MOCVD of Noble Metal Film Materials for Medical Implants: Microstructure and Biocompatibility of Ir and Au/Ir Coatings on TiNi

Noble metals such as Ir, Pt, Au are promising as coatings for metal medical implants to improve biocompatibility and corrosion resistance. Moreover, these coatings can be used as a basis for the further formation of bimetallic hetero-structures with enhanced antibacterial properties. In this work, we develop an approach to obtain such coatings by metal-organic chemical vapor deposition (MOCVD). We have been focused on the formation of Ir coating with developed morphology and subsequent discrete Au coating onto the titanium nickelide (TiNi) implant material. Iridium was deposited in an oxidizing atmosphere from the volatile precursor [Ir(cod)(acac)] (cod = cyclooctadiene-1,5, acac = acetylacetonate-anion). The effects of the deposition temperature (290–350 °C) and amount of introduced oxygen on the composition (Ir, Ir + IrO2) and microstructure of the samples were studied. Hetero-metallic Au/Ir coatings were obtained using [(CH3)2Au(thd)] precursor (thd = dpm = dipivaloylmethanate-anion) at a deposition temperature of 240 °C in the presence of oxygen. To assess the biocompatibility, the toxicity of Ir/TiNi, Au/Ir/TiNi, and uncoated TiNi in relation to human embryonic stem cell line Man-1 was examined after 1, 3, and 5 days of incubation. The results obtained were explained based on the coating microstructures.


Introduction
Nowadays titanium nickelide (TiNi) alloy is widely used for the manufacturing of implants and other devices for orthopedics and reconstructive surgery [1,2]. The main reasons for medical application of TiNi are its high biological compatibility and unique properties such as shape memory effect, superelasticity, and high damping capacity. Moreover, the Young's modulus of TiNi is only three times the elastic modulus of human bone, which is the closest match among the non-biodegradable materials. Such biomechanical properties allow minimizing undesirable reactions caused by high rigidity of the inserted structure, for example necrosis under pressure and slow healing. Therefore, TiNi is recognized as an ideal material for a number of load-devices, for example, compression devices The product was purified by zone sublimation (140 • C, 5 × 10 −2 Torr) with 80% yield. Elemental analysis (CARLO-ERBA-11008, wt.%): for C 13 H 19 O 2 Ir found C, 39.3; H, 4.9, calculated C, 39.1; H, 4.8. Melting point: 155 • C. Saturated vapor pressure is given by the equation ln(p, atm) = 23.03 − 12,817/T (K) in the temperature range 100-150 • C [24].
Iridium coatings were deposited at fixed parameters of mass transfer: source temperature (110 °C, partial pressure of precursor vapor = 0.02 Torr) and carrier gas flow rate (argon, 17 mL/min). The deposition temperature and the flow rate of the reagent gas (oxygen) were varied in the ranges of 290-350 °C and 33-133 mL/min, respectively. Gold was deposited at an evaporator temperature of 70 °C (partial vapor pressure of the precursor = 0.3 Torr), a deposition temperature of 240 °C, argon and oxygen flow rates of 17 mL/min. Growth rates (nm/min) were calculated as the quotients of the coating thicknesses and the deposition times.
Iridium coatings were deposited at fixed parameters of mass transfer: source temperature (110 • C, partial pressure of precursor vapor = 0.02 Torr) and carrier gas flow rate (argon, 17 mL/min). The deposition temperature and the flow rate of the reagent gas (oxygen) were varied in the ranges of 290-350 • C and 33-133 mL/min, respectively. Gold was deposited at an evaporator temperature of 70 • C (partial vapor pressure of the precursor = 0.3 Torr), a deposition temperature of 240 • C, argon and oxygen flow rates of 17 mL/min. Growth rates (nm/min) were calculated as the quotients of the coating thicknesses and the deposition times.

