Alternative technique for calcium phosphate coating on titanium alloy implants.

As an alternative technique for calcium phosphate coating on titanium alloys, we propose to functionalize the metal surface with anionic bath containing chlorides of palladium or silver as activators. This new deposition route has several advantages such as controlled conditions, applicability to complex shapes, no adverse effect of heating, and cost effectiveness. A mixture of hydroxyapatite and calcium phosphate hydrate is deposited on the surface of Ti-6Al-4V. Calcium phosphate coating is built faster compared with the one by Simulated Body Fluid. Cell morphology and density are comparable to the control one; and the results prove no toxic compound is released into the medium during the previous seven days of immersion. Moreover, the cell viability is comparable with cells cultivated with the virgin medium. These experimental treatments allowed producing cytocompatible materials potentially applicable to manufacture implantable devices for orthopedic and oral surgeries.

Introduction

n class="Chemical">Metallin class="Chemical">c artificial implants are widely used as medical devices to replace, support or enhance an existing biological structure. Most of the orthopedic or dental implants are made of titanium alloys due to their advanced mechanical properties and excellent biocompatibility. Per-Ingvar Brånemark first discovered the principles of osteointegration in the 1950s. Pure titanium (Ti) was placed in a rabbit fibula, connected directly to the bone and did not induce any severe inflammatory or other reactions in skin or bone tissues. This unique property of titanium, allowing direct contact and connection to be developed between living bone and metal surface without soft tissue interface, has now multiple applications such as dental implants, craniofacial prosthesis, hip replacements, etc.

Besides n class="Chemical">titanium itself, n class="Chemical">Ti-6Al-4V is one of the most commonly used materials for medical implants because of its suitable mechanical properties, its corrosion resistance and relative inertia toward living tissues. Despite that, there are some uncertainties concerning the release of Al and V ions under certain conditions. Nevertheless, the essential requirements for any artificial implanted material, especially metallic ones, are demonstrated to possess not only biocompatibility, but also bioactivity with the host. Indeed, several months are required to get a sufficient osteointegration, and in numerous cases, the osteointegration is incomplete. Thus, considerable work has been done to investigate various surface modification methods to increase the osteointegration rate such as surface-roughening (e.g., sandblasting and/or acid etching) and coating with hydroxyapatite (HA) Ca10(PO4)6(OH)2. Several physical and chemical methods were reported in the literature to modify titanium implants surfaces such as the plasma spray, magnetron sputtering, dip- and spin- coating sol-gel suspensions, soaking, electrophoresis, pulsed laser deposition, and biomimetic routes.

Sinn class="Chemical">ce the 1960s, n class="Chemical">calcium phosphates (Ca-P) and mainly hydroxyapatite (HA) have been widely used in bone surgeries as they easily bond with bone once they are implanted, in spite of their poor mechanical characteristics. The elaboration of bioactive ceramic coatings on metallic and non-metallic substrates implanted into the human body is, therefore, a way to combine the mechanical properties of the substrate and the bioactivity of the ceramic layer. To achieve fast and long-term bioactive surfaces, implant surfaces should be coated with a developed Ca-P for its superior biocompatibility and bioactive acting as a barrier against ion release.

In the late 80s, the den class="Chemical">velopmenpan>t of osteoconductive HA plasma spray coatings on metallic prostheses allowed to improve the biological behavior of the metallic surface. Although plasma spray technique is currently employed to produce clinically-used coatings, the long-term stability of the coating/implant is questionable. The poor coating, substrate adherence and the lack of uniformity of the coating affect the implant long-term lifetime. These drawbacks have been shown to arise from the important residual internal stress induced in thermally sprayed coating during deposition. Indeed, a network of micro-crack developed in the ceramic coating volume in interface failures was induced.

