Sol-gel TiO2-based coatings in 3D printed porous Ti-6Al-4V alloy structures as efficient antibacterial drug delivery systems: Thorough structural and biological characterization

Full Length Article Sol-gel TiO 2 -based coatings in 3D printed porous Ti-6Al-4V alloy structures as efficient antibacterial drug delivery systems: Thorough structural and biological characterization Anna Knaislov´ a a,* , Diana Horkavcov´ a b , Eva Jablonsk´ a c , Dalibor Vojtˇ ech a a Department of Metals and Corrosion Engineering, University of Chemistry and Technology, Prague. Technick´ a 5 166 28 Prague, Czech Republic b Department of Glass and Ceramics, University of Chemistry and Technology, Prague. Technick´ a 5 166 28 Prague, Czech Republic c Department of Biochemistry and Microbiology, University of Chemistry and Technology, Prague. Technick´ a 5 166 28 Prague, Czech Republic ARTICLE INFO Keywords: Dodethick Gyroid Sol-gel coatings 3D printing Hydroxyapatite silver ABSTRACT The application of sol-gel coatings on titanium-based materials offers a promising approach for enhancing their bioactivity, antibacterial properties, and adhesion, particularly for biomedical applications. This study focuses, for the first time, on the preparation and characterization of sol-gel TiO 2 -based coatings containing hydroxyapatite and silver in 3D-printed porous gyroid and dodethick structures. TiO 2 -based coatings on the standard wrought Ti-Al-V alloy rods were used as a reference. The coatings were applied via the specific dip-coating process developed by the author team. The microstructural analysis revealed that the sol-gel coatings on the reference wrought rod samples were homogeneous and well-adhered. The coatings on the porous gyroid and dodethick structures exhibited some localized cracking due to the complex geometry of the porous structures. Bioactivity was evaluated through the standard in vitro simulated body fluid tests, confirming hydroxyapatite precipitation on HA-containing coatings. Antibacterial properties were assessed against Escherichia coli, demonstrating nearly 100 % bacterial inhibition for Ag-containing coatings. Cytotoxicity tests with L929 fibroblast cells indicated that coatings with lower Ag concentrations in sol were non-toxic, while higher Ag concentrations in sol resulted in reduced cell viability, particularly in gyroid structures. 1. Introduction Titanium and its alloys, particularly Ti–6Al–4 V, are widely used in biomedical applications due to their excellent mechanical properties and biocompatibility, making them suitable for long-term implantation [1]. Sol-gel titania-based coatings applied to Ti-6Al-4 V biomedical alloy fabricated via 3D printing represent a growing area of research, especially in biomedical and engineering fields. These coatings are primarily utilized to enhance surface properties, including bioactivity, adhesion, corrosion resistance, antibacterial performance and photocatalytic activity [2,3]. While they have been applied to compact, planar implant surfaces, there are no scientific reports, to our knowledge, that explore their application on 3D-printed porous Ti-6Al-4 V structures. Porous architectures, however, offer a very high specific surface area, which makes them suitable candidates for hosting drugs or antibacterial agents. Gyroid and dodecahedron lattice structures are particularly advantageous due to their ability to mimic the architecture of natural trabecular bone [4,5]. Gyroids, being triply periodic minimal surfaces, provide a high surface-area-to-volume ratio and excellent mechanical strength, [6] making them ideal for bone scaffold applications [7,8]. Likewise, dodecahedron (dodethick) structures are complex porous networks that offer high load-bearing capacity and tunable porosity, improving mechanical stability, fluid permeability, and reducing stress shielding [9,10]. The sol-gel process involves the transformation of a colloidal solution (sol) into a three-dimensional gel network, allowing precise control over coating composition, thickness, and surface characteristics [11]. Titania-based sol-gel coatings are particularly advantageous for biomedical applications due to their biocompatibility, osteointegration capability, antibacterial activity, and resistance to corrosion [12–14]. Coatings doped with hydroxyapatite (HA) and silver nitrate (AgNO₃) can further enhance bone integration and introduce antibacterial properties [15]. Post-deposition heat treatment improves coating adhesion and crystallinity, converting amorphous