As shown in Figure 1, All the specimen powders and samples were made from PEEK material and its composites from Jugao-AM Tech. (Xi’an, China). After mixing the PEEK (550 PF), nHA and CF powders with the blender in the R&D lab, all the samples were prepared into 1.75 ± 0.5 mm filaments by using a twin-screw extruder. The porous PEEK scaffold modeling was designed with 3D Max software, and the composite scaffolds samples were manufactured using the FDM. The printing parameters were 420°C print temperature, 30 mm/s print speed, 0.4 mm nozzle diameter, and 0.2 mm layer thickness for filaments that were dried at 60°C for 8 h (Table 1). The scaffolds used in each group for the in vitro experiments had a diameter of 1.4 mm and a height of 2 mm, as illustrated in Figure 2A. The diameter of the porous scaffolds for compression of the sample are 1.4 mm and the height is 5 mm. The tensile experiment samples were 150 mm in length and 4 mm in height (Figure 4A), following ISO 527-2: 2012 requirements (Wang P et al., 2020). As shown in Table 2, different groups of porous scaffolds were named P, P-10C, P-10H10C, P-20H10C, and P-30H10C based on the variation of PEEK, nHA, and CF content.

Scanning electron microscopy (SEM, TESCAN MIRA LMS, Czech) and energy dispersive spectroscopy (EDS) were used to analyze the surface micromorphology and elements of PEEK, nHA, and CF powders and the different scaffolds. The scaffold groups were then sprayed with Pt before the samples were photographed using an SEM at a 3 kV accelerating voltage for the morphology images and a 15 kV accelerating voltage for the EDS shots. PEEK, nHA, and CF powders were dispersed in ethanol, ultrasonically dispersed for 20 min, dropped on silicon wafers, dried naturally, and the surface morphology was observed by SEM. The PEEK composite scaffold’s distinctive peaks were observed using X-ray diffraction (XRD, Rigaku MiniFlex600). We explored the special functional groups of the PEEK composite scaffolds using Fourier infrared spectroscopy (FTIR; Thermo Scientific Nicolet iS20) in the spectral scan range of 500–4,000 cm⁻¹. The contact angles of the porous scaffolds were photographed and recorded using an optical contact angle meter (Biolin Scientific Co., Ltd., Sweden). The surface roughness of the porous scaffolds was measured using atomic force microscopy (AFM, Dimension Icon, America). Thermogravimetric analysis (TGA, TGA55, America) was performed on a Netzsch thermal analyzer TG, and the samples were examined under an air atmosphere at a constant rate of 20°C min⁻¹ in the scanning range of room temperature to 900°C with an empty aluminum pan as a reference.

Tensile samples and 3D printed porous scaffolds were tested and analyzed in a universal testing machine (MTS Systems Ltd., China) at room temperature. The tensile speed of the samples is at a speed of 5 mm/s. Tensile strength was computed from the final data, and a stress-strain curve was shown. PEEK porous scaffolds of the compression speed are set at 0.5 mm/min. The load-displacement curve is recorded automatically, and the stress-strain curve is plotted based on the data obtained.

Essentially, bioactivity refers to the creation of a layer like that of HA on the surface of a substance, which predicts its adherence to bone. For the prepared 3D printed porous scaffolds, biological activity was evaluated after immersion for 7 days in Simulated Body Fluid (SBF, Solarbio). The scaffolds were gently rinsed with distilled water and dried overnight at 60°C. The precipitation of apatite particles formed on the surface of the scaffold was observed by SEM. In addition, MC3T3-E1 cells were evaluated for their biocompatibility and capacity for osteogenic differentiation with different scaffolds. In a cell incubator at a culturing temperature of 37°C in an atmosphere of 5% CO₂ and 95% humidity, cells were incubated in a complete medium (Dulbecco’s modified Eagle medium containing 10% fetal bovine serum and 1% penicillin and streptomycin). The osteogenic medium was prepared for the osteogenic differentiation assessments by adding 10 mM sodium glycerol-phosphate, 50 μg/mL L-ascorbic acid, and 10 nM dexamethasone medium to the above mediums. In addition, the scaffolds in each group were washed with acetone, ethanol, and sterile deionized water, sterilized in an autoclave, and dried before cell experiments (He et al., 2019).