Many n class="Chemical">Ca-P coatings on titanium have been done by immersing chemically-treated metal and polymer implants, in an aqueous solution containing calcium and phosphate ions at pH and physiological temperatures in order to simulate the biological process. Among chemical routes, the alkali–heat treatment of Ti followed by the biomimetic process seemed to be one of the most popular and effective methods. Dipping in alkali solution produces a Na-containing titanate layer, able to induce the deposition of Ca-P ceramics in vitro as well as in vivo. Other reports indicated that acid–alkali treatment on Ti implants- might be more effective for the deposition of a homogeneous bone-like apatite layer on the surface of Ti when immersed in simulated body fluid (SBF). This biomimetic process imitates the growth of in situ-formed bone-like HA crystals. Thus, the HA coating is more readily degraded by osteopath overcoming many of the drawbacks associated with other deposition techniques, such as the processes utilizing high temperature treatment. It has also been proven that the chemical pre-treatment in alkali solution can improve the bonding between titanium substrate and calcium phosphate coatings fabricated by the subsequent biomimetic deposition in SBF.,

As an n class="Chemical">alternpan>an class="Chemical">tive technique for Ca-P coating, Oliviera, Leonor and Reis proposed to treat the surface of non-conducting materials such as polymers in an autocatalytic bath.- This method is based on redox reactions with electroless plating using palladium chloride (PdCl2) as a catalyst. In this process, the PdCl2 works as an arm of ionic change between the substrate and the solution.

In the present study, we slightly modify this method and extend it to the growth of n class="Chemical">Ca-P layer on the n class="Chemical">metal substrate. After a mechanical polishing and a chemical treatment (e.g., using Kroll reagent) on the titanium substrate, a ceramic interlayer (sodium titanate, HT) is chemically generated by the action of an alkaline bath (NaOH), covalently linked to the metal substrate and ceramic coating. The alkaline pre-treatment on the Ti substrate and the further thermal treatments are performed in order to obtain a micrometric network layer (HT). Indeed, the Ca-P layer is produced using palladium chloride (PdCl2, ACM1) or silver chloride (AgCl, ACM2) to activate the surface, according to Reis’s method. This new deposition route has several advantages over plasma spray, pulsed laser deposition, as well as magnetron sputtering offering: an easy way to control the experimental conditions, the applicability to complex shapes, and especially no adverse effect of heating. Furthermore, this method is cost effective.

Results and Discussion

After the n class="Chemical">alkn class="Chemical">aline-heat treatment, a layer of titanium oxide (TiO2) and sodium titanate (Na2TinO2n+1, n = 3, 5, 6) was created on the surface of the metal according to previous works., The structure of sodium titanate can be schematized in Scheme 1.

Scheme 1. Structure of Na2TinO2n+1.

Sn class="Chemical">cheme 1. Strun class="Chemical">cture of Na2TinO2n+1.

The n class="Chemical">Ca-P layer was thenpan> deposited by n class="Chemical">dipping the substrates into the baths ACM1 and ACM2. Their compositions are given in the experimental part. The layers were analyzed by XRD and the morphologies were observed by TEM and SEM techniques. No difference between both XRD patterns was observed; the patterns exhibit only the diffraction peaks of α-Ti and β-Ti structures showing that the sodium titanate and Ca-P layers are either too thin or made of so small crystallites that the peaks are too broad. The surface morphologies were examined by SEM (Figs. 1C and 2C). A dense coating made of nanoparticles is clearly observed on ACM1 (Fig. 1C). The analysis shows that the layer is composed mainly of Ca and P. No sodium titanate network is detected. On the contrary, the ACM2 coating is less dense than the previous one as proved by EDS-X analyses. Indeed, Ca and P are detected as well as Na and Ti.

Figure 1. ACM1: (A) TEM observations on the scratched coating and the corresponding SAED pattern (insert). (B) XRD pattern, where TS is sodium titanate, R rutile, and S substrate. (C) Morphology by SEM and (D) corresponding EDS-X analyses.

Figure 2. ACM2: (A) TEM observations on the scratched coating and the corresponding SAED pattern (insert). (B) XRD pattern. (C) Morphology by SEM and (D) the corresponding EDS-X analyses.

Figure 1. n class="Chemical">ACM1: (A) TEM obsern class="Chemical">vations on the scratched coating and the corresponding SAED pattern (insert). (B) XRD pattern, where TS is sodium titanate, R rutile, and S substrate. (C) Morphology by SEM and (D) corresponding EDS-X analyses.

