Magnetron sputtered ß-Ti coatings for biomedical application: A HiPIMS approach to improve corrosion resistance and mechanical behavior

Full Length Article Magnetron sputtered ß-Ti coatings for biomedical application: A HiPIMS approach to improve corrosion resistance and mechanical behavior Juan Carlos S´ anchez-L´ opez a,* , Vanda Godinho a,b , Carmen L´ opez-Santos a,c , Paula Navarro b,d , Luisa Marleny Rodríguez-Albelo b , Miriam S´ anchez-P´ erez a , Emilio Jim´ enez-Piqu´ e e , Yadir Torres b a Instituto de Ciencia de Materiales de Sevilla (ICMS), CSIC-US, Avda. Am´ erico Vespucio 49, E-41092 Sevilla, Spain b Departamento de Ingeniería y Ciencia de Los Materiales y del Transporte, Universidad de Sevilla, Escuela Polit´ ecnica Superior, Virgen de ´ Africa 7, E-41011 Sevilla, Spain c Departamento de Física Aplicada I, Escuela Polit´ ecnica Superior, Universidad de Sevilla, Virgen de ´ Africa 7, E-41011 Sevilla, Spain d Instituto de Microelectr´ onica de Sevilla, IMSE-CNM, Avda. Am´ erico Vespucio 28, 41092 Sevilla, Spain e Center for Structural Integrity, Reliability, and Micromechanics of Materials (CIEFMA), Department of Materials Science and Engineering, Universitat Polit` ecnica de Catalunya (UPC) - BarcelonaTECH, 08019, Barcelona, Spain ARTICLE INFO Keywords: TNZT Magnetron sputtering Mechanical properties Wetting Surface chemistry Electrical impedance ABSTRACT This work presents the surface modification of commercially pure Ti specimens (c.p.-Ti) prepared by conventional powder metallurgy by depositing a thin film of a ß-Ti alloy (Ti-35Nb-7Zr-5Ta, wt. %, TNZT). Two types of pulsed technologies: conventional (p-DC) and high-power impulse magnetron sputtering (HiPIMS), with and without bias assistance (−60 V) under similar power conditions (250 W) were applied on titanium specimens and silicon substrates leading to different film morphologies and functional properties. Microstructural, X-ray diffraction, nanoindentation, surface wetting, XPS and electrochemical impedance measurements were done to characterize their functionality. All the coatings presented a reduced Young’s Modulus (E ≤80GPa) compared to the bulk Ti, representing a reduction of more than 30 %. This decrease can significantly contribute to the reduction of the stress-shielding effect, mitigating the risk of implant loosening and failure. The hardness values of TNZT coatings, slightly lower than c.p.-Ti substrate, range from 4.1 to 4.7 GPa. XPS analysis shows a passivation layer of TiO 2 , Nb 2 O 5 , and ZrO 2 , which offers high impedance and excellent corrosion resistance. The best compromise between mechanical and corrosion properties is achieved with the HiPIMS technology, thanks to its compact film microstructure with high electrical resistance, despite its limited thickness of about 1 μ m. 1. Introduction Titanium (Ti) and its alloys have emerged as a preferred choice for orthopedic and dental implants over the last few decades [1,2]. This preference relies on the combination of adequate properties, including excellent biocompatibility, high corrosion resistance, and suitable mechanical properties when compared to other materials. Although Ti implants combine numerous advantages with high success rates, still 5–10 % result in failure [3], which considering the high number of implants placed every year is deemed unacceptable. There are three main reasons for Ti implants failure. Firstly, the stress-shielding phenomenon, caused by the substantial disparity in stiffness between the implant and the bone, leads to bone resorption and increases the risk of fracture of the adjacent bone. Secondly, problems associated with excessive mechanical loads or fatigue failures resulting from inappropriate material selection or implant design. Thirdly, the corrosion resistance of Ti implants still requires improvement. The corrosion resistance of titanium is mainly ascribed to the spontaneous formation of a very thin (2–5 nm) amorphous layer of titanium dioxide (TiO 2 ), which is crucial for the biocompatibility of Ti implants. This passivation layer serves as a barrier, converting Ti and its alloys into bioinert materials, thereby minimizing the release of ions from the implant into the surrounding tissues and preventing adverse reactions in the body. However, when exposed to aggressive biological media and subjected to repetitive or varying loads, these oxide layers, which usually contain many defects, can be compromised [4]. * Corresponding author at: Instituto de Ciencia de Materiales de Sevilla (CSIC-Univ. Sevilla), Avda. Am´ erico Vespucio 49, E-41092 Sevilla, Spain. E-mail address: [email protected] (J.C. S´ anchez-L´ opez). Contents lists available at ScienceDirect Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc https://doi.org/10.1016/j.apsusc.2024.161366 Received 4 July 2024; Received in revised form 16 September 2024; Accepted 26 September 2024 Applied Surface Science 680 (2025) 161366 Available online 27 September 2024 0169-4332/© 2024 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/bync-nd/4.0/ ). Hence, the development of materials with appropriate biomechanical behavior, such as low compliance, good tribological resistance, and biofunctional characteristics like corrosion resistance and bone ingrowth. This is particularly important for implant applications where a high strength and fatigue resistance of the implant structure are required together with low elastic modulus and better biocompatibility close to the bone is desired. β-Ti alloys, with Young’s Modulus (E) falling within the range of 50–95 GPa [1,5,6] are promising candidates to avoid the stress shielding phenomenon, as their modulus is closer to that of cortical bone. In particular, β-Ti alloy with chemical composition of Ti35Nb-7Zr-5Ta (in wt.