All experimental data were statistically analyzed using GraphPad Prism (version 9.0) software. One-way analysis of variance (ANOVA) was used for multiple group comparisons and Tukey’s multiple comparison test was used between group comparisons. P < 0.05, differences were considered statistically significant.

In the conducted experiment, a variety of composite PEEK scaffolds were fabricated through the utilization of FDM technology. These scaffolds consisted of diverse combinations of nHA, CF, and PEEK, denoted as P, P-10C, P-10H10C, P-20H10C, and P-30H10C, as outlined in Table 2. As depicted in Figure 2A, the utilization of FDM technology in the fabrication of the hybrid stent structure results in the formation of an intricate three-dimensional configuration characterized by a grid-like structure composed of minute pores and channels. Figure 2E shows the surface morphology of PEEK, HA, and CF powders. PEEK powder has an irregular shape with a particle size of 20–70 μm. nHA has a needle-like morphology with a particle size of 50–400 nm, and agglomerates appear on the surface of the material. CF has a rod-like distribution with a length of 10–90 μm. The SEM pictures, when magnified by factors of 100, revealed the presence of a highly porous and linked network structure in five distinct sample groups (Figure 2B). Further, these photos exhibited intricate surface textures and intricate features. The results of the SEM analysis indicated that the surface of sample P exhibits a smooth texture. Nevertheless, it was observed that sample P-10C exhibit the existence of CF protrusions. The other three scaffold groups, on the other hand, demonstrated that with the addition of nHA, particles or protrusions developed on the scaffold surface, affecting the surface roughness. The average pore size of the hybrid stent was determined using ImageJ software, as depicted in Figure 2B. The P, P-10C, P-10H10C, P-20H10C, and P-30H10C obtained measurements indicated that the pore size ranges from 632.97 ± 21.85 μm, 637.25 ± 44.51 μm, 471.63 ± 99.66 μm, 560.23 ± 70.07 μm, and 554.43 ± 44.91 μm. The pore size data that has been acquired is of utmost importance in understanding the characteristics of the scaffold. This is because the pore size plays a critical role in facilitating vascularization throughout the process of tissue regeneration, as well as influencing cell proliferation and migration (Qian et al., 2023a). According to initial research findings, it has been shown that a hole size of approximately 400–600 μm is deemed optimal for facilitating complete mineralized bone development and vascularization, hence promoting the process of osteogenesis (Qin et al., 2022). Figure 2C shows the cross section of the porous scaffolds. The surface of the scaffolds is rougher except for group P, where the surface is dense and smoother and a cross-section of rod-like carbon fibers can be seen embedded inside the scaffolds. The distribution of white nHA particles can be seen for P-10H10C, P-20H10C, and P-30H10C.

The element distribution of five groups of PEEK scaffolds can be depicted by the process of element mapping in an EDS diagram (Figure 2D). Both PEEK and CF contain a significant amount of carbon (C) and small amounts of phosphorus(P). As a result, the materials P and P-10C exhibit regions with high carbon content. In the case of mixed PEEK incorporating nHA, the EDS analysis may reveal that calcium (Ca) and phosphorus (P) are the predominant element, alongside carbon. It is worth noting that Ca and P serves as the primary constituent of calcium phosphate (CaPO₄) inside the nHA composition.

The XRD patterns of CF powder, HA powder, and five groups of scaffolds are shown in Figure 3D. There were about four reflection peaks corresponding to the PEEK crystal structure, located at 18.8°, 20.7°, 22.7°, and 28.8°, respectively. The incorporation of CF did not influence the crystal structure of PEEK. In P-10H10C, P-20H10C, and P-30H10C, nHA diffraction peaks were seen. The typical modes of the nHA characteristic peaks include (002), (211), (300), (222), and (213) (Liu et al., 2019; Cui et al., 2021). The positions of these peaks coincide with the data of TCPDS Card No. 09-0432 on the standard peaks of HA. It was demonstrated that pure HA was present in P-10H10C, P-20H10C, and P-30H10C, indicating that nHA was successfully incorporated into the PEEK composite porous scaffolds (Zheng et al., 2022).

