ISSN 00360295, Russian Metallurgy (Metally), Vol. 2010, No. 10, pp. 917–935. © Pleiades Publishing, Ltd., 2010. Original Russian Text © E.A. Levashov, D.V. Shtansky, Ph.V. KiryukhantsevKorneev, M.I. Petrzhik, M.Ya. Tyurina, A.N. Sheveiko, 2009, published in Deformatsiya i Razrushenie Materialov, 2009, No. 11, pp. 19–36.
PROMISING MATERIALS AND TECHNOLOGIES
Multifunctional Nanostructured Coatings: Formation, Structure, and the Uniformity of Measuring Their Mechanical and Tribological Properties E. A. Levashov, D. V. Shtansky, Ph. V. KiryukhantsevKorneev, M. I. Petrzhik, M. Ya. Tyurina, and A. N. Sheveiko National University of Science and Technology “MISIS”, Leninskii pr. 4, Moscow, 119049 Russia email:
[email protected] Abstract—The state of the art of the studies in the field of the development and certification of novel multi functional nanostructured coatings having a wide spectrum of applications is reviewed. The main tendencies in the optimization of the compositions and properties of the coatings are described, and the modern meth ods of diagnostics of the nanostructured coatings are considered. DOI: 10.1134/S0036029510100113
INTRODUCTION
MODERN FUNCTIONAL COATINGS
The development and use of multifunctional nano structured coatings (MNCs) is one of the challenging problems of modern materials science. The demand for MNCs is caused by their high service properties, which are much higher than the properties of the cor responding bulk materials and microcrystalline coat ings. MNCs are characterized by a significant volume fraction of interfaces, a high interfacial binding energy, the absence of dislocations inside nanocrystallites, deformation via grainboundary sliding, the presence of amorphous layers between crystallites, and a change in the mutual solubility of components in interstitial phases. These specific features result in unique prop erties, such as a high hardness (H > 30 GPa), elastic recovery (R > 70%), high thermal stability, high oxida tion resistance at temperatures above 1000°C, a high electrical resistance, and so on.
Hard WearResistant Coatings
MNCs are also irreplaceable for creating nextgen eration biocompatible materials, including orthopedic and dental implants for skull–jaw–face surgery, and for fixing the neck and loin parts of the spinal column, in which the multifunctional (chemical, mechanical, tribological, biological) properties of thinfilm mate rials fully manifest themselves. In this review, we consider the structure and prop erties of promising MNCs based on refractory com pounds designed for the needs of machine building and medicine and the modern methods of their pro duction and certification.
MNCs are used to protect the surface of products and tools subjected to various types of wear and a simultaneous action of high temperatures and aggres sive media. These products and tools include cutting and die tools, rolls, parts of aviation engines, parts of gas turbines and compressors, plain bearings, and extrusion nozzles for glass and mineral fibers. To with stand wear, a coating must have high hardness H, good elastic recovery R, and low elastic modulus E. When estimating the wear resistance of the coating, one should also take into account the long elastic strain to failure H/E and the resistance to plastic deformation H 3/E 2 [1]. It should be noted that parameter H 3/E 2 can be used to predict the type of fracture during local ized deformation of MNCs [2]. The hardness and elastoplastic characteristics of coatings can be significantly increased by the forma tion of a nanostructured state in them. When addi tional elements (such as silicon or boron) are intro duced into the composition of a wellknown titanium nitride TiN coating, the crystallite size decreases from several hundred to several nanometers [3, 4]. This effect is caused by the fact that TiNbased crystalline phases that have a limited solubility of a third element (Si, B) form during deposition of TiSiN and TiBN coatings. The introduced elements segregate along crystallite boundaries, recrystallization is retarded, and an amorphous phase forms. The introduction of modifying additions decreases the crystallite size and increases the coating hardness according to the Hall– Petch relation. Moreover, the formation of a columnar structure, which degrades the mechanical properties, is suppressed [4, 5].
917
918
LEVASHOV et al. (a) Vw × 5 10
10–7,
0
(b)
mm3
N–1
m–1
15
20
25
0
30 Ti–N
10
Tool life, min 20 30 40 50
60
Without coating
Cr–Ti–Al–C–N Ti–Si–N
TiN
Ti–Cr–Si–C–N Ti–Al–C–N
Ti–B–N
Ti–C–N Cr–B–N
Ti–Si–B–N
Ti–Cr–Cu–B–N Ti–Cr–B–N
Ti–Cr–B–N
Ti–B–N
Fig. 1. (a) Reduce wear Vw of coatings tested according to the pinondisk scheme and (b) wear resistance of VK6 hardalloy cut ting plates coated with TiN layers during dry turning of steel 321 (Cr18Ni10MnTi, 170 HB).
TiSiN coatings are characterized by a high hard ness (25–45 GPa) [4, 6] and a rather low elastic mod ulus (170–250 GPa) [2, 4]. The mechanical properties of such coatings depend substantially on the silicon content and become maximal at 5–10 at % Si [4, 6]. There exist data on TiSiN coatings whose hardness exceeds 100 GPa [7, 8]. The authors of [7, 8] proposed a concept that the maximum hardness is achieved due to the formation of a nanocomposite structure, which consists of hardphase nanocrystallites separated by thin amorphousphase layers. These phases should be immiscible and have a high interfacial energy and a high melting temperature. Tribological tests demonstrated that the friction coefficient of TiSiN coatings decreases with increasing silicon content [4, 9], which is related to the formation of SiO2 or Si(OH)2based tribological layers playing the role of a solid lubricant. The study of the wear resistance of TiNbased coatings using the pinon disk scheme (the normal load was 5 N, the speed was 10 cm/s, and the counterbody consisted of a WC–Co ball 3 mm in diameter) showed that wear Vw of a TiSiN coating is half that of a TiN coating (Fig. 1a). TiBN coatings are characterized by a high hardness (30–55 GPa) [2, 10–12] and high resistance to impactdynamic actions [13] and abrasive wear [12]. The hardness of such coatings can be ultrahigh if they contain the BN phase in the form of amorphous inter crystalline layers [7, 8]. As compared to TiN coatings, TiBN coatings are characterized by a lower friction coefficient, and a positive effect of the amorphous a BN phase was established (hereafter, an amorphous state is designated by letter a) [10]. In contrast to expectation, hexagonal boron nitride in TiBN coat ings does not provide a lubricating effect [14].
When chromium is introduced into TiBN coatings, substantial structure refinement takes place and the friction coefficient decreases significantly [10]. TiCrBN coatings can increase the life of cutting tool by many times (Fig. 1b). The addition of aluminum to boron nitride coatings favors the formation of a TiAlBN coating with an extremally small crystallite size (0.3–0.8 nm) [15]. Superhard TiSiBN coatings with a hardness of 70 GPa were fabricated in [12, 16]. The hardness of these coatings was maximal at a strictly stoichiometric composition of hexagonal phase Ti(B, N)2. The con tribution of residual compressive stresses to the total hardness was 10–15%. The resistance of cutting tool coated with TiSiBN was three to four times higher than that coated with TiN (see Fig. 1b), and the cut ting time was two times shorter [17]. Carboncontaining MNCs, such as TiBCN, TiSiCN, and TiAlCN, are characterized by a low fric tion coefficient (0.2–0.3), which is related to a posi tive effect of carbon (which plays the role of a solid lubricant). The coating hardness decreases with increasing carbon content. TiAlCN and TiSiCN coat ings have high abrasive wear resistance [3, 12]. Apart from coatings based on refractory titanium compounds, researchers have recently devoted close attention to designing nanostructured WSiN, VSiN, and ZrSiN coatings, whose hardness and elastoplastic characteristics are higher than those of the corre sponding base coatings [7, 8, 18]. Nanostructured CrBN and CrTiAlCN coatings, which have high tribo logical properties (see Fig. 1a), have recently been formed [19, 20].
