Materials Science and Engineering C 49 (2015) 159–173

Contents lists available at ScienceDirect

Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec

Bioactivity of coatings formed on Ti–13Nb–13Zr alloy using plasma electrolytic oxidation Maciej Sowa a, Magdalena Piotrowska a, Magdalena Widziołek b, Grzegorz Dercz c, Grzegorz Tylko b, Tadeusz Gorewoda d, Anna M. Osyczka b, Wojciech Simka a,⁎ a

Faculty of Chemistry, Silesian University of Technology, B. Krzywoustego Street 6, 44-100 Gliwice, Poland Faculty of Biology and Earth Sciences, Jagiellonian University, Gronostajowa Street 9, 30-060 Kraków, Poland Institute of Materials Science, University of Silesia, 75 Pulku Piechoty Street 1a, 41-500 Chorzow, Poland d Institute of Non-Ferrous Metals, Sowińskiego Street 5, 44-100 Gliwice, Poland b c

a r t i c l e

i n f o

Article history: Received 18 July 2014 Received in revised form 9 December 2014 Accepted 20 December 2014 Available online 23 December 2014 Keywords: Titanium alloy Surface modification Anodic oxidation Osteogenesis Adult mesenchymal stem cells

a b s t r a c t In this work, we investigated the bioactivity of anodic oxide coatings on Ti–13Nb–13Zr alloy by plasma electrolytic oxidation (PEO) in solutions containing Ca and P. The bioactive properties of the films were determined by immersion in simulated body fluid (SBF), and their biocompatibility was examined using adult human bone marrow derived mesenchymal stem cells (hBMSCs). The oxide layers were characterised based on their surface morphology (SEM, AFM, profilometry) as well as on their chemical and phase compositions (EDX, XRF, XRD, XPS). We report that anodic oxidation of Ti–13Nb–13Zr led to the development of relatively thick anodic oxide films that were enriched in Ca and P in the form of phosphate compounds. Furthermore, the treatment generated rough surfaces with a significant amount of open pores. The surfaces were essentially amorphous, with small amounts of crystalline phases (anatase and rutile) being observed, depending on the PEO process parameters. SBF soaking led to the precipitation of small crystals after one week of experiment. During culturing of hBMSCs on the bioactive Ti–13Nb–13Zr surfaces the differentiation of human mesenchymal stem cells toward osteoblasts was promoted, which indicated a potential of the modified materials to improve implant osseointegration. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Currently, bone reconstruction strategies often involve the use of metallic biomaterials. Titanium and its alloys are commonly used because of their good mechanical strength, low density, chemical inertness in the tissue environment, biocompatibility, ability to osseointegrate and the lowest elastic modulus (i.e., 48–112 GPa) compared to other metallic biomaterials [1–3]. Ti–6Al–4V alloy is widely used, despite the fact that it was initially developed for aeronautical use. However, problems regarding the use of this alloy in biomedical engineering have been reported. Poor wear resistance and a relatively high elastic modulus (110 GPa) can be listed among these problems [4,5]. Moreover, Ti–6Al–4V alloy is considered to be toxic due to the presence of carcinogenic vanadium and its oxide (V2O5) and possibly neurotoxic aluminium [6–10]. As a result, a new generation of vanadium-free alloys has been engineered [11–13], including Ti–13Nb–13Zr, which has been identified as a potentially non-toxic alloy that is capable of supporting bone growth. Anodic oxidation is a relatively non-expensive method for surface modification. It can be performed via two distinct modes of operation. The conventional mode, whereby the metal workpiece is polarised ⁎ Corresponding author. E-mail address: [email protected] (W. Simka).

http://dx.doi.org/10.1016/j.msec.2014.12.073 0928-4931/© 2014 Elsevier B.V. All rights reserved.

anodically at voltages that do not exceed the potential corresponding to dielectric breakdown of the oxide layer, and the second mode — plasma electrolytic oxidation (PEO) or micro-arc oxidation (MAO). The latter name comes from the plasma- and thermochemical reactions that occur on the surface of the treated metal, when the threshold of dielectric breakdown voltage is breached. Relatively thin, homogeneous oxide coatings of barrier type can be produced by conventional anodising. In contrast, PEO coatings are highly porous and considerably thicker than the conventional anodic films. The control of thickness and surface morphology of the oxide coating permits indirect manipulation over corrosion resistance of the anodised material. Thick (micrometer range), tight and compact oxide layer, free from deeply penetrating cracks, slows down the rate of corrosion of underlying metal or metal alloy. Furthermore, increased amounts of electrolyte species can be incorporated into the structure of the oxide layer in the course of PEO treatment [14–18]. Adult human bone marrow derived mesenchymal stem cells (hBMSCs) are the multipotential cells capable of differentiating into osteoblasts, adipocytes, chondrocytes and myoblasts and other cell types [19–21]. The process of bone regeneration requires the recruitment and migration of osteoblast progenitors via the blood and epithelium of blood vessels to specific locations within the organism or to the MSC niche in bone marrow [22]. Next, osteoprogenitors

160

M. Sowa et al. / Materials Science and Engineering C 49 (2015) 159–173

begin to produce collagen-rich extracellular matrix, which is subsequently mineralised [22,23]. Biomaterials designed to be used in bone reconstruction should at least be osteoconductive or osteoinductive. Thus, they should not only stimulate the migration of osteoprogenitor cells toward bone defects, but also support or stimulate the differentiation of these progenitors into osteoblasts to repair the damaged bone tissue [2,19,24,25]. Surface of a biomaterial which has been enriched with calcium and phosphorus can be characterised in terms of its chemical composition (Ca/P ratio), surface topography and crystallinity. It is believed that calcium phosphate coatings with Ca/P atomic ratio are similar to that of stoichiometric hydroxyapatite (approximately 1.67) in order to induce apatite precipitation in simulated body fluid (SBF). However, it was determined that higher Ca/P ratios increase the adsorption of adhesion proteins onto modified surface. Nanotopography of material has a direct impact on adsorption of adhesion proteins, which in turn affects cell adhesion and proliferation, while microtopography provides socalled niches with conditions suitable for cell growth. High crystallinity of CaP in the coatings promotes slower dissolution, which has a detrimental effect on cell attachment, spreading and proliferation [26]. Although Ti–13Nb–13Zr alloy has been widely described in the recent literature [16,27–29], to the authors' best knowledge, this is the first report on the ability of this alloy to support the osteogenic differentiation of hBMSCs in vitro. The surface modification of Ti–13Nb–13Zr alloy using PEO in solutions containing calcium and phosphorus compounds was performed to produce anodic oxide layers, which exhibited bioactive properties. The obtained coatings were characterised based on their surface morphology as well as their chemical and phase compositions. The bioactivity of these coatings was determined by an SBF immersion test, and the biocompatibility of the coatings was studied in short- and long-term osteogenic hBMSC cultures. 2. Materials and methods

Table 2 Sample labels, treatment conditions and roughness parameters of the titanium alloy specimens. Sample

Ca(H2PO2)2, mol dm−3

H3PO4, mol dm−3

U, V

Ra, μm

TNZ-I-200 TNZ-I-400 TNZ-II-200 TNZ-II-400 TNZ-III-200 TNZ-III-400 TNZ-IV-200 TNZ-IV-400

0.01

–

0.1

–

1

–

0.1

0.1

200 400 200 400 200 400 200 400

0.10 ± 0.03 0.14 ± 0.02 0.09 ± 0.02 0.43 ± 0.09 0.51 ± 0.08 1.50 ± 0.18 0.13 ± 0.03 0.91 ± 0.15

were rinsed with distilled water and ultrasonically cleaned in a deionised water/2-propanol mixture for 5 min. A DC power supply (PWR 800H, Kikusui, Japan) was used throughout these treatments. The process was performed in a water-cooled electrolysis cell with a titanium mesh cathode and magnetic stirrer. The titanium alloy specimen served as the anode. 2.3. Surface property characterisation 2.3.1. Scanning electron microscopy and energy-dispersive X-ray analysis The modified Ti–13Nb–13Zr surfaces were examined in terms of their elemental compositions and morphology by the use of a scanning electron microscope (SEM, Hitachi S-3400N, Japan) operating at an accelerating voltage of 25 kV. 2.3.2. Surface roughness characterisation The Ra parameter of the oxide layers on the Ti–13Nb–13Zr specimens was determined using a Mitutuyo Surftest SJ-301 profilometer, in accordance with the EN ISO 4287:1997 standard [30]. The Ra parameter was calculated according to the formula:

2.1. Materials and pretreatment Ra ¼ Rectangular samples of Ti–13Nb–13Zr alloy (BIMO metals, Wrocław, Poland) were prepared from a sheet of alloy, and their chemical composition is presented in Table 1. The surface area of the face that was modified was equal to 5 cm2. In addition, disc-shaped samples were cut from Ti–13Nb–13Zr alloy rods (diameter = 8 mm). The specimens were mechanically ground using waterproof abrasive paper up to a #800-grit size. They were then chemically etched in a solution that contained 1 mol dm− 3 HF and 4 mol dm− 3 H2SO4 for 1 min. Afterwards, the samples were rinsed with distilled water and ultrasonically cleaned in a deionised water/2-propanol mixture for 5 min. Then, the titanium alloy specimens were electropolished in a solution composed of concentrated H2SO4 (10 mol dm− 3), ethylene glycol (5.4 mol dm− 3) and NH4F (0.4 mol dm− 3). The process was carried out at a constant current density of 1 A cm−2 for 210 s. Then, the samples were cleaned ultrasonically. 2.2. Anodisation of the Ti–13Nb–13Zr alloy The sample labels and process conditions are summarised in Table 2. The pretreated specimens (TNZ-EP) were subjected to anodisation in electrolytic baths I–IV, which contained different amounts of Ca(H2PO2)2 and H3PO4 (Table 2). The anodic oxidation was performed under a constant current of 0.31 A cm− 2 and up to final voltage of either 200 or 400 V for 10 min. The anodised specimens Table 1 Chemical composition of Ti–13Nb–13Zr alloy (wt.%). Nb

Zr

O

Fe

C

H

N

Ti

13.3

13.0

0.10

0.08

0.05

0.009

0.004

Balance

1 l

Zl jZ ðxÞjdx;

ð1Þ

0

where l is the elementary length in the x coordinate, used to determine the irregularity of the investigated surface, and |Z(x)| is the absolute ordinate value inside l. 2.3.3. Atomic force microscopy (AFM) The nanoscale surface topography of selected samples was examined using an atomic force microscope (NanoScope E, Digital Instruments) operating in contact mode. Standard silicon nitride cantilevers with a spring constant of 0.12 N m−1 were used. 2.4. Characterisation of the anodic oxide film composition Four samples were selected based on the preliminary surface studies (described above), for detailed characterisation. Microscopic and spectral methods were used, as described below. 2.4.1. Cross-sectional analysis The cross-section of the anodic oxide films on the titanium alloy samples was examined using a high-resolution field emission scanning electron microscope (HR-FESEM, Inspect F, FEI Company, USA) equipped with an EDX system (EDAX, Oxford Instruments, UK). 2.4.2. X-ray diffractometry The phase compositions of the oxide layers formed on the Ti–13Nb– 13Zr alloy specimens were determined using an X-Pert Philips PW 3040/60 diffractometer operating at 30 mA and 40 kV. A vertical goniometer and Eulerian cradle were used throughout the experiments. The wavelength of the radiation source (λCuKα) was 0.154178 nm.

M. Sowa et al. / Materials Science and Engineering C 49 (2015) 159–173

The grazing incidence X-ray diffraction spectra were recorded over the range of 10–80° for the 2θ angle with a 0.05° step size and an incident angle α = 0.25°. 2.4.3. X-ray photoelectron spectroscopy The XPS spectra of the titanium alloy samples were measured using a Physical Electronics PHI 5700 photoelectron spectrometer. A monochromatic Al Kα radiation source (1486.6 eV) was used, and the ultimate energy resolution of the spectra was estimated to be 0.35 eV. 2.4.4. X-ray fluorescence spectrometry Further examinations of the elemental composition of the oxide layers grown on the Ti-13Nb–13Zr alloy samples were conducted on a ZSX Primus X-ray fluorescence spectrometer (Rigaku) using a Rh tube with the equipment necessary for surface preparation. The instrument was equipped with a scintillation counter and a flow proportional counter. Semi-quantitative XRF analysis was carried out using a set of UniQuant (Omega Data System NL) reference materials. 2.5. Immersion tests in simulated body fluid (SBF) The bioactivity of the selected modified titanium alloy surfaces was determined in protein-free simulated body fluid, in which the concentration of ions corresponded to that of human blood plasma (Na+: 1.42; K+: 0.05; Mg2+: 0.015; Ca2+: 0.025; Cl−: 1.478; HCO− 3 : 0.042; HPO24 − 0.010; and SO24 −: 0.005 mol dm− 3). The SBF was prepared according to the method of Kokubo [31]. The Ti–13Nb–13Zr alloy specimens were immersed in SBF for 4 weeks at 36.5 ± 0.5 °C. The surfaces of the SFB-treated samples were examined by SEM weekly.

161

Bovine Serum and 1% antibiotics (penicillin/streptomycin) at 37 °C in a humidified atmosphere containing 5% CO2. The growth medium was exchanged every 3–4 days. The cells were cultured until a confluent monolayer was achieved. The cells were detached with 0.25% trypsin– EDTA, centrifuged, counted and re-suspended in a fresh growth medium. Cells from passages 1–5 were used for the experimental cultures on selected titanium alloy samples. 2.6.3.1. Short-term (10-days) culture analysis. Prior to cell seeding, the titanium alloy specimens were rinsed with α-MEM to increase the surface wettability and to facilitate cell adhesion. hBMSCs were seeded on the titanium alloy samples at a density of 1.5 × 104 cells cm−2 in α-MEM supplemented with 10% FBS and 1% antibiotics. After 1 day of initial culture, the alloy samples were transferred to new culture plates and cultured in α-MEM supplemented with 10% FBS, 1% antibiotics, 100 μg cm− 3 ascorbate-2-phosphate (Asc) and 10 − 7 mol dm− 3 dexamethasone (Dex). The growth media and supplements were exchanged every 2–3 days. After 10 days, the cells were analysed for viability, alkaline phosphatase (ALP) activity and osteogenic gene expression. 2.6.3.2. Long-term (21-days) culture analyses. hBMSCs were seeded on the titanium alloy specimens, as described for the short-term cultures. After 1 day of initial culturing, the alloy samples were transferred to new culture plates and cultured as described above, except that the culture medium was supplemented with 10 nmol dm− 3 β-glycerophosphate (BGP), in addition to Asc and Dex. The media and supplements were exchanged every 2–3 days. After 3 weeks, the cells were analysed for viability, collagen production and calcification of the extracellular matrix (ECM).

2.6. Biological experiments Two samples were chosen (TNZ-III-400 and TNZ-IV-400), for further in vitro examination using hBMSC. 2.6.1. Biological reagents Unless otherwise stated, all chemicals were purchased from SigmaAldrich. Phosphate buffered saline solution (PBS) and 0.25% trypsin in 1 mmol dm− 3 of tetrasodium EDTA were obtained from Hyclone. Mesenchymal stem cells (MSC)-Qualified Foetal Bovine Serum (FBS), α-modified Eagle's medium (α-MEM), phenol-free α-MEM, penicillin/ streptomycin solution, nuclease-free water, SuperScript III 1st Strand Synthesis System, SYBR Green Polymerase Chain Reaction (PCR) Master Mix, TaqMan Gene Expression Master Mix and TaqMan Gene Expression Assays were obtained from Life Technologies. CellTiter 96 Aqueous One Solution was purchased from Promega, and Ficoll-Paque Premium was purchased from GE Healthcare. TRIzol reagent was obtained from Ambion. Recombinant human bone morphogenic protein-2 (rhBMP2) was purchased from R&D Systems. Primers were obtained from Genomed. 2.6.2. Preparation of titanium alloy specimens All samples used in the cell culture studies were in the form of 8 × 4 mm discs or 15 × 8 × 2 mm rectangular blocks. Prior to cell culture, the samples were rinsed with distilled water and sterilised at 180 °C for 30 min. 2.6.3. Cell culture Adult human bone marrow-derived mesenchymal stem cells (hBMSCs) were obtained from the iliac crest of 4 patients (age 29–68, both genders) according to an Institutional Review Board approved protocol (KBET/17/L/2007). Ficoll-Paque Premium was used to isolate the mononuclear cell fraction, as described in previous studies [32]. The fraction was collected and washed with balanced saline solution (BSS) according to the original Ficoll-Paque Premium protocol. The cells were then cultured in α-MEM supplemented with 10% MSC-Qualified Foetal