The FTIR spectra labeled in Figure 3E demonstrate the characteristic absorption peak of PEEK in all five composite scaffolds. It showed the carbonyl stretching vibration of C=O at a peak at 1,646 cm⁻¹. The peaks at 1,598 cm⁻¹ and 1,501 cm⁻¹ are caused by vibrations in the C=C benzene plane. As the nHA content increased, the intensity of the PO₄ ^3- and OH⁻ peaks increased (Zheng et al., 2022; Zheng et al., 2021). The peak of the P-10H10C, P-20H10C, and P-30H10C composite scaffolds was caused by vibrations at 1,047 cm⁻¹(PO₄ ^3-) and 3,575 cm⁻¹ (OH⁻). The thermal stability of the 3D printed porous scaffolds was analyzed by TGA. The TGA thermal spectra of each group of scaffolds are given in Figure 4F. When the temperature exceeded 470°C, the materials of each group began to decompose. The initial decomposition temperature of P and P-10C was 562°C. The initial decomposition temperature of P-10H10C, P-20H10C, and P-30H10C was 470°C. The initial decomposition temperature of P-10H10C, P-20H10C, and P-30H10C was 470°C. At 800°C, the residual weight percentage of P group scaffolds was 1.73%; at 900°C, the residual weight percentage of P-10C was 7.34%; and the residual weight percentages of scaffolds after addition of nHA were 21.05% for P-10H10C, 21.86% for P-20H10C, and 32.94% for P-30H10C, respectively. The decomposition temperature of the porous scaffolds was greater than the set temperature of 420°C for 3D printing, indicating that the prepared scaffolds did not decompose during the printing process and had good thermal stability.

As shown in Figures 3A, B, we analyzed the roughness of the material surface by AFM. The surface of the P scaffold exhibited a smooth texture. However, upon the addition of 10% CF powder, an internal network structure was generated within the scaffold. This is attributed to the interaction between PEEK and CF during the mixing process. Consequently, the surface micro texture of the P-10C scaffold was enhanced, accompanied by a little increase in roughness. Nevertheless, the incorporation of nHA particles could lead to the formation of particle distribution inside scaffold materials, resulting in various interactions between nHA particles and PEEK. The incorporation of nHA into a PEEK scaffold resulted in alterations in surface roughness, the impact might vary depending on the quantity of nHA added. The addition of nHA in P-10H10C, P-20H10C, and P-30H10C composites led to a moderation of the interfacial tension and an enhancement in material fluidity, as compared to P-10C alone. The increased roughness of P-20H10C might be attributed to the elevated concentration of nHA particles, which have the tendency to deposit or aggregate on the surface of the scaffold. This deposition and aggregation process leads to the formation of irregular protrusions, thus resulting in an augmented roughness of the scaffold surface. Nevertheless, the potential explanation for the decreased roughness seen in P-30H10C could be attributed to the presence of nHA particles, which have an impact on the unrestricted motion of fluid molecules and thus lead to a reduction in surface roughness (Wang et al., 2022).

The hydrophilicity analysis of the material is shown in Figure 3C. As the amount of nHA in the material increases, the hydrophilic angle of the surface decreases. P-20H10C exhibits the highest hydrophilicity in comparison to PEEK. It is possible to determine the reason for this phenomenon because, when exposed to water, nHA may form a hydrophilic coating on the material’s surface (Jang et al., 2020). This hydrophilic layer can lower the material’s surface tension, facilitating water diffusion on the material’s surface and enhancing the material’s hydrophilicity (Durham et al., 2016). At the same time, the hydrophilic layer can absorb the nearby water molecules, resulting in surface wetting and enhancing the material’s hydrophilicity even more. According to earlier research, the calcium and phosphate ions on the surface of nHA may induce biological reactions and cell signaling by attaching to receptors on cell membranes. These signaling systems can increase cell differentiation, proliferation, and secretory activity, which promotes cell adhesion and proliferation (Zhang et al., 2022; Zhong et al., 2022).

