Bioactive Glasses as Alternatives to Hydroxyapatite for Bone Implant Coatings: A Comparative Study (2025)

2.1 Powder and Substrate

Four distinct powders were employed in the fabrication of the bioactive coatings. From the SiO₂-CaO-Na₂O-P₂O₅ system, two commercial glasses were selected: the 45S5 powder (Denfotex Research, United Kingdom) and the S53P4 manufactured at the Instituto de Cerámica y Vidrio (ICV) from a mixture of SiO₂, (NH₄)₂HPO₄, Na₂CO₃, and CaCO₃. Another glass from the SiO₂-CaO-P₂O₅-MgO system, the 62W, was also obtained at ICV [22] through the reagents SiO₂, Ca₃(PO₄)₂, CaCO₃, and MgO. Additionally, HA powder, used as a comparative material, was sourced as commercial sintered HA powder Captal30 from Plasma Biotal Limited, United Kingdom. The specific composition of the bioactive glasses [10] is detailed in Table1.

Bioactive glass powders were synthesized via the conventional melt-quenching method. Each composition was prepared by mixing the aforementioned reagents. The mixtures were melted using a platinum crucible at high temperatures (1450°C for S53P4 and 1500°C for 62W) in an electric furnace for 2 h, with a heating rate of 5°C/min. The molten material was poured onto cold water to form glass frits and prevent crystallization. The fritted glass was subsequently milled using a tungsten-carbide vibrating cup mill to obtain glass powder. The resulting powder was sieved using 63 and 100 μm mesh sizes with a vibratory sieve shaker (AS200, Retsch, Haan, Germany). Powders ranging between 63 and 100 μm were collected and utilized for coating production. Additionally, 0.7% wt of Aerosil was added to the final powder to enhance its flowability. The chemical composition of the produced powders was quantified using x-ray fluorescence spectrometry (XRF) (MagicX PW-2424, Malvern Panalytical Ltd., Malvern, UK) with an RX tube and a generator of 2.4 kW, using a Pt–Au crucible.

Ti6Al4V was the material selected for the substrates. For the invitro tests, discs measuring 2 mm thick and 9 mm in diameter (Tamec, Spain) were coated. Substrates for the bond strength test (Ibermetal, Spain) were 25 mm in diameter and 5 mm in height. Rectangular substrates measuring 100 × 20 × 5 mm (Ibermetal, Spain) were utilized for the metallographic characterization of the coatings.

2.2 Coating Deposition

The coatings were fabricated employing APS equipment (Plasma-Technik A3000S, Sulzer Metco AG, Wohlen, Switzerland) utilizing an F4 plasma torch. Argon served as the primary carrier gas, while hydrogen facilitated plume formation as the secondary gas. During APS, the powders are (at least partially) melted and accelerated to the substrate. The temperatures reached during the process are adjusted by varying the argon and hydrogen flow rates to reach the melting temperature of the powder material to be deposited. Thus, the coating is formed by flattened particles that adhere to the substrate and to the already adhered particles [50].

The powders were applied onto substrates that had been previously grit-blasted (MAB-4, MAB industrial, Barcelona, Spain) with corundum G24 (grit size 800 μm) at 0.5 MPa, which produces a roughness surface of R_a = 5.1 ± 0.5 μm and R_z = 29.8 ± 4.0 μm. This procedure generates more surface irregularities that favor the particles' mechanical anchoring [51, 52]. The parameters employed for coating deposition are outlined in Table2. These conditions were determined following optimization of the coating formation for each powder, which involved adjusting the spray distance, current, and gas flow rates to achieve a well-adhered coating with similar thickness.

In particular, for the deposition of HA, this powder has a minor particle size distribution compared to bioactive glasses, so the spraying distance must be adjusted to ensure that the particles reach the substrate in a molten state at short distances. If the spraying distance is too long, the particles may cool down too much before reaching the substrate, reducing adhesion. Therefore, for this material, the spraying distance resulting from the optimization of the coating is consistent. It needs to be mentioned that, for 45S5 glass, when the spray distance increases, a less porous microstructure is obtained that results in a better-adhered coating; this phenomenon was analyzed in detail by Rojas [53]. Although S53P4 glass has a similar composition to 45S5, it has a different amount of each component, which changes the thermal behavior and viscosity of the glass, allowing a better-adhered coating at a shorter distance, keeping the same plasma enthalpy. For the 62W powder, a longer distance resulted in better adhesion with the substrate.