Coating Characterization
The phase composition of the coatings was determined by X-ray diffraction (XRD, Shimadzu XRD-7000 diffractometer, OneSight linear detector, Shimadzu, Kyoto, Japan). The diffraction patterns were recorded with a step of 2θ = 0.0143 • and 5 s accumulation per a point on CuKα radiation (Ni filter, range 2θ = 10-65 • ) or on CoKα radiation (Fe filter, range 2θ = 10-80 • ). For ease of comparison, the results were converted to CuKα radiation. Diffraction patterns were indexed according to the PDF file [29]. The unit cell parameters of Ir and IrO 2 were calculated using the PowderCell 2.4 program [30] with the internal (Si substrate) or external (polycrystalline silicon) standard. The unit cell parameter of Au was calculated from the position of the (111) peak. Coherent scattering regions (CSR) were calculated using the Scherrer formula taking into account the FWHM (full widths at half maximum) of the polycrystalline silicon standard.
The microstructural features of surface and cross-section of the samples were investigated using scanning electron microscopy (SEM, microscope JEOL-ISM 6700F, JEOL, Kyoto, Japan). The elemental composition was estimated by energy dispersive analysis (EDX, analyzer EX-2300BU, JEOL, Tokyo, Japan). EDX detector error was ±5%. The adhesive characteristics of the samples were evaluated using a tape test (adhesive force 2.97 N/10 mm).
Quantitative data on the composition were obtained using X-ray photoelectron spectroscopy (XPS, PHOIBOS-150 spectrometer, SPECS Surface Nano Analysis GmbH, Berlin, Germany, PHOIBOS-150-MCD-9 analyzer, FOCUS-500 monochromator, AlKα radiation, hν = 1486.74 eV, 200 W, analyzer transmission energy 20 eV). The binding energy (E b ) scale was calibrated using the positions of the energy level peaks of the Au 4f 7/2 (E b = 84.0 eV) and Cu 2p 3/2 (E b = 932.67 eV) atoms. To remove the surface layer, the samples were bombarded with Ar + ions with energy of 2.5 keV for 10 min. The spectra were processed in the CASA program (Japan) using the Doniach-Sanjich asymmetric function. The background was taken into account according to the Shirley method. The Ir 4f spectra are 4f 5/2,7/2 doublets with the spin orbit splitting 2.98 eV with the area ratio 4:3. These peaks were fitted into the components using the parameters proposed in [31].
Quantitative data on the amount of deposited gold were obtained by inductively coupled plasma atomic emission spectroscopy (ICP-AES) using the High-Resolution Spectrometer iCAP 6500 (Thermo Fisher Scientific, Waltham, MA, USA). The sample solution was injected into the plasma through a nebulizer of SeaSpray type using a peristaltic pump with the rate of 0.7 mL/min. Analysis conditions: cooling argon flow = 12 L/min, secondary = 0.5 L/min; registration time on the first slit = 15 s; on the second slit = 5 s. The power supplied to ICP inductor was 1150 W (recommended by the manufacturer of the spectrometer). The registration of emission spectra was carried out at the axial observation of plasma.
Sample dissolution was performed using mixt concentrated hydrochloric and nitric acids (3:1). Sample preparation was performed using disposable plastic tubes with volume of 15 and 50 mL, polypropylene container with volume of 20 mL, and automatic pipette with variable volume (1.00-5.00 mL, 100-1000 µL, 10-100 µL). For determining Au the most intense spectral lines were used (without the spectral influence of the matrix), namely, 208.209, 242.795, 267.595 nm. The validation of technique by spike experiment was provide.