%), hereafter referred to as TNZT, exhibits one of the lowest E among β-Ti alloys, around 55 GPa [6]. Moreover, the passivation layer formed on these alloys includes ZrO 2 and Nb 2 O 5 , improving the corrosion resistance properties when compared to that formed on pure Ti [6–9]. Additionally, alloying elements with low electrochemical reaction potentials, such as Ta, Nb, and Zr could lead to the reduction of the anode activity and improvement of passive properties [8]. However, TNZT alloy presents relatively low strength and poor wear behavior, limiting its application as implant material [5,6]. For bulk alloy materials, different techniques were applied to increase the strength of these alloys, either by inclusion of other alloy elements [10,11], severe plastic deformation [12] or ageing treatments [13]. Surface modification techniques can be used to customize the implant design to meet the specific medical requirements of each situation. In our previous work [14], we introduced the employ of magnetron sputtering-based techniques to modify the surface of Ti implants manufactured through conventional powder metallurgy with a Ti-6Al4V coating, resulting in improved mechanical properties. Plasmabased methods such as magnetron sputtering not only allow for the growth of crystalline materials at room temperature under nonequilibrium conditions but also provide control over the deposition parameters, enabling the creation of different surface nanostructures that could enhance osteoblast maturation, increase bone-to-implant contact and improve the implant success rates. This study employed the high-power impulse magnetron sputtering technique (HiPIMS), which is widely recognized as an advantageous variant for achieving improved mechanical, tribological, and anti-corrosive properties [15,16]. This technique has been shown to yield superior results when compared with DC magnetron sputtering [14–16]. The high-energy pulses delivered to the target material during the sputtering process result in a denser and more adherent coating on the implant able to withstand handling and its insertion in the patient body as well as guarantee a good performance in service [14]. In this work, we explore a similar approach to tailor the surface of Ti implants depositing a β-Ti alloy (TNZT) with the aim of decreasing the mismatch between the bone and the implant, improving osseointegration. Additionally, alloying Ti with Nb, Zr, and Ta avoids the potential drawbacks associated with toxicity from elements such as Al and V, with excellent mechanical properties and workability [1,6]. In this work, we investigate the ability of high plasma ionization of HiPIMS technique and ion energy (via substrate biasing) on the morphology, wettability and mechanical properties of TNZT coatings. In addition, electrical impedance measurements are performed to semi-quantitatively evaluate potential changes in the coatings, which can be related to the corrosion resistance of the surfaces and any alterations that may occur at the implant/bone interface. The achieved functionality of the coating is compared with an analogous one deposited via conventional DC sputtering in pulsed mode, designed to prevent arc formation. The combination of surface tailoring and a promising biomaterial like TNZT could play a crucial role in influencing and modulating cell-surface interactions, thereby enhancing the coating’s potential for biomedical application. 2. Experimental 2.1. Thin film deposition The TNZT coatings were fabricated via magnetron sputtering (MS) from a commercial 2 inches diameter by 3 mm-thickness target with the following nominal chemical composition (Ti 53 wt%, Nb 35 wt%, Zr 7 wt%, Ta 5 wt%) provided by Photon Export. The substrates were titanium cylinders (diameter 12 mm; height 5 mm) prepared from commercially pure Ti −grade IV (c.p.-Ti) through conventional powder metallurgy. The procedure followed for the preparation of the titanium specimens is further described in [14]. Replica coatings, deposited simultaneously on silicon (100) substrates were also employed for microstructural and chemical analysis. The pressure of the vacuum chamber was ≈1x10 −4 Pa prior to synthesis. Then argon was fed into the chamber at a flow rate of 25 sccm until the pressure reached 0.75 Pa, serving as sputtering gas. A circular 2 inches balanced magnetron was used, providing a total area of 20.27 cm 2 . The power applied to the magnetron head was 250 W using three different methodologies: pulsed direct current (p-DC), HiPIMS, and HiPIMS with simultaneous biasing of the substrates at −60 V. ENI DC-pulsed and HiPIMS Solvix equipment were used as power sources. The rotary sample holder (30 r.p.m.) was heated at 250 ◦C during film deposition (duration: 150 min). The distance between the substrate and the target was consistently maintained at 10 cm for each deposition. The time characteristics of the pulses are summarized in Table 1. The operation duty cycle represents the on/off time ratio of the pulses and typically are less than 10 % in HiPIMS processes. This ensures an ultrahigh peak power density and a higher ionization percentage of sputtered species [16]. With an average sputtering power of 250 W, the voltage during pulses was measured at 625 V, and the peak current was approximately 25 A. The corresponding peak power density and peak current density were estimated to be around 0.77 kW/cm 2 and 1.2 A/cm 2 , respectively. The bias voltage (negative average value of 60 V) was applied with the p-DC source using the same pulse features (frequency 250 kHz; duration 496 ns). 