Coral-Inspired Bacterial Cellulose/Calcium–Phosphorus Composites via Synchronous In Situ Biomineralization

Abstract

Inspired by the natural synchronization of organic matrix biosynthesis and inorganic mineral deposition in coral reefs, this study introduces a novel synchronous in situ biomineralization method. This strategy aims to overcome the challenges of weak interfacial bonding and complex processing in the fabrication of bacterial cellulose/calcium–phosphorus composites (BC/Ca–P). Calcium gluconate was used as a dual-functional precursor by substituting d-glucose in culture medium to simultaneously supply a carbon source for BC biosynthesis and calcium ions (Ca²⁺) for mineralization at different substitution ratios (10% for BC-CaG₁₀ and 100% for BC-CaG₁₀₀). We achieved concurrent biosynthesis of the BC matrix and deposition of Ca–P phases, leading to a unique inorganic–organic architecture in which Ca–P nanoparticles were uniformly anchored within the BC nanofibrous network and strongly integrated by interfacial interactions. Consequently, the moderate mineralization in BC-CaG₁₀ significantly enhanced the mechanical strength through a synergistic reinforcement mechanism, reaching 12.03 MPa, representing a 115% increase compared to that of pure BC. Conversely, the high mineral content of BC-CaG₁₀₀ facilitated the formation of a bioactive hydroxyapatite (HAP) phase, providing abundant active sites. In preliminary biomedical evaluations, the fabricated composites exhibited favorable biocompatibility. Notably, BC-CaG₁₀₀ exhibited efficient hemostatic performance in a rat tail amputation model, reducing hemostasis time by 71% (to 142 s) and blood loss by 97% (to 15 mg). This study presents a facile, scalable, and green approach for preparing novel biomimetic materials with promising potential for biomedical applications, including tissue engineering and wound management.

1. Introduction

Biomineralization is a complex physiological process that refers to the formation of inorganic–organic composites with hierarchical and ordered structures by organisms in vitro or ex vivo in a benign environment. , The exoskeletons of corals are typical examples of these biocomposites. Their formation is more than a simple physical and chemical precipitation, it is a biologically controlled crystallization process. Critically, the formation of coral skeletons is a highly coordinated and synchronized process in which coral polyps secrete an organic matrix and actively regulate the nucleation and growth of calcium carbonate through physiological mechanisms. , This dynamic and synchronous biomineralization strategy accurately coordinates the organic and inorganic phases, imparting the resulting structure with integration and functions superior to those obtained via conventional mineralization. Thus, mimicking the natural coordination of matrix biosynthesis and mineral deposition presents a key strategy for developing advanced organic–inorganic hybrid materials.

Bacterial cellulose (BC) is a natural nanofiber biomaterial produced by specific microorganisms, such as Gluconacetobacter xylinus. During fermentation, bacteria convert sugar into cellulose chains, which then self-assemble into a three-dimensional (3D) interwoven nanofiber network. This biosynthesized nanomaterial has excellent purity and biocompatibility, ensuring the safety of the tissue interface. The nanofiber structure closely mimics the natural extracellular matrix (ECM), providing an ideal scaffold for cell adhesion and proliferation, and holds significant potential in tissue engineering. , BC also possesses superior water retention and vapor permeability properties, aiding in exudate absorption and maintenance of a moist wound environment. These characteristics make BC an outstanding material for advanced medical dressing applications. ,

Calcium–phosphorus (Ca–P) compounds are important bioactive inorganic materials, mainly calcium phosphate and its derivatives. Hydroxyapatite (HAP), β-tricalcium phosphate (β-TCP), and amorphous calcium phosphate (ACP) have attracted particular attention. , This prominent status stems from the homology in composition and structure between them and the minerals found in human bone and teeth. These biomimetic properties make these compounds the basic biological materials for bone tissue engineering. Furthermore, Ca²⁺ released from the Ca–P matrix can accelerate blood coagulation and promote wound healing. Therefore, the integration of Ca–P crystallization with BC networks offers a promising strategy for fabricating bacterial cellulose/calcium–phosphorus (BC/Ca–P) composites. This method is designed to overcome the inherent limitation of limited biological activity in native BC, endowing the BC/Ca–P composite with superior biological activity and hemostatic capabilities. At the same time, this simultaneous in situ mineralization process generates strong interfacial interactions between the Ca–P and BC fibers, firmly anchoring the inorganic Ca–P phase to the BC nanofibers.

Current strategies for the fabrication of BC/Ca–P composites are predominantly classified into two-step or multistep postmineralization techniques and in situ physical incorporation (Table ). Postmineralization generally involves a sequential process, beginning with the synthesis of the organic matrix, followed by the deposition of the inorganic phase, which requires intricate secondary treatments. For example, the stepwise biomimetic mineralization strategy adopted by Xun et al. and Shi et al. necessitates the repetitive immersion of the BC matrix in precursor solutions or simulated body fluids (SBF). , Despite their efficacy, these protocols are time-consuming, and frequent buffer exchanges result in significant reagent consumption and environmental burden. Similarly, Busuioc et al. utilized ultrasound-assisted precipitation to accelerate mineral deposition, however, this process still involves complex preoxidation and ultrasound treatment steps. Conversely, in situ physical incorporation involves the direct introduction of presynthesized minerals during bacterial fermentation. For example, Niamsap et al. supplemented the culture medium with presynthesized hydroxyapatite coated cellulose nanocrystals (HAP/CNCs). However, this remains inherently a two-step approach requiring nanoparticle presynthesis. During the prolonged fermentation process, presynthesized inorganic minerals inevitably undergo gravitational settling within the liquid culture medium. This physical settling typically leads to particle aggregation of minerals, consequently resulting in weak interfacial bonding within the BC network. Therefore, developing a simplified, synchronized one-step strategy that does not rely on external nucleating agents, presynthesis, or complex postprocessing is critical for the large-scale fabrication and clinical translation of BC-based medical devices.

