Plasma sprayed fluoride and zinc doped hydroxyapatite coated titanium for load-bearing implants

Abstract

Titanium (Ti) alloys show excellent fatigue and corrosion resistance, high strength to weight ratio, and no toxicity; however, poor osseointegration ability of Ti may lead to implant loosening in vivo. Plasma spraying of hydroxyapatite [HA, Ca₁₀ (PO₄)₆ (OH)₂] coating on Ti surfaces is commercially used to enhance osseointegration and the long-term stability of these implants. The biological properties of HA can be improved with the addition of both cationic and anionic dopants, such as zinc ions (Zn²⁺) and fluoride (F⁻). However, the hygroscopic nature of fluoride restricts its utilization in the radiofrequency (RF) plasma spray process. In addition, the amount of doping needs to be optimized to ensure cytocompatibility. We have fabricated zinc and fluoride doped HA-coated Ti6Al4V (Ti64) to mitigate these challenges using compositional and parametric optimizations. The RF induction plasma spraying method is utilized to prepare the coatings. Multiple parametric optimizations with amplitude and frequency during the processing result in coating thicknesses between 80 and 145 μm. No adverse effects on the adhesion properties of the coating are noticed because of doping. The antibacterial efficacy of each composition is tested against S. aureus for 24, 48, and 72 h, and showed that the addition of zinc oxide and calcium fluoride to HA leads to nearly 70 % higher antibacterial efficacy than pure HA-coated samples. The addition of osteogenic Zn²⁺and F⁻ leads to 1.5 times higher osteoblast viability for the doped samples than pure HA-coated samples after 7-days of cell culture. Zn²⁺ and F⁻ doped HA-coated Ti64 with simultaneous improvements in anti-bacterial efficacy and in vitro biocompatibility can find application in load-bearing implants, particularly in revision surgeries and immune-compromised patients.

1. Introduction

Current treatment options for bone disorders consist of skeletal reconstruction with implants. Titanium and its alloys are widely used to repair load-bearing skeletal defects, owing to their high strength to weight ratio, excellent corrosion and fatigue resistance, and nontoxicity. However, the bioinert nature of Ti hinders new bone formation or osseointegration after implantation [1,2]. Several surface modification methods, including coating with bioactive hydroxyapatite [HA, Ca₁₀(PO₄)₆(OH)₂] ceramics, are employed to improve the in vivo biocompatibility of Ti alloys. The compositional similarities of HA with bone help in early-stage osseointegration [3,4]. The biological properties of HA-coated Ti can be further improved using various dopants [5,6]. The flexible crystal chemistry of HA allows substitution with different cations and anions in the Ca²⁺ and OH⁻ sites, respectively [7,8]. Previous studies report the fabrication of Ag⁺, Cu²⁺, Mg²⁺, Zn²⁺, and F⁻ doped HA by solid-state or wet precipitation methods [9–11]. Among these dopants, Zn²⁺ is an essential trace element which helps in new bone formation, protein, and DNA synthesis [12]. However, over intake of Zn²⁺ leads to toxicity [13]. Fluoride (F⁻) is another trace element in natural bone, blood, and dental enamel that accelerates new bone formation and growth [14]. Fluorine is the most reactive non-metal in the periodic table, with very high electronegativity. Due to its chemical nature, fluorine is susceptible to agglomeration and moisture absorption and limits bulk production using fluoride-containing powders [15,16]. Fluoride helps prevent dental caries, but an over intake may lead to dental and skeletal fluorosis [17]. Additionally, both Zn²⁺ and F⁻ have antibacterial properties. Most published research reports assess the effects of Zn²⁺ and F⁻ doping levels in HA individually [14,18,19]. However, there is a knowledge gap in understanding the influence of the combined system of Zn²⁺ and F⁻ on cytocompatibility and anti-bacterial efficacy, which are the primary focus of this study.

Plasma spraying is the only US food and drug administration (FDA) approved commercial manufacturing method to fabricate ceramic coated metallic implants for load-bearing applications [20]. However, a severe bacterial infection to load-bearing implants leads to many revision surgeries. Local antibiotic delivery to the infection site is also challenging, and it further complicates the treatment [21]. In this regard, designing an implant with inherent anti-bacterial resistance is crucial. There is no available literature regarding the parametric optimizations and F⁻ doped HA coating fabrication on Ti64 substrate by RF induction plasma spraying. In addition, the combined biological effects of Zn²⁺ and F⁻ co-doped HA-coated Ti-6Al-4V (Ti64) are yet to be explored. The scientific question posed in the current work is, can we fabricate optimized zinc and fluoride co-doped HA-coated Ti64 with enhanced biocompatibility and inherent anti-bacterial resistance utilizing RF induction plasma spraying? A detailed parametric and compositional optimization is necessary to maximize the advantages of zinc and fluoride-containing HA coating without affecting the adhesion properties between the substrate and the coating. The novelty of this work lies in the compositional modifications, parametric optimizations, and successful fabrication of Zn²⁺ and F⁻ co-doped HA-coated Ti64, showing superior cytocompatibility and inherent anti-bacterial resistance over pure HA-coated Ti64.

