Flexible hard nanocomposite coatings

Abstract

The article reports on flexible hard nanocomposite coatings prepared by magnetron sputtering. It is shown that the flexible hard nanocomposite coatings (i) represent a new class of coatings which are simultaneously hard, tough and resistant to cracking, (ii) exhibit high values of the hardness H and effective Young's modulus E* ratio H/E* ≥ 0.1, elastic recovery W_e ≥ 60%, compressive macrostress σ < 0 and dense, void-free microstructures, and (iii) are formed in the zone T of the Thornton's Structural Zone Model (SZM); here E* = E/(1 − ν²), E is the Young's modulus and ν is the Poisson's ratio. The magnetron sputtering, which is a very powerful process used in the preparation of nanocomposite coatings, is described in detail. The basic principles of the formation of the flexible hard coatings are also described in detail. It is shown that the key parameters which determine the formation of these coatings are (1) the energy E_p = E_ca + E_bi delivered to the growing coating by condensing atoms (E_ca) and bombarding ions (E_bi) (the non-equilibrium heating), (2) the substrate heating controlled by the substrate temperature T_s (the equilibrium heating) and (3) the melting temperature T_m of the coating material. The flexible hard coatings have a huge application potential. Four examples of flexible coatings are given: (1) flexible protective coatings, (2) flexible functional coatings, (3) flexible over-layer preventing cracking of brittle coating and (4) flexible multilayer coating. Also, the principle of low-temperature sputtering of flexible nanocomposite coatings is described in detail. Finally, trends for future development of these nanocomposite coatings with unique properties are given.

---

1. Introduction

Hard nanocomposite coatings represent a new generation of coatings [ref. 1–44 and references therein]. A typical example of hard nanocomposite coatings is the two-phase nc-TiN/a-Si₃N₄ composite. The main advantage of hard nanocomposite coatings is an enhanced hardness H. Therefore, at first, the main attention was devoted to the explanation of the origin of this enhanced hardness of the nanocomposite coatings and to the achievement of as high as possible hardness H. The nanocomposite coatings with hardness H > 40 GPa are called superhard nanocomposite coatings. At present, there are two large groups of two-phase nanocomposite coatings with enhanced hardness H: (1) hard phase/hard phase nanocomposites and (2) hard phase/soft phase nanocomposites.^10,41,42 Here it is necessary to underline that (i) both groups of the nanocomposite coatings can exhibit enhanced hardness H and (ii) the second group, i.e. the hard phase/soft phase nanocomposites, can also exhibit further unique physical and functional properties.

The thermal stability and oxidation resistance of the nanocomposite coatings at high temperatures, T, above 1000 °C are also very important. Recently, two kinds of X-ray amorphous hard coatings – (1) the nc-TMN/a-Si₃N₄ composite coatings with high (>20 at%) Si content and (2) the Si–B–C–N coatings with covalent bonds – thermally stable and oxidation resistant up to ∼1500 and ∼1700 °C, respectively, – were successfully developed.³⁴ Besides, it was understood that the enhanced hardness H is not a single important property of the hard nanocomposite coatings. For many applications the toughness of the hard coating is more important than its extremely high hardness H considerably exceeding 40 GPa. Therefore, at present, a great attention is concentrated on the development of the flexible coatings which are simultaneously hard, tough and resistant to cracking in bending.^41,44–52 These coatings represent a huge potential for many new advanced applications, for instance, in the flexible electronics, the flat panel displays, the micro-electro-mechanical systems (MEMS), the formation of functional coatings on flexible substrates such polymer foils, thin sheet glass, textile etc.

The development of flexible nanocomposite coatings is a subject of this article. It reports on a new class of advanced flexible hard coatings with enhanced resistance to cracking prepared by reactive magnetron sputtering. This article shows (i) the correlations between the coating microstructure, mechanical properties of coatings (the hardness H, the effective Young's modulus E*, the elastic recovery W_e and macrostress σ) and the energy E delivered to the growing coating, (ii) the conditions under which the flexible hard nanocomposite coatings can be formed, (iii) how cracking of brittle coatings can be prevented, (iv) four examples of flexible hard coatings and (v) the principle of low-temperature sputtering of flexible coatings; here the effective Young's modulus E* = E/(1 − ν²), E is the Young's modulus and ν is the Poisson's ratio. At the end, the trends of next development are outlined.

2. Magnetron sputtering of nanocomposite coatings

The magnetron sputtering is a very efficient tool for the formation of nanocomposite coatings.^53–80 Today, this method is well mastered and the sputtering of coatings can be realized in direct current (DC), radio frequency (RF), and DC or AC pulse mode of sputtering process; here the AC pulse is the pulse composed of negative half-sinusoids. The DC sputtering is used for the deposition of metallic and electrically conductive coatings. The RF and pulsed sputtering is suitable particularly for sputtering of electrically insulating coatings. The sputtering of coatings can be realized in pure argon or in a mixture of argon and reactive gas (RG). In the first case the coatings containing elements of sputter target can be formed. The sputtering of coatings in a mixture of argon and reactive gas is called reactive magnetron sputtering.

The reactive magnetron sputtering is very important process because it forms coatings containing not only elements of sputtered target but also elements of reactive gas. This way the compound coatings such as nitrides, oxides, carbides, borides, chlorides, etc., and their combinations can be formed. However, the reactive magnetron sputtering can exhibit two undesirable effects: (1) the hysteresis effect and (2) arcing on the surface of sputtered target. The both effects occur in consequence of a poisoning of the sputtered target and need to be eliminated to form defects free coatings in a reproducible way. The hysteresis effect can be eliminated in the case when the pumping speed for the reactive gas S_RG is greater than a critical pumping speed of the pumping system of deposition device.⁵⁵ The arcing on the sputtered target occurs when electrically insulating coatings are sputtered in consequence of charging of insulating layers formed on its uneroded areas. There are two ways allowing eliminate the arcing: (1) to use magnetrons with full target erosion (not perfectly mastered yet) or (2) to remove the accumulated charge from insulated layers on uneroded areas.^56,80 Very efficient tool removing accumulated charge from the sputtered target is pulsed sputtering with a positive voltage on the target during pulsed off time. Typical example of such pulsed sputtering system is the dual magnetron operating in bipolar mode.^63,77,80 The principle of the dc pulse dual magnetron sputtering is displayed in Fig. 1. The voltage polarity of magnetron cathodes periodically changes from negative to positive. When the cathode voltage U_d is negative the target material is sputtered. On the contrary, when U_d is positive the charge accumulated at the surface of insulating layer formed on the uneroded areas of target is discharged as a result of the electron bombardment. Here, it is also worthwhile to note that a distribution of the discharge above magnetron sputtered targets strongly depends on orientation of magnets in individual magnetrons,⁸¹ see Fig. 2. The knowledge of this fact is of key importance for a placing of the substrate on which the coating is sputtered in front of the dual magnetron.