1. A Strategic Approach to the Manufacture of BC/Ca–P Composites .

reference	mineralization strategy	precursors/key additives	procedure	time consumption	limitations/challenges
Xun et al. (2024)	stepwise biomimetic mineralization	CaCl2 + 1.5 × SBF	multistep, 1. ferm., 2. CaCl2 immersion, 3. SBF immersion	ferm. + 1–14 days (post treatment)	most time-consuming, high environmental burden due to frequent buffer exchange
Shi et al. (2022)	post-treatment in situ mineralization	Ca-gluconate + SBF	two-step, 1. ferm., 2. SBF immersion	ferm. + 1–5 days (SBF soaking)	requires secondary chemical treatment to induce crystal nucleation and growth
Busuioc et al. (2022)	ultrasound-assisted postmineralization	DABC + Ca/P salts	multistep, 1. ferm., 2. oxidation, 3. ultrasound assisted mineralization	ferm. + 7–8 days (modification & mineralization)	high energy consumption, complex alternating immersion steps
Niamsap et al. (2019)	in situ physical incorporation	nano-HAP/CNC	two-step, 1. nanoparticle synthesis, 2. ferm	∼10 days (additive preparation & ferm.)	nanoparticle aggregation, weak interfacial bonding
this work	simultaneous in situ biomineralization	Ca-gluconate (C/Ca source) + phosphate Buffer (P source)	one-step, 1. simultaneous synthesis during fermentation	∼7 days (ferm. only)	maximum mineral loading is constrained by the solubility limit of Ca2+ in the culture medium

^aNote: ferm. = fermentation. SBF = simulated body fluid. DABC = dialdehyde bacterial cellulose. CNC = cellulose nanocrystals.

In this study, we introduce a novel bioinspired strategy that integrates coral biomineralization principles into BC biosynthesis. Mimicking the synchronous growth of organic and inorganic phases typical of coral reefs, we developed a synchronous in situ biomineralization protocol (Figure ). Specifically, calcium gluconate was employed as a dual-functional precursor by replacing d-glucose in the culture medium. This metabolic design concurrently provide the carbon source for BC biosynthesis and calcium ions (Ca²⁺) for the biomineralization of Ca–P at varying substitution ratios. Subsequently, the micromorphology, physicochemical properties, biocompatibility, and hemostatic performance of the synthesized BC/Ca–P composites were systematically characterized.

1.

Schematic of the synchronous in situ biomineralization of BC/Ca–P composites.

2. Materials and Methods

2.1. Materials

The bacterial strain G. xylinus 1.1812 was obtained from the Chinese Academy of Sciences. d-glucose and calcium gluconate were purchased from Aladdin Reagent Co., Ltd. Fishmeal peptone, yeast extract, potassium dihydrogen phosphate, and sodium hydroxide were acquired from Sinopharm Chemical Reagent Group Co., Ltd., and citric acid was supplied by Shanghai Een Chemical Technology Co., Ltd. The cell Counting Kit-8 (CCK-8) was purchased from Dojindo Molecular Technology Co., Ltd. L929 mouse fibroblasts were purchased from Shanghai Institute of Biochemistry and Cell Biology. All chemicals used were of analytical grade, unless otherwise specified.

2.2. Preparation of BC/Ca–P Composites

BC and BC/Ca–P composites were synthesized using a static fermentation method. The basal medium for all groups contained 0.5 wt % fishmeal peptone, 0.5 wt % yeast extract, 0.2 wt % disodium hydrogen phosphate, 0.1 wt % potassium dihydrogen phosphate, and 0.1 wt % citric acid. The experimental groups were differentiated based on the carbon source composition. The control group (BC) consisted of 5.0 wt % d-glucose. The composite groups were formulated by substituting d-glucose with calcium gluconate: BC-CaG₁₀ contained 4.5 wt % d-glucose and 0.5 wt % calcium gluconate (10% replacement), whereas BC-CaG₁₀₀ contained 5.0 wt % calcium gluconate in place of d-glucose (100% replacement). Prior to inoculation, the medium pH was adjusted to 5.5 using NaOH and sterilized at 121 °C for 30 min. The media were then inoculated with the activated seed culture and incubated statically at 35 °C for 8 days. At the end of fermentation, the harvested membrane samples were treated with a 2 wt % NaOH solution at 80 °C for 70 min to remove bacterial cells and residues. Subsequently, the samples were rinsed thoroughly with deionized water until neutral and lyophilized at −12 °C for 24 h.

2.3. Characterization Methods

2.3.1. Scanning Electron Microscopy (FE-SEM)

The samples were characterized using a field-emission scanning electron microscope (Regulus 8100, Hitachi, Japan). The dried sample surface was secured to an aluminum sample stage using conductive carbon tape, followed by gold sputtering for 180 s under vacuum conditions to enhance conductivity. The microscopic morphology was observed at an accelerating voltage of 5 kV, and the elemental compositions were analyzed using energy dispersive X-ray spectroscopy (EDS).