2. Materials and methods

2.1. Fabrication of doped powders and physical characterizations

Commercial HA powder (NEI, USA) was sieved through 175–212 μm mesh to get powder with a particle size in that range. The Zn²⁺ and F⁻ doping with HA were performed using the precursors ZnO and CaF₂, respectively. The mixed powders were ball milled for 2 h at a speed of 80, with a 2:1 powder to ball ratio [5]. The water content of the prepared powders was measured using the ASTM C373 standard [22]. The exact amount of powder was oven-dried at 110 °C for 24 h and immediately kept inside the desiccator after drying. The weight of the dried powders was measured, and the water content was calculated as a % change of weight after drying. The angle of repose test was carried out as per published literature [23]. Briefly, the powders were poured into a 1 L water-filled graduated cylinder, then the cylinder was tilted and brought back to the normal position. Once the powders settled in the water, the measured dimensions were used to calculate the angle. This method was performed multiple times to ensure repeatability, and an average value was reported.

2.2. Compositional optimization

The undoped and doped powders with different compositions were uniaxially pressed at 165 MPa, with 4 min holding. The obtained green discs of 12.5 mm diameter and 2 mm height were sintered at 1250 °C with 2 h holding in a muffle furnace as reported in our previous study [24]. The ZnO doping level was varied as 0.10, 0.25, and 0.5 wt%. For the CaF₂ doping, the different levels were selected as 0.10, 0.3, and 0.6 wt%. 9 different compositions were tested for a day 3 cytotoxicity assessment and 24 h anti-bacterial assay. Based on the obtained results, the samples with 0.25 wt% ZnO and 0.3 wt% CaF₂ were chosen for further study. The detailed nomenclature and compositions of the samples, used for the optimization study are mentioned in Table 1a and b.

Table 1.

(a) Tested compositions with various ZnO and CaF₂ doping amounts to optimize the biological properties (b) compositions of the powders, utilized for plasma spray coating and further biological study.

(a)
Sample ID (for optimization)	ZnO doping level (wt%)	CaF2 doping level (wt%)
Z1F1	0.1	0.1
Z1F2	0.1	0.3
Z1F3	0.1	0.6
Z2F1	0.25	0.1
Z2F2	0.25	0.3
Z2F3	0.25	0.6
Z3F1	0.5	0.1
Z3F2	0.5	0.3
Z3F3	0.5	0.6
(b)
Sample ID (for plasma spray coating)	ZnO doping level (wt %)	CaF2 doping level (wt %)
HA	0	0
Zn-HA	0.25	0
F-HA	0	0.3
Zn – F - HA	0.25	0.3

2.3. Coating preparation using RF induction plasma spray

The HA and doped HA powders were coated on sandblasted Ti64 discs (12.2 mm diameter and 2 mm thickness) using a 30 kW inductively coupled radio frequency plasma spray system (Tekna Plasma System, Canada) [6,25]. A series of optimization studies were performed with the input parameters, i.e., amplitude and frequency. Other process variables such as power, spray distance from the nozzle, and the gas flow rate was kept constant during the optimizations. Our previous studies utilized 25 kW power, 110 mm spray distance, and a carrier gas flow rate of 25 standard 1 min⁻¹ (s.L.p.m.) [5,25]. The complete parametric optimization is reported in Table 2. To compute the adhesion strength of the coating, another set of coated samples was prepared on Ti64 discs of 25.4 mm diameter and 2 mm height.

Table 2.

The RF induction plasma spray process parameter optimization for each parametric combination. The presence of fluoride leads to no powder flow region for a set of amplitude and frequency. Six samples are tested per optimization step and an average value of the thickness is reported.

