Ti sheets (10 × 10 × 1 mm³) and Ti rods (Φ2 × 6 mm³) were used as substrate materials, and all specimens were pretreated to remove the oxide layer and surface contaminants. First, the Ti samples were degreased by ultrasonic cleaning in acetone solution for 30 min to eliminate surface oil. Then, the Ti samples were wet-polished using SiC sandpaper with progressively finer grits (#600, #800, #1200, and #2000) until the surface was smooth with no visible scratches. After polishing, the samples were sequentially ultrasonically cleaned in acetone, deionized water, and 75% medical ethanol for 30 min each, and finally dried in a constant-temperature oven at 30 °C for subsequent MAO treatment.

The MAO experiment was conducted in a specialized electrolyte cell, with Ti samples as the anode and stainless steel as the cathode. The electrolyte composition is presented in Table 1, and the electrolyte was continuously stirred during the oxidation process to maintain uniform temperature and composition. All reagents were purchased from Harbin Yanjie Technology Co., Ltd. The key MAO parameters were set as follows: applied voltage of 400 V, pulse frequency of 600 Hz, duty cycle of 8%, and oxidation time of 5 min.

The MAO-treated samples were subjected to MH post-treatment using an MH synthesis instrument (XH-800s, China). The samples were fixed in a hydrothermal reactor kettle and completely immersed in 0.5 mol/L NaOH solution. The MH treatment parameters were set as a temperature of 200 °C, microwave power of 800 W, and treatment times of 10 and 60 min, respectively. The detailed treatment procedures have been reported in previous studies^[29-31]. After the MH reaction, the samples were immediately removed, soaked in deionized water for 10 min to remove residual alkali, and dried in an oven at 30 °C for 2 h. For convenience of description, the samples were labelled as shown in Table 2.

The apatite-inducing ability of the samples was evaluated by soaking in simulated body fluid (SBF)^[38]. The samples were immersed in 30 mL of SBF solution, and the SBF was refreshed every 48 h to maintain a stable ion concentration. The soaking time was set to 3, 5, and 7 days, and the experiment was conducted in a constant-temperature incubator at 37 °C. The SBF solution was prepared by dissolving analytical-grade NaCl (8.030 g/L), NaHCO₃ (0.352 g/L), KCl (0.225 g/L), K₂HPO₄·3H₂O (0.230 g/L), MgCl₂·6H₂O (0.311 g/L), CaCl₂ (0.293 g/L), and Na₂SO₄ (0.0720 g/L) in deionized water. The pH value of the solution was adjusted to 7.4 using 6.06 g/L Tris hydroxymethylaminomethane ((CH₂OH)₃CNH₂) and 1 mol/L HCl. All reagents were purchased from Harbin Yanjie Technology Co., Ltd.

The antibacterial performance of the samples was evaluated using Escherichia coli (E. coli, Gram-negative) and Staphylococcus aureus (S. aureus, Gram-positive), which are the main pathogenic bacteria causing implant-related infections. The cultivation procedures for E. coli and S. aureus have been reported in previous studies^[20,39]. First, the bacterial strains were inoculated onto solid bacterial media and cultured at 37 °C for 24 h. Subsequently, the cultures were transferred to fresh solid media every 24 h. After two rounds of solid-phase cultivation, a small aliquot of viable bacteria was transferred to liquid bacterial media and incubated at 37 °C for an additional 24 h. The bacterial suspension was diluted with phosphate-buffered saline (PBS) to a concentration of 1.75 × 10⁵ cfu/mL for subsequent experiments.

The samples were placed in a 24-well plate and sterilized using an autoclave at 121 °C for 30 min after a 2 h immersion in 70 v/v% ethanol within a sterile culture hood. Then, 1 mL of the bacterial suspension was added to each well and incubated at 37 °C for 24 h without shaking. Subsequently, bacteria at a concentration of 1.75 × 10⁵ cfu/mL were seeded onto the samples, and the samples were placed back into the 24-well plate and incubated at 37 °C for 24 h. After incubation, the samples were rinsed with PBS to remove non-adhered bacteria, and the adhered bacteria were eluted by ultrasonic vibration for 5 min. The eluent was diluted to an appropriate concentration, spread on solid LB agar medium, and cultured at 37 °C for 24 h. The number of bacterial colonies was counted under an optical microscope, and the antibacterial rate (R) was calculated according to Equation (1):

(1) $$ \begin{equation} \begin{aligned} R=(\mathrm{A}-\mathrm{B}) / \mathrm{A}^{\star} 100 \% \end{aligned} \end{equation} $$

where A is the number of bacterial colonies for the blank group, and B is the number of bacterial colonies for the MAO and MHC05T60 samples.