RUSSIAN METALLURGY (METALLY)
Vol. 2010
No. 10
MULTIFUNCTIONAL NANOSTRUCTURED COATINGS
SelfLubricating Coatings The friction coefficients of coatings based on the refractory compounds of transition metals can be decreased as a result of the formation of nanocompos ite or layered structures containing solid and self lubricating phases or layers. When the volume fraction of a selflubricating phase increases, both the friction coefficient and hardness decrease correspondingly. Chameleontype MNCs intended for spacecraft parts are described in [21]. (They are called chame leon coatings, since the chemical composition of tri bological layers on the coating surface changes as a function of the ambient composition and temperature during friction.) Hard phases WC, TiC, and Al2O3 pro vide high wear resistance, and selflubricating phases, such as aC, nsWS2 (ns designates a nanostructured state of a material), aBN, and Au, decrease the fric tion coefficient in a contacting pair. Diamondlike carbon decreases the friction coefficient of a coating to 0.1–0.15 in humid air at low temperatures; chalco genides decreases the friction coefficient to less than 0.1 in vacuum or dry nitrogen; and metals and chalco genides surrounded by a ceramic matrix decrease the friction coefficient to 0.2 at high temperatures (about 500°C). MoS2 is most often used as a lubricating phase in nanocomposites and layered coatings. Moreover, Ti/MoS2, TiN/MoS2, TiSiN/MoSx, and TiB2/MoS2 are well known. However, molybdenum and tungsten diselenides are more promising in some cases due to high oxidation resistance and high friction coefficient stability at a high air humidity [22]. Twolayer coatings WSex /TiC, WSex /TiCN, and WSex /TiSiN, in which the upper layer consists of nanocrystalline phase nsWSe2 or nsW3O embedded into amorphous matrix aWSex [23], have a stable low friction coefficient in both air (0.015–0.05) and water (0.06–0.07). These coatings are characterized by the absence of a high friction coefficient during the runin period in a contacting pair, which is typical of coatings with MoS2. According to [24], the friction coefficients of synthesized nanocomposite and nanolayered TiCrBN/WSex coatings decrease with increasing lubricating phase WSex phase content: at 10 at % WSex, it is 0.2, which is lower than that of the base TiCrBN coating by a factor of 2.5. It should be noted that the wear resistance and hardness of TiCrBN/WSex coatings are identical to those of TiCrBN coatings. HeatResistance Coatings The lifetime of parts operating at high temperatures depends on their thermal stability and heat resistance. These characteristics are controlled by the elemental and phase compositions of coatings and the thermal stability and barrier properties of the phases making up their base. RUSSIAN METALLURGY (METALLY)
Vol. 2010
919
The hardness of widely used microcrystalline TiN, TiC, TiCN, and TiAlN coatings decreases monotoni cally with increasing temperature, which is related to recrystallization and stress relaxation. However, the hardness of MNCs can remain high up to 1000°C [11]. For example, the hardness of TiCrBN coatings is 30 GPa and remains unchanged up to 1000°C inclu sively, an the hardness of TiBN coatings increases by 30% after heating to 600–800°C [25]. This hardening effect can be explained by a change in the thickness of intergranular amorphous layers and the concentration separation of supersaturated solid solutions (metasta ble phases) induced by spinodal decomposition. In the temperature range 20–1000°C, TiBN and TiCrBN coatings have a stable structure with an average crys tallite size of 2–5 nm [25]. Intergranular amorphous layers hinder recrystallization. An important property of thin coatings consists in their ability to withstand diffusion of such elements as Co, Ni, Fe, and Cr from a substrate material during heating. It was found in [25] that MNCs containing a crystalline phase with a cubic lattice hinder the diffu sion of metallic elements from a substrate to a greater extent as compared to coatings with a hexagonal phase, in which impurity atoms can easily diffuse between basal planes. Microcrystalline coatings are oxidized via reactive diffusion of oxygen along grain boundaries, which is the main cause of the relatively low heat resistance of a TiN coating (it is restricted to 550°C). The heat resis tance of TiSiN, CrBN, TiAlCN, and TiSiBN coatings, whose structure consists of metal nitride nanocrystal lites separated by amorphous aSi3N4, aBN, or aAlN layers, is 800–1000°C [20, 26], which is explained by the fact that oxygen diffusion through disordered regions is strongly hindered. In turn, the oxides (e.g., SiO2) that form during the oxidation of the amorphous phases are good diffusion barriers. When the amorphous phase content in a coating increases, the degree of nanocrystallite separation increases and the heat resistance also increases [27]. The heat resistance of a coating can be increased by introducing the elements that form protective oxide lay ers on the coating surface, such as aluminum, chro mium, silicon, and yttrium. For example, oxygen dis solves in the cubic lattice of a TiAlCN coating at 550°C, and aluminum diffuses to the coating surface at 800°C and forms a protective Al2O3 layer, which prevents fur ther oxidation [12]. The authors of [5, 25, 28] showed that the introduction of chromium, silicon, and alumi num into the composition of a TiBN coating increases its heat resistance to 900°C due to the formation of pro tective Cr2O3 and TixAlySiOz oxides on the coating sur face. CrAlCN and CrAlTiCN coatings withstand oxi dation up to 1000°C owing to the formation of protec tive CrOx, AlOx, CrCx, and AlNxOy layers on their surfaces; these layers were detected by glow discharge optical emission spectroscopy (Fig. 2).
No. 10
920
LEVASHOV et al.
O Cr
50 40 30
C
20
Al N
10 2 3 Depth, μm
4
5
Fig. 2. Elemental depth profile of a CrAlCN coating after steplike annealing (600°C, 5 h; 800°C, 1 h; 1000°C, 1 h) according to glow discharge optical emission spectroscopy.
CorrosionResistant Coatings The corrosion resistance of coatings is mainly determined by their chemical composition, the type of structure, and (to a lesser extent) the crystallite size. For example, the authors of [12] showed that the cor rosion resistance of nanostructured TiSiCN coatings in a 5 N H2SO4 solution is comparable with that of microcrystalline TiN, TiC, and TiCN coatings and that the corrosion resistance of MNCs containing alu minum and boron is lower by a factor of 5–10. The highest negative potential and the highest corrosion
Bioactive Coatings Titanium nitride is widely used as a coating for prostheses in stomatology, traumatology, and cardio surgery. This is caused by a combination of its high hardness, relatively high wear and corrosion resistance in physiological media, and bio and hemocompati bility. The main disadvantage of TiN is its low bioactiv ity, which limits its application for implants. To increase the bioactivity, a coating made of hydroxyla patite (HAP) of the composition Ca10(PO4)6(OH)2 is often deposited onto the surface of medical implants. A HAP layer contains calcium and phosphorus and initiates the formation of a bonelike apatite layer on its
E, mV 800 600 400 200 0 –200
Ti–Cr–Si–C–N
Ti–C–N
Cr–Ti–Al–C–N
Ti–N
Cr–Al–C–N
Ti–Al–C–N
Ti–Si–N
Ti–Cr–B
Ti–B–N
Ti–B
Cr–B
Ti–Si–B–N
Ti–Si–B
Ti–Al–Si–B–N
Ti–Al–Si–B
–400 Ti–Si
1
Cr–B–N
0
W
Ti–Cr–B–N
Concentration, at %
60
rate were found to be characteristic of TiSiBN coatings with a hexagonal structure. TiCrBN coatings have a higher positive free corro sion potential and corrosion resistance that is four times lower than that of TiBN [5]. The corrosion prop erties of TiCrBN coatings improve with the chromium content. The corrosion rates of TiSiBN and TiAlSiBN coatings are higher than that of a TiBN coating by a factor of 6 and 14, respectively. CrB2 is a promising corrosionresistant material. CrBN coatings consisting of CrB2 crystallites and amorphous regions based on boron nitride have higher corrosion resistance than CrB coatings [20]. Figure 3 shows the steadystate free corrosion potential of coat ings with various elemental compositions in a 1 N H2SO4 solution.
Ni
70
Fig. 3. Free corrosion potentials of nanostructured coatings in a 1 N H2SO4 solution. RUSSIAN METALLURGY (METALLY)
Vol. 2010
No. 10
Number of cells
MULTIFUNCTIONAL NANOSTRUCTURED COATINGS 100 90 80 70 60 50 40 30 20 12 3 4 5 10 0 1
(а)
921
(b)
0.1 mm 3
5
7
Days Fig. 4. (a) Proliferation of MS3T3E1 osteoblasts on the surface of various bioactive nanostructured coatings and (b) the interface between a bone and the surface of a titanium implant with a TiCaCPON coating in 1 month after implantation: (1) reference (medical glass without a coating), (2) TiCaCOP, (3) TiCaCPON, (4) TiCaCO, and (5) TiCaCON.
