View PDF VersionPrevious ArticleNext Article
 
DOI: 10.1039/C6RA02370C (Paper) RSC Adv., 2016, 6, 53043-53053

PVD multilayer VN–VN/Ag composite coating with adaptive lubricious behavior from 25 to 700 °C

Yongtao Mua, Ming Liu*a, Yongxin Wangb and Eryong Liub
aSchool of Mechatronics Engineering, Harbin Institute of Technology, Harbin 150001, China. E-mail: liuming@hit.edu.cn; Tel: +86 451 86403799
bKey Laboratory of Marine Materials and Related Technologies, Zhejiang Key Laboratory of Marine Materials and Protective Technologies, Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo 315201, China

Received 26th January 2016 , Accepted 19th May 2016

First published on 19th May 2016

Abstract

The doping of Ag nanoparticles in a VN matrix was developed on Inconel 718 and Si (100) substrates using a multi-arc ion plating technique. The microtopographies and microstructures of the as-deposited multilayer VN–VN/Ag and VN coatings were studied using a scanning electron microscope, an X-ray diffractometer, and a transmission electron microscope. The mechanical properties were investigated using a scratch test and a nanoindentation test. The results show that the incorporation of Ag-dopant into a VN coating can promote the decrease of average grain size, and the formation of a diffuse structure in the VN/Ag layer. The lamellar structure of the VN–VN/Ag coating can achieve an enhanced adhesion strength in comparison to the monolayer VN coating, but it has a lower hardness and elastic modulus because of the formation of a high content of soft metallic Ag. Furthermore, tribological testing was performed on a UMT-3 ball-on-disk tribometer, with Al2O3 balls as the counterpart, at room temperature (RT), 300, 500, or 700 °C in ambient air. The VN–VN/Ag coating had an excellent self-adaptive lubricating behavior with a decrease of friction coefficient as the temperature increased, but its wear rate showed a drastic increase especially at the high temperature region because of oxidization and a tribochemical reaction. The phase composition of the worn area was investigated using a micro-Raman spectrometer, which indicated a smooth Ag, Magnéli phase (V2O5) and bimetallic oxides (Ag3VO4 and AgVO3) were responsible for the decrease of the coefficient of friction in (i) room to mid-range temperatures and (ii) mid-range to high temperatures.

1. Introduction

Reliable lubrication and wear resistance at elevated temperatures have gained more and more attention because of their role in energy efficiency in turbomachinery and in the aerospace industry.1–11 However, common oil or grease lubrication is known to volatilize or undergo chemical degradation upon overheating.12,13 Conventional solid lubricants such as graphite and molybdenum disulfide have also been subjected to oxidation, and they too lose efficacy rapidly at T > 350 °C.14 Seeking effective lubricating materials for the referred applications is a challenging task for the tribology community. Two-phase composite coatings consisting of inclusions of soft metals (e.g., Ag, Cu, Au, or Pb) in the nitride matrix of transition metals have recently become a subject of investigation as solid lubricant materials for high temperature environments because the production of lubricious phases requires oxidation.15–18

The specific lubrication mechanism of a transition/noble metal combination at elevated temperatures has been studied for many years, and the Magnéli phase and graphitization bimetallic oxides have been found to play an important role in the decrease of the friction coefficient. For example, the lubricious Ag3VO4 and AgVO3 phases were generated to attain a friction coefficient of 0.1 via tribotesting a VN/Ag adaptive coating at a high temperature of 1000 °C.14 Lubricative NbO6 and silver niobate were synthetized to provide a decrease of the friction coefficient from 0.30 to 0.15, as the nanocomposite coating consisting of Nb nitride with Ag nanoinclusions was tribotested using a Si3N4 counterface in the range of 700 to 1000 °C.19 Furthermore, the Mo2N/Ag coating showed an obvious effect in the reduction of friction with temperatures above 400 °C because of the creation of lubricous silver molybdate and MoO3.17 The use of the Mo2N/Cu coating gave a reduced coefficient of friction of 0.2 from 0.5 because of its conversion to CuMoO4 on the rubbed surfaces after exposure to temperatures exceeding 530 °C, and sustained the lowest value (μ = 0.2) for 3000 cycles at 650 °C.20