2.3 Powder and Coating Characterization

An examination was carried out to analyze the initial powders and the resulting coatings, utilizing a scanning electron microscope (SEM) with energy-dispersive x-ray diffraction capabilities (Phenom ProX, Phenom-World BV, Eindhoven, The Netherlands). Prior to observation, the specimens were coated with gold to improve conductivity. This coating process was accomplished using a sputtering coating system (E-5000, Polaron, Watford, England).

The device (Surftest 301, Mitutoyo, Kawasaki, Japan) was used to measure roughness parameters R_a and R_z, with R_a representing the average absolute deviation of the roughness irregularities from the mean line over one specific length and R_z representing the difference in height between the average of the five highest and lowest peaks and valleys in a specific length of [54]. For this evaluation, five measurements were done for each coating type.

To investigate the cross-sections of the coatings, a grinding and polishing procedure was utilized. Initially, specimens were embedded in resin and then subjected to abrasion using silicon carbide papers up to a grit size of 5 μm. Following this step, the samples were polished up to 1 μm using diamond slurry. Furthermore, the composition of the sprayed coatings was determined from the cross-section, using an SEM (Quanta-200, FEI, Eindhoven, The Netherlands) equipped with an energy-dispersive x-ray detector (UltraDry, Thermo Fisher Scientific, Madison, USA).

The bond strength of the coatings was assessed utilizing a mechanical testing machine (ME-402/10, Servosis, Madrid, Spain) in accordance with ASTM C633-13 standard.

Three samples of each composition were glued to counter-test pieces using HTK ULTRA BOND 100 glue (HTK, Germany). Subsequently, a normal tensile stress was applied to the coating at a displacement rate of 0.02 mm/s until fracture occurred. Furthermore, the coatings were tested after one day of immersion in Hank's balanced salt solution (HBSS) (Sigma-Aldrich, Germany) at 37°C.

2.4 Bioactivity and Degradation Assessment

To assess the coatings' ability to generate an HCA layer, each coated sample was immersed vertically in 50 mL of HBSS in a thermostatic bath with agitation at 37°C. To prevent ionic saturation of the solution, it was refreshed twice a week. Each coating type's samples were tested in triplicate for 3, 7, 14, and 21 days. Following each period, the coated samples were rinsed with ultra-pure water and air-dried at room temperature for 24 h. The sample surfaces were examined via SEM for each time interval to investigate the kinetics of apatite layer formation. The crystallographic structure of the as-sprayed coatings and samples immersed in HBSS for 14 days was analyzed using x-ray diffraction (XRD) with a diffractometer (X'Pert PRO MPD, PANalytical, Cambridge, UK). The analysis was done using the radiation source α-line of Cu (l = 1.5418 Å), with a tube voltage of 45 kV and a current of 40 mA. The scan was done in a 2θ range between 20° to 70° with a step size of 0.017° and a measuring time of 80 s per step. The International Centre for Diffraction Data (ICDD) was used as reference data. To determine the thickness of the formed layer, the cross-sections of the coatings on samples immersed for 21 days were analyzed. The samples were embedded in resin, ground using silicon carbide papers, and polished with diamond slurry down to 1 μm. Finally, the prepared samples were gold-coated before SEM observation.

The dissolution process of the different compositions was evaluated by a degradation test. Samples were immersed in 10 mL of a Tris–HCl solution at 37°C ± 1°C for 120 h, with pH adjusted to 7.4 ± 0.1. Following immersion, the samples were rinsed with ultra-pure water and dried overnight at 120°C. To evaluate degradation behavior, the weight of the dried samples was recorded before and after the test using a high-precision scale (CPA225D, Sartorius, Goettingen, Germany). The elements released into the solution were quantified by measuring ion concentrations at various intervals using inductively coupled plasma optical emission spectrometry (ICP-OES) (Optima 8300, PerkinElmer, Waltham, USA).

2.5 Cell Culture Studies

Three coatings from each series were seeded with osteoblasts to investigate their capacity to support cell adhesion, growth, and proliferation. The experiments were conducted in triplicate using various human osteoblast lines (obtained from knee trabecular bone post-prosthesis replacement [55]) ranging from passages 3 to 6 to ensure consistency in results. Approval for the study was obtained from the Parc de Salut Mar Ethics Committee. To mitigate inter-experimental variability, results were standardized against tissue culture plastic (TCP) at three days within each experiment, which served as a control.