surface, through which it connects with a bone tissue. However, HAP cannot be used in the case of implants operating under a load, since it has limited plasticity and a low strength and cracking resistance as com pared to the bone tissue. A radically new approach to designing MNCs with a high bioactivity has recently been developed: additional elements (Ca, Zr, Si, O, P) are introduced into coatings based on refractory compounds TiC, Ti(C, N), (Ti, Ta)C, and (Ti, Ta)(C, N) in order to improve the mechanical and tribological properties of a material and to provide bioactive properties and biocompatibil ity of the material surface. Inorganic additions CaO, ZrO2, Si3N4, TiO2, Ca3(PO4)2, and Ca10(PO4)6(OH)2 are introduced at the stage of selfpropagating high temperature synthesis of composite targets for vacuum deposition of coatings [26, 29–32]. As follows from a combination of the elastoplastic and tribological char acteristics, the developed MNCs substantially exceed such medicalpurpose materials as Ti, TiAlV, TiN, ZrO2, TiO2, SiO2, Al2O3, and HAP, since they have a high hardness (25–35 GPa), a relatively low elastic modulus (less than 300 GPa), high elastic recovery (70–80%), long elastic strain to failure of at least 0.1, plastic deformation resistance higher than 0.2 GPa, a friction coefficient less than 0.25 in air and a physiological solution, and a reduced wear less than 10–5 mm3 N–1 m–1. In particular, in vitro cytocompatibility estimated using cells of connective (fibroblasts, epithelial cells) and bone (osteoblasts) tissues showed that TiCaCON and TiCaCOPN coatings ensure satisfactory spread ing, proliferation (growth), and differentiation (for mation of a special phenotype) of cells [29, 31]. A simultaneous introduction of calcium and phosphorus in the composition of coatings is most efficient to intensify bone tissue growth (Fig. 4a) [32]. It was also found that the introduction of certain elements, e.g., niobium or zirconium (at high contents), destroys an actinic cytoskeleton of the cells [30]. RUSSIAN METALLURGY (METALLY)
Vol. 2010
In vivo experiments showed that TiCaCO(P)N coatings are nontoxic and do not cause inflammatory reactions [30, 31]. The results of implantation of coated samples into a bone tissue in order to study osteointegration according to the degree of bone resorption around an implant and the structure and maturity of a connective tissue and newly formed cell structures are described in [30]. Figure 4b illustrates the results obtained in one month after the introduc tion of an implant coated with TiCaCPON. Regions with close contact between the implant and bone with out coarse fibrous structures, gaps, and voids are visi ble. The introduction of additional elements such as silicon and tantalum into the composition of bioactive coatings is discussed in [30, 32–34]. The introduction of tantalum increases the coating hardness to 40 GPa [33, 34]. METHODS OF DEPOSITION OF NANOSTRUCTURED COATINGS Magnetron Sputtering of SHS Targets To produce MNCs, researchers widely apply chemical deposition, including plasmaassisted depo sition, and physical deposition methods, such as mag netron sputtering, acr evaporation, electronbeam evaporation, and ionbeam sputtering. The advantage of magnetron sputtering consists in insignificant (by 50–250°C) heating of a substrate [16], which makes it possible to deposit coatings on almost any materials. Moreover, with this technology, one can deposit hard and superhard MNCs with vari ous levels of elastoplastic characteristics [18]. The possibilities of magnetron sputtering can be substantially extended using composite multicompo nent cathode targets produced by selfpropagating hightemperature synthesis (SHS) [35]. SHS is efficient for the production of a wide range of targets based on ceramics, metalloceramics, and intermetallics. The radical difference between the processes of sputtering of
No. 10
LEVASHOV et al. (a)
(b) N
2
3
8 7 6 5
S
S
N
N
S
S
3
4
3
1
2
4 N
S
N S N
1
1
4
2 N S N
9
S
S N S
N
6
S N S
922
5
N
Fig. 6. Schematic diagrams of commercial setups for mag netron sputtering of nanostructured coatings: (a) close field unbalanced magnetron system and (b) dual magne tron systems: (1) magnetron, (2) target, (3) lines of mag netic field, (4) support for fixing substrates, (5) ion source for substrates cleaning, and (6) heater.
10
implantation is used as an individual technology oper ation before the main stage of sputtering or after dep osition of coatings. This method ensures a high adhe sion strength of coatings. The significant increase in the adhesion strength induced by ion implantation is caused by the presence of extended transition layers between a substrate and a coating and a significant decrease in the residual internal stresses [34].
Fig. 5. Schematic diagram of a setup for magnetron and ion sputtering with simultaneous ion implantation: (1) magnetron, (2) ion source for sputtering, (3) ion source for substrates cleaning, (4) ion implanter, (5) substrates, (6) substrate holder, (7) heater, (8) thermocouple, (9) rota tion drive, and (10) pumping system.
composite and metallic targets consists in the fact that, in the former case, a homogeneous flow of metallic and nonmetallic atoms and ions is transferred to a substrate. All elements necessary for coating formation, including nonmetallic elements (C, O, N, P), can be sputtered from one target [4, 10]. Both disklike and extended rectangular segment SHS targets can be used in sputter ing devices [36]. SHS targets have been successfully tested in setups of various types, namely, dc magnetron systems [2–5, 23–26, 28–34], highfrequency [37] and pulsed [36] magnetron systems, magnetron systems with an addi tional inductively coupled plasma [38], and arc evap orators [39]. Magnetron sputtering of SHS targets was used to produce the following hard coatings: TiSiN [2, 4, 9], TiBN [3, 10, 12, 17], TiSiBN [3, 12, 16], TiSiCN [3, 12], TiAlCN [3, 12], TiCN [40], TiMoCN [40], TiAlBN [15], TiAlSiBN [17, 25, 28], TiCrBN [2, 5, 10, 17, 25], CrB [2, 10, 19, 20], TiZrCON [29], TiTaCaCPON [33, 34], etc. Magnetron and Ion Sputtering Assisted by Ion Implantation The authors of [17] developed a combined method of magnetron sputtering and ion implantation. In this method, highenergy metal ions are implanted into a substrate and a coating during coating formation by magnetron sputtering of a composite target, in con trast to the wellknown technologies in which ion
Combined magnetron and ion sputtering makes it possible to deposit coatings from both conductive and nonconductive (oxides, polymers, etc.) targets. The authors of [24] described the deposition of nanocom posite and nanolayer TiCrBN/WSex coatings during simultaneous and alternate operation of a magnetron and an ion source. The schematic diagram of the setup is shown in Fig. 5. Magnetron Sputtering in Setups with Pulsed Power Supply and a Complex Configuration of Magnetic Fields One of the promising methods for the deposition of hard coatings is pulsed magnetron sputtering, in which a magnetron is powered at a frequency of 20–400 kHz. Deposited coatings have a dense structure owing to a high degree of ionization [41]. An increase in the den sity and a decrease in the imperfection of a coating lead to an increase in their physicomechanical proper ties. For example, when pulsed power supply was used for magnetrons, the hardness of coatings in the TiSiBN system increased from 20 to 40 GPa [36]. To produce MNCs on a commercial scale, special ists now widely use unbalanced magnetron systems with closed magnetic fields, which form a stable plasma in a large volume and can substantially increase the degree of ionization of a sputtered mate rial [41, 42]. For example, such systems are applied to deposit Ti/MoS2, TiSiN/MoSx, TiC/C, CrAlN, and Cr/C coatings (Fig. 6a).
RUSSIAN METALLURGY (METALLY)
Vol. 2010
No. 10
MULTIFUNCTIONAL NANOSTRUCTURED COATINGS
The use of dual unbalanced magnetron sputtering systems is rather promising (Fig. 6b). When a pulsed bipolar voltage is applied to dual magnetrons, one magnetron operates as a cathode and the other oper ates as an anode in the first halfcycle and vice versa in the second halfcycle. This magnetron operation regime makes it possible to obtain a high degree of plasma ionization, to exclude microparticle formation (which is possible in a conventional magnetron dis charge), to decrease the degree of target contamina tion during operation in a reaction mode, and to increase the deposition rate of reaction coatings.
923
(10
0)
0) (01
2 nm
MODERN METHODS FOR THE EXAMINATION OF THE STRUCTURE AND COMPOSITION OF THIN NANOSTRUCTURED COATINGS XRay Diffraction and Electron Microscopy As a rule, the examination of MNCs includes the determination of their elemental and phase composi tions, the lattice parameters, the grain size of a crystal line phase, surface morphology and topography, and internal stresses. Xray diffraction analysis is one of the most widely used methods for the diagnostics of the structure of coatings. The study of thin MNCs encounters difficulties related to a low line intensity, a substantial line broadening of a dispersed crystalline component, the superposition of diffraction lines of several phases, a variable composition of a crystalline component characterized by different lattice parame ters, and the presence of macro and microstresses. Moreover, it is often difficult to interpret Xray diffrac tion patterns because of a strong texture in a coating. To solve these problems, one can take additional Xray diffraction patterns from a tilted sample or from refer ence samples with a simpler chemical composition [4, 9]. A glancing Xray beam method is promising, since it can minimize a substrate signal [23]. Xray dif fraction analysis is often applied to determine internal stresses using the sin2ψ technique. The contribution of thermal stresses can be estimated when coatings deposited onto substrates having strongly different thermal expansion coefficients are compared [4, 9]. It should be noted that there is an alternative method for determining stresses from the bending flexure of a coated substrate [19, 34]. Transmission electron microscopy (TEM) can rather accurately determine the phase composition of a coating and to estimate the crystallite size in MNCs using darkfield images. However, if the crystallite size is about 1 nm or smaller (which coincides with the scale of amorphous contrast), it is impossible to unambiguously interpret TEM results. Highresolu tion transmission electron microscopy (HRTEM) is a powerful tool for determining the crystalline structure of individual grains smaller than 1 nm [43]. Most modern microscopes have a resolution of 0.2 nm, RUSSIAN METALLURGY (METALLY)
Vol. 2010
Fig. 7. Crystallite 4–5 nm in size in a TiCrBN coating, HRTEM [2].
which allows one to observe twodimensional contrast from a nanocrystallite situated in a reflecting position (Fig. 7). The angle between atomic planes and the interatomic distances unambiguously indicate a cubic structure in the nanocrystallite in Fig. 7. The image around the nanocrystallite has contrast with a random intensity distribution on the atomic scale, which is an indication of a random (amorphous) structure. It should be noted that such an image can also be caused by other factors, such as local stresses, the superposi tion of nanocrystallites, the neighboring grain situated in a nonreflecting position, and impurity segregation. In [28], HRTEM was used to study the oxidation of TiCrBN and TiAlSiCN coatings. TEM combined with other investigation methods, such as electronenergy loss spectroscopy, can more accurately determine the phase composition of MNCs [15, 44]. Scanning electron microscopy (SEM) is necessary to study the morphology of coatings and can be used to analyze the deformation of MNCs [2, 45]. It is often combined with energy dispersive spectroscopy and Rutherford backscattering spectroscopy in order to estimate the elemental composition of coatings and to comprehensively investigate their wear, oxidation, and corrosion [24]. Spectroscopic Methods The most widely used spectroscopic methods of inves tigating MNCs are represented by Xray photoelectron spectroscopy (XPS), Raman spectroscopy, and infrared Fourier spectroscopy. XPS is successfully used to deter mine the elemental and phase compositions of MNCs containing fine crystalline and amorphous phases, when the application of diffraction methods is limited. It is opti mal to use XPS in combination with Xray diffraction and TEM. For example, the authors of [4, 9, 10, 19] employed this combination of methods to study nsTiN/aTiSix, nsTiN/aTiSixNy, nsTiN/aBN, nsTiN/aCrB2, nsTiN/aTiB2/aBN, and nsCrB2/aBN MNCs.