In addition, the morphology of adhered bacteria on the sample surface was observed by SEM to further evaluate the antibacterial effect. The samples with adhered bacteria were fixed with 2.5% glutaraldehyde solution for 4 h, dehydrated using a gradient ethanol series (30%, 50%, 70%, 80%, 90%, 95%, 100%), and then dried and sputtered with gold for SEM observation.

The electrochemical corrosion behavior of MAO and MAO-Nb coatings was evaluated using a Reference 600 electrochemical workstation (Gammry, USA) in 0.5 mol/L NaOH solution. A conventional three-electrode system was adopted: the sample, with an exposed area of 0.3 cm², served as the working electrode, a platinum plate acted as the auxiliary electrode, and a saturated calomel electrode (SCE) was used as the reference electrode. Before the test, the sample was immersed in the electrolyte for 5 min to reach a stable open-circuit potential (OCP).

Potentiodynamic polarization curves were measured in the potential range from +1 V to -1 V (vs. SCE) with a scanning rate of 1 mV·s^-1, and the corrosion potential (E_corr), corrosion current density (I_corr), and polarization resistance (R_p) were obtained by fitting the polarization curves using Gamry Echem Analyst software. Electrochemical impedance spectroscopy (EIS) was performed in the frequency range of 10^-2 Hz to 10⁵ Hz with an alternating current amplitude of 10 mV, and the EIS data were fitted and analyzed using ZSimpWin software to determine the impedance parameters of the coating.

All experimental data were expressed as mean ± standard deviation ($$\overline{x}$$ ± s), and statistical analysis was performed using SAS 6.0 software. One-way analysis of variance (ANOVA) was used to compare differences between groups, and P < 0.05 was considered statistically significant. All in vitro experiments and surface characterizations were repeated at least three times to ensure reproducibility of the results. OriginPro 2018C and Microsoft Office 2024 were used to generate the experimental figures and tables.

The phase composition and surface microstructure of MAO, MHC05T10, and MHC05T60 samples were characterized by XRD and SEM, and the results are shown in Figure 1. As observed from the XRD patterns [Figure 1A], the MAO sample mainly exhibits diffraction peaks of the Ti substrate and a weak anatase TiO₂ peak, and no obvious HA phase diffraction peak is detected, indicating that the MAO coating is primarily composed of an amorphous structure and anatase TiO₂ phase. After MH treatment, the diffraction intensity of anatase TiO₂ in the MHC05T10 and MHC05T60 samples is significantly enhanced, and new diffraction peaks appear at 2θ = 31.9°, 33.1°, 46.7°, and 49.5°, which correspond to the (211), (300), (222), and (213) crystal planes of the HA phase (PDF: 73-0294), indicating that MH treatment can induce the formation of HA crystals on the Nb-doped MAO coating. With increasing MH treatment time, the diffraction intensity of HA phase peaks in the MHC05T60 sample is higher than that in the MHC05T10 sample, indicating that the content of HA crystals increases with prolonged treatment time.

Figure 1B presents SEM images of MAO, MHC05T10, and MHC05T60 samples. The MAO sample exhibits a typical porous surface with microscale dimensions, with pore sizes ranging from several micrometres to tens of micrometres, and the pore surface appears relatively smooth. After MH treatment, the MHC05T10 and MHC05T60 samples continue to exhibit a porous structure, with numerous short rod-like HA crystals observed on their surfaces. Magnified SEM images indicate that these crystals form a well-ordered HA submicron pillar structure on the MAO coatings, and the crystals are densely arranged. The length of the short rod-like HA crystals in the MHC05T60 sample is slightly greater than that in the MHC05T10 sample, which is consistent with the XRD analysis results, indicating that the growth of HA crystals is promoted with increasing MH treatment time.

