Dataset and publication "Multilayer Design of Sustainable Multifunctional Zr-Cu-N Coatings: A Route for Enhanced Mechanical and Antibacterial Performance"

Multilayer design of sustainable multifunctional Zr–Cu–N coatings: A route for enhanced mechanical and antibacterial performance Rostislav Daniel a,* , Tobias Ziegelwanger a , Michal Zítek a , Michaela ˇ Cerven´ a b , Stanislav Haviar b , Michael Meindlhumer a , Pavel Baroch b , Jozef Keckes c , Petr Zeman b a Department Materials Science, Montanuniversit¨ at Leoben, Franz-Josef-Straße 18, 8700 Leoben, Austria b Department of Physics and NTIS – European Centre of Excellence, University of West Bohemia in Pilsen, Univerzitní 8, 301 00 Pilsen, Czech Republic c Erich Schmid Institute for Materials Science, Austrian Academy of Sciences, Jahnstrasse 12, 8700 Leoben, Austria ARTICLE INFO Keywords: Zr–Cu–N Metallic glass Multilayer architecture Multifunctional coating Damage-tolerant material Antibacterial activity Micromechanical testing Sputter deposition ABSTRACT Wear-resistant protective coatings with antimicrobial activity are essential for durability and hygiene in healthcare, public spaces, food industry, consumer products, and industrial environments. This study developed sustainable multifunctional Zr–Cu–N coatings with exceptional damage tolerance, and antibacterial properties using non-reactive and reactive sputtering of only two elemental Zr and Cu targets without external heating. The coatings’ superior performance stems from a sophisticated multilayer architecture combining elastic ZrCu metallic glass, hard and stiff ZrN ceramic, and hard and tough ZrN–Cu nanocomposite coatings. Each constituent was optimized for composition and mechanical properties before integration into multilayer structures to provide high damage tolerance and antibacterial functionality. Antibacterial efficacy was tested in a high-traffic environment over 60 days, showing consistent antimicrobial performance. Fracture stress and toughness were assessed through in situ bending experiments on microcantilever beams fabricated by focused ion beam milling. Results revealed that optimizing the thicknesses of ductile and stiff sublayers significantly enhances damage tolerance while maintaining high hardness and wear resistance. The incorporation of Cu in an unbonded state within the ZrN–Cu nanocomposite facilitates sustainable and scalable production of these multifunctional coatings with antibacterial properties, making them ideal for large surface applications in high-traffic environments like hospitals, office buildings, and public transport. 1. Introduction The growing demand for sustainable production and consumption necessitates innovative scientific strategies to ensure that modern materials maintain high mechanical performance and thermal stability, even with fewer alloying elements. While multi-principal element alloys exhibit exceptional properties under a variety of challenging conditions [1,2], their production can have significant environmental, economic, and societal impacts due to their complex compositions, often using alloying elements with problematic sourcing [3]. Reducing the number of elements in alloys and compounds can decrease the need for extensive mining operations, which are often environmentally detrimental, and also facilitate more efficient and cost-effective recycling, resulting in lower energy consumption and reduced greenhouse gas emissions during production. The strategy of reducing elemental complexity in modern materials, aligned with sustainable production, can only be effective if the materials’ properties are maintained or enhanced. Complex microstructures have proven highly effective in enhancing the mechanical strength and ductility of multi-phase, multi-principal element alloys [4–7] as well as advanced steels [8,9] Al- [10,11] or Ti-alloys [12,13]. These microstructures feature engineered grain boundaries and interfaces, with coexisting phases contributing to controlled deformation across multiple length scales. However, synthesizing such structurally complex materials using traditional thermodynamic processes or thermal annealing techniques is challenging. In contrast, physical vapor deposition (PVD) is a well-established and widely used technique for synthesizing nanostructured materials with complex microstructures and phase * Corresponding author. E-mail address: [email protected] (R. Daniel). Contents lists available at ScienceDirect Materials & Design journal homepage: www.elsevier.com/locate/matdes https://doi.org/10.1016/j.matdes.2025.114037 Received 10 February 2025; Received in revised form 16 April 2025; Accepted 30 April 2025 Materials & Design 254 (2025) 114037 Available online 2 May 2025 0264-1275/© 2025 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). compositions [14–18]. Specifically, magnetron sputter deposition employs thermodynamically non-equilibrium processes associated with high-energy particle fluxes and high cooling rates at the substrate to create structurally heterogeneous coatings with high hardness and thermal stability [19–22]. Zirconium nitride (ZrN) is a typical representative of transition metal nitrides synthesized by PVD used in various applications due to its excellent properties. High hardness, corrosion resistance, wear resistance and thermal stability of ZrN make it suitable as a protective coating for cutting tools, molds and dies [23–25]. Additionally, ZrN is used as a protective coating for biomedical implants and devices, such as orthopedic and dental implants, due to its high biocompatibility [26,27]. However, ZrN coatings can fail through cracking, delamination, spallation, and substrate deformation under heavy mechanical loading, primarily due to their brittle ceramic nature. Furthermore, ZrN lacks the multifunctionality required for many advanced applications. Alloying with immiscible elements that segregate to grain boundaries has proven very effective in promoting the formation of nanocrystalline structures by inhibiting grain growth during synthesis [28,29]. Specifically, the microstructure of ZrN coatings can be manipulated through alloying with Cu, which induces a transformation from a columnar to a fine-grained nanocomposite structure by encapsulating ZrN crystallites with Cu. The crystallite size