Research shows that the combination of hard ceramic nitrides with soft metallic particles is a viable alternative for self-lubricating coating materials with the presence of oxidation in high-temperature environments. However, it is challenging to control the lubricant transport rate, which may cause a faster depletion of the lubricating layer than required. A feasible scheme for controlling lubricant out-diffusion is the application of a barrier layer or multilayer structure, which limits the metal transport to areas within the wear track.21–24 In view of this, a multilayer VN–VN/Ag structure with a nanoscale bilayer thickness (Λ) was synthesized in the research described in this paper in order to provide an approach to control lubricant transport to the surface prior to wear. The hypothesis is that an alternating VN layer gives the coating a favorable bearing capacity and serves as the diffusion barrier to suppress lubricant transport, whereas the VN/Ag layer gives a good lubrication performance under high-temperature oxidation. Not only that, the multilayer coating can confine the initiation and propagation of the micro-cracks within the origination layer during the loading process, and thus, can effectively improve its adhesion strength and tribological properties.25,26 The main goal of the current article is to develop a good candidate to qualify the adaptive lubricant, by studying the tribological behavior of the innovative multilayer VN–VN/Ag coating at a temperature range of 25 to 700 °C.

2. Experimental details

2.1 Coating deposition

Mechanically polished Inconel 718 (20 mm × 20 mm × 3 mm) and single crystal Si (100) wafers were used as the substrates, and were ultrasonically cleaned in acetone and ethanol for 15 min consecutively. The coatings were prepared using a multi-arc ion plating system (Hauzer Flexicoat 850) to sputter three V targets (99.95 wt% purity) and three Ag targets (99.99 wt% purity), which were 63 mm in diameter. The coatings were prepared in a pure N2 (99.99%, volume fraction) atmosphere with a discharge of 350 sccm. The main technological steps and deposition parameters are summarized in Table 1. As shown in Table 1, prior to deposition, the chamber was pumped down to a base pressure below 5 × 10−5 mbar, and the substrates were etched subsequently by Ar+ bombardments for 2 min with a pulsed bias voltage of −900, −1100, or −1200 V, to remove the thin oxide layer and other adherent impurities. Then deposition was conducted at a bias voltage of −100 V and at a distance of 100 mm from the substrate to the target with a rotational speed of the specimen holder of 3 rpm. During deposition the substrate temperature was increased to 400 °C, and the total pressure of the sputtering gas was controlled at 5 × 10−2 mbar. For the multilayer coating, the VN layer was first deposited by applying V to the targets with a constant current of 60 A for 16 min, followed by the VN/Ag layer with a Ag target current of 50 A and unaltered V target parameter for 4 min. The VN layer and the alternating VN/Ag layer were setup as one block. The deposition was repeated 11 times to form the multilayer coating. For comparison, a monolayer VN coating was also prepared as the Ag targets were shut off, whereas other deposition parameters remained unchanged relative to the VN–VN/Ag.
Table 1 The main technological steps and deposition parameters for the multilayer VN–VN/Ag coating obtained in pure N2 using a multi-arc ion plating apparatus equipped with individual V and Ag targets
  VN–VN/Ag
VN layer VN/Ag layer
Base pressure (mbar) 5.0 × 10−5
Ar plasma etching 2 min under a bias voltage of −900, −1100, or −1200 V
Substrate bias (V) −100
Substrate temperature (°C) 400
Working pressure (mbar) 5.0 × 10−2
V target current (A) 60
Ag target current (A) − 50
Sputtering time (min) 16 4

