Calcium chloride (ACS reagent, ≥96.0%), sodium dihydrogen phosphate (ACS reagent, ≥98.0%), sodium hydroxide (ACS reagent, ≥97.0%), polyvinyl alcohol (PVA, Mw 9,000–10,000), and sodium alginate (Alg) were all purchased from Sigma-Aldrich, St. Louis, MO, United States. Deionized water (18.3 MΩ cm) was used to prepare all aqueous solutions. All ILs 1-butyl-3-methylimidazolium bromide [C₄mim]Br; 1-dodecyl-3-methylimidazolium bromide [C₁₂mim]Br; and 1-tetradecyl-3-methylimidazolium bromide [C₁₄mim]Br were prepared and purified using the similar procedure described previously by Zhao et al. (2009), and their structure was illustrated in Figure 2A.

The mineralization of HAP crystals was carried out using the hydrothermal precipitation method similar to the literature (Hao et al., 2013). In all experiments, the initial Ca²⁺ and PO₄ ^3- concentrations were fixed at 0.1 M and 0.06 M, respectively. CaCl₂ and NaH₂PO₄•·2H₂O were used as Ca²⁺ and PO₄ ^3- reservoirs, respectively. A series of CaCl₂/NaH₂PO₄.2H₂O/ILs solution was prepared, and the concentration of ILs was fixed at 10 mg mL^-1. In a typical procedure, 1.0 g [C₁₄mim]Br and 0.48 g NaOH were dissolved in 50 mL water. Subsequently, a solution of CaCl₂ (25 mL, 0.1 M) and the other solution of NaH₂PO₄.2H₂O (25 mL, 0.06 M) were added. The obtained mixtures were stirred using vortex mixing (IKA, Vortex, Genius 3) for 10 min. Then, the resultant mixtures were transferred to a 200 mL teflon-lined stainless autoclave container at 200°C for 7 h. After that, the container cooled naturally to room temperature prior to harvest the crystals by centrifugation. Finally, the collected crystals were rinsed several times with deionized water and air-dried for further analyses. All experiments were repeated at least twice. A control experiment was also carried out in the absence of ILs. The feed composition and designations of the obtained apatite crystals in the absence or presence of ILs are listed in Table 1.

The morphologies of the products were investigated by scanning electron microscopy (SEM, Hitachi, S4800, Tokyo, Japan). To identify the composition of the synthetic products, Fourier-transform infrared spectroscopy (FTIR) was performed by using a SHIMADZU spectrum system (SHIMADZU, Kyoto, Japan) with a resolution of 4.00 cm⁻¹. X-ray diffraction (XRD) patterns were collected on a SHIMADZU X-lab 6000 X-ray powder diffractometer with Cu Kα radiation to confirm the phase and structure of the formed crystals. High-resolution transmission electron microscopy (HRTEM, FEI, Tecnai G² F20 STwin, USA) was adopted to investigate the microstructure together with the Fourier-transform (FFT) method. The differential scanning calorimetry (DSC) curves were performed on TGA/DSC1/1,100 (Mettler Toledo, Switzerland).

To prepare the hydrogel, 0.53 g PVA was dissolved in 10 mL of distilled water and placed in an oven (80°C) for 1 h according to the previous work (Zhang et al., 2020; Zhang et al., 2021; Ren et al., 2023). Then, 0.1 g Alg was added into the PVA solution and the mixture was stirred for 1 h under constant temperature of 70°C to form a homogenous solution. HAP@H was synthesized by adding 0.05 g lyophilized HAP powder into the obtained solution under stirring.

Rheological performances of different groups, including PVA solution, PVA-Alg hydrogel, and HAP@H, were carried out by investigating the changes in flowability of tilted small bottles according to the reported work (Chen et al., 2020; Ding et al., 2021). Macroscopic morphologies of pure hydrogel and HAP@H were observed. In order to investigate the porous structure of the obtained hydrogel, the freeze-dried gel was sliced and observed by micrography. The tensile and elastic properties of HAP@H were also tested by stretching and knotting treatments. Compression and swelling properties were performed on the hydrogel and the lyophilized hydrogel.

The structures of the ILs ([C₄mim]Br, [C₁₂mim]Br, and [C₁₄mim]Br) used in controlling HAP crystallization are illustrated in Figure 2A. The alkyl chain length of imidazolium cations ranged from C4 to C14. A control experiment was also carried out in the absence of ILs. The feed composition and designations of the groups are listed in Table 1.

The SEM image of HA crystals templated by [C₁₄mim]Br (ILHA-3) is shown in Figure 2B, and it is clear that these ILHA-3 were nanofibers with lengths ranging from 20 to 50 μm. A single ILHA-3 fiber is shown in Figure 2C, as indicated by the arrows. This fiber had a length of about 45 μm and a diameter of about 0.3 μm. These results demonstrated that [C₁₄mim]Br could successfully template HAP into highly c-axis-oriented nanofibers with length of tens of microns.

In order to investigate the effect of the alkyl chain length of ILs cations on the mineralization of HAP, the morphology of the samples from the groups (listed in Table 1) of HA-0, ILHA-1, ILHA-2, and ILHA-3 was characterized by SEM and TEM (Figure 3). In the blank control group (HA-0), the products were aggregates of nanoneedle (about 300 nm in length and 40 nm in width). This result suggests that it is difficult to get high-aspect ratio HAP crystals in the absence of the crystallization template. Also, in the group of ILHA-1, flake-shaped HA crystals were obtained. This result suggests that [C₄mim]Br could inhibit the growth orientation of the a-axis; thus, the HAP would grow along the b-axis and c-axis orientation to form thin flake-shaped crystals. Figure 3C indicates that the HAP crystals obtained in the ILHA-2 group were mainly conjugated fibers. It could be inferred that these conjugated fibers might be generated by incomplete flaky crystal cracking. Also, this was proved by Figure 3D, indicated by the arrows, that the front end of the flaky crystal cracked into fibrous crystals. Moreover, in the ILHA-3 group, the SEM image in Figure 3E proved that the obtained HAP crystals were nanofibers. The HRTEM image in Figure 3F further suggests that the <001> orientation (c-axis) and 0.342 nm crystal lattice are ascribed to (002) crystal plane. These results were further supported by Fourier-transform diffractogram inserted in Figure 3F.