2.6.3.3. SEM analysis. The morphology of hBMSCs cultured on the titanium alloy samples was investigated using SEM at day 5 of the short-term cultures. The cells were fixed with a 2.5% glutaric aldehyde in a 0.1 mol dm−3 cacodylic buffer. After fixation, the alloy samples were rinsed with 0.1 mol dm− 3 cacodylic buffer several times, subjected to osmium post-fixation for 1 h, and then washed 3 times in 0.1 mol dm− 3 cacodylic buffer. The samples were subsequently dehydrated in increasing concentrations of ethanol (50–100%), rinsed with 100% acetone, dried at the critical point (− 200 °C for 40 min) and left overnight under vacuum. The dehydrated samples were coated with a thin film of gold (JFC-1100F, Tokyo, Japan) and observed by SEM (JEOL JSM5410, Tokyo, Japan). 2.6.4. Cell viability hBMSC viability was analysed using the CellTiter 96 Aqueous One Solution Cell Proliferation Assay. Briefly, the cultures were rinsed twice with PBS and then immersed in 3-(4,5-dimethylthiazol-2-yl)-5(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) solution, which was diluted 1:10 in phenol-free α-MEM. The cultures were incubated for 10–20 min at 37 °C in a humidified 5% CO2 atmosphere. After this incubation period, 200-μl aliquots of culture medium containing MTS were transferred from each well to separate wells of a 96-well plates, and the absorbance at 490 nm was measured using a plate reader. 2.6.5. ALP activity Alkaline phosphatase activity was measured after 10 days of cell culture. hBMSCs were washed twice with PBS and then lysed in digestion buffer, as previously described [32]. The protein lysates were extracted overnight at 4 °C. The protein extracts were then transferred to 1.5 cm3 centrifuge tubes and separated from the cell debris by centrifugation at 12,000 rpm for 10 min at room temperature. Supernatants were used for the enzymatic reaction. For the assay, 100-μl aliquots of protein extracts were transferred to a spectrophotometer cuvette containing 900 μl of a substrate solution (50 mmol dm−3 p-nitrophenyl phosphate

162

M. Sowa et al. / Materials Science and Engineering C 49 (2015) 159–173

in cell assay buffer comprising: 1.5 mol dm−3 Tris, 1 mmol dm−3 ZnCl2 and 1 mmol dm−3 MgCl2·6H2O). Changes in the absorbance at 405 nm were monitored for 6 min at 1 min intervals using a spectrophotometer (SHIMAZU UV-1650 PC). Total ALP activity was determined as the nmol p-nitrophenol/min/total volume of protein extract. ALP activity was normalised to the number of viable cells estimated from the MTS assay and expressed relative to the control (i.e., cultures on the TNZEP samples). 2.6.6. Collagen production and extracellular matrix calcification To examine collagen production and extracellular matrix calcification, after 21 days of hBMSC culture, the cells were first analysed for cell viability, washed twice with PBS and then fixed with an 8% formalin solution. For collagen detection, Sirus Red solution in saturated picric acid (1 g dm−3) was added to each culture well. Then, staining was performed for 18 h at room temperature with gentle rocking. Afterwards, the staining solution was removed, the cultures were extensively washed with distilled water and then de-stained for 20 min using 1:1

solution of 0.2 mol dm−3 sodium hydroxide and methanol. Extracellular matrix calcification was analysed using Alizarin Red S solution in distilled water (1 g dm−3), with the pH adjusted to 5.5 using ammonium hydroxide. Staining was performed for 30 min at room temperature with vigorous shaking. De-staining was performed with gentle rocking for 20 min using 5% perchloric acid aqueous solution. The extracted dye was removed, and 200-μl aliquots of the recovered staining solutions were transferred to a clean 96-well plate to measure the absorbance at 490 nm using a plate reader. The results were normalised to the number of viable cells obtained from the cell viability assay and expressed relative to the collagen production or ECM calcification by hBMSCs cultured on control TNZ-EP samples. 2.6.7. Osteogenic gene expression The mRNA levels of selected osteogenic markers were analysed 10 days after the hBMSC cultures were established on the selected modified titanium alloy samples (TNZ-III-400 and TNZ-IV-400). The results were compared to the gene expression levels of cultures

Fig. 1. SEM micrographs of the titanium alloy surfaces modified under various conditions.

M. Sowa et al. / Materials Science and Engineering C 49 (2015) 159–173

established on pretreated Ti–6Al–7Nb alloy samples (TNZ-EP). RNA was extracted using TRIzol reagent, and 1-μg aliquots of the total RNA from each culture were reverse-transcribed to cDNA (complementary DNA) using the SuperScript III 1st Strand Synthesis System. The gene expression levels were analysed using a StepOnePlus Real Time PCR instrument (Applied Biosystems) and the Delta–Delta-Ct (ddCT) method using 1-μl aliquots of cDNA from the first-strand reactions. TaqMan Gene Expression Assays were used for the analysis of alkaline phosphatase (ALP Hs01029144 m1) and Tata binding protein (TBP Hs00427620 m1; house-keeping gene) mRNA levels. Each reaction mixture (10 μl of total volume) contained cDNA, a specific TaqMan probe and TaqMan Gene Expression Master Mix, as recommended by the manufacturer. For the analysis of the collagen type I (ColI) and osteocalcin (OC) mRNA levels, the following primer sets were used: Collagen type I (ColI): forward: 5′-GTCTAGACATGTTCAGCTTTGT GGA-3′, reverse: 5′-CTTGGTCTCGTCACAGATCACGTCAT-3′; Osteocalcin (OC): forward: 5′-AAGAGACCCAGGCGCTACCT-3′, reverse: 5′-AACTCGTCACAGTCCGGATTG-3′. Each reaction mixture (15 μl of total volume) contained cDNA, a primer pair, SYBR Green I, AmpliTaq Gold DNA Polymerase and the reaction buffer, as recommended by the manufacturer (SYBR Green PCR Master Mix). The PCR reactions were performed for 40 cycles using a denaturation step at 95 °C for 30 s, annealing at 60 °C for 1 min and elongation at 72 °C for 30 s. The results were expressed relative to the mRNA levels in the cultures grown on the pretreated titanium alloy samples (TNZ-EP). 2.6.8. Statistical analyses The results are expressed as the mean ± SD. All experiments were repeated at least three times, but less than ten times. Student's t-tests and post hoc Kruskal–Wallis tests were used to analyse the differences between the studied groups. Statistically significant differences (p b 0.05) compared to TNZ-EP are indicated by *.

163

3. Results 3.1. Characterisation of the anodic oxide coating morphologies The pretreated Ti–13Nb–13Zr alloy samples were subjected to anodising in solutions that contained different amounts of Ca(H2PO2)2 and H3PO4 at 200 and 400 V (Table 2). The surface morphologies of the resulting oxide coatings are presented in Fig. 1. An SEM image representing the surface of the pretreated Ti–13Nb–13Zr alloy sample (TNZ-EP) is not included. However, the surface morphology of the electropolished alloy was investigated earlier [16,33]. After anodising in bath I at a voltage of 200 V (TNZ-I-200), the surface appearance of the Ti alloy sample did not change markedly. The surface morphology of this sample resembled that of the electropolished sample. A similar effect was observed for the samples treated at a higher concentration of Ca(H 2 PO2 ) 2 (0.1 mol dm− 3 — TNZ-II-200), and with the addition of H 3 PO 4 (0.1 mol dm − 3 — TNZ-IV-200) no significant changes in the surface appearance were observed. Once the concentration of calcium hypophosphite in the anodisation bath had been increased to 1.0 mol dm − 3 while keeping the process voltage at 200 V (TNZ-III-200), the surface of the oxide layer became highly developed, with characteristic valleys dotting the coating surface. As expected, notable alterations in the surface morphology of the titanium alloy samples were observed after conducting the treatment at 400 V in baths I–IV. The oxide layers obtained during the course of these treatments were covered with numerous pores that ranged in size from approximately 2 (TNZ-I-400) to 10 μm (TNZ-III-400). Significant surface cracking was discovered for the TNZ-III-400 sample. The number of pores on the surface of this sample was found to be lower than that of the other samples treated at 400 V. Surface profiles of the oxide coatings obtained on the Ti–13Nb–13Zr alloy are shown in Fig. 2, while the Ra parameters calculated for each sample are presented in Table 2. The electropolished surface of the titanium alloy was found to be very smooth, as its Ra parameter was equal to 0.15 μm. The surface of the alloy was not altered significantly when the treatment was conducted at 200 V in baths I, II and IV. The Ra parameters of the samples treated under those conditions were determined to

Fig. 2. Surface profiles measured for the titanium alloy surfaces modified under various conditions.

164

M. Sowa et al. / Materials Science and Engineering C 49 (2015) 159–173

be 0.10, 0.09 and 0.13 μm, respectively. A slight increase in surface roughness (Ra = 0.51 μm) was observed for the TNZ-III-200 sample (Fig. 2). An increase in the Ra parameter was noted for each of the samples anodised at 400 V, regardless of the electrolytic bath chosen, compared with the samples treated at 200 V. The surface roughness increased with the concentration of the baths, with the Ra parameter equal to 0.14, 0.43, 0.91 and 1.50 μm for samples treated in baths I, II, IV and III, respectively. The surface topographies of the titanium alloy samples that were anodised in bath III at 200 and 400 V were analysed using AFM. The results of the investigations are presented in Fig. 3. AFM image of the electropolished sample (Fig. 3a) served as a reference. Its topographical features were discussed in one of our previous works [33]. The sample anodised at 200 V (Fig. 3b) showed the typical rough topography, similar to the one depicted in Fig. 1. The surface of the TNZ-III-400 sample (Fig. 3c) was dotted with pores and was characterised by a wavy topography.