To further investigate the microstructure and elemental distribution of the MHC05T60 sample, the results are shown in Figures 2-4. The cross-sectional TEM image in Figure 2A clearly shows that the MHC05T60 coating exhibits a typical three-layer structure: (I) an outermost growth layer with short rod-like HA crystals, (II) a middle porous layer containing anatase TiO₂ clusters and an amorphous structure, and (III) an interface layer with TiO₂ nanocrystals. In region I, Ca, P, and O elements are mainly concentrated in the outermost growth layer [Figures 2C-E]. In contrast, the middle porous region (region II) exhibits lower concentrations of Ca and P, with uneven distributions of Ca, P, Na, and Nb elements [Figures 2D-G]. Additionally, several nanocrystal clusters with higher concentrations of Ti and O elements are observed in the middle porous region, and similar features are present in adjacent regions of the porous layer [Figures 2B and C]. Furthermore, in region III, the boundary between the porous layer and the Ti substrate exhibits a strong bonding interface.

The magnified TEM image, along with the corresponding HRTEM and FFT images of regions II and III presented in Figure 2A, is shown in Figure 3. The magnified TEM image [Figure 3A] reveals that the average size of the crystal clusters in region II is about 50 nm. The HRTEM and FFT patterns [Figures 3B and C] show that the lattice fringes with d-spacings of 0.243 and 0.342 nm correspond to the (103) and (202) crystal planes of anatase, and the FFT pattern exhibits typical diffraction spots of anatase TiO₂ along the [010] zone axis [Figures 3B and C], indicating that the nanocrystalline clusters are anatase. The magnified TEM image further indicates that region III comprises an oxidized layer, an interface layer, and the Ti substrate, and the lattice fringes of the Ti substrate correspond to the (110), (10-1), and (011) crystal planes of Ti with the [-11-1] zone axis, as shown in Figures 3D-G. The metallurgical bonding between the interface layer and the Ti substrate is tight, without obvious gaps, which ensures high bonding strength of the coating.

The magnified TEM image and EDS line-scanning results of the outermost growth layer of the MHC05T60 coating are shown in Figure 4. The EDS line-scanning results [Figure 4B] show that the outermost layer can be divided into three typical regions: areas M1, M2, and M3. In area M1, the main elemental compositions are 20 at% Ti, 55 at% O, 20 at% Ca, and 2.5 at% P and 2.5 at% Na, while the atomic concentrations of Si and Nb are nearly zero. In area M2, the atomic concentrations of Ti and O decrease significantly to 7.5 at% and 47.5 at%, respectively, whereas those of Ca and P increase notably to 27.5 at% and 10 at%, respectively. The contents of Si, Nb, and Na are about 2.5 at% each. In area M3, the main elemental compositions are Ca, P, and O, along with small amounts of Si and Nb, with concentrations of 30 at%, 15 at%, 50 at%, 2.5 at%, and 2.5 at%, respectively. The HRTEM and FFT results show that area M1 exhibits an amorphous structure [Figures 4C and E], region M2 contains a polycrystalline HA phase [Figure 4C and D], and region M3 consists of a single HA phase with the [0-11] zone axis [Figures 4F and G], indicating that the outermost growth layer exhibits a gradient structure from amorphous phase to HA crystal.

Figure 7 shows the potentiodynamic polarization curves and electrochemical impedance spectra (EIS) of the MAO and MAO-Nb samples in 0.5 mol/L NaOH solution, and the results are summarized in Table 3. In Figure 7A, the corrosion potential (E_corr) of the MAO-Nb sample (-0.280 ± 0.014 V) is slightly lower than that of the MAO sample (-0.250 ± 0.0125 V), and the corrosion current density (I_corr) of the MAO-Nb sample ((5.64 ± 0.082) × 10^-9 A/cm²) is much lower than that of the MAO sample ((2.28 ± 0.035) × 10^-8 A/cm²). Furthermore, the polarization resistance (R_p) of the MAO-Nb sample ((7.09 ± 0.38) × 10⁶ Ω·cm²) is more than four times that of the MAO sample ((1.81 ± 0.21) × 10⁶ Ω·cm²), indicating that Nb doping significantly improves the corrosion resistance of the MAO coating.

The EIS test results [Figures 7B and C] are consistent with the polarization curve results. The Nyquist plots of the MAO-Nb coating show a larger capacitive arc radius (about 1,500 kΩ·cm²) than that of the MAO coating (about 750 kΩ·cm²), indicating that the MAO-Nb coating has higher charge-transfer resistance and a lower corrosion rate. Figure 7C shows the Bode modulus plots of MAO and MAO-Nb coatings. Previous corrosion studies indicate that the impedance modulus at low frequency can represent the polarization resistance^[40,41]. Therefore, the impedance modulus values at a frequency of 0.01 Hz can be considered as the polarization resistance (R_p). In this study, the impedance modulus value of MAO-Nb at 0.01 Hz (6.5 × 10⁶ Ω·cm²) is slightly higher than that of the MAO coating (6.25 × 10⁶ Ω·cm²), which further confirms that Nb doping improves the corrosion resistance of the MAO coating.