TiO₂ into more stable anatase or * corresponding author. E-mail address: [email protected] (A. Knaislov´ a). Contents lists available at ScienceDirect Applied Surface Science Advances journal homepage: www.sciencedirect.com/journal/applied-surface-science-advances https://doi.org/10.1016/j.apsadv.2025.100816 Received 27 May 2025; Received in revised form 21 July 2025; Accepted 6 August 2025 Applied Surface Science Advances 29 (2025) 100816 Available online 14 August 2025 2666-5239/© 2025 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). rutile phases, thereby improving long-term durability and biological performance [16,17]. On 3D-printed titanium substrates, sol-gel coatings foster cell attachment and proliferation while forming a protective layer against oxidation and degradation [16,18]. The inclusion of silver also provides an antimicrobial effect, helping prevent post-implantation infections [19]. Several studies conducted at our university have focused on the development and characterization of titanium-based coatings on titanium substrates for biomedical applications. These included the use of sol-gel and dip-coating techniques with various additives such as silver, calcium, and phosphate to improve implant performance. The coatings were evaluated for adhesion, antibacterial activity, bioactivity, and cytocompatibility, showing promising results such as strong adhesion, antimicrobial effects against Escherichia coli and S. epidermidis, promotion of hydroxyapatite formation, and non-toxicity to cells. Some of our research also explored coatings on porous titanium, TiSi alloys, and the effects of calcium titanate powder additions to sol-gel formulations [20–27]. However, the application of sol-gel coatings to 3D-printed porous structures remains unstudied. This is particularly challenging due to the need for complete and uniform contact between the viscous sol and the internal surfaces of fine pores. Therefore, the aim of this study was to prepare and characterize sol-gel titania-based coatings, with silver (Ag) and commercial hydroxyapatite (HA, Ca₁₀(PO₄)₆(OH)₂), on three different titanium-based substrates: a Ti-6Al-4 V alloy rod, a 3D-printed porous Ti-6Al-4 V gyroid structure, and a 3D-printed porous titanium dodethick structure. The coatings were assessed in terms of bioactivity and antibacterial properties to evaluate their potential for biomedical applications. 2. Material and methods 2.1. Sample preparation The initial substrate for the sol-gel coating was a machined wrought TiAl6V4 alloy rod, 15 mm in diameter and 3 mm in height, with a surface area of 4.94 cm 2 . This rod underwent grinding with P400, P600, and P800 grit silicon carbide abrasive papers (grinder Buehler MetaServ 250). Subsequently, the rod was cleaned ultrasonically in ethanol for 10 min. The second substrate for the sol-gel coating was a 3D-printed TiAl6V4 alloy gyroid, 15 mm in diameter and 3 mm in height, with a surface area of 9.90 cm 2 , calculated average porosity 66 % (Fig. 1a) [28–31]. The gyroid was ultrasonically cleaned in ethanol for 3 ×10 min. The third substrate for the sol-gel coating was a 3D-printed titanium dodethick (lattice: 1.35), 15 mm in diameter and 3 mm in height, with a surface area of approximately 14 cm 2 , calculated average porosity 63 % (Fig. 1b) [28–31]. The dodethick was ultrasonically cleaned in ethanol for 2 ×10 min. 2.2. Composition and conditions for the preparation of sols The basic titanium sol (T sol) was prepared by gradually mixing tetra-n‑butyl‑orthotitanate (Sigma Aldrich Merck), Triton X-100 (Roth), 1 mol/l HNO 3 (p.a., Lach-Ner), acetylacetone (p.a., Lach-Ner), and ethanol (den.), followed by stirring at laboratory temperature for 24 h and aging at laboratory temperature for 24 h. The aged T sol had a viscosity of 4.8 mPas (viscometer RV1). The basic sol was then divided into three 25 ml aliquots: the first aliquot remained unmodified (T sol), AgNO 3 with a silver concentration of 0.04 mol/l was added to the second, and AgNO 3 with a silver concentration of 0.06 mol/l was added to the third. After dissolving the AgNO 3 , commercial hydroxyapatite with a concentration of 0.2 mol/l was added to the sols (TAN04HA sol, TAN06HA sol). A dip-coater (IDLab) was used for coating. The dipping speed was 20 cm/min, the withdrawal speed was 6 cm/min, and the dwell time in the sol was 30 s, with a sol stirring speed of 200 rpm (IKA). Coating was performed at laboratory temperature. 