2.2 Characterization

The surface morphologies of the as-deposited coatings were characterized using a field emission scanning electron microscope (SEM, FEI Quanta FEG 250), which was equipped with an energy-dispersive spectroscopy system (EDS, Oxford Instruments X-Max) for elemental analysis. The roughness measurement was carried out using an atomic force microscope (AIST-NT). The microstructures of both the samples were identified using X-ray diffraction apparatus (XRD, Bruker D8 diffractometer). All patterns were measured using Cu-Kα radiation (wavelength λ = 0.154 nm) with an incidence angle of 2° at a current of 40 mA and an accelerating voltage of 40 kV. The scanning angle 2θ ranged from 10° to 90° at a scanning speed of 4° per min with 0.02° step size. Further characterization of the coating microstructure was performed using a high-resolution transmission electron microscope (TEM, FEI TF20) operated at 300 kV.

The scratching response of the total coatings was examined using a scratch tester (CSM Instruments Revetest, Switzerland) with a conical diamond indenter with a 0.2 mm tip radius and 120° taper angle. The sample table moved horizontally along the sample to attain a scratch length of 5 mm at a predetermined speed of 5 mm min−1. An increasing normal load, starting from zero, was applied on the tip stylus at a preset loading rate of 118 N min−1. The adhesion force (LC) was detected using the acoustic emission signal, which consisted of the critical load LC1 at the first surface fracture and the critical load LC2 at the whole film delamination from the substrate.

The hardness and elastic modulus were obtained using a nano indenter (MTS Nano Indenter G200) system equipped with a Berkovich diamond tip using the continuous stiffness option. The maximum indentation depth was limited to about 400 nm, which was around one-tenth of the film thickness. A total of six indentations were configured on different areas of each sample to obtain reliable statistics. The desired values were deduced by analyzing the load–displacement curves using the Oliver–Pharr method.27

The tribological test was performed using the high temperature module of a ball-on-disk tribometer (CETR UMT-3, USA) under dry sliding conditions in ambient air, which runs in unidirectional sliding mode using a rotating sample holder. Various temperatures (RT, 300, 500, 700 °C) were used in the research and the Al2O3 balls with a diameter of 10 mm, a surface roughness of 53 nm and a hardness value of 9 GPa were adopted as the counterpart. A constant normal load of 10 N, a sliding speed of 100 rpm, a sliding track diameter of 6 mm and a sliding time of 30 min were used in the experiments. The friction coefficient was continuously recorded during the tribotesting. Furthermore, based on the cross-sectional profiles of the wear track detected using a profiler (KLA Tencor Alpha-Step IQ), the wear loss could be obtained after the sliding tests were completed, then the wear rate was calculated using the following classical wear equation:28

 
W = V/(S × L) (1)
where W is the wear rate of the coating, V is the wear loss, S is the accumulated sliding distance and L is the applied normal load. The surface topographies of the wear track and wear scar were characterized using SEM, in addition, the phase composition of the worn area was detected using a micro-Raman spectrometer (Renishaw plc, UK) equipped with a 50× objective lens to locate the regions of interest.

3. Results and discussion

3.1 Coating morphology and microstructure

The surface morphologies of the as-deposited VN and multilayer VN–VN/Ag coatings are shown in Fig. 1. As can be seen from Fig. 1a, the VN coating shows a pronounced roughness because of the generation of μm and sub-μm sized droplets during the deposition process. In general, the droplets display a shape associated with a liquid drop with the pointed end to the specimen surface. These small sized micro particles were dispersed over the coating surface and are impossible to avoid with the vacuum arc deposition technique. However, the co-deposition of Ag and V results in a roughening of the otherwise relatively smooth VN coating (Fig. 1b). By measuring, the average roughness of the multilayer VN–VN/Ag coating is found to be about 153 nm, which is higher than the value of 120 nm for the VN coating alone. This roughening can be attributed to the segregation of a large quantity of metallic silver from the VN lattice in tandem with subsequent migration to the specimen surface under a relatively high processing temperature. As in Fig. 1c and d, the EDS analysis shows that the large droplets are mainly elemental V for the VN coating, whereas a substantial Ag content is exhibited for the multilayer VN–VN/Ag coating.
image file: c6ra02370c-f1.tif
Fig. 1 SEM images and EDS results of VN (a and c) and multilayer VN–VN/Ag (b and d) coatings.