Biological Investigation
Man-1 cells were provided from the cell culture collection of Laboratory of Developmental epigenetics of the Federal Research Center Institute of Cytology and Genetics SB RAS [32]. The cells were cultured in Dulbecco's modified Eagle's medium (DMEM) (Thermo Fisher Scientific, Waltham, MA, USA), supplemented with 10% fetal bovine serum (Thermo Fisher Scientific, Waltham, MA, USA), 100 U/mL penicillin (Gibco, Waltham, MA, USA), 100 U/mL streptomycin (Gibco, Waltham, MA, USA) and 2 mmol/L L-glutamine (Invitrogen, Carlsbad, CA, USA) at 37 • C in a humidified atmosphere of 5% CO 2 . The cytotoxicity tests were carried out by indirect contact. Extracts were prepared using DMEM medium as the extraction medium with the surface area to extraction medium ratio 1.25 mL/cm 2 in a humidified atmosphere with 5% CO 2 at 37 • C for 72 h [33]. The control groups involved the use of DMEM medium as negative controls. Cells were incubated in 96-well flat-bottomed cell culture plates at 1 × 10 4 cells per 200 µL in each well and incubated for 24 h to allow attachment. The medium was then replaced with 200 µL of extracts. After incubating the cells in a humidified atmosphere with 5% CO 2 at 37 • C for 1, 3, and 5 days, respectively, cell morphologies were observed by optical microscopy (Nikon Ti-E microscope, Nikon, Tokyo, Japan) and the cytotoxic effect was measured using the Cell Proliferation Kit XTT (Applichem, PanReac Applichem, Barcelona, Spain) according to the manufacturer's instructions. The optical density of the well contents was measured at a wavelength of 450 nm and a reference wavelength of 655 nm using an iMark plate photometer (Bio-Rad Laboratories, Inc., Hercules, CA, USA) with DMEM medium as a reference solution. Cell viability was calculated using the equation: where A is the difference between the optical density of the samples and the corresponding reference solution.
The pH in the extracts was measured using pH-meter Anion 4102 with a combined glass electrode ESK-10601/7 (Infraspack-Analit, Novosibirsk, Russia). The electrode was calibrated against a standard buffer solutions with pH 6.86 and 9.18. The content of nickel and other metals in the extracts was determined by ICP-AES as described in par. 2.2 using multielement standard solutions MES 1,2,3,4 [34].

Statistical Analysis
Statistical processing of the results was carried out using the STATISTICA 8.0 software (StatSoft Inc., Tulsa, OK, USA). Normality of data distribution was assessed using the Shapiro-Wilk test. Mann-Whitney U-test was used to identify differences between groups. Differences between groups were considered significant at p < 0.05. Results are presented as median and interquartile range.

Composition and Microstructure of Coatings Obtained by MOCVD from [Ir(cod)(acac)] in Oxidizing Atmosphere
Traditionally, metallic iridium coatings have been obtained through the MOCVD in a reducing atmosphere [23,[35][36][37]. These experiments should be carried out at relatively high temperatures (≥400 • C) to achieve effective vapor decomposition of most volatile precursors in hydrogen presence. However, lower temperatures (≤350 • C) are recommended to coat TiNi substrates via a MOCVD approach [17][18][19][20]. The use of oxygen as a reagent gas is an opportune way both to reduce the deposition temperature and ensure a high growth rate. However, the process is complicated by the possibility of oxidized iridium formation [35,36]. The studies of thermal decomposition of iridium complexes vapors [38] and MOCVD processes at various oxygen concentrations [39] have demonstrated that the presence of cyclooctadiene-1,5 (cod) as a part of the precursor leads to the formation of the metal phase.
In this work, a complex [Ir(cod)(acac)] characterized by availability and storage stability [40] was selected as the iridium precursor. Since this compound was used only in a reducing atmosphere [23,41], we started the work with a series of MOCVD experiments varying deposition temperature and oxygen flow rate using model Si substrates to study the composition and microstructural features of the coatings obtained. The deposition conditions for the studied samples are shown in Table 1.
