Low-Temperature Hydrothermal Treatment Surface Functionalization of the Ultrafine-Grained TiMo Alloys for Medical Applications

Hydroxyapatite (HAp) is the most widely used material for bio coating. The functional layer can be produced by many methods, however, the most perspective by its utility, easy to scale up, and simplicity aspects remains a hydrothermal treatment approach. In this work, an HAp coating was produced by low-temperature hydrothermal treatment on the ultrafine-grain beta Ti-xMo (x = 23, 27, 35 wt.%) alloys. The proposed surface treatment procedure combines acid etching, alkaline treatment (AT), and finally hydrothermal treatment (HT). The uniqueness of the approach relies on the recognition of the influence of the molar concentration of NaOH (5 M, 7 M, 10 M, 12 M) during the alkaline treatment on the growth of hydroxyapatite crystals. Obtained and modified specimens were examined structurally and microstructurally at every stage of the process. The results show that the layer after AT consist of titanium oxide and phases based on sodium with various phase relations dependent on NaOH concentration and base composition. The AT in 7 M and 10 M enables to obtain the HAp layer, which can be characterized as the most developed in terms of thickness and porosity. Finally, selected coated samples were investigated in terms of surface wettability test managed in time relation, which for the results confirm high hydrophilicity of the surfaces. Conducted research shows that the low-temperature hydrothermal processing could be considered for a possible adaptation in the drug encapsulation and delivery systems.


Introduction
Titanium and its alloys are the most commonly used metallic materials in the biomedical field. These biometals have found application in orthopedics, not only maxillofacial but also, dental implants under load-bearing conditions [1]. Their use is justified by low Young's modulus, low density, and high corrosion resistant properties [1,2].
Contemporary researchers have demonstrated that the properties of Ti-Mo alloys can be adjusted depending on the Mo content. As the Mo concentration increases, the hardness increases, and the rod-or plate-like structures appear under acidic or neutral conditions. Furthermore, hydrothermal treatment can differ in the chemical composition of the solution. Calcium chelating agents and various organic surfactants are the most common HAp sources for hydrothermal treatment. Although, EDTA, as the most used calcium chelating agent, found the application in HAp production. Ca-EDTA is simple to decompose under hydrothermal conditions. This modifier enables the formation of longer HAp crystals at a lower temperature.
The present study investigates the formation of the hydroxyapatite layers on the TixMo (x = 23, 27, 35 wt.%) ultrafine-grained alloys obtained by mechanical alloying and hot pressing approach. The HAp covers were obtained after the proposed low-temperature procedures. Surface treatments rely on acidic etching, alkaline, and finally hydrothermal treatment. All process stadiums were investigated structurally and was depictured on the schematic process illustration in Figure 1. temperature the results are better. Regarding the relation between pH and HAp crystals morphology, the high pH value facilitates the growth of a spherical or short rod-like form of HAp [16]. However, rod-or plate-like structures appear under acidic or neutral conditions. Furthermore, hydrothermal treatment can differ in the chemical composition of the solution. Calcium chelating agents and various organic surfactants are the most common HAp sources for hydrothermal treatment. Although, EDTA, as the most used calcium chelating agent, found the application in HAp production. Ca-EDTA is simple to decompose under hydrothermal conditions. This modifier enables the formation of longer HAp crystals at a lower temperature. The present study investigates the formation of the hydroxyapatite layers on the TixMo (x = 23, 27, 35 wt.%) ultrafine-grained alloys obtained by mechanical alloying and hot pressing approach. The HAp covers were obtained after the proposed low-temperature procedures. Surface treatments rely on acidic etching, alkaline, and finally hydrothermal treatment. All process stadiums were investigated structurally and was depictured on the schematic process illustration in Figure 1. A low-temperature surface treatment processing applied for sintered ultrafine-grained TixMo alloys remain attractive for further recognition in drug encapsulation and delivery systems. Processing temperatures and applied for treatment solutions for the above stay relevant factors.
The starting sample compositions and in the same resulting phase relation with its inherent structure and size, influence as the research shows a step-by-step processing relation. The phase grain boundaries and the primary powder particle grains that could be noticed for the sinters obtained in a specific way (revealed after etching), by its energetically favorable state influence further processing stadiums that mimic the substrate structure. Analyzed in this work, various molar concentrations of NaOH (5M, 7M, 10M, 12M) alkaline solutions shows its influence on the treatment products, necessary however for further proper HAp layer formation and growth. The above points out also the dependence between the substrate activity characteristics and the applied solution concentration. Analyzed additional wettability characteristics of the obtained HAp layers correlate with not modified surfaces and base microcrystalline titanium, confirming its meaning and usability in a broad medical applications range.

