA amount represents the expected BA content based on the formulation composition.

All measurements were performed in triplicate and reported as mean ± SD (n = 3).

Results and discussion

Differential scanning calorimetry (DSC)

The DSC thermograms of pure BA, β-CD, their physical mixture, and the prepared BA–β-CD inclusion complexes (Complexes A, B &C) and (Complexes 1, 2 & 3) are presented in Figs. 3 and 4. Pure BA displayed a sharp endothermic melting peak at 275 °C, corresponding to its crystalline nature. In contrast, β-CD showed a broad endothermic event in the range of 80–120 °C, attributed to the loss of bound water molecules, which is typical for β-cyclodextrins²⁸.

Fig. 3.

DSC thermograms of pure baicalein, β-cyclodextrin, the physical mixture, and the binary inclusion complexes (Complex A:1:1, Complex B: 1:2, Complex C: 1:3).

Fig. 4.

DSC thermograms of pure baicalein, β-cyclodextrin, chitosan, physical mixture, and the ternary complexes (Complex 1: 1:1, Complex 2: 1:2, Complex 3: 1:3).

The physical mixture presented the characteristic thermal events of both BA and β-CD, with the BA melting endotherm remaining clearly detectable, indicating the absence of significant solid-state interaction upon simple mixing. However, notable changes were detected in the thermograms of the prepared inclusion complexes. In the complexes, the sharp melting endotherm of BA was either markedly reduced in intensity or completely absent, accompanied by broadening of the thermal profile²⁹.

The disappearance and attenuation of the characteristic BA melting peak in the inclusion complexes indicate reduced crystallinity and successful molecular inclusion of BA within the β-CD cavity. This effect became more pronounced with increasing β-CD content, suggesting more efficient complex formation at higher molar ratios. Overall, the DSC findings provide strong evidence for the formation of true BA–β-CD inclusion complexes rather than simple physical mixtures, supporting solid-state interaction between BA and β-CD²⁹. The disappearance or significant reduction of the characteristic melting endotherm of crystalline baicalein suggests loss of long-range molecular order and transition toward an amorphous or molecularly dispersed state within the β-CD matrix. This behaviour is consistent with host–guest inclusion complex formation, where baicalein molecules become encapsulated within the hydrophobic cavity of β-CD, thereby disrupting crystalline lattice packing. Such amorphization or decrease in crystallinity contributes directly to enhanced solubility and dissolution performance by reducing lattice energy barriers to dissolution. Similar findings have been reported for polymeric and cyclodextrin-based systems, where amorphization enhances solubility and dissolution performance by improving molecular mobility and wettability³⁰.

Fourier transform infrared (FTIR) spectroscopy

The FTIR spectra of pure BA, β-CD, their physical mixture, and the BA–β-CD inclusion complexes (Complexes A, B, and C) are presented in Fig. 5. Pure BA exhibited characteristic absorption bands corresponding to phenolic O–H stretching in the region of 3200–3500 cm⁻¹, aromatic C = O stretching near 1650 cm⁻¹, and aromatic C = C vibrations in the range of 1500–1600 cm⁻¹³⁰. β-CD showed a broad O–H stretching band around 3300–3400 cm⁻¹, along with characteristic C–O–C and C–O stretching vibrations in the 1000–1150 cm⁻¹ region³¹. The physical mixture retained the principal absorption bands of both BA and β-CD with minimal shifts, indicating the absence of significant chemical interaction upon simple blending. In contrast, the FTIR spectra of Complexes A, B, and C demonstrated noticeable changes in the characteristic BA bands. Specifically, the intensity of BA’s carbonyl and aromatic bands was markedly reduced or broadened, accompanied by slight shifts in the O–H stretching region.

Fig. 5.

FTIR spectra for pure baicalein, β-CD, the physical mixture, and the binary inclusion complexes (Complex A: 1:1, Complex B: 1:2, Complex C:1:3).