Figure 2. n class="Chemical">ACM2: (A) TEM obsern class="Chemical">vations on the scratched coating and the corresponding SAED pattern (insert). (B) XRD pattern. (C) Morphology by SEM and (D) the corresponding EDS-X analyses.

Typin class="Chemical">cn class="Chemical">al TEM examinations of the scratched coatings are presented in Figures 1A and 2A for ACM1 and ACM2, respectively. The ACM1 collected particles appear to be in shape of rods of about 100–200 nm long while the ACM2 ones present more isotropic shapes in the 200–300 nm size range. Selected area electron diffraction (SAED) patterns, recorded on each sample (inserts in Figs. 1A and 2A), are characteristic for polycrystalline samples. The experimental lattice distances are given in Table 1 and interpreted in terms of a mixture of hydroxyapatite (Ca5[PO4]3[OH], ICDD 9–432) and calcium phosphate hydrate (calcium hydrogen phosphate hydrate, Ca4H[PO4]3·3H2O, ICDD 11–184) on both ACM1 and ACM2.

Table 1. Experimental lattice distances measured on SAED (Figs. 1A and 2A) recorded on ACM1 and ACM2 samples

ACM1dexp(± 0.005 nm)	ACM2dexp(± 0.005 nm)	Hydroxyapatite,(Ca5(PO4)3(OH), ICDD 09–432),dtheo (nm)	Calcium hydrogen phosphate hydrate, (Ca4H(PO4)3·3H2O, ICDD 11–184)d theo(nm)
	0.370		0.374
0.350	0.352		0.348
0.342		0.344	0.343
0.336			0.338
0.310	0.311	0.308/0.317	0.305
0.278	0.275	0.281/0.279/0.272	0.283/0.277/0.275
	0.207	0.207	0.208
0.194	0.197	0.194	0.195
	0.189	0.188	0.186
0.182		0.184
0.175		0.175

Figure 3 shows the morphology of HT (Fig. 3A), n class="Chemical">ACM1 (Fig. 3B) and n class="Chemical">ACM2 (Fig. 3C) samples after immersion in SBF for 7 d. Ca and P concentrations after immersing the samples in SBF1 for 2, 4 and 7 d are given in Figure 3D and E. It was assumed that any change by decrease/increase in Ca or P concentrations, compared with the SBF reference solution, would represent a deposition/release on/from the sample surfaces respectively.

Figure 3. Morphology by SEM after one week in SBF (A) alkaline-heat treated titanium alloy (HT), HT treated after (B) ACM1 and (C) ACM2 baths. Concentrations of (D) Ca and (E) P ions in SBF1 after for 2, 4 and 7 d of the samples HT, ACM1 and ACM2.

Figure 3. Morphology by SEM after one week in SBF (A) n class="Chemical">alkn class="Chemical">aline-heat treated titanium alloy (HT), HT treated after (B) ACM1 and (C) ACM2 baths. Concentrations of (D) Ca and (E) P ions in SBF1 after for 2, 4 and 7 d of the samples HT, ACM1 and ACM2.

Ca or P ions release from SBF is obsern class="Chemical">ved after 2, 4 and 7 d for ACM1 and ACM2, (Fig. 3D and E). A sharp decrease in the ionic concentration of both Ca and P ions is recorded giving as result a dense Ca-P layer during all soaking periods (Fig. 3B and C). Differently, Ca ions for HT sample (Fig. 3A) are not modified, while P ions are slightly absorbed onto the surface.

Both tested spen class="Chemical">cimenpan>s (n class="Chemical">ACM1 and ACM2) were not toxic for human osteoblasts. In fact, as reported in Figure 4A, the cell viability did not differ (P > 0.05) between control and test groups. Viability was observed onto experimental materials always higher than 95% with respect to control.

Figure 4. Direct cytocompatibility assay. Osteoblasts cultivated directly onto specimens’ surfaces; the viability values are comparable to the control ones (A) and no statistical differences were calculated (P > 0.05). Bars represent mean and standard deviations. Moreover, the test and control cell morphology (phalloidin in red and DAPI in blue) is comparable as shown in (B). Bar scale = 50 μm.