Fig. 1 Schematic diagram of symmetric bi-polar dc pulsed dual magnetron sputtering.⁸⁰

Fig. 2 Photos of the discharge of the dual magnetron operated in a bipolar mode at repetition frequency of pulses f_r = 100 kHz, discharge current I_d = 0.5 A in nitrogen at p_N2 = 0.5 Pa. (a) Closed magnetic B field and (b) mirror magnetic B field.⁸²

The energy E delivered to the growing coating has a crucial effect on its structure, microstructure, elemental and phase composition, and physical properties.^83–88 The energy can be delivered by (i) the substrate heating E_sh, (ii) the conversion of the kinetic energy E_p of bombarding ions (E_bi) and fast neutrals (E_fn) incident on the surface of growing film, (iii) the heat evolved during formation of the compound E_ch (the energy released in exothermic chemical reactions), (iv) the heating from the sputtered magnetron target E_mt which almost always is not perfectly cooled, and (v) the radiation from the plasma E_rad. The total energy E_T delivered to the growing coating can be expressed by the following formula⁸⁸

E_T = E_sh(T_s, t_d) + E_p(U_s, i_s, a_D, p_T, t_d) + E_ch(t_d, T_a) + E_mt(W_d, t_d, d_s–t) + E_rad(t_d) (1)

where t_d is the time of the film deposition, T_a is the annealing temperature, p_T = p_Ar + p_RG is the total pressure of sputtering gas mixture, p_Ar and p_RG are the partial pressures of argon and reactive gas (RG), respectively, W_d = (U_dI_d)/S is the magnetron target power density, I_d and U_d are the magnetron current and voltage, respectively, S is the whole area of magnetron target and d_s–t is the substrate-to-target distance. The energy delivered to the growing coating by incident particles E_p is composed of two terms

E_p = E_bi + E_fn (2)

In the simplest case of a collisionless, fully ionized plasma when E_fn = 0, the energy E_p can be expressed in the following form^84–88

E_p [J cm⁻³] = E_bi = E_i(ν_i/ν_ca) = e(U_s − U_p)i_s/a_D ≈ (U_si_s)/a_D (3)

where E_i is the energy of one ion, ν_i and ν_ca is the flux of ions and condensing atoms, respectively, and e is the electron charge, U_s is the substrate bias, U_p is the plasma potential, i_s is the substrate ion current density and a_D is the film deposition rate.

Despite the fact that eqn (3) is very simplified it is very useful because it can use the measured values of U_s, i_s and a_D not only to assess easily the energy E_bi delivered to the growing coating but also to determine necessary electron density of the plasma in the magnetron discharge to reach a required value of E_bi. Moreover, this formula clearly shows a very strong effect of the deposition rate a_D on the value E_bi. This effect is very important particularly in a reactive magnetron sputtering when the deposition rate a_D decreases with increasing partial pressure of reactive gas p_RG at the same discharge current I_d of the magnetron discharge, see Fig. 3. From this figure it is clearly seen that in formation of the stoichiometric and overstoichiometric Ti(Fe)N_{x ≥ 1} nitride coatings a greater energy E_bi is delivered during their growth compared with that delivered to the substoichiometric Ti(Fe)N_{x < 1}; here x = N/(Ti + Fe). It is very important fact because the magnitude of the energy E_bi decides on the microstructure and the macrostress σ generated in growing coating. More details are given below in the Section 4.

Fig. 3 (a) Deposition rate a_D of Ti(Fe)N_x films and (b) energy E_bi delivered to them during their growth by bombarding ions as a function of p_N2. The films were sputtered using a DC magnetron equipped with a TiFe (90/10 at%) alloy target of 100 mm in diameter at (i) I_d = 1 A, i_s = 0.5 mA cm⁻², (ii) I_d = 2 A, i_s = 1 mA cm⁻², (iii) I_d = 3 A, i_s = 1 mA cm⁻², and U_s = −100 V, T_s = 300 °C, d_s–t = 60 mm and p_T = p_Ar + p_N2 = 0.5 Pa.⁸⁰

The value of the E_bi is controlled by three parameters: (1) the substrate bias U_s, (2) the substrate ion current density i_s and (3) the deposition rate of coating a_D. The substrate bias U_s is negative and should be low (|U_s| ≤ 50 V) to avoid the generation of defects in growing coating. The substrate ion current density i_s should be greater than 1 mA cm⁻² to efficiently influence the growth mechanism of coating. The deposition rate a_D should be selected according to the requirement what coating – porous coating with columnar microstructure and tensile macrostress (σ > 0) or dense coating with voids-free microstructure and compressive macrostress (σ < 0) – should be formed. For the formation of porous coatings a lower energy E_bi is sufficient. On the other hand, a higher energy E_bi is necessary to form dense, voids-free coatings; for details see below the Section 4.

The achievement of a higher energy E_bi in sputtering systems is a quite difficult problem. The magnitudes of i_s and a_D strongly depend on the discharge power P_d and the distance d_s–t between the substrate and the sputtered target. The energy E_bi can be increased by decreasing the deposition rate a_D. The decrease of a_D at constant P_d can be achieved by increasing d_s–t because a_D ∼ 1/(d_s–t)². However, increase of d_s–t means that the substrate is located into less intensive discharge from which lower ion current I_s can be extracted to the substrate at a given value U_s. It results in no increase but on the contrary in the decrease of the energy E_bi. Therefore, it is necessary to increase the ionization of the sputtering gas. It can be achieved by an independent ionization of sputtering gas in the sputtering system. The independent ionization of sputtering gas can be realized, for instance, either using a hot cathode electron beam or hollow cathode discharge or their combination, see Fig. 4. These sputtering systems are perspective for the sputtering of coatings at higher values of the energy E_bi and the formation of new advanced nanocomposite coatings. Recently, high power impulse magnetron sputtering (HIPIMS) process was developed.^89–102 In HIPIMS process when extremely high target power densities W_t = U_dI_d/S_t ranging from ∼100 W cm⁻² to ∼several kW cm⁻² are used at low values of the duty cycle τ/T ≤ 0.1 very high deposition rates a_D during pulse on time and high ionization of sputtered atoms (up to 90%) are achieved. It means that the HIPIMS process can substitute the cathodic arc evaporation process and produce coatings without macroparticles. It is main advantage of the HIPIMS process. However, the high deposition rate a_D of coating about several hundreds of nm min⁻¹ during its growth, i.e. during the pulse on time, results in the decrease of the energy E_bi delivered by ions to growing coating. At present, the effect of the energy E_bi on properties of coatings produced by the HIPIMS process is not still understood well and thus it is intensively investigated in many laboratories worldwide. For the low-temperature (T_s ≤ 100 °C) sputtering of nanocomposite coatings new magnetrons operating at low pressures p ≤ 0.1 Pa of sputtering gas are necessary to be developed.⁷⁴ These examples clearly show that for the creation of the advanced nanocomposite coatings with new unique properties new magnetrons and sputtering systems need to be developed.

Fig. 4 Magnetron with additional ionization of sputtering gas using (a) the electron beam, (b) the hollow cathode discharge and (c) the inductively coupled RF discharge. Adapted after ref. 80.

3. Flexible hard nanocomposite coatings and their main characteristics

The principle of formation of the flexible hard coatings can be explained with the aid of the Hooke's law

σ = εE (4)

where σ is the stress (load), ε is the strain (deformation) and E is the Young's modulus. If we need to form the material which allows a high elastic deformation prior to a failure, i.e. the material with a high value ε at a given value of stress σ, its Young's modulus E must be reduced. It means that the materials with the lowest value of the Young's modulus E at a given hardness H (σ = const) need to be developed. It is a simple solution but a quite difficult task. A possible behaviour of the material can be best illustrated in the stress σ vs. strain plane. The stress σ vs. the strain ε dependences for brittle, tough, resilient and ductile hard coatings are schematically displayed in Fig. 5.

Fig. 5 Schematic illustration of stress σ vs. strain ε curves of super-hard (brittle), hard (tough), hard (resilient) and ductile coatings. Resilient coatings exhibit no plastic deformation (line 0A). Adapted from ref. 41.

