Additively Manufactured Biomedical Ti-15Mo Alloy with Triply Periodical Minimal Surfaces and Functional Surface Modification

Abstract

Ti and Ti alloys are being widely used as bone tissue repair materials. Progress on mechanical properties and bio-functionality is required for their applications due to the large difference in elastic modulus between bone and Ti implants and the fact that the Ti materials themselves are biologically inert. In this work, a low-modulus, β-phase Ti-15Mo alloy based on a triply periodical minimal surfaces (TPMS) structure was fabricated using a Powder Bed Fusion-Laser Beam (PBF-LB) under optimized printing parameters into implant samples with controllable porous structures. The selection of TPMS, lattice unit cell size, and relative density was based on a combination of mechanical properties and cytocompatibility. Surface modifications were used to further impart antibacterial, antioxidant, and osteogenesis properties to the implants. Broad-spectrum antibacterial Ag, antioxidant tannic acid (TA), and highly stable fluorinated hydroxyapatite ((F)HA) were applied as an advanced coating on a microporous TiO₂ surface modification layer formed by micro-arc oxidation. Ultimately, porous Ti-15Mo implant samples with a biofunctional coating were obtained with Young’s modulus 15–50 GPa, a yield strength of approximately 100 MPa, and good cytocompatibility, hemocompatibility, and bactericidal effects. This study provides a systematic scheme for the preparation and surface modification of β Ti alloy implants for subsequent studies.

1. Introduction

Titanium (Ti) alloys have excellent biomechanical properties and good biocompatibility and are commonly used as hard tissue grafting materials in medical applications. However, the widely used Ti-6Al-4V alloy and commercially pure (CP) Ti have a stress shielding effect due to the mismatch between their elastic modulus of approximately 105–110 GPa and the bone tissue [1,2,3], which can lead to degradation of the original bone tissue and ultimately to implant failure [4]. Moreover, Ti alloys are biologically inert materials and have difficulty forming strong biological bonds with bone tissue. β-phase Ti alloys, on the other hand, are considered to have great potential to become a new generation of Ti alloys for medical applications due to their low modulus of elasticity, high biocompatibility, and corrosion resistance [5], of which the Ti-Mo family is represented by Ti-15Mo, with an elasticity modulus of 75–80 GPa [6,7,8].

The elastic modulus of the biomaterial is key to the force distribution and stress shielding effect of the implant, and in this regard, porous structures are ideal candidates. Compared to dense materials, porous structures greatly reduce stiffness [9,10] and are particularly suitable for applications in β Ti alloys where the modulus still does not match the modulus of human bone (0.022 to 21.8 GPa) [11]. At the same time, porous structures with through holes also facilitate biological cell behavior, such as cell adhesion, proliferation, and bone tissue regeneration [12,13,14,15], which can lead to greater biological bonding between implant and tissue, thus enhancing implant fixation and improving implant success. Additive manufacturing techniques, such as Laser Based Powder Bed Fusion (PBF-LB) are considered to be one of the most favorable methods for the fabrication of complex porous structures for artificial bone implants, as they are capable of producing alloy components with complex internal structures [16,17,18]. Compared to conventional subtractive manufacturing methods, PBF-LB offers the flexibility to modulate the porosity of porous alloys and to produce complex shapes, facilitating the creation of customized porous bone implant scaffolds.

In fact, the inner surface of real bone is mostly composed of irregular surfaces, and studies have shown that the surface curvature of the scaffold microstructure has an important impact on the regeneration process of bone tissue [19,20]. Bidan et al. [21,22] proposed that the average curvature of the structural surface in trabecular bone is approximately 0, which has drawn attention to the submicroscopic structure of the bone and the method of constructing porous bone implant scaffolds with minimal curved surfaces. Unlike conventional microstructures in porous scaffold designs, since the curvature on triply periodical minimal surfaces (TPMS) is 0 everywhere, the local surface energy of TPMS bone scaffolds is minimal [23], and the thermodynamic state is stable [20]. An implant of TPMS lattice has good connectivity, relatively high specific surface area, and relatively tortuous internal channels. Studies have shown that when the porosity of the implant of TPMS lattice is similar to that of trabecular bone, the structure has a higher yield stress and a lower elastic modulus [24]. In addition, functional gradient-type scaffolds with morphological features such as variable density can be generated by adjusting the surface equations [25,26]. The best pore size, however, has been determined differently by various studies. For instance, in a biomaterial structure, pores between 100 and 350 μm were deemed to be appropriate for cell colonization and vascularization and aiding bone tissue penetration [27]. Capillary growth will be encouraged in the bone scaffold by pores larger than 140 μm [28], and cell proliferation will be aided by pores larger than 300 μm [14]. According to research by Murphy et al. [29], bone production peaked at 325 μm pore size. When the pore size was between 300 and 1200 μm, bone production did not differ noticeably, as demonstrated by Hollister et al. [30]. Moreover, rats with femoral defects experienced a positive healing effect through a cylindrical bone scaffold of 80% porosity with pore sizes between 400 and 600 μm [31].