Good mechanical properties are essential for human implant materials. The analysis of the mechanical properties of each group of tensile samples was shown in Figure 4. The porous structure of the 3D printing reduced the elastic modulus of the PEEK material compared to prior research (Wang P et al., 2020). However, Figure 4B demonstrated that P-10C has the best mechanical qualities and the greatest tensile strength. In comparison to P, the yield strength of P-10C increased from 42.35 ± 0.42 MPa to 56.15 ± 0.54 MPa (Figure 4E). It is important to note that the mechanical properties of PEEK/CF were less affected by the addition of just a small portion of nHA. The results of sample compression are consistent with the results of tensile experiments (Figure 4D). However, as the nHA content increased to 30%, the tensile and yield strengths of the composites declined. This might be related to the binding mechanisms of PEEK and nHA. It has been demonstrated that adding too much nHA can cause stress concentration and brittle fracture at the interface between both materials (Kang et al., 2022). When the nHA particle content was excessively high, the interaction force between particles increased, which led to particle aggregation and accumulation (Kim et al., 2020; Hwangbo et al., 2022). This phenomenon results in a concentration of interface tensions between the nHA particles and the PEEK matrix, weakening the material’s strength and toughness.

In the cell live-dead staining experiment, after incubating the cells indirectly with the scaffolds for 3 days, a large number of live cells with green fluorescence and a small number of dead cells with red fluorescence could be seen in all groups of materials (Figure 5A). The ratio of living to dead cells showed no statistically significant variation (Figure 5D), demonstrating the PEEK composite scaffolds’ high biocompatibility, and the cells exhibited activity on the surface of the scaffold and demonstrated a high rate of proliferation.

The SEM images of cell adhesion experiments revealed that MC3T3-E1 cells adhered to the scaffold’s surface in all groups (Figure 5B). Different scaffolds displayed various patterns of cell adhesion. The majority of the cells in P and P-10C were connected in a spherical pattern. The cells in P-10H10C and P-20H10C showed filamentous pseudopods, which expanded in quantity and spread across the surface of the materials. The main reason was that adding nHA and CF to PEEK composites could improve their surface features, making them more biocompatible and bioactive. nHA could interact with cells to foster cell adhesion because it has a microstructure and chemical composition that is close to bone tissue. Because of its high specific surface area and superior mechanical qualities, CF could offer more areas of adhesion and support, which may be helpful in cell growth and adhesion (Rodzeń et al., 2021; Zhong et al., 2021). Figure 5C shows the precipitated particles formed by each group of scaffolds after 7 days of immersion in SBF. A large distribution of precipitated material was visible on the scaffold surface after nHA was added. By incorporating HA nanoparticles into the composite scaffolds, bone regeneration was promoted, which led to greater biomineralization capacity.

The CCK-8 assay was used to measure the proliferation ability of cells after 24h, 48h, and 72 h of culture, respectively (Figure 5E). The results indicated that P and P-10C increased less rapidly than the blank group. After 72 h of culture, cells with the nHA groups exhibited higher OD values, indicating the cells’ ability to proliferate most effectively. This confirmed that the addition of nHA could improve PEEK’s biological activity. Researchers have demonstrated that nHA could significantly roughen up the surface of the material, promoting cell proliferation and osteogenic differentiation (Chen et al., 2020). It was noted in the literature that Sari constructed porous HA scaffolds by precipitating hydroxyapatite and honeycomb (HCB) together, proving the material’s excellent capacity for cell proliferation and metabolic activity (Sari et al., 2021). Furthermore, nHA released calcium ions that can impact cell signaling, the production and breakdown of extracellular matrix, and the process of cell proliferation and differentiation.

The osteogenic differentiation of MC3T3-E1 cells on the material’s surface could demonstrate the material’s bone-enhancing capabilities. Previous research has focused on the osteogenic activity of small amounts of nHA (Chen et al., 2019). Using ALP qualitative and quantitative assays, ARS staining and quantitative analysis, and PCR experiments, we confirmed that PEEK composite porous scaffolds have great osteogenic differentiation ability. Through the utilization of ALP qualitative and quantitative assays, ARS staining and quantitative analysis, as well as PCR investigations, we have successfully validated that PEEK composite porous scaffolds exhibit a significant enhancement in the capacity to promote cellular osteogenic differentiation.