Superhard materials are very brittle, exhibit almost no plastic deformation and a very low strain at failure ε = ε₁. Hard and tough materials exhibit both the elastic and plastic deformation. The material withstanding a higher strain ε₁ ≪ ε ≤ ε_max without cracking exhibits a higher toughness. The hardness of tough materials is higher in the case when a greater value of ε_max is achieved at higher values of σ_max. Such materials exhibit a high strength. On the contrary, the materials with σ < σ_max and ε_σmax < ε < ε₃ are ductile, exhibit a lower strength and a high ductility. Hard and tough materials with no plastic deformation (blue line 0A) are the resilient hard coatings with 100% elastic recovery W_e. The hardness H of hard, tough and highly elastic coatings, ranging from about 15 to 25 GPa, is, however, sufficient for many applications. The main advantage of the highly elastic (flexible) coatings with a low plastic deformation is their enhanced resistance to cracking. These are reasons, why the hard, tough and flexible hard coatings are now intensively developed in many labs worldwide. The hard and simultaneously tough and flexible coatings represent a new generation of advanced hard nanocomposite coatings.

The analysis given above shows that a new task in the development of advanced hard flexible nanocomposite coatings with enhanced toughness is to master the formation of coatings with (i) the low value of the effective Young's modulus E* and (ii) a high value of the elastic recovery W_e. The hard flexible coatings with enhanced toughness, low values of E* and high values of W_e were already prepared by magnetron sputtering.^41,44–52 Also, it was found that these hard nanocomposite coatings with enhanced toughness exhibit (i) the enhanced resistance to cracking, (ii) the high ratio H/E* ≥ 0.1 and (iii) the high elastic recovery W_e ≥ 60%. More details are given below and in the ref. 41.

4. Design of flexible hard nanocomposite coatings

During many investigations of the interrelationships between the deposition parameters of coatings and their properties the general rules enabling to design the hard coatings resistant to cracking were discovered. It was found that these coatings should exhibit simultaneously the following properties^41,44

1. The low effective Young's modulus E* resulting in a high ratio H/E* ≥ 0.1 and the high elastic recovery W_e ≥ 60%.

2. The dense, voids-free microstructure.

3. The compressive macrostress (σ < 0).

All these properties (H, E*, W_e, H/E* ratio, microstructure, macrostress σ, elemental and phase composition) of hard coatings resistant to cracking are controlled by (1) the energy E_p = E_ca + E_bi delivered to the growing coating by condensing atoms (E_ca) and bombarding ions (E_bi) and the substrate heating (T_s) and (2) the kind and the amount of elements added in the base material of coating what moreover decides also on the value of the melting temperature T_m of the coating material. The relationships between the energy E_bi, the microstructure of coating, the macrostress σ generated in growing coating and the melting temperature T_m of the coating material are clearly illustrated in Fig. 6 and 7.

Fig. 6 Structural zone model (SZM) of sputtered metallic films developed by J. A. Thornton. Adapted after ref. 104.

Fig. 7 Macrostress σ in sputtered α-Ti(N) and δ-TiN_{x ≈ 1} films as a function of energy E_bi = U_si_s/a_D delivered to them by bombarding ions at p_T = p_Ar + p_N2 = 5 Pa and T_s = 350 °C, i.e. at T_s/T_m = 0.32 and 0.19 for the α-Ti(N) film and the δ-TiN_x≈1 film, respectively. Adapted after ref. 24 and 106.

Fig. 6 shows the evolution of the film microstructure in 3D representation as a function of the T_s/T_m ratio (the equilibrium heating of substrate) and the sputtering pressure of argon p_Ar, i.e. as a function of the energy E_p delivered to the growing film by incident condensing of neutral particles (non-equilibrium heating). The evolution of the film microstructure displayed in Fig. 6 is called the structural zone model (SZM). The SZM of the sputtered metallic films was developed by J. A. Thornton in 1977.¹⁰⁴ The SZM is divided in four zones: (1) the zone 1 composed of tapered crystallites separated by voids, (2) the zone T composed of fibrous grains embedded in a voids-free amorphous matrix, (3) the zone 2 composed of columnar grains separated by dense inter-crystalline boundaries¹⁰⁵ and (4) the zone 3 composed of recrystallized grain structure. The boundary between the zone 1 and the zone T corresponds to zero macrostress σ = 0 generated in sputtered films and separates the films with a low-density microstructure composed of columns separated by voids (the zone 1) from the films with a dense, voids-free microstructure composed of fibrous grains embedded in an amorphous inter-grain phase (the zone T). Very different microstructure in the zone 1 and the zone T is a main reason why properties of the films formed in the zone 1 and the zone T so strongly differ and why the films formed in the zone T exhibit unique properties.

Fig. 7 shows the macrostress σ as a function of the energy E_bi. The effect of the melting temperature T_m of film material on the magnitude and kind of macrostress σ and on the microstructure of film are also clearly seen in this figure. The macrostress σ generated in the growing film also very well correlates with its microstructure; see Fig. 6 and photos inserted in Fig. 7. From Fig. 7 four general and very important issues can be drawn.

1. The macrostress σ changes from tensile (σ > 0) to compressive (σ < 0) with increasing energy E_bi.

2. There is a critical value of E_bi = E_c at which films exhibit zero macrostress σ = 0.

3. The values of the macrostress σ and the critical energy E_c depend on the melting temperature T_m of the film material.

4. The macrostress σ and the critical energy E_c increase with increasing T_m of coating material.

The magnitude of the critical energy E_c depends on (i) the substrate bias U_s and the substrate ion current density i_s, (ii) the deposition rate a_D of film, (iii) the sputtering gas pressure p, (iv) the substrate temperature T_s, (v) the kind and amount of the element added in the film what determines the melting temperature T_m of the film material, and (vi) the contamination of film by oxygen O and nitrogen N from a residual atmosphere in the deposition chamber when the film is deposited at low deposition rates a_D,¹⁰⁷ i.e. that the critical energy E_c = f(E_bi, elemental composition, T_m, p, p₀) is a very complex function depending on many deposition and film material parameters; here p is the sputtering gas pressure and p₀ is the base pressure in deposition chamber.

Many experiments performed so far show that the films formed in the zone T exhibit not only dense, voids-free microstructure composed of tiny fibrous grains embedded in an amorphous matrix and compressive macrostress σ < 0 but also a high ratio H/E* ≥ 0.1 and high values of the elastic recovery W_e ≥ 60%. It means that films formed in the zone T can be flexible and thereby automatically can exhibit an enhanced resistance to cracking if the sufficient energy E > E_c is delivered in films during their growth. Three examples of flexible films are given below in the Sections 5–7.

In summary, it can be concluded that the formation of films with microstructure corresponding to the zone T of the Thornton's SZM requires deliver to films during their growth a sufficiently high energy E, which must exceed the critical energy E_c.

5. Flexible protective coatings

In many cases, the surface of materials needs to be protected against damage, for instance, scratching, oxidation, corrosion, erosion, etc. In these cases protective coatings are used. Such coatings have to be simultaneously hard and flexible, and particularly when they are deposited on flexible substrates they must exhibit an enhanced resistance to cracking during the substrate bending. Such the properties exhibit flexible hard films described above.

The resistance of the protective flexible coating to cracking is demonstrated in the Zr–Al–O system. In this system the flexibility of coating and its resistance to cracking is controlled by the Zr/Al ratio, i.e. by the elemental composition of coating. The Zr/Al ratio strongly influences the mechanical properties of the Zr–Al–O coating and thereby its resistance to cracking. The Zr–Al–O coatings with a low ratio Zr/Al < 1 exhibit a low ratio H/E* < 0.1 and a low elastic recovery W_e < 60% are brittle and easily crack in bending. On the contrary, the Zr–Al–O coatings with a high ratio Zr/Al > 1 exhibit a high ratio H/E* ≥ 0.1 and a high elastic recovery W_e ≥ 60% are flexible and well resists to cracking in bending. This fact is illustrated in Fig. 8. More detail is given in ref. 46 and 48.