Sample name	Amplitude	Frequency (Hz)	Average coating thickness (μm) (n = 6)
HA	20	106.8	90–110
HA	20	107.3	95–115
HA	25	106.8	100–115
HA	25	107.3	100–125
HA	27	106.8	100–120
HA	27	107.3	105–135
HA	30	106.8	110–130
HA	30	107.3	120–150
Zn-HA	20	106.8	80–100
Zn-HA	20	107.3	90–110
Zn-HA	25	106.8	90–105
Zn-HA	25	107.3	110–130
Zn-HA	27	106.8	85–100
Zn-HA	27	107.3	110–135
Zn-HA	30	106.8	100–125
Zn-HA	30	107.3	120–145
F-HA	20	106.8	No powder flow
F-HA	20	107.3	No powder flow
F-HA	25	106.8	No powder flow
F-HA	25	107.3	No powder flow
F-HA	26	107.3	No powder flow
F-HA	26	112	60–75
F-HA	26	113.7	90–110
F-HA	24.5	112	100–125
F-HA	24.5	113.7	120–145
Zn-F-HA	20	106.8	No powder flow
Zn-F-HA	20	107.3	No powder flow
Zn-F-HA	25	106.8	No powder flow
Zn-F-HA	25	107.3	No powder flow
Zn-F-HA	30	106.8	No powder flow
Zn-F-HA	30	107.3	No powder flow
Zn-F-HA	20	111	70–95
Zn-F-HA	25	111	80–100
Zn-F-HA	25	111.9	95–115
Zn-F-HA	25	112.5	110–135
Zn-F-HA	30	112.5	115–130

2.4. Phase identification and microstructural characterization

Phase identification of coatings was performed using the X-Ray diffraction method (Siemens D500 diffractometer) with a 2θ range of 20 to 60 degrees at a step size of 0.05 degrees using a Cu Kα radiation at 35 kV, and 30 mA. The microstructural characterization was carried out with field emission scanning electron microscopy (FESEM; FEI Apreo volume scope, FEI company, USA). The images were taken at 20 kV and 0.8 nA beam current. From now onwards, the hydroxyapatite-coated Ti64 will be represented as HA, fluoride-doped hydroxyapatite-coated Ti64 will be represented as F-HA, zinc doped hydroxyapatite-coated Ti64 will be represented as Zn-HA, and Zn-F-HA will be used to indicate the zinc and fluoride-doped hydroxyapatite-coated Ti64.

2.5. Assessment of mechanical properties

The bond strength test of the prepared coatings was performed as per the ASTM C633 standard of tensile adhesion test [5,26]. As per the recommended ASTM standard, the samples with a coating thickness of ~285 ± 20 μm were used for this characterization. Each sample was glued to the posts with Armstrong A-12 epoxy resin as an adhesive and cured at 93.3 °C (200 °F) for 30 min, followed by slow cooling to room temperature according to the manufacturer’s recommendation. A constant crosshead speed of 0.0013 cm s⁻¹ was used during the tensile test until the coating failed. The failure load/sample area data was used to compute the adhesion strength.

2.6. In vitro cell – material interaction study

2.6.1. Cell seeding on the coated surface

The cytocompatibility of the fabricated compositions up to day 7 was tested with Human fetal osteoblast cells (hFOB, obtained from ATCC, Manassas, VA) per our previous work [24]. A mixture (1:1 ratio) of Ham’s F12 medium and Dulbecco’s Modified Eagle’s Medium (DMEM/F12, Sigma, St. Louis, MO) was used to prepare the cell media. The basal medium was prepared using sterilized 2.5 mM l-glutamine (without phenol red) solution. This media was mixed with 10% fetal bovine serum (FBS, ATCC, Manassas, VA) and 0.3 mg/mL G418 (Sigma, St. Louis, MO) and used in the cell culture experiment. The samples were autoclaved at 121 °C for 1 h followed by drying. The sterilized samples were kept in a 24 well plate inside the cell culture hood. On top of each sample, 40 × 10³ cells were seeded, followed by 1 mL growth media addition. Each well plate was incubated at 37 °C under a 5% CO₂ atmosphere.

2.6.2. Cell viability assessment by MTT assay

MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bro-mide) assay was utilized to measure the cell viability after each time point of cell-material interaction. The samples were moved to a new 24-well plate. On top of each sample, 100 μL of MTT (Sigma, St. Louis, MO) solution was added, followed by 900 μL of cell media addition. The well plates were kept inside an incubator at 37 °C for 2 h. The solution was removed from each well, and 600 μL of MTT solubilizer solution (a mixture of 10% Triton X-100, 0.1 M HCl, and isopropanol) was added per sample to ensure complete dissolution of the purple formazan crystals generated during the MTT assay. This solution was used to quantify the cell viability. 100 μL of the obtained solution was put in each well of a 96-well plate, and the optical density was measured at 570 nm with a UV–Vis microplate reader (BioTek) [42]. Tests were carried out in triplicate to ensure repeatability.