decreases with the amount of Cu in the coating, allowing control over the coating microstructure and, consequently, its mechanical properties [30–32]. The primary effect of microstructure on hardness is associated with the grain-boundary-mediated processes (grain boundary sliding and rotation) and restricted dislocation movement at the grain boundaries under mechanical loading [33]. Concurrently, the presence of the ductile Cu phase enhances toughness [34]. High fracture toughness in brittle materials can also be achieved by combining them with elastic/soft constituents in a multilayer architecture [35–37]. Such composite materials, with microscale structural heterogeneity and spatial variations in mechanical properties, exhibit higher damage tolerance due to the effective reduction of crack driving force at the interfaces with the elastic/soft constituents and the deflection of propagating cracks. This effect increases with the increasing difference in the elastic modulus of the hard and soft constituents [38], making the selection of the elastic material crucial. Metallic glasses are known for their unique combination of high strength and excellent elasticity, arising from their disordered atomic structure [39,40]. This lack of crystallinity prevents the movement of dislocations that typically lead to severe plastic deformation in crystalline metals. Hence, Zr–Cu thin-film metallic glass is a suitable candidate as the elastic constituent in combination with ZrN to form a damage-tolerant multilayer architecture. Attention should also be given to both Zr–Cu–N and Zr–Cu material systems due to known antimicrobial activity of Cu against microorganisms [41]. This activity can result from physical interaction of Cu atoms with cell or virus plasma membranes or through secondary effects [42]. Physical interaction is particularly effective if nano-sized Cu particles attach to the membrane and infiltrate the cell [43]. Reactive oxygen species, formed by the reduction of copper, can directly damage lipids, proteins, and DNA through oxidative interactions [44]. Alternatively, Cu ion release can promote the formation of reactive oxygen species, causing cell damage either directly by damaging cell membranes or indirectly by causing oxidative stress [42]. However, this antimicrobial activity is only present if Cu can undergo ionization. Zr–Cu thin-film metallic glasses prepared by magnetron sputter deposition, with Cu fully bonded and homogeneously distributed in the amorphous matrix, do not exhibit antibacterial effects [45]. In contrast, ZrN–Cu nanocomposite coatings with a Cu content higher than 10 at. % have been successfully tested for antibacterial activity [34]. These results indicate the necessity of a heterogeneous distribution of Cu in the matrix with Cu-rich regions where not all Cu atoms are bonded to Zr or N. This explains the different behavior in reported antibacterial activity of Zr–Cu and ZrN–Cu materials. In summary, this study demonstrates a novel and sustainable materials design strategy based on the use of only two metallic elements, Zr and Cu, combined in both low-temperature non-reactive and reactive magnetron sputtering processes using argon and nitrogen-containing atmospheres. This streamlined approach enables the formation of a variety of binary and ternary compounds with a broad spectrum of compositions and microstructures. It not only simplifies material selection and deposition procedures but also reduces the environmental and economic impact of production, supporting industrial scalability and efficient large-area coating. The key innovation lies in the development of a hierarchical multilayer architecture that integrates elastic ZrCu thin-film metallic glass, hard nanocrystalline ZrN, and nanocomposite ZrN–Cu into a single coating system. This architecture results in a multifunctional material that combines high hardness, mechanical damage tolerance, and long-term antibacterial activity. The optimization of the deposition conditions, multilayer architecture, and Cu content has been demonstrated as an essential workflow to synthesize these sustainable, compositionally, and structurally complex yet scalable coatings with balanced functional properties. The demonstrated longterm mechanical durability and antibacterial performance highlight the ZrN–Cu/ZrCu/ZrN multilayer coating system as a promising candidate for surface protection in heavy-traffic environments such as hospitals, public transportation, office buildings, and other public infrastructure. 2. Materials and methods 2.1. Coating synthesis Monolithic and multilayer coatings from the ternary Zr–Cu–N system were deposited by magnetron sputtering using an AJA International ATC 2200-V system, equipped with four unbalanced magnetrons with circular, indirectly cooled sputtering targets (50.8 mm in diameter). The system was evacuated with a turbomolecular pump (1200 l/s) backed by a multi-stage roots pump (27 m 3 /h). The base pressure achieved before each deposition was below 5 ×10 −5 Pa. The depositions were performed at a total pressure of 0.533 Pa (4 mTorr) on polished and ultrasonically pre-cleaned single-crystalline Si(100) substrates without any external heating. The substrates were rotated at a speed of 40 rpm at a target-tosubstrate distance of 150 mm. A radio frequency (RF) bias power of 50 W was applied to the substrate holder. For Cu-containing monolithic coatings and sublayers, two magnetrons with 6 mm thick Zr and Cu targets of 99.5 % and 99.99 % purity were used, respectively. Cu-free monolithic coating and sublayers were deposited using only the Zr target. For N-containing monolithic coatings and sublayers, reactive depositions were performed in a gas mixture of argon and nitrogen with constant flow rates of 50 sccm and 5 sccm, respectively. N-free monolithic coating and sublayers were deposited non-reactively, at an argon flow rate of 50 sccm only. The magnetron with the Zr target was powered by an Advanced Energy Pinnacle Plus +5/5 kW power supply operating in a DC regime, while the magnetron with the Cu target was powered by a Hüttinger Elektronik TruPlasma Highpulse 4002 power supply operating in a highpower impulse regime at an average target pulse power density of 1000 W/cm 2 . The deposition rates were monitored by a quartz crystal sensor (SQM-160, Inficon) and adjusted before each deposition by varying the DC discharge power on the Zr target or the repetition frequency on the Cu target to achieve the desired elemental composition of the monolithic coatings and sublayers. 