Figure 8 presents the colony-counting results and SEM images of the MAO and MHC05T60 samples against E. coli and S. aureus. In Figure 8A, the colony-counting results show that the number of viable bacteria on the MHC05T60 sample is significantly lower than that on the MAO sample. Based on these results, the antibacterial rates for S. aureus and E. coli on the MHC05T60 sample reach 96.1% and 87.3%, respectively, whereas those on the MAO coating are only 31.9% and 26.8%, respectively, indicating that the MH-treated Nb-doped MAO coating exhibits excellent broad-spectrum antibacterial performance.

As shown in Figure 8B, E. coli and S. aureus adhered to the MAO coating remain in a normal proliferative state with intact cell morphology. In contrast, after MH treatment, the number of adhered bacteria on the MHC05T60 sample is very low, and the remaining bacteria exhibit clear cell damage, with several deformed and lysed cells, which directly confirms the excellent antibacterial effect of the MHC05T60 sample.

MH treatment is a novel and efficient surface modification technique that introduces a microwave field into the conventional hydrothermal process. Its unique heating characteristics, such as volumetric heating, rapid heat transfer, and selective heating, enable rapid formation of HA crystals on the MAO coating surface. In this study, Nb was doped into the MAO coating by adding C₁₀H₅NbO₂₀ to the electrolyte, and MH treatment was conducted in NaOH solution to induce the formation of HA crystals. The formation mechanism of the three-layer structure of the MHC05T60 coating and the evolution of HA crystals are shown in Figure 13.

Based on the TEM results, the Nb-doped MAO coating is mainly composed of an amorphous structure containing Ca, P, Si, Na, and Nb elements, along with a small amount of anatase TiO₂ nanocrystal clusters, and its microstructure is similar to that reported in previous studies^[29-31]. During MH treatment, the microwave field acts as a heat source and rapidly heats the NaOH solution to the target temperature (200 °C), forming a high-temperature and high-pressure reaction environment in a specialized autoclave. Under these conditions, the incorporated Ca, P, and Si elements diffuse toward the coating surface, resulting in the formation of SiO₄^4-, Ca²⁺, and PO₄^3- species at the solution-coating interface. Hydroxide ions (OH^-) in the MH treatment solution facilitate HA nucleation by interacting with these species. The Si element in the coating forms SiO₄^4- at the interface, which acts as a heterogeneous nucleation site for HA crystals and promotes rapid nucleation. As the MH treatment time increases, HA nuclei are rapidly formed on the surface and subsequently grow in situ toward the solution, thereby forming a three-layer structure of the MHC05T60 sample, including an outer growth layer with rod-like HA crystals, a porous middle layer containing an amorphous structure and numerous TiO₂ nanocrystal clusters, and a dense interface layer.

Due to the interconnected porous structure of the MAO coating, some HA crystals are formed on the surfaces of internal pores. However, owing to the limitations imposed by the porous structure, only a limited amount of OH^- ions can diffuse into these pores, resulting in the formation of HA nuclei on the pore surfaces. Furthermore, the HA crystals observed in this study exhibit a short rod-like morphology, which differs from the nanowire structures with smaller diameters and greater lengths formed on the surface of MH-treated MAO coatings without Nb, as previously reported^[29-31]. Nb incorporated into the coating plays a key role in regulating the morphology of HA crystals. The Nb-doped MAO coating exhibits enhanced corrosion resistance in NaOH solution, suggesting a reduced reaction rate between OH^- and the MAO coating, which weakens the Ca and P concentration gradient at the coating-solution interface. As a result, fewer HA nuclei are formed on the Nb-incorporated MAO coating within a short time compared to the Nb-free MAO coating. Consequently, the formation of finer and longer HA crystals on the Nb-doped MAO coating is less favorable due to the insufficient Ca and P concentration gradient. In contrast, short rod-like HA crystals with larger diameters are formed, which may be more favorable for inducing bone formation, as this morphology resembles HA crystals in natural bone.