The surface of the coatings obtained at a low oxygen flow rate (33 mL/min, samples 1-3) is formed by small particles of an irregular shape (Figure 1a,b). The deposition temperature increasing in the range of 290-350 • C leads to the changes in the size of these particles (from 10-20 nm in sample 1 and 2 to 40-90 nm in sample 3), and their shape converts to the X-shape (sample 3). The smallest particles (samples 1 and 2) are grouped into agglomerates (Figure 1a). The cross-section study has showed that the columnar-dendritic growth of coatings is realized under these MOCVD conditions (Figure 1a,b). At low deposition temperatures (samples 1 and 2), the columnar formations are very densely packed and formed by small pieces close in size to those observed on the surface (Figure 1a). With an increase in the deposition temperature (sample 3), the boundaries of the columnar formations appear more clearly and their width rise in accordance with the growth of their constituent particles (Figure 1b). The two-fold increase in the growth rate was observed when the deposition temperature changed from 290 to 350 • C. In general, the observed columnardendritic microstructure is typical for the metallic iridium coatings obtained by MOCVD at high growth rates and low ratios of the deposition temperature (T dep. ) to the melting temperature (T m ) of the metal [37]. In our experiments, this ratio was T dep. /T m = 0.12-0.14 (T m (Ir) = 2446 • C [42]). According to XRD, the investigated coatings are formed by polycrystalline metallic iridium ( Table 1). The diffraction patterns contain reflections (111)    According to XRD, the investigated coatings are formed by polycrystalline metallic iridium ( Table 1). The diffraction patterns contain reflections (111) and (200)  . The average crystallite sizes calculated from CSR correspond to the particle sizes estimated from the SEM data. The unit cell parameter (Table 1) is close to that for pure iridium (3.839 Å, PDF 010-87-0715 [29]). It should be noted that the presence of a small shoulder in the region 2θ = 39.2-40.2 • in the diffraction pattern of sample 3 obtained at the highest deposition temperature (350 • C) may indicate the presence of an iridium oxide phase (Figure 2). for randomly oriented crystallites). The average crystallite sizes calculated from CSR correspond to the particle sizes estimated from the SEM data. The unit cell parameter (Table  1) is close to that for pure iridium (3.839 Å, PDF 010-87-0715 [29]). It should be noted that the presence of a small shoulder in the region 2θ = 39.2-40.2° in the diffraction pattern of sample 3 obtained at the highest deposition temperature (350 °C) may indicate the presence of an iridium oxide phase ( Figure 2). An increase in the oxygen flow rate strongly promotes the formation of the oxidized iridium. For example, its four-fold increase is sufficient for the formation of crystalline IrO2 even at a deposition temperature of 310 °C (sample 4, Figure 2). With a two-fold increase, the oxide phase becomes dominant already at 330 °C (sample 5, Figure 2). The strong preferred orientation of IrO2 crystallites along the (101) direction is observed for this sample. The unit cell parameters of iridium oxide calculated here (a = 4.497(2) Å, c = 3.155(3) Å) correspond to pure tetragonal IrO2 (a = 4.505 Å, c = 3.159 Å, PDF 010-88-0288 [29]). According to CSR, the average size of these (101)-oriented crystallites is 130(13) nm and the (110)-oriented crystallites is 85(9) nm.
The appearance of a crystalline oxide phase in the coatings causes corresponding changes in their microstructure. Specifically, three types of particles appear on the surface of the mixed samples: (1) square nanotubes, (2) incomplete/scrolled nanotubes, (3) wedgeshaped rods. The types (1) and (2) are observed in sample 4 with the diameters about 30-50 nm and up to 100 nm, respectively (Figure 1c). The surface of sample 5 is mainly formed by particles (2) with increased sizes (up to 400 nm); particles (3) are also present ( Figure  1d). In general, the observed changes in the particle shapes depending on the deposition conditions are consistent with the findings described in the study of the morphology evolution of IrO2 crystals in MOCVD processes using [Ir(cod)(Cp Me )] precursor and oxygen reagent gas (Cp Me = methylcyclopentadienyl) [43]. An increase in particle size is associated with an increase in the deposition temperature and growth rate.
Samples 3 and 5 were selected for a more detailed investigation of the composition An increase in the oxygen flow rate strongly promotes the formation of the oxidized iridium. For example, its four-fold increase is sufficient for the formation of crystalline IrO 2 even at a deposition temperature of 310 • C (sample 4, Figure 2). With a two-fold increase, the oxide phase becomes dominant already at 330 • C (sample 5, Figure 2). The strong preferred orientation of IrO 2 crystallites along the (101) direction is observed for this sample. The unit cell parameters of iridium oxide calculated here (a = 4.497(2) Å, c = 3.155(3) Å) correspond to pure tetragonal IrO 2 (a = 4.505 Å, c = 3.159 Å, PDF 010-88-0288 [29]). According to CSR, the average size of these (101)-oriented crystallites is 130(13) nm and the (110)-oriented crystallites is 85(9) nm.