2.2. Structural analysis The structure of the coatings was characterized by X-ray diffraction (XRD) using a PANalytical X’Pert PRO MPD diffractometer with a CuK α (λ =0.154187 nm) radiation and scanned with an ultrafast detector R. Daniel et al. Materials & Design 254 (2025) 114037 2 X’Celerator. The measurements were performed in the Bragg-Brentano geometry at a scanning speed of 0.1◦/s. The microstructure of the coatings in cross-section was observed by a Hitachi SU-70 scanning electron microscope operated at a primary electron energy of 5 keV. The Si(100) substrate with the deposited coating was fractured at ambient conditions to enable imaging of the microstructure. 2.3. Elemental composition The elemental composition of the coatings was measured using a scanning electron microscope (Hitachi SU-70) equipped with wavelength-dispersive spectroscopy. The instrument was operated at a primary electron energy of 15 keV. For quantitative analysis, Zr, Cu, and nitrides standards were utilized, and a conservative error margin of 1 at. % was considered. 2.4. Micromechanical testing The hardness and elastic modulus of the coatings were evaluated from load-displacement curves measured at room temperature using a microindentation system (Fischerscope H100) equipped with a Vickers diamond tip. To mitigate the impact of thermal drift, the load–displacement curves were recorded during thermal equilibrium. Indentations were performed in a load-controlled manner with a load of 10 mN. Microcantilevers for micromechanical testing were fabricated using focused ion beam (FIB) milling (FEI Versa 3D), with dimensions of 1.8 ± 0.2 μ m in thickness (t, corresponding to the coating thickness), 2.4 ± 0.1 μ m in width (B), and 15.4 ±1.2 μ m in length (L). The initial milling was performed with a Ga +ion beam (30 keV, 7 nA), followed by fine milling (30 keV, 1 nA). For fracture toughness evaluation, microcantilevers were notched using a lower milling current (30 keV, 50 pA) approximately 2 μ m from the microcantilever support. The notch does not span the entire width of the microcantilever, leaving material bridges less than 100 nm wide along the sidewalls. These bridges create a sharp pre-crack after fracture under load. In situ micromechanical testing was performed in a ZEISS LEO 982 SEM, using a Hysitron PI 85 nanoindenter, equipped with a 700 nm sphero-conical indenter tip (Synton MDP AG). The cantilevers were loaded at a quasi-static speed of 20 nm/s using a displacementcontrolled feedback loop at a distance l of approximately 10 μ m from the microcantilever support (precisely measured for each experiment). The measured load (P) vs. displacement (w) curves were corrected for the instrument compliance and the elastic modulus (E) of the coatings was calculated as E=4 B×dP dw ×(l t)3 .(1) Using linear fracture mechanics, the fracture stress σ F was calculated using the maximum applied load at fracture according to σ F=6Pmaxl Bt2.(2) After testing, the fracture cross-sections were analyzed in a ZEISS LEO 1525 SEM and the notch depth (a) was measured. This enabled the calculation of the fracture toughness (KIC)by KIC = σ F π a √×Y(a t),(3) where Y(a t)is the dimensionless shape factor accounting for the notch shape in a freestanding cantilever geometry [46]. To obtain reliable values for Young’s modulus (E), fracture stress ( σ F), and fracture toughness (K IC ), at least four notched and unnotched microcantilevers of each coating were tested and analyzed. 2.5. Antibacterial activity The antibacterial activity of the Zr–Cu–N coatings was evaluated based on the growth and survival of the enterobacteria E. coli (NiCo21 (DE3)) placed as a solution on the coating surface of the tested sample. Before each application of the bacterial solution, the coating surface was cleaned with a 70 % ethanol solution and dried. Subsequently, 5 µl of the bacterial solution was applied to the coating surface and a cover glass was attached. The prepared sample was then placed in a humidity chamber for a specified time (90 min). Afterwards, the coating surface and the cover glass were rinsed with 5 ml of physiological saline solution. From this solution, 150 µl was taken after dilution and transferred to a preheated Petri dish with nutrient agar and incubated for 12 h at 37 ◦C. After incubation, the colonies grown on each Petri dish were counted, and the antibacterial efficacy was calculated using the following formula: E [%] =(A −B)/A * 100, where A is the number of bacterial colonies grown on the negative reference (bare substrate), and B is the number of colonies grown on the tested sample (coated substrate). Note that an efficacy of 100 % indicates a complete absence of bacterial colonies after incubation, meaning all bacteria were eliminated. The entire procedure for measuring antibacterial activity was repeated three times. 3. Results Three monolithic and three multilayer Zr–Cu–N coatings of different elemental compositions, microstructures (Fig. 1), crystallographic structures (Fig. 2) and mechanical properties (Fig. 3) were selected and subjected to systematic investigation. The obtained results are presented in this section. 