These spectral modifications suggest the involvement of hydrogen bonding and host–guest interactions between BA and the β-CD cavity, confirming the successful formation of inclusion complexes. The extent of band attenuation became more pronounced with increasing β-CD content, indicating enhanced inclusion efficiency at higher molar ratios.

The FTIR spectra of the ternary complexes (Complexes 1, 2, and 3), along with pure chitosan and the corresponding physical mixture, are shown in Fig. 6. Chitosan exhibited characteristic absorption bands, including a broad O–H and N–H stretching band around 3200–3500 cm⁻¹, amide I (C = O stretching) near 1650 cm⁻¹, and amide II (N–H bending) around 1550 cm⁻¹. In the spectra of the ternary complexes, the characteristic absorption bands of BA were further attenuated or became indistinguishable, while noticeable broadening and shifting of the O–H/N–H stretching region were observed. Additionally, changes in the amide I and amide II bands of chitosan were evident, indicating intermolecular interactions between chitosan and the BA–β-CD inclusion complex.

Fig. 6.

FTIR spectra of pure baicalein, β-CD, chitosan, the physical mixture, and the ternary complexes (Complex 1: 1:1, Complex 2: 1:2, Complex 3: 1:3).

Compared with the binary systems, the ternary complexes exhibited more pronounced band broadening and reduced peak definition, suggesting stronger hydrogen bonding and the formation of a polymer-coated inclusion system. These findings confirm that chitosan was successfully incorporated onto the BA–β-CD complexes, forming a stable ternary system without evidence of drug degradation. The observed shifts and broadening of hydroxyl and carbonyl stretching bands indicate intermolecular hydrogen bonding interactions between baicalein and β-CD. The attenuation, shifting, and broadening of characteristic baicalein peaks in the complex systems further confirm the formation of intermolecular interactions, particularly hydrogen bonding between baicalein and the carrier components. Such interactions are known to stabilize the drug within the carrier matrix and reduce drug crystallinity. Previous studies have demonstrated that the disappearance or weakening of drug-specific FTIR bands is indicative of successful encapsulation or inclusion complex formation, rather than simple physical mixing³⁰.

Powder X-ray diffraction (PXRD) analysis

The diffractogram of pure BA exhibited numerous sharp and intense diffraction peaks throughout the scanned 2θ range, confirming its highly crystalline nature Fig. 7. Prominent reflections were observed within the 10–30° region, characteristic of the ordered crystalline lattice of baicalein. The high intensity and sharpness of these peaks indicate long-range molecular order and strong crystal lattice packing, which are consistent with its limited aqueous solubility observed in equilibrium solubility studies. The PXRD pattern of the physical mixture Fig. 7 retained the characteristic diffraction peaks of pure BA with only slight reduction in intensity^32,33. No significant peak shifting or disappearance was observed. This indicates that simple blending of BA with β-CD (and chitosan where applicable) did not induce structural transformation or inclusion complex formation. The persistence of crystalline reflections confirms the absence of strong solid-state interactions in the physical mixture.

Fig. 7.

Powder X-ray diffraction (PXRD) patterns of (A) pure baicalein (BA) and (B) physical mixture.

The binary systems exhibited noticeable reduction in peak intensity compared with pure BA Fig. 8. Partial broadening of characteristic BA peaks was observed, suggesting disruption of crystalline order. Among the binary systems, Complex B (1:2 BA:β-CD) demonstrated the greatest reduction in diffraction intensity, indicating more effective interaction between BA and β-CD at this molar ratio. However, residual crystalline peaks were still detectable in all binary formulations, suggesting incomplete amorphization and partial inclusion complex formation. The attenuation and broadening of diffraction peaks indicate reduction in crystallinity, likely resulting from encapsulation of BA within the hydrophobic cavity of β-CD. Such host–guest interactions disrupt the regular crystal lattice of BA and reduce long-range molecular order.

Fig. 8.

Powder X-ray diffraction (PXRD) of the binary complexes.