2.3.2. Transmission Electron Microscopy (TEM)

The BC-CaG₁₀₀ sample was first immersed in deionized water for 24 h and then subjected to ultrasonic treatment for 30 min. The released Ca–P inorganic particles in the resulting suspension were observed using HRTEM. The sample suspension was dropped onto a carbon-coated copper grid and dried. The microscopic morphology of the samples was observed at higher magnifications using a field-emission transmission electron microscope (Talos F200X, Thermo Fisher Scientific, USA). The elemental composition was determined using energy-dispersive X-ray spectroscopy (EDS).

2.3.3. Fourier Transform Infrared Spectroscopy (FTIR)

Spectra were recorded in the range of 4000–500 cm^–1 using a Fourier transform infrared spectrometer (Vertex 70, Bruker, Germany) to analyze chemical bonds. The cellulose I _α content was quantified by the following formula (eq ).

$$f_{\alpha }(\%)=\frac{A_{750}}{A_{750}+{KA}_{710}}$$ 1

where K = 0.16, and A is the area of the corresponding peak.

2.3.4. X-ray Diffraction Characterization (XRD)

Diffraction patterns were recorded using an X-ray diffractometer (Smartlab SE, Nippon Rigaku, Japan) in the 2θ range of 10–70° with a scanning speed of 10°/min. The crystallinity (Crl^XRD) was calculated using the Segal equation, as in the following formula (eq ).

$${\mathrm{C}\mathrm{r}\mathrm{l}}^{\mathrm{X}\mathrm{R}\mathrm{D}}(\%)=\frac{I_{200}-I_{\mathrm{a}\mathrm{m}}}{I_{200}}\times 100$$ 2

where I ₂₀₀ represents the maximum intensity of the peak corresponding to the 200 crystalline plane and I _am represents the intensity of the amorphous region.

The apparent crystallite size (Cs) perpendicular to the (200) plane was calculated using the following formula (eq ).

$$\mathrm{C}\mathrm{s}(\mathrm{n}\mathrm{m})=\frac{0.89\times \lambda }{\beta \mathrm{C}\mathrm{O}\mathrm{S}\theta }$$ 3

where λ is the X-ray wavelength (0.154 nm), β is the full width at half-maximum (FWHM) of the diffraction peak (in radians), and θ is the Bragg diffraction angle.

2.3.5. X-ray Photoelectron Spectroscopy (XPS)

The surface chemical composition was analyzed using an X-ray photoelectron spectrometer (K-Alpha, Thermo Scientific, USA) in the range of 1500–0 eV at a scan step of 1.0 eV with Al Kα irradiation (1486.6 eV) at 12 kV and 6 mA.

2.3.6. Mechanical Properties

The dry samples were cut into strips (40 mm × 10 mm) for mechanical testing. The sample thickness was measured before the analysis. The tensile properties were evaluated using a universal testing machine (Instron 5982, Instron, USA) with a 50 N load cell at a rate of 10 mm/min.

2.3.7. Porosity

The porosity of Samples was measured using the ethanol displacement method. A freeze-dried scaffold was weighed to obtain W ₁. It was then completely submerged in 10 mL of anhydrous ethanol. After soaking for 1 h, the volume of anhydrous ethanol was recorded as V ₁. The scaffold was removed, excess liquid was gently blotted from the surface with filter paper, and it was weighed again to obtain W ₂. The porosity (%) was determined using the following formula (eq ).

$$\mathrm{P}\mathrm{o}\mathrm{r}\mathrm{o}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{y}(\%)=\frac{W_2-W_1}{{\rho }_{\mathrm{C}}{V}_1}\times 100$$ 4

where W ₁ is the dry weight, W ₂ is the wet weight, V ₁ is the volume of anhydrous ethanol after 1 h of soaking, and ρ_C is the density of anhydrous ethanol.

2.3.8. Cytotoxicity Assay

The cytocompatibility of the samples was evaluated using the CCK-8 assay. L929 cells were cultured in Dulbecco’s modification of Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 100 U/mL of penicillin, and 100 μg/mL of streptomycin at 37 °C in a humidified atmosphere with 5% CO₂. Material extracts were prepared according to ISO 10993.5 (GB/T 16886.5). Briefly, the samples (BC, BC-CaG₁₀, and BC-CaG₁₀₀) were sterilized by autoclaving. Extracts were prepared by immersing the samples in complete DMEM at a ratio of 0.1 g/mL and incubating at 37 °C for 24 h. The supernatant was collected and filtered through a 0.22 μm membrane filter to obtain the final extract. L929 cells were seeded into 96-well plates at a density of 1 × 10⁴ cells/well and cultured for 24 h to allow for attachment. Subsequently, the prepared extracts were used to replace the culture medium, and fresh DMEM with 10% FBS was used as the control. Cell viability was assessed at day 1, 3, and 5 by adding 10 μL of CCK-8 solution to each well and incubating for 2 h. Absorbance at 450 nm was measured using a microplate reader. Relative cell viability was calculated based on the ratio of sample absorbance to control absorbance. Cell morphology was observed using an inverted optical microscope after 72 h.