Prior to cellular testing, the samples underwent sterilization in 70% ethanol for three hours to prevent contamination during experimentation. Subsequently, a preconditioning stage was conducted, immersing all samples in Dulbecco's modified Eagle's medium (DMEM) (Invitrogen, USA) supplemented with 10% fetal bovine serum for 24 h. This precondition step was necessary to prevent cell death from the abrupt pH increase caused by glass ion release [56, 57], since the test is done in a static cell culture system. A cell suspension was prepared and seeded onto the coated samples and TCPs at a density of 6.5 × 10³ cells per sample in supplemented DMEM. Incubation was carried out at 37°C in a humidified atmosphere containing 5% CO₂, with media changed every three days.

Cell proliferation was assessed using the MTS assay (CellTiter 96 AQueous One Solution Cell Proliferation, Promega, USA). This colorimetric method measures viable cells by the reduction of MTS tetrazolium into a colored formazan product, which is soluble in the cell culture medium. At each time point (3, 7, and 14 days), the medium was aspirated, and the samples were transferred to fresh wells. Subsequently, the samples were incubated for 90 min in a solution containing 50 μL of MTS reagent and 250 μL of supplemented medium. Following incubation, absorbance was measured at 490 nm using a microplate reader (Infinite 200, Tecan, Männedorf, Switzerland).

Quantitative data obtained from the MTS assay were subjected to one-way analysis of variance (ANOVA), followed by Tukey's post hoc test for group comparisons. Statistical significance was defined as p < 0.05.

SEM was employed to examine cell attachment and morphology. Osteoblasts were seeded onto the coatings at a density equivalent to that used in the MTS assay. After seven days of incubation, samples were rinsed twice with phosphate-buffered saline (PBS) buffer (pH 7.4) to eliminate unattached cells. The remaining cells were fixed in 2.5% glutaraldehyde in PBS for 3 h, followed by another PBS rinse. The cells were dehydrated through successive ethanol baths, each lasting 15 min, with increasing concentrations: 50%, 65%, 70%, 80%, 90%, 95%, and finally 100%. Subsequently, the samples underwent drying using a critical point dryer (CPD) (K850, Emitech, Lewes, UK) and were carbon-coated for SEM observation using a high-vacuum carbon evaporator (K950X, Emitech, Lewes, UK).

Cell response can be influenced by surface roughness [58], therefore, additional tests were conducted on smooth samples to ensure that any variances in cell response were attributed to composition rather than surface roughness. To analyze surfaces without significant topographical features, the coatings were abraded until a surface roughness (R_a) of less than 5 μm was reached.

3.1 Powder and Coating Characterization

The hydroxyapatite (HA) powder comprises micrometric spherical particles (below 60 μm) arranged in small aggregates (Figure1A). On the other hand, the three bioactive glass powders exhibit the characteristic morphology of crushed powders [50], featuring irregular and sharp particles (Figure1B–D). The chemical composition of the glass powders is detailed in Table1; the composition of the obtained powders is close to the theoretical ones.

Figure2 shows cross-sectional images of the coatings obtained using different feedstock powders. Overall, thick and well-adhered coatings were achieved for all materials. However, thermal stresses led to the formation of some cracks in the glass coatings. Microstructural analysis revealed porosity in the 45S5 and S53P4 coatings, with larger pores observed in the 45S5 coating. Such microstructures are typical for bioactive glasses produced by APS [17, 59, 60]. Porosity can arise from particles that incompletely melt, not achieving good flattening and creating voids during coating build-up, while volatilization of sodium and phosphorus oxides at high temperatures can also contribute to porosity [53]. In contrast, the 62W composition exhibited a dense, non-porous coating. This result can be explained by the absence of sodium in its composition but also by the presence of magnesium oxide, which favors the good melting of the particles, since the 62W composition exhibits a wide temperature window due to exhibiting a dual Tg behavior [61].

Table3 presents the roughness and thickness results, indicating increased values of thickness and roughness when glass powders were used instead of HA. These results can be explained by the glass powders' larger size and sharp morphology compared to HA powders. As the size of the glass powders is larger than that of the HA ones, even though the particles melt during the deposition process, the flattened particles generate a less uniform thickness.

3.2 Bond Strength Study

To assess the quality of the deposited coatings, bond strength tests were conducted on both the as-sprayed coatings and those immersed in a physiological solution for one day. Figure3 shows the results for each type of coating. The investigation revealed that HA-coated samples exhibited superior adhesion before immersion. Among the glass coatings, the highest values were observed for the coating with less porosity in its microstructure, the 62W, followed by the S53P4 and 45S5 coatings, respectively.