In addition to modification of the substrate of Ti alloy implants, surface modification and loading with various functional physiochemical coatings to improve osseointegration has been the focus of research [3,32,33]. Ti surface roughening and micro-nano treatment can effectively improve the bonding performance with bone tissue [34,35]. Micro-arc Oxidation (MAO) is a new modification technique based on the anodic oxidation of valve metals (Al, Ti, Ag, Nb, etc.) to produce a layer of porous TiO₂ ceramics grown in situ on the surface [35,36,37]. It uses a high working voltage and introduces the working area into the high voltage discharge zone, instead of the Faraday zone of anodic oxidation [38]. The TiO₂ ceramic layer, which is firmly bonded to the substrate, can be modified by adjusting the power supply parameters and electrolyte formulation to change the hardness, wear resistance, and roughness of the Ti alloy surface, thus improving the bonding properties to the bone tissue [39,40].

To prevent bacterial adhesion, antibacterial bioactive coatings can be prepared on the surface of Ti implants. Nano Ag has broad-spectrum antimicrobial properties, but it has a dose-dependent cytotoxicity, so the appropriate concentration is important for striking a balance between biocompatibility and antimicrobial properties. Hydroxyapatite (HA) has good bioactivity and osteogenic properties and can be compounded with various antimicrobial agents. Studies have shown that the doping of HA with F can improve its stability [41,42,43,44], and the corrosion resistance of fluoride-hydroxyapatite ((F)HA) increases with increasing F content [45]. Cellular experiments have shown that F-doping can improve the biological properties of HA [46,47], and a small amount of F ions can improve osteoblast activity [48] and promote cell proliferation [45]. Coating Ag nanoparticles with HA reduces the risk to human tissues while maintaining the good antimicrobial properties of Ag. Hence, in order to further improve the osteogenicity on a cell adhesion scale and make the bonding of HA and Ti stronger, microporous TiO₂ prepared by MAO is designed as an intermediate layer between Ti alloy and coating. In the presence of inflammation or certain pathologies (e.g., diabetes), excessive production and accumulation of reactive oxygen species (ROS) can trigger an oxidative stress response that can affect the biological performance of surrounding osteoblasts, which in turn can lead to deterioration and failure of the implant. Tannic acid (TA) with many pyrogallol groups as part of a composite coating can increase the antibacterial and antioxidant capacity of the specimen, which has been used in the dental field [49,50,51].

In this paper, pre-alloyed Ti-15Mo powder was used, and TPMS lattices were designed and selected with full consideration of the mechanical properties and biocompatibility. After PBF-LB preparation and annealing, the microstructure and mechanical properties of the samples were characterized. Samples with superior properties were selected for surface modification to enhance bio-functionality. In order to impart antioxidant properties to Ti alloy implants, this study prepared a layer of TA-Ca complex on the surface of MAO produced microporous TiO₂ as a calcium site for (F)HA formation using strongly reducing TA as an antioxidant, and then induced the formation of (F)HA coating by a hydrothermal method; the prepared coating is expected to have good antioxidant activity and biological properties. Considering the antimicrobial and anti-infective capacity, Ag nanoparticles are covered by (F)HA in the composite coating to create a slow release effect.

2. Materials and Methods

2.1. Raw Materials

The Ti-15Mo powder (Figure 1) used in this experiment was purchased from Northwest Institute for Non-ferrous Metal Research (Xi’an, China) and produced by the gas atomization method. A laser particle size analyzer was used to assess the particle size, and the results revealed that the particle size was Dv (50) = 38.4 μm. The as-cast sample used for comparison was taken from the powder supplier (Keruikerui Co., Ltd., Wuxi, China).

2.2. Lattices Design and Samples Design

Uniform TPMS structures were designed by MSLattice 1.0 [52], a software based on MATLAB R2020a, applying the mathematical description of those surfaces to produce 3D models. The nodal coordinates that represent a minimal surface have been described using a number of different mathematical techniques. The level-set approximation strategy, the most straightforward and popular technique, uses a set of trigonometric functions satisfying F(x,y,z) = const (constant for an iso-surface). The following are the level-set equations describing the minimal surfaces chosen in this work. Figure 2a represents the design and preparation process of the porous Ti-15Mo lattices. Both solid and sheet lattices derived from the surfaces are shown. Gyroid (G), sheet IWP (ISH), and solid IWP (ISO) structures were chosen because studies [53,54] have shown that they use less material and have better mechanical properties under the same unit cell size.

$$\mathrm{G}\mathrm{y}\mathrm{r}\mathrm{o}\mathrm{i}\mathrm{d}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\sin X\cos Y+\sin Y\cos Z+\sin Z\cos X=const$$

(1)

$$\mathrm{I}\mathrm{W}\mathrm{P}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}2\left(\cos X\cos Y+\cos Y\cos Z+\cos Z\cos X\right)-\left(\cos 2X+\cos 2Y+\cos 2Z\right)=const$$

(2)

where X = 2απx, Y = 2βπy, Z = 2πγz, in which α, β, and γ are constants associated with the lattice unit cell parameters in the x, y, and z directions, respectively.