2.3.9. Hemostatic Properties

The hemostatic activity of various samples (gauze, BC, and BC-CaG₁₀₀) was evaluated using a rat tail amputation model. All animal experimental procedures were conducted in accordance with the Guidelines for the Care and Use of Laboratory Animals and were approved by the Animal Ethics Committee of Tongji Hospital, Tongji University (No.: 20241225-20250228-DW-130). Male Sprague–Dawley (SD) rats (0.3–0.4 kg) were randomly divided into three groups (n = 5 per group). Rats were anesthetized via intraperitoneal injection of 50 μg zolazepam combined with xylazine (0.1 mL/100 g). To ensure consistency in the initial bleeding conditions, all rats were deeply anesthetized and immobilized, and the distal 6 mm of the tail tip was precisely excised using a standardized surgical scalpel. Following amputation, the tail was allowed to bleed freely for 15 s to establish normal blood flow. Subsequently, each material sample was immediately applied to the wound site with light pressure to control bleeding. The hemostasis time was recorded using a stopwatch the moment bleeding completely ceased and a stable blood clot formed at the wound site, as confirmed by continuous visual observation without any signs of rebleeding. To completely eliminate subjective bias from a single observer’s visual estimation, hemostasis time was recorded simultaneously by two independent investigators using stopwatches in a double-blinded manner, and the average of the two recorded times was taken as the final data. Additionally, to prevent severe underestimation of blood loss due to rapid water evaporation, the blood-soaked samples were immediately transferred into preweighed 50 mL centrifuge tubes and sealed after confirming hemostasis. All samples were then weighed collectively. To observe the microscopic interactions between the dressing materials and blood components, the surfaces of the samples were examined using SEM following the in vivo hemostatic tests.

2.4. Statistics

In this study, all experiments in each test group had at least three parallel replicates. The data are presented as mean ± standard deviation. Statistical analyses were performed using GraphPad software. A one-way analysis of variance (ANOVA) was used to compare groups and determine the significance of the results. *p < 0.05 is the level of significance and **p < 0.01 is the level of high significance.

3. Results and Discussion

3.1. Micromorphological Analysis

The surface morphologies of pure BC and BC/Ca–P composites, synthesized by synchronous in situ biomineralization at varying calcium gluconate concentrations in the culture medium, are shown in Figure a. The pure BC sample exhibited a characteristic 3D porous network composed of randomly interwoven nanofibers. In the composite samples, observable mineral deposits were primarily distributed along the nanofibers and were embedded within the BC network. The mineral morphology and dimensions were dependent on the Ca²⁺ concentration in the culture medium. The particles in the BC-CaG₁₀ group appeared predominantly columnar. As the Ca²⁺ concentration increased, the particle diameter increased progressively. Notably, the BC-CaG₁₀₀ sample exhibited distinct spherical and agglomerated morphologies. This trend indicates that higher Ca²⁺ availability facilitates Ca–P mineralization for crystal growth and accumulation during BC biosynthesis. EDS mapping (Figure b) revealed a uniform distribution of calcium (Ca) and phosphorus (P) throughout the BC network, confirming the successful deposition of the Ca–P phase.

2.

SEM and EDS images of BC and BC/Ca–P composites. (a) FE-SEM images of pure BC, BC-CaG₁₀, and BC-CaG₁₀₀. (b) EDS elemental mapping of BC-CaG₁₀₀ samples.

To further investigate the nanostructure of the synthesized BC/Ca–P composite, the supernatant obtained from the ultrasonicated BC-CaG₁₀₀ suspension was analyzed using HRTEM. The micrographs (Figure a,b) reveal mineralized Ca–P deposits embedded within the BC network. The Ca–P deposits formed irregular hierarchical clusters ranging from 100 to 300 nm in size. As shown in the high-magnification TEM image (Figure b), the larger Ca–P particles consist of densely aggregated needle-like nanocrystals, indicating a bottom-up assembly process. Remarkably, even under the significant mechanical shear forces induced by ultrasonication, the Ca–P particles remained firmly anchored to the BC nanofibers. This observation demonstrates a strong interfacial affinity between the in situ grown mineral phase and the BC matrix. Such strong interfacial anchoring unequivocally distinguishes our metabolically driven synchronous in situ biomineralization from BC/Ca–P composites fabricated via physical incorporation or multistep postmineralization strategies, where simply blended or postdeposited inorganic particles are highly susceptible to detachment and leaching under similar shear stress. ,

3.

TEM image of the BC-CaG₁₀₀ sample. (a,b) TEM images at different magnifications. (c) Bright-field TEM micrograph and corresponding elemental distribution maps for carbon (C), oxygen (O), phosphorus (P), and calcium (Ca).

To characterize the elemental composition and spatial distribution within the composite, TEM-EDS mapping was conducted. The elemental maps show a clear colocalization of Ca and P signals, which are selectively concentrated in the particulate regions, as shown in the bright field image (Figure c). The elemental maps also revealed that carbon (C) was predominantly distributed along the BC fibers, consistent with the BC matrix. Oxygen (O) was detected throughout the sample, and its signal intensity was markedly enhanced within the mineralized regions. The precise spatial correlation among Ca, P, and enriched O confirms that the particulate deposits are Ca–P compounds. Furthermore, the preferential deposition of Ca–P within the BC network suggests strong interactions between Ca–P and BC fibers, potentially arising from electrostatic forces, hydrogen bonding, or other specific interactions. ,,

3.2. Chemical Composition and Crystal Structure Characterization

The chemical structures of the pure BC and BC/Ca–P composites were characterized using FTIR (Figure a). The full-range spectra of pure BC exhibited characteristic bands at 3348 cm^–1 (O–H stretching), 2891 cm^–1 (C–H stretching), and 1637 cm^–1 (O–H in-plane bending vibration), which are consistent with typical BC. − For the BC-CaG₁₀₀ composite, two new absorption bands emerged at 1024 cm^–1 and 964 cm^–1, which are characteristic of phosphate (PO₄ ) stretching vibrations, confirming the in situ mineralization reaction between Ca²⁺ and phosphate ions to form Ca–P compounds. Notably, the relative intensities of the characteristic cellulose bands were attenuated in the BC-CaG₁₀₀ sample. This observation suggests that at high precursor concentrations of Ca²⁺, the abundant Ca–P deposits form a coating layer on the BC nanofibers, partially shielding the cellulose signal. Conversely, the absence of discernible phosphate peaks in the BC-CaG₁₀ spectrum implies that the mineral content was below the FTIR detection threshold.