The ternary systems displayed a more pronounced reduction in crystallinity compared with the binary complexes Fig. 9. Diffraction peaks became significantly broadened, and several characteristic BA reflections were either markedly diminished or nearly absent. In particular, Complex 3 showed a predominantly diffuse halo pattern with minimal sharp peaks, indicating substantial amorphous character. The enhanced reduction in crystallinity in ternary systems may be attributed to the combined effect of cyclodextrin inclusion and additional intermolecular interactions with chitosan. Chitosan likely contributes to: Further disruption of BA crystal packing, hydrogen bonding interactions, inhibition of recrystallization. This synergistic interaction promotes greater molecular dispersion of BA within the matrix. The PXRD results provide solid-state evidence supporting host–guest inclusion between BA and β-CD. Reduction in diffraction intensity without formation of new crystalline peaks suggests molecular dispersion rather than formation of a new crystalline phase. The additional amorphization observed in ternary systems further supports the role of polymer-assisted stabilization.

Fig. 9.

Powder X-ray diffraction (PXRD) of the ternary complexes.

Collectively, PXRD analysis confirms progressive disruption of the crystalline structure of BA upon complexation, consistent with thermodynamic (phase-solubility) and dissolution data.

In vitro dissolution

The in vitro dissolution profiles of pure BA, BA–β-CD binary inclusion complexes, and BA–β-CD–chitosan ternary complexes are illustrated in Figs. 10 and 11.

Fig. 10.

In vitro dissolution profiles of pure BA and BA–β-CD binary inclusion complexes (Complex A: 1:1, Complex B: 1:2, Complex C: 1:3) in phosphate buffer pH 6.8, Error bars representing mean ± SD (n = 3).

Fig. 11.

In vitro dissolution profiles of pure BA and BA–β-CD–chitosan ternary complexes (Complex 1: 1:1, Complex 2: 1:2, Complex 3: 1:3) in phosphate buffer pH 6.8, error bars representing mean ± SD (n = 3).

Pure BA exhibited very limited dissolution throughout the study period, confirming its poor aqueous solubility and dissolution-limited behaviour. In contrast, all binary BA–β-CD complexes showed a pronounced enhancement in dissolution rate and extent compared with the pure drug (p < 0.05) (Fig. 10). Among the binary systems, Complex B (1:2 BA:β-CD) demonstrated significantly higher (p < 0.05) All ternary formulations showed higher drug release than the pure drug dissolution, followed by Complex C (1:3) and Complex A (1:1). The rapid increase in dissolved drug concentration during the initial phase (0–30 min) is attributed to improved wettability, partial amorphization, and host–guest inclusion of BA within the β-CD cavity³⁴.

However, the binary complexes exhibited a relatively fast release pattern, with peak concentrations reached at approximately 30–40 min, followed by a gradual decline. This behaviour indicates enhanced drug availability but limited control over sustained release.

In contrast, the ternary BA–β-CD–chitosan complexes displayed a distinct and more controlled dissolution behaviour (Fig. 11). All ternary formulations showed significantly higher drug release than the pure drug (p < 0.05) and a smoother release profile over time. The ternary systems exhibited an initial release phase followed by a prolonged plateau, indicating sustained-release characteristics imparted by the chitosan coating. Among these, Complex C (1:3 BA:β-CD) achieved the highest overall release, while Complexes A and B demonstrated comparable controlled-release behavior. The dissolution behaviour is consistent with the intrinsic solubility value (7.50 µg/mL) determined in the equilibrium solubility study, confirming dissolution-limited drug performance.

The improved dissolution and sustained release observed in the ternary systems can be attributed to the synergistic effect of β-CD-mediated solubilization and chitosan-induced diffusion control. Chitosan likely formed a polymeric barrier around the inclusion complex, moderating drug diffusion into the dissolution medium and preventing rapid depletion.