2.6.3. Investigation of cellular morphology with FESEM

Cellular attachment and morphology were analyzed by field-emission scanning electron microscope (FESEM) (FEI Inc., Hillsboro, OR, USA) [27]. Samples were fixed with 2% paraformaldehyde and 2% glutaraldehyde in 0.1 M phosphate buffer and refrigerated overnight at 4 °C. After refrigeration, each sample was rinsed three times with 0.1 M phosphate buffer solution, and post-fixation was carried out with 2% osmium tetroxide (OsO₄) [41]. In the following steps, the samples were dehydrated in an ethanolic series with one time 30%, 50%, 70%, 95%, and three times with 100% ethanol, respectively. Post-dehydration, hexamethyldisilane (HMDS) was added to each sample top, and the well plates were kept overnight inside a desiccator for drying, per our previous work [24]. The dried samples were gold coated and analyzed under the FESEM.

2.7. Assessment of antibacterial properties

2.7.1. Modified ISO 22196: 2011 standard protocol

The antibacterial efficacy of the coatings was tested against S. aureus, according to the modified ISO 22196: 2011 Standard [14]. Optical densities of the activated bacterial suspensions were measured at 625 nm using a UV–Vis spectroscopy microplate reader (BioTek) after a serial dilution and compared with the McFarland standard. Based on this quantification, each sterilized sample in a 24 well plate was loaded with 10⁶ CFU of bacteria with 1 mL broth media. The well plates were incubated at 37 °C and 90% humidity. After 24, 48, and 72 h of sample – bacteria interaction, the samples were moved to a glass vial containing 1 mL phosphate buffer solution (PBS). Vortexing of each vial was performed for 15 s, and 10 μL of bacterial suspension was plated on the agar plates, followed by 24 h incubation at 37 °C. The bacterial colony formation was checked from the photographed plates after 24 h incubation. The following equations were used to calculate each treatment’s antibacterial potential (R %).

$$\text{R}\%=\left({\text{N}}_{\text{control}}-{\text{N}}_{\text{treatment}}\right)/{\text{N}}_{\text{control}}$$ (A)

$$\text{Bacterial}\text{cell}\text{viability}(\%)={\text{N}}_{\text{treatment}}/{\text{N}}_{\text{control}}\times 100\%$$ (B)

where N_control and N_treatment represent the number of colonies formed in the control and treatment sample agar plates, respectively, at each time point.

2.7.2. Live-dead staining, confocal microscopy, and evaluation of bacterial morphology

The live and dead bacteria on sample surfaces were quantified using the live-dead staining. After each time point of bacterial interaction, samples were moved to a new well plate. The stain was prepared with a 1:1 mixture of calcein M (Biolegend, CA, USA) and propidium iodide (Invitrogen, MA, USA) solution in PBS. 800 μL of this solution was poured on top of each sample, and then the well plate was kept for 30 min incubation at 37 °C. Confocal microscopy was performed immediately after staining (using Leica TCS SP5 confocal laser microscope) to ensure accurate live and dead bacterial count. Calcein M was detected with a laser of wavelength range 485 nm–535 nm, and another laser of 530 nm–620 nm wavelength range was used to identify the propidium iodide stain. In the images obtained using confocal microscopy, the green color indicates live bacteria whereas the dead bacteria are identified by red color. Bacterial attachments on sample surfaces were characterized with FESEM using the same method as for osteoblast study

3. Results

The tested compositions for evaluation of initial cytotoxicity and antibacterial potential are listed in Table 1a. The obtained cytotoxicity results after 3 days of cell-material interactions are shown in Fig. 1a. It can be seen that any composition having a ZnO doping level of 0.5 wt% shows a decrease in cell viability than control HA. A similar observation is noticed for compositions with a CaF₂ level of 0.6 wt%. The rest of the compositions show no cytotoxic effects after 3 days of culture. After 24 h of interaction with S. aureus, initial antibacterial test results are shown in Fig. 1b. Z1F1 shows ~20% antibacterial efficacy among the tested compositions, Z1F2 shows ~46%, and ~35% antibacterial efficacy is noticed for Z2F1. The highest antibacterial efficacy of ~70% is noticed for the Z2F2 composition. Based on these optimization results, a composition of 0.25 wt% ZnO and 0.3 wt% CaF₂ doped HA was selected for the plasma coating and further biological characterizations.

Fig. 1.

Compositional optimization based on initial cytotoxicity and antibacterial efficacy (a) All compositions with ZnO doping level 0.5 wt% and CaF₂ doping level 0.6 wt% show toxicity with the hFOB cells at day 3. Rest compositions show similar cell viability to that of control. Based on these results, 4 compositions are selected for antibacterial property assessment (b) tested compositions show antibacterial efficacy against S. aureus after 24 h of sample – bacteria interaction. The composition with 0.25 wt% ZnO and 0.3 wt% CaF₂ doped HA shows the highest antibacterial efficacy of ~70%. Due to the best antibacterial efficacy, this composition is utilized for further study; ** indicates p < 0.001.