3.1. Monolithic coatings Two of the monolithic coatings are binary ZrCu and ZrN coatings. The ZrCu coating contained 46 at. % Cu and 54 at. % Zr and the ZrN coating was stoichiometric with equal amounts of Zr and N. Apart from their different elemental compositions, these coatings possess distinctly different structural and mechanical properties, which is desirable for their combination in structurally complex hierarchical multilayer Zr–Cu–N coatings. As shown in Figs. 1a and 2a, the ZrCu coating is X-ray amorphous with a microstructure typical for metallic glasses. It is characterized by vein-like features and striations on cross-sectional micrographs. Additional evidence for metallic glass behavior is the glass transition observed using differential scanning calorimetry, as reported in our previous papers [47,48]. In contrast, the ZrN coating exhibits a crystalline structure with a face-centered cubic (fcc) rock salt lattice (Fig. 2a) and a columnar microstructure (Fig. 1b). Near the coating/substrate interface, the columnar grains are fine and develop rapidly, maintaining an almost constant width throughout the entire coating thickness. This behavior is attributed to the competitive growth of crystalline islands formed during coalescence and the establishment of growth conditions in subsequent stages, during which crystallites with the preferred (111) orientation predominantly form, overgrowing crystallites with other orientations [49–51]. Consequently, the crystallites with the (111) orientation dominate the coating structure (evidenced by the peak at 33.5◦2θ in Fig. 2a), with only minor contributions of the (200)-oriented (peak at 39.1◦2θ), (220)-oriented (peak at 56.7◦ 2θ) and (311)-oriented crystallites (peak at 68.0◦2θ). The difference in the nature of these two alloys is also reflected in their mechanical properties (Fig. 3). Due to its metallic character with delocalized electrons, the ZrCu metallic glass coating exhibits moderate hardness (H =5.5 GPa) and high elasticity (E f =99 GPa). In contrast, stoichiometric ZrN is a hard and stiff ceramic material with combined metallic and covalent character of bonding and hardness of 21 GPa and effective elastic modulus of 207 GPa. Figs. 4a, b and 5a, b show the results of microcantilever beam R. Daniel et al. Materials & Design 254 (2025) 114037 3 bending experiments on the ZrCu and ZrN coatings, respectively. While the ZrN coating fractured in a brittle manner within elastic deformation (Fig. 4b), the ZrCu coating deformed without fracturing (Fig. 4a) but with the emission of shear bands. This behavior clearly emphasizes a high inelastic deformation capacity of the ZrCu metallic glass at high strains. During bending, two distinct stress zones form in the microcantilever beam: a tensile stress zone on the outer side near the surface (localized close to the microcantilever support) and a compressive stress zone below the neutral axis. When localized shear strain exceeds the critical value for inelastic deformation in ZrCu, shear bands form and propagate from the tensile stress zone toward the compressive stress zone (Fig. 4a). The initiation and propagation of these shear bands accommodate inelastic deformation in the compressive stress zone and contribute to a high damage tolerance of the material, while cracks develop in the upper part of the microcantilever beam (Fig. 4a), where the highest tensile stress develop. Although the formation of shear bands and subsequent cracks at high loads typically leads to brittle fracture, the critical stress was not reached at the maximum displacement of this experiment until a pre-crack was induced in the microcantilever by FIB. Even then, the crack initiated at the pre-crack tip in the tensile zone did not propagate through the entire coating thickness (Fig. 5a), as energy dissipation through shear band formation in the compressive zone prevented further crack propagation. The evaluation of the stress–strain curves and fracture surfaces (considering that the unnotched microcantilever beam did not fracture at the maximum displacement) yields fracture stress values of 3.9 GPa and fracture toughness values of 3.1 MPa•m 1/2 . In contrast, the covalent bonding of ZrN and its columnar microstructure contribute to brittle fracture, which is significantly promoted by introducing of a pre-crack in the microcantilever beam (Fig. 5b), resulting in Fig. 1. Cross-sectional SEM micrographs showing the microstructure and the architecture of (a–c) monolithic and (d–f) multilayer Zr–Cu–N coatings. Fig. 2. XRD patterns of (a) monolithic and (b) multilayer Zr–Cu–N coatings, with reference powder diffraction standards for fcc ZrN (PDF card #00-035-0753) and fcc Cu (PDF card #00-004-0836) included for comparison. R. Daniel et al. Materials & Design 254 (2025) 114037 4 a fracture stress of 3.5 GPa and a fracture toughness of 1.9 MPa•m 1/2 . The elastic modulus of 82 GPa for ZrCu and 197 GPa for ZrN (Fig. 6a), as evaluated from the microcantilever beam bending experiment, are consistent with the indentation results in Fig. 3a. The reactive sputter deposition in nitrogen-containing plasma discharge from Zr and Cu targets enabled to form a nanocomposite microstructure composed of crystalline ZrN and Cu (Fig. 1c). The deposition conditions were optimized to incorporate 10 at. % Cu into the ZrN matrix, resulting in a Zr 45 Cu 10 N 45 nanocomposite composed of approximately 80 mol. % ZrN and 20 mol. % Cu (assuming N is preferentially bonded to Zr). This amount of Cu is sufficiently high to encapsulate ZrN crystallites, which are subsequently small in size (Fig. 1c), globular in shape [31] and randomly oriented (Fig. 2a). Previous TEM investigations revealed a uniform distribution of Cu, which forms a thin disordered tissue embedding the ZrN crystallites [32], which is also indicated by a low-intensity broad peak in the ZrN-Cu diffractogram at 43.3◦2θ. The segregation of Cu to the grain boundaries of ZrN resulted in a reduced effective elastic modulus of the ZrN-Cu coating (E f =180 GPa) compared to ZrN, and the hardness was also slightly reduced (H =19 GPa), see Fig. 3. The hardness comparable to that of ZrN is attributed to the high N content (45 at. %), which also preserves the covalent character of bonding, as well as the nanocomposite structure. These mechanical properties of the ZrN–Cu nanocomposite are further evident in its mechanical response during the microcantilever beam bending experiment. Unlike the fracture surfaces of ZrN, the unnotched microcantilever beams of the nanocomposite exhibit a distinctive fracture behavior. The crack propagates through the microcantilever but abruptly changes direction in the compressive stress zone, as indicated by the curved fracture path in the lower half of the cantilever. This is due to the crack deflecting to regions with lower compressive stress (Fig. 4c). Consequently, the fracture stress significantly increases to 5.2 GPa (Fig. 6b). However, this effect is greatly suppressed when stress is localized at the pre-crack tip (Fig. 5c), resulting in a flat fracture surface similar to that of ZrN and comparable fracture toughness values (Fig. 6c). 