The enhanced dissolution behavior observed in the binary and ternary systems can be attributed to multiple synergistic mechanisms, including reduced crystallinity, improved wettability, and increased surface area resulting from inclusion complex formation. In addition, polymer incorporation contributes to controlled and sustained drug release by forming a diffusion barrier around the drug molecules. Similar biphasic or enhanced release profiles have been reported in polymer-based drug delivery systems, where an initial rapid release is followed by a sustained release phase governed by diffusion and matrix relaxation mechanisms³⁵.

Solubility

The equilibrium solubility of pure baicalein (BA), BA–β-cyclodextrin (β-CD) binary inclusion complexes (A–C), and BA–β-CD–chitosan ternary complexes (1–3) was assessed using the shake-flask method (24 h, 1 mL purified water). The calculated solubility results are summarized in Table 1.

Table 1.

Equilibrium solubility of BA and prepared complexes (1 mL Water, 24 h).

Sample	Solubility (µg/ml)	± SD (n = 3)
Pure baicalein	7.50	0.0875999
Complex A (1:1)	63.02	0.076589
Complex B (1:2)	65.99	0.06570
Complex C (1:3)	13.38	0.011246
Complex 1 (1:1)	58.48	0.076539
Complex 2 (1:2)	8.49	0.023850
Complex 3 (1:3)	21.37	0.038975

Pure BA showed very low aqueous solubility (7.50 µg/mL), proving its intrinsic hydrophobicity and dissolution-limited behaviour. In contrast, all prepared complexes exhibited significantly enhanced solubility relative to the pure drug (p < 0.05), indicating successful molecular interaction and structural modification.

Among the binary systems, complex B (1:2 BA:β-CD) illustrated the highest solubility (65.99 µg/mL), followed closely by complex A (1:1; 63.02 µg/mL), while complex C (1:3) displayed comparatively lower solubility (13.38 µg/mL). The pronounced enhancement detected for complex B indicates that the 1:2 molar ratio offers an optimal balance between inclusion efficiency and drug loading. At this ratio, BA is efficiently accommodated within the hydrophobic β-CD cavity, resulting in improved aqueous dispersion and reduced crystallinity, consistent with the DSC and FTIR findings.

The comparatively lower solubility detected for the 1:3 system may be attributed to excessive cyclodextrin presence, potentially leading to aggregation phenomena or formation of less efficient inclusion assemblies, which may limit effective drug release into the aqueous phase³².

The ternary BA–β-CD–chitosan systems also demonstrated improved solubility compared with pure BA. Complex 1 (1:1 + chitosan) demonstrated the highest solubility among ternary formulations (58.48 µg/mL), followed by Complex 3 (21.37 µg/mL) and Complex 2 (8.49 µg/mL). The incorporation of chitosan likely contributes additional hydrogen bonding and electrostatic interactions, as supported by the docking and MM-GBSA analysis. The improvement in BA solubility can be mechanistically attributed to multiple complementary factors. First, β-CD inclusion reduces the exposure of the hydrophobic BA structure to the aqueous environment by accommodating the aromatic scaffold within its hydrophobic cavity, while the hydrophilic exterior of β-CD improves apparent aqueous dispersion. Second, DSC and PXRD findings indicate reduced crystallinity and partial amorphization of BA after complex formation, which lowers the crystal lattice energy required for dissolution. This explanation is consistent with recent solubility-enhancement literature, where incorporation of a poorly water-soluble drug into a carrier matrix converted the drug from a crystalline to an amorphous state and produced marked enhancement in equilibrium solubility and dissolution³⁶. Therefore, in the present study, the higher solubility of Complex B is likely due to an optimal balance between host–guest inclusion, reduced crystallinity, improved wettability, and molecular dispersion of BA within the β-CD matrix.