No. 10
924
LEVASHOV et al.
Intensity, arb. units
μm 0
198
2 4
1
6
2 194
190 186 Binding energy, eV
182
226 μm
8 297 μm 10
Fig. 8. 1s boron (B1s) spectra of TiBN coatings deposited in (1) a mixture of argon with nitrogen and (2) pure argon. The vertical solid and dashed lines indicate the binding energies of the TiB2 and BN phases, respectively [10].
Fig. 9. Threedimensional image of the fracture zone in a nanostructured TiCrBN coating.
Figure 8 shows the highresolution energy spectra of TiBN coatings fabricated in argon and a gas mixture of argon with nitrogen. In the nonreactive coating, Ti and B are bound in a phase whose composition is close to TiB2; in the reactive coating, B is also bound with N in the BN phase. Thus, the coatings have substantially different phase compositions, which eventually causes their different mechanical and tribological character istics [10, 17]. It should be noted that Xray diffraction and TEM were able to reveal only a TiNbased crystal line phase in the coatings [10, 17]. Raman spectroscopy can be used to estimate a phase composition [24, 46], which gives information on the thermal stability of MNCs [46] or the type of tribochemical reactions [24]. Infrared Fourier spec troscopy can give additional information on the func tional groups on the surface of bioactive MNCs that form during storage in air [34]. Among the spectroscopic methods of analysis, glow discharge optical emission spectroscopy should also be noted. With this method, one can rapidly obtain information on the elemental composition and distribution across MNCs and comprehensively study the diffusion, oxidation, and corrosion of MNCs (see Fig. 2). Scanning Probe Microscopy and Optical Profilometry These methods are usually employed to study the surface topography and the roughness parameters of MNCs [2, 4, 10, 31, 34]. In [2, 45], scanning probe microscopy was used to analyze the mechanisms of localized deformation in MNCs. Scanning probe microscopy can be combined with sclerometry, and this combination can be use to reveal the specific fea tures of the morphology of MNCs that cannot be detected with SEM and TEM [2]. The replacement of contact profilometry by con tactless one can eliminate the effect of the needle tip radius on the measurement results. A contactless pro
filometer can be used to analyze the mechanisms of adhesion and cohesion fracture of MNCs, the corro sion of MNCs, and the tribological processes occur ring in them [10, 20, 24]. Figure 9 shows a three dimensional image of the fracture zone in a TiCrBN coating during adhesion tests. UNIFORMITY OF ASSURANCE OF MEASUREMENT MECHANICAL AND TRIBOLOGICAL PROPERTIES OF THE NANOSTRUCTURED SURFACES OF NANOINDUSTRY PRODUCTS The leading manufacturers of scientific and research equipment have recently developed high accuracy devices that are intended for quantitative measurements and complex interpretation of the mechanical and tribological properties of the nano structured surfaces of bulk materials and coatings on the micro and nanolevels and also are based on nanoindentation, instrumented scratching, impact cyclic and tribological measurements, and optical and contact profilometry. However, a metrologically pure analysis and the commercialization of nanoindustry products require not only apparatus, but also standardmethodological support of the uniformity of measurements (measure ment techniques, calibration techniques, standards, practical recommendations); calibration and reference samples; certified consumable materials (indenters, counterbodies); and qualified staff, which can perform measurements, processing of results, and calibration and service functions (i.e., the set of procedures regu lating mechanical and tribological tests). To solve this challenging problem and to provide the uniformity of measuring the hardness, elastic modulus, elastic recovery, adhesion strength, friction coefficient, wear, fatigue limit, and surface topology, researchers at the Moscow State Institute of Steel and Alloys have started to create a metrological set that
RUSSIAN METALLURGY (METALLY)
Vol. 2010
No. 10
MULTIFUNCTIONAL NANOSTRUCTURED COATINGS
925
Table 1. Surface information methods under conditions of mechanical contact with an indenter (counterbody) Characteristics Realtime parame ters
Scheme of sample–counterbody mechanical contact indentation Load–indentation depth
scratching Load–displacement– indentation depth
Calculated properties Hardness, elastic mod Mohs hardness, adhe ulus, elastic recovery sion/cohesion strength
includes the following modern equipment: nanoin denter (nanohardness tester), scratch tester (adhesion meter), tribometer, atomic force microscope, optical and contact nanoprofilometer, impact tester, and scanning nanoindenter. This metrological set should provide measurements over wide ranges and uncer tainties of the physical quantities measured on the nanoscale. The creation and certification of reference samples of nanomaterials is also one of the most important metrology problems. We now consider the possibilities of the modern methods of studying the functional surfaces solid phase materials under mechanical contact conditions for measuring the hardness, elastic modulus, elastic recovery, adhesion strength, friction coefficient, wear, fatigue limit, and surface topography, i.e., the param eters that control the service properties of nanoindus try products. Contact of Two Solids in the Theory of Elasticity and Its Applications The methods under study are based on an analyti cal solution to the socalled Hertz problem of the mutual deformation of two solid balls during compres sion [47], which was comprehensively considered in [48]. One of the technically important formulations of this problem is the hemisphere–plane contact scheme (which is often observed in practice), i.e., the interac tion of spheres of finite and infinite radii. The use of Hertz’s model for the description of the interaction of an indenter and a sample is allowable and helpful for an analytical consideration, since the vertex of any indenter can be described as a hemi sphere of a certain radius using the socalled radius of curvature as a parameter. This approach is considered to be correct in the case where the surface forces are negligibly low as compared to the total interaction forces and the contact radius is substantially smaller than the radius of curvature. These conditions make up the basis of the methods of estimating the func tional properties of surface layers, and they were used in the measuring devices developed in the last few years. The modern methods of studying the response of a surface to the mechanical contact with an indenter (counterbody) include the following stages: a test with RUSSIAN METALLURGY (METALLY)
Vol. 2010
sliding
impact cyclic loading
Friction coefficient–dis Number of impacts placement under a constant under a fixed load load Wear resistance of coating Fatigue limit (from fractographic data)
realtime recording of experimental data by a com puter; an analysis of these data; and the examination of the traces of the mechanical contact by optical, probe, and electron microscopy and contact and con tactless profilometry (Table 1). As a rule, indentation and scratching methods use a diamond indenter, and its geometry is verified by cal ibration before tests. In tests according to a sliding scheme, the counterbody is only conventionally con sidered as an indenter, since it is made of the required material (steel, ceramics, hard alloy, etc.) rather than a superhard material (diamond). In impact loading tests, hardalloy balls are employed. Such a counter body undergoes both elastic deformation under a load and wear, which results in a change in its geometry. However, Hertz’s model can be used to calculate the initial stresses (Hertz stresses) that appear in a sam ple–counterbody pair before testing [49]. This is important to understand how far the material of the chosen counterbody and the load meet the conditions of a real friction couple. Instrumented Indentation Hertz’s works served as a basis for the development of a theory and methods of hardness measurements. The hardness is often measured in science and engi neering in spite of the fact that the physical meaning of this quantity and correct hardness measurement meth ods are still being discussed [50, 51]. Hardness is assumed to mean the property of a surface layer to with stand elastic and plastic deformation (or fracture) dur ing local contact actions from another harder body (indenter), which has a certain shape and size [52]. As follows from the definition given in ISO 145771:2002, hardness is the response of a material to the indentation of a body of a given shape [53]. Instrumented indentation is now widely used to determine the hardness and elastic modulus of surface layers [54]. It should be noted that the foundations of this method were developed in the 1960s–1970s in the Soviet Union, and this method was known as the kinetic hardness method [55]. Nevertheless, this method had received wide acceptance after work [54] and was named after the authors of that work. Accord ing to ASTM E 254607 [56], instrumented indenta tion is taken to be indentation during which the force applied to an indenter and the indentation depth are
No. 10
926
LEVASHOV et al.