When the above unit cell size was set to 1.5 mm and the relative density (RD) was set to 30–40% (Figure 3a), the lattice derived using MSLattice presented a suitable layer thickness and pore size. As introduced before, these lattices facilitate cell growth with the desired strength at lower densities and lower moduli, while being within the limits of the MSLattice design and PBF-LB fabrication. Samples of various shapes for testing mechanical properties and performing in vitro experiments were designed after determining the lattices. Figure 3b,c show two sets of the designs of samples. The samples are named by lattices and densities for G30, G35, G40; ISH30, ISH35, ISH40; ISO30, ISO35, and ISO40, respectively. It needs to be mentioned that based on the NumPy library, which supports complex computations in multiple dimensions, some programs were written in Python 3.8.8 to parse .stl files to calculate the surface area of the .stl models by summing up all polygonal regions of the mesh.

2.3. PBF-LB Processing and Process Optimization

The PBF-LB printer used in the experiments was an EOS M290 3D printer (EOS GmbH Electro Optical Systems, Munich, Germany) equipped with a 400 W fiber laser (1064 nm). Its spot diameter is about 70 μm, and it can prepare materials in the range of 250 × 250 × 325 mm³ in a high-purity argon atmosphere. The interlayer rotation angle was 67°, and the scanning strategy was Chess Scan. The volume energy density (VED) was used for understanding the laser energy input. Equation (3) defines the calculation of this parameter [55]:

$$VED={\displaystyle \frac{P_L}{v_s\times h_s\times t_l}}$$

(3)

where P_L is laser power, v_s is scanning speed, h_s is scanning pitch, and t_l is single-layer layer thickness.

The layer thickness was set to 50 μm according to the average powder particle size. After multiple trials, the optimum sample relative density of 99.3% was determined when printing a dense bulk sample utilizing the drainage method. At this moment, the corresponding VED = 100 J/mm³, and the lattice’s VED was designed for 40% of it. The optimized printing parameters are shown in Figure 3a.

All as-printed samples were annealed in a vacuum heat treatment furnace (MH(BR)-669-HV, FULCRUM, Shanghai, China) at a rate of 5 °C/min to 500 °C, held for 3 h [56,57]. After the samples were naturally cooled to 300 °C in the furnace, Ar was passed through the furnace to reduce to room temperature.

2.4. Microstructural Analysis

An optical metallographic microscope (OM) (Zeiss, Oberkochen, Germany) was used for the metallographic observation, which also required polishing and etching before observation. Afterwards, the alloy surface was etched using Kroll solution (3 mL HF:5 mL HNO₃:100 mL H₂O). The samples were subsequently ultrasonically cleaned with anhydrous alcohol and dried completely before further study by microscopy.

For the X-ray diffractometer (XRD, SmartLab, Rigaku, Tokyo, Japan) analysis, the 2θ angle range was set to 30–80°, the step size was set to 0.02°, and the test speed was set to 6°/min. The grazing incidence small-angle X-ray scattering (GISAXS) patterns were obtained using the same X-ray diffractometer and 2 h scanning from 10–30°. The morphology of the unit cells of the printed samples and sample microstructures was examined using the secondary electron mode of a TESCAN MIRA3 (TESCAN, Brno, Czech Republic), HITACHI SU8230 (Hitachi High-Tech Corporation, Tokyo, Japan), and Zeiss Merlin (Carl Zeiss AG, Oberkochen, Germany) scanning electron microscope (SEM, accelerating voltage 5 kV, operating current 100 pA). The energy-dispersive X-ray spectroscopy (EDS) mode (accelerating voltage 25 kV, operating current 2 nA) was used to analyze the surface chemical composition.

The electron backscatter diffraction (EBSD) accessory in the Zeiss Merlin SEM characterizes the phase distribution, grain size, and grain boundary information at a 20 kV voltage, 5000 pA current, and 15 mm working distance. The test data were analyzed by the OIM Analysis 7 software 7.3.1.1.

2.5. Mechanical Property Testing

With the use of the HX-1000TM/LCD automatic turret microhardness tester (Shanghai optical instrument factory, Shanghai, China), the specimens’ hardness was determined. In the studies, 100 gf of load and 15 s of loading time were the parameters employed. To lessen the experimental error, each sample was measured three times. To mitigate the impact of surface contaminants and scratches on the measurement results, the sample’s surface was ground, polished, and ultrasonically cleaned with anhydrous ethanol before the hardness test was conducted.

A universal mechanical testing equipment (Instron 68TM-30, Norwood, MA, USA) with a non-contacting video extensometer was used to conduct a room temperature tensile and compression test. Figure 3b depicts the form of the tensile specimens. The elongation gauge of the tensile and compression sample used in the experiments has a scale of about 30 mm, and the set strain rate for the experiments was 0.002/s. Triplicate specimens were prepared for each distinct structure. The tensile bars’ fractography images were detected by SEM.