Fig. 8 Comparison of surface morphology of (a) brittle and (b) flexible ∼3000 nm thick Zr–Al–O coating reactively sputtered at U_s = U_fl, T_s = 500 °C and p_T = 1 Pa on Mo strip after bending around steel cylinder of radius r = 12.5 mm.^44,48

Recently, the resistance of the film against cracking was demonstrated also in the following coating systems: (1) Al–Cu–O oxide/oxide nanocomposite coating,⁴⁵ (2) Zr–Al–O oxide/oxide nanocomposite coating,^46,48 (3) Al–O–N nitride/oxide nanocomposite coating,⁴⁷ (4) Si–Zr–O oxide/oxide nanocomposite coating,⁴⁹ (5) Ti–Ni–N nitride/nitride nanocomposite coating,⁵⁰ (6) Al–Cu–N nitride/nitride nanocomposite coating⁵¹ and (7) (Ti,Al,V)N_x nitride/nitride nanocomposite coating.⁵² All these coatings exhibit the high ratio H/E* ≥ 0.1 and high elastic recovery W_e ≥ 60%. This finding clearly indicates that the high ratio H/E* ≥ 0.1 and high W_e ≥ 60% are key parameters necessary for the formation hard films resistant to cracking. For more details see the papers.^45–52

6. Flexible functional coatings

Many functional coatings are brittle, easily crack and thus they lose its function during the operation. Therefore, it is vitally important to develop flexible functional coatings with an enhanced resistance to cracking. The possibility to form functional coatings resistant to cracking is demonstrated in two systems Cr–Cu–O¹⁰³ and Al–Cu–N⁵¹ of antibacterial coatings with different elemental composition. The Cu content in the coating decides on its antibacterial function, i.e. on the efficiency of the killing of bacteria on the coating surface. The efficiency of killing of bacteria increases with increasing Cu content in the coating, see Fig. 9.

Fig. 9 Photos of Petri dishes with Endo agar and colonies of Escherichia coli bacteria cultivated from bacterial suspensions which were in contact with a-(Cr–Cu–O) coatings with various at% of Cu in the dark for 5 hours. The films were sputtered on Si(100) substrate at floating potential (U_s = U_fl) and T_s = 500 °C.¹⁰³ Dark areas in this figure are colonies of living Escherichia coli bacteria.

From Fig. 9 it is seen that a high amount of Cu of about 20 at% in the Cr–Cu–O coatings is necessary to kill all bacteria on the coating surface. However, such high content of Cu in the coating results in low values of H, E*, W_e and H/E* ratio, see Table 1. It means that the Cr–Cu–O film with ∼20 at% Cu exhibits a low mechanical protection against fretting and cracking in bending, particularly in the case when the coating is deposited on a flexible substrate. This coating exhibits 100% efficiency of killing of E. coli bacteria, see Fig. 9 but very easily cracks, see Fig. 10 and Table 1. Therefore, it was highly desirable to develop an antibacterial coating which simultaneously efficiently kills bacteria on its surface and exhibits an enhanced resistance against mechanical damages and cracking. It was found that the Cr–Cu–N coating perfectly fulfils both requirements. The replacement of O by N makes it possible (i) to reduce Cu content from ∼20 at% to ∼10 at% for 100% killing of all E. coli bacteria and (ii) to increase the hardness H from ∼3 GPa to ∼20 GPa, the elastic recovery W_e from 36 to 74%, H/E* ratio from 0.046 to 0.122, and to convert the macrostress σ from tensile to compressive (σ < 0), i.e. to reach conditions which are necessary to produce highly flexible antibacterial coating with enhanced resistance to cracking in bending. The enhanced resistance of the Cr–Cu–N antibacterial coating to cracking was confirmed by its deposition on Mo strip (55 × 9 × 0.15 mm³) and bending around a fixed cylinder of radius r = 10 mm, see Fig. 10. It means that while the Cr–Cu–O coating is one-functional antibacterial coating, the Al–Cu–N coating is two-functional antibacterial/flexible coating.

Table 1 Thickness h, deposition rate a_D, deposition parameters, physical and mechanical properties of sputtered Cr–Cu–O and Al–Cu–N coatings and the assessment of their resistance to cracking by bending illustrated in Fig. 10^a

Coating	h [nm]	aD [nm min^−1]	Ts [°C]	Usp [V]	isp [mA cm^−2]	σ [GPa]	Cu [at%]	H [GPa]	E* [GPa]	We [%]	H/E*	Cracks bending
a Here, isp is the averaged substrate ion current density over negative pulse of substrate bias Usp.^44,51,103
Cr–Cu–O	2190	18.3	500	Ufl	—	0.1	19.5	3.2	70	36	0.046	Yes
Al–Cu–N	2730	63.5	400	−100	1.38	−1.7	9.6	21.9	180	74	0.122	No

---

Fig. 10 Surface morphology of (a) Cr–Cu–O oxide coating with H/E* < 0.1 and (b) Al–Cu–N coating with H/E* > 0.1 deposited on Mo strip (55 × 15 × 0.15 mm³) after bending around cylinder of radius r = 10 mm, respectively.^44,51

At present, one of the hottest tasks is, for instance, to develop and master the production of three-functional flexible/transparent/electrically conductive coatings.

In summary it is necessary to conclude that the formation of two- and multi-functional coatings with maximally optimized functions is quite difficult but possible process. The formation of such films is controlled by a combined action of the elemental composition of coating and the energy E delivered to the growing coating.^41,44

7. Flexible over-layer preventing cracking of brittle coatings

Sometimes it is impossible to form flexible functional coatings because these coatings are brittle. Therefore, it is important to find a way how to prevent cracking of the brittle coatings. The principle preventing the brittle coatings from cracking is based on over-coating of the brittle coating by the elastic over-layer with high ratio H/E* ≥ 0.1, high elastic recovery W_e ≥ 60% and compressive macrostress σ < 0.⁴⁹

The experimental evidence of the validity of this principle is illustrated in Fig. 11. This figure shows the surface morphology of (a) the single-layer brittle Zr–Si–O coating with low ratio H/E* = 0.08, low elastic recovery W_e = 50% and tensile macrostress σ = 0.25 GPa and (b) the two-layer coating composed of the same brittle Zr–Si–O coating (the bottom layer) and the elastic Zr–Si–O film with high ratio H/E* = 0.1, high elastic recovery W_e = 68% and compressive macrostress σ = −1.5 GPa (the top layer) sputtered on a Mo strip (60 × 10 × 0.1 mm³) after bending of the coated strip around the cylinder with diameter of r = 12.5 mm. As expected, the single-layer coating with the low ratio H/E* < 0.1, low elastic recovery W_e < 60% and tensile macrostress σ > 0 easily cracks in bending around the cylinder of radius r = 12.5 mm already at a small bending angle α ≈ 30°, see Fig. 11a. On the contrary, the two-layer coating in which the same layer 1 as in the single-layer coating is covered by the layer 2 with high ratio H/E* ≈ 0.1, high elastic recovery W_e > 60% and compressive macrostress σ < 0 exhibits no cracks even after bending up to angle α = 180° around the cylinder with the same radius r = 12.5 mm, see Fig. 11b.