2.3. Heat treatment conditions for coated substrates The coated samples were dried at laboratory temperature, and then heat-treated in air at 400 ◦C for 120 min using a heating rate of 2 ◦C/min (Clasic). The samples were cooled to room temperature overnight. 2.4. Measurement of the selected properties The adhesion of the coatings to the machined wrought rod substrate was assessed using an adhesive tape test, according to ASTM D3359–02 [32]. A 6 ×6 grid of scratches was made on the coatings. Following tape application (Permacel 99), loading, and peeling, the coating surfaces were visually characterized, and the degree of adhesion was determined using a classification scale. The bioactivity of all coating types on the three substrate types was evaluated according to ISO 23,317 [33]. Coated samples (n =2 per condition) were immersed in a simulated body fluid solution, maintaining a constant surface area-to-volume ratio (S/V), for 21 days at 36.5 ◦C. The simulated body fluid solution was replaced every 7 days. After the in vitro test, the samples were dried at laboratory temperature and characterized. Using atomic absorption spectrophotometer (AAS, Agilent 280 FS) the concentration of Ca 2+ was determined, and UV–VIS (Shimadzu UV2450) was used to measure the concentration of (PO 4 ) 3ions in SBF solutions before and after the 21-day in vitro test. The antibacterial activity of the coatings was measured statically in a flow box (SafeFast Elite) for 24 h at laboratory temperature using E. coli (strain DMB3138). The coated substrates were immersed into suspension of physiological solution (NaCl, 9 g.l -1 ) with bacteria concentration of 10 4 CFU.ml -1 for 24 h at the laboratory temperature in the dark. A consistent surface area-to-volume ratio (S/V) was maintained across Fig. 1. (OM) Surfaces of: a) 3D printed TiAl6V4 alloy gyroid; b) 3D printed titanium dodethick. A. Knaislov´ a et al. Applied Surface Science Advances 29 (2025) 100816 2 sample types. Two independent tests were performed. Following the interaction period, 100 µL aliquots of the suspension were taken and spread onto two Petri dishes with agar (LB). Four dishes from each coating and substrate type were incubated in a biological thermostat at 36.5 ◦C for 24 h. After incubation, the dishes were photographed, and the colonies of surviving bacteria were counted using the computer program NIS-Elements AR 3.10 and visually compared to the reference sample (suspension without contact with the samples). Cytotoxicity measurements for all three coating types on both substrate types were performed according to ISO 10,993–5 [34]. Samples were sterilized in 70 % ethanol for 2 h. After drying, the samples were immersed in MEM +5 % FBS medium (Minimum Essential Medium, Fetal Bovine Serum) with antibiotics in 50 mL tubes at 37 ◦C on an orbital shaker (130 rpm) for 24 h. The extraction ratio was 1.25 cm 2 /ml. Each material was tested in triplicate. L929 cells (ATCC CCL-1) were seeded in a 96-well plate in MEM +10 % FBS medium (10,000 cells/- well) and cultured for 24 h in an incubator (37 ◦C, 5 % CO 2 ). After 24 h, the material extracts were added to the subconfluent cell layer. Six technical replicates (wells) were used for each sample. The extraction medium alone (MEM +5 % FBS with antibiotics) served as the negative control, and the medium with detergent (0.5 % Triton X-100) served as the positive (toxic) control. The silver concentration in the extracts was measured using ICP-OES. Following 24 h of incubation with the extracts, the metabolic activity of the cells was assessed using the resazurin test. The extracts were removed, and a solution of resazurin (final concentration 25 µg/ml) in MEM medium without phenol red was pipetted onto the cells. After 1.5 h, the fluorescence of the formed resorufin (ex/em 560/590 nm) was measured. The relative metabolic activity was expressed as a percentage of the negative (unaffected) control. 2.5. Characterization of the substrates and coatings The substrates and coatings before and after tests were characterised by scanning electron microscope (Hitachi S4700, 15 kV, 20 µA) and by the optical microscope (Olympus BX 51). The phase composition of the TAN04HA, TAN06HA coatings after firing and after in vitro testing was analysed by XRD on the device PANalytical X´Pert PRO with software High Score Plus (30 mA, 40 kV, Cu). 