Table 1b denotes the compositions utilized to fabricate the plasma-coated samples. Multiple parametric optimizations with each composition were carried out to obtain coatings. The parametric optimization details are given in Table 2. The schematic of the plasma spraying process and a summary of the obtained biological properties of coated samples are shown in Fig. 2a. The XRD results in Fig. 2b show the characteristic peaks of hydroxyapatite (JCPDS # 09-0432) [14]. Fig. 2c represents the water content of each powder. For HA and Zn-HA, only 0.2 ± 0.05%, and 0.4 ± 0.08% water content are measured, respectively. In contrast, the F-HA shows a water content of 4.1 ± 0.4%, and the Zn-F-HA shows 3.4 ± 0.2%. An angle of repose test has been carried out to investigate the effects of fluoride on powder flowability [23]. The F-HA shows an angle of repose ~46 ± 3° [23,28]. Fig. 2d and e represents the coating thickness as a function of amplitude and frequency, respectively. The lower and upper limit of each coating thickness is mentioned in Table 2 and Fig. 2d and e.

Fig. 2.

(a) Demonstration of plasma spraying process schematic to fabricate Zn-F doped HA coated Ti64 and associated biological properties as a result of F- and Zn²⁺ doping (b) XRD patterns of the prepared coating corroborate well with hydroxyapatite (JCPDS # 09-0432) (c) water content analysis shows 0.2 ± 0.05% water for HA, 0.35 ± 0.08% water for the Zn – HA, 4.1 ± 0.4% for F - HA, and 3.4 ± 0.2% water for the Zn – F – HA composition. ** indicates p < 0.001 (d, e) upper bound and lower bound thickness as a function of amplitude and frequency respectively.

The thickness, frequency, and amplitude correlation for each composition are plotted on a 3D plane (Fig. 3a–d) utilizing MATLAB 2020b. HA and Zn-HA show a coating thickness of 80–145 μm for amplitude and frequency range of 20–30, and 106.8–107.3 Hz, respectively. In contrast, similar range of amplitude and frequency leads to no powder flow for both F-HA and Zn-F-HA (Table 2 and Fig. 3c, d).

Fig. 3.

3D plots of resultant coating thickness as a function of frequency and amplitude optimization for each composition as represented in Table 2. (a) HA and (b) Zn – HA shows a coating thickness of 80–145 μm for a similar frequency and amplitude range of 106.8–107.3 Hz, and 20–30 respectively (c) F-HA results in no powder flow in the frequency, and amplitude region of 106.8–107.3, Hz and 20–24 respectively. Parametric optimizations with this composition result in a coating thickness of 60–145 μm for 112–113.7 Hz frequency and 24.5–26 amplitude (d) Zn-F – HA shows a no powder flow region for frequency range 106.8–107.3 Hz. In addition, 70–130 μm coating thickness is observed for a frequency region 111–112.5 Hz.

Fig. 4a shows the SEM images of the coated surfaces. A rough and porous surface is visible, and the addition of dopants does not cause any significant changes in the coating morphology. Fabricated coating before parametric optimization is shown in Fig. 4b, which indicates a coating failure and poor adhesion on Ti64 substrate. However, good adhesion between the coating and Ti64 substrate is noticed after parametric optimization (Fig. 4b). The adhesion strengths of HA and Zn-HA were measured as 19.6 ± 3 MPa, 18.3 ± 2.2 MPa. Whereas, 19 ± 2 MPa adhesion strength is noticed for the Zn-F-HA sample. Similar bond strength for all tested compositions indicates that the doping does not cause any adverse effects on the coating adhesion. Fig. 4c shows the coating failure after the tensile test. The sample and post after tensile testing are shown in Fig. 4d. A mixed-mode of failure is observed in this test. Predominantly cohesive failure is seen in the coating. However, some adhesive failure is also noticed from the top post and epoxy resin [5].

Fig. 4.

(a) FESEM images of the coated surfaces. Similar morphology is observed for all surfaces. (b) representative image of samples before and after optimization. Parametric optimization leads to adhere coating on Ti64 surface (c) coating failure after the tensile test (d) the post and substrate after tensile testing.

The results are shown in after 24, 48, and 72 h of sample–bacteria interaction. It is evident from Fig. 5a that all the doped HA coating leads to a statistically significant reduction in bacterial colony formation at each time point compared to control. According to Fig. 5b, the Zn-HA shows ~24 ± 3% less bacterial cell viability than the control after 24 h. After 48 and 72 h, the same composition results in ~28 ± 4% and ~26 ± 2.5% antibacterial efficacy, respectively. The F-HA composition show ~35 ± 3% more antibacterial efficacy than the control after 24 h. This composition leads to ~37 ± 4% and 40 ± 2% antibacterial efficacy after 48 and 72 h, respectively. The Zn–F–HA shows an antibacterial efficacy of ~70 ± 4% after 24 h, which further changes to ~67 ± 2% after 48 h and 68 ± 4% after 72 h.