Materials and Specimens
For the hydrothermal treatment surface functionalization, an ultrafine-grain base Ti23Mo, Ti27Mo, and Ti35Mo (wt.%) alloy samples were fabricated by mechanical alloying (MA) and powder metallurgy (PM) methods, which for the details of preparation could be found in our earlier work [5,7]. The sintering procedure proposed in this work was the same for all samples and relay on the A low-temperature surface treatment processing applied for sintered ultrafine-grained TixMo alloys remain attractive for further recognition in drug encapsulation and delivery systems. Processing temperatures and applied for treatment solutions for the above stay relevant factors.
The starting sample compositions and in the same resulting phase relation with its inherent structure and size, influence as the research shows a step-by-step processing relation. The phase grain boundaries and the primary powder particle grains that could be noticed for the sinters obtained in a specific way (revealed after etching), by its energetically favorable state influence further processing stadiums that mimic the substrate structure. Analyzed in this work, various molar concentrations of NaOH (5 M, 7 M, 10 M, 12 M) alkaline solutions shows its influence on the treatment products, necessary however for further proper HAp layer formation and growth. The above points out also the dependence between the substrate activity characteristics and the applied solution concentration. Analyzed additional wettability characteristics of the obtained HAp layers correlate with not modified surfaces and base microcrystalline titanium, confirming its meaning and usability in a broad medical applications range.

Materials and Specimens
For the hydrothermal treatment surface functionalization, an ultrafine-grain base Ti23Mo, Ti27Mo, and Ti35Mo (wt.%) alloy samples were fabricated by mechanical alloying (MA) and powder metallurgy (PM) methods, which for the details of preparation could be found in our earlier work [5,7]. The sintering procedure proposed in this work was the same for all samples and relay on the hot pressing in 800 • C in vacuum conditions which was schematically shown in Figure 2. To prepare the (β) titanium alloys, the commercial Ti (Alfa Aesar, 99.9% purity, CAS:7440-32-6, Karlsruhe, Germany) and Mo (44 µm, 99.6%, CAS:7430-08-7 Sigma Aldrich, Karlsruhe, Germany) powders were used. The dimensions of the hot-pressed cylinder shape bulk samples were 5 mm high and 8 mm in diameter. In the next step, samples were mechanically abraded with 180 to 2500-grit silicon carbide (SiC) grinding papers followed by polishing with 0.5 µm diamond paste and cleaning in ethanol. Prepared samples were chemically etched in an acid solution of 3% HF + 10% H 3 PO 4 (Poch S.A., Gliwice, Poland) for 3 min and afterward washed in distilled water. hot pressing in 800 °C in vacuum conditions which was schematically shown in Figure 2. To prepare the (β) titanium alloys, the commercial Ti (Alfa Aesar, 99.9% purity, CAS:7440-32-6, Karlsruhe, Germany) and Mo (44 μm, 99.6%, CAS:7430-08-7 Sigma Aldrich, Karlsruhe, Germany) powders were used. The dimensions of the hot-pressed cylinder shape bulk samples were 5 mm high and 8 mm in diameter. In the next step, samples were mechanically abraded with 180 to 2500-grit silicon carbide (SiC) grinding papers followed by polishing with 0.5 μm diamond paste and cleaning in ethanol. Prepared samples were chemically etched in an acid solution of 3% HF + 10% H3PO4 (Poch S.A., Gliwice, Poland) for 3 min and afterward washed in distilled water.