3. Results and discussion 3.1. Microstructure of the prepared samples For uniformity, the surfaces of the machined wrought TiAl6V4 alloy rods underwent mechanical grinding (Fig. 2a). However, traces of scratches remained after cleaning. Subsequent cleaning with ethanol revealed small spherical spherulites, ranging from units to tens of micrometers, on the gyroid samples. While the gyroid surface area was calculated based on cell shape and number, electron microscope images showed a larger actual surface area due to the pervasive presence of these spherulites (Fig. 2b). These surface irregularities may significantly influence the quality and properties of subsequently prepared coatings. Further characterization and testing of these novel gyroid-shaped biomaterials will be essential. Following ultrasonic cleaning with ethanol, the surfaces of the dodethick substrates displayed 3D-printed grid layers (1.35 mm hole size) arranged in a scaffold structure (Fig. 2c). The theoretical substrate surface area was calculated from the grid shape (1.35 mm), the number of printed layers, and the wafer shape. However, electron microscope images revealed that the actual surface area was larger than this approximation. This discrepancy was attributed to the presence of inhomogeneously distributed spherical structures (units to tens of micrometers) on the grid layers. These surface irregularities may significantly affect the quality and properties of subsequently prepared coatings. Following firing, the coatings on the alloy machined shaped bars Fig 2. (SEM) Microstructure of the samples after cleaning: a) the machined wrought rod after mechanical treatment and subsequent cleaning; b) gyroid 5 mm - 3D; c) dodethick 3D. A. Knaislov´ a et al. Applied Surface Science Advances 29 (2025) 100816 3 presented isolated cracks (Fig. 3). These cracks did not have a negative impact on the properties measured. The coating are very thin, and perfectly copies the surface of the substrate, as the morphology typical of grinding is visible. The commercial hydroxyapatite particles, with particle sizes in the micrometer range, and the silver particles, with particle sizes in the nanometer range for both concentrations, were distributed relatively homogeneously throughout the coating surface. Following firing, the coatings on the 3D-printed gyroid alloy exhibited significantly more cracking than those on the machined rod substrates (Fig. 4). This increased cracking is likely due to the irregular surface of the gyroid, which hindered uniform coating flow during withdrawal, compared to the smoother rod samples. The presence of a plate/doughnut structure in the lower part of the gyroid sample further contributed to uneven sol flow during withdrawal. However, all coatings accurately replicated the dissected surface of the gyroid, with no significant alteration of its dimensions, indicating optimal sol viscosity. The distribution of commercial hydroxyapatite particles within the coatings was relatively uniform. Conversely, a closer examination revealed an uneven distribution of silver particles in both concentration cases. The thickness of the basic coating (T, Fig. 4a) was low because small and larger beads of the original gyroid surface were still visible after coating. The coatings with Ag and HA (TAN04HA, TAN06HA) exhibited a significantly greater thickness, as these coatings covered most of the gyroid surface features (Fig. 4b, 4c). Coatings of the basic T sol applied to the 3D dodethick substrate showed cracking (Fig. 5). This is probably due to the irregular surface of the bead grid and the wafer, which resulted in complex coating flow during the extraction process. While cracks were present, the coating remained adhered and precisely replicated the highly structured surface. The coating thickness was visually estimated to be in the range of 350 to 400 nm. Despite perfectly replicating the substrate’s complex surface, the coating containing Ag and HA exhibited an irregular distribution of nano-sized silver (Ag) and micro-sized hydroxyapatite (HA) particles (Fig. 6). The structured surface limited the ability to achieve a homogeneous particle distribution. Cracks were observed in the coating (TAN04HA); however, they did not negatively impact its quality or properties. The coating demonstrated excellent adhesion to the substrate. Functionally, the TAN06HA coating was comparable to the TAN04HA coating, differing only in its higher Ag concentration. Although local cracks were present, particularly around the HA beads and particles, the coating accurately replicated the grid’s structured surface without dimensional changes (Fig. 7). An uneven distribution of silver nanoparticles was observed. The coating exhibited high adhesion