2.2. Phase composition, microstructural and mechanical characterization of coatings X-ray diffraction analysis (XRD) was employed to study the crystalline phase composition of the coatings, the uncoated Ti substrates, along with the target material utilized for the sputtering deposition. The equipment employed was an X’Pert Pro PANALYTICAL diffractometer with Cu K α radiation at 1◦of incidence angle. The average crystalline domain sizes were estimated using the Debye-Scherrer formula based on the (110) and (211) reflections. X-ray photoemission spectroscopy (XPS) was used to investigate the composition and chemical state of the coatings surface, with a PHOIBOS 100 spectrometer, working with the Mg K α radiation as excitation source. The pass energy was set at 50 eV for the survey spectra, and 30 eV for the high energy resolution spectra of the detected elements. An ionic sputtering with Ar + ions (2 keV, 5 min, 10 -6 mbar) of Table 1 Pulse parameters used and deposition rates for each sample preparation. Sputtering methodology Pulse characteristics Duty Cycle Deposition rate (nm/min) p-DC 250 kHz; 496 ns 88 % 19.3 HiPIMS 500 Hz; 40 µs 2 % 7.6 HiPIMSþbias (─60 V) 500 Hz; 40 µs 2 % 7.7 J.C. S´ anchez-L´ opez et al. Applied Surface Science 680 (2025) 161366 2 approximately 1.0 µA was made to remove partially the surface contamination. The spectra were calibrated using the binding energy (BE) of the main oxide component of Ti 2p peak attributable to TiO 2 phase at 458.6 eV. Scanning electron microscopy (SEM) was conducted using a SEMFEG Hitachi S4800 microscope with an energy-dispersive X-ray (EDX) detector. Morphological characterization and chemical analysis were performed at 5 and 20 kV, respectively. The coated titanium specimens were directly observed for top analysis while silicon wafers (100) were used for cross-section analysis. Several SEM images underwent image analysis to ascertain the equivalent diameter (Deq) of columnar structures and pores, and shape factor (Ff), calculated by the formula Ff = 4 π A/PE 2 , where A represents the area of columns/pores, and PE stands for the experimental perimeter of pseudoelliptic columns/pores. ImageProPlus 6.2 software (Mediacibernectic, Bethesda, MD, USA) was employed for image analysis measurements. Optical interferometry was employed to determine the film surface roughness (Sa and Sq) on silicon coated substrates using a S-Neox 090 microscope from Sensofar at 50×magnification. The analyzed area corresponds to approximately 300 μ m 2 . Wettability was assessed through static contact angle (CA) measurements acquired using an OCA 20 Dataphysics instrument. Measurements were done with milli-Q water (1–2 μ L, 3 replicas, 72.8 mN/m) and bovine albumin serum from Sigma Aldrich (5 μ L, 1 test, 50.0 mN/m [17]), on coated c.p.-Ti substrates. The measured values were compared with those obtained for an uncoated and coated c.p.-Ti specimen with the biomedical alloy Ti6Al4V[14]. Nanoindentation experiments were carried out using an MTS nanoindenter XP with a Berkovich diamond tip at a constant strain rate of 0.05 s −1 . A matrix of 4 ×4 indentations was drawn for each coating and the results were analyzed by the Oliver and Pharr method [18]. A fused silica standard was used for calibration. Film hardness value was estimated at 10 % of the thickness of the coating and the elastic modulus was estimated by extrapolation to zero penetration depth, according to ISO 14577. Measurements were done utilizing the continuous stiffness measurement (CSM) option, enabling the calculation of both properties during loading period relative to the penetration depth. The influence of the investigated TNZT coatings on the electrical impedance of the c.p.-Ti cylinders was evaluated using a HewlettPackard 4395A (Agilent Technologies) equipment and methodology described in earlier works [14,19]. A ß-Ti bulk TNZT alloy was also prepared and measured in similar way for comparison purposes. To place the samples in the analyzer, the device features a retractable accessory, the HP19042 module, which was used to secure each sample while measurements were conducted. The contact area of the output port is 12 mm 2 . Once the HP 4395A is calibrated, fixture compensation Table 2 Chemical composition of the target material and deposited coatings by EDX analysis at 20 kV. Nominal composition of the target is Ti53Nb35Zr7Ta5. Ti Nb Zr Ta Nb/Ti Zr/Ti Ta/Ti wt.% (at.%) wt.% (at.%) wt.% (at.%) wt.% (at.%) wt.% (at.%) wt.% (at.%) wt.% (at.%) Target 55.3 (71.9) 34.0 (22.8) 5.0 (3.4) 5.7 (2.0) 0.61 (0.32) 0.09 (0.05) 0.10 (0.03) p-DC 59.6 (75.0) 29.6 (19.2) 6.7 (4.4) 4.1 (1.4) 0.50 (0.26) 0.11 (0.06) 0.07 (0.02) HiPIMS 57.3 (73.0) 32.6 (21.4) 6.4 (4.3) 3.7 (1.2) 0.57 (0.29) 0.11 (0.06) 0.06 (0.02) HiPIMSþbias 55.9 (71.9) 33.9 (22.5) 6.4 (4.3) 3.7 (1.3) 0.61 (0.31) 0.11 (0.06) 0.07 (0.02) Fig. 1. XRD scans for the three TNZT films under the three different synthesis conditions (p-DC, HiPIMS and HiPIMS+bias) compared to those of the used TNZT target and the uncoated c.p.