Alkaline Treatment
Alkaline treatment (AT) of TixMo (x = 23, 27, 35) alloys were performed in four different molar concentrations of NaOH (Poch S.A., Gliwice, Poland). Each alloy set of Ti23Mo, Ti27Mo, and Ti35Mo samples was treated in 5 M, 7 M, 10 M, and 12 M NaOH solution. The samples were anchored to the bottom of the glass beakers and filled with the 250 mL of NaOH solution. Subsequently, the beakers were put into the furnace and treated for 24 h at 90 °C conditions. At the end of the alkaline treatments, samples were washed in distilled water. After AT, the samples were marked as Table 1 shows.

Hydrothermal Treatment
Hydroxyapatite coatings were synthesized from CA-EDTA solution under hydrothermal conditions. The preparation of the solution was divided into two stages. Firstly, Ca-EDTA (Sigma-Aldrich, CAS: 62-33-9, Saint Louis, MO, USA) and KH2PO4 (Sigma-Aldrich, CAS: 7778-77-0, Saint Louis, MO, USA) were dissolved in deionized water with 0.25 mol/L and 0.15 mol/L, respectively, to prepare 200 mL of solution. Secondly, the pH of the solution was titrated by 1 mol/L NaOH to the value of 8.9. The fresh resulting mixture was transferred into glass beakers with placed earlier prepared AT samples, that were next sealed and put into the furnace. The hydrothermal treatment of the samples was conducted at 130 °C for 6 h (time was measured from the time when the boiling point was reached) with the next cooling to room temperature with the furnace. The sample was removed from the beakers and dried at room temperature. After hydrothermal treatment (HT), the samples were marked as presented in Table 2.

Alkaline Treatment
Alkaline treatment (AT) of TixMo (x = 23, 27, 35) alloys were performed in four different molar concentrations of NaOH (Poch S.A., Gliwice, Poland). Each alloy set of Ti23Mo, Ti27Mo, and Ti35Mo samples was treated in 5 M, 7 M, 10 M, and 12 M NaOH solution. The samples were anchored to the bottom of the glass beakers and filled with the 250 mL of NaOH solution. Subsequently, the beakers were put into the furnace and treated for 24 h at 90 • C conditions. At the end of the alkaline treatments, samples were washed in distilled water. After AT, the samples were marked as Table 1 shows.

Hydrothermal Treatment
Hydroxyapatite coatings were synthesized from CA-EDTA solution under hydrothermal conditions. The preparation of the solution was divided into two stages. Firstly, Ca-EDTA (Sigma-Aldrich, CAS: 62-33-9, Saint Louis, MO, USA) and KH 2 PO 4 (Sigma-Aldrich, CAS: 7778-77-0, Saint Louis, MO, USA) were dissolved in deionized water with 0.25 mol/L and 0.15 mol/L, respectively, to prepare 200 mL of solution. Secondly, the pH of the solution was titrated by 1 mol/L NaOH to the value of 8.9. The fresh resulting mixture was transferred into glass beakers with placed earlier prepared AT samples, that were next sealed and put into the furnace. The hydrothermal treatment of the samples was conducted at 130 • C for 6 h (time was measured from the time when the boiling point was reached) with the next cooling to room temperature with the furnace. The sample was removed from the beakers and dried at room temperature. After hydrothermal treatment (HT), the samples were marked as presented in Table 2.