However, polymer association may also influence diffusion behaviour and equilibrium partitioning, explaining the differences observed between binary and ternary rankings. Essentially, the equilibrium solubility trend for the binary complexes is consistent with the previously reported dissolution results, where the 1:2 BA:β-CD system demonstrated superior dissolution performance^8,9. This correlation confirms that enhanced inclusion efficiency directly contributes to increased aqueous availability. Although the ternary systems exhibited sustained-release characteristics in dissolution testing, their equilibrium solubility values reflect the combined effects of molecular inclusion and polymer-mediated interaction.

Overall, the solubility findings reinforce the conclusion that β-CD inclusion significantly improves BA aqueous solubility, with the 1:2 molar ratio representing the most effective binary formulation. The addition of chitosan modulates system behaviour, contributing to controlled-release characteristics while maintaining improved solubility relative to the pure drug.

Phase-solubility analysis

Phase-solubility analysis of baicalein (BA) with β-cyclodextrin (β-CD) was conducted according to the Higuchi–Connors method. The solubility of BA increased linearly with increasing β-CD concentration over the investigated range (0–12 mM), producing an AL-type diagram (Fig. 12).

Fig. 12.

Higuchi–Connors phase-solubility diagram of baicalein (BA) in the presence of increasing concentrations of β-cyclodextrin (β-CD) (0–12 mM) at 37 ± 0.5 °C after 24 h equilibration. The linear increase in BA solubility indicates an AL-type profile, suggesting formation of a predominant 1:1 inclusion complex. The regression equation was \:y=3.9732x+7.5179, with the y-intercept corresponding to the intrinsic solubility of BA in water, error bars representing mean ± SD (n = 3).

where y represents BA solubility (µg/mL) and x represents β-CD concentration (mM). The regression demonstrated excellent linearity (R² > 0.99), confirming first-order dependence of BA solubility on β-CD concentration.

The intrinsic solubility of BA (S₀), determined from the y-intercept, was 7.52 µg/mL. The linear AL-type profile indicates predominant formation of a 1:1 inclusion complex between BA and β-CD.

The apparent stability constant (K_s), calculated using the Higuchi–Connors equation, was found to be approximately:

This moderate stability constant indicates thermodynamically favourable complex formation while maintaining sufficient reversibility to support drug release. The obtained K_s value falls within the typical range reported for flavonoid–β-CD inclusion systems and is consistent with the enhanced solubility and dissolution behaviour observed for the binary complexes.

Overall, the phase-solubility findings provide quantitative confirmation of host–guest inclusion complexation and strongly support the solid-state (DSC, FTIR) and molecular docking results presented in this study.

Practical yield and drug content (assay)

Practical yield and baicalein drug content of the prepared binary BA–β-CD inclusion complexes and ternary BA–β-CD–chitosan complexes were determined to support formulation reproducibility and batch quality.

Drug content was quantified using the validated HPLC method described in Sect.  2.2.6, and results are expressed as percentage of label claim (mean ± SD, n = 3).

As shown in Table 2.

Table 2.

Practical yield (%) and baicalein drug content (HPLC assay, % of label claim) of the prepared binary BA–β-cyclodextrin inclusion complexes (Complexes A–C) and ternary BA–β-cyclodextrin–chitosan complexes (Complexes 1–3); results are presented as mean ± SD (n = 3).

Formulation code	System	Nominal composition	Drug content (assay, % label claim) (mean ± SD, n = 3)
Complex A	Binary	BA:β-CD 1:1	92 ± 1.78532
Complex B	Binary	BA:β-CD 1:2	90 ± 0.87643
Complex C	Binary	BA:β-CD 1:3	87 ± 0.27683
Complex 1	Ternary	BA:β-CD (1:1) + chitosan coating	82 ± 0.89765
Complex 2	Ternary	BA:β-CD (1:2) + chitosan coating	80 ± 0.23896
Complex 3	Ternary	BA:β-CD (1:3) + chitosan coating	84 ± 0.12467

The measured drug content values were close to the theoretical label claim across formulations, indicating efficient drug incorporation and acceptable content uniformity; the practical yield reflected process-related losses during solvent removal and freeze-drying, with ternary complexes expected to show slightly lower recovery due to additional coating/handling steps.