Figure 4. Direct n class="Chemical">cytocompatibility assay. Osteoblasts cultivated directly onto specimens’ surfaces; the viability values are comparable to the control ones (A) and no statistical differences were calculated (P > 0.05). Bars represent mean and standard deviations. Moreover, the test and control cell morphology (phalloidin in red and DAPI in blue) is comparable as shown in (B). Bar scale = 50 μm.

The n class="Chemical">viability assay results were confirmed by the morphological analysis; in fact, by immunofluorescence staining it is possible to appreciate the correct spread and density of cells cultivated onto experimental specimens surface that are comparable with the control (Fig. 4B). Also, SEM images confirmed the absence of morphological differences between cells cultivated onto treated samples and those onto the control material (Fig. 5).

Figure 5. SEM images of osteoblasts seeded directly onto specimen surfaces after 24 h Control (upper panel, left) and treated (ACM1and ACM2) samples. The control samples morphology can be stated to be comparable.

Figure 5. SEM images of osteoblasts seeded diren class="Chemical">ctly onto spen class="Chemical">cimen surfaces after 24 h Control (upper panel, left) and treated (ACM1and ACM2) samples. The control samples morphology can be stated to be comparable.

n class="Chemical">Cells cultivated with the medium previously kept in direct contact for one week with treated samples showed a viability comparable with cells cultivated with the virgin medium (Fig. 6A). In fact, no statistical significant differences were observed between control and test samples. The cell viability was always registered higher than 98% of the control.

Figure 6. Non-direct cytocompatibility assay. Osteoblasts cultivated with medium stored one week in direct contact with treated samples; viability values are comparable with controls (A) and no statistical differences were noticed (P > 0.05). Bars represent means and standard deviations. Moreover, cells morphology and density investigated by light microscopy were comparable between test and control samples as shown in (B). Bar scale = 50μm.

Figure 6. Non-direct n class="Chemical">cytocompatibility assay. Osteoblasts cultivated with medium stored one week in direct contact with treated samples; viability values are comparable with controls (A) and no statistical differences were noticed (P > 0.05). Bars represent means and standard deviations. Moreover, cells morphology and density investigated by light microscopy were comparable between test and control samples as shown in (B). Bar scale = 50μm.

n class="Chemical">Visun class="Chemical">al investigation using light microscopy confirmed the viability assay results. As shown in Figure 6B, the cell morphology and density are similar between control and test samples, suggesting that no toxic compounds were released into the medium by the treated specimens during the previous seven days of immersion.

These ann class="Chemical">alyses show that n class="Chemical">Ca-P nanostructured layers made of hydroxyapatite Ca5(PO4)3(OH) and calcium phosphate hydrate (Ca4H[PO4]3·3H2O) cover the titanium substrate well. The layer is thicker and denser when Pd is involved. The hypothesis made previously by Reis et al. to explain the mechanism leading to the Ca-P deposition on polymers is schematized in Figure 7. First, palladium ions link onto the surface through the hydroxyde groups (Fig. 7A, step 1). Then, the negatively charged phosphate ions and the positively charged Ca ions are successively attracted so that this stacking results in the formation of calcium phosphate compounds (Fig. 7B, step 2). When silver is involved, a similar mechanism can be proposed. However, according to Reis et al., Pd ions are not eliminated. In case of Ag, the bath has the advantage to introduce ions known to have efficient antimicrobial properties; moreover, as recently demonstrated by Alan B.G. Lansdown, the pharmaceutical and toxicological profile of Ag in biomaterials shows that health risks associated with systemic absorption of silver as Ag are low.

Figure 7. Hypothesis mechanism of Ca-P bioactive coating in the ‘activating’ baths by using PdCl2 as the catalyst.

Figure 7. Hypothesis men class="Chemical">chanism of n class="Chemical">Ca-P bioactive coating in the ‘activating’ baths by using PdCl2 as the catalyst.