Fig. 5.

(a) Antibacterial efficacy against S. aureus reveals all doped samples have a significantly lesser number of bacterial colony formation on the agar plate, compared to control after 24, 48, and 72 h of sample – bacteria interaction. (b) Zn- HA shows ~28% less bacterial cell viability compared to control, F-HA composition shows ~40% less bacterial cell viability compared to control HA. In addition, ~70% less bacterial cell viability is observed in the Zn – F – HA coated surface. ** indicates p < 0.001.

The confocal microscopic images (Fig. 6) indicate that the treatment samples consist of many dead bacteria (red dots) compared to the control. The green dots on control samples indicate no antibacterial efficacy for the control at any time point. The highest number of red dots are observed in the Zn–F–HA-coated surfaces.

Fig. 6.

Confocal microscopic images after each time point of antibacterial efficacy assessment indicate that the treatment samples show a significant number of dead bacteria (red dots), compared to the control. The control samples at all time points indicate a high concentration of live bacteria (green dots). The Zn – F – HA samples show the highest concentration of dead bacteria among the tested compositions, indicating superior antibacterial efficacy for this composition.

The FESEM images in Fig. 7a indicate the embedding of bacterial cells throughout the coating. High magnification inset images for HA at 24 h show the S. aureus morphology on the sample surface. The combined effects of Zn²⁺ and F⁻ lead to bacterial cell rupture, as evident from the high magnification inset FESEM image of Zn–F–HA. It is interesting to note that the HA coating shows a very similar S. aureus morphology at 72 h, compared with 24 h. In contrast, the Zn–F–HA leads to bacterial cell damage and debris formation, as evident in the high magnification inset FESEM image.

Fig. 7.

(a) The FESEM results indicate embedding of bacterial cells throughout the coating. The high magnification inset FESEM images are shown for HA and Zn – F – HA samples. The Zn – F – HA samples show a dilated bacterial cell wall when compared with the control. In the FESEM images, bacteria are marked with the “→” symbol. The inset high magnification Zn – F – HA sample image at 72 h shows the presence of debris, generated due to bacterial cell death. The debris is marked with a dotted circle. (b) proposed antibacterial mechanism, due to ROS generation and contact killing of S. aureus in presence of Zn²⁺ and F⁻ in the coating.

The proposed antibacterial mechanism in corroboration with agar plate, FESEM, and confocal microscopy results are represented in Fig. 7b. The presence of zinc and fluoride in the coating leads to reactive oxygen species (ROS) generation and causes bacterial cell death by contact killing. Zn²⁺ also binds with PsaBCA transporters and hinders resistance against oxidative stress, leading to bacterial cell death [29].

The MTT assay results (Fig. 8a) indicate that on day 3, Zn-F-HA shows slightly increased cell viability compared to control HA. Cytocompatibility of all tested compositions is confirmed on day 3, as per the ISO 10993 standard [30]. On day 7, all compositions show increased cell viability than on day 3. In addition, the doped compositions show higher cell viability compared to HA at day 7. The Zn-F-HA shows ~1.5 times higher cell viability than that of pure-HA. The SEM images in Fig. 8b represent healthy cellular attachment, covered with an apatite layer on the sample surface. Extended cellular filopodia are noticed on the surface.

Fig. 8.

(a) MTT assay with osteoblast cells until day 7 on the coated surface. No compositions are cytotoxic. Zn – F – HA shows slightly higher cell viability at day 3, compared to control. On day 7, cell viability for Zn-HA and F-HA increases significantly compared to control HA, and the Zn – F – HA composition shows ~1.5 times higher cell viability compared to control (b) FESEM images at days 3 and 7 show healthy osteoblast cells on each coated surface, the cells are covered with apatite layer. The identified cells are marked with the “→” symbol.