Pmax Ac
load
Loading
Unloading
S
hf
hc hf hi hc hmax indenter displacement
hmax
Fig. 10. General form of a loading–unloading curve and the scheme of indenter–surface contact.
recorded in the course of loading–unloading and used for further calculations of the hardness (during load ing) and the elastic modulus (during unloading). The Oliver–Pharr method consists in choosing the parameters of a power function that describes the experimental dependence of indentation depth h and the contact area on applied load P (Fig. 10). Elastic modulus E is calculated from the indentation projec tion area, contact stiffness S = dP/dh (which is deter mined from the slope of the upper third of an unload ing curve), the Poisson ratio, and the indenter param eters. The imprint projection area is determined from maximum indentation depth hm of a diamond indenter on the assumption that it is undeformed during a test. Thus, in instrumented indentation, the hardness is determined as in the Rockwell method, from the indentation depth, but without preliminary loading of an indenter. This is the main difference of this method from the Vickers and Brinell hardness tests, in which the imprint parameters (diagonal or diameter, respec tively) are measured. As shown in Fig. 11, a complex state of stress, which is close to uniform compression, forms near the contact zone during indentation, and the deformation propagating toward the bulk of the material has both elastic (reversible) and plastic (irreversible) compo nents. Therefore, instrumented indentation gives information on the hardness and Young’s modulus and (a)
can be used to estimate the fraction of the elastic com ponent in the total strain, which is characterized by the elastic recovery R = (hm – hf)/hm, where hm is the maximum imprint depth and hf is the indentation depth after unloading. It should be noted that the method of hardness measurements from the indentation area was devel oped in the early XXth century for metallic polycrys talline alloys, in which the elastic recovery is at most 10–20%. For materials with a higher fraction of elastic deformation, this method overestimates the hardness [50, 51]. A large elastic recovery manifests itself in a decrease in the indentation sizes after unloading. Therefore, it is impossible to correctly determine the hardnesses of materials with high values of R (which is typical of most nanomaterials) using the traditional methods of measuring the indentation diagonal. The traditional methods of hardness measurements cannot also be applied to study hard and superhard materials and coatings, since the indentation size in the case of small loads is so small that it is often invisible in an optical microscope and heavy loads cause cracking. Instrumented indentation makes it possible to deter mine the hardness, Young’s modulus, and the elastic recovery of both superhard and soft materials at small loads (several mN). The possibilities of instrumented indentation are illustrated in Fig. 12 and Table 2, which demonstrate how different the properties of bulk polycrystalline materials, single crystals, quasicrystalline materials, metallic glasses, and nanostructured films are. The elastic recovery of a nanocrystalline TiSiN film or a sil icon single crystal is at least three times higher and the elastic recovery of a ZrCuNiTi metallic glass or an AlCuFe single crystal is two times higher than that of polycrystalline copper or an OT41 titanium alloy. The ZrCuNiTi metallic glass is characterized by a low Young’s modulus and a high elastic recovery. These experimental data were obtained with a Nanohardness Tester (CSM Instruments, Switzer land), which combines a precise hardness tester and an optical microscope using the same microscopic stage with a programmable digital motorized drive. The motion in lateral (motion to positions for indentation’
(b)
elastic region
(c)
plastic region hard region
elastic region
Fig. 11. Region of contact of an incompressible spherical indenter with a flat surface of an (a) absolutely elastic, (b) absolutely plastic, and (c) elastoplastic material when a normal load is applied. RUSSIAN METALLURGY (METALLY)
Vol. 2010
No. 10
MULTIFUNCTIONAL NANOSTRUCTURED COATINGS 1
2
4 Load, mN
927
TiSiN
3
Si (100)
2 1
0
40 60 80 20 Indentation depth, nm
8
9
Cu
4 μm
100
Fig. 12. Experimental indentation curves for copper, silicon single crystal, and magnetron TiSiN coating on silicon.
Fig. 13. Shape of imprints on an aluminum matrix disper sionstrengthened by AlCuFe quasicrystal line particles.
observation with an optical or scanning probe micro scope) and vertical (measurement) planes is con trolled with a computer using precise transducers and software at a positioning accuracy much better than one micron. Nanoindentation is a particular case of instru mented indentation. It is performed at very small loads (several mN). At such loads, the indentation depth is several tens of nanometers. Nanoindentation is indis pensable for studying thin nanostructured films and multilayer nanostructures. Hardness measurements in which the indentation depth is no more than 10–12% of the coating thickness are considered to be correct (without the effect of a substrate) [57]. Stable experi mental curves are obtained at an indentation depth
more than 25–30 nm. Therefore, nanoindentation is successfully used to estimate the mechanical proper ties of thin films with nanometer and submicron thick nesses. A high positioning precision also makes it possible to perform selective instrumented indentation, namely, indentation of micron, submicron, and nanometer structural constituents that can be distinguished in an optical or scanning probe microscope and, thus, to esti mate their mechanical properties when studying mul tiphase, composite, and gradient materials. Figure 13 shows the difference in the identifications in an alumi num matrix (indentation 2, H = 1.6 GPa, E = 74 GPa, R = 11.9%) and in the region of accumulation of hard
Table 2. Properties of materials and coatings calculated from sensing indentation Material
H, GPa
E, GPa
R, %
Copper
2.1
121
14
Titanium (OT41)
4.1
130
19
Multilayer Ti/aC:H* film
8.0
128
34
Amorphous ZrCuTiNi ribbon
11.5
117
42
Singlecrystal (100) silicon
11.8
174
62
Thin PVD TiSiN film
28.4
295
62
0.72
56
7
dispersionstrengthened by AlCuFe quasicrystal line microparticles (70%)
1.9
106
14
dispersionstrengthened by AlCuFe quasicrystal line nanoparticles (30%)
2.5
98
17
AlCuFe quasicrystal microparticles in dispersionstrengthened sintered ASD4 aluminum
10.1
175
32
AlCuFe quasicrystal line in a thin PVD film
13.7
286
32
ASD4 aluminum, sintered ASD4 aluminum
* TiaC:H means titanium/hydrogenised amorphous carbon. RUSSIAN METALLURGY (METALLY)
Vol. 2010
No. 10
480
R, %
(a) 80 R
360
Hardness, GPa
60 240
40 E 20
120
H 400 600 200 Indentation depth, nm
0
480
(b) 80
R 360
60
E 240
40
120
20
800
Young’s modulus, GPa
LEVASHOV et al. Young’s modulus, GPa Hardness, GPa R, %
928
H 400 600 200 Indentation depth, nm
0
800
Hardness, GPa
200
40 20 0
E H 400 600 800 200 Indentation depth, nm
100
0 1000
80
400
(b)
300 60 200
40 E
H 20 0
400 600 800 200 Indentation depth, nm
100
Young’s modulus, GPa
300 60
R R, %
400
(a)
Hardness, GPa
R
80
Young’s modulus, GPa
R, %
Fig. 14. Effect of the indentation depth on hardness H, elastic modulus E, and elastic recovery R of a TiCaCPON coating (h = 1.8 µm) on oxidized substrates of (a) fused silica and (b) sapphire.
0 1000
Fig. 15. Effect of the indentation depth on hardness H, elastic modulus E, and elastic recovery R of a TiCaCPON coating (h = 1.8 µm) on metallic substrates of (a) microcrystalline and (b) nanostructured Grade 4 titanium.
ening quasicrystalline particles (indentation 8, H = 4.8 GPa, E = 92 GPa, R = 27.4%). To estimate the mechanical properties of mul tiphase materials, researchers use matrix indentation at a step chosen so that different structural constituents are indented. At a sufficiently large number of inde pendent measurements and a significant difference in the properties of individual structural constituents, one can obtain average values of hardness, elastic modulus, and elastic recovery; quantitatively estimate their fractions; and map a distribution of the mechan ical properties of multiphase and composite materials. Instrumented indentation when the load increases (depthsensing indentation). One of the fundamental problems related to determining the mechanical prop erties of the coating–substrate system is searching for an optimum indentation depth range at which the effect of the interface on the measured properties is minimal. Foreign standards only give general recom mendations for a correct estimation of the mechanical properties of a coating, which consist in a limitation of the indentation depth to 10% of the coating thickness [53]. However, this recommendation cannot be
equally valid for such different systems as hard coat ing–soft substrate and soft coating–hard substrate. Therefore, to describe correctly the mechanical prop erties of coatings on different substrates, one has to perform measurements when the load increases, i.e., when the indentation depth changes over a wide range. Such studies are especially important for develop ment of reference samples of the nanostructured sur faces of nanoindustry products. We chose substrates made of fused silica, fianit, sapphire, singlecrystal sili con, and Grade 4 titanium in microstructured and nanostructured states (with an average grain size of 15– 20 μm and less than 200 nm, respectively). (Nstita nium was produced by equalchannel angular pressing in UGATU [58]). TiCaCPON films were deposited onto all substrates in one technological cycle of magne tron sputtering, and their mechanical and tribological properties are shown in Figs. 14 and 15. All experimen tal curves have three specific ranges characterized by different changes in the properties of the coating: in the first range, they are controlled by the surface state; in the second range, they are relatively constant; and in the third range, they are controlled by the state and proper
RUSSIAN METALLURGY (METALLY)
Vol. 2010
No. 10
MULTIFUNCTIONAL NANOSTRUCTURED COATINGS
ties of the substrate and the stresses at the coating–sub strate interface. A comparative analysis of the nanoindentation data allowed us to reveal that a certain optimum indentation depth (for correct measurements) exists for the same coating deposited onto different substrates. For a TiCaCPON film (h = 1.8 μm), it is 60–300 nm, i.e., 3.3–16.7% of its thickness (Table 3). In this case, the hardness, elastic modulus, and elastic recovery of the film have similar values, namely, 25 ± 2 GPa, 210 ± 20 GPa, and 65 ± 5%, respectively, and are independent of the substrate material. (It should be noted that the hardness and elastic modulus of the coating on the sub strate made of nanostructured titanium are 10–30% higher than those on the initial titanium.) These data support the necessity of careful mea surements for every substrate–coating system and determination of the depth profiles of the mechanical properties of films, which are necessary to find reason able applications of nanostructured coatings. The curves in Figs. 14 and 15 demonstrate that the effect of relatively hard oxide substrates begins to man ifest itself at a depth of 100–150 nm (6–8% of the film thickness) and that the effect of more plastic metallic substrates is noticeable only at a depth of 215–300 nm (12–17%). It is interesting that the film on annealed nanostructured titanium has characteristic maxima of the hardness, elastic modulus, and elastic recovery at a depth of 200 nm (see Fig. 15b). These maxima are likely to be caused by high residual stresses in the film or a physicochemical interaction during its ion plasma deposition on a severe deformed substrate. Instrumented Scratching Specialpurpose equipment like scratch testers (adhesion meters) is used to determine the adhe sion/cohesion strength, scratching resistance, and the mechanisms of fracture of nanostructured coatings and thin films. A REVETEST (CSM Instrumets) device is among them (Fig. 16). In this device, a sur face is scratched with a diamond conical Rockwell type indenter at a constant or increasing (stepwise or continuously) load. The testing according to ASTM C 162405 consists in sliding of a diamond indenter with a vertex angle of 120° and a radius of curvature of 20– 200 μm on a coating surface, with the load applied to the indenter increasing continuously. The readings of transducers (loading force, acoustic emission inten sity, friction force, friction coefficient, indentation depth) are stored on a hard disk in a controlling com puter. The minimum (critical) load Lc resulting in fracture is determined from the shape of property– load curves and coating fracture traces (optical microscopy). Figures 17 and 18 show the results of testing 1.8μmthick hard nanostructured TiCrBN coatings with a crystallite size less than 100 nm. The coating RUSSIAN METALLURGY (METALLY)
Vol. 2010
929
Table 3. Optimum indentation depth hopt to obtain stable values of the mechanical properties if a TiCaCPON coating (h = 1.8 µm) Substrate
hopt, nm (%)
Fused silica
100–150 (5.5–8.3)
Sapphire
60–100 (3.3–5.6)
Grade 4 titanium microstructured
150–215 (8.3–11.9)
nanostructured
90–300 (5–16.7)
Note: The ratios of the indentation depth to the coating thickness are given in parentheses.