4.

FTIR patterns of BC and BC/Ca–P composites. (a) Full-range FTIR spectra of BC, BC-CaG₁₀, and BC-CaG₁₀₀. (b–d) Deconvoluted spectra in the 800–650 cm^–1 region for determining the I _α mass fraction (f _α) of (b) BC, (c) BC-CaG₁₀, and (d) BC-CaG_100..

To further investigate the effect of in situ mineralization on the cellulose crystal structure, the spectral region of 800–650 cm^–1 was deconvoluted (Figure b–d). The bands near 750 cm^–1 and 710 cm^–1 are assigned to the out-of-plane bending vibration of –OH groups embedded in the triclinic (I _α) and monoclinic (I _β) allomorphs of cellulose, respectively. Quantitative analysis based on these bands revealed that the mass fraction of the I _α allomorph (f _α) increased from 62.1% in pure BC to 65.7% in BC-CaG₁₀, and further to 71.1% in BC-CaG₁₀₀. This progressive increase indicates that the in situ mineralization process promoted by calcium gluconate favors the formation of the cellulose I _α phase. We believe this is primarily the result of physical forces. According to classical studies on cellulose crystallography, the I _α crystal form is more likely to form in larger microfibril bundles, which are formed through the further aggregation of smaller subfibrils secreted by the cells. In our comineralization system, the introduction of calcium gluconate alters the ionic environment of the culture medium, accelerating the physical aggregation of extruded cellulose chains and thereby promoting the formation of large microfibrils rich in the I _α phase. Consequently, driven by the physics of “high-speed extrusion,” the system is strongly directed toward a kinetically controlled pathway, thereby promoting the formation of a higher proportion of the metastable I _α phase.

It has been established that cellulose I _α serves as the fundamental structural component of BC biomaterial properties, and that an increase in the proportion of cellulose I _α crystalline forms significantly influences their physical properties. Specifically, as the relative content of I _α increases, the crystallinity of the material decreases, the toughness increases, conferring better deformability and energy dissipation mechanism, and the Young’s modulus decreases accordingly, indicating that the rigidity of the material is reduced. The triclinic I _α phase is characterized by a metastable structure with a unique hydrogen bonding network. From a biomedical standpoint, this metastability results in increased surface energy and enhanced accessibility of active hydroxyl groups, thereby creating an optimal microenvironment for protein adsorption and cell adhesion. Consequently, this structural modulation not only affects mechanical properties but also plays a crucial role in determining the performance of cellulosic biomaterials. − These crystallographic results further confirm the gradual increase in the proportion of the I _α phase in the samples. We would also like to note that, although the spectroscopic data strongly suggest these chemical interactions, computational simulations and in situ kinetic monitoring will still be required in the future to provide definitive proof.

The crystalline structure and phase composition of the samples were analyzed using XRD (Figure ). The pure BC displayed characteristic diffraction peaks of cellulose I with 2θ values of about 14.1°, 16.4°, and 22.5°, corresponding to the (100) (010), and (110) crystal planes, respectively. , For the composite samples (BC-CaG₁₀ and BC-CaG₁₀₀), the retention of these cellulose I peaks indicates that the fundamental crystalline architecture of the BC matrix was preserved during in situ mineralization. Notably, new diffraction reflections were observed in the BC/Ca–P composites, most prominently in BC-CaG₁₀₀, at 2θ values of 25.9°, 31.9°, 33.2°, 40.0°, 46.8°, 49.6°, and 53.4°. These reflections correspond to the (002), (211), (300), (130), (222), (213), and (004) planes of hydroxyapatite (HAP), respectively, fully matching the standard JCPDS Card No. 98-000-0251. , It is worth noting that the main reason why the (211) and (300) diffraction peaks are not prominent in the BC-CaG₁₀ samples is the low concentration of calcium and phosphorus precursors in this group, resulting in a low content of the in situ synthesized HAP phase. BC itself possesses extremely high crystallinity, which produces exceptionally intense diffraction peaks and a broad background baseline. Consequently, in the BC-CaG₁₀ sample, the faint diffraction signals generated by the low-content inorganic phase were completely masked by the strong diffraction peaks of the highly crystalline cellulose matrix. In contrast, in the BC-CaG₁₀₀ sample, as the precursor concentration increased significantly, the degree of mineralization in the material increased markedly. At this point, the absolute content of the HAP mineral phase exceeds the detection threshold of XRD, overcoming the masking effect of the BC matrix. Consequently, characteristic diffraction peaks such as (211) and (300) become very distinct. The partial XRD spectrum in Figure b shows that the (010) diffraction peak of pure BC is at 16.62°, that of BC-CaG₁₀ is at 16.7°, and that of BC-CaG₁₀₀ is at 16.76°. Studies have shown that as the degree of mineralization increases, the characteristic XRD peaks shift toward higher 2θ angles, which is an important indicator of an increase in the proportion of cellulose I _α.