to the substrate. Cross-sectional analysis was performed to examine the microstructure of dodethick (Fig. 8). To assess whether the viscosity of the sols (T, TAN04HA, and TAN06HA) facilitated their penetration through the scaffold lattice to the plate during coating under constant stirring, the coated substrates were sectioned and visually characterized post-firing. Evaluation of both optical microscopy (OM) images and scanning electron microscopy (SEM) details revealed that the coatings precisely replicated the intricate inner surface of the lattice, thus confirming the penetration of all sols to the plate. Hydroxyapatite (HA) particles and silver (Ag) particles were observed within the TAN04HA and TAN06HA coatings inside the coated scaffolds (Fig. 8b, 8c). Interestingly, the coatings showed little evidence of cracking within the substrates. Nevertheless, the concentration and distribution of the incorporated elements (micro-HA, nano-Ag) within the lattice cavities require further investigation. The microstructure of the sol-gel-coated titanium and Ti6Al4V alloy samples exhibited distinct characteristics depending on the substrate geometry and coating composition. Un [35] found through extensive experimental research that the composition of the sol, dip-coating conditions and thermal treatment can significantly affect the properties of sol-gel coatings. The stiffness discrepancy (better bonding) between the implant and the bone tissue can be addressed by making porous Fig 3. (SEM) Microstructure of the machined wrought rod after firing: a) T coating; b) TAN04HA coating; c) TAN06HA coating. A. Knaislov´ a et al. Applied Surface Science Advances 29 (2025) 100816 4 materials. In addition, the generated pores could improve bone ingrowth towards the interior of the implant (with an optimal pore size) [36–38]. The machined wrought Ti6Al4V alloy rod displayed a relatively uniform and smooth surface after mechanical grinding and ultrasonic cleaning. Following sol-gel coating and thermal treatment, the coatings demonstrated good adhesion with minimal cracking, ensuring homogeneous coverage. Hydroxyapatite (HA) and silver (Ag) particles were evenly distributed across the surface, and no significant delamination was observed. Sol penetration was optimal, resulting in a stable and adherent coating. In contrast, the 3D-printed gyroid structures presented greater challenges in achieving homogeneous coatings due to their complex geometry [39]. Although the coatings adhered well to the substrate, significant cracking was observed, likely due to uneven sol flow over the intricate surface during dip-coating and subsequent thermal expansion stresses. However, the coatings maintained good adhesion, and their bioactive and antibacterial functionalities were retained. The silver nanoparticle distribution was less uniform compared to the rod samples, which could influence the localized antibacterial performance. The gyroid structures, with their highly porous architecture, provided an increased surface area, potentially enhancing interactions with biological environments [40]. The 3D-printed dodethick structures presented an even more intricate surface morphology, composed of overlapping grid layers and embedded beads. The coatings on these structures exhibited localized cracking, particularly around the beads, but maintained excellent adhesion. The sol effectively penetrated through the scaffold to the underlying plate, indicating that the viscosity was optimized for such geometries. The micro-HA and nano-Ag additives were distributed throughout the surface, although achieving homogeneity was Fig 4. (SEM) Microstructure of the Ti6Al4V gyroid 3D after firing: a) T coating; b) TAN04HA coating; c) TAN06HA coating. Fig 5. (SEM) Microstructure of the Ti dodethick 3D with T coating after firing. A. Knaislov´ a et al. Applied Surface Science Advances 29 (2025) 100816 5 challenging. Cross-sectional analysis confirmed that the coatings successfully penetrated the lattice, effectively covering the complex internal surfaces. Interestingly, fewer cracks were observed within the inner regions of the dodethick scaffolds, suggesting that surface curvature and confined geometries influenced stress distribution during thermal treatment. The study demonstrated that sol-gel coatings effectively conformed to different titanium-based substrates, but substrate geometry significantly impacted coating homogeneity and integrity. While smooth surfaces such as the machined rod facilitated uniform coatings with