4. Discussions

4.1. Requirement of inherent biological properties for the HA-coated Ti implants

Increasing demand for skeletal reconstruction among younger adults due to active modern lifestyle fuels the need to design implants with longer service life. It is predicted that by 2030, total hip arthroplasty will increase by ~71% [5,31]. In addition, more than 50% of total hip and knee prostheses will be implanted in patients below 65 years of age. This number is ~20% higher than the current statistics. Our work aims to design doped HA-coated Ti64 implants with enhanced biological properties for load-bearing applications. We have successfully prepared Zn²⁺ and F⁻ co-doped HA-coated Ti64 using RF induction plasma spray to achieve this goal. XRD results (Fig. 2b) indicate coating fabrication with the retention of HA phases. During the high-temperature plasma spraying process, HA often degrades to the tricalcium phosphate phase, which results in faster coating degradation. No such co-existing phases are observed. Utilization of supersonic plasma nozzle contributes to-wards this phase purity [25]. From a clinical point of view, a bacterial infection is one of the significant causes of implant failure. Inherent antibacterial properties of the coatings will increase the chances of long-term implant stability [32]. Doped samples show antibacterial efficacy against S. aureus, due to the bactericidal potential of Zn²⁺ and F⁻. This will broaden the application of these coated implants.

4.2. Bridging the gap between doping level and biological properties

Both Zn²⁺ and F⁻ are essential to maintain healthy musculoskeletal health. However, over-exposure to these elements leads to cytotoxicity and related complications. The US food and drug administration (FDA) recommended intake limit for Zn²⁺ is 8 mg per day per adult woman and 11 mg per day per adult man [13]. The daily tolerable upper intake level of fluoride is 10 mg, per adult [33]. We have performed a detailed compositional optimization study to find a balance between this co-doped HA system’s antibacterial efficacy and cytocompatibility. Table 1 shows the detailed composition of each tested sample. Initial cytotoxicity assessment on day 3 (Fig. 1a) indicates that ZnO doping level of 0.5 wt% and CaF₂ level of 0.6 wt% leads to osteoblast cell death because of overexposure to Zn²⁺ and F⁻, and these compositions are eliminated from further studies. The antibacterial potential of the remaining compositions (Fig. 1b) indicates that the presence of Zn²⁺ and F⁻ results in antibacterial efficacy against S. aureus within 24 h. The highest antibacterial efficacy of ~70% is obtained for a ZnO doping level of 0.25 wt% and CaF₂ level of 0.3 wt%. This composition is selected for further studies based on the combinations of initial cytocompatibility and antibacterial efficacy results.

4.3. Parametric optimizations to mitigate the challenges with dopants

Among the chosen dopants, Zn²⁺ has an ionic radius of ~0.74 Å and it is expected to replace the Ca1 or Ca2 sites of the HA crystal structure. The ionic radius of F⁻ is 1.33 Å and it goes into the OH⁻ sites of the HA [7,11]. Few recent works report that fluoride-doped hydroxyapatite is emerging as an alternative to hydroxyapatite due to its higher stability in the biological medium, excellent osteogenic properties, better osseointegration, and protein adsorption, compared to HA [3,34]. One previous work reports atmospheric plasma spraying of fluoride-doped hydroxyapatite on Ti64 substrate. However, the possibility of fabricating Zn²⁺ and F⁻ co-doped HA-coated Ti64 using RF plasma spray has not been reported yet [35]. The current work provides a detailed parametric optimization strategy to mitigate the challenges with F⁻. Table 2 reports the parametric combinations which lead to a successful coating. Interestingly, both the HA and Zn-HA require minimal optimizations, and a similar parametric combination for both powders results in a coating thickness of 80–145 μm. However, both the fluoride-containing compositions require extensive optimizations. The anomalous behavior of F-HA and Zn-F-HA can be attributed to the reactivity and hygroscopic nature of F⁻. This observation is corroborated well with the measured water content of ~4.1 ± 0.4% for F-HA compared to only 0.35 ± 0.08% for the HA powder. The angle of repose test indicates 46 ± 3° repose angle for F-HA. According to previous works, this range shows the cohesive nature of the powder with poor flowability [23,28]. The water content and angle of repose together indicate that fluoride’s hygroscopic and agglomerating nature adversely affects the powder flowability during plasma spraying. This challenge is mitigated by multiple parametric optimizations, as mentioned in Table 2. We have reported the workable parameters and obtained coating thicknesses as a function of frequency and amplitude in Fig. 2d, e, and Table 2. This workable parametric database will guide future researchers to generate new algorithms and support industry 4.0 automation [36].

4.4. Importance of adhesion strength measurement

The adhesion strength between the coating and substrate plays an important role in determining load-bearing implants’ clinical success. According to the ISO 13779-2 standard, the adhesion strength of the coating should be ≥15 MPa for clinical approval [37]. In our work, the adhesion strength of the fabricated coatings is higher than the minimum requirement mentioned in the ISO standard. The similar adhesion strength of all tested compositions indicates that the dopants do not have any adverse effects on the coating property. Adhesion properties of zinc and fluoride co-doped HA-coated Ti64 have not been reported elsewhere.