-Ti substrate. The positions of the cubic [c-Ti (□) (PDF #44–1288), c-Nb ( ○) (PDF #2–1108), c-Ta (×) (PDF #1–1309)], and hexagonal crystalline references [h-Ti (✰) (PDF #44–1294)] are also included. J.C. S´ anchez-L´ opez et al. Applied Surface Science 680 (2025) 161366 3 is performed to calibrate it along with the HP19042 to eliminate errors occurring between the electrode of the test fixture and the output port of the impedance test kit. Electrical impedance measurements were performed under dry conditions in the ambient environment, as the device does not require specific conditions for measurements. The frequency ranged between 100 MHz and 500 MHz This range was specifically chosen due to the absence of notable differences observed among the curves obtained for each coated disc at lower frequencies. The impedance measurements were performed three times for every sample, the mean value was then calculated and graphically represented. 3. Results and Discussion The chemical composition of the coatings was obtained by EDX analysis on three silicon coated specimens and compared to that obtained for the target material. The results are outlined in Table 2, expressed in wt.% (with corresponding values in at. % provided in parenthesis). The composition of the coatings closely reproduces the target composition with slight differences. The most significant feature corresponds to the Nb/Ti ratio, which increases from 0.50 to 0.57 (in wt.%) when changing from p-DC to HiPIMS, and reaching 0.61 in the HiPIMS+bias, reproducing the value measured for the target. Regarding Zr and Ta elements, the differences were not significant based on the deposition mode. The crystalline structure of the coatings was established by grazing angle XRD at 1◦of incidence angle. Fig. 1 depicts the diffractograms obtained for the three sputtering conditions together with those corresponding to the initial TNZT alloyed target and the uncoated Ti specimen. The diffraction patterns of the TNZT films agree with cubic titanium (in contrast to the hexagonal phase of the c.p.-Ti substrate) whose positions are slightly shifted towards higher angles due to the presence of Ta, with lower lattice parameter (a 0 =0.3298 vs. 0.3306 nm of Ti). No evidence of hexagonal phases of titanium or zirconium is observed, confirming that after sputtering the cubic form (ß-phase) of Table 3 Coating thickness given by SEM cross-section, surface roughness values by optical interferometry and crystallite size calculated by XRD data from peaks (110) and (211). Coating thickness Sa Crystal size (110) Crystal size (211) (nm) (nm) (nm) (nm) p-DC 2900 10.2 ± 1.2 54 52 HiPIMS 1140 3.0 ± 0.5 68 21 HiPIMSþbias 1150 2.7 ± 0.8 22 12 Fig. 2. Cross-section (left column) and top (right column) micrographs obtained by SEM for the three TNZT coatings deposited on silicon substrates and c.p.-Ti specimens, respectively. J.C. S´ anchez-L´ opez et al. Applied Surface Science 680 (2025) 161366 4 the initial target alloy is maintained in the three coatings. Nevertheless, the crystallographic orientations are modified from a predominant (110) growth towards (211) in the TNZT thin films. The polycrystallinity, characterized by a more random orientation, is enhanced in the p-DC compared to HiPIMS, where the occurrences of the (200) and (220) orientations are very minor, although it might be influenced by its higher thickness. Additionally, there is a noticeable peak broadening, particularly evident when HiPIMS and biasing of the substrates are utilized simultaneously. The determination of the peak broadening, using the FWHM data of (110) and (211) reflections, allows the estimation of the crystalline domain sizes. The values are summarized in Table 3 and confirm that under conditions of high ion bombardment, achieved combining HiPIMS discharges and substrate biasing [16,20], the crystal sizes are minimized. Specifically, when employing argon as sputtering gas, Ti is an element characterized by a high ionization efficiency due to a relatively large electron impact ionization cross-section and a low ionization potential. Bohlmark and coworkers [21,22] demonstrated the higher ionic contribution and broad energy distribution in the HiPIMS vs. DC plasmas. They found that the ionic contribution of Ti exceeded 90 % (compared to 8 % in DC), and approximately half of the ionized Ti had energies higher than 20 eV, in contrast to approximately 2 eV in DC sputtering discharge. Fig. 2 summarizes cross and top sections SEM micrographs of the three TNZT coatings. The cross-section images were obtained from silicon coated samples after cleavage while the top-views were captured from the coated c.p.