was deposited by magnetron sputtering of composite SSTTD 31 cathode targets on a sapphire substrate. The experimental conditions were as follows: the load increased from 0.9 to 30 N at a rate of 1 mm/min, and the scratch length was 5 mm. Three scratches were formed with a Rockwell C diamond indenter having a radius of curvature of 200 μm, and the averaged criti cal loads were determined. The fracture of the coating during scratching by a diamond indenter can be conventionally divided into three stages (Fig. 17). At stage A, the indenter pene trates monotonically into the coating: the indentation depth (ID) decreases weakly, the friction force (FF) and the friction coefficient (FC) increase weakly, and acoustic emission (AE) signal remains unchanged. At stage B, the AE level and amplitude increase and the slope of the ID curve changes. Stage C begins with a jumplike change in the ID, FC, and AE and the FF increases monotonically. A comparative analysis of the structural features of the scratches suggests that, at stage A (the load is smaller than 9 N), the indenter leaves no traces on the
No. 10
PN Rod with indenter
Indentation depth sensor
Ff
Acoustic emission transducer
Sample displacement (dx/dt) Fig. 16. Schematic diagram for tests with a REVETEST scratch tester.
930
LEVASHOV et al. A
B
C
60 1.5
100 1
ID
48 1.2
80
5
36 0.9
60
11
40
17
20
23
24 0.6
AE
12 0.3 FC 0N
0 0.90N 0 mm
FF
6.72
12.54
18.36
24.18
0% 30.00
1
2
3
4
5
29 μg
Fig. 17. Results of testing TiCrBN coatings with a REVETEST scratch tester at a load increasing from 0.9 to 30 N. ID is the inden tation depth; AE, the acoustic emission; FC, the friction coefficient; and FF, the friction force.
(а)
10 μm (b)
10 μm
(c)
10 μm (d)
10 μm
Fig. 18. Micrographs of scratch regions at an indenter load of (a) 6.7, (b) 9.7 N, (c) 12.4, and (d) 28.4 N (×800).
coating (Fig. 18a). The diamond indenter slides on the hard nanostructured coating at a very low friction coefficient (less than 0.07). A load more than 9 N (stage B) leads to chevron cracks on the crack bottom (see Fig. 18b). The formation of these cracks is accom panied by an increase in the AE level and amplitude and a monotonic increase in the FC to 0.1. The third stage of fracture (stage C) takes place at a load more than 12.5 N and is related to local and, then, continu ous spalling of the substrate under the indenter pres sure (Figs. 18c, 18d). The appearance of spalled regions in the coating is accompanied by sharp out bursts in the AE, FC, and ID curves (see Fig. 17). An analysis of these data suggests that the adhesion strength of the TiCrBN coating is about 9 N, which is
clearly visible in Fig. 18b. At a load more than 12.5 N applied to the indenter, the substrate undergoes local fracture, and a load more than 13.5 N causes continu ous fracture of the substrate over the entire contact area and spalling of the coating near the scratch. Tribological Tests The tribological tests of functional coatings according to the pinondisk scheme are performed using a specialpurpose automated friction machine in both air and a liquid medium (Fig. 19). These tests also allow to use Hertz’s model, they correspond to inter national standards (ASTM G99959, DIN50324), and are optimal for estimating the wear resistance of a
RUSSIAN METALLURGY (METALLY)
Vol. 2010
No. 10
MULTIFUNCTIONAL NANOSTRUCTURED COATINGS (a)
F
0.8 Friction coefficient
R
931
ω
0.7
TiCaCPON coating
Al2O3 counterbody
0.6 0.5 0.4
Nstitanium
Sapphire
0.3 0.2 0.1
Mstitanium
Fused silica
1000 2000 3000 4000 5000 6000 7000 8000 9000 Path length, rev. (b) 1.0 WC–Co counterbody TiCrBN coating 0.9 0.8 Nstitanium Fused silica 0.7 0.6 0.5 Sapphire 0.4 Mstitanium 0.3 0.2 0.1 0
sample and the counterbody [59]. The friction coeffi cient of a friction couple is determined during tests. A ball counterbody is made of a certified material (steels of various grades, Al2O3, Si3N4, WC–Co, etc.). During a test, a ball is fixed in a holder, which is made of a cor rosionresistant steel, transfers a load to the ball, and is connected to a friction force transducer. An analysis of the wear products and the structures of the wear groove on a sample and the wear spot on the counterbody gives important information on the coating fracture mechanism. For this purpose, micro scopic observation is used. For example, the structure of grooves and the spot diameter are observed in an optical microscope, and a profilometer is used to determine the average crosssectional area and groove depth. With these data, researchers calculate the vol ume losses of the sample and counterbody. Reduced wear I (which is inversely proportional to the wear resistance) is calculated using volume loss ΔV normal ized by path length N (m) and applied load P (N), I = ΔV/(NP). The reduced wear of the coating and coun terbody is independent of the path length and load and can be used to compare the results of tribological tests performed under different conditions. In the framework of the works dealing with creating reference samples of nanostructured surfaces, we measured the tribological properties (friction coeffi cient, reduced wear) of 1.8μmthick TiCaCPON and TiCrBN coatings deposited onto substrates made of fused silica, fianit, sapphire, singlecrystal silicon, and Grade 4 titanium in both structural states. Each com position was deposited in one technological cycle. The results obtained demonstrate that these coatings have a stable friction coefficient, which weakly depends on the substrate material (Fig. 20). The average friction coefficient of the TiCaCPON coatings is 0.25, which is sufficient for them to be recommended as antifric tion coatings. The average friction coefficient of the TiCrBN coatings reaches 0.7, which implies that they can be recommended as friction coatings. The coat RUSSIAN METALLURGY (METALLY)
Vol. 2010
Friction coefficient
Fig. 19. Schematic diagram for tests with a TRIBOMETER setup: F is the normal load and R is the wear ring radius.
0
1000 2000 3000 4000 5000 6000 7000 8000 9000 Distance, rev.
Fig. 20. Friction coefficient vs. the distance for (a) TiCaCPON and (b) TiCrBN coatings on fused silica, sap phire, and Grade 4 titanium substrates in microstructured (mstitanium) and nanostructured (nstitanium) states.
ings deposited onto hard sapphire substrates have the minimum wear. The reduced wear of the hard nanostructured TiCrBN coatings deposited onto a nanostructured tita nium substrate is almost an order of magnitude lower than that of the coatings on a traditional Grade 4 tita nium substrate (8 × 10–6 and 1 × 10–5 mm3 N–1 m–1, respectively). Similar features are also observed for the nanostructured TiCaCPON coatings (Fig. 21). This supports the new trend in creating composite materials, e.g., medical implants made of nanostructured titanium coated with a biocompatible and bioactive nanostruc tured TiCaCPON layer. Cyclic Impact Tests Fatigue strength tests of coatings are performed during cyclic impact loading according to the scheme shown in Fig. 22 [60]. This scheme is used in an ImpactTester (CemeCon, Germany) device. The tests are carried out under conditions of multicycle impacts of a hardalloy ball on a coating to be tested
No. 10
932
LEVASHOV et al. (a)
(b) Reduced wear I, mm × H–1 m–1 1.0 × 10–5 TiCaCPON coating 3
Reduced wear I, mm × H m 1.4 × 10–5 TiCrBN coating 3
–1
–1
Mstitanium
1.2 × 10–5 1.0 × 10–5 6 × 10–6 4 × 10–6
Nstitanium
Fused silica
2 × 10–6
Mstitanium
6 × 10–6
Nstitanium
8 × 10–6
8 × 10–6
4 × 10–6 2 × 10–6
Sapphire
Fused silica Sapphire
Sample
Sample
Fig. 21. Reduced wear of (a) TiCrBN and (b) TiCaCPON coatings on fused silica, sapphire, and Grade 4 titanium substrates in microstructured (mstitanium) and nanostructured (nstitanium) states.