3.2. Multilayer coatings The aim of the study was also to combine hard and brittle compounds formed from the Zr and Cu targets by either reactive or non-reactive processes in various multilayer architectures, and to explore their potential to enhance the damage resistance of the Zr–Cu–N system without compromising its high strength and hardness. For this purpose, three Zr–Cu–N multilayer coatings were deposited, featuring varying thicknesses of metallic ZrCu sublayers (~20 and ~50 nm) and compositions of nitride sublayers (ZrN and ZrN–Cu), with approximately the same thickness of 200 nm for the nitride sublayers (Fig. 1d–f). The internal stress for all these coatings, measured from the bending of coating/ substrate couples using a modified Stoney’s formula, was compressive and approximately the same (−2.2 GPa). This value is also comparable to the stress in the monolithic ZrN and Zr–Cu–N coatings (−2.3 and Fig. 3. Mechanical properties of monolithic and multilayer Zr–Cu–N coatings including values of (a) effective Young’s modulus and (b) hardness, both measured by the indentation technique. Fig. 4. Fracture surfaces of unnotched cantilevers for (b and c) monolithic and (d–f) multilayer Zr–Cu–N coatings after the in situ SEM microbending tests. Image (a) shows the monolithic ZrCu coating, which deformed without fracturing up to reaching the instrument’s maximum displacement. R. Daniel et al. Materials & Design 254 (2025) 114037 5 −2.1 GPa, respectively), while it is much higher compared to the metallic ZrCu coating (−0.2 GPa). A complete overview of the architecture and overall thicknesses of the multilayer coatings, which were designed to be consistent, is provided in Table 1. Fig. 2b shows the effect of ZrCu sublayers on the growth of ZrN and ZrN–Cu. Due to its amorphous structure, ZrCu does not exhibit a pronounced templating effect, unlike the growth of transition metal nitrides on crystalline metals [49]. The difference in the crystallographic orientation of ZrN in the multilayer structure compared to the monolithic coating can primarily be attributed to the dimensional restrictions imposed by the sublayer thickness. Our previous studies have shown that a complete change in the crystallographic orientation from (2 0 0) to (1 1 1) during competitive crystallite growth requires a minimum thickness of about 1.5–2.0 μ m [49,50,52]. The 200-nm thickness of the ZrN sublayers only allows for a partial texture crossover (see the diffractograms of the ZrN monolithic and M01 multilayer coatings in Fig. 2a and b, respectively). The difference in the crystallographic structure of the M01 and M02 multilayer coatings suggests, however, that ZrCu affects adatom mobility during the initial growth stages and microstructure development as the sublayer thickness increases. Since the orientation of the crystallites in the ZrN–Cu nanocomposite is very similar to that of ZrN in the M02 multilayer coating, the diffractograms of the M02 and M03 multilayer coatings do not significantly differ (Fig. 2b). A distinct change in the microstructure of these two multilayer coatings is, however, visible in the cross-sectional micrographs in Fig. 1e and f. While the ZrN sublayers exhibit a pronounced columnar microstructure that repeatedly re-nucleates at each ZrCu sublayer and rapidly Fig. 5. Fracture surfaces of notched cantilevers of (b and c) monolithic and (d–f) multilayer Zr–Cu–N coatings after the in situ SEM microbending tests. Image (a) shows the monolithic ZrCu coating, which deformed without fracturing up to reaching the instrument’s maximum displacement. Fig. 6. Mechanical properties of monolithic and multilayer Zr–Cu–N coatings including values of (a) Young’s modulus, (b) fracture stress and (c) fracture toughness, all measured by in situ SEM microbending tests. Table 1 Overview of the architecture of multilayer Zr–Cu–N coatings summarizing relevant structural parameters. M01 M02 M03 Coating architecture ZrN/ZrCu ZrN/ZrCu ZrN–Cu/ZrCu/ZrN ZrN sublayer thickness 215 nm 205 nm 185 nm ZrCu sublayer thickness 20 nm 50 nm 50 nm Total thickness 2.1 µm 2.0 µm 1.8 µm Number of ZrN sublayers 9 8 6 Number of ZrCu layers 9 8 8 Number of ZrN–Cu layers −−2 ZrCu volume fraction 8.5 % 20.0 % 22.2 % R. Daniel et al. Materials & Design 254 (2025) 114037 6 develops within the first 10–20 nm of thickness, both top ZrN–Cu nanocomposite sublayers of the M03 multilayer coating exhibit a uniform fine-grained microstructure throughout their thickness (Fig. 1f). The size of the columnar ZrN grains in all three multilayer coatings is very similar, which is also indicated by the similar width of the peaks in the diffractograms (Fig. 2b). Indentation of the multilayer coatings provides a composite value that is affected by the hardness and elastic modulus of all constituents, which plastically and elastically deformed during the experiment. Increasing the volume fraction of ZrCu in the multilayer coatings from 8.5 % (M01) to 20.0 % (M02) and 22.2 % (M03) (determined by the thickness and number of ZrCu sublayers) reduced the hardness of the multilayer coatings from 14.0 GPa to 12.0 GPa and 11.5 GPa, respectively (Fig. 3b). The decrease in hardness of the multilayer coatings compared to the ZrN monolithic coating corresponds to this increase in volume fraction and is attributed to the relatively low hardness of ZrCu (Fig. 3b). Similarly, the effective elastic modulus of ZrCu of 99 GPa contributed to the reduction of the effective elastic modulus of the multilayer coatings to 180 GPa (M01), 160 GPa (M02) and 157 GPa (M03). In general, the optimized process conditions established in our previous studies ensured good mechanical performance and adhesion of the ZrN–Cu films [30,31]. Moreover, the reduction of intrinsic compressive stress −along with effective stress relaxation facilitated by the alternating hard and elastic layers in the multilayer architecture −contributed to an overall enhancement of the mechanical properties. Unlike the relatively smooth fracture surfaces observed in the monolithic coatings, the fracture surfaces of all multilayer coatings exhibit a distinctly irregular, rough, faceted to terraced appearance. These terraces, visible across the entire fracture surface of the multilayer coatings (Fig. 4d–f), are indicative of crack deflection at the interfaces where the ZrN (or ZrN–Cu) sublayers meet the ZrCu sublayers. In addition to