minimal defects, complex 3D-printed structures, especially gyroid and dodethick designs, required careful optimization of sol viscosity and deposition techniques to minimize cracking. Future work should focus on refining coating application methods, potentially incorporating multi-step layering or modified drying protocols to enhance coating uniformity and mechanical stability across highly structured surfaces. The phase composition of the TAN04HA, TAN06HA coatings on rod, gyroid and dodethick after firing is showed on Fig. 9. The diffractograms of the coatings after firing need to be viewed from two angles. First, from the material point of view, it is obvious that the intensity of the peak for the basic titanium phase is greatly influenced by the shape of the substrate. The peaks are most intense for the rod, then for the gyroid, and the smallest peak belongs to the dodethick. Second, from the composition of the coatings point of view, it is obvious that the phase for commercial hydroxyapatite (HA) was detected mainly for the rod and gyroid. But the intensity is very small compared to the intensity for titanium. It was again confirmed that the shape of the substrate also influences the detection of commercial hydroxyapatite and silver. In both types of coatings, neither hydroxyapatite nor silver was detected on the dodethick. Moreover, it is obvious from the SEM images that the particles in both types of coatings cover the entire surface of the structured dodethick substrate. 3.2. Properties of the prepared samples 3.2.1. Adhesion test The coatings on the alloy samples of the machined wrought rod showed very good adhesion, evaluated according to the classification table as grade 5B (possibly even 4B). After visual characterization, it was found that particles of commercial hydroxyapatite and silver (for both concentrations) remained fixed in the coatings after the tape was peeled off (Fig. 10). Even some of the adhesive from the tape remained on the sample after it was torn off. Coatings on the machined wrought Ti6Al4V alloy rod demonstrated outstanding adhesion, with minimal particle detachment observed in the tape test. This strong adhesion can be attributed to the rod’s relatively smooth surface, which facilitated uniform coating deposition and strong interfacial bonding during thermal treatment. Milella [41] prepared titanium sol-gel coatings containing HA and found cracks on their surface after firing. The cracks do not influence the mechanical and adhesive properties (by tensile test) of the coating. In contrast, the 3D-printed gyroid and dodethick structures maintained good adhesion, although some surface irregularities and cracks were present. The coatings adhered well even to the highly porous and intricate lattice structures, indicating their potential for biomedical applications. However, the gyroid and dodethick samples exhibited more pronounced cracking, likely due to internal stresses introduced during drying and firing by their complex geometries. Fig 6. (SEM) Microstructure of the Ti dodethick 3D with TAN04HA coating after firing. Fig 7. (SEM) Microstructure of the Ti dodethick 3D with TAN06HA coating after firing. A. Knaislov´ a et al. Applied Surface Science Advances 29 (2025) 100816 6 3.2.2. Bioactivity test The alloy coatings containing commercial hydroxyapatite (TAN04HA, TAN06HA) showed excellent bioactive properties, with a visually detected thick layer of precipitated bone hydroxyapatite (HAp) after the in vitro test (Fig. 11b, 11c). The hydroxyapatite globules, composed of plate-like nanocrystals, grew and clustered with increasing exposure time in SBF. The layer cracked as a result of drying (during preparation for SEM measurements), which allowed for a rough estimation of the layer thickness in the micrometre range. It is important to consider that, in vivo, the layer would not be air-exposed, suggesting that Fig. 8. (SEM) Microstructure of Ti - dodethick (3D printing; grid: 1.35) – cross section: a) T coating after firing; b) TAN04HA coating after firing; c) TAN06HA coating after firing. Fig. 9. XRD patterns of the TAN04HA, TAN06HA coatings on rod, gyroid and dodethick after firing. A. Knaislov´ a et al. Applied Surface Science Advances 29 (2025) 100816 7 the cracking was induced by laboratory measurements. In comparison, the T coating (without HA and Ag additives) exhibited very low bioactivity, indicated by the precipitation of isolated, very small HAp globules on the coating surface (Fig. 11a). The coatings on the 3D gyroid behaved