2.6. Surface Modification and Characterization

2.6.1. Micro-Arc Oxidation (MAO) Process

Small cylindric samples measuring 6 mm in diameter and 4.5 mm in height (Figure 3c) with 30% RD were used for surface modification. Prior to treatment, each sample was ultrasonically cleaned for 10 min in acetone, alcohol, and deionized (DO) water. Temperatures were kept below 40 °C throughout the MAO process using the porous Ti-15Mo samples, which served as the cathode and a stainless-steel container as the anode. In this experiment, the MAO was carried out in a 10 g/L (NaPO₃)₆ (Sinopharm Chemical 5Reagent, Shanghai, China) basic electrolyte solution [58] with the following parameters: 450–500 V voltage, 10% duty ratio, 500 Hz pulse frequency, and 12 min oxidation time. The oxidized samples (denoted G(M) group, ISH(M) group and ISO(M) group) were then cleaned using ultrasonic vibration and distilled water for 15 min, followed by an overnight dry at 60 °C.

2.6.2. Manufacturing of Multi-Biofunctional Coating

When the introduction of bio-functionality was considered, lattice structures with better mechanical properties were selected. The antimicrobial properties were first introduced into the MAO-treated porous Ti-15Mo scaffolds. Deionized (DI) water was used for the preparation of the solution and for cleaning thoroughly. The samples were immersed in 0.05 M, 0.1 M, and 0.2 M high-concentration AgNO₃ solutions [59,60], protected from light, and shaken with a magnetic stirrer for 1 h. The samples were then immersed in 10 g/L aqueous TA solution and shaken for 1 h. The samples were then removed and rinsed with DI water and sonicated for 20 min to remove the loose Ag nanoparticles and dried overnight in an oven at 55 °C.

A mixture of TA at a concentration of 30 g/L, CaCl₂ at 0.5 M, and NH₄F at (1/18) M [61] was prepared and heated to 90 °C. The corresponding amount of Na₃PO₄ at a concentration of (1/3) M was quickly added to the mixture solution and magnetically stirred. Each sample with reduced Ag nanoparticles on the surface was suspended and soaked in the middle of the solution with a thin wire for 10 h of constant temperature cross-linking reaction. The samples were then removed and rinsed with deionized water, sonicated, and dried overnight in an oven at 55 °C.

A summary of the experimental methods of surface modification is shown in Figure 2b. Since the Ag loading was the only variable in the coating for all the scaffolds, the subsequent sets of samples were named accordingly, e.g., ISH0.05, ISH0.1, and ISH0.2. The names of samples with a complete coating remained unchanged.

2.6.3. Surface Characterization

The morphology of the multi-biofunctional coating was observed by SEM, and the elemental composition and distribution were detected and analyzed by SEM-EDS. By using an auto scratch coating tester (WS-2005, Zhongke Kaihua, Dongguan, China), the coatings underwent scratch testing, and their adhesion force to substrates was assessed. The load applied by the tester was set from 0.2 N to 16 N at a loading speed of 3 N/min and a scratching speed of 3 mm/min [61,62].

2.7. Bio-Property Test: In Vitro Test

2.7.1. Cytotoxicity Test and Ion Precipitation Analysis

The samples were immersed in acetone, alcohol with a 95% alcohol concentration, and DI water for 10 min ultrasonically. The samples were dried before being sanitized for later use at a high temperature and pressure.

Preparation of leachate: Each sterilized sample was immersed in 2 mL of serum-containing cell culture medium with sodium alginate spheres. The samples were extracted at 37 °C for 24, 48, and 72 h. The leaching solutions were obtained by removing the samples.

Cytotoxicity assay: MG-63 cells were prepared into a cell suspension at a concentration of 5 × 10⁴ cells/mL, inoculated into 96-well culture plates at 200 μL per well, and cultured in an incubator. When the cells grew to 80% confluence, different sample leachates (experimental group), sodium alginate sphere leachates (negative control), 6.4% phenol (positive control), and normal culture medium (blank control) were added sequentially. 5 replicates were set up for each group and incubated for 24 h. The number of cells in each well was determined by absorbance (OD) on a microplate reader at 450 nm. Cell viability was characterized by cell relative growth rate (RGR, Equation (4) [63]).

$$RGR\%={\displaystyle \frac{{OD}_{sample}-{OD}_{blank}}{{OD}_{control}-{OD}_{blank}}}\times 100\%$$

(4)

Ion precipitation analysis: Quantitative determination of ion precipitation by inductively coupled plasma mass spectrometry (ICP-MS). The 7700 Series ICP-MS (Agilent, Lexington, MA, USA) was turned on with Ar gas. After tuning, the mixed standard solutions were drawn, and the standard curves were made with the elements Ti, Mo, and Ag in the standard sample. The nitrified leaching solutions and blank control solution were drawn one by one, and the content of each element was determined by the full quantitative method. Finally, the element concentrations of Ti, Mo, and Ag in each leachate were calculated according to the standard curves.