Materials Characterization
At every stage of sample preparation, an X-ray diffraction (XRD) (Empyrean, Panalytical, Almelo, Netherlands) analysis was conducted by the Panalytical Empyrean equipment with the copper anode (CuKα-1.54 Å) at a Brag-Brentano reflection mode configuration with 45 kV and 40 mA parameters. The measurement parameters were set up for 20 • -90 • with a 15 s per step 0.0167 • in all cases.
The following structural models were used: (1) Base specimens: (2) Alkali treatment: • For detailed structure analysis, a collation of the pairs of the highest intensity peak relations of processing product with the substrate was revealed. The data refers to obtained patterns of the processing products depicted in a graph relation, which for a proportion of the highest intensity peak of the substrate were related to the highest intensity peak of the product phases. Indirectly presented data represent a quantitative collation, which for diminished base substrate intensity with increasing intensity of products corresponds to a layer thickness growth. Additionally, the growing intensity of the product phases corresponds to its higher crystallinity and volumetric amount related to the collected instrument data. Further conclusions could also be drawn from the presented data.
Scanning electron microscopy in the SE-mode (SEM, MIRA3 Tescan) was used to characterize the surface after all stages of modification.
The contact angles (CA) of analyzed surfaces were determined by the sessile drop method. The measurements were performed by a Drop Shape Analyzer-DSA25 instrumentation (KRÜSS GmbH, Hamburg, Germany) and the KRÜSS ADVANCE 1.5.1.0 software instrumentation (KRÜSS GmbH, Hamburg, Germany). The used parameters were as follows: •

Results and Discussion
In the present study, the hydroxyapatite cover on the ultrafine-grained beta titanium alloys was obtained and afterward evaluated focusing on its formation criteria. Using the various molar concentrations of NaOH (5 M, 7 M, 10 M, 12 M) the influence of the alkali treatment on HAp layer growth were investigated, while the rest of the parameters were constant. Furthermore, the chemical composition impact of beta titanium ultrafine-grained alloys on the surface modification approach was also investigated. Figure 3 shows the XRD patterns of Ti-xMo bulk sample ultrafine-grained alloys. In all of the analyzed cases, the main phase corresponds to Ti(β) type structure-Ti 0.9 Mo 0.1 . However, for all systems, a small amount of the Ti(α) phase could be noticed. The amount of Ti(α) phase remains strictly connected with the beta stabilizing Mo content in the starting powder mixtures. Additionally, an increase in molybdenum content causes also a decrease in the beta phase unit cell, which is revealed by a peak shift to larger angles. The reductions in the interplanar distance (∆d hkl ) between references phase and MA alloys set as follows 0.002513, 0.004263, and 0.014648 by increasing molybdenum content in the alloy (Figure 3). The above relation could be only explained by the bcc unit cell transition from titanium-based one to molybdenum which for the cell dimension is lower. In the specific compared relation of Ti 0.9 Mo 0.1 to Mo lattice parameters dimension present as fallowed 325.7 to 314.7 pm. A similar effect was already observed in our earlier research [5] which details structural estimation confirming such behavior; however, other intermediate beta phases with growing molybdenum content were confirmed like MoTi.

Results and Discussion
In the present study, the hydroxyapatite cover on the ultrafine-grained beta titanium alloys was obtained and afterward evaluated focusing on its formation criteria. Using the various molar concentrations of NaOH (5M, 7M, 10M, 12M) the influence of the alkali treatment on HAp layer growth were investigated, while the rest of the parameters were constant. Furthermore, the chemical composition impact of beta titanium ultrafine-grained alloys on the surface modification approach was also investigated. Figure 3 shows the XRD patterns of Ti-xMo bulk sample ultrafine-grained alloys. In all of the analyzed cases, the main phase corresponds to Ti(β) type structure-Ti0.9Mo0.1. However, for all systems, a small amount of the Ti(α) phase could be noticed. The amount of Ti(α) phase remains strictly connected with the beta stabilizing Mo content in the starting powder mixtures. Additionally, an increase in molybdenum content causes also a decrease in the beta phase unit cell, which is revealed by a peak shift to larger angles. The reductions in the interplanar distance (Δdhkl) between references phase and MA alloys set as follows 0.002513, 0.004263, and 0.014648 by increasing molybdenum content in the alloy (Figure 3). The above relation could be only explained by the bcc unit cell transition from titanium-based one to molybdenum which for the cell dimension is lower. In the specific compared relation of Ti0.9Mo0.1 to Mo lattice parameters dimension present as fallowed 325.7 to 314.7 pm. A similar effect was already observed in our earlier research [5] which details structural estimation confirming such behavior; however, other intermediate beta phases with growing molybdenum content were confirmed like MoTi.  A hot pressing of the prepared by MA blends allows obtaining alloys with an ultra-fine grain structure for all proposed compositions, as shown in Figure 4. The average grain size of Ti23Mo oscillates 1.097 ± 0.216 (Figure 4). The microstructure remains homogenous and shows a polygonal grain morphology suggesting fine kinetics between the components and stable state of the obtained systems. A hot pressing of the prepared by MA blends allows obtaining alloys with an ultra-fine grain structure for all proposed compositions, as shown in Figure 4. The average grain size of Ti23Mo oscillates 1.097 ± 0.216 (Figure 4). The microstructure remains homogenous and shows a polygonal grain morphology suggesting fine kinetics between the components and stable state of the obtained systems.