3.2. Voltage–time response of the treatment Based on the preliminary characterisation, four samples were selected for further study: TNZ-II-400, TNZ-III-200, TNZ-III-400 and TNZ-IV400. The U = f(t) curves recorded during the anodic oxidation of those specimens are shown in Fig. 4. It is well known that the final voltage and the slope at which the voltage increases during anodising are tightly related to the properties of the obtained oxide films [17,18,34, 35]. The voltage–time progression registered during the first 30 s of anodising of the TNZ-III-200 sample (Fig. 4a) showed three distinct stages: (1), (2) and (5). Deteriorations from the initial voltage–time progression during stage (1) indicate the commencing of the oxide breakdown phenomenon. The voltage increments in those stages were as follows: 20.5 and 4.6 V s−1 for the regions (1) and (2), respectively. In stage (5), the current dropped to values that oscillated close to 0 A, and the voltage was maintained at 200 V. Another stage, (3), can be distinguished by analysing the curves recorded for the samples anodised at 400 V. Anodisation carried out in the same bath (III) proceeded similarly up to the voltage of 200 V (Fig. 4b). However, at a voltage of 240 V there was an additional stage (3) at which the rate of voltage increase changed, and the rise continued at a constant rate until it reached 400 V. The values of the voltage increments in stages (1), (2) and (3) were calculated to be 20.3, 4.2 and 1.0 V s−1, respectively. Yet another characteristic stage could be discerned for the samples anodised in baths II and IV at 400 V (Fig. 4c). The slopes of the stages (1)–(4) measured for the anodisation of the TNZ-II-400 sample were 20.5, 2.1, 0.9 and 1.8 V s− 1, respectively. During anodisation of the TNZ-IV-400 sample, the final voltage of the treatment was attained in the shortest time period. The voltage increments for this sample were 20.6, 2.7, 1.4 and 2.6 V s−1 for stages (1)–(4), respectively. 3.3. Composition of the anodic oxide coatings

Fig. 3. AFM images showing the surfaces of the titanium alloy samples subjected to electropolishing (a) and anodising at (b) 200 V and (c) 400 V in a 1.0 mol dm−3 Ca(H2PO2)2 solution.

The internal structure of the oxide coatings grown on the Ti alloy, their thickness and elemental composition were analysed using FESEM. The cross-sectional micrographs and the corresponding EDX spectra of the selected titanium alloy specimens are depicted in Fig. 5. The atomic ratios of Ca to P are also given. The results demonstrate that all of the Ti alloy samples were completely covered by relatively uniform oxide layers containing all of the alloying elements (Ti, Nb and Zr), as well as all the components of the electrolytic baths (Ca and P). The spectrum of the TNZ-III-400 sample revealed an additional signal for Na, which was regarded as contamination. The thickness of the oxide layers was found to be related to both the voltage of treatment and the composition of the electrolytic bath. For the sample anodised at 200 V in bath III, the thickness of the oxide layer ranged from 2.43 to 4.74 μm. The treatment conducted at 400 V in the same bath led to the development of an oxide layer that was 9.37–11.6 μm thick. Thinner oxide layers were obtained from the baths that were more dilute. The thickness of the coating on the TNZ-II-400 sample was determined to be 6.46–7.65 μm, while that of the TNZ-IV-400 sample ranged from 5.90 to 8.49 μm. A significant amount of closed pores could be observed within those layers. The relative amounts of Ca and P were found to be approximately the same for the specimens anodised at 400 V. A different amount of Ca/P (1.0) was determined for the TNZ-III-200 sample. The elemental composition of the coatings determined by XRF is summarised in Table 3. According to these results, oxygen was a dominant component of the coatings, suggesting that the alloying elements (Ti, Nb and Zr) were in the form of oxides. The relative amounts of calcium and phosphorus were analogous to those determined via EDX spectral analysis of the cross-sections of the coatings. The incorporation of Ca and P species into the oxide layer proceeded to a higher extent when the treatment was conducted at a higher concentration of Ca(H2PO2)2 and a higher voltage. The addition of phosphoric acid into the electrolyte led to only a minor increase of the amount of P in the coatings.

M. Sowa et al. / Materials Science and Engineering C 49 (2015) 159–173

165

Fig. 4. The voltage–time dependencies recorded during anodic oxidation of the titanium alloy samples under selected experimental conditions. Insets: the voltage–time progression of the (a) TNZ-III-200, (b) TNZ-III-400 and (c) TNZ-II-400 samples at shorter time scales.

The XRD measurements were conducted to determine the phase composition of the films produced on the Ti–13Nb–13Zr alloy. Fig. 6 shows the diffraction patterns of the selected Ti alloy specimens. From these results, it can be concluded that the oxide layers are primarily amorphous, with a small amount of crystalline phases present on the surface of the TNZ-II-400 (TiO2 — anatase) and TNZ-III-200 (Ti and TiO2 — rutile) samples. The chemical composition of the oxide film on the titanium alloy obtained by anodic oxidation at 400 V in a bath that contained 0.1 mol dm−3 Ca(H2PO2)2 and 0.1 mol dm−3 H3PO4 was investigated with XPS. The survey spectrum of the TNZ-IV-400 sample is shown in Fig. 7, and the deconvoluted spectra of P 2p, Ca 2p, Ti 2p, O 1s, Nb 3d and Zr 3d core excitations are presented in Fig. 8. Analysis of the results indicated the presence of: C, N, O, P, Ca, Ti, Zr and Nb. Carbon and nitrogen were regarded as contaminants derived from the adsorption of organic molecules from the air. The relatively large amount of oxygen suggested that the titanium, zirconium and niobium existed in the oxide forms. The peak found at a binding energy (BE) of 458.95 eV in the Ti 2p core excitation spectrum confirmed that the titanium was present as TiO2. In addition, the spectra corresponding to Nb 3d and Zr 3d showed that niobium existed as Nb2O5, due to the peak at a BE of 207.67 eV, and zirconium was present as ZrO2, indicated by the peak at 182.2 eV. The P 2p core excitation spectrum revealed that phosphorus existed in the form of phosphates. 3.4. Preliminary bioactivity studies of the anodic oxide coatings The bioactivity of the coatings prepared on Ti–13Nb–13Zr alloy was studied by immersion tests in protein-free SBF. SEM micrographs showing the results are presented in Fig. 9. Small crystallites of a newly formed phase were observed on the surface of the TNZ-II-400 sample after the first week of immersion (not shown). The crystallites became more abundant over time. After four weeks of immersion, the crystallites developed clusters (encircled area in Fig. 9). In the case of the sample anodised at 200 V in bath III, the surface morphology of the oxide layer was altered after the immersion tests. A crystalline phase precipitated on the borders of the valley structures covering the sample's surface. Crystallites were present on the surface of the TNZ-III-400 sample after one week of immersion in SBF. Although, the number of crystallites did not increase significantly after additional exposure, a globular structure was identified on the surface of this sample after four weeks of immersion. The analysis of the surface of the TNZ-IV-400 sample

revealed that the precipitation of the SBF components occurred after the first week of immersion. The crystallites were uniformly spread across the sample surface. Further exposure of the sample to SBF led to an increase in both the number and size of the crystalline structures. XRD experiments were conducted to identify the structures formed on the surface of the titanium alloy specimens immersed in SBF for four weeks. Fig. 10 shows a diffraction pattern recorded for the TNZ-III-200 sample. The compound that was precipitated on the surface of this sample during immersion in SBF was identified as tricalcium phosphate (Ca3(PO4)2 — β-TCP). A signal ascribed to crystalline titanium was also registered in the pattern. The precipitates observed on the surfaces of the other samples were too scarce to identify peaks corresponding to their crystalline phases (data not shown). 3.5. In vitro hBMSC cultures on the anodic oxide coatings Cell culture studies using hBMSCs were limited to the TNZ-III-400 and TNZ-IV-400 samples. These samples were chosen on the basis of high surface porosity (Fig. 1), roughness (Fig. 2) and significant Ca and P content (Table 3). The oxide coatings on those samples were relatively thick (Fig. 5), which has an impact on their corrosion resistance in physiological conditions [16,27–29]. The results of the SBF assay did not provide much meaningful data, upon which one could distinguish a sample more suitable for further biological investigations. The electropolished titanium alloy sample (TNZ-EP) was used as a control. The morphology of the hBMSCs cultured on the modified titanium alloy surfaces was investigated by SEM (Fig. 11). Cell morphology was similar on all examined surfaces. However, better surface coverage by the cells was observed for the two anodised titanium alloy specimens compared to the TNZ-EP sample. Fig. 12 shows the viability of hBMSCs cultured on the selected titanium alloy samples. The data were collected after 10 and 21 days of culture. Notably, cell viability on the TNZ-IV-400 sample was slightly higher compared to the control sample after 10 days, but it decreased to the same level as the control after 21 days. There was no significant effect of the TNZ-III-400 sample surface on hBMSC viability after 10 days of culture. Nevertheless, cell viability on this sample decreased greatly after 21 days of culture. The hBMSC cultures were tested for ALP activity after 10 days and for collagen production and ECM calcification after 21 days of culture (Fig. 13). The oxide coating produced on the TNZ-III-400 samples decreased the ALP activity of hBMSCs compared to the control. However,