Fig. 11 Schematic illustration of (A) coating/substrate geometry, (B) method of bending of the coated Mo strip and (C) photos of surface morphology of (a) single-layer Zr–Si–O coating with low ratio H/E* < 0.1, low elastic recovery W_e < 60% and tensile macrostress σ > 0 (the layer 1) and (b) two-layer Zr–Si–O coating composed of the bottom layer with low ratio H/E* < 0.1, low elastic recovery W_e < 60% and tensile macrostress σ > 0 (the layer 1) and the upper layer with high ratio H/E* ≥ 0.1, high elastic recovery W_e > 60% and compressive macrostress σ < 0 (the layer 2) after bending around the cylinder with diameter of r = 12.5 mm.⁴⁹

This experiment clearly demonstrates that the brittle coating can be protected against cracking by the highly elastic over-layer with high ratio H/E* ≥ 0.1, high elastic recovery W_e ≥ 60% and compressive macrostress σ < 0. Here, it is worthwhile to note that the compressive macrostress (σ < 0) of the elastic over-layer plays a very important role preventing the brittle coating (the layer 1) from cracking. However, more experiments are needed to be done to find the best way how to reach a maximal resistance of the brittle coating to cracking.

8. Flexible multilayer coatings resistant to cracking

Recent experiments indicate that the alternation of the macrostress σ in multi-layered coatings could be a new approach in formation of very thick (≥10 μm) coatings with enhanced resistance to cracking. This hypothesis was tested in three-layer and four-layer Zr–Si–O coatings composed of alternating layers in tension (σ > 0) and compression (σ < 0) sputtered on a Mo strip (60 × 10 × 0.1 mm³). The surface morphology of these Zr–Si–O coatings together with the one-layer and two-layer Zr–Si–O coatings after bending around the fixed cylinder of radius r = 12.5 mm is shown in Fig. 12. The layer in tension is denoted as the layer 1 (L1) and the layer in compression is denoted as the layer 2 (L2). The first layer deposited at the substrate was always the layer 1. The thicknesses h_n of individual layers in the coating and its mechanical properties are given in Table 2; the index n denotes the number of layer in the coating.

Fig. 12 Surface morphology of (a) one-layer, (b) two-layer, (c) three-layer and (d) four-layer Zr–Si–O coating deposited on Mo strip after bending around fixed cylinder of radius r = 12.5 mm.⁴⁹

Table 2 Thickness h_n of individual layers in the multilayer Zr–Si–O coatings, their mechanical properties, macrostress σ in the top layer, and cracks created in multilayer coating during bending of coated Mo strip around fixed cylinder of radius r = 12.5 mm⁴⁹

Coating	Content of layers	h1 [nm]	h2 [nm]	h3 [nm]	h4 [nm]	hT [nm]	H [GPa]	E* [GPa]	We [%]	H/E*	Macrostress in top layer	Cracks in bending
Coating A	L1	3000	—	—	—	3000	12.6	161	50	0.078	Tension	Yes
Coating B	L1 + L2	3000	3000	—	—	6000	17.6	165	70	0.107	Compression	No
Coating C	L1 + L2 + L3	2500	2500	2500	—	7500	12.2	149	56	0.082	Tension	Yes
Coating D	L1 + L2 + L3 + L4	2400	2400	2400	2400	9600	16.6	157	70	0.106	Compression	No

---

From Fig. 12 it is seen that the alternation of the layers in tension and compression strongly influences the resistance of the coating to cracking. Always, when the top layer is in tension (σ > 0) the coating cracks. On the contrary, when the top layer is in compression (σ < 0) the coating does not cracks, i.e. the coating exhibits an enhanced resistance to cracking. This is a very important phenomenon which can be used in the formation of new advanced very thick (10 to 100 μm) coatings. Also, it is worthwhile to note that the layer 1 is X-ray amorphous and the layer 2 is crystalline; more details are in ref. 44 and 49. Based on this experiment new very thick multilayer coatings composed of many crystalline/X-ray amorphous bi-layers can be realized, see Fig. 13. At present, such thick multilayer coatings are under investigation in our labs. Moreover, it can be expected, that these coatings will exhibit new unique properties.

Fig. 13 Multilayer coating composed of alternating X-ray amorphous and crystalline layers.

However, many questions still remain open. For instance, what is a minimum thickness of the layer in compression which is sufficient to prevent the cracking of the multilayer coating, what is the effect of layers with alternating macrostress σ on physical properties of the multilayer coating, what is the thermal stability and the thermal conductivity of the multilayer coating composed of layers with macrostress σ alternating from tension (σ > 0) to compression (σ < 0), what should be the structure and the microstructure of the layer on the substrate/coating interface to achieve maximum adhesion of the coating to the substrate, etc. These tasks are a subject of our next investigation.

9. Low-temperature sputtering of flexible nanocomposite coatings

In the Section 4 above, it was shown that the flexible films can be formed in the zone T of the SZM. However, to form coatings in the zone T it is necessary to overpass the boundary line between the zone 1 and the zone T, which in the 2D representation of the SZM is a function of the ratio T_s/T_m and the sputtering gas pressure p, see Fig. 14. Therefore, it is important to show, how the location of the boundary line between the zone 1 and zone T can be controlled. The principle of the low-temperature sputtering of films with microstructure of zone T is based on replacement of the equilibrium substrate heating (T_s) by the non-equilibrium atomic scale heating (ASH).^41,44,74

Fig. 14 Thornton's structural zone model (SZM) of sputtered films extended to the region of low sputtering gas pressures. Adapted after ref. 74.

The location of the boundary line between the zone 1 and the zone T in the SZM can be influenced by (1) the energy E_bi delivered to the growing film by bombarding ions when the films are formed at negative substrate bias U_s < 0 (ref. 108 and 109) and (2) by the energy E_fn delivered to the growing film by fast neutral particles controlled by the pressure p of sputtering gas and increasing with decreasing p. The boundary line between the zone 1 and the zone T shifts to lower values of the ratio T_s/T_m with increasing negative substrate bias U_s and decreasing sputtering gas pressure p, see Fig. 14. It means that the films with the microstructure of the zone T can be formed at low values T_s/T_m, i.e. at low substrate temperatures T_s, even at unheated substrates (T_s = RT). Here, it is important to underline that the deposition of the films with microstructure corresponding to the zone T of the Thornton's SZM at low values of T_s/T_m ratio is possible also without using a negative substrate bias U_s, i.e. at the substrate held at the floating potential (U_s = U_fl), when the films are sputtered at low pressures p, see Fig. 14. However, new low-pressure magnetrons need to be developed for the formation of films at very low pressures p. More details are given in ref. 42, 61, 74 and 80.

The possibility to form the films with the microstructure corresponding to the zone T of the Thornton's SZM at low substrate temperatures T_s ≤ 100 °C is of a great potential for many advanced applications, namely for flexible electronics, flat panel displays, MEMS systems, formation of functional films on polymer foils and fabrics, etc. The formation of nanocrystalline or even crystalline films with the microstructure corresponding to the zone T at low substrate temperatures T_s ≤ 100 °C can be realized also by the addition of selected elements in the film because the melting temperature T_m of such films can be decreased compared to the pure films and thereby the T_s/T_m ratio can be increased.

In summary it can be concluded that the films with dense, voids-free microstructure, compressive macrostress σ < 0 and enhanced resistance to cracking can be formed in the zone T of the Thornton's SZM also at low substrate temperatures T_s ≤ 100 °C, i.e. at low values of T_s/T_m ratio, when sufficient energy E_bi is delivered to the growing film by bombarding ions and/or when a low sputtering gas pressure p is used. Very important role in the low-temperature sputtering of the flexible films plays also the melting temperature T_m of the film material which (i) decides on the value of T_s/T_m ratio and (ii) can be controlled by the addition of selected elements in the film.

10. Conclusions

Up to now, four important milestones have been achieved in development of the hard nanocomposite coatings: (1) the hard/hard two-phase nanocomposites with enhanced hardness H composed of two hard phases, (2) the hard/soft nanocomposites with enhanced hardness H composed of one hard and one soft phase, (3) X-ray hard nanocomposites with high thermal stability and oxidation resistance up to ∼1500 °C and (4) flexible hard nanocomposite coatings with high toughness and enhanced resistance to cracking. The principle of formation of the flexible hard nanocomposite coatings and several applications are reported in this article. However, there are further tasks which have to be mastered in the development of new advanced hard nanocomposite coatings. These tasks can be briefly summarized as follows.