the terraces that span the coating thickness from the surface to the substrate, pronounced irregularities with a stepped appearance formed within individual ZrN sublayers in the in-plane direction, aligned with the columnar microstructure of ZrN. A similar stepped morphology is evident in the fracture surfaces of notched microcantilever beams from the multilayer coatings (Fig. 5d–f). Consequently, the fracture stress and fracture toughness values, derived from the stress–- strain curves, were tens of percent higher than those of the ZrN and ZrN–Cu monolithic coatings (except for fracture stress of the ZrN–Cu coating, which was higher; Fig. 6b). Specifically, the M01 multilayer coating exhibited fracture stress and toughness of 4.6 GPa and 2.5 MPa⋅m 1/2 , respectively; the M02 multilayer coating, 3.9 GPa and 2.1 MPa⋅m 1/2 ; and the M03 multilayer coating, 4.2 GPa and 2.2 MPa⋅m 1/2 . The elastic modulus of the multilayer coatings, reaching 257 GPa (M01), 188 GPa (M02), and 172 GPa (M03), was also significantly affected by the multilayer architecture. 4. Discussion The protection of surfaces against wear and corrosion by thin films and coatings is a well-established approach used to improve the performance and lifetime of tools [53–55], medical devices [56–58], engine components, bearings and gears [59–61], and have also found applications in aerospace and energy sector [62–64]. For these applications, hard ceramic coatings are commonly used as protective coatings due to their high hardness and wear resistance. However, their widespread use usually encounters the problem of catastrophic failure under mechanical stress in the form of brittle fracture due to the lack of plastic deformation. ZrN is a typical example of a hard ceramic material that has a high hardness of 21 GPa, and a high stiffness (E f =207 GPa) (Fig. 3), but it fractures in a brittle manner under mechanical stress (Fig. 7). This is due to its ceramic nature, given by covalent component of the bonding, which causes the energy required to form dislocations to be enormously high. Thus, before dislocation formation can occur, a single crystal fractures along a specific cleavage plane or in case of polycrystalline ceramic materials brittle fracture occurs along grain boundaries, due to their relatively low cohesive strength [65]. Since the columnar ZrN grains are oriented normal to the coating/substrate interface (Fig. 1b), intergranular brittle fracture causes rapid crack growth along the grain boundaries through the entire coating thickness immediately after crack initiation (Fig. 5b). The preferential crack propagation along the boundaries of the columnar grains is clearly visible on the fracture surfaces of the microcantilever beams, which form a characteristic rough, highly irregular morphology with vertically oriented topology features and height and depth variations on the surface (Fig. 4b). The abrupt fracture upon exceeding the critical stress value for each microcantilever beam is evidenced by the typical linear elastic stress–- strain curves (Fig. 7) and low fracture toughness values (Fig. 6c). The brittle behavior with low fracture toughness and lack of multifunctionality makes ZrN ceramics rather unsuitable for highly demanding applications. In contrast, ZrN–Cu coatings deform more plastically (Fig. 7) due to the presence of the ductile Cu phase segregated at the grain boundaries [32]. If its amount is sufficiently high, the Cu-rich grain boundaries can absorb and dissipate stress during deformation. The ductile Cu phase also facilitates dislocation movement and grain boundary sliding, which further contributes to plastic deformation. In order to maintain the usual hardness of ZrN while increasing the plasticity of ZrN–Cu, the Cu content must be precisely optimized. A low Cu content in the nanocomposite ensures its high hardness by preserving the fine-grained ZrN matrix, but at the same time limits the effect of the ductile Cu phase on deformation. Conversely, a higher Cu content effectively increases the ductility but reduces the overall hardness due to pronounced grain boundary sliding and limited dislocation pile-up, as the grain size is inversely proportional to the thickness of the Cu tissue embedding the ZrN crystallites [32]. Therefore, in this work, ZrN–Cu has been developed with an optimized Cu content of 10 at. %, which ensures a hardness comparable to ZrN (Fig. 3b) and at the same time a ductile behavior during deformation with sufficient plasticity (Fig. 7). This is evidenced by the fracture surface of the ZrN–Cu nanocomposite (Fig. 4c) with spherical, uniformly distributed nano-sized grains showing a tortuous crack propagation path with evidence of ductile tearing, reflecting the interplay between deformation of the brittle ZrN and ductile Cu phases. The fine-grained nanocomposite microstructure of ZrN–Cu also contributes to the mechanical isotropy, which contrasts sharply with the mechanical anisotropy of the columnar microstructure of sputterdeposited transition metal nitrides [66,67]. This is particularly beneficial in situations of multidirectional loading, allowing for a more Fig. 7. Representative experimental stress–strain curves of unnotched cantilevers for monolithic and multilayer Zr–Cu–N coatings. The curve of a monolithic ZrCu coating, which deformed without fracturing, is characterized by frequent shear band events. R. Daniel et al. Materials & Design 254 (2025) 114037 7 versatile use of this type of material, which is also important in meeting the high demands for surface protection in heavy traffic environments. However, its damage tolerance in the presence of a pre-crack is still insufficient (Fig. 6c) (although its fracture stress is about 1.5 times higher than that of ZrN, see Fig. 6b). To improve the damage tolerance of ZrN, we combined it with the elastic ZrCu in a multilayer architecture (Fig. 1d and e) to better control the crack propagation after its initiation. In contrast to ZrN, ZrCu exhibits an amorphous microstructure with shortand medium-range order (Figs. 1a and 2a). The lack of long-range order and regular lattice structure means that there are fewer obstacles to impede atomic movement under applied stress, resulting in lower hardness and elastic modulus (Fig. 3). With its high plastic deformation capability (Fig. 