The X-ray diffraction patterns of the VN and multilayer VN–VN/Ag coatings are presented in Fig. 2. For the VN coating, the data show a tetragonal structure with a strong (200) peak around the diffraction angle of 44.3°, and other weaker peaks at 38.2°, 64.4°, 77.4° and 81.7°. For the multilayer VN–VN/Ag coating, Fig. 2 shows a similar pattern but the strongest (111) peak at 38.2° with the emergence of the co-deposition of V and Ag elements in the VN/Ag layer. Furthermore, the intensity and breadth of the total diffraction peaks of the VN–VN/Ag coating increase in comparison with those of the VN coating, which can be explained by the superposition of the patterns of the VN phase and Ag phase because of their close interplanar spacing. Also, the grain refinement, which results from the compounding of VN and Ag particles, may be another reason for the increase in peak breadth and intensity. Eqn (2) is an approach to calculating the average grain size of VN and multilayer VN–VN/Ag coatings, which is:29

 
Dc = Kλ/β[thin space (1/6-em)]cos[thin space (1/6-em)]θ (2)
where Dc is the average grain size (nm), K is the Scherrer constant (K = 0.89), λ is the X-ray wavelength, β is the half-width of the diffraction peak and θ is the Bragg angle (°). So the average grain sizes of the monolayer VN coating and multilayer VN–VN/Ag coating were 17.3 nm and 12.4 nm, respectively. It is obvious that the grains are refined in the composite structure by doping Ag into the VN lattice. Furthermore, the strong (111) preferred orientation at 38.2° implies a dense crystalline structure because the plane (111) is a densely packed plane.30 Furthermore, the Ag2O (111) peak is also observed in the multilayer VN–VN/Ag coating, indicating the oxidization of silver element.


image file: c6ra02370c-f2.tif
Fig. 2 XRD patterns of VN and multilayer VN–VN/Ag coatings.

The TEM micrographs further confirm the microstructures of the as-deposited VN and multilayer VN–VN/Ag coatings, as shown in Fig. 3. In Fig. 3a, the VN (111) and VN (200) crystal faces of the VN coating are exhibited by measuring their interspaces in the high-resolution image, and validated by the subsequent selected area electron diffraction patterns. Whereas in Fig. 3b, the lamellar structure of the multilayer characteristics can be observed clearly, and the average modulation period measured from the TEM observation is about 330 nm, and the modulation ratio (VN layer to VN/Ag layer) is about 1.54. In Fig. 3c, the bright field image at high magnification indicates an obvious diffuse structure in the VN/Ag layer. The EDX (Energy Dispersive X-ray) analyses have produced the specific element percentage in different regions of the VN–VN/Ag coating, as shown in Table 2. The test point A is located in the VN/Ag layer, which shows an element composition mainly including V, N, Ag, and the ratio of atom percent of V and N is greater than 1[thin space (1/6-em)]:[thin space (1/6-em)]1 (i.e., V[thin space (1/6-em)]:[thin space (1/6-em)]N > 1[thin space (1/6-em)]:[thin space (1/6-em)]1), suggesting that the VN/Ag layer is a vanadium rich structure. In contrast, test point B shows that the VN layer is mainly composed of V and N, and V[thin space (1/6-em)]:[thin space (1/6-em)]N < 1[thin space (1/6-em)]:[thin space (1/6-em)]1. The test results reveal that the metallic silver particles impede the reaction of the N element in the VN/Ag layer. So it is deduced that the intermittent implanting of elemental silver in the VN matrix has a substantial impact on the microstructure of the coating, thus forming a diffuse structure in the VN/Ag layer.