-Ti cylinders. The features obtained on both type of substrates are consistent, confirming that the film morphology is independent of the nature of the substrate, depending mainly on the deposition parameters. The measured thickness values are 2900, 1140, and 1150 nm for the films deposited with p–DC, HiPIMS and HiPIMS+bias, respectively. The decrease in thickness of HiPIMS samples is consistent with the reduction of deposition rates achieved by this technology in comparison to p–DC [23] and consistent with our previous results [24]. This reduction relies on the decrease of operation time (duty cycle decreased from 88 % in pDC to 2 % in HiPIMS), and the back–attraction of the larger proportion of positively charged ions from the target present in HiPIMS discharges [25]. The film growth is clearly columnar in p-DC but the film microstructure evolves to vain-like in the HiPIMS samples. This type of microstructure is typical of metallic glass-like behavior [26]. Despite the larger thickness of the p-DC sample, we can conclude that the columnar structure with very well-defined borders is already visible at a thickness comparable to that of HiPIMS samples. As commented previously, the higher ion density achieved in the HiPIMS plasma favored a denser film morphology, while in p-DC less energetic conditions resulted in a microstructure characterized by shadowing effects, leading to an open intercolumnar morphology [16]. The additional substrate bias facilitates the acceleration of Ar + ions at higher energies contributing also to densify the film surface and to reduce the deposition rate [24]. The top view of the samples clearly illustrates this trend, with a marked reduction of the open porosity at the column boundaries when changing to HiPIMS conditions. Further surface smoothing is achieved with the assistance of negative bias although the column shape becomes more irregular. The bigger structures correspond to pillars formed by various subunits (minor columns), which coalesce under the influence of additional ion bombardment. Image analysis carried out on these micrographs allows to determine the average columnar size and the pore size distribution. The estimated intercolumnar porosities were found to be 4, 2, 0.3 % (corresponding equivalent pore diameter 16, 10 and 1.5 nm) for p-DC, HiPIMS and HiPIMS+bias, respectively. These results highlight the increasing compaction structures obtained under more energetic conditions. Detailed results are summarized in supporting information, including histograms (Fig. S1a) and image analysis (Fig. S1b). Using various top view micrographs, the equivalent column diameters were estimated to be 150 nm for p-DC, 95 nm for HiPIMS and 154 nm for HiPIMS+bias. These values are probably overestimated due to the high degree of coalescence and surface smoothening influenced by ion bombardment, making difficult to evaluate accurately the size of columns. Attending to the top view (cf. Fig. 2, right column), it becomes evident that the Fig. 3. Average surface roughness measured by interferometric analysis on the silicon coated samples. Fig. 4. Elastic modulus (a) and hardness values (b) of the different coatings (pDC, HiPIMS and HiPIMS+bias) deposited on silicon compared to the uncoated c.p.-Ti substrate. J.C. S´ anchez-L´ opez et al. Applied Surface Science 680 (2025) 161366 5 column shape of the HiPIMS+bias sample is more heterogeneous, alternating larger features with minor ones. The result is a mosaic-like pattern whose space between tiles is minimized resulting in smoother surface overall. The estimation of surface roughness by means of interferometric microscopy analysis confirmed this result. Fig. 3 summarizes the mean values (Sa and Sq) revealing a marked step down when transitioning from p-DC to HiPIMS, from 10 to ≤3 nm, although a certain contribution from the larger thickness of p-DC sample cannot be disregarded. The 2D and 3D images used for this analysis can be found in supplementary material (Figure S2). The mechanical properties of the different coatings were assessed by nanoindentation, and the resulting elastic modulus (E) and hardness (H) values are comparatively included in Fig. 4a and 4b, respectively, together with the uncoated substrate material (c.p.-Ti). The graphics show an initial regime where both H and E values are lower than those of the pristine substrate. This regime aligns with the 10 % rule of thumb commonly used for assessing the mechanical properties, minimizing the influence of the substrate. Thus, the elastic moduli for the three coatings obtained by extrapolation to zero are similar (80 ±5 GPa). This represents a reduction of more than 30 % of Young’s modulus of the c.p.-Ti specimen (120 ±5 GPa), which would contribute to decrease the stress shielding phenomena between the Ti implant and the bone. Moreover, comparing this result with our previous work [14], the use of this TiNbZrTa alloyed target proves to be advantageous compared to TiAlV coatings, prepared by the same sputtering deposition technique [3]. The reduction in elastic modulus values, in that case, varied between 8–20 % [14] compared to the values presented by the Ti substrate, depending on whether the p-DC or HiPIMS mode was used. This trend is coincident with the work of Cordeiro et al. whose values were 102, 139 and 82 GPa por c.p.-Ti, TiAlV and TNZT, respectively [27]. In terms of hardness, the developed coatings exhibit values that are comparable to the values reported for TNZT alloys [28,29], and slightly lower than pure titanium surface, (ranging from 4.1 to 4.7 ±0.1 GPa), with the highest values corresponding as expected to the film produced using HiPIMS and bias assistance. An important parameter to be considered for the implant’s lifetime is its wear resistance. The material resistance to abrasion, fatigue and adhesion forces are directly determined by the mechanical properties while corrosive wear is also conditioned by the electrochemical behavior. Thus, the formation of the Nb oxide containing passivation layer on β-Ti alloys has been reported as a positive factor to improve wear resistance [30]. Table 4 summarizes the measured mechanical properties, H/E and H 3 /E 2 values of the different coatings vs. the uncoated titanium specimen. The H/E ratio correlates with elastic strain to failure and wear resistance, and the H 3 /E 2 ratio is connected to the resistance against plastic deformation [31]. In the case of the coatings prepared in this work, H/E varied between 0.05 and 0.06, and H 3 /E 2 from 0.011 to 0.016. These values are comparable to those found in the literature for β-Ti alloys and Ti6Al4V alloys subjected to laser treatments to improve wear resistance [28,32] and represents a significant improvement in respect to c.p.