To n class="Chemical">create an osteoconductive titanium alloy, a new process route was applied. It consists of an immersion of Ti-6Al-4V in alkaline solution and a further heat treatment, followed by the immersion in a PdCl2 or AgCl -containing baths. The obtained surface morphologies were optimized by the growth of porous sodium titanate layer. It was observed that the formation of HA and calcium phosphate hydrate was induced after 3 h immersion in ACM1 and ACM2 baths. The structure and morphology of the so-formed layers were characterized using several techniques.

These experimentn class="Chemical">al treatmenpan>ts allowed producing cytocompatible materials potentially applicable to manufacture implantable devices for orthopedic and oral surgeries.

Materials and Methods

The samples were n class="Chemical">cut from a bar of commercially available Ti-6Al-4V ELI titanium alloy (chemical composition in wt.%: C, 0.013; Fe, 0.16; N, 0.014; H2, 0.006; O, 0.11; Al, 6.05; V, 4.0; Ti, bal.). The specimens of 20 mm diameter and 1.5 mm thickness were abraded with SiC abrasive papers 240, 400, 1000, 2500, 4000 and polished to mirror with non-crystallizing colloidal silica suspension; then cleaned by successive ultrasonic treatments in acetone, ethanol and distilled water for 30 min. After mechanical polishing, the substrates were etched for 10 min in Kroll solution’s reagent, a mixture of 0.045 M HF and 0.058 M HNO3, at room temperature. The specimens were rinsed with distilled water and dried at 40 °C in an electric oven.

To form the n class="Chemical">sodium titanate hydrogel layer on the n class="Chemical">Ti-6Al-4V surface, the samples were dipped into a NaOH 10 M solution at 60 °C for 24 h and then thermally treated at 630 °C for 1 h with a heating rate of 5 °C/min.

To produn class="Chemical">ce n class="Chemical">Ca-P coatings, the pre-treated specimens were soaked in two baths containing calcium chloride (CaCl2) and phosphoric acid (H3PO4) and either palladium chloride (PdCl2, ACM1) or silver chloride (AgCl, ACM2) under magnetic stirring for 3 h at (80 ± 2) °C and pH 5.3 ± 0.1. CaCl2 acted as the source of calcium, while H3PO4served as a source of phosphate. The molar Ca/P ratio was 1.67. Then, the samples were cleaned in distilled water and dried in environmental conditions.

To en class="Chemical">vn class="Chemical">aluate the in vitro formation of Ca-P, the samples were soaked in simulated body fluid (SBF) solution at pH 7.4 using a buffer solution consisting of Tris hydroxymethylaminomethane and HCl, at body temperature (37 °C). Ion concentrations of SBF solution and human blood plasma are shown in Table 2.

Table 2. Ion concentrations of SBF

	Concentrations of ions, mM
	Na+	K+	Mg2+	Ca2+	Cl-	HCO3-	HPO42-	SO42-	pH
Plasma body human	142.0	5.0	1.5	2.5	103.0	27.0	1.0	0.5	7.2- 7.4
SBF (type c)	142.0	5.0	1.5	2.5	147.8	4.2	1.0	0.5	7.4

The n class="Chemical">change inpan> n class="Chemical">Ca and P concentrations after 2, 4, 7 d soaking periods of samples was analyzed using the UV-Visible spectroscopy and Lambda 950 UV/VIS spectrometer (Perkin Elmer). After soaking, the specimens were washed with distilled water and dried at room temperature.

The n class="Chemical">Ca-P coatings were analyzed by X-ray diffraction (XRD) using a Bruker D5000 diffractometer equipped with a quartz monochromator and Cu Kα radiation (λ = 0.154 06 nm). Diffraction scans were recorded in the 20–60° angular range with a 0.03° 2θ-step and a counting time of 4 s.

Sn class="Chemical">canninpan>g electron microscopy (SEM) observations of the coating morphologies were performed using a JEOL6700F equipped with an Energy Dispersive X-ray spectrometer (EDS-X). Images were recorded at 3 kV and EDS-X analysis at 10 kV on 10 × 10 mm2 area and repeated three times at least.