image file: c6ra02370c-f3.tif
Fig. 3 TEM results for (a) monolayer VN coating, (b) alternately-layered structure at low magnification of the multilayer VN–VN/Ag coating, and (c) the diffuse structure of the VN/Ag layer at high magnification.
Table 2 Percentages of elements in different regions of multilayer VN–VN/Ag coating (at%)
Point V N Ag
A 49.54 43.12 7.34
B 47.78 52.22 —

3.2 Mechanical properties

The hardness H and elastic modulus E of the VN and multilayer VN–VN/Ag coatings are presented in Table 3. For the VN coating, the average hardness and modulus are about 30.2 GPa and 350.5 GPa, respectively. After silver is doped into the VN coating to form the multilayer structure, the average hardness and modulus are reduced to 12.6 GPa and 220.0 GPa, respectively. The lower hardness and modulus of the multilayer VN–VN/Ag coating in comparison to the VN coating may be related to the solid solution effect of soft metallic Ag in the VN lattice, which leads to a diffuse structure and layered form in the VN–VN/Ag coating. These characteristics may contribute to an effective release of residual stress because it resists crack growth, thus resulting in a decline of the hardness and modulus in the VN–VN/Ag coating. Furthermore, for the tribological coatings, the H3/E2 ratio is always regarded as a key parameter to predict the plastic deformation resistance of a material:31–33
 
Py = 0.78r2(H3/E2) (3)
where Py is the yield stress of a coating, r is the radius of contact area. It is observed that the higher the H3/E2 ratio, the better the plastic deformation resistance of a coating. The value of H3/E2 for the multilayer VN–VN/Ag coating is 0.0413 GPa, which is lower than that of 0.2242 GPa for the VN coating, indicating an unfavorable plasticity resistance of the multilayer VN–VN/Ag coating, which is probably because of the ductility of malleable Ag.
Table 3 The hardness, elastic modulus, E/H and H3/E2 of VN and multilayer VN–VN/Ag coatings
Sample H (GPa) E (GPa) E/H H3/E2 (GPa)
VN 30.2 350.5 11.606 0.2242
VN–VN/Ag 12.6 220.0 17.4603 0.0413

The scratch morphologies and acoustic signals of VN and multilayer VN–VN/Ag coatings are shown in Fig. 4. The first cracking (LC1) of the multilayer VN–VN/Ag coating occurred at 76.5 N, and the first cracking of the VN coating appeared at 51.2 N. The whole delamination (LC2) of the multilayer VN–VN/Ag and VN coatings occurred at 132.3 N and 84.6 N, respectively. It is obvious that both LC1 and LC2 of the multilayer VN–VN/Ag coating are higher than those of the VN coating, indicating that the adhesion of the multilayer VN–VN/Ag coating is better than that of the VN coating. The higher adhesion strength of the multilayer VN–VN/Ag coating may be attributed to its higher fracture toughness, which will make the coating more difficult to be broken after delamination or buckling. The fracture toughness of a coating was calculated using eqn (4):34

 
Kc = α1(E/H)1/2(P/c3/2) (4)
where Kc is the fracture toughness of a coating, P is the maximum applied load, α1 is an experience constant, c is the radial crack length and E/H is the ratio of elastic modulus to hardness. Combined with the results from Table 3, the E/H ratio of VN coating is found to be about 11.606, while that of the multilayer VN–VN/Ag coating increases to 17.4603, indicating that the fracture resistance of the VN–VN/Ag coating is higher than that of the VN coating.


image file: c6ra02370c-f4.tif
Fig. 4 The scratch morphologies and acoustic signals of the VN and VN–VN/Ag coatings.