2.7.2. Hemocompatibility Tests

Each experimental group of G0.05-ISH0.2, the negative control group without a sample, and the positive control group using distilled water were kept in a 24-well plate at a constant temperature at 37 °C. A fresh mouse eyeball blood solution containing 3.8% sodium citrate diluted with normal saline at a ratio of 1:5 was added to each well and incubated at 37 °C for 1 h. After the liquid was centrifuged, the OD of the supernatant was detected at a wavelength of 545 nm. Hemocompatibility was characterized as Equation (5) [64]:

$$Hemocompatibility \%={\displaystyle \frac{{OD}_{sample}-{OD}_{negative control}}{{OD}_{positive control}-{OD}_{negative control}}}\times 100\%$$

(5)

2.7.3. Total Antioxidant Capacity Test

Due to the presence of TA, the antioxidative capacity of the coated samples was determined by a total antioxidant capacity assay kit (BEYOTIME, Haimen, China) with the ferric reducing ability of plasma (FRAP) method [65]. The kit is based on the principle that under acidic conditions, the antioxidants reduce Fe³⁺-tripyridyltriazine (TPTZ) to produce a blue Fe²⁺-TPTZ, which is then measured at 593 nm using a microplate reader to obtain the total antioxidant capacity of the sample. The antioxidant capacity of Trolox, a positive control with similar antioxidant capacity to that of vitamin E, was used as a benchmark for calibration. Each sample was incubated with FRAP reagent for 15–90 min at 37 °C.

2.7.4. Antibacterial Test

As a semi-quantitative assay, the inhibition zone test was utilized to assess the antimicrobial effect of the multi-coatings against Escherichia coli (E. coli) and Staphylococcus aureus (S. aureus). Both the experimental group containing the coating and the control group treated with MAO only were autoclaved. Chilled dormant E. coli and S. aureus were activated to obtain bacterial suspensions, and 100 μL of each was pipetted onto plates poured with cooled sterile solid LB medium. The bacterial suspensions were spread evenly with a spreader. Each sample was sterilized again with an alcohol lamp and pressed firmly to embed in the agar in the center of one plate. A camera was used to capture digital pictures of the samples after the plates had been removed from an incubator where they had been incubated for 24 h at 37 °C. The size of the zone was then measured.

A summary of the experimental methods of in vitro biocompatibility and the biomedical function tests is shown in Figure 2c.

2.8. Statistical Analysis

All data were analyzed by one-way ANOVA using SPSS 17.0 software (SPSS, Chicago, IL, USA) and presented as average ± standard deviation. Experiments were performed in triplicate with statistical values indicated in relevant analyses. Statistical significance was set at p < 0.05.

3. Results and Discussion

3.1. Results and Discussion of the As-Printed Dense and Porous Ti-15Mo

Under the parameters of VED = 100 J/mm³, P_L = 200 W, v_s = 400 mm/s, h_s = 100 μm, and t_l = 50 μm, dense Ti-15Mo alloy was printed with a result of 99.29% RD. Figure 4a,b gives an example of the microstructure morphology of dense samples printed with pre-alloyed Ti-15Mo powder by OM. The average melt channel width of 117.0 μm can be measured from the channels showing the full width in Figure 4a. The defects in the fusion channels and micro poles resulting from the laser scan are visible in the figure, as are the grain boundaries in the martensite and β phases after deep etching.

Figure 4c,d give an example of the surface morphology and compositions of a dense sample printed with pre-alloyed Ti-15Mo powder and an as-casted sample as a control group by the SEM SE mode. The large grain boundaries of the as-cast sample are very evident when comparing the moderately etched as-printed and as-cast samples under the SE of the SEM. The EDS results show that the elemental composition of the sample is indeed 85 wt% Ti and 15 wt% Mo.

For the dense samples, tensile tests were carried out, and the results are shown in

Table 1. Mechanical properties of dense Ti-15Mo with and without heat treatment. compared with the literature.

Samples	σ0.2 (MPa)	UTS (MPa)	Elongation (%)	E (GPa)	Hardness (Hv)	Ref.
As-printed, dense	816 ± 4.5	910 ± 9.3	19 ± 3.0	82.1 ± 2.5	391.9 ± 7.0	This article
As-cast	802 ± 3.3	905 ± 8.7	15 ± 5.0	78.3 ± 1.7	319.8 ± 7.9	This article
Ti-15Mo (in situ alloyed by PBF-LB, 1100 °C treated)	1170 ± 17	1180 ± 22	1.6 ± 0.2	86.3 ± 1.9	379.3 ± 13	This article
Ti-15Mo	584	934	14	/	321	[66,67]
Ti-15Mo (cast)	745 ± 13	921 ± 19	25 ± 2	84 ± 1	310 ± 10	[68]

along with others. The PBF-LB laser gave a large increase in hardness to the alloy and maintained a high level of strength while keeping the modulus low and the elongation ideal. The numbers in parentheses are standard deviations.

Regarding the porous samples, their SEM images are shown in Figure 3d. The obtained 300–550 μm pores are favorable for cell growth. A major problem with the preparation of porous structures from metal powders by PBF-LB is that partially molten powder is on the lattice surface, making the scaffold morphology less smooth, and a morphology similar to spheroidization appears. For the purpose of in vivo bone defect applications, physical treatment such as sandblasting cannot be used to treat the surface at a cell size of 1.5 mm. The reason for using only prolonged ultrasound to remove the loose powder and not acid etching or electrolytic polishing to give the scaffold a smooth appearance is that our previous studies [69] have demonstrated to a certain extent that the metal powder in this case does not have a significant negative effect in vivo. Methods such as acid etching are more suitable for porous structures with larger cell sizes, i.e., thicker struts and larger pores. Therefore, on balance, a smaller unit cell with a simpler initial surface treatment is a more desirable combination.