Binary docking [BA (ligand) +β-CD (receptor)

Docking analysis demonstrated favourable inclusion of BA within the hydrophobic cavity of β-CD, with GlideScores gives reasonable inclusion pose (Fig. 13). The predicted binding mode is consistent with classical β-cyclodextrins inclusion complexation, which is primarily driven by hydrophobic accommodation of the flavone scaffold within the cavity and stabilized by hydrogen bonding at the β-cyclodextrins rims^12,37,38. In the top-ranked pose, BA is expected to insert its largely hydrophobic aromatic framework into the β-CD cavity, while the polar hydroxyl groups remain oriented toward the rim, enabling hydrogen-bond interactions with the primary/secondary hydroxyl groups of β-CD and/or interfacial water molecules. Prime MM-GBSA rescoring yielded a binding free energy (ΔG_bind) of − 23.34 kcal/mol, supporting the formation of a stable host–guest inclusion complex (Table 3).

Fig. 13.

Top-ranked BA–β-CD inclusion pose from binary docking.

Table 3.

Docking and MM-GBSA binding energies of binary and ternary complexes.

System	GlideScore (kcal/mol)	Prime MM-GBSA ΔG_bind (kcal/mol)
Baicalein–β-cyclodextrin (Binary)	–4.30 to − 4.35	–23.34
Baicalein–β-cyclodextrin–Chitosan (Ternary)	–1.79 to − 2.86	–40.50
ΔΔG (Ternary – Binary)	–	–17.16 kcal/mol

Ternary complex [BAI–βCD (ligand) + chitosan (receptor)

The BA–β-CD inclusion complex was docked onto chitosan to evaluate ternary complexation. The top-ranked ternary poses produced less negative GlideScores (≈ − 1.79 to − 2.86 kcal/mol) (Fig. 14). This difference is not necessarily indicative of reduced stability, but likely reflects limitations of empirical docking scores in polymeric, surface-mediated interaction models³⁹. GlideScore is optimised for well-defined binding pockets and may underestimate diffuse hydrogen bonding networks and electrostatic contributions that often dominate polymer–carrier associations^40,41.

Fig. 14.

Best pose obtained from ternary docking (BA–β-CD–chitosan system).

In contrast, Prime MM-GBSA yielded a markedly more favourable binding free energy (ΔG_bind = − 40.50 kcal/mol), indicating stronger association in the ternary system. In the binary complex, the ternary system exhibited an additional stabilisation of 17.16 kcal/mol (ΔΔG = − 17.16 kcal/mol; Table 3), suggesting that chitosan introduces intermolecular interactions that enhance overall complex stability. MM-GBSA is generally more appropriate for polymer-associated systems because it accounts for solvation effects and permits post-docking relaxation^40,41. The enhanced stabilisation is consistent with additional hydrogen bonding and electrostatic interactions, potentially involving protonated amino groups of chitosan and the hydroxyl groups on the external surface of β-CD.

Mechanistic insight into baicalein–β-cyclodextrin inclusion

Molecular docking indicated that BA can be favourably accommodated within the β-CD hydrophobic cavity, with the aromatic scaffold positioned within the hydrophobic interior and polar hydroxyl groups oriented toward the rim. The corresponding MM-GBSA binding free energy (ΔG_bind = − 23.34 kcal/mol) supports thermodynamically favourable host–guest complex formation. Together with DSC and FTIR findings, these results support successful inclusion/association and reduced crystallinity, which are expected to improve dissolution performance.