-Ti., particularly in the case of HiPIMS deposition with bias assistance (H/E =0.06; H 3 /E 2 =0.016). Another crucial parameter contributing to the success of an implant is the surface wettability. Hydrophilic titanium-based implant surfaces facilitate improved cell adhesion while hydrophobic surfaces are commonly known for the repellent response to bacteria adhesion. However, when the material is working for long-term or even under cycling in physiological environments, surface wettability through the interaction with water may not be a good reference. In this work, wettability has been evaluated both in water and bovine serum media. The coatings developed using different deposition techniques exhibit a quite similar surface response to the interaction with liquids of different nature as presented in Fig. 5. Wettability of the coatings using milli-Q water droplets results in comparable WCA values around 80◦, as Table 4 Values of Young’s modulus (E), hardness (H), and the ratios H/E and H 3 /E 2 measured by nanoindentation for the coatings deposited via p-DC, HiPIMS and HiPIMS+bias, and the bare substrate (c.p.-Ti). E (GPa) H (GPa) H/E H 3 /E 2 p-DC 80 ±5 4.1 ±0.1 0.05 0.011 HiPIMS 80 ±5 4.1 ±0.1 0.05 0.011 HiPIMSþbias 80 ±5 4.7 ±0.1 0.06 0.016 c.p.-Ti 120 ±5 5.1 ±0.1 0.04 0.009 Note: The H values were estimated at ≈300 nm for p-DC samples and ≈100 nm for HiPIMS samples, following the 10 % rule of thumb. E values were determined by extrapolating the linear portion of the curve to zero nm. Fig. 5. Measured contact angles on the surfaces coated by TNTZ using water and bovine albumin serum as wetting agents. The uncoated c.p.-Ti is also included as reference. J.C. S´ anchez-L´ opez et al. Applied Surface Science 680 (2025) 161366 6 shown in Fig. 5a, being relatively lower in the case of the HiPIMS+bias deposition. This quasi-hydrophobic state, compared with the marked hydrophilic response of the uncoated c.p.-Ti implant (WCA of 48◦), is consistent with previous findings obtained with the same fabrication techniques for Ti6Al4V coatings [14]. When the TNZT surfaces are wetted with droplets of a biological medium simulating fluid, such as bovine serum albumin protein (Fig. 5b), the measured contact angles varied more significantly depending on the technique. They ranged from 62◦for the coating fabricated with the p-DC mode to around 75◦for HiPIMS deposition, regardless of whether bias is applied or not. This result indicates that the wetting properties of TNZT coatings prepared by HiPIMS have little effect on changing the substrate’s affinity for wetting with bovine serum while the p-DC surface (with higher average roughness) improves this affinity. This is unlike Ti6Al4V coatings in our previous work, which presented higher contact angle values [14]. Coatings with lower surface tension would align with presenting some resistance to bacteria interaction and inhibiting thrombus formation by preventing platelet accumulation [33,34]. Thus, by disregarding the differences that may arise from the chemical nature developed using p-DC or HiPIMS techniques (as discussed in the next section), it becomes apparent that the less rough surface with fewer defects produced by HiPIMS seems to induce a slightly more hydrophilic and less proteinphilic character, suggesting better bioand hemo-compatible behavior [35]. Additionally, surface chemistry significantly influences the interactions between surfaces and their environment, affecting bio and wear response. Analysis of the surface chemical composition by XPS is summarized in Table 5, including both the initial data and after Ar + etching to investigate beneath the outermost surface region. The most significant conclusion is the high contribution of oxygen (more than 50 at.%) which is not removed after mild Ar + bombardment whilst carbon content tends to decrease. This is indicative of the formation of a surface oxide layer by outdiffusion of the metals forming the TNZT alloy, mainly Ti, Nb and Zr as expected [27]. The low atomic concentration of Ta (<2 at.% as measured by EDX) along with the high oxide coverage formed through the major elements present in the alloy rendered this element undetectable. Comparing the Nb/Ti and Zr/Ti among the three coatings, differences in the outward diffusion of metals are noticeable depending on the type of sputtering process. Thus, the highest values are obtained for the sample prepared by HiPIMS with bias assistance, indicating enrichment in Nb and Zr in the passivation layer. Comparing with the bulk atomic Nb/Ti and Zr/Ti ratios measured by EDX (cf. Table 2, at.