Acid Etching
The chemical etching step was applied in the proposed procedure to develop a surface roughness. Figure 5 shows the SEM microphotographs of Ti23Mo, Ti27Mo, and Ti35Mo surface after etching in 3% HF + 10% H3PO4 acid solution. The etching process runs firstly along the primary powder particle grains, some of the effects could be also observed for the phase grain boundaries. In the case of acid etching of Ti27Mo, the obtained effect remains inhomogeneous. Etching takes place on the entire surface along the powder particle grains, but the grain boundaries etching can be indicated in some areas. Ti35Mo alloy is the most resistant to acidic etching in the proposed process solution. The primary powder particles cannot be pointed out. The etching process runs along the grain boundaries, but the final effect remains fine.

Acid Etching
The chemical etching step was applied in the proposed procedure to develop a surface roughness. Figure 5 shows the SEM microphotographs of Ti23Mo, Ti27Mo, and Ti35Mo surface after etching in 3% HF + 10% H 3 PO 4 acid solution. The etching process runs firstly along the primary powder particle grains, some of the effects could be also observed for the phase grain boundaries. In the case of acid etching of Ti27Mo, the obtained effect remains inhomogeneous. Etching takes place on the entire surface along the powder particle grains, but the grain boundaries etching can be indicated in some areas. Ti35Mo alloy is the most resistant to acidic etching in the proposed process solution. The primary powder particles cannot be pointed out. The etching process runs along the grain boundaries, but the final effect remains fine.

Alkali Treatment in NaOH Solution with Different Molar Concentration
Figures 6-8 shows the XRD patterns of the alkali-treated surface of the TixMo alloys. The diffraction patterns point out mostly Ti(β) and Ti(α) phases derived from the substrate. However, the titanium oxide and phases base on sodium (Na2Ti2(OH)4 and Na0.55Mo2O4) can be indicated. The peaks from the layer phases are characterized by low intensity. This indicates a slight thickness of the created layers. Ti35Mo sample possesses the highest resistance to alkaline treatment. XRD patterns of Ti35Mo do not reveal Na2Ti2(OH)4 phase regardless of the NaOH concentration. The largest quantitative share of the Na2Ti2(OH)4 phase can be observed for the samples 23-7AT and 27-7AT.
The surface morphology analysis after alkaline treatment with the various NaOH solution for 24 h at 90 °C, was shown in Figure 9. The results indicated that the surface morphology depends on the molar concentration of NaOH solution and, also, the chemical composition of the base materials. In all cases, the created layers are characterized by surface cracking. Formed structure imitates the primary powder and phase grains because of the privileged higher energy regions with easier kinetics. For the abovementioned, it is worth noticing that the reaction kinetics indirectly depend on a substrate relation and solution concentration, which may lead to an unexpected product phase stability diminishment, which, in our case, is visible in an intense layer cracking and delamination from the substrate. The depth of cracks and the size of formed structures have the smallest value in the case of Ti35Mo alloy. Additionally, NaOH-soaked samples (at all concentrations) exhibit a porous cellular-looking morphology. Only the sample Ti35Mo 10 M possesses a unique microstructure morphology, which can be determined as an island-like morphology.      The surface morphology analysis after alkaline treatment with the various NaOH solution for 24 h at 90 • C, was shown in Figure 9. The results indicated that the surface morphology depends on the molar concentration of NaOH solution and, also, the chemical composition of the base materials. In all cases, the created layers are characterized by surface cracking. Formed structure imitates the primary powder and phase grains because of the privileged higher energy regions with easier kinetics. For the abovementioned, it is worth noticing that the reaction kinetics indirectly depend on a substrate relation and solution concentration, which may lead to an unexpected product phase stability diminishment, which, in our case, is visible in an intense layer cracking and delamination from the substrate. The depth of cracks and the size of formed structures have the smallest value in the case of Ti35Mo alloy.