The results of the XRD analysis of powder stock, as-printed scaffolds, and as-cast alloy are shown in Figure 5a. Only the β phase was detected in the interval 30–80°, with the highest peaks occurring on the β $\left\{110\right\}$ crystal plane. This indicates that the martensite content is below the XRD detection limit, which can demonstrate the phase composition of the as-printed scaffolds. In the presence of a large amount of β-phase stabilizing element Mo, the transformation of the β-phase into the α-phase is suppressed in the melting pool with a fast cooling rate, and only a small amount of the β-phase is transformed into a supersaturated solid solution (martensite) of the same composition but with a different crystal structure. This martensitic phase transformation is a diffusion-free phase transformation in which no atomic diffusion takes place; only lattice reconstruction occurs [70].

The results of the EBSD analysis of the as-printed scaffolds and as-cast alloy are shown in Figure 5b–f. As can be seen from Figure 5b, the as-printed scaffolds exhibited fine polycrystalline microstructures with grain sizes significantly smaller than those of the as-cast alloy. The average particle size of the former was about 40 μm, while that of the latter was about 100 μm, which was also consistent with the mapping results of Figure 5c,d. This phenomenon is due to the extremely short existence time of each melting pool of the PBF-LB technique, which effectively refines the Ti-15Mo alloy grains. The kernel average misorientation (KAM) results are shown in Figure 5e,f. The method is based on the deviation of the orientation gradients between pixels and therefore corresponds to the geometric dislocation density and local residual stresses for assessing the stress state inside a metallic material. As the results showed very low orientation gradients, only 40% of the standard color scale bar in the range 0–5 was used here for display purposes. The as-cast sample had almost no dislocations, while the as-printed Ti-15Mo showed yellow-red aggregations at the low-angle grain boundaries, indicating relatively high residual stresses in these areas.

The results of the compression and tensile tests on the as-printed porous Ti-15Mo samples are summarized in

Table 2. Tensile tests results of as-printed Ti-15Mo of G, ISH, and ISO structures.

Structures	Tensile Tests Results
	RD	YS (MPa)	UTS (MPa)	ε (%)	E (GPa)
G	30%	158.84 ± 2.62	160.10 ± 3.02	1.10 ± 0.12	18.16 ± 0.32
	35%	216.50 ± 4.17	217.98 ± 1.69	1.15 ± 0.17	23.24 ± 0.38
	40%	300.11 ± 5.18	313.95 ± 2.66	1.15 ± 0.09	33.53 ± 0.45
ISH	30%	139.26 ± 3.31	143.04 ± 1.02	1.07 ± 0.07	15.80 ± 0.26
	35%	167.62 ± 1.44	168.73 ± 5.71	1.00 ± 0.04	21.10 ± 0.08
	40%	170.04 ± 4.75	173.34 ± 2.54	0.61 ± 0.07	35.62 ± 0.17
ISO	30%	131.12 ± 5.29	131.21 ± 4.63	1.24 ± 0.10	12.65 ± 0.13
	35%	154.75 ± 3.85	158.96 ± 2.08	1.08 ± 0.06	18.38 ± 0.61
	40%	151.85 ± 5.58	159.27 ± 0.94	1.02 ±0.11	22.84 ± 0.73

and

Table 3. Compression tests results of as-printed Ti-15Mo of G, ISH, and ISO structures (with 30%, 35%, and 40% RD).

Structures	Compression Tests Results
	RD	σ0.2 (MPa)	Final Strength (MPa)	ε (%)	E (GPa)
G	30%	90.99 ± 2.11	167.85 ±3.24	20.2 ± 0.2	31.59 ± 0.26
	35%	117.40 ± 4.54	192.20 ± 4.72	20.1 ± 0.1	36.60 ± 0.07
	40%	129.91 ± 2.86	213.62 ± 3.39	20.0 ± 0.2	40.37 ± 0.85
ISH	30%	78.82 ± 1.15	125.52 ± 2.13	39.3 ± 0.1	31.10 ± 0.06
	35%	107.10 ± 4.63	157.62 ± 2.41	39.3 ± 0.1	43.00 ± 0.73
	40%	118.51 ± 3.75	183.83 ± 6.87	39.1 ± 0.3	47.3 ± 0.55
ISO	30%	82.31 ± 5.97	113.51 ± 2.24	39.2 ± 0.4	16.80 ± 0.34
	35%	98.92 ± 2.42	147.64 ± 4. 79	39.0 ± 0.2	19.90 ± 0.31
	40%	120.10 ± 1.86	178.08 ± 4.58	39.0 ± 0.3	24.20 ± 0.17

. Figure 6 shows the results of the compression tests on the three TPMS lattices with different RD values. The yield strength σ_0.2 ranged from 78 to 130 MPa, and Young’s modulus ranged from 16 to 48 GPa for all experimental samples. Figure 6d–f show that the yield strength and Young’s modulus of TPMS lattices increase with increasing RD. Comparing the three groups with the same structural parameters and different TPMS, such as G35, ISH35, and ISO35, the Young’s modulus and yield strength are different, indicating that the different structures have a significant effect on the mechanical properties, given the same structural parameters.