To promote effective osseointegration between the implant and bone, it is crucial for the coating to exhibit good stability in the initial stage. However, the HA coating experienced the most substantial percentage of bond strength loss after one day of immersion (81%), indicating its inadequate suitability in this regard. Similarly, the 45S5 glass coating demonstrated the lowest adhesion value both before and after immersion, with a corresponding 50% loss in adherence. Consequently, from a mechanical standpoint, this coating fails to meet the necessary criteria (considering that 15 MPa is the minimum required for thermally sprayed HA coatings according to ISO 13779-2:2018). The most promising outcomes emerged from the S53P4 and 62W coatings, which experienced reductions in bond strength of 67% and 40%, respectively, after the immersion test. Particularly, the 62W coating retained a bond strength of 19 MPa after immersion, being significantly superior to the other coatings.

3.3 In Vitro Bioactivity Study—Ability to Form Apatite

The ability of the coatings to enhance bonding with bone tissue was assessed for all samples. The test results are presented in Figures4-6, which correspond to the top surface of the samples, the XRD patterns, and the cross-sectional coatings after immersion, respectively.

After three days in HBSS, the surface of the coated samples revealed a uniform apatite layer on all glass coatings. In contrast, the HA coatings displayed only isolated apatite deposits at this stage. By 7 and 21 days, all coatings exhibited a dome-like morphology of precipitated apatite across the entire surface, indicating the formation and growth of an apatite layer. Figure5 shows the XRD patterns of the as-sprayed coatings and those after three weeks of immersion. The results indicate that the bioactive glass coatings are completely amorphous due to the rapid cooling of the deposited material after spraying. Some peaks corresponding to the apatite layer (ICDD—Reference code: 00-024-0033) were identified in all coating types after three weeks, confirming the bioactive capability of all the studied coatings. A minor peak corresponding to calcite (ICDD—Reference code: 00-047-1743) was detected for 45S5 and 62W coatings, which can appear due to the high amount of calcium ions in the solution [62].

After 21 days of immersion, the cross-sectional images confirmed the formation of a continuous apatite layer across the surface of the coatings (Figure6). A reduction in the thickness of the sprayed coatings was also observed, attributed to its dissolution and the formation of the apatite layer. As seen in Table4, the measured thickness of the formed layer was thinner for HA coating, and it also experienced less coating reduction compared to the glass coatings during the test. Thick apatite layers were observed on the surfaces of the three bioactive glass coatings, which had more than 60% reduction in glass coating thickness. Notably, the 62W composition showed the most pronounced results, with a growth of an apatite layer of 16 μm and a significant reduction in coating thickness from 105 to 37 μm, indicating that the elements in this glass can promote rapid osseointegration.

EDS point scans along the cross-section of the coatings after immersion in HBBS showed the effect of interaction with a simulated body fluid. Bioactive glasses generate a silicon-rich layer at the interface with body fluids by the migration of ions from the glass, which promotes the nucleation of apatite [63]. Consequently, for the glass coatings, a high percentage of silicon is detected before the apatite layer begins, which is attributed to this phenomenon. On top of the formed apatite layer, only calcium and phosphorus were detected, as expected.

3.4 Degradation Study

The degradation rate and ion release of bioactive glasses are influenced by the network's structure, which is closely linked to the bioactive behavior of the coatings. When bioactive glasses are immersed in solution, an ion exchange occurs between the protons in the solution and the modifier cations in the glass, increasing the pH [64] because water molecules can easily enter a disrupted silicate network. Consequently, ion exchange happens more rapidly for glass materials with fewer network formers and lower network connectivity, resulting in greater weight loss and a higher release of ions from the coatings into the solution.

In the degradation test, the coated samples were immersed in a Tris–HCl solution for 120 h. Figure7 illustrates the percentage of weight loss for the coatings over different time periods. All glass compositions showed a gradual increase in weight loss over time, with a more pronounced rise in the initial stage. A variation in degradation among the compositions was also observed.

Specifically, 45S5 glass exhibited the highest dissolution among the glasses studied. This composition experienced a significantly higher weight loss (Figure7) due to its fewer network-forming oxides (Table1) and, consequently, a more disrupted structure. In contrast, S53P4 glass, composed of the same oxides as 45S5 but with fewer modifier oxides and more network formers, demonstrated a much lower dissolution rate and less degradation due to its more connected structure. These findings align with results obtained by other researchers studying these compositions [65].