We cultivated MC3T3-E1 cells for 7 and 14 days to evaluate their ability to differentiate into osteoblasts. The MC3T3-E1 cells were subjected to osteoblast induction media for a duration of 7 and 14 days in order to assess their early-stage osteoblast differentiation potential. At day 7, the difference in ALP activity between P-10H10C and P-20H10C was not statistically significant, but they were both higher than that of the blank group and the P group (Figures 6A, C). After 14 days, ALP staining revealed that the surface colors of the early osteogenic differentiation of MC3T3-E1 cells in the porous scaffolds in the various groups were darker than those of the cells at day 7. P-20H10C was more apparent in cells (Figure 6A).

ARS staining revealed that MC3T3-E1 cells had late osteogenic differentiation in different porous scaffolds (Figure 6B). At 14 days, only a small amount of calcium nodule deposition was observed in the pore plates. At 21 days, a significant increase in calcium nodule deposition was observed with the increase of nHA content. ARS Quantitative analysis verified this (Figure 6D). By using the FDM technology, Babilotte combined nHA with medical-grade poly (lactic-co-glycolic) acid (PLGA) and discovered that the material with the additional nHA had a greater level of cell mineralization and a significant increase of calcium nodule deposition (Babilotte et al., 2021). This further demonstrated nHA’s capacity to promote matrix mineralization and osteoblast development.

We assessed the expression of osteogenic genes in the osteogenic differentiation of MC3T3-E1 cells. Runx-2 and ALP were indicators of early osteoblast differentiation. Runx-2 was primarily involved in the transcription of osteogenic genes (Pérez-Campo et al., 2016). The expression levels of Runx-2 in each scaffold group on days 7, 14, and 21 are shown in Figure 7A. The expression levels of P-10H10C and P-20H10C were higher than P, which was statistically different. ALP primarily participates in phosphate hydrolysis processes and encourages bone matrix mineralization (Sardiwal et al., 2013). ALP activity is represented as a biomarker of bone formation; it primarily participates in phosphate hydrolysis processes and encourages bone matrix mineralization. It can also indicate the stage of osteoblast development and activity. The expression of the ALP gene reached its highest level on day 14, and the expression decreased on day 21 (Figure 7B). Col-1 is a protein that predominantly promotes osteoblast adhesion and differentiation, facilitating bone repair (Bou Assaf et al., 2019). At day 21, the cellular expression of the P-10H10C, P-20H10C, and P-30H10C scaffolds was higher than that of P, which might be attributed to the inclusion of nHA, which roughened the surface of the scaffolds (Figure 7C). This might be related to the ability of nHA to offer a surface for cell adhesion and penetration, giving MC3T3-E1 cells an environment favorable for cell adhesion, proliferation, and differentiation. In this situation, Col-1 can be generated more frequently by MC3T3-E1 cells. By increasing collagen matrix synthesis, nHA indirectly influences Col-1 expression. However, the large amount of agglomeration of 30% nHA in the composites restricts the interaction of cells with the material and thus reduces the osteogenic differentiation ability of the cells. In addition, OCN is a marker indicating the maturation stage of osteoblasts (Li F et al., 2022). It is crucial for maintaining the mechanical characteristics of bone and bone density and was primarily engaged in the mineralization process of the bone matrix. As the amount of time increased, OCN expressed itself proportionately (Figure 7D). The expression of osteogenic genes like Col-1 and OCN was indirectly impacted by nHA because it enhanced Runx-2 expression and activity, which in turn encouraged the differentiation of MC3T3-E1 cells toward osteogenesis. These findings supported the ALP and ARS results, which indicate that P-10H10C, P-20H10C, and P-30H10C all enhance osteogenic differentiation, with P-20H10C having the strongest effects. Therefore, it could be assumed that the addition of nHA particles stimulated the expression of relevant osteogenic genes and promotes osteoblast mineralization.