7) and fracture toughness value of 3.1 MPa.m 1/2 (Fig. 6c), the coating demonstrates a strong ability to repeatedly deflect cracks at the interfaces between the brittle ZrN and elastic ZrCu sublayers. This is evidenced in Fig. 8, which shows the distinct steps formed on the fracture surface (Fig. 8a) and the crack deflection occurring at individual interfaces throughout the entire coating thickness (Fig. 8b). The primary mechanism driving crack deflection at the interface between the soft elastic and stiff brittle sublayers is the difference in energy release rates required for crack deflection along the interface versus penetration into the stiff brittle ZrN sublayer [38,68]. The higher toughness of ZrCu, coupled with its ability to absorb the energy released during crack growth, further enhances energy dissipation. This increased energy absorption reduces crack driving forces, contributing to a greater fracture toughness of the multilayer structure [69]. Moreover, each interface acts as an inherent microstructural barrier to crack propagation due to the differences in microstructure between nanocrystalline ZrN and amorphous ZrCu. Although the elastic modulus of the multilayer structures would be expected to lie between ZrN (197 GPa) and ZrCu (82 GPa) according to the rule of mixture, the modulus of the M02 multilayer coating is practically identical to that of ZrN, and in the case of the M01 multilayer coating, is even higher than that of the stiffest constituent (Fig. 6a). The reason for this is a synergistic effect of (i) interfacial confinement, which limits the elastic response of the individual sublayers, (ii) strong interfacial bonding, which enhances load transfer across the sublayers with a subsequent more uniform distribution when subjected to stress, increasing the stiffness of the multilayer system, and (iii) residual stress distribution, which increases the resistance of the material to deformation [52]. The contributions of these effects obviously vary depending on the multilayer architecture. By increasing the thickness of the ZrCu sublayers from 20 to 50 nm, while keeping the overall coating thickness constant, the volume fraction of the soft, elastic ZrCu phase increases and the number of interfaces decreases (Fig. 1d and e), which subsequently reduces the effect of the interface-rich multilayer architecture on the mechanical response, including the elastic modulus, fracture stress and toughness (Fig. 6). These results suggest an optimal architecture where the thickness of the ZrCu sublayers should not exceed 20 nm. The aim of our future studies is to identify the minimum thickness of the elastic constituent and the optimum total number of sublayers in the multilayer architecture for the best damage tolerance, as this study obviously shows that the ability of a multilayer system to stop or slow down crack propagation by its deflection is very much dependent on the coating design. On the other hand, the values of both fracture stress and fracture toughness of all multilayer coatings are much higher than those of monolithic ZrN. In addition to these advantageous mechanical properties, it is necessary to combine them with resistance to chemical corrosion and oxidation, as well as the ability to reduce microbes such as bacteria, viruses and fungi responsible for the spread of disease in urbanized environments. This can be achieved by replacing the top two ZrN sublayers of the M02 multilayer coating with ZrN–Cu sublayers. The incorporation of at least 10 at. % Cu into ZrN has been reported to impart antibacterial functionality to the ZrN–Cu nanocomposite [34]. Furthermore, as demonstrated in our recent experiments, surfaces in heavy-traffic environments can be effectively protected by the ZrN–Cu nanocomposite over extended periods. We simulated this real-world scenario by applying the ZrN–Cu monolithic coating on a stainlesssteel substrate (AISI 304) with dimensions of 20 ×10 mm, testing its antibacterial performance, and integrating it into the door handle of a restroom (Fig. 9a). Subsequently, the coated steel substrate was exposed to frequent daily contact with human hands and repeatedly removed for retesting after 30 and 60 days as outlined in Section 2.4. The initial antibacterial efficacy was 100 %. The primary antibacterial effect of Cu is attributed to the ability of copper ions to interact with and penetrate cell membranes, causing significant damage and loss of membrane potential. This membrane disruption leads to the leakage of essential cellular contents, ultimately resulting in cell death [70]. Additionally, the oxidative properties of Cu and the generation of reactive oxygen species (ROS) have been identified as crucial mechanisms. These ROS induce oxidative stress by damaging proteins, lipids, and DNA through redox cycling and the production of hydroxyl radicals, leading to cellular dysfunction and death [71]. Both antibacterial effects rely on the sufficient release of Cu ions, which occurs only when Cu is not fully bonded with Zr or N. This also explains the lack of antibacterial activity observed in Zr–Cu metallic glass coatings [45]. Fig. 9b shows the variation in results among individual measurements (with three measurements conducted after each time period) and over time. The observed high sensitivity of the tests to specific conditions under which they are performed is primarily attributed to potential differences in the bacterial cell state, growth phase, or initial concentration, which can influence the bacteria’s response to antibacterial agents [41]. The sustained antibacterial efficacy of approximately 90 % after 60 days of regular hand contact clearly demonstrates the long-term antibacterial performance and protective capability of the ZrN–Cu coating. The slight reduction in efficacy compared to the initial test on the virgin surface of the coating (prior to hand contact) is attributed to a decrease in surface roughness and, consequently, a reduction in surface area. However, no declining trend is observed over time, as evidenced by the comparable results after 30 and 60 days, indicating that the coating remains wear resistant under daily hand contact. Based on these results, the top two ZrN–Cu sublayers were integrated into the M03 multilayer coating with approximately the same thickness