5.

XRD patterns of BC and BC/Ca–P composites. (a) XRD full spectrum and the inset shows the HRTEM image of BC-CaG₁₀₀. (b) XRD spectrum in the 2θ range from 15.5° to 18°.

Furthermore, the inset HRTEM analysis in Figure shows clear lattice fringes with an interplanar spacing of 0.281 nm, corresponding to the (211) plane of HAP. Crystallinity analysis (Table ) revealed that pure BC exhibited the highest crystallinity index (89.71%), reflecting a highly ordered packing of cellulose chains. However, with the increasing dosage of calcium gluconate in the fermentation medium, the crystallinity index declined progressively, reaching 79.94% for BC-CaG₁₀ and 64.08% for fully substituted BC-CaG₁₀₀. This reduction resulted from the synchronous in situ biomineralization process, wherein the formation and deposition of HAP particles within the cellulose matrix disrupted the crystalline structure by disturbing the original orderly chain arrangement of cellulose molecules. The consequent formation of amorphous regions at the cellulose-HAP interface further lowered overall crystallinity. Concurrently, the cellulose crystallite size decreased from 8.52 nm in BC to 5.87 and 5.78 nm in BC-CaG₁₀ and BC-CaG₁₀₀, respectively. This reduction indicates that the in situ formation of HAP imposes a confinement effect that restricts the lateral growth of the cellulose crystallites.

2. d-Spacings, Crystallinity, and Crystallite Size of BC, BC-CaG₁₀, and BC-CaG₁₀₀ .

	d(100)/nm	d(010)/nm	d(110)/nm	CrlXRD/%	crystallite size/nm
BC	0.614	0.532	0.392	89.71	8.52
BC-CaG10	0.613	0.531	0.393	79.94	5.87
BC-CaG100	0.612	0.535	0.393	64.08	5.78

The surface chemical composition and valence states of the samples were analyzed using X-ray photoelectron spectroscopy (XPS). As shown in Figure a, the survey spectrum of pure BC is dominated by the C 1s (285.0 eV) and O 1s (532.0 eV) peaks, which are characteristic of the cellulosic structure. In contrast, evident Ca 2p and P 2p peaks appeared in the BC-CaG₁₀₀ spectrum, confirming the deposition of a calcium phosphate phase on the BC surface. , For BC-CaG₁₀, these signals were present but significantly weaker, consistent with the lower Ca²⁺ concentration in the culture medium. High-resolution spectra of BC-CaG₁₀₀ further confirmed the chemical states (Figure b,c). The P 2p peak at ∼134 eV is characteristic of PO₄ , and a Ca 2p doublet at ∼348 eV and ∼352 eV (corresponding to Ca 2p3/2 and Ca 2p1/2, respectively) confirms the presence of Ca²⁺ in the crystal lattice. Furthermore, the Ca/P ratio of the BC-CaG₁₀₀ sample calculated from the XPS data was 1.66, which closely matches the theoretical value of 1.67 for HAP. These results corroborate the FTIR and XRD findings and conclusively demonstrate that HAP is formed within the BC matrix in BC-CaG₁₀₀ via synchronous in situ mineralization.

6.

XPS surface analysis of BC and BC/Ca–P composites. (a) Wide-scan survey spectra. (b) High-resolution P 2p spectrum of BC-CaG₁₀₀. (c) High-resolution Ca 2p spectrum of BC-CaG₁₀₀.

The high-resolution C 1s spectra were deconvoluted into three separate peaks corresponding to C–C/C–H, C–O, and O–C–O bonds (Figure a–c). As quantified in Figure d, the pure BC sample showed a C–O bond proportion of 43.2%, which is characteristic of the polysaccharide backbone of cellulose. For the BC-CaG₁₀ sample, this proportion increased markedly to 70.9%. We suggest that this significant increase is due to Ca²⁺ acting as a “nucleation site” or “inducer,” disrupting the original strong intermolecular hydrogen bonds in the BC region and exposing more previously hidden C–O groups on the fiber surface, thereby increasing the proportion of C–O bonds. , In contrast, the proportion of C–O bonds in BC-CaG₁₀₀ decreased to 59.4%, which is intermediate between the values for BC-CaG₁₀ and pure BC. This reduction can be attributed to the formation of a more ordered and thermodynamically stable HAP crystal structure. The increased lattice perfection likely diminishes the density and/or reactivity of the interfacial sites, thereby weakening the specific interactions with cellulose and leading to a lower extent of C–O bond formation at the organic–inorganic interface.

7.

Analysis of the carbon bonding states. (a–c) High-resolution C 1s spectra after deconvolution for (a) BC, (b) BC-CaG₁₀, and (c) BC-CaG₁₀₀. (d) Quantitative distribution of functional groups derived from spectral deconvolution.