The appearance of a crystalline oxide phase in the coatings causes corresponding changes in their microstructure. Specifically, three types of particles appear on the surface of the mixed samples: (1) square nanotubes, (2) incomplete/scrolled nanotubes, (3) wedgeshaped rods. The types (1) and (2) are observed in sample 4 with the diameters about 30-50 nm and up to 100 nm, respectively (Figure 1c). The surface of sample 5 is mainly formed by particles (2) with increased sizes (up to 400 nm); particles (3) are also present (Figure 1d). In general, the observed changes in the particle shapes depending on the deposition conditions are consistent with the findings described in the study of the mor-phology evolution of IrO 2 crystals in MOCVD processes using [Ir(cod)(Cp Me )] precursor and oxygen reagent gas (Cp Me = methylcyclopentadienyl) [43]. An increase in particle size is associated with an increase in the deposition temperature and growth rate. Samples 3 and 5 were selected for a more detailed investigation of the composition using XPS. The spectra after Ar + etching are shown in Figure 3a. Both coatings contain only iridium, oxygen, and carbon (wt.%): Ir, 95.1; O, 3.2; C, 1.7 (sample 3); Ir, 93.6; O, 5.3; C, 1.1 (sample 5). The analysis of the Ir 4f peaks shows that in both samples iridium is observed in two states: Ir 0 (E b (Ir 4f 7/2 ) = 60.7 eV) and Ir IV belonging to IrO 2 (E b (Ir 4f 7/2 ) = 61.4 eV). It should be noticed that the content of oxidized iridium in sample 5 is significantly higher in comparison with sample 3. Namely, IrO 2 is the dominant component in sample 5 (Ir 0 /Ir IV at.% ratio = 0.8). The broadening of the FWHM parameter is observed for the Ir IV 4f 7/2 component (1.50 compared to 0.90 for the Ir 0 component). According to [31], this may indicate the presence of an amorphous oxide. In the case of sample 3, both components of the Ir 4f 7/2 peak are close in the considered parameter and its values (FWHM = 0.90 for Ir 0 and 0.96 for Ir IV ) correspond to crystalline phases [31]. The Ir 0 /Ir IV at.% ratio in sample 3 is 6.1. Thus, the general trends in the formation of the sample compositions, determined from the XRD and XPS data, namely an increase in the oxygen content and the predominance of the IrO 2 phase in sample 5 compared to 3, correlate with each other. However, to establish correlations of the XPS data with the bulk composition of the resulting coatings, additional studies of the local composition over the layer thickness are required (XPS data are collected in Ir samples from a depth of about 5 nm).

Iridium and Gold-Iridium Coatings on TiNi
The deposition conditions for sample 2 were selected to obtain metallic Ir coatings on titanium nickelide because here the crystalline oxide phase does not appear yet. However, the growth rate is high enough to provide the developed morphology formed by agglomerates of dendritic columnar structures.

Iridium and Gold-Iridium Coatings on TiNi
The deposition conditions for sample 2 were selected to obtain metallic Ir coatings on titanium nickelide because here the crystalline oxide phase does not appear yet. However, the growth rate is high enough to provide the developed morphology formed by agglomerates of dendritic columnar structures.
According to SEM and XRD study of the samples obtained in a same experiment on Si and TiNi, the substrate does not affect the microstructure of the forming coating as well as the orientation and the size of the crystallites (Figures 4 and 5a). In particular, the CSR estimated for Ir coatings on TiNi shows sizes of 21(2) nm and 11(1) nm for the crystallites oriented in (111) and (200) directions, respectively. Within the margin of error, this correlates with the data for Ir obtained on Si substrate ( Table 1). The relative intensity of the I (111) /I (200) reflections in the coatings on TiNi also varies within 4-5. The unit cell parameter is insignificantly higher than on Si (3.847(2) Å). According to cross-section SEM, the coating thicknesses in this series of samples (Ir/Si) is 680-700 nm. Note that the previous studies of metal coatings on the same TiNi materials have shown that a layer thickness from 400 nm to 2 µm does not prevent superelastic deformation of the substrate [44,45]. Evaluation of the adhesion of the obtained Ir coatings to TiNi by the adhesive tape method (scotch test) showed the absence of peeling.