166

M. Sowa et al. / Materials Science and Engineering C 49 (2015) 159–173

Ca/P 0.4

Ca/P 1.0

Ca/P 0.4

Ca/P 0.3

Fig. 5. Cross-sectional FESEM micrographs of the titanium alloy samples anodised under selected experimental conditions. The thicknesses of the oxide coatings and the EDX spectra, along with the Ca/P surface atomic ratios measured for the corresponding coatings, are shown.

collagen production on this surface was significantly increased by approximately 25-fold relative to the control, while ECM mineralisation was increased by approximately 60-fold. The TNZ-IV-400 samples did not affect ALP activity, collagen production or ECM mineralisation of the hBMSCs. The modified titanium alloy surfaces studied here were also analysed for the mRNA expression levels of various osteogenic differentiation markers after 10 days of hBMSC culture (Fig. 14). The ALP mRNA levels were higher for cells cultured on the modified surfaces of TNZ-IV400 and TNZ-III-400 compared to the control. The collagen type I mRNA levels of cells cultured on the TNZ-III-400 samples were significantly

Table 3 Chemical composition of the surfaces of the titanium alloy specimens (at.%), measured by XRF. Sample

O

P

Ca

Ti

Nb

Zr

TNZ-II-400 TNZ-III-200 TNZ-III-400 TNZ-IV-400

47.5 41.2 36.7 41.5

12.3 12.1 18.3 13.2

3.6 5.5 17.9 5.4

25.9 29.4 16.3 27.6

4.9 5.4 5.0 5.7

5.9 6.4 5.8 6.7

higher compared to TNZ-IV-400 and the control. The mRNA expression levels of osteocalcin were comparable for all of the samples studied. 4. Discussion Anodic oxidation carried out beyond the voltage of oxide breakdown, known as plasma electrolytic oxidation, leads to the development of surface structures with unique characteristics [14]. Aside from the obvious modifications in the surface morphology of the treated workpiece, this method allows for the introduction of various bioactive elements in the form of different chemical compounds within the boundary of the anodic oxide film, which comes into contact with physiological environment upon implantation. This oxide layer, which protects the underlying metal or metal alloy from corrosion, gives rise to a series of biological responses from the host cells. The nature of these interactions determines whether the implant will be successfully incorporated into the living organism and perform its intended function [36]. The effect of ceramic surfaces containing calcium and phosphorus on the effectiveness of bone growth is well established [37–40], and PEO has been reported for the successful development of such coatings [15,16,41–44]. Additionally, the three-dimensional surface with large

M. Sowa et al. / Materials Science and Engineering C 49 (2015) 159–173

Counts

Rutile

167

Anatase

Titanium

TNZ-II-400

100 50 0 10

20

30

50

40

60

70

80

2θ, ° Counts

200

TNZ-III-200

100 0 10

20

30

40

50

60

70

80

2θ, ° Counts

150

TNZ-III-400

100 50 0 10

20

30

40

50

60

70

80

Counts

2θ, °

TNZ-IV-400

100 50 0 10

20

30

50

40

60

70

80

2θ, ° Fig. 6. TL-XRD patterns of the oxide coatings prepared on the titanium alloy samples anodised under selected experimental conditions. Relevant phases are designated in the spectra.

open pores obtained by this method mimics the bone structure and leads to a high surface area contact, which is necessary to achieve good bonding strength between an implant and bone. Nevertheless, such architectures are also attractive to bacteria, that form biofilms on the surface of the implant, leading to infections and eventually implant rejection [45]. Anodic oxide films obtained by PEO surface treatment of Ti–13Nb– 13Zr in the electrolyte solutions containing Ca(H2PO2)2 and H3PO4 at voltages of 200 and 400 V were characterised in this study. The oxide coatings obtained at lower voltages were relatively smooth, with the

Fig. 7. XPS survey spectrum of a titanium alloy sample that was anodised at 400 V in a solution that contained 0.1 mol dm−3 of Ca(H2PO2)2 and 0.1 mol dm−3 of H3PO4 (TNZIV-400 sample).

exception of that obtained in the solution that contained the highest concentration of dissolved species (bath III). This result suggests that the dielectric breakdown of the oxide layer did not occur in the case of these samples. Breakdown was observed in the case of the TNZ-III-200 sample, because the resistivity depends on the concentration of electrolytes. In this case, the potential drop across the oxide layer was sufficiently high to initiate the dissolution of the oxide. The morphology of the samples anodised above the potential of the dielectric breakdown of the oxide layer resembled that reported for titanium and its alloys [15,42–44]. Of the samples that were anodised at 400 V, only TNZ-III400 was noteworthy. Fewer pores were present on the surface of this sample, and the overall topography was wavy, as shown by the AFM investigations (Fig. 3b). A study by Krupa et al. [46,47] on PEO of titanium reported similar surface characteristics. In order to explain the differences in the surface morphology of the titanium alloy samples anodised under various conditions, one needs to examine the voltage progression curves (Fig. 4). Different stages (1)–(5) can be distinguished during the PEO process. Voltage–time curves recorded for the samples anodised conventionally (not by PEO) have shown only one region (1) of linear voltage progression. In contrast, the curve corresponding to the TNZIII-200 sample (Fig. 4a) showed another stage (2) in which the rate of the voltage rise decreased. This second stage was associated with the vigorous evolution of bubbles and partial dissolution of the oxide coating, which was reflected in the sample surface morphology (Figs. 1 and 3a). When the voltage of anodisation rose above 200 V, as for the TNZ-III-400 sample, the next region (3) emerged (Fig. 4b), and observable micro-discharges appeared on the surface during this period. High temperature micro-arcs [34,36] may have led to the fusion of the oxide surface layer with the electrolyte components and, thus, to the development of a wavy topography with visible craters dotting the surface. The appearance of another region (4) in the voltage–time curve for the samples TNZ-II-400 (Fig. 4c) and TNZ-IV-400 was most likely due to oxide

168

M. Sowa et al. / Materials Science and Engineering C 49 (2015) 159–173

450

450 400

P 2p

350

350

300

300

250

250

c/s

c/s

400

200

200

150

150

100

100

50

50 138

136

134 132 Binding Energy (eV)

130

128

138

Ti 2p

800

Ca 2p

136

C

134 132 Binding Energy (eV)

130

128

O 1s

2500

700 2000

c/s

c/s

600 1500

500 1000

400

300

500

200 468

466

464 462 460 Binding Energy (eV)

458

534

456

532 530 Binding Energy (eV)

528

526

350

250 240

Nb 3d

230

Zr 3d

300

220 250 c/s

c/s

210 200

200

190 180

150

170 160 100

150 212

211

210

209 208 207 Binding Energy (eV)

206

205

195

190 185 Binding Energy (eV)

180

Fig. 8. Deconvoluted spectra of the P 2p, Ca 2p, Ti 2p, O 1s, Nb 3d and Zr 3d core excitations determined for the titanium alloy sample that was anodised at 400 V in a solution that contained 0.1 mol dm−3 of Ca(H2PO2)2 and 0.1 mol dm−3 of H3PO4 (TNZ-IV-400 sample).

recrystallisation and the appearance of defects in the crystal lattice [17]. XRD patterns of the sample (Fig. 6) showed that the crystalline titania phase (anatase) was detected in the TNZ-II-400 sample. However, there were no distinct signals corresponding to this phase in the pattern recorded for the TNZ-IV-400 sample. This result may be explained by the short duration of stage (4) during the PEO treatment of this sample, where only negligible recrystallisation could occur. The thickness of the anodic coatings was controlled by the process voltage and the concentration of the electrolyte components (Fig. 5). The effect of voltage is well understood [14,42,44,48]. The increase in voltage during the treatment occurs due to the thickening of the dielectric oxide layer, whose high resistivity accounts for much of the potential. The role of the

concentration of electrolyte can be attributed to the duration of the various anodisation stages (1)–(5) [35,49,50]. The thickness of the coatings seemed to be related to the time needed to reach the final treatment voltage (stage — 5), with thicker coatings being obtained when this time was prolonged. This effect can be explained by the different amounts of charge delivered during the processing because once the final anodisation voltage is reached, the current density was allowed to drop to retain that voltage value. Plasma- and thermo-chemical reactions that occur during PEO treatment of titanium in solutions containing Ca and P can lead to the formation of the following ions: Ti4 +, Ca2 +, PO34 −, HPO24 − and OH− due to ionisation of substrate and electrolyte components [51].