3.3. Characterization of Mechanical Properties

The typical tensile stress–strain curves for the pure BC and BC/Ca–P composites are shown in Figure a, indicating a significantly improved mechanical performance of BC-CaG₁₀ in the dry state. The elongation at break results are shown in Figure b. BC-CaG₁₀ and BC-CaG₁₀₀ were 4.16 ± 0.71% and 4.30 ± 0.23%, respectively. Both values were slightly higher than the 3.58 ± 0.49% for pure BC, but the difference was not statistically significant. This indicates that HAP deposition provided a limited enhancement to the ductility of the BC matrix. The tensile strength revealed a dramatic improvement for BC-CaG₁₀, which showed a remarkable 115% increase to 12.03 ± 1.13 MPa (vs 5.59 ± 0.61 MPa for pure BC; p < 0.01), whereas that of BC-CaG₁₀₀ decreased to 4.84 ± 0.68 MPa, falling slightly below the pure BC (Figure c). The Young’s modulus followed a trend similar to that of the tensile strength, peaking at 335.4 ± 27.9 MPa for BC-CaG₁₀ (1.6 times that of pure BC at 206.8 ± 20.9 MPa) and dropping to the lowest value of 130.7 ± 15.2 MPa for BC-CaG₁₀₀ (Figure d). Collectively, these results indicate a Ca²⁺ concentration-dependent effect of HAP mineralization on the mechanical properties of BC, where moderate deposition reinforces the tensile performance, whereas a higher level of mineralization induces slight degradation. ,

8.

Mechanical properties of BC and BC/Ca–P composites. (a) Stress–strain curves. (b) Elongation at break. (c) Tensile strength. (d) Young’s modulus. Data are presented as mean ± SD. Statistical significance is indicated by asterisks (**p < 0.01; ns: not significant).

Mechanistically, the integrity of the 3D BC nanofibrous network and the strength of the organic–inorganic interfacial bonding are the primary determinants of the mechanical performance of BC/Ca–P composites. In BC-CaG₁₀, moderate mineralization produces beneficial synergistic effects. The Ca–P phase is uniformly dispersed as nanoparticles, acting as a rigid filler (see Figure a). As indicated by the XPS results mentioned above, the originally tight intermolecular hydrogen bonds in the BC phase have been disrupted. At the same time, Ca²⁺ and C–O groups form coordination bonds that result in strong interfacial interactions, firmly anchoring the in situ grown HAP particles to the BC fibers. It is precisely these strong chemical interfacial bonds at the microscopic level that successfully trigger the macroscopic “synergistic enhancement mechanism,” promoting stress transfer and significantly enhancing the strength and toughness of the composite material. In contrast, the high mineralization of BC-CaG₁₀₀ had a detrimental effect on its structure. The reduced crystallinity of cellulose (XRD), coupled with the uneven aggregation of HAP particles observed by SEM, disrupted the continuity of the BC nanofiber network, reduced the effective interfacial contact area between BC and HAP, and physically shielded the exposed active C–O bonding sites, leading to a large-scale failure of strong interfacial bonding at the microscopic level. At the same time, loose particle aggregates formed significant stress concentration points within the composite structure, preventing the effective transfer and dispersion of stress. This manifested macroscopically as the disappearance of synergistic reinforcement, a decline in tensile properties, and increased brittleness.

In particular, the rapid deposition of HAP crystals in BC-CaG₁₀₀ physically prevents newly formed cellulose chains from assembling into highly ordered microfibrils via hydrogen bonding, ultimately leading to a loss of crystallinity and mechanical strength. In future work, we will optimize the concentration of the phosphate buffer to stabilize the pH or shorten the fermentation time to limit excessive mineral growth, to investigate whether this can mitigate such structural degradation.

3.4. Porosity

To further elucidate the impact of mineral aggregation on the fluid absorption capacity of the scaffolds, quantitative porosity measurements were conducted (Figure ). The pure BC exhibits the highest porosity, approximately 81.6 ± 2.5%. With the introduction of HAP, the porosity shows a slight downward trend. The porosity of the BC-CaG₁₀ sample showed no significant change compared to pure BC, at 78.9 ± 1.7% (ns), whereas the BC-CaG₁₀₀ sample exhibited a more significant decrease, with its porosity dropping to approximately 72.3 ± 3.1% (p < 0.01). We speculate that the decrease in porosity is primarily attributed to the physical filling effect of HAP nanoparticles. When HAP is incorporated into the BC matrix, the inorganic particles occupy the interstitial voids within the complex three-dimensional nanofiber network of cellulose. As the HAP loading gradually increased, more pore spaces were filled or blocked, resulting in a more compact and dense composite structure. From a rapid absorption of blood and exudate in clinical wound management perspective, although the addition of HAP reduced the total void volume, the porosity of the BC-CaG₁₀₀ composite remained above 70%. This retained ideal porous architecture is entirely sufficient to support robust capillary action for the rapid absorption of blood and exudates, while also providing unobstructed fluid pathways for the continuous release of bioactive Ca²⁺.

9.

Porosity of BC and BC/Ca–P composites.

3.5. In Vitro Cytocompatibility

The cytocompatibility of materials is crucial in biomedical applications. Therefore, the CCK-8 assay, which reflects the number of viable cells based on metabolic activity, was used to evaluate the biocompatibility of the samples with the L929 fibroblasts. As shown in Figure a, the absorbance of L929 cells from the BC, BC-CaG₁₀, and BC-CaG₁₀₀ samples increased continuously over days 1, 3, and 5, indicating that the L929 cells could carry out normal propagation behavior on BC materials. Furthermore, no significant differences were observed in the absorbance values among the groups at each time point, demonstrating a consistent and desirable growth trend. Figure b shows the relative viability of L929 cells. The cell viability values of BC-CaG₁₀ and BC-CaG₁₀₀ groups were higher than 100% on days 1, 3, and 5 compared to the BC group, indicating that these materials are not cytotoxic and promote cell proliferation. The ANOVA results showed no significant difference between the groups (ns). These results confirm the good biocompatibility of pure BC and BC/Ca–P composites.

10.