Correlation between binary docking and dissolution enhancement

The enhanced dissolution observed for all binary BA- β-CD systems can be attributed to the molecular inclusion predicted by docking. Encapsulation within β-CD reduces crystal lattice energy, improves wettability, and promotes rapid drug–medium interaction, explaining the pronounced initial dissolution enhancement compared with pure BA. Among the tested ratios, the 1:2 BA:β-CD complex exhibited the highest dissolution, suggesting an optimal balance between complexation efficiency and drug loading. However, despite improved dissolution, the relatively rapid release and subsequent decline in drug concentration indicate that β-Cyclodextrins inclusion alone does not provide adequate control over drug diffusion, consistent with previous reports on binary β-Cyclodextrins systems.

Role of chitosan in ternary complex stabilisation and release control

Incorporation of chitosan introduced a second level of stabilization. Although ternary docking GlideScores were less negative—likely reflecting the limitations of empirical docking metrics for surface-mediated polymeric interactions—MM-GBSA predicted a substantially more favourable binding free energy (ΔG_bind = − 40.50 kcal/mol). The additional stabilization energy (ΔΔG = − 17.16 kcal/mol) suggests the chitosan contributes extra hydrogen bonding and electrostatic interactions, plausibly involving protonated amino groups of chitosan and the hydroxyl-rich -CD exterior. Such interactions have been reported to enhance the structural integrity and functional performance of β-cyclodextrins -polymer delivery systems.

Docking–dissolution relationship in ternary complexes

The superior stabilisation predicted by MM-GBSA calculations directly correlates with the experimentally observed sustained dissolution profiles of the ternary complexes. Unlike the binary systems, which exhibited rapid drug release, the ternary formulations showed smoother and more controlled release patterns. This behaviour reflects the formation of a polymer-coated inclusion system, in which chitosan acts as a diffusion-modulating barrier while β-CD maintains enhanced drug solubility. Notably, the 1:3 ternary system containing the highest β-CD ratio (1:3) achieved the highest overall release, indicating that higher β-CD content can support stronger inclusion and polymeric stabilisation. This finding underscores the importance of combining molecular encapsulation with polymeric control to achieve both enhanced dissolution and sustained release. It should be noted that molecular docking provides a static representation of molecular interactions and does not fully account for dynamic solution behaviour, polymer flexibility, or in-vivo physiological complexity. Therefore, docking results are interpreted as supportive mechanistic evidence rather than definitive predictors of pharmacokinetic or biological performance. The computational findings complement, but do not replace, experimental validation.

Conclusion

This study systematically developed and characterized binary (BA–β-cyclodextrin) and ternary (BA–β-cyclodextrin–chitosan) inclusion systems to address the poor aqueous solubility of baicalein. Equilibrium solubility analysis demonstrated significant enhancement compared with the pure drug, with the optimized 1:2 BA:β-CD system achieving approximately 8.8-fold improvement. The Higuchi–Connors phase-solubility study revealed an AL-type profile and an apparent stability constant (K_s) of 5.3 × 10² M⁻¹, confirming predominant 1:1 host–guest complex formation. Dissolution studies further demonstrated enhanced release for binary systems and modulated release behavior for ternary formulations.

Collectively, the integrated solid-state, thermodynamic, and computational findings provide mechanistic evidence supporting cyclodextrin-assisted solubilization of baicalein. While in-vivo pharmacokinetic evaluation is required to confirm bioavailability enhancement, the present results establish a strong physicochemical foundation for further development of baicalein oral delivery systems.

Author contributions

Ruba Malkawi; Study design, Methodology, writing the first draft of the manuscript, Data analysis. Baraa Jarwa; Docking study and molecular structure validation. Maram Hawarri; Docking validation, and revise the draft, 4 Israa H Isawi, Study validation and design, Katreen Banisalman; Revise the manuscript and structural validation, Osaima Alghodran; Methodology and validation, Qusai Al-Shdiefat ; Edit and revise the final copy of the manuscript.

Funding

This research was funded by Jadara University, Jordan, under research grant number 20245001.

Data availability

The data supporting the findings of this study are available from the corresponding authorupon reasonable request. Dr. Ruba Malkawi, Email: R.malkawi@jadara.edu.jo.

Declarations

Competing interests

The authors declare no competing interests.