M. Sowa et al. / Materials Science and Engineering C 49 (2015) 159–173

169

Fig. 9. SEM micrographs of titanium alloy samples anodised under selected experimental conditions after IV weeks of immersion in SBF. The encircled area shows a cluster of crystallites. The globular crystalline structure is indicated by an arrow.

Some of the reactions in which these ions can participate are listed below [15,51]: −

þ

2þ

þ OH þ H 2 O→ CaO þ H3 O

2þ

þ Ti

Ca

4þ

Ca

3 Ca

2þ

þ

þ 3 OH þ 3H 2 O→ CaTiO3 þ 3H 3 O

ð3Þ

2−

þ 2 PO4 → Ca3 ð PO4 Þ2

2þ

10 Ca

−

ð2Þ

ð4Þ þ

2−

þ 6 PO4 þ 4H 2 O→ Ca10 ð PO4 Þ6 ð OHÞ2 þ 2H 3 O :

ð5Þ

Similar reactions can also occur when the oxide layer is exposed to SBF. The XRD experiments did not show any crystalline calciumcontaining species. However, Ca and P were successfully introduced into the oxide layers, as determined by XRF and EDX analysis of the coatings (Fig. 5 and Table 3), leading to the conclusion that Ca and P compounds were amorphous. As shown by XRF, the relative amounts of calcium and phosphorus on the surface of the investigated samples indicated that the Ca/P ratio was highest for TNZ-III-400. This result

can be explained by the considerably longer time of the stage (3) for this sample, in which oxide fusion with the electrolyte components may have occurred on the coating surface. The Ca/P atomic ratios determined via EDX of the cross-sections of the oxide coatings were not exactly the same as those obtained from XRF. This difference indicates that Ca and P were not uniformly distributed throughout the oxide films, because XRF probed only the surface oxide, while the EDX analysis was performed on the coating's cross-section. Ca and P were relatively more abundant within the oxide layer with increasing voltage and concentration of Ca(H2PO2)2 and H3PO4 (Table 3). This effect was reported earlier by Han et al. on pure titanium [44]. However, research conducted by Yao et al. [43] show that higher process voltage during PEO may lead to a decrease in relative amount of Ca and P in the oxide layer, when additional ions are present in the solution (in this case Si species). This indicates how important the composition of the electrolytic bath is. The bioactivity of the oxide coatings was assessed by the SBF assay, but it was determined to be not high, due to small amounts of apatite precipitates formed on the surface of the modified Ti alloy (Figs. 9 and 10). However, the reliability of bioactivity testing by this method has

β-TCP

600

Titanium

Counts

400

200

0 10

20

30

40

50

60

70

2θ, ° Fig. 10. Exemplary TL-XRD pattern of a titanium alloy sample that was anodised at 200 V in a 1.0 mol dm−3 Ca(H2PO2)2 solution (TNZ-III-200 sample) after IV weeks of immersion in SBF. Relevant phases are designated in the spectrum.

170

M. Sowa et al. / Materials Science and Engineering C 49 (2015) 159–173

a)

b)

c)

d)

e)

f)

Fig. 11. SEM micrographs of cells after 5 days of culture on the surfaces of (a, b) an electropolished Ti–13Nb–13Zr alloy sample and (c, d) a titanium alloy that was anodised at 400 V in a solution that contained 1.0 mol dm−3 of Ca(H2PO2)2 or (e, f) 0.1 mol dm−3 of Ca(H2PO2)2 and 0.1 mol dm−3 of H3PO4. Images are shown at various magnifications.

been questioned [52,53]. Thus, the results obtained from the SBF tests were not taken into consideration for the choice of samples for biological experiments. In this study, the anodic oxide coatings obtained on Ti– 13Nb–13Zr alloy led to scarce amounts of crystals, which were determined to be β-TCP crystals in the case of the TNZ-III-200 sample. After 4 weeks of immersion in SBF, the surface of the TNZ-III-400 sample was observed to develop a minute amount of globular crystals, similar to what has been reported by others [15,42,46,47]. Currently, vanadium-free titanium alloys containing elements such as Zr, Nb, Ta or Mo (widely accepted to be non-toxic) are characterised by having the highest biocompatibility among the metallic biomaterials used in clinical applications. Furthermore, their hydroxides have been confirmed to have a major role in the hydroxyapatite precipitation process [1,54–56]. In the present work, viability studies of hBMSCs during short- (10 days) and long-term (21 days) culture revealed that two

different surface modifications of the titanium alloy had distinct effects on the cell growth (Fig. 12). An initial increase in cell viability was noted on the TNZ-IV-400 sample, but there was a significant decrease in cell viability after 21 days of culture. A significant decrease in cell viability was also noted in the case of the TNZ-III-400 sample. According to the results for the gene expression of ALP, ColI and OC (Fig. 14), the hBMSCs were in the advanced stages of osteogenesis after 10 days of culture. Low levels of ALP activity in the cells (Fig. 13) indicated that the protein production rate was not yet high at day 10 of culture, although the ALP gene expression levels were significantly increased. This result may also suggest that the ALP activity should be determined at a later time point. In contrast, very high levels of collagen production and ECM calcification were noted for the TNZ-III-400 sample after 21 days of culture. Therefore, it is probable that the hBMSCs had decreased their rate of proliferation due to differentiation toward osteoblasts [32,

M. Sowa et al. / Materials Science and Engineering C 49 (2015) 159–173

A L P m R N A g e n e e x p r e s s io n ,

2 .5

f o l d c h a n g e r e la t i v e t o c o n t r o l ( 1 . 0 )

a)

171

2 .0

1 .5

1 .0

0 .5

0 .0

100

75

c)

50

25

0

C o l l a g e n p r o d u c t io n ,

b)

% c h a n g e r e la t i v e t o c o n t r o l ( 1 0 0 % )

T N Z-E P

M in e r a lis a tio n ,

% c h a n g e r e la t i v e t o c o n t r o l ( 1 0 0 % )

T N Z-I V -4 0 0

*

4500 3500 2500 1500 1000

500

T N Z-I V -4 0 0

6 5

4 3

2 1 0

T N Z-E P

T N Z-I I I -4 0 0

T N Z-I V -4 0 0

T N Z-E P

T N Z-I I I -4 0 0

T N Z-I V -4 0 0

1 .5

1 .0

0 .5

0 .0

Fig. 14. Real-time PCR analysis of gene expression levels of hMSC cultures after 10 days on the electropolished (TNZ-EP) and anodised (TNZ-III-400 and TNZ-IV-400) Ti–13Nb–13Zr alloy samples. The results are shown as the mean value ± SD of the fold change relative to TNZ-EP (control sample). Differences that are statistically significant (p b 0.05) are designated with *.

150 100 50 0

T N Z-E P

c)

T N Z-I I I -4 0 0

f o l d c h a n g e r e la t i v e t o c o n t r o l ( 1 . 0 )

125

T N Z-I I I -4 0 0

*

7

f o l d c h a n g e r e la t i v e t o c o n t r o l ( 1 . 0 )

150

O s t e o c a lc in m R N A g e n e e x p r e s s i o n ,

A L P a c tiv ity ,

% c h a n g e r e la t i v e t o c o n t r o l ( 1 0 0 % )

a)

b)

C o ll a g e n I m R N A g e n e e x p e r s s i o n ,

T N Z-E P

Fig. 12. Cell viability of hMSC cultures after 10- and 21-days of culture on the electropolished (TNZ-EP) and anodised (TNZ-III-400 and TNZ-IV-400 samples) Ti– 13Nb–13Zr alloy samples. The results are shown as the mean value ± SD of the percent change relative to TNZ-EP (control sample).

T N Z-I I I -4 0 0

T N Z-I V -4 0 0

*

10000 9000 8000 7000 6000 3000 2500 200 150 100 50 0

T N Z-E P

T N Z-I I I -4 0 0

T N Z-I V -4 0 0

Fig. 13. ALP activity (a) and collagen production (b) of hMSC cultured for 10 days and extracellular matrix mineralization (c) of hMSCs cultured for 21 days on electropolished (TNZ-EP) and anodised (TNZ-III-400 and TNZ-IV-400) Ti–13Nb–13Zr alloy samples. The results are shown as the mean value ± SD of the percent change relative to TNZ-EP (control sample). Differences that are statistically significant (p b 0.05) are designated with *.