Coatings 2021, 11, x FOR PEER REVIEW 10 of 17 400 nm to 2 μm does not prevent superelastic deformation of the substrate [44,45]. Evaluation of the adhesion of the obtained Ir coatings to TiNi by the adhesive tape method (scotch test) showed the absence of peeling.
[(CH3)2Au(thd)] was chosen as a volatile precursor for subsequent deposition of discrete gold film [46]. In the same experiment, Si and TiNi wafers were used as substrates as well as Ir/Si and Ir/TiNi samples described above. The results of studying the obtained samples (Au/Si, Au/TiNi, Au/Ir/Si, and Au/Ir/TiNi) using XRD and SEM are shown in Figures 4 and 5b,c.  [(CH 3 ) 2 Au(thd)] was chosen as a volatile precursor for subsequent deposition of discrete gold film [46]. In the same experiment, Si and TiNi wafers were used as substrates as well as Ir/Si and Ir/TiNi samples described above. The results of studying the obtained samples (Au/Si, Au/TiNi, Au/Ir/Si, and Au/Ir/TiNi) using XRD and SEM are shown in Figures 4 and 5b,c. According to XRD data, the Au/Si and Au/TiNi samples are characterized by a strong preferred orientation of the fcc-Au phase (Figure 4). In particular, the ratio of the intensities of the (111) (2θ = 38.  SEM analysis of the sample surface confirms the similarity of the Au coatings formed on the Si and TiNi surfaces (Figure 5b). In particular, the coating on Si is formed by two types of particles namely nanoparticles with the size of 8-40 nm and irregularly shaped agglomerates up to 400 nm in length. The latter type of particles is preferred. Apparently, these agglomerates are formed by several crystallites. In the case of TiNi, according to the XRD data, a decline in the size of the particles forming the coating surface is observed. The length of the agglomerates does not exceed 200 nm.
The larger type of Au particles is also clearly observed on the surface of the predeposited Ir layers on Si and TiNi substrates (Figure 5c). Thus, the formed gold coatings retain the discrete character. XRD analysis allows identifying the fcc-Au (111) and (200) reflexes ( Figure 4). The Au texture in the (111) direction noted above is retained, however, it becomes less pronounced: I (111) /I (200) = 4-10. This make it possible to estimate by CSR calculations the sizes of Au crystallites oriented in the (200) direction: 22(2)-28(3) nm. The sizes of Au crystallites oriented in (111) direction are comparable or smaller than those that were formed on TiNi substrate: 28(3)-40(4) nm. The unit cell parameters of Au for Au/Ir/Si and Au/Ir/TiNi samples also corresponds to pure metal indicating the absence of the carbon inclusion in the gold lattice [47,48]. Note that the unit cell parameter and CSR of the Ir phase do not change after Au deposition.
EDX data also confirm the presence of Au in the obtained samples. In particular, the amounts of gold have been evaluated to be 10 at.% and 18 at.% for the samples Au/Ir/Si and Au/Ir/TiNi, respectively. However, the calculation of the elemental composition from the EDX spectra is complicated by the superposition of the CKα (0.277 keV) and AuMα (2.123 keV) bands with iridium bands (N45N67 and Mα, respectively). This can lead to an overestimation of the calculated gold and carbon content. The latter was confirmed by comparing XPS and EDX data for Ir/Si samples. To quantify the deposited gold, we dissolved the obtained coatings in aqua regia and analyzed the resulting solutions by ICP-AES method. The average total gold concentrations were estimated to be 42 and 96 mcg/cm 2 for Au/Si and Au/Ir/Si samples, respectively and 56 and 144 mcg/cm 2 for Au/TiNi and Au/Ir/TiNi samples, respectively. Thus, the pre-deposited iridium layer favors the deposition of gold in the subsequent MOCVD experiment. Similar effect was observed during electrodeposition of Au nanoparticles on Ir coatings obtained by atomic layer deposition [49]. Is was explained in terms of the additional nucleation centers that formed on the Ir substrate and increase the density of Au islands.