Hydrothermal Treatment in Ca-EDTA Solution
The XRD patterns of hydrothermally treated TixMo alloys confirm the presence of HAp on the sample surfaces, as shown in Figures 10-12. The XRD patterns obtained for the samples after AT in 7 M, 10 M, and HT are characterized by sharp and high-intensity HAp peaks. Additionally, no peaks Additionally, NaOH-soaked samples (at all concentrations) exhibit a porous cellular-looking morphology. Only the sample Ti35Mo 10 M possesses a unique microstructure morphology, which can be determined as an island-like morphology.

Hydrothermal Treatment in Ca-EDTA Solution
The XRD patterns of hydrothermally treated TixMo alloys confirm the presence of HAp on the sample surfaces, as shown in Figures 10-12. The XRD patterns obtained for the samples after AT in 7 M, 10 M, and HT are characterized by sharp and high-intensity HAp peaks. Additionally, no peaks correspond to any other crystalline calcium phosphate phases. Disodium phosphate (Na 2 HPO 4 ) impurities could be indicated in the layer, which derives from the Ca-EDTA solution. This phase forms a binder between the hydroxyapatite crystals. The patterns obtained for the samples after AT in 12 M and HT are characterized by a low intensity of the peaks originating from the HAp phase.        The overall morphology of the obtained after HT process surfaces, indicates the existence of many uniform cavernous-like HAp microstructure with needle-like HAp particles, as is presented in Figure 13. However, preceding alkaline treatment has a significant influence on HAp formation. SEM microphotographs of 23-7HT and 35-7HT show HAp covers on Ti(β) alloys, which are characterized by high surface development and open porosity. HAp layers formed on samples after 10 M AT are consistent and possess pores with a larger dimension than samples after 7 M AT. It is relevant for the biomedical application when surfaces that contact directly with a tissue possess the possibilities to be colonized and overgrow by bone tissue. As the molar concentration of NaOH increased from 10 M to 12 M, the HAp phase does not cover the entire surface and exhibits an island-like morphology. The abovementioned conclusion indicates low stability of the AT products in the case of 12 M NaOH solution, which does not provide a proper anchor for the HT product phase.
Structural collated data in the graph (Figure 14), represents the ratio of the intensity between the peak of the post-products phases and the Ti(β) phase, which shows the differences in the reaction after AT and the next HT. It is worth noticing that the peak ratio relates to the highest intensity peaks for each phase that corresponds to the following family of the plane: (111) for TiO-cubic, (020) for Na 2 Ti 2 O 4 (OH) 2 -orthorhombic, (−102) for Na 0.55 Mo 2 O 4 -monoclinic, (121) for HAp-hexagonal, and also (110) for Ti(β) phase. The abovementioned data confirm the earlier conclusion about the processing products, that mimic the substrate, which was also confronted with the SEM analysis discussion.
Regarding the AT product ratio presents a complex overview. Although, it does not reveal the relation between solution concentration or substrate composition and surface response. Locally, for the same substrate composition, some tendency could be observed. For Ti23Mo example, the TiO and Na 2 Ti 2 O 4 (OH) 2 product phase amount grows when the solution concentration increases from 5 M to 7 M. However, further 10 M and 12 M treatment exposes a decrease or even absence in the case of TiO product phase. A familiar relation could be observed in the case of the Ti27Mo sample, although only for the Na 2 Ti 2 O 4 (OH) 2 product phase. Structural collated data in the graph (Figure 14), represents the ratio of the intensity between the peak of the post-products phases and the Ti(β) phase, which shows the differences in the reaction after AT and the next HT. It is worth noticing that the peak ratio relates to the highest intensity peaks for each phase that corresponds to the following family of the plane: (111) for TiO-cubic, (020) for Na2Ti2O4(OH)2-orthorhombic, (−102) for Na0.55Mo2O4-monoclinic, (121) for HAp-hexagonal, and also (110) for Ti(β) phase. The abovementioned data confirm the earlier conclusion about the processing products, that mimic the substrate, which was also confronted with the SEM analysis discussion. The represented data confirms the HAp structural response after the HT in all analyzed experimental examples. However, in some cases, in a very low degree. In the case of the HT products, it stays important that the related value of I HAp(121) /I β(110) ratio depends indirectly on earlier AT treatment. Although, AT products were not noticed in the patterns after HT even when an HAp intensity was very low. The reason for this behavior is probably connected with the low stability of products after AT. The obtained results were assessed in terms of structural relations, understood as the highest HAp intensity ratio. This criterion indicates the samples after 10 M AT. Regarding the AT product ratio presents a complex overview. Although, it does not reveal the relation between solution concentration or substrate composition and surface response. Locally, for the same substrate composition, some tendency could be observed. For Ti23Mo example, the TiO and Na2Ti2O4(OH)2 product phase amount grows when the solution concentration increases from 5 M to 7M. However, further 10 M and 12 M treatment exposes a decrease or even absence in the case of TiO product phase. A familiar relation could be observed in the case of the Ti27Mo sample, although only for the Na2Ti2O4(OH)2 product phase.
The represented data confirms the HAp structural response after the HT in all analyzed experimental examples. However, in some cases, in a very low degree. In the case of the HT products, it stays important that the related value of IHAp(121)/Iβ(110) ratio depends indirectly on earlier AT treatment. Although, AT products were not noticed in the patterns after HT even when an HAp intensity was very low. The reason for this behavior is probably connected with the low stability of products after AT. The obtained results were assessed in terms of structural relations, understood as the highest HAp intensity ratio. This criterion indicates the samples after 10 M AT.