57]. The biological response of cells to Ti-based biomaterials is dependent on the surface features of the material, such as its structure, morphology and chemical composition [58–60]. Kim et al. proposed a reciprocal relationship between surface roughness and MG63 cell viability [57]. In our samples, this effect could not be easily confirmed, which may suggest an additional effect of chemical composition of the surface. The analysis of the SEM micrographs of the hBMSCs on the surfaces of the modified titanium alloy (Fig. 11) indicated that the cells were highly flattened on all the surfaces, and numerous filopodia were visible in the case of the anodised samples. Therefore, the surface topography had a significant effect on the morphology of the cells and could have affected their viability and differentiation [26,61]. As mentioned earlier, the amount of calcium and phosphorus incorporated into the oxide layer was different and may have had a significant effect on cell morphology, adhesion and differentiation [39,42]. This effect was most likely visible here because the collagen production and ECM mineralisation levels of hBMSCs were markedly increased when cultured on the surface of the TNZ-III-400 sample, which had the highest Ca/P surface atomic ratio among the investigated specimens.

172

M. Sowa et al. / Materials Science and Engineering C 49 (2015) 159–173

5. Conclusions Our study has shown that the proposed method of surface modification of a vanadium-free candidate material for surgical biomaterials Ti– 13Nb–13Zr alloy provides a material surface with bioactive properties. Furthermore, these surfaces were capable of inducing the differentiation of hBMSCs toward osteoblasts. Surface modification was performed using plasma electrolytic oxidation at various voltages in a number of different electrolytic baths containing calcium and phosphorus. This treatment led to the successful incorporation of Ca and P into the oxide coatings, which had a beneficial effect on the differentiation of hBMSCs into osteoblasts. An additional factor that may have affected the bioactivity of the oxide layer was its surface topography, which resembled the structure of porous bone. The rougher surface was observed to promote cell spreading, with numerous cytoplasmic cell projections being visible in the adhered cells. Acknowledgements The authors would like to thank Prof. T. Niedźwiedzki, the head of the Polytrauma, Orthopaedic and Neuroorthopaedic Department of the Rydygier Hospital in Krakow, for procuring human bone marrow samples, and the Laboratory of Scanning Electron Microscopy for Biological and Geological Sciences of the Institute of Zoology at Jagiellonian University, which is supported in part by the Foundation for Polish Science Subin 94 Programme, for the use of the JEOL JSM 5410 scanning electron microscope. Additionally, J. Michalska, O. Woźnicka, J. Szade and A. Winiarski are gratefully acknowledged for their contributions. This work was supported by the Polish Ministry of Science and Education under research project nos. IP 2012 0459 72 (WS) and 2011/01/ B/NZ4/00664 (AMO). References [1] W. Simka, A. Iwaniak, G. Nawrat, A. Maciej, J. Michalska, K. Radwański, J. Gazdowicz, Modification of titanium oxide layer by calcium and phosphorus, Electrochim. Acta 54 (2009) 6983–6988. [2] S. Yu, Z. Yu, G. Wang, J. Han, X. Ma, M.S. Dargusch, Biocompatibility and osteoconduction of active porous calcium-phosphate films on a novel Ti–3Zr– 2Sn–3Mo–25Nb biomedical alloy, Colloids Surf. B 85 (2) (2011) 103–115. [3] X. Liu, P.K. Chu, C. Ding, Surface modification of titanium, titanium alloys, and related materials for biomedical applications, Mater. Sci. Eng. R 47 (3–4) (2004) 49–121. [4] D.H. Kohn, Metals in medical applications, Curr. Opin. Solid State Mater. Sci. 3 (1998) 309–316. [5] S. Ankem, C.A. Greene, Recent developments in microstructure: property relationships of beta titanium alloys, Mater. Sci. Eng. A 263 (1999) 127–131. [6] E. Kobayashi, T. Wang, H. Doi, Mechanical properties and corrosion resistance of Ti– 6Al–7Nb alloy dental castings, J. Mater. Sci. Mater. Med. 9 (1998) 567–574. [7] J.J. Ramsden, D.M. Allen, D.J. Stephenson, J.R. Alcock, G.N. Peggs, G. Fuller, G. Goch, The design and manufacture of biomedical surfaces, Ann. CIRP 56 (2007) 687–711. [8] Z.X. Chen, W.X. Wang, Y. Takao, Y. Matsubara, L.M. Ren, Microstructure and shear fracture characteristics of porous anodic TiO2 layer before and after hot water treatment, Appl. Surf. Sci. 257 (2011) 7254–7262. [9] J.J. Jacobs, A.K. Skipor, J. Black, R. Urban, J.O. Galante, Release and excretion of metal in patients who have a total hip-replacement component made of titanium-base alloy, J. Bone Joint Surg. 73 (1991) 1475–1486. [10] C.H. Kua, D.P. Piolettia, M. Browne, P.J. Gregson, Effect of different Ti–6Al–4V surface treatments on osteoblasts behavior, Biomaterials 23 (2002) 1447–1454. [11] Y. Tsutsumi, M. Niinomi, M. Nakai, H. Tsutsumi, H. Doi, N. Nomura, T. Hanawa, Micro-arc oxidation treatment to improve the hard-tissue compatibility of Ti–29Nb–13Ta–4.6Zr alloy, Appl. Surf. Sci. 262 (2012) 34–38. [12] R. Banerjee, S. Nag, J. Stechschulte, H.L. Fraser, Strengthening mechanisms in Ti–Nb– Zr–Ta and Ti–Mo–Zr–Fe orthopaedic alloys, Biomaterials 25 (2004) 3413–3419. [13] W.S. Lee, C.F. Lin, T.H. Chen, H.H. Hwang, Effects of strain rate and temperature on mechanical behaviour of Ti–15Mo–5Zr–3Al alloy, J. Mech. Behav. Biomed. Mater. 1 (2008) 336–344. [14] A.L. Yerokhin, X. Nie, A. Leyland, A. Matthews, S.J. Dowey, Plasma electrolysis for surface engineering, Surf. Coat. Technol. 122 (1999) 73–93. [15] S. Durdu, Ö.F. Deniz, I. Kutbay, M. Usta, Characterization and formation of hydroxyapatite on Ti6Al4V coated by plasma electrolytic oxidation, J. Alloys Compd. 551 (2013) 422–429. [16] W. Simka, A. Krząkała, M. Masełbas, G. Dercz, J. Szade, A. Winiarski, J. Michalska, Formation of bioactive coatings on Ti–13Nb–13Zr alloy for hard tissue implants, RSC Adv. 3 (2013) 11195–11204.