Cell proliferation results of L929 on different samples on days 1, 3, and 5. (a) Absorbance at 450 nm for different samples. (b) Cell viability of different samples. Data are presented as mean ± SD. Statistical significance is indicated by asterisks (ns, not significant).

3.6. In Vivo Hemostatic Performance

In accordance with widely accepted ethical guidelines for animal experimentation (the 3R principles: replacement, reduction, and refinement), it is strictly required to minimize the number of animals used. Although BC-CaG₁₀ exhibits excellent mechanical properties, its content of Ca–P mineral phases is extremely low. Since rapid hemostasis primarily relies on the release of Ca²⁺ from the HAP phase to activate the coagulation cascade, a mechanism that allows for the prediction of its hemostatic efficacy, we selected animal samples for the control group (pure BC) and the experimental group (BC-CaG₁₀₀), which exhibits optimal biological activity and hemostatic potential, to ensure that in vivo experiments are both efficient and ethically sound. This comparison allows for in vivo verification of the central role of the Ca–P phase introduced by in situ mineralization in the hemostatic mechanism.

Figure shows the results of the hemostatic performance evaluation of gauze, BC, and BC-CaG₁₀₀ using a rat tail amputation model. As shown in Figure a,b, the gauze group showed the maximum blood loss of 532.1 ± 31.2 mg and the slowest hemostasis, requiring 496.5 ± 30.7 s to achieve hemostasis. The BC group showed significantly reduced blood loss and hemostasis time compared with the gauze group (p < 0.01), however, the results were still unsatisfactory, with a blood loss of 117.8 ± 10.6 mg and hemostasis time of 324.1 ± 10.6 s. In contrast, the BC-CaG₁₀₀ group had a significant advantage in hemostasis (p < 0.01), with a blood loss of 16.1 ± 2.1 mg, which was only 3% of that of the gauze group and 13.7% of the BC group. Hemostasis was the fastest and effectively controlled within 142.5 ± 7.3 s. These results indicate that BC-CaG₁₀₀ exhibits superior hemostatic efficacy among the three samples, confirming that mineralized BC-CaG₁₀₀ significantly promotes hemostasis at the wound site. ,

11.

Representative images of a rat tail amputation model with quantitative statistics of hemostasis time and blood loss. (a) Representative images of hemorrhage. (b) Hemostasis time. (c) Blood loss. Data are presented as mean ± SD. Statistical significance is indicated by asterisks (**p < 0.01; ns: not significant).

Microscopic morphological characterization of samples obtained from a rat tail amputation hemostasis experiment using SEM further elucidated the hemostatic mechanism of the material (Figure ). In the pure BC group, a regular and cross-linked nanofiber network structure was observed at all magnifications, with relatively smooth fiber surfaces and only a minimal amount of plasma protein adhesion. This indicates that pure BC primarily relies on the physical adsorption of the fiber network to absorb exudate and lacks active sites to actively induce blood coagulation, resulting in lower hemostatic efficiency. In the BC-CaG₁₀₀ composite group, an extremely strong interaction with blood components was observed. SEM images clearly show that a large number of erythrocytes were effectively retained within the BC composite network, and the fiber surfaces were covered with a dense, membrane-like distribution of coagulated plasma proteins, forming a robust thrombus scaffold structure.

12.

SEM images of the hemostatic dressings obtained immediately after the in vivo rat tail amputation model. (a) BC, and (b) BC-CaG₁₀₀ composite.

These results indicate that the HAP particles were not merely “blocked” but actively participated in the coagulation reaction. The mechanism behind its superior hemostatic performance is as follows: blood containing water rapidly enters the highly porous BC network, comes into contact with widely distributed HAP, and induces the release of Ca²⁺. As a key coagulation factor (Factor IV) in the coagulation cascade, free Ca²⁺ accelerates the production of thrombin and the conversion of fibrinogen to fibrin, thereby triggering rapid coagulation simultaneously within and outside the entire 3D network of the BC/Ca–P composites.

4. Conclusion

In conclusion, this study successfully developed a coral-inspired synchronous in situ biomineralization strategy to overcome the challenges of weak interfacial bonding and complex processing in the fabrication of BC/Ca–P composites. Unlike conventional physical incorporation or multistep postmineralization methods, this in situ approach utilizes calcium gluconate as a dual-functional precursor to achieve the concurrent biosynthesis of BC and deposition of Ca–P phases, thereby constructing a unique inorganic–organic architecture. Structural and chemical characterizations confirmed the uniform anchoring of Ca–P inorganic nanoparticles within the 3D network of BC and their strong integration with the organic matrix through interfacial interactions. This structural integration imparts distinctive physicochemical properties to the composites. Specifically, moderate mineralization (BC-CaG₁₀) significantly enhanced the mechanical strength through a synergistic reinforcement mechanism, whereas a high mineral content (BC-CaG₁₀₀) offered abundant bioactive HAP sites. In preliminary biomedical evaluations, the fabricated composites exhibited favorable biocompatibility. Moreover, BC-CaG₁₀₀ demonstrated efficient hemostatic performance in a rat tail amputation model, where the uniformly distributed Ca–P particles effectively accelerated the coagulation cascade. Our study provides a facile, scalable, and green route for preparing novel BC/Ca–P composites and demonstrates the promising potential of these biomimetic materials for biomedical applications, including tissue engineering and wound management. To advance clinical translation, future studies will focus on overcoming precursor solubility limits to increase calcium content, and tailoring the in vivo biodegradation behavior to match natural tissue regeneration.

The authors declare no competing financial interest.