1. The low-temperature sputtering of flexible, hard, two- and multi-functional nanostructured and nanocomposite coatings.

2. The formation of single-phase crystalline coatings thermally stable in a wide range of temperatures from the room temperature up to very high temperatures approaching the melting temperature T_m of the coating material.

3. The detailed investigations of interrelationships between the coating properties and the energy E delivered to the coating during its growth or in a post-deposition treatment process.

4. The development of new deposition systems operating under new physical conditions.

There is no doubt that every even a small progress in the solution of these tasks will strongly contribute to deepen present knowledge in the field of advanced hard nanocomposite coatings and will represent a huge potential for realization of new advanced applications.

Acknowledgements

This work was supported in part by the Grant Agency of the Czech Republic under project No. P108/12/0393.

References

1. H. Gleiter, Nanocrystalline materials, Prog. Mater. Sci., 1989, 33, 223–315.
2. R. Birringer, Nanocrystalline materials, Mater. Sci. Eng., A, 1989, 117, 33–43.
3. R. W. Siegel, Cluster-assembled nanophase materials, Annu. Rev. Mater. Sci., 1991, 21, 559–579.
4. R. W. Siegel, What do we really know about the atomic-scale structures of nanophase materials?, J. Phys. Chem. Solids, 1994, 55(10), 1097–1106.
5. R. W. Siegel and G. E. Fougere, Grain size dependent mechanical properties in nanophase materials, in Proc. Mater. Res. Soc. Symp, ed. H. J. Grant, R. W. Armstrong, M. A. Otooni and K. Ishizaki, Warrendale, PA, 1995, vol. 362, pp. 219–229.
6. S. Vepřek and S. Reiprich, A concept for the design of novel superhard coatings, Thin Solid Films, 1995, 265, 64–71.
7. H. Gleiter, Nanostructured materials: state of the art and perspectives, Nanostruct. Mater., 1996, 6, 3–14.
8. S. Yip, The strongest size, Nature, 1998, 391, 532.
9. A. A. Voevodin and J. S. Zabinski, Superhard, functionally gradient, nanolayered and nanocomposite diamond-like carbon coatings for wear protection, Diamond Relat. Mater., 1998, 7, 463–467.
10. J. Musil, Hard and superhard nanocomposite coatings, Surf. Coat. Technol., 2000, 125, 322–330.
11. H. Gleiter, Nanostructured materials: basic concepts and microstructure, Acta Mater., 2000, 48, 1–29.
12. A. Leyland and A. Matthews, On the significance of the H/E ratio in wear control: a nanocomposite coating approach to optimised tribological behaviour, Wear, 2000, 246, 1–11.
13. L. Hultman, Thermal stability of nitride thin films, Vacuum, 2000, 57(1), 1–30.
14. R. Hauert and J. Patscheider, From alloying to nanocomposites – Improved performance of hard coatings, Adv. Mater., 2000, 2(5), 247–259.
15. S. Vepřek, Nanostructured superhard materials, in Handbook of Ceramic Hard Materials, ed. R. Riedel, WILEY-VCH, Weinheim, 2000, ch. 4, pp. 104–139.
16. R. Hauert and J. Patscheider, From alloying to nanocomposites – Improved performance of hard coatings, Adv. Mater., 2000, 2(5), 247–259.
17. H. Gleiter, Tuning the electronic structure of solids by means of nanometer-sized microstructures, Scr. Mater., 2001, 44, 1161–1168.
18. R. A. Andrievskii, Thermal stability of nanomaterials, Russ. Chem. Rev., 2002, 71(10), 853–866.
19. G. M. Demyashev, A. L. Taube and E. E. Siores, Superhard nanocomposite coatings, in Handbook of Organic-Inorganic Hybrid Materials and Nanocomposite, ed. H. S. Nalwa, American Scientifuc Publishers, 2003, vol. 1, pp. 1–61.
20. J. Patscheider, Nanocomposite hard coatings for wear protection, MRS Bull., 2003, 28(3), 180–183.
21. S. Zhang, D. Sun, Y. Fu and H. Du, Recent advances of superhard nanocomposite coatings: a review, Surf. Coat. Technol., 2003, 167, 113–119.
22. R. A. Andrievski, Nanomaterials based on high-melting carbides, nitrides and borides, Russ. Chem. Rev., 2005, 74(12), 1061–1072.
23. S. Zhang, D. Sun, Y. Fu and H. Du, Toughening of hard nanostructured thin films: a critical review, Surf. Coat. Technol., 2005, 198, 2–8.
24. J. Musil, Physical and mechanical properties of hard nanocomposite films prepared by reactive magnetron sputtering, in Nanostructured Coatings, ed. J. T. M. DeHosson and A. Cavaleiro, Springer Science + Business Media, LCC, New York, 2006, ch. 10, pp. 407–463.
25. L. Hultman and C. Mitterer, Thermal stability of advanced nanostructured wear-resistnt coatings, in Nanostructured Coatings, ed. J. T. M. DeHosson and A. Cavaleiro, Springer Science + Business Media, New York, 2006, ch. 11, pp. 464–510.
26. P. H. Mayrhofer, C. Mitterer and L. Hultman, Microstructural design of hard coatings, Prog. Mater. Sci., 2006, 51, 1032–1114.
27. C. Lu, Y. W. Mai and Y. G. Shen, Recent advances on understanding the origin of superhardness in nanocomposite coatings: a critical review, J. Mater. Sci., 2006, 41, 937–950.
28. C. S. Sandu, F. Medjani, R. Sanjines, A. Karimi and F. Levy, Structure, morphology and electrical properties of sputtered Zr-Si-N thin films: from solid solution to nanocomposite, Surf. Coat. Technol., 2006, 201, 4219–4229.
29. Y. H. Lu and Y. G. Shen, Nanostructure transition: from solid solution Ti(N,C) to nanocomposite nc-Ti(N,C)/a-(C,CN_x), Appl. Phys. Lett., 2007, 90, 221913.
30. C. S. Sandu, F. Medjani and R. Sanjines, Optical and electrical properties of Zr-Si-N thin films: from solid solution to nanocomposite, Rev. Adv. Mater. Sci., 2007, 15, 173–178.
31. J. Musil and M. Jirout, Toughness of hard nanostructured ceramic thin films, Surf. Coat. Technol., 2007, 201, 5148–5152.
32. A. Raveh, I. Zukerman, R. Shneck, R. Avni and I. Fried, Thermal stability of nanostructured superhard coatings: a review, Surf. Coat. Technol., 2007, 201, 6136–6142.
33. J. Musil, Properties of hard nanocomposite thin films, in Nanocomposite films and coatings, ed. S. Zhang and N. Ali, Imperial College Press, London, 2007, ch. 5, pp. 281–328.
34. J. Musil, J. Vlcek and P. Zeman, Advanced amorphous non-oxide coatings with oxidation resistance above 1000°C, sSpecial Issue on Nanoceramics, Adv. Appl. Ceram., 2008, 107(3), 148–154.
35. J. Musil, P. Baroch and P. Zeman, Hard nanocomposite coatings. Present status and trends, in Plasma Surface Engineering Research and its Practical Applications, ed. R. Wei, Research Signpost, Kerala, 2008, ch. 1, pp. 1–34.