Wettability Tests
The properties of the surface, including its hydrophilicity or hydrophobicity, have a significant influence on the interaction between the surface of the biomaterial and the biological environment, in particular, on the number of adsorbed proteins and their adhesion to the surface. In biomedical applications, a desirable situation is defined by a material that exhibits hydrophilic surface properties. The abovementioned properties have a positive effect on the adsorption, adhesion, and cell proliferation activity, which remain important in osseointegration characteristics improvement.
Selected HAp layers on TixMo alloys have been characterized in terms of wettability by glycerol testing fluid presented in Figure 15. All modified sample surfaces possess high hydrophilicity. The contact angle (CA) in time when the drop was placed (t = 0) was 30°, 26°, and 31.5° for 23-10HT, 27-10HT, and 35-10HT, respectively. In the case of 23-10HT after 4.5 s the CA decreased to approximately 7.5°, the same effect was achieved after 6.5 s for 27-10HT and 35-10HT. Additionally, the unmodified surfaces of Ti-xMo alloys and microcrystalline titanium were analyzed for comparison purposes. The HT confirms a visible CA decrease for selected covered samples in comparison to untreated samples.

Wettability Tests
The properties of the surface, including its hydrophilicity or hydrophobicity, have a significant influence on the interaction between the surface of the biomaterial and the biological environment, in particular, on the number of adsorbed proteins and their adhesion to the surface. In biomedical applications, a desirable situation is defined by a material that exhibits hydrophilic surface properties. The abovementioned properties have a positive effect on the adsorption, adhesion, and cell proliferation activity, which remain important in osseointegration characteristics improvement.
Selected HAp layers on TixMo alloys have been characterized in terms of wettability by glycerol testing fluid presented in Figure 15. All modified sample surfaces possess high hydrophilicity. The contact angle (CA) in time when the drop was placed (t = 0) was 30 • , 26 • , and 31.5 • for 23-10HT, 27-10HT, and 35-10HT, respectively. In the case of 23-10HT after 4.5 s the CA decreased to approximately 7.5 • , the same effect was achieved after 6.5 s for 27-10HT and 35-10HT. Additionally, the unmodified surfaces of Ti-xMo alloys and microcrystalline titanium were analyzed for comparison purposes. The HT confirms a visible CA decrease for selected covered samples in comparison to untreated samples.

Conclusions
The authors proposed a schematic course of the process presented in Figure 16. The attention should be focused on the substrate composition, structure, size, morphology, homogeneity, and physiochemical properties that follow, influencing HAp layer state and final properties through the intermediate alkaline treatment, that mimics the substrate.
The obtained in this work results indicate that the proposed process enables the preparation of

Conclusions
The authors proposed a schematic course of the process presented in Figure 16. The attention should be focused on the substrate composition, structure, size, morphology, homogeneity, and physiochemical properties that follow, influencing HAp layer state and final properties through the intermediate alkaline treatment, that mimics the substrate.
Ti35Mo samples in a time function.