Biological Characteristics of Noble Metal Coatings on TiNi
To assess the biocompatibility, a series of all-side-coated Ir/TiNi and Au/Ir/TiNi samples were obtained at the conditions discussed in par. 3.2. Human embryonic stem cell line Man-1 [50] were used for primary biological tests. We studied the dynamics of their survival for five days in extracts obtained by keeping the TiNi, Ir/TiNi and Au/Ir/TiNi samples in a culture medium ( Figure 6). In the case of uncoated TiNi, in the first 24 h, cell viability was comparable to that observed for control group (86%) and further a trend to a moderate cell proliferation was observed. The extracts obtained using Ir/TiNi samples clearly showed a dynamic cytotoxic effect leading to the death of almost all cells within 5 days. In the case of Au/Ir/TiNi samples, the initial increase in proliferation was observed (up to 130%) and then it was near constant in time.
Coatings 2021, 11, x FOR PEER REVIEW 12 of 17 samples in a culture medium ( Figure 6). In the case of uncoated TiNi, in the first 24 h, cell viability was comparable to that observed for control group (86%) and further a trend to a moderate cell proliferation was observed. The extracts obtained using Ir/TiNi samples clearly showed a dynamic cytotoxic effect leading to the death of almost all cells within 5 days. In the case of Au/Ir/TiNi samples, the initial increase in proliferation was observed (up to 130%) and then it was near constant in time.  Figure 7 showed the morphologies of Man-1 cells cultured in extraction media for 1, 3, and 5 days. The cell morphologies in TiNi and Au/Ir/TiNi extracts were normal and healthy, which is similar to that of the control group, and exhibited a healthy morphology of cells with spindle shape. In the Ir/TiNi group, by the end of the first day of observation, the number of living cells decreased markedly, and by the end of the experiment, only a single number remained.  Figure 7 showed the morphologies of Man-1 cells cultured in extraction media for 1, 3, and 5 days. The cell morphologies in TiNi and Au/Ir/TiNi extracts were normal and healthy, which is similar to that of the control group, and exhibited a healthy morphology of cells with spindle shape. In the Ir/TiNi group, by the end of the first day of observation, the number of living cells decreased markedly, and by the end of the experiment, only a single number remained. To explain these observations, we performed the analysis of extracts. Measurement of pH showed that all the extracts obtained by keeping the metal samples are close to each other and differ from the control insignificantly. In particular, pH = 7.22 was obtained for control, while in the case of TiNi, Ir/Ti, and Au/Ir/TiNi it was 7.71, 7.58, 7.64, respectively. Thus, there were no negative effect of the medium basicity on cell growth. AES analysis showed that the extracts obtained using the Ir/TiNi samples have a higher nickel content (0.86 mcg/mL). In the case of Au/Ir/TiNi samples, the nickel concentration was an order of magnitude lower (0.096 mcg/mL), while for TiNi samples as well as control extract, no nickel was found. These results correlate with the pronounced cytotoxic effect observed for the Ir/TiNi sample extracts.
The appearance of nickel in the extracts is caused by the electrochemical effect in the Ni-Ir galvanic pair, where Ni becomes a sacrificed anode. The contact of this galvanic pair To explain these observations, we performed the analysis of extracts. Measurement of pH showed that all the extracts obtained by keeping the metal samples are close to each other and differ from the control insignificantly. In particular, pH = 7.22 was obtained for control, while in the case of TiNi, Ir/Ti, and Au/Ir/TiNi it was 7.71, 7.58, 7.64, respectively. Thus, there were no negative effect of the medium basicity on cell growth. AES analysis showed that the extracts obtained using the Ir/TiNi samples have a higher nickel content (0.86 mcg/mL). In the case of Au/Ir/TiNi samples, the nickel concentration was an order of magnitude lower (0.096 mcg/mL), while for TiNi samples as well as control extract, no nickel was found. These results correlate with the pronounced cytotoxic effect observed for the Ir/TiNi sample extracts.