%), there is a clear increased outdiffusion of Zr (a factor between 2.5 and 5) for the three coatings, and Nb for the HiPIMS+bias sample (factor 1.5 to 2). These chemical variations should be considered together with the nature of the formed oxide on the corrosion and biocompatibility of the surfaces. XPS survey analysis was therefore conducted at the Ti 2p, Zr 3d and Nb 3d photoelectron peaks to establish the oxidation state of these elements in the outermost layer of the three coatings. Fig. 6 exhibits the high-resolution Ti 2p, Zr 3d and Nb 3d photopeak doublets for p-DC, HiPIMS and HiPIMS+bias, both in the initial state and after Ar + etching. No metal–metal bonding was identified in none of the deposited coatings and the binding energy positions align with oxidized states in all cases. Fig. 6a shows the main component of XPS Ti 2p peak centred at 458.6 eV of binding energy. This value is typical for titanium bonded to Table 5 Atomic composition of TNZT surface coatings fabricated by p-DC, HiPIMS and HiPIMS+bias before and after a mild Ar + etching treatment. Initial % at. (TNZT) C O Ti Zr Nb N Nb/Ti Zr/Ti p-DC 32.0 48.3 13.9 2.4 3.4 −0.24 0.17 HiPIMS 32.2 47.8 12.9 2.4 3.3 1.4 0.26 0.19 HiPIMSþbias 41.1 44.1 7.6 2.0 3.4 1.8 0.45 0.26 After Ar + sputtering % at. (TNZT) C O Ti Zr Nb N Nb/Ti Zr/Ti p-DC 28.7 48.8 15.7 3.9 2.9 −0.18 0.25 HiPIMS 25.0 52.9 14.9 2.3 4.0 1.0 0.27 0.15 HiPIMSþbias 9.4 63.7 12.4 3.7 7.7 3.3 0.62 0.30 Fig. 6. XPS high energy resolved spectra of the different coatings (p-DC, HiPIMS and HiPIMS+bias): a) Ti 2p, b) Nb 3d and c) Zr 3d. J.C. S´ anchez-L´ opez et al. Applied Surface Science 680 (2025) 161366 7 oxygen in TiO 2 phase. This assignment is further supported with the corresponding O 1 s photoelectron peaks shown in Fig. S3a from the supporting information. Comparing these oxygen signals (Fig. S3a) in the as-deposited state with those obtained with Ti6Al4V surfaces [14], it is evident that the hydroxide component is not present in this case. Other sub-oxidized titanium species, such as Ti 2 O 3 , appear just below the outermost surface region as a broad contribution at lower binding energies (over 456.2 eV) for the HiPIMS+bias sample, as well as to a lesser extent, for the p-DC one [36]. The presence of Ti 3+ in the nonstoichiometric TiO 2 structure generates oxygen vacancies in order to achieve the electroneutrality, which influence negatively the corrosion properties. However, it is reported that the presence of Nb in the TNZT alloys helps to decrease the content of these anionic vacancies, improving the passivation characteristics of the oxide film [37]. Moreover, the radii similarity of Ti and Nb ions contributes that the replacement in the crystal lattice does not generate point defects, which also favour the corrosion rate. Fig. 6b presents the main contribution of the Zr 3d region centred at 182.3 eV. This binding energy is slightly lower than those typically reported for ZrO 2 species [27]. A comparable pattern is observed in the position of the main peak of the Nb 3d photoelectron peak (207.0 eV), which is located at energies slightly below that corresponding to Nb 2 O 5 (BE ~ 207.3 eV) [38,39]. These chemical species are stable after the ion bombardment cleaning except for the case of the HiPIMS+bias method, which displays a wide XPS signal in the Nb 3d peak (Fig. 6c), which extends to lower binding energies involving sub-oxidized contributions (NbO x around 203.9 eV) [40]. As a result, similar to the observations in the analysis of the titanium spectrum, the HiPIMS+bias conditions favoured the presence of sub-stoichiometric Nb oxides beneath the topmost layer, non-reaching the maximum oxidation state and most thermodynamically stable Nb 2 O 5 phase. In summary, the assembly of XPS results are indicative of the formation of a passivation oxide layer with TiO 2 as predominant phase, considering its atomic concentration in the alloy, which is nearly three times higher than the values corresponding to the other elements. Additionally, the Zr and Nb contents are increased in the oxide layer in respect to their ratio in the bulk alloy composition. Stable surface oxidation through the generation of a mixture of TiO 2 , Nb 2 O 5 , ZrO 2 and Ta 2 O 5 layer has been previously reported like biocompatible [28] and corrosion resistance [6–9] when compared to Ti or TiAlV alloys used for orthopedic implants. Nevertheless, the presence of anion vacancies, point defects and non-stoichiometric under certain synthesis conditions may induce changes in the electrochemical behaviour as we will discuss next. Electrical impedance spectroscopy (EIS) measurements were performed to semi-quantitatively evaluate potential changes in the electrochemical properties of the surfaces determining the bio-corrosion resistance [41,42].Fig. 7 illustrates a schematic representation of the electrical impedance measurements, considering the system c.p.