Conclusions
The authors proposed a schematic course of the process presented in Figure 16. The attention should be focused on the substrate composition, structure, size, morphology, homogeneity, and physiochemical properties that follow, influencing HAp layer state and final properties through the intermediate alkaline treatment, that mimics the substrate.
The obtained in this work results indicate that the proposed process enables the preparation of a hydroxyapatite coating. However, the HAp layer is the most developed for samples treated with alkaline treatment in 7 M and 10 M NaOH solution. The above statement confirms also stable intermediate phase forms obtained on the substrates after AT, necessary as Figure 16 shows for HAp layer formation and growth. The sample comparison in terms of starting chemical composition does not reveal any significant differences between the resulting coatings for the samples after treatment with 7 M and 10 M NaOH. The differences are noticeable only for the samples after treatment with 12 M NaOH, the least developed coating was obtained for the Ti27Mo 12 M sample. Coatings on Ti23Mo 12 M and Ti35Mo 12 M samples have similar characteristics and island-like morphology. Furthermore, conducted research confirms the positive effect of the obtained HAp layers on the wettability surface characteristics in the time managed tests.
The obtained results prove that a proper process control allows overcoming mentioned problems for other technological approaches of HAp deposition, also in large scale and complex parts geometry at a low demandable approach. The hydrothermal treatment approach remains to grow in interest by the research community, and in the future, we hope by its simplicity, also by the manufacturers. Conducted research remains also the first step in further consideration and field The obtained in this work results indicate that the proposed process enables the preparation of a hydroxyapatite coating. However, the HAp layer is the most developed for samples treated with alkaline treatment in 7 M and 10 M NaOH solution. The above statement confirms also stable intermediate phase forms obtained on the substrates after AT, necessary as Figure 16 shows for HAp layer formation and growth.
The sample comparison in terms of starting chemical composition does not reveal any significant differences between the resulting coatings for the samples after treatment with 7 M and 10 M NaOH. The differences are noticeable only for the samples after treatment with 12 M NaOH, the least developed coating was obtained for the Ti27Mo 12 M sample. Coatings on Ti23Mo 12 M and Ti35Mo 12 M samples have similar characteristics and island-like morphology. Furthermore, conducted research confirms the positive effect of the obtained HAp layers on the wettability surface characteristics in the time managed tests.
The obtained results prove that a proper process control allows overcoming mentioned problems for other technological approaches of HAp deposition, also in large scale and complex parts geometry at a low demandable approach. The hydrothermal treatment approach remains to grow in interest by the research community, and in the future, we hope by its simplicity, also by the manufacturers. Conducted research remains also the first step in further consideration and field recognition in a possible adaptation of low-temperature hydrothermal processing in the drug encapsulation and delivery systems.