consistently with the coatings on the machined wrought rod (Fig. 12). Across all coating types, surface images showed that the layer of precipitated hydroxyapatite faithfully replicated the fired coating, even with cracking present. The nanoplatelike globules presented a consistent structure, with sizes in the micrometer range. The outer and inner spaces of the gyroid demonstrated uniform behavior throughout the test, with no localized changes in behavior detected, despite the sample’s complex geometry. Dodethick with coating T (base coating without additives) showed no changes after 21 days of exposure in simulated body fluid (SBF) solution (Fig. 13). The coating’s surface displayed no changes in morphology or composition, and no phases or precipitates were detected. This coating did not demonstrate bioactive properties following in vitro testing. The TAN04HA coating on dodethick, containing lower silver (Ag) and commercial hydroxyapatite (HA) content, showed excellent bioactive properties (Fig. 14). At the conclusion of the 21-day in vitro test, a visually detected layer of precipitated hydroxyapatite (HAp) was observed, composed of globules of platelet-shaped nanocrystals. This layer experienced cracking due to drying, a procedure necessary for SEM measurements. The TAN06HA coating on dodethick, containing a higher concentration of silver (Ag) and the same concentration of commercial hydroxyapatite (HA) as the TAN04HA coating, showed similar behavior to the TAN04HA coating (Fig. 15). After 21 days of in vitro testing in simulated body fluid (SBF), the coated dodethick scaffold surface was almost completely covered with a layer of precipitated hydroxyapatite (HAp). This precipitated phase was characterized by globules composed Fig. 10. (SEM) The adhesion of the coatings of the machined wrought rod after firing: a) TAN04HA; b) TAN06HA. Fig. 11. (SEM) Microstructure of the machined wrought rod after in vitro test in SBF: a) T coating; b) TAN04HA coating; c) TAN06HA coating. A. Knaislov´ a et al. Applied Surface Science Advances 29 (2025) 100816 8 of platelet-like nanocrystals. Detailed SEM analysis revealed that the precipitated HAp platelets grew directly from the commercial HA particles within the TAN06HA coating. As the in vitro testing time increased, the HAp globules (with sizes in the micrometer range) grew and interconnected, forming a continuous layer. This layer completely covered the original coating, filled firing cracks, and filled voids in the scaffold lattice. At low magnification (x30), SEM images showed that the precipitated hydroxyapatite layer perfectly replicated the morphology of the coated scaffold. The HAp nanoplatelet globules exhibited the same structure for both types of coatings (TAN04HA, TAN06HA). The phase composition of the TAN04HA, TAN06HA coatings on rod, gyroid and dodethick after in vitro test is showed on Fig. 16. The analysis demonstrated the presence of precipitated hydroxyapatite (HAp) phase from SBF on the surface of both types of coatings on all types of substrates. Again, the influence of the substrate shape on both the intensity and shape of the HAp peaks and the detected amount was evident. From the diffractograms, it is obvious at first glance that the most HAp is on the TAN04HA and TAN06HA coatings on the gyroid (55 - 65 %) and on the rod (15 - 20 %). XRD analysis of the coatings on dodethick detected only trace amounts of the HAp phase, although the SEM images clearly show the entire coverage of this shaped substrate with the new precipitated phase. Although XRD analysis cannot distinguish commercial hydroxyapatite from precipitated hydroxyapatite (Ca 10 (PO 4 ) 6 (OH) 2 applies to both), by comparing the peak intensities, their shapes and the amount ( %) of the phase of both types of coatings TAN04HA and TAN06HA after firing and after in vitro testing, the difference between the hydroxyapatites is obvious at first glance. Using AAS, the concentration of Ca 2+ was determined, and UV–VIS was used to measure the concentration of (PO 4 ) 3ions in SBF solutions before and after the 21-day in vitro test. For biomaterial predictive testing, monitoring changes in the concentrations and pH of the SBF Fig. 12. (SEM) Microstructure of the gyroid after in vitro test in SBF: a) T coating; b) TAN04HA coating; c) TAN06HA coating. Fig. 13. (SEM) Microstructure of Ti - dodethick (3D printing; grid: 1.35) with T coating after in vitro test in SBF: a) area; b) detail. A. Knaislov´ a et al. Applied Surface Science Advances 29 (2025) 100816 9