(а)
(b)
Fimp
Ball–indenter coating 2α
tcoat 150 μm
substrate
Fig. 22. (a) Schematic diagram for a cyclic impact test and (b) the imprint formed upon the test.
(а)
50 μm (b)
50 μm
Fig. 23. Microstructure of an imprint in a TiCaCPON coating after (a) 105 and (b) 106 impact loading cycles at a force of 800 N.
until its fatigue fracture. The purpose of these tests is to determine the critical load and the number of impacts to the onset of cohesion/adhesion fracture of the coat ing. To this end, the imprint that forms upon impact
loading is subjected to fractographic analysis after every test cycle. Figure 23 shows the results of testing TiCaCPON coatings deposited onto a Grade 4 titanium substrate.
RUSSIAN METALLURGY (METALLY)
Vol. 2010
No. 10
MULTIFUNCTIONAL NANOSTRUCTURED COATINGS Fimp 1600 1200
without fracture fracture
800 400 0
0.5 × 106 1.0 × 106 1.5 × 106 N
Fig. 24. Fatigue fracture diagram for a TiCaCPON coating on a Grade 4 titanium substrate plotted with an Impact Tester device.
It is seen that numerous localized cracks form on the indentation bottom without fracture of the coating after 105 impacts at a load of 800 N (Fig. 23a). The coating is seen to fail after 106 cycles at the same load (Fig. 23b). The results of fractographic analysis are used to plot a fatigue fracture diagram, which makes it possible to predict the material life. As an example, Fig. 24 shows such a diagram for a TiCaCPON coating on a Grade 4 titanium substrate. CONCLUSIONS Further progress in the field of creating nanostruc tured functional surfaces and coatings is related to searching for new compositions, implementing coating deposition technologies, and the development of tech niques for measuring the properties of nanomaterials. The modern methods of studying functional sur faces under conditions of mechanical contact during indentation, scratching, sliding, and an impact of a counterbody provide data on structuresensitive prop erties that could not be obtained earlier, since extremely small volumes (nanovolumes) of a material are used for an investigation. These properties are important for designing novel functional surfaces and coatings, since they allow one to predict their possible fracture mechanisms controlled by the structural state of the surface layers. A metrologically pure analysis and the commer cialization of nanoindustry products require a metro logical set that includes both apparatus and standard methodological support of the uniformity of measure ments, calibration and reference samples, certified consumable materials, and a qualified staff. REFERENCES 1. A. Leyland and A. Matthews, “On the Significance of the H/E Ratio in Wear Control: A Nanocomposite Coating Approach to Optimized Tribological Behav iour,” Wear 246 (1/2), 1–11 (2000). RUSSIAN METALLURGY (METALLY)
Vol. 2010
933
2. D. V. Shtansky, S. A. Kulinich, E. A. Levashov, and J. J. Moore, “Structure and PhysicalMechanical Properties of Nanostructured Thin Films,” Fiz. Tverd. Tela 45 (6), 1122–1129 (2003) [Phys. Sol. State 45 (6), 1055–1062 (2003)]. 3. D. V. Shtansky, E. A. Levashov, and A. N. Sheveiko, “Miteinem SHSLegierangsTarget Abgeschiedene Mehrkomponentenschichten Ti–B–N, Ti–Si–B–N, Ti–Si–C–N und Ti–Al–C–N for Unterschiedliche Technologische Anwendungen,” Galvanotechnik, No. 10, 3368–3378 (1997). 4. D. V. Shtansky, I. V. Lyasotsky, E. A. Levashov, et al., “Comparative Investigation of Ti–Si–N Films Magne tron Sputtered Using Ti5Si3 + Ti and Ti5Si3 + TiN Tar gets,” Surf. Coat. Technol. 182 (2/3), 210–220 (2004). 5. D. V. Shtansky, Ph. V. KiryukhantsevKorneev, E. A. Levashov, et al., “Hard Tribological Ti–Cr–B–N Coatings with Enhanced Thermal Stability, Corrosion and Oxidation Resistance,” Surf. Coat. Technol. 202 (4–7), 861–865 (2007). 6. K. H. Kim, S. Choi, and S. Yoon, “Superhard Ti–Si–N Coatings by Hybrid System of Arc Ion Plating and Sputtering Techniques,” Surf. Coat. Technol. 298 (2/3), 243–248 (2002). 7. S. J. Veprek, “The Search for Novel, Superhard Mate rials,” Vac. Sci. Technol. 5, 2401–2418 (1999). 8. S. J. Veprek, M. G. J. VeprekHeijman, P. Karvankova, and J. Prochazka, “Different Approaches to Superhard Coatings and Nanocomposites,” Thin Solid Films 476 (1), 1–29 (2005). 9. F. V. KiryukhantsevKorneev, D. V. Shtansky, E. A. Levashov, et al., “Structure and Properties of Ti– Si–N Coatings Produced by Magnetron Sputtering of SHS Targets,” Fiz. Met. Metalloved. 97 (3), 96–103 (2004) [Phys. Met. Metallogr. 97 (3), 376–383 (2004)]. 10. D. V. Shtansky, F. V. KiryukhantsevKorneev, E. A. Levashov, et al., “Structure and Properties of Ti– B–N, Ti–Cr–B–(N), and Cr–B–(N) Coatings Deposited by Magnetron Sputtering of Targets Pre pared by SelfPropagating HighTemperature Synthe sis,” Fiz. Tverd. Tela 47 (2), 242–251 (2005) [Phys. Sol. State 47 (2), 230–2339 (2005)]. 11. R. A. Andrievskii, “Nanomaterials: Concept and Modern Problems,” Ros. Khim. Zh. 46 (5), 50–56 (2002). 12. D. V. Shtansky, E. A. Levashov, A. N. Sheveiko, and J. J. Moore, “The Structure and Properties of Ti–B–N, Ti–Si–B–N, Ti–Si–C–N and Ti–Al–C–N Coatings Deposited by Magnetron Sputtering Using Composite Targets Produced by SelfPropagating HighTempera ture Synthesis (SHS),” J. Mater. Synth. Process 6 (1), 61–72 (1998). 13. C. Rebholz, H. Ziegete, A. Leyland, et al., “Structure, Mechanical and Tribological Properties of Ti–B–N and Ti–Al–B–N Multiphase Thin Films Produced by ElectronBeam Evaporation,” Vac. Sci. Technol. A 16 (5), 2851–2857. 14. T. P. Mollart, J. Haupt, R. Gilmore, and W. Gisster, “Tribological Behaviour of Homogeneous Ti–B–N, Ti– B–N–C and TiN/h–BN/TiB2 Multilayer Coat ings,” Surf. Coat. Technol. 86/87 (1), 231–236 (1996). 15. D. V. Shtansky, K. Kaneko, Y. Ikuhara, and E. A. Levashov, “Characterization of Nanostructured
No. 10
934
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
LEVASHOV et al. Multiphase Ti–Al– B–N Thin Films with Extremely Small Grain Size,” Surf. Coat. Technol. 148 (2/3), 206–215 (2001). D. V. Shtansky, E. A. Levashov, A. N. Sheveiko, and J. J. Moore, “Optimization of PVD Parameters for the Deposition of Ultra Hard Ti–Si–B–N Coatings,” J. Mater. Synth. Process. 7 (3), 187–193 (1999). D. V. Shtansky, A. N. Sheveiko, E. A. Levashov, et al., “Hard Tribological Ti–B–N, Ti–Cr–B–N, Ti–Si– B–N and Ti–Al–Si–B–N Coatings,” Surf. Coat. Technol. 200 (1–4), 208–212 (2005). J. Musil and M. Jirout, “Toughness of Hard Nano structured Ceramic Thin Films,” Surf. Coat Technol. 201 (1–4), 5148–5152 (2007). Ph. V. KiryukhantsevKorneev, J. F. Pierson, M. I. Petrzhik, et al., “Effect of Nitrogen Partial Pres sure on the Structure, Physical and Mechanical Prop erties of CrB2, and Cr–B–N Films,” Thin Solid Films 517 (8), 2675–2680 (2009). Ph. V. KiryukhantsevKorneev, J. F. Pierson, J. P. Baner, et al., “Structure and Properties of Hard Nanostructured Coatings in Cr–B–N System,” in Pro ceedings of III FranceRussia Seminar on New Achieve ments in Materials and Environmental Sciences, Metz, France, Ed. by A. Postnikov (EDP Sciences, Paris 2008), pp. 11–14. A. A. Voevodin and J. S. Zabinski, “Nanocomposite and Nanostructured Tribological Materials for Space Applications,” Compos. Sci. and Technol. 65 (5), 741– 748 (2005). T. Kubart, T. Polcar, L. Kopecky, et al., “Temperature Dependence of Tribological Properties of MoS2, and MoSe2 Coatings,” Surf. Coat Technol. 193 (1–3), 230–233 (2005). D. V. Shtansky, T. A. Lobova, E. A. Levashov, et al., “Structure and Tribological Properties of WSex, WSex/TiN, WSex/TiCN and WSex/TiSiN Coatings,” Surf. Coat Technol. 183 (2/3), 328–336 (2004). D. V. Shtansky, A. N. Sheveyko, D. I. Sorokin, et al., “Structure and Properties of Nanocomposite and Mul tilayer TiCrBN/WSex Coatings Deposited by Ion Implantation Assisted Sputtering of TiCrB and WSe2, Targets,” Surf. Coat Technol. 202 (24), 5953–5961 (2008). F. V. KiryukhantsevKorneev, M. I. Petrzhik, E. A. Levashov, et al., “Effect of Al, Si, and Cr on the Thermal Stability and Resistance to HighTemperature Oxidation of Coatings Based on Titanium Boroni tride,” Fiz. Met. Metalloved. 104 (2), 176–183 (2007) [Phys. Met. Metallogr. 104 (8), 1220–1227 (2007)]. E. A. Levashov and D. V. Shtansky, “Multifunctional Nanostructured Films,” Usp. Khim. 76 (5), 501–509 (2007). J. B. Choi, C. Kurn, Y. Kim, et al., “Microstructure Effect on the HighTemperature Oxidation Resistance of Ti–Si– N Coating Layers,” Jpn. J. Appl. Phys. 2, 6556–6559 (2003). C. Paternoster, A. Fabrizi, R. Cecchini, et al., “Ther mal Evolution and Mechanical Properties of Hard Ti– Cr– B–N and Ti–Al–Si–B–N Coatings,” Surf. Coat. Technol. 203, 736–740 (2008).