Figure 4 provides the IR and XRD characterization of the samples. The IR spectra in Figure 4A indicate that the absorption peaks at ∼3,500 cm^-1, which are ascribed to the OH of HAP crystals, had an obvious increase in intensity from group HA-0 to ILHA-3. The increased intensity of OH absorption usually indicates an increase in crystallinity or orientation of HAP. This inference was further proved by the XRD patterns shown in Figure 4B. It is clear that the intensity of the (002) crystal plane had been obviously increased, which signifies the increased orientation of HAP along the c-axis (<001>).

In order to further discuss the role of the ILs template in controlling the crystallization of HAP, the schematic crystal growth processes of HAP nanoflakes templated by [C₄mim]Br and nanofibers templated by [C₁₄mim]Br are shown in Figure 5. It is suggested that the [C₄mim]Br might have adhered to the surface of the crystals; therefore, it could inhibit the growth along the a-axis. Thus, the ILHA-1 crystals, controlled by [C₄mim]Br, grew along the b-axis and c-axis to form nanoflakes. However, for using [C₁₄mim]Br as a template, these ILs would inhibit the growth both along the a-axis and b-axis; thus, the ILHA-3 group was controlled to grow along the c-axis to form nanofibers.

Rheological performance testing of PVA solution, PVA-Alg hydrogel, and HAP@H was carried out by testing flowability changes in tilted small bottles, as shown in Figure 6. It is clear that the pure PVA solution had good flowability (Figure 6A). Although the freshly prepared pure hydrogel also presented flowability (Figure 6B), the gel hanging on the bottom and wall of the bottle indicates that the viscosity of gel was much higher than that of pure PVA. It should be noted that pure hydrogel was still not solidified and needed aging time. However, the newly prepared HAP@H presented rapid gel curing performance and did not require a long gel aging time (Figure 6C).

In order to investigate the mechanism behind the rheological performance difference between pure hydrogel and HAP@H, the DSC test was performed on the samples. The DSC curves (Figure 7) proved that HAP@H had an increase (from 98°C to 107°C) in the water loss temperature compared to pure hydrogel. This might suggest that the strong interaction of hydrogen bonds between the HAP crystals and hydrogel network molecules enhanced the moisture retention ability.

After gel aging, pure hydrogel and HAP@H, placed in 24-well plates, presented several differences, as shown in Figure 8. First, HAP@H presented a more even and compact hydrogel appearance than that of pure hydrogel. Unlike the compact and uniform appearance for HAP@H shown in Figure 8B(3) and Figure 8B(4), the photos for pure hydrogel showed a less compact (Figure 8A(1)) and not heterogeneous appearance (Figure 8A(2)). Also, these appearance differences were further proved by the lyophilized gel’s morphology comparison. Figure 8A(5,7) shows that the lyophilized pure hydrogel did not have a uniform porous structure, while lyophilized HAP@H presented a good and uniform porous structure, as shown in Figure 8B(6, 8).

The tensile and elastic properties of HAP@H were also tested by stretching and knotting treatments, as indicated in Figure 9. The HAP@H in 3.0 cm length could be starched to the length of 10.5 cm. After releasing, the sample restored to the original 3.0 cm length (Figures 9A,C). Also, the knotting treatment for HAP@H (shown in Figures 9B,D) revealed that the knot–untie operation did not change the shape and length of the gel sample.

To investigate the compression properties of the obtained hydrogel, the obtained pure hydrogel and HAP@H were tested by a 200 g weight, as presented in Figure 10. It can be observed that the hydrogel sample had an original size of 16 mm in diameter and 18 mm in height [shown in Figure 10A(1)]. However, after the weight having compressed on the sample, Figure 10A(2) indicates that the gel block was crushed. After removing the weight, Figure 10A3) shows the gel sample’s structure had already been destroyed. The detailed information has been presented in Figure 10A(4) and Figure 10A(5). Unlike the structure of pure hydrogel block, which was crushed, after being compressed, the HAP@H gel block (indicated in Figure 10B) remained a complete structure. Except for the water in the gel being extruded, the structure of the gel remained intact.

Figure 11 provides the swelling properties of the freezing dried hydrogel block. It is clear that the pure hydrogel swelled a lot, to a rate of 627%, by adsorbing plenty of water (Figure 11A (1–5) and Figure 11B). Especially, it was difficult for the swollen gel block to maintain a complete shape. After being squeezed (200 g weight), most of the water would be discharged (Figure 11C). Also, at the same time, the structure of pure hydrogel was broken. As for the lyophilized HAP@H, it had an appropriate swelling rate (Figure 11B). Furthermore, the photos in Figure 11A (6–10) clearly indicate that the HAP@H gel block maintained an intact shape under pressure of 200 g weight, and only little water had been squeezed out (Figure 11A(10) and Figure 11C). Especially, compared with Figure 10B2), Figure 11A7) clearly proves that the lyophilized HAP@H presented stronger compressive mechanical properties than freshly prepared HAP@H.