The 62W coatings presented a higher weight loss rate during the initial periods, indicating a rapid release of ions when in contact with the physiological solution, which tended to stabilize over more extended periods. By the end of the test, the 62W glass exhibited a higher weight loss rate than S53P4 glass but lower than 45S5 glass. The 62W glass has a similar content of network-forming oxides to 45S5 (Table1) and lacks alkaline oxides and instead contains higher amounts of alkaline earth oxides (CaO and MgO). Since alkaline oxides favor solubility, it is consistent that 62W coatings experienced less degradation than 45S5. Lastly, the HA coatings showed the least dissolution due to the low degradation rate of this biomaterial [66].

ICP analysis provided insights into the ion release from each biomaterial during the degradation test. Figure8 illustrates the concentration of silicon, phosphorus, calcium, sodium, and magnesium in the solution.

For the glass coatings, fewer network-forming elements (silicon and phosphorus) were detected in the solution compared to modifier elements (calcium, sodium, and magnesium). This is because network formers are covalently bonded, which also explains the relatively low release of silicon, despite its high content in the glass compositions.

For the HA samples, only calcium and phosphorus were detected in the solutions, indicating their release from the coating. The 45S5 glass exhibited significant dissolution, particularly of the modifier ions. Despite similar amounts of calcium and sodium oxides in this glass, a higher release of sodium was detected due to its weaker bonding within the structure.

In the case of S53P4, the release of calcium and sodium was similar to 45S5, with more sodium than calcium identified by ICP analysis. The silicon levels measured after the test were comparable across all glass compositions. However, less silicon was released from the S53P4 coatings despite its higher silica content, likely due to the greater number of network formers in this glass. These results can also be correlated with the network connectivity of the glass composition, which describes the degree of polymerization in a glass network, based on the number of bridging oxygens (BOs) per network-forming cation. The S53P4 composition has a network connectivity of 2.54 as it has a more cross-linked network, resulting in slower degradation and controlled ion release, while the 45S5 has a network connectivity of 2.11, which has a less polymerized structure, allowing faster dissolution and ion exchange[10].

The 62W coatings showed a rapid release of silicon in the initial periods, a trend also observed in the weight loss rate during the early stages. The amount of calcium detected for 62W glass was significantly higher than for the other compositions, reflecting its higher calcium oxide content. Magnesium was also detected in the 62W solutions, as it is present in the composition in a small proportion (see Table1). The role of magnesia in silicate glasses is not entirely clear and may vary depending on its proportion in the composition. Some studies suggest it acts as an intermediate oxide [67, 68], while others suggest it functions as a modifier oxide [69]. The amount of magnesium released during the degradation test in the 62W composition suggests that it might be acting as a modifier.

3.5 Cell Culture Study

The composition of the coatings affects both their degradation and their bioactive response. Therefore, the cellular response is also expected to be influenced by the different elements present in each composition.

Osteoblasts were seeded on coated samples, and TCP was used as a positive control. The MTS results for osteoblast proliferation at different time periods are shown in Figure9. After three days, comparable cellular activity was obtained among the bioactive coatings, indicating similar initial adhesion of osteoblasts for the different coatings. However, after seven days, there was a significant increase in the values of the glass coatings compared to the HA coatings. After 14 days, the 45S5 and 62W glasses promoted higher cell proliferation than the other studied coatings. Other researchers have also reported a stimulating effect on osteoblast proliferation for compositions containing small amounts of magnesia [21, 39]. Based on the MTS results, it can be clearly seen that the glass composition significantly affects the growth and proliferation of osteoblasts.

The morphology of attached osteoblasts on the surface of coatings after seven days is shown in Figure10. For all coatings, well-adhered and spread cells with long filopodia were observed, maintaining the typical morphology of osteoblasts. Since coating roughness can affect osteoblast interaction with the surface, additional tests were conducted on smooth samples (R_a < 5 μm). This helps clarify if differences in cell response are merely due to the composition. Figure11 shows cells attached to smooth bioactive coatings after seven days. From these images, it can be concluded that cell response is not influenced by the roughness of the coatings, as well-adhered and spread cells were also observed on smooth surfaces. Additionally, a higher number of cells were distinguished on the 62W coatings for both rough and smooth surfaces, suggesting a preference for this glass composition, as reflected in the MTS analysis.