[17] L.O. Snizhko, A.L. Yerokhin, A. Pilkington, N.L. Gurevina, D.O. Misnyankin, Anodic processes in plasma electrolytic oxidation of aluminium in alkaline solutions, Electrochim. Acta 49 (2004) 2085–2095. [18] E. Matykina, A. Berkani, P. Skeldon, G.E. Thompson, Real-time imaging of coating growth during plasma electrolytic oxidation of titanium, Electrochim. Acta 53 (2007) 1987–1994. [19] R.J. Miron, Y.F. Zhang, Osteoinduction: a review of old concepts with new standards, J. Dent. Res. 8 (2012) 736–744. [20] B.V. Afanasyev, E.E. Elstner, A.R. Zander, A.J. Friedenstein, founder of the mesenchymal stem cell concept, Cell. Ther. Transplant. 1 (3) (2009) 35–38. [21] Z. Liu, Y. Zhuge, O.C. Velazquez, Trafficking and differentiation of mesenchymal stem cells, J. Cell. Biochem. 106 (6) (2009) 984–991. [22] H. Li, K.A. Khor, P. Cheang, Impact formation and microstructure characterization of thermal sprayed hydroxyapatite/titania composite coatings, Biomaterials 24 (6) (2003) 949–957. [23] R. Nishimura, K. Hata, F. Ikeda, F. Ichida, A. Shimoyama, T. Matsubara, M. Wada, K. Amano, T. Yoneda, Signal transduction and transcriptional regulation during mesenchymal cell differentiation, J. Bone Miner. Metab. 26 (3) (2008) 203–212. [24] K.J.L. Burg, S. Porter, J.F. Kellam, Biomaterial developments for bone tissue engineering, Biomaterials 21 (23) (2000) 2347–2359. [25] F. Di Palma, A. Guignandon, A. Chamson, M. Lafage-Proust, N. Laroche, S. Peyroche, L. Vico, A. Rattner, Modulation of the responses of human osteoblast-like cells to physiologic mechanical strains by biomaterial surfaces, Biomaterials 26 (20) (2005) 4249–4257. [26] R.A. Surmenev, M.A. Surmeneva, A.A. Ivanova, Significance of calcium phosphate coatings for the enhancement of new bone osteogenesis — a review, Acta Biomater. 10 (2014) 557–579. [27] W. Simka, R.P. Socha, G. Dercz, J. Michalska, A. Maciej, A. Krza̧kała, Anodic oxidation of Ti–13Nb–13Zr alloy in silicate solutions, Appl. Surf. Sci. 279 (2013) 317–323. [28] L.T. Duarte, S.R. Biaggio, R.C. Rocha-Filho, N. Bocchi, Preparation and characterization of biomimetically and electrochemically deposited hydroxyapatite coatings on micro-arc oxidized Ti–13Nb–13Zr, J. Mater. Sci. Mater. Med. 22 (7) (2011) 1663–1670. [29] L.T. Duarte, S.R. Biaggio, R.C. Rocha-Filho, N. Bocchi, Surface characterization of oxides grown on the Ti–13Nb–13Zr alloy and their corrosion protection, Corros. Sci. 72 (2013) 35–40. [30] EN ISO 4287, Standard: Geometrical Product Specification (GPS) – Surface Texture: Profile Methods – Terms, Definitions and Surface Texture Parameters, 1997. [31] T. Kokubo, Apatite formation on surfaces of ceramics, metals and polymers in body environment, Acta Mater. 46 (1998) 2519–2527. [32] E. Pamula, J. Kokoszka, K. Cholewa-Kowalska, M. Laczka, L. Kantor, L. Niedzwiecki, G.C. Reilly, J. Filipowska, W. Madej, M. Kolodziejczyk, G. Tylko, A.M. Osyczka, Degradation, bioactivity, and osteogenic potential of composites made of PLGA and two different sol–gel bioactive glasses, Ann. Biomed. Eng. 39 (8) (2011) 2114–2129. [33] W. Simka, M. Mosiałek, G. Nawrat, P. Nowak, J. Żak, J. Szade, A. Winiarski, A. Maciej, L. Szyk-Warszyńska, Electrochemical polishing of Ti–13Nb–13Zr alloy, Surf. Coat. Technol. 213 (2012) 239–246. [34] R.O. Hussein, D.O. Northwood, X. Nie, Coating growth behavior during the plasma electrolytic oxidation process, J. Vac. Sci. Technol. A 28 (2010) 766–773. [35] M. Petković, S. Stojadinović, R. Vasilić, Lj. Zeković, Characterization of oxide coatings formed on tantalum by plasma electrolytic oxidation in 12-tungstosilicic acid, Appl. Surf. Sci. 257 (2011) 10590–10594. [36] M. Geetha, A.K. Singh, R. Asokamani, A.K. Gogia, Ti based biomaterials, the ultimate choice for orthopaedic implants — a review, Prog. Mater. Sci. 54 (2009) 397–425. [37] Y.-T. Sul, E.-S. Byon, Y. Jeong, Biomechanical measurements of calcium-incorporated oxidized implants in rabbit bone: effect of calcium surface chemistry of a novel implant, Clin. Implant. Dent. Relat. Res. 6 (2) (2004) 101–110. [38] Y.-T. Sul, C.B. Johansson, Y. Kang, D.G. Jeon, T. Albrektsson, Bone reactions to oxidized titanium implants with electrochemical anion sulphuric acid and phosphoric acid incorporation, Clin. Implant. Dent. Relat. Res. 4 (2) (2002) 78–87. [39] S.N. Nayab, F.H. Jones, I. Olsen, Effects of calcium ion implantation on human bone cell interaction with titanium, Biomaterials 26 (23) (2005) 4717–4727. [40] D. Krupa, J. Baszkiewicz, J.A. Kozubowski, A. Barcz, J.W. Sobczak, A. Biliński, M.D. Lewandowska-Szumieł, B. Rajchel, Effect of calcium-ion implantation on the corrosion resistance and biocompatibility of titanium, Biomaterials 22 (15) (2001) 2139–2151. [41] S.R. Paital, N.B. Dahotre, Calcium phosphate coatings for bio-implant applications: materials, performance factors, and methodologies, Mater. Sci. Eng. R 6 (2009) 1–70. [42] L.-H. Li, Y.-M. Kong, H.-W. Kim, Y.-W. Kim, H.-E. Kim, S.-J. Heo, J.Y. Koak, Improved biological performance of Ti implants due to surface modification by micro-arc oxidation, Biomaterials 25 (2004) 2867–2875. [43] Z. Yao, L. Li, Z. Jiang, Adjustment of the ratio of Ca/P in the ceramic coating on Mg alloy by plasma electrolytic oxidation, Appl. Surf. Sci. 255 (2009) 6724–6728. [44] Y. Han, S.-H. Hong, K. Xu, Structure and in vitro bioactivity of titania-based films by micro-arc oxidation, Surf. Coat. Technol. 168 (2003) 249–258. [45] M. Wilson, Bacterial biofilms and human disease, Sci. Prog. 84 (3) (2001) 235–254. [46] D. Krupa, J. Baszkiewicz, J. Zdunek, J. Smolik, Z. Słomka, J.W. Sobczak, Characterization of the surface layers formed on titanium by plasma electrolytic oxidation, Surf. Coat. Technol. 205 (2010) 1743–1749. [47] D. Krupa, J. Baszkiewicz, J. Zdunek, J.W. Sobczak, W. Lisowski, J. Smolik, Z. Słomka, Effect of plasma electrolytic oxidation in the solutions containing Ca, P, Si, Na on the properties of titanium, J. Biomed. Mater. Res. B 100B (2012) 2156–2166. [48] A.L. Yerokhin, X. Nie, A. Leyland, A. Matthews, Characterisation of oxide films produced by plasma electrolytic oxidation of a Ti–6Al–4V alloy, Surf. Coat. Technol. 130 (2000) 195–206.

M. Sowa et al. / Materials Science and Engineering C 49 (2015) 159–173 [49] X. Nie, A. Leyland, H.W. Song, A.L. Yerokhin, S.J. Dowey, A. Matthews, Thickness effects on the mechanical properties of micro-arc discharge oxide coatings on aluminium alloys, Surf. Coat. Technol. 116–119 (1999) 1055–1066. [50] G. Sundararajan, L. Rama Krishna, Mechanisms underlying the formation of thick alumina coatings through the MAO coating technology, Surf. Coat. Technol. 167 (2003) 269–277. [51] J. Sun, Y. Han, X. Huang, Hydroxyapatite coatings prepared by micro-arc oxidation in Ca- and P-containing electrolyte, Surf. Coat. Technol. 201 (2007) 5655–5658. [52] T. Kokubo, H. Takadama, How useful is SBF in predicting in vivo bone bioactivity? Biomaterials 27 (2006) 2907–2915. [53] M. Bohner, J. Lemaitre, Can bioactivity be tested in vitro with SBF solution? Biomaterials 30 (2009) 2175–2179. [54] T. Kokubo, H. Kim, M. Kawashita, Novel bioactive materials with different mechanical properties, Biomaterials 24 (13) (2003) 2161–2175. [55] T.A.G. Donato, L.H. de Almeida, R.A. Nogueira, T.C. Niemeyer, C.R. Grandini, R. Caram, S.G. Schneider, A.R. Santos Jr., Cytotoxicity study of some Ti alloys used as biomaterial, Mater. Sci. Eng. C 29 (4) (2009) 1365–1369.

173

[56] E. Eisenbarth, D. Velten, M. Müller, R. Thull, J. Breme, Biocompatibility of β-stabilizing elements of titanium alloys, Biomaterials 25 (2004) 5705–5713. [57] M. Kim, C. Kim, Y. Lim, S. Heo, Microrough titanium surface affects biologic response in MG63 osteoblast-like cells, J. Biomed. Mater. Res. A 79 (4) (2006) 1023–1032. [58] L.F. Cooper, A role for surface topography in creating and maintaining bone at titanium endosseous implants, J. Prosthet. Dent. 84 (5) (2000) 522–534. [59] L. Le Guehennec, M. Lopez-Heredia, B. Enkel, P. Weiss, Y. Amouriq, P. Layrolle, Osteoblastic cell behaviour on different titanium implant surfaces, Acta Biomater. 4 (3) (2008) 535–543. [60] K. Anselme, M. Bigerelle, Topography effects of pure titanium substrates on human osteoblast long-term adhesion, Acta Biomater. 1 (2) (2005) 211–222. [61] X. Zhu, J. Chen, L. Scheideler, R. Reichl, J. Geis-Gerstorfer, Effects of topography and composition of titanium surface oxides on osteoblast responses, Biomaterials 25 (18) (2004) 4087–4103.