-Ti substrate +TNZT coating as series resistors in a circuit. Fig. 8 shows the electrical impedance values |Z| for the bare titanium substrate and the effect of the different TNZT coatings at frequencies from 100 to 500 MHz. Previous works have also utilized a similar range of frequencies in EIS to evaluate the porosity in implants [41], to study the cellular growth on porous Ti implants [42] or to correlate the changes in the impedance values with the mechanical properties of cortical or trabecular bones [43]. At these frequencies, it is possible to detect variations in biological materials in the so-called γ–region [44], in particular on the aqueous content of biological species and small molecules. As expected, the impedance values are higher in the coated discs, as illustrated in Fig. 8a, and exhibit a linear dependence with the frequency, varying from approximately 150 to 500 mΩ[19,41]. These values are comparable to a fully dense bulk β−TNZT alloy, included in the same graph for comparison purposes, despite of their limited thickness (1 to 3 μ m). This demonstrates that the coatings we are obtaining are quite dense and achieve consistent chemical compositions, resulting in similar values of electrical impedance. However, the differences among the different β-Ti coatings and the bare Ti implant become more apparent at higher frequency values, when the effects of the microstructure and chemical composition are more relevant, as has been observed in previous works [19,41]. To better evaluate the differences in impedance properties of the TNZT coatings, the impedance value of the bare Ti substrate were firstly removed from the original measurements (Fig. 8b). Then these values were normalized by dividing them by the corresponding coating thickness (Fig. 8c). Upon removing the effect of the thick c.p.-Ti substrate, a significant decrease in impedance values is observed. Once normalized by the coating thickness, very high impedance values are observed (cf. Fig. 8c). These data treatments enable a more thorough evaluation of the coatings’behavior. The results indicate that p-DC deposited coating shows the lowest impedance values, whereas the HiPIMS deposited TNZT coatings demonstrate significantly higher impedance, multiplying by a factor of 4 to 7 depending on the process and testing frequency. On the other hand, in Fig. 8d, the electrical impedance results are normalized by the electrical impedance response of the titanium substrate, yielding a normalized impedance ratio. In this type of results analysis protocol, the effect of frequency on electrical impedance measurements is disregarded, denoting the same behavior of the coatings in all ranges of measured frequencies. In Fig. 8f, the impedance values are also dividing by the coating thicknesses to obtain the electrical resistance by length unit. Once again, higher values of impedance are observed for the HiPIMS coatings. There is a direct relationship between the surface state and the electrical properties of the coated implants, or in other words between electrical impedance results and corrosion resistance. The XPS surface analysis proved the formation of a surface protective passivation layer comprising a mixture of Ti, Nb and Zr oxides, serving to prevent the corrosion in biological media. Besides, the high values of WCA angles measured on TNZT coatings, around 80◦vs. 50◦in the case of c.p.-Ti, helps to reduce surface wetting phenomena, thereby inhibiting corrosive mechanisms. These experimental evidences justify the increased electrical impedance values of ß-coated Ti specimens. However, other factors as film microstructure, oxide type, oxide stoichiometry and oxide ratio inside the passivation layer must be considered to fully explain the distinct behavior depending on the method of synthesis employed. Thus, p-DC coating presents the lowest impedance values, which can be Fig. 7. Scheme of electrical impedance measurements considering the system c.p.-Ti substrate +TNZT coating as in series resistors in a circuit (M.S.: magnetron sputtering). J.C. S´ anchez-L´ opez et al. Applied Surface Science 680 (2025) 161366 8 related to the columnar microstructure with high intercolumnar porosity, in opposition to the denser morphology observed in the HiPIMS samples. The open structure easies the access of the oxygen to the titanium substrate, and therefore the behavior demonstrated by the p-DC is more similar to that of bare c.p.-Ti. Between the HiPIMS samples, a slightly poorer behavior is found if simultaneous bias voltage is applied to the substrate. The increased Nb and Zr diffusion in the passivation layer and the presence of suboxides NbO x and Ti 2 O 3 may account for this decrease in corrosion resistance as previously reported in non-stoichiometric and point defects oxides [37,45]. 4. Conclusions In this work, a β-Ti alloy composed was deposited by magnetron sputtering of Ti–35Nb–7Zr–5Ta (wt. %) on c.p.-Ti specimens prepared by conventional powder metallurgy for implant applications. Pulsed magnetron sputtering technologies (p-DC and HiPIMS, with and without bias assistance) under similar power conditions were applied leading to different film morphologies and functional properties (mechanical, surface chemistry and corrosion resistance). All the coatings presented a reduced Young’s Modulus (≤80GPa) compared to the c.p.-Ti specimen, representing a reduction of E of more than 30 %. This decrease can Fig. 8. a) Electrical impedance measurements (|Z|) of the bare c.p.-Ti substrate vs. TNZT coated c.p.-Ti substrates. A fully dense bulk ß-TNZT alloy (FD TZNT) is included for comparison purposes; b) Electrical impedance values of the TNZT coatings after subtraction of bare c.p.-Ti substrate, |Z C |=|Z|-|Z c.p.-Ti |; c) Normalized values of impedance for the TNZT coatings considering the coating thickness (t), |Z C/t |=|Z C |/t; d) Normalized impedance ratios considering the impedance of the bare c.p.-Ti substrate, |Z N |=|Z|/|Z c.p.-Ti |; e) Normalized impedance ratios of the TNZT after subtraction of the bare c.p.-Ti contribution, |Z NC |=|Z N |-1; f) Normalized impedance ratios of the TNZT coatings by the coating thickness, |Z NC/t |=|Z NC |/t. J.C. S´ anchez-L´ opez et al. Applied Surface Science 680 (2025) 161366 9