Furthermore, the data in Table 2 and Table 3 show that the elongation of the porous structures remained relatively constant in the tensile and compression tests, but there was some degree of difference in Young’s modulus and yield strength. Considering that porous implants are mainly subjected to compressive stresses in vivo, the relatively low-strength structure of ISO will be removed from subsequent in vitro experiments and used as a control group. The mechanical properties of the G structure are slightly stronger than those of the ISH. The strut thicknesses of 0.15 mm and 0.11 mm for G30 and ISH30, respectively, are also consistent with the results of the mechanical properties analyzed in the literature [71,72], i.e., the mechanical properties are stronger with a wider support structure in the load direction, all other things being equal.

Figure 7 shows the tensile test results for the three TPMS lattices with different RD values. All flat tensile specimens show similar mechanical behavior trends to the compression specimens, with yield strengths ranging from 131 to 300 MPa and Young’s modulus ranging from 12 to 36 GPa. Figure 7g–i show the fracture analysis of three sets of tensile specimens, where the joint appearance of the cleavage planes and dimples demonstrates that all three tensile failure mechanisms were a mixture of brittle and ductile fracture.

3.2. Results and Discussion of the Surface Modification of Ti-15Mo Lattices

Figure 8 represents the initial stages of surface modification. All samples designed for in vitro tests were treated by MAO and named as G(M), ISH(M), and ISO(M), respectively. The 0.1–1 μm holes were evenly distributed and tightly covered on the substrate scaffold. The composite structure of the 100-micron-scale macropore substrate scaffold and the micron-scale ceramic coating laid a foundation for the subsequent attachment of multi-biofunctional coatings with stable thermodynamic properties, larger surface area, and easier adhesion. Based on the experimental results of the mechanical properties, G(M) and ISH(M) were selected as the experimental group, and ISO(M) was henceforth set as the control group. The SEM image of the reduced Ag particles is shown in Figure 8a. The smallest individual Ag particle size was about 20 nm, with some Ag particles agglomerated up to 400 nm. Nano Ag particles were distributed evenly inside and outside the micron-scale TiO₂ pores. Figure 8c,d show the EDS results for the corresponding color star symbols in Figure 8a, where the Cl peak was derived from the electrolyte of MAO. The bright white particles in the SEM were thus confirmed to be all nano Ag and attached to the microporous TiO₂.

Figure 9 shows the (F)HA-TA-Ag-TiO₂ coating after completion of the surface modification steps. Figure 9a shows flake HA covering the outside and inside of micron-scale TiO₂ pores, while Figure 9b shows another kind of rough surface spherical HA aggregated morphology, both with a tiny amount of uncovered nano Ag. Figure 9c,d add the Ca and P peaks representing HA and the weak peak of the reactant Na based on the previous EDS. Figure 9e–g show the evidence of all the results of surface modification. TiO₂, nano Ag, and HA have been distributed as the corresponding element mappings. According to Figure S1, the MAO coating was composed of an amorphous phase, whose amorphous bulge was at 20–30° in the results of GISAXS. An XRD peak corresponded to the hydroxyapatite (HA) phase at 2θ = 25.36°. These results align with the SEM observations and theoretical expectations for hydrothermal crystallization behavior [58].

The change of the tangential friction force was measured by the acoustic emission method to determine the bonding force between the coating and the substrate. The indenter gradually increased the pressure on the surface. When the relative displacement of the coating and the substrate occurred, an acoustic signal would be generated, as shown in Figure 10. When the coating failed, there would be acoustic feedback, and the corresponding normal critical load (Lc) was used to characterize the bonding force between the film and the substrate. The larger the critical load, the stronger the bonding between the coating and the substrate. The average binding force (Lc) of TiO₂ prepared by MAO was ~30N, and the average bonding force between the complete coating and the substrate was 9.98 N, which is comparable to the conclusions of some existing literature [61]. The data indicate that MAO can indeed prepare a well-bonded ceramic layer; the bonding force of the composite coating is greater than the literature reference value of ~6.5 N for HA, also indicating the advantage of F-doping on the stability of the outermost HA layer.

3.3. Results and Discussion of the In Vitro Analysis of Coated Ti-15Mo Scaffolds

When testing the biologically relevant properties of coated porous Ti-15Mo, the experimental groups involved were G and ISH samples treated with all surface modification steps (i.e., G0.05 to ISH0.2), and the control group was mainly the MAO-treated ISO(M) only. The ion release amounts of Ti, Mo, and Ag from the leachates for 0–24 h, 24–48 h, and 48–72 h for each group of samples were calculated and are presented in the histograms in Figure 11. The ionic release of Ti and Mo was minimal for all groups of samples, and no Ag was detected for ISO(M), as no Ag was added. For each experimental group, the line graph in Figure 11c systematically reflects the overall precipitation of Ag from the coating over the three days, with a steady increase in the amount released. This demonstrates not only the feasibility of releasing the effective ions from the coating in the incubation environment, but also the slow release effect. The minimum total Ag release was about 13.5 μg, and the G0.05 and ISH0.05 groups released about 18.5 μg and 17 μg, respectively, which is high compared to the literature results of several micrograms [59,73].