29. D. V. Shtansky, E. A. Levashov, N. A. Gloushankova, et al., “Structure and Properties of ZrO2, and CaO doped TiCxNy Coatings for Biomedical Applications,” Surf. Coat. Technol. 182 (1), 101–111 (2004). 30. D. V. Shtansky, N. A. Gloushankova, I. A. Bashkova, et al., “Multifunctional Biocompatible Nanostructured Coatings for LoadBearing Implants,” Surf. Coat Technol. 201 (7), 4111–4118 (2006). 31. D. V. Shtansky, N. A. Gloushankova, I. A. Bashkova, et al., “Multifunctional Ti(Ca, Zr)–(C, N, O, P) Films for LoadBearing Implants,” Biomaterials 27, 3519–3531 (2006). 32. D. V. Shtansky, N. A. Gloushankova, A. N. Sheveiko, et al., “Design, Characterization and Testing of TiBased Multicomponent Coatings for LoadBearing Medical Application,” Biomaterials 26, 2909–2924 (2005). 33. D. V. Shtansky, I. A. Bashkova, F. V. Kiryukhantsev Korneev, et al., “Bioactive Ceramic TantalumCon taining Films for Implants,” Dokl. Akad. Nauk 418 (1), 121–124 (2008). 34. D. V. Shtansky, N. A. Gloushankova, I. A. Bashkova, et al., “TaDoped Multifunctional Bioactive Nano structured Films,” Surf. Coat. Technol. 202 (15), 3615–3624 (2008). 35. E. A. Levashov, A. S. Rogachev, I. P. Yukhvid, et al., Physicochemical and Technological Foundations of Self Propagating HighTemperature Synthesis (BINOM, Moscow, 1999) [in Russian]. 36. M. Andronis, A. Leyland, E. Levashov, et al., “The Structure and Mechanical Properties of Ti–Si–B Coatings Deposited by DC and PilsedDC Unbalanced Magnetron Sputtering,” Plasma Process. Polym. 4, 687–692 (2007). 37. D. Zhong, E. Sutter, E. A. Levashov, et al., “Mechani cal Properties of Ti–B–C–N Coatings Deposited by Magnetron Sputtering,” Thin Solid Films. 398– 399, 320–325 (2001). 38. W. Kulisch, P. Colpo, P. N. Gibson, et al., “ICP Assisted Sputter Deposition of TiC/CaO Nanocom posite Films,” Surf. Coat. Technol. 188/189 (11/12), 735–740 (2004). 39. Z. Werner, J. Stanislawski, E. Levashov, et al., “New Types of MultiComponent Hard Coatings Deposited by ARC PVD on Steel Pretreated by Pulsed Plasma Beams,” Vacuum 70 (2/3), 263–267 (2003). 40. D. V. Shtansky, E. A. Levashov, N. N. Khavskii, and J. J. Moore, “Prospects of Creating Composite Wear Resistant Films Using SHS Cathodes,” Izv. Vyssh. Uchebn. Zaved., Tsvetn. Metall., No. 1, 59–68 (1996). 41. P. J. Kelly and R. D. Arnell, “Magnetron Sputtering: A Review of Recent Developments and Applications,” Vacuum 56 (3), 159–172 (2000). 42. D. G. Teer, “Technical Note: A Magnetron Sputter IonPlating System,” Surf. Coat, Technol. 39/40 (2), 565–572 (1989). 43. D. V. Shtansky, “HighResolution Transmission Elec tron Microscopy in Nanotechnology Investigations,” Ros. Khim. Zh. 46 (5), 81–89 (2002). 44. D. V. Shtansky and E. A. Levashov, “Multicomponent Nanostructured Thin Films: problems and Solutions
RUSSIAN METALLURGY (METALLY)
Vol. 2010
No. 10
MULTIFUNCTIONAL NANOSTRUCTURED COATINGS
45. 46.
47. 48. 49. 50. 51.
52.
(Review),” Izv. Vyssh. Uchebn. Zaved., Tsvetn. Metall., No. 3, 52–62 (2001). D. V. Shtansky, S. A. Kulinich, E. A. Levashov, et al., “Localized Deformation of Multicomponent Thin Films,” Thin Solid Films 420/421, 330–337 (2002). Ph. V. KirynkhantsevKorneev, D. V. Shtansky, E. A. Levashov, et al., “Thermal Stability and Oxida tion Resistance of Ti–B–N, Ti–Cr–B–N, Ti–Si–B– N and Ti–Al–Si–B–N Films,” Surf. Coat. Technol. 201 (13), 6143–6147 (2007). H. R. Hertz, “Uber die Bemhrung Fester Elastischer Korper,” Journal fur die Reine und Angewandte Math ematik, No. 92, 156 (1882). L. Landau and M. Lifshitz, Theory of Elasticity (Nauka, Moscow, 1987) [in Russian]. G. M. Gamilton, “Explicit Equations for the Stresses beneath a Sliding Spherical Contact,” Proc. Inst. Mech. Engrs. C 197, 53 (1983). J. Musil, H. Zeman, F. Kunc and J. Vlcek, “Measure ment of Hardness of Superhard Films by Microinden tation,” J. Mater. Sci. Eng. A 340 (1/2), 281 (2002). S. Veprek, S. Mukherjee, H. D. Mannling and J. L. He, “Hertzian Analysis of the SelfConsistency and Reli ability of the Indentation Hardness Measurements on Superhard Nanocomposite Coatings,” Thin Solid Films A 436 (2), 292 (2003). V. S. Zolotarevskii, Mechanical Properties of Metals (MISiS, Moscow, 1998) [in Russian].
RUSSIAN METALLURGY (METALLY)
Vol. 2010
935
53. ISO 1457714:2002. Metallic materials. Instrumented Indentation Test for Hardness and Materials Parameters. 54. W. C. Oliver and G. M. Pharr, “An Improved Technique for Determining Hardness and Elastic Modulus Using Load and Displacement Sensing Indentation Experi ments,” J. Mater. Res., No. 7, 1564 (1992). 55. S. I. Bulychev and V. P. Alekhin, Testing of Materials by Continuous Penetration of Indenter Materials (Mashi nostroenie, Moscow, 1990) [in Russian]. 56. ASTM E 2546–07. Standard Practice for Instrumented Indentation Testing. 57. G. M. Pharr, “Measurement of Mechanical Properties by Ultralow Load Indentation,” Mater. Sci. Eng. A 253 (1/2), 151 (1998). 58. R. Z. Valiev, I. P. Semenova, V. V. Latysh, et al., “Nano structured Titanium for Biomedical Applications: New Design and Commercialization Prospects,” Rossiiskie nanotekhnologii 3 (9/10), 80–89 (2008). 59. M. I. Petrzhik, M. R. Filonov, K. A. Pecherkin, et al., “Wear Resistance and Mechanical Properties of Medi cal Alloys,” Izv. Vyssh. Uchebn. Zaved., Tsvetn. Metall., No. 6, 62 (2005). 60. K.D. Bouzakis, N. Vidakis, T. Leyendecker, et al., “Determination of the Fatigue Behaviour of Thin Hard Coatings Using the Impact Test and a FEM Simula tion,” Surf. Coat. Technol. 86/87 (2), 549–556 (1996).
No. 10