The coated layers of samples n class="Chemical">ACM1 and ACM2 were scratched and collected on a microscope copper grid. Transmission electron microscope images (TEM) were recorded using a TOPCON 002B transmission electron microscope operating at 200 kV and equipped with a CCD camera and an EDS-X to analyze the chemical composition.

n class="Species">Human n class="Disease">osteosarcoma cells (MG63, ATCC: CRL-1427) were cultivated at 37 °C, 5% CO2 in Dulbecco’s modification Minimal Essential Medium (DMEM, Sigma-Aldrich) supplemented with 10% fetal bovine serum (FBS, Lonza) and 1% penicillin-streptomycin. When the cells reached 85–90% confluence, they were detached by trypsin/EDTA (EDTA, Sigma-Aldrich), harvested and used for experiments.

Round spen class="Chemical">cimenpan>s (15 mm diameter) were plan class="Chemical">ced into 12 multi well plates (CellStar, PBI International); 2 × 104 cells/sample were seeded directly onto the surfaces of each specimen in a low volume of medium (200 µl) and allowed to adhere for 4 h prior to adding 2 ml of fresh medium to each well; plates were then incubated for 24 h at 37 °C, 5% CO2. Afterwards, the cell viability was evaluated with the (3-[4,5-Dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) colorimetric assay (MTT, Sigma-Aldrich). Briefly, 100 µl of MTT solution (3 mg/ml in phosphate buffered saline [PBS]) were added to each sample and incubated 4 h in the dark at 37 °C; afterwards, formazan crystals were solved with 100 µl of dimethyl sulphoxyde (DMSO, Sigma) and 50 µl were collected and centrifuged to remove eventually debris. Surnatants optical density (o.d.) was evaluated at 570 nm with a spectrophotometer (Spectra Count, Packard Bell). Uncoated titanium o.d. was used as a control and considered as 100% cell viability while coated samples (named ACM1 and ACM2) viability was calculated as follows: (sample o.d. / titanium control o.d.) × 100. Experiments were performed 6 times for controls and each different coating.

Furthermore, the immunofluoresn class="Chemical">cenpan>n class="Chemical">ce staining was performed in order to investigate the cells morphology after 24 h of adhesion onto the biomaterials surface. Cells were fixed for 20 min with 4% paraformaldehyde at room temperature (RT) and then washed 3 times with PBS. Phalloidin (rhodamine B tetramethylisothiocynate, 1/2000 in PBS, AbCam) solution was added for 45 min (RT) in order to investigate cells cytoskeleton. Afterwards, the samples were washed 3 times with PBS and co-stained with 4',6-diamidino-2-phenylindole (DAPI, Sigma-Aldrich). The cells morphology was optically evaluated using a fluorescence microscope (Leica AF 6500, Leica Microsystems).

Serum free n class="Chemical">DMEM was inpan>cubated without cells for 1 wk at 37 °C, 5% CO2 in direct contact with controls or coated samples in a ratio of 2 ml/specimens for a total of 10 ml/specimen. Afterwards, eluates were collected, supplemented with 10% FBS and used to cultivate MG63. Cells were seeded in a defined number (2 × 104 /well) into 24 wells plates (Cell Star, PBI International) and cultivated for 24 h at 37 °C, 5% CO2. Afterwards, the cell viability was evaluated by the 3-(4,5 Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide colorimetric assay (MTT, Sigma) as described for the direct cytocompatibility assay.

SEM was used to inn class="Chemical">vesn class="Chemical">tigate the morphology of cells seeded directly onto the specimens’ surface. Briefly, after fixing, dehydration with the ascendant ethanol scale (70%, 80%, 90%, 100%), and hexamethyldisilazane, the specimens were mounted on aluminum stubs, using a conductive carbon tape, coated with a thin carbon layer (10 nm) and observed with a StereoScan 360 SEM (AbCam) at 10 kV with various magnifications, using secondary electrons.

Furthermore, the n class="Chemical">cells morphology was n class="Chemical">visually investigated after one week of cultivation by light microscopy (Leica AF 6500, Leica Microsystems).

Stan class="Chemical">tisn class="Chemical">tical analysis was done using Levene’s test and ANOVA one-ways followed by Sheffè’s post-hoc test using IBM SPSS® Statistics 20 (IBM).