The appearance of nickel in the extracts is caused by the electrochemical effect in the Ni-Ir galvanic pair, where Ni becomes a sacrificed anode. The contact of this galvanic pair with the solution that leads to Ni 2+ release becomes possible due to the microstructure of the obtained coatings. In fact, the high rate of coating growth in the vertical direction and the defects on the intrinsic TiNi surface relief (Figure 8) lead to the formation of the clearances or cracks between dendrite agglomerates (Figure 8b). Subsequent Au deposition appears to cover these clearances partially (Figure 8c). This prevents Ni release and therefore improves the biocompatibility of the material. The increased proliferation in this case could be associated with the formation of reactive oxidized species (ROSs) promoted by the gold surface. In fact, ROS can stimulate cellular proliferation and act as a second messenger in cellular signaling [51,52]. This issue will be investigated in detail in future work.
with the solution that leads to Ni 2+ release becomes possible due to the microstructure of the obtained coatings. In fact, the high rate of coating growth in the vertical direction and the defects on the intrinsic TiNi surface relief (Figure 8) lead to the formation of the clearances or cracks between dendrite agglomerates (Figure 8b). Subsequent Au deposition appears to cover these clearances partially (Figure 8c). This prevents Ni release and therefore improves the biocompatibility of the material. The increased proliferation in this case could be associated with the formation of reactive oxidized species (ROSs) promoted by the gold surface. In fact, ROS can stimulate cellular proliferation and act as a second messenger in cellular signaling [51,52]. This issue will be investigated in detail in future work.

Conclusions
This work focused on the formation of Ir and Au/Ir coatings with developed morphology onto the TiNi substrates by MOCVD method and the assessment of their biocompatibility.
To obtain target Ir coatings, the [Ir(cod)(acac)] precursor was tested at deposition temperatures 290-350 °C and oxygen flow rates 33-133 mL/min. At lowest oxygen amount, the crystalline metal coatings are mainly formed while the iridium oxide phase appears at higher temperatures (350 °C). The higher oxygen flow (67, 133 mL/min) promotes the formation of a crystalline IrO2 even at lower temperatures. Ir coatings with a well-developed surface morphology formed by dendritic columns of 10-20 nm particles were obtained on TiNi substrates at 310 °C. The surface of these samples was decorated with discrete Au coatings using [(CH3)2Au(thd)] precursor at deposition temperature 240 °C and oxygen flow rate 17 mL/min. The Ir/TiNi samples exhibited a pronounced cytotoxic effect on Man-1 fibroblasts, which is associated with the nickel release due to the clearances between dendrite agglomerates. The subsequent discrete Au layer reduces nickel leaching and promotes cell proliferation. Therefore, it is of interest to further develop such bimetallic coatings and study their mechanical characteristics.
Alternatively, the development of a sub-layer layer with a compact microstructure could be a further research direction to permit the deposition of noble metal coatings with developed morphologies onto TiNi by MOCVD. It should be also noted that the materials obtained here could be applied directly to other implant materials such as carbon and polymer ones.

Conclusions
This work focused on the formation of Ir and Au/Ir coatings with developed morphology onto the TiNi substrates by MOCVD method and the assessment of their biocompatibility.
To obtain target Ir coatings, the [Ir(cod)(acac)] precursor was tested at deposition temperatures 290-350 • C and oxygen flow rates 33-133 mL/min. At lowest oxygen amount, the crystalline metal coatings are mainly formed while the iridium oxide phase appears at higher temperatures (350 • C). The higher oxygen flow (67, 133 mL/min) promotes the formation of a crystalline IrO 2 even at lower temperatures. Ir coatings with a welldeveloped surface morphology formed by dendritic columns of 10-20 nm particles were obtained on TiNi substrates at 310 • C. The surface of these samples was decorated with discrete Au coatings using [(CH 3 ) 2 Au(thd)] precursor at deposition temperature 240 • C and oxygen flow rate 17 mL/min. The Ir/TiNi samples exhibited a pronounced cytotoxic effect on Man-1 fibroblasts, which is associated with the nickel release due to the clearances between dendrite agglomerates. The subsequent discrete Au layer reduces nickel leaching and promotes cell proliferation. Therefore, it is of interest to further develop such bimetallic coatings and study their mechanical characteristics.
Alternatively, the development of a sub-layer layer with a compact microstructure could be a further research direction to permit the deposition of noble metal coatings with developed morphologies onto TiNi by MOCVD. It should be also noted that the materials obtained here could be applied directly to other implant materials such as carbon and polymer ones.