36. R. Wei, Plasma enhanced magnetron sputtering deposition of superhard, nanocomposite coatings, in Plasma Surface Engineering Research and its Practical Applications, ed. R. Wei, Research Signpost Publisher, Kerala, 2008, ch. 3, pp. 87–115.
37. A. D. Pogrebjank, A. P. Shpak, N. A. Azarenkov and V. M. Beresnev, Structure and properties of hard and superhard nanocomposite coatings, Phys.-Usp., 2009, 52, 29–54.
38. A. Matthews and A. Leyland, Materials related aspects of nanostructured tribological coatings, SVC Bull., 2009, 40–44.
39. A. D. Korotaev, B. D. Borisov, Yu. V. Moshkov, S. V. Ovchinnikov, Yu. P. Pinzhin and A. N. Tyumentsev, Elastic stress state in superhard multielement coatings, Phys. Mesomech., 2009, 12(5–6), 269–279 , in Russian.
40. J. Musil, Recent progress in hard nanocomposite coatings, part 1, Galvanotechnik, 2010, 8, 1856–1867; J. Musil, Recent progress in hard nanocomposite coatings, part 2, Galvanotechnik, 2010, 9, 2116–2121.
41. J. Musil, Hard nanocomposite coatings: thermal stability, oxidation resistance and toughness, Surf. Coat. Technol., 2012, 207, 50–65.
42. J. Musil, P. Zeman and P. Baroch, Hard Nanocomposite Coatings, in Comprehensive Materials Processing, Coatings and Films, ed. D. Cameron, Elsevier, 2014, vol. 4, ch. 4.13, pp. 325–353.
43. S. Veprek, Recent search for new superhard materials: go nano!, J. Vac. Sci. Technol., A, 2013, 31(5), 050822.
44. J. Musil, Advanced hard nanocomposite coatings with enhanced toughness and resistance to cracking, in Thin Films and Coatings: Toughening and Toughening Characterization, ed. S. Zhang, CRC Press, USA, 2015, ch. 7, pp. 377–463.
45. J. Blažek, J. Musil, P. Stupka, R. Čerstvý and J. Houška, Properties of nanocrystalline Al-Cu-O films reactively sputtered by dc pulse dual magnetron, Appl. Surf. Sci., 2011, 258, 1762–1767.
46. J. Musil, J. Sklenka and R. Cerstvy, Transparent Zr-Al-O nanocomposite coatings with enhanced resistance to cracking, Surf. Coat. Technol., 2012, 206, 2105–2109.
47. J. Musil, M. Meissner, R. Jilek, T. Tolg and R. Cerstvy, Two-phase single layer Al-O-N nanocomposite films with enhanced resistance to cracking, Surf. Coat. Technol., 2012, 206, 4230–4234.
48. J. Musil, J. Sklenka, R. Čerstvý, T. Suzuki, M. Takahashi and T. Mori, The effect of addition of Al in ZrO₂ thin film on its resistance to cracking, Surf. Coat. Technol., 2012, 207, 355–360.
49. J. Musil, J. Sklenka and J. Prochazka, Protective over-layer coating preventing cracking of thin films deposited on flexible substrates, Surf. Coat. Technol., 2014, 240, 275–280.
50. J. Musil, R. Jílek and R. Čerstvý, Flexible Ti-Ni-N thin films prepared by magnetron sputtering, J. Mater. Sci. Eng. A, 2014, 4(2), 27–33.
51. J. Musil, J. Blazek, K. Fajfrlik and R. Cerstvy, Flexible antibacterial Al-Cu-N films, Surf. Coat. Technol., 2015, 264, 114–120.
52. J. Prochazka, R. Cerstvy and J. Musil, Interrelationship between mechanical properties and resistance to cracking of magnetron sputtered (Ti,Al,V)Nx nitride films, Paper B5-1-4, 42nd international Conference on Metallurgical Coatings and Thin Films (ICMCTF 2015), April 20–24, 2015, San Diego, CA, USA, Book of Abstracts, p. 80.
53. Thin Films Processes, ed. J. L. Vossen and W. Kern, Academic Press, Inc., New York, 1978.
54. B. Window and N. Savvides, Unbalanced magnetrons as sources of high ion fluxes, J. Vac. Sci. Technol., A, 1986, 4, 504.
55. S. Kadlec, J. Musil and J. Vyskocil, Hysteresis effect in reactive sputtering: a problem of system stability, J. Phys. D: Appl. Phys., 1986, 19, L187–L190.
56. S. Schiller, U. Heisig, C. Korndorfer, G. Beister, J. Reske, K. Steinfelder and J. Strumpfel, Reactive DC high-rate sputtering as production technology, Surf. Coat. Technol., 1987, 33, 405–423.
57. J. Musil, S. Kadlec, J. Vyskocil and V. Valvoda, New results in dc reactive magnetron deposition of TiN_x films, Thin Solid Films, 1988, 167, 107–119.
58. Handbook of Plasma Processing Technology, ed. S. M. Rossnagel, J. J. Cuomo and W. D. Westwood, Noyes Publications, Park Ridge, New Jersey, USA, 1990.
59. B. Window and N. Savvides, Ion-assisting magnetron sources: principles and uses, J. Vac. Sci. Technol., A, 1990, 8, 1277.
60. J. Musil, S. Kadlec, V. Valvoda, R. Kuzel Jr. and R. Cerny, Ion-assisted sputtering of TiN films, Surf. Coat. Technol., 1990, 43/44, 259–269.
61. J. Musil, S. Kadlec and W.-D. Munz, Unbalanced magnetrons and new sputtering systems with enhanced plasma ionization, J. Vac. Sci. Technol., A, 1991, 9(3), 1171–1177.
62. E. Kusano, An investigation of hysteresis effects as a function of pumping speed, sputtering current, and O₂/Ar ratio, in Ti–O₂ reactive sputtering processes, J. Appl. Phys., 1991, 70(11), 7089.
63. S. L. Rohde, L. Hultman, M. S. Wong and W. D. Sproul, Dual-unbalanced magnetron deposition of tiN films, Surf. Coat. Technol., 1992, 50, 255–262.
64. J. Musil, S. Kadlec and J. Vyskocil, Hard coatings prepared by sputtering and arc evaporation, in Physics of Thin Films, Mechanic and dielectric properties, ed. M. H. Fracombe and J. L. Vossen, Academic Press, Inc., San Diego, CA, 1993, vol. 17, pp. 80–139.
65. S. Schiller, K. Goedicke, J. Reschke, V. Kirchhoff, S. Schneider and F. Milde, Pulsed magnetron sputter technology, Surf. Coat. Technol., 1993, 61(1–3), 331–337.
66. W. D. Sproul, Ion-assisted deposition in unbalanced-magnetron sputtering systems, Mater. Sci. Eng., A, 1993, 163, 187–192.
67. S. M. Rossnagel, Directional and preferential sputtering-based physical vapour deposition, Thin Solid Films, 1995, 263, 1–12.
68. W. D. Sproul, Advances in reactive sputtering, 39th Annual Technical Conference Proceedings, Philadelphia, USA, 1996, pp. 3–6.
69. W. D. Sproul, New routes in the preparation of mechanically hard films, Science, 1996, 273, 889–892.
70. B. Window, Issues in magnetron sputtering of hard coatings, Surf. Coat. Technol., 1996, 81, 92–98.
71. S. Kadlec and J. Musil, Low pressure magnetron sputtering and selfsputtering discharges, Vacuum, 1996, 47(3), 307–311.
72. J. Musil, Basic properties of low-pressure plasma, in Proceedings of the International School of Physics Enrico Fermi, Course CXXXV, ed. A. Paoleti and A. Tucciarone, IOS Press, Amsterdam, 1997, pp. 145–177.
73. W. D. Sproul, High-rate reactive DC magnetron sputtering of oxide and nitride superlattice coatings, Vacuum, 1998, 51(4), 641–649.