Figure 12 shows the results of cell and blood compatibility. As Ti-15Mo and TiO₂ have not been shown to be antibacterial and cytotoxic, the 94.2% RGR result of the control group ISO(M) allowed for a rough estimate of the cell proliferative effect of MAO-treated as-printed Ti-15Mo lattices. The cytotoxicity results showed that the RGR of each experimental group was higher than 94.2%, indicating that the high Ag ion release concentration and the TiO₂ ceramic layer of the composite coating were non-toxic to cells, and the overall cytocompatibility of the composite coating was better than that of microporous TiO₂. In the hemolysis assay, the sample was in direct contact with the blood cells, and hemolysis is related to osmotic pressure and hemoglobin degeneration caused by ions. As can be seen from Figure 12b, G0.1 had the highest hemolysis rate, followed by G0.2 and ISH0.1, all of which were higher than the 5% hemolysis safety line. ISH0.2 had a relatively low hemolysis rate, but still reached 3.498% and should not be preferred. G0.05 and ISH0.05 reflected ideal safety.

After the comprehensive ion precipitation, cytocompatibility, and hemocompatibility tests results, G0.05 and ISH0.05 were selected for superiority. The two groups of samples with AgNO₃ concentration of 0.1 M at the time of preparation showed the highest Ag concentration in both ion release and hemocompatibility assays, while G0.2 and ISH0.2 did not show the most significant toxicity. The reason for such results may be that at high Ag levels, the Ag ion in the leaching solution reaches an appropriate amount to bind to proteins such as serum in the medium, and hemoglobin is directly bound by the excess Ag in the sample, in order to form a stable precipitate without exerting significant toxicity.

The results of antioxidant properties are shown in Figure 13. Figure 13a shows that the absorbance values used to measure the total antioxidant capacity to ROS of the (F)HA-TA-Ag-TiO₂ coating in G0.05 and ISH0.05 gradually increase with incubation time, due to the slow sustained release of TA from the coating. Therefore, the prepared (F)HA-TA-Ag-TiO₂ coating has long-term anti-oxidation properties. The antioxidant capacity of the coatings with different TPMS structures was found to be similar, further indicating that the antioxidant properties of the composite coatings are conferred by the TA.

The antibacterial test results of the samples are shown in Figure 14. As a semi-quantitative test, the size of the inhibition zone is summarized in

Table 4. Results of the inhibition zones (expressed in millimeters).

Inhibition Zone/mm	G0.05	ISH0.05	ISO(M)
S. aureus	0.75 ± 0.02	1.50 ± 0.01	0
E. coli	0.55 ± 0.01	1.00 ± 0.02	0

. The inhibition zone pattern of the ISH0.05 group was complete, the size was relatively larger, and the higher silver ion release radius showed better antibacterial performance. In contrast, the G0.05 group also had a good performance of the inhibition zone, but the edge of the zone was not smooth, and the size was relatively small. The control group had no bactericidal properties because it lacked Ag. The results reflect that different TPMS structures affect the amount of Ag released, with almost a twofold difference in the radial length of the corresponding inhibition circles.

When the structural parameters are the same, it means that the same amount of pre-alloyed powder is used. At this time, it can be assumed that the larger the surface area, the smaller/thinner the struts of the lattice. In order to explain the gap of doubling the antimicrobial capacity of ISH0.05 versus G0.05, the surface area of two CAD models with the same structural parameters, G30 and ISH30, was analyzed using Python 3.8.8 to obtain a result of 542.98 mm² for G30 and 629.33 mm² for ISH30, consistent with the previous results for the thickness of struts. With similar ion precipitation results, a difference in surface area of approximately 16% may not be sufficient to determine the antimicrobial efficiency gap alone, so the pore structure of the ISH structure may be more favorable to the coating of the active ingredient for bactericidal applications.

4. Conclusions

In this work, porous β-phase Ti-15Mo TPMS alloys based on Gyroid, I-WP Sheet, and I-WP solid lattices were fabricated by PBF-LB. The unit cell size was 1.5 mm, the relative density (RD) was 30–40%, the compressive modulus was 16–48 GPa, and the yield strength σ_0.2 ranged from 78 to 130 MPa, which was effectively close to the mechanical parameters of human bone. After using MAO to prepare functional modification layers, the antibacterial, antioxidative and osteopromoting composite coatings were successfully prepared by using broad-spectrum antibacterial Ag, antioxidative tannic acid and (F)HA with high stability, respectively. It was demonstrated that the G and ISH groups, with an Ag ion release of about 6 μg/D and an RD of 30% possessed better antioxidant and biocompatible properties, with the ISH group having better antibacterial ability. For the same structural parameters, ISH lattices have somewhat weaker mechanical properties than G lattices and are more biofunctional with the multi-coatings.