of 200 nm as the ZrN sublayers in order to keep the number of sublayers identical to the M02 multilayer coating. As we have shown that the ZrN–Cu nanocomposite exhibits better overall mechanical performance than ZrN (Fig. 6), so its integration into the less fracture-tough M02 ZrN/ ZrCu multilayer coating was intended to investigate the potential of the ZrN–Cu nanocomposite to improve the damage tolerance of Zr–Cu–N multilayer coatings while simultaneously providing the antibacterial Fig. 8. Detailed view of (a) fracture surface and (b) crack deflection at individual sublayer interfaces in a notched cantilever of an M01 multilayer Zr–Cu–N coating. Distinct steps formed on the fracture surface and cracks repeatedly deflecting at the interfaces between the brittle ZrN and elastic ZrCu sublayers are observed. R. Daniel et al. Materials & Design 254 (2025) 114037 8 activity. The results show that both the fracture stress and toughness of the M03 multilayer coating are higher than those of the M02 multilayer coating (Fig. 6b and c), demonstrating the positive effect of the ZrN–Cu nanocomposite on damage tolerance. Based on these results, we assume that replacing the hard, brittle ZrN sublayers in the M01 multilayer coating with ZrN–Cu would similarly increase its fracture stress and toughness. In addition, the M03 multilayer coating shows a lower elastic modulus after integration of the elastic ZrN–Cu nanocomposite, indicating a higher elasticity of this multilayer coating (Fig. 3a). However, the integration of the ZrN–Cu sublayers did not affect the hardness of the multilayer coating (Fig. 3b). The stress–strain curves of the unnotched microcantilevers in Fig. 7 also show their more ductile behavior compared to the M02 multilayer coating, which is related to the presence of the ductile ZrN–Cu nanocomposite. The antibacterial performance of the M03 coating was verified in a single-test measurement and compared with that of the M02 coating (Fig. 10). As shown, the M03 multilayer coating demonstrates a strong ability to effectively eliminate E. coli bacteria on its surface, outperforming the M02 coating and thereby achieving the desired multifunctionality. By comparing the mechanical properties of all monolithic and multilayer coatings, it is evident that there is still significant potential to enhance the overall damage tolerance of the Zr–Cu–N multilayer system through further optimization of the architecture. Our ongoing research aims to develop a multilayer architecture combining ZrN–Cu, ZrN and ZrCu sublayers that exhibit hardness, fracture stress and fracture toughness compared to the specimens tested in this study. The results suggest that these objectives can be achieved by keeping the thickness of the ZrCu sublayers below 50 nm and the ZrN and ZrN–Cu sublayers below 100 nm. This configuration is expected to enhance the strengthening effect of the interfaces and simultaneously promote the deflection and potential arrest of cracks. Furthermore, replacing ZrN with ZrN–Cu or periodically alternating these sublayers within the multilayer structure is anticipated to increase both fracture stress and fracture toughness. This is supported by our observations that the ZrN–Cu coating exhibits higher fracture stress than the multilayer coatings and its incorporation in the multilayer structure enhances fracture toughness (Fig. 6b and c), along with a reduction in the elastic modulus. Increased fracture toughness and decreased elastic modulus are critical for improving the damage tolerance of multilayer coatings. Our previous work has also demonstrated that the mechanical properties of the Zr–Cu–N nanocomposite can be significantly enhanced by optimizing its phase composition. Formation of a heterogeneous dual-phase nanocomposite structure, where ZrN nanocrystallites are enveloped by a very thin amorphous ZrCu phase (optimally 3–5 nm thick), significantly enhances the hardness of the nanocomposite. In this complex structure, plastic deformation is primarily mediated by dislocation movement within the nanocrystals, which is impeded at grain boundaries filled with the amorphous ZrCu phase [72]. Furthermore, at higher deposition temperatures and target power densities (all coatings in this study were synthesized without external heating), the segregation of ZrN and ZrCu in the new type of the nanocomposite Zr–Cu–N coatings was observed to be significantly more pronounced, which leads to further enhancement of the hardness and fracture toughness of the nanocomposite coatings. Implementing the discussed strategies and optimizing the multilayer architecture is the focus of our ongoing studies. 5. Conclusions Combining ZrCu, ZrN and ZrN–Cu materials in various multilayer architectures has demonstrated the feasibility of synthesizing hard and damage-tolerant multilayer coatings with complex phase compositions and antimicrobial properties using only two elemental Zr and Cu targets in non-reactive Ar and reactive Ar/N 2 -containing plasma discharges. The high hardness of the ZrN/ZrCu and ZrN–Cu/ZrCu/ZrN multilayer coatings is provided by the hard and stiff ZrN ceramic, and hard and tough ZrN–Cu nanocomposite sublayers. The elastic and ductile ZrCu metallic glass sublayers enhance damage tolerance by effectively deflecting propagating cracks at the interfaces with the elastic/soft constituents. These interfaces in the multilayer coatings also contribute to increased strength by suppressing dislocation movement under applied stress. Despite the presence of the soft ZrCu sublayers, the Fig. 9. (a) Photograph showing a ZrN–Cu monolithic coating deposited on a stainless-steel substrate (AISI 304) with dimensions of 20 ×10 mm and integrated into a door handle exposed to daily contact with human hands for 60 days. (b) Antibacterial efficacy of the coating measured against E. coli (NiCo21 (DE3)) in three consecutive tests at 30-day intervals. Fig. 10. Photographs of Petri dishes with nutrient agar and diluted E. coli bacteria solutions after the contact with (a) M02 and (b) M03 multilayer coatings. A strong elimination of the bacteria colonies in (b) demonstrates an enhanced antibacterial efficacy of the M03 coating with a top ZrN–Cu sublayer compared to the M02 coating with a top ZrN sublayer. R. Daniel et al. Materials & Design 254 (2025) 114037 9