74. J. Musil, Low-pressure magnetron sputtering, Vacuum, 1998, 50(3–4), 363–372.
75. R. D. Arnell and P. J. Kelly, Recent advances in magnetron sputtering, Surf. Coat. Technol., 1999, 112, 170–176.
76. I. Safi, Recent aspects concerning DC reactive magnetron sputtering of thin films: a review, Surf. Coat. Technol., 2000, 127, 203–218.
77. P. J. Kelly and R. D. Arnell, Magnetron sputtering: a review of recent developments and applications, Vacuum, 2000, 56, 159–172.
78. J. Musil, Hard nanocomposite films prepared by reactive magnetron sputtering, in Nanostructured Thin Films and Nanodispersion Strengthened Coatings, ed. A. A. Voevodin, D. V. Shtansky, E. A. Levashov and J. J. Moore, NATO Science Series, Kluwer Academic Publishers, Dordrecht, The Netherlands, 2003, vol. 155, ch. 5, pp. 43–57.
79. W. D. Sproul, D. J. Christie and D. C. Carter, Control of reactive sputtering processes: review, Thin Solid Films, 2005, 491, 1–17.
80. J. Musil, J. Vlček and P. Baroch, Magnetron discharges for thin films plasma processing, in Materials Surface Processing by Directed Energy Techniques, ed. Y. Pauleau, 2006, Elsevier Science Publisher B.V., Oxford, UK, ch. 3, pp. 67–106.
81. J. Musil and P. Baroch, High-rate pulse reactive magnetron sputtering of oxide nanocomposite coatings, Vacuum, 2013, 87, 96–102.
82. J. Musil and P. Baroch, Discharge in dual magnetron sputtering system, IEEE Trans. Plasma Sci., 2005, 33(2), 338–339.
83. H. Polakova, J. Musil, J. Vlcek, J. Alaart and C. Mitterer, Structure-hardness relations in sputtered Ti-Al-V-n films, Thin Solid Films, 2003, 444, 189–198.
84. J. Musil, Sputtering systems with enhanced ionization for ion plating of hard wear resistant coatings, Proc. of the 1st Meeting on the Ion Engineering Society of Japan (IESJ-92), Tokyo, Japan, 1992, pp. 295–304.
85. H. Poláková, J. Musil, J. Vlček, J. Alaart and C. Mitterer, Structure- hardness relations in sputtered Ti-Al-V-N films, Thin Solid Films, 2003, 444, 189–198.
86. J. Musil, H. Polaková, J. Šůna and J. Vlček, Effect of ion bombardment on properties of hard reactively sputtered single-phase films, Surf. Coat. Technol., 2004, 177–178, 289–298.
87. J. Musil and J. Šůna, The role of energy in formation of sputtered nanocomposite films, Mater. Sci. Forum, 2005, 502, 291–296.
88. J. Musil, J. Šícha, D. Heřman and R. Čerstvý, Role of energy in low-temperature high-rate formation of hydrophilic TiO₂ thin films using pulsed magnetron sputtering, J. Vac. Sci. Technol., A, 2007, 25(4), 666–674.
89. J. Musil, J. Lestina, J. Vlcek and T. Tolg, Pulsed dc magnetron discharge for high-rate sputtering of thin films, J. Vac. Sci. Technol., A, 2001, 19, 420–424.
90. J. Vlcek, A. D. Pajdarova and J. Musil, Pulsed dc magnetron discharges and their utilization in plasma surface engineering, Contrib. Plasma Phys., 2004, 44(5–6), 426–436.
91. R. D. Arnell, P. J. Kelly and J. W. Bradley, Recent developments in pulsed magnetron sputtering, Surf. Coat. Technol., 2004, 188–189, 158–163.
92. A. P. Ehiasarian, Fundamentals and applications of HIPIMS, in Plasma Surface Engineering Research and its Practical Applications, ed. R. Wei, Research Signpost Publisher, Kerala, 2008, ch. 2, pp. 35–86.
93. A. Anders, J. Andersson and A. Ehiasarian, High power impulse magnetron sputtering: current-voltage-time characteristics indicate the onset of self-sputtering, J. Appl. Phys., 2007, 102, 113303.
94. U. Helmersson, M. Lattemann, J. Bohlmark and A. P. Ehiasiaran, Ionized physical vapor deposition (IPVD): A review of technology and applications, Thin Solid Films, 2006, 513(1–2), 1–24.
95. J. Vlcek, P. Kudlacek, K. Burcalova and J. Musil, High-power pulsed sputtering using a magnetron with enhanced plasma confinement, J. Vac. Sci. Technol., A, 2007, 25, 42–47.
96. J. Vlcek, P. Kudlacek, K. Burcalova and J. Musil, Ion flux characteristics in high-power pulsed magnetron sputtering discharges, EPL, 2007, 77, 45002.
97. D. Horwat and A. Anders, Spatial distribution of average charge state and deposition rate in high power impulse magnetron sputtering of copper, J. Phys. D: Appl. Phys., 2008, 41, 135210.
98. J. W. Bradley and T. Welzel, Physics and phenomena in pulsed magnetrons: an overview, J. Phys. D: Appl. Phys., 2009, 42, 093001.
99. A. Anders, Deposition rates of high power impulse magnetron sputtering: physics and economics, J. Vac. Sci. Technol., A, 2010, 28(4), 783–790.
100. P. Poolcharuansin and J. Bradley, Shot- and long-term plasma phenomena in a HIPIMS discharge, Plasma Sources Sci. Technol., 2010, 19, 025010.
101. A. Anders, M. Panjan, R. Franz, J. Andersson and P. Ni, Drifting potential humps in ionization zones: the “propeller blades” of high power impulse magnetron sputtering, Appl. Phys. Lett., 2013, 103, 144103.
102. P. M. Barker, E. Lewin and J. Patscheider, Modified high power impulse magnetron sputtering process for increased deposition rate of titanium, J. Vac. Sci. Technol., A, 2013, 31(6), 060604.
103. J. Musil, J. Blažek, K. Fajfrlík, R. Čerstvý and Š. Prokšová, Antibacterial Cr-Cu-O films prepared by reactive magnetron sputtering, Appl. Surf. Sci., 2013, 276, 660–666.
104. (a) J. A. Thornton, Recent developments in sputtering – magnetron sputtering, Met. Finish., 1979, 77(5), 83–87; (b) J. A. Thornton, High rate thick films growth, Annu. Rev. Mater. Sci., 1977, 7, 239–260.
105. B. A. Movchan and A. V. Demchishin, Study of the structure and properties of thick vacuum condensates of nickel, titanium, tungsten, aluminum oxide and zirconium oxide, Phys. Met. Metallogr., 1969, 28, 83–90.
106. J. Musil, V. Poulek, V. Valvoda, R. Kužel Jr., H. A. Jehn and M. E. Baumgartner, Relation of deposition conditions of Ti-N films prepared by dc magnetron sputtering to their microstructure and macrostress, Surf. Coat. Technol., 1993, 60, 484–488.
107. P. Pokorný, J. Musil, P. Fitl, M. Novotný, J. Lančok and J. Bulíř, Contamination of magnetron sputtered metallic films by oxygen from residual atmosphere in deposition chamber, Plasma Processes Polym., 2015, 12, 416–421.
108. R. Messier, A. P. Giri and R. A. Roy, Revised structure zone model for thin film physical structure, J. Vac. Sci. Technol., A, 1984, 2(2), 500–503.
109. A. Anders, A structure zone diagram including plasma-based deposition and ion etching, Thin Solid Films, 2009, 518, 4087–4090.

---