4.5. Enhancement of biological properties due to doping

Fig. 8a shows that both Zn-HA and F-HA show significantly enhanced osteoblast cell viability at day 7, compared to control HA. The Zn-F-HA shows a ~1.5 folds increase in the cell viability after 7- days of culture. This increase in cell viability can be attributed to the osteogenic potential of dopants. Recent works suggest that mitogen-activated protein (MAP) kinase and tyrosine kinase pathways get activated due to Zn²⁺ and lead to the maturation and growth of osteoblast cells [38]. Whereas F⁻ upregulates the tyrosine-phosphorylated proteins and helps in new osteoblast formation. However, the exact process of activating the phosphorylated proteins by F⁻ is debated in the literature. One work suggests that activation of MAP kinases, guanine nucleotide-binding proteins (G proteins), and phosphatases leads to tyrosine-phosphorylated protein activation [39]. In contrast, another work suggests that calcium influx into the cell in the presence of fluoride acts as a signal for osteoblast growth [40]. The potential of the coating for enhanced osteoblast growth is further supported in FESEM (Fig. 8b), which shows healthy cellular attachment covered with apatite layers.

S. aureus is the predominant bacteria responsible for osteomyelitis, leading to implant failure. To ensure the inherent antibacterial effects of the coating, three-time points are chosen as 24, 48, and 72 h. Fig. 5a and b shows that each dopant reduces the number of bacterial colonies after the coating and bacterial interaction. The Zn-F-HA coating shows the highest antibacterial efficacy of ~70% due to the presence of antibacterial Zn²⁺ and F⁻. The retention of antibacterial efficacy at 48 and 72 h is also confirmed. Fig. 6 shows the live and dead bacteria. The control sample shows only live bacteria as green dots. However, the treatment samples show a significant number of dead bacteria as red dots at each time point. The highest number of dead bacterial colonies is noticed for the Zn–F–HA composition. Previous works suggest that interaction of bacterial cell membrane with Zn²⁺and F⁻ leads to reactive oxygen species (ROS) generation, which ruptures the bacterial cell wall and causes the death of the bacteria. The high magnification inset FESEM images (Fig. 7a) for this composition show a dilated bacterial cell wall after 24 h. In addition, debris that originated due to bacterial cell death is visible in the inset FESEM image of Zn–F–HA after 72 h. One recent work reports that manganese ion (Mn²⁺) protects bacterial cells against oxidative stress [29]. Interaction of bacterial cells with Zn²⁺ results in the binding of PsaBCA transporters with Zn²⁺, which restricts the Mn²⁺ intake and kills the bacteria. Hence, a combination of agar plate results, live dead assay, and FESEM confirms the antibacterial efficacy of the doped HA compositions. Fig. 7b shows the proposed antibacterial mechanism due to ROS generation followed by oxidative cell damage and bacterial cell death in the presence of Zn²⁺ and F⁻.

The contributions to science from this work can be outlined as (a) successful fabrication of Zn²⁺ and F⁻ co-doped HA-coated Ti64 using RF induction plasma spraying, with eliminating the associated challenges due to F⁻ (b) reporting of detailed plasma spraying process parameters to guide future researchers, and (c) demonstration of anti-bacterial properties and enhanced cell viability on the coated surface. Future works may be aimed to develop a generic algorithm with our data and other reported works to computationally establish the correlation between the dopant level and obtained biological properties as a function of process variables. Those automated approaches will foster new material design and development, focusing on customized orthopedic applications.

5. Summary

Our work indicates that the anti-bacterial and biological properties of HA can be enhanced with optimized Zn²⁺ and F⁻ co-doping. This co-doped HA is successfully deposited on Ti6Al4V substrate using RF induction plasma spraying technique. The agglomeration and hygroscopic tendency of fluoride hinder the powder flow during the coating process. Multiple parametric optimizations are required to achieve the proper coating thickness (80–145 μm) for the fluoride-containing compositions. In contrast, pure HA coatings require minimal optimization. Frequency and amplitude control are the two most important parameters that lead to a successful coating preparation. The fabricated coatings show an adhesive strength of >15 MPa, indicating the applicability of these coatings as a load-bearing implant. The presence of osteogenic Zn²⁺ and F⁻ shows enhancement of osteoblast viability after 7 days of cell-material interaction. The doped HA coated surfaces show antibacterial efficacy against S. aureus. The workable parametric combinations reported in our work to fabricate plasma sprayed zinc and fluoride doped HA coatings are expected to guide future researchers, and support industry 4.0 automation without multiple trial and error approaches. These cytocompatible and antibacterial coatings can find application as a load-bearing orthopedic implant material such as total hip replacement, particularly in revision surgeries and immune-compromised patients.