3.3 Tribological properties

The friction coefficients of the VN and multilayer VN–VN/Ag coatings sliding against Al2O3 balls are presented in Fig. 5. The first part of the friction coefficients represents the run-in period with a large fluctuation of the value, while the second part obviously levels off as the contact surfaces of the friction pair become smoother as the sliding proceeds. It is observed that the steady friction coefficients of the monolayer VN and the multilayer VN–VN/Ag coatings are about 0.59 and 0.44, respectively, at room temperature. When the temperature is increased to 300 °C, the friction coefficient of the VN–VN/Ag coating shows a slight decrease as compared with the results obtained at RT, but for the VN coating the value of the CoF almost remains the same. The CoF values of the multilayer VN–VN/Ag coating are about 25% and 31% lower than those of the VN coating at RT and 300 °C respectively, which indicates the lubricating action of metallic Ag in the VN matrix. After 300 °C, the friction coefficient of the VN sample shows a drastic descending trend until 700 °C because of the formation of the lubricious Magnéli phase, but the multilayer VN–VN/Ag coating gives an insignificant friction coefficient value at 500 °C, which is probably ascribed to the surface roughening caused by the adhesive wear. At 700 °C, the friction coefficient of VN–VN/Ag coating attains its lowest value of 0.19, which is about 32% lower than the lowest one of 0.28 for the VN coating. The decrease of the friction coefficient at elevated temperatures for the VN–VN/Ag coating is mainly because of the occurrence of graphitization and the liquid lubricating phases. Overall, the multilayer structure with the addition of Ag particles into the VN lattice displays enhanced frictional properties in contrast to the VN reference film especially at high temperatures.
image file: c6ra02370c-f5.tif
Fig. 5 The friction coefficients of the (a) VN and (b) multilayer VN–VN/Ag coatings sliding against Al2O3 balls at different temperatures in ambient air.

The worn surface morphologies of VN and multilayer VN–VN/Ag coatings sliding against Al2O3 balls under different temperatures are shown in Fig. 6. At room temperature, a large number of shallow cavities and some flakes and delamination pits can be seen on the worn surface of the VN coating (Fig. 6a), suggesting that the wear mechanism is dominated by fatigue wear. After silver is doped into the VN coating to form the multilayer VN–VN/Ag structure, there is a lubrication layer rich in silver above the worn surface (Fig. 6b), which causes the coefficient of friction to be lower than that of the VN coating. When the friction temperature is increased to 300 °C, the worn surface of the VN coating is comparatively smooth and flat, and accompanied by a small amount of wear debris piling up on both sides of the wear track (Fig. 6c). However, the multilayer VN–VN/Ag coating has some shallow scratch grooves on the bottom of the worn surface (Fig. 6d), indicating that the wear mechanism is micro ploughing. The wear product accumulated at both edges of the wear track is increased in amount and its particle sizes are distinctly enlarged. At 500 °C, for the VN coating (Fig. 6e), the worn surface is glossy and compact, meanwhile there are trace amounts of adhesion plaques attached to the friction interface, and the wear debris transferred to the edges coheres in clumps. In contrast, the multilayer VN–VN/Ag structure displays a gradient slope related to the wear track (Fig. 6f), as the geometrical sizes of the wear debris become obviously larger and its morphology is loose clusters on both sides of the contact area. With the increase of temperature to 700 °C, plenty of wear debris is produced during the tribological testing of the VN coating, some of which is laminated to cover most of the wear track, while some is agglomerated in a blocky structure on both of its sides (Fig. 6g). This caking for the VN coating may be caused by the liquid oxidative product (i.e., V2O5, with a melting point of 674 °C (ref. 35)) agglutinating during the cooling process. When the multilayer VN–VN/Ag coating slides against Al2O3 balls under the same high temperature condition, the frictional surface brings out some rugged asperities and ridges (Fig. 6h), indicating the solidification of a liquid phase (i.e., AgVO3).


image file: c6ra02370c-f6.tif
Fig. 6 Worn surface morphologies of VN and multilayer VN–VN/Ag coatings at different temperatures.

Fig. 7 shows the cross-sectional profiles of the wear track on VN and VN–VN/Ag coatings at room temperature, 300 °C and 500 °C in ambient air. At RT, the maximum depth of the wear track for the VN coating is about 529 nm, while that for the VN–VN/Ag coating is about 440 nm, both of which are the lowest values found with these temperatures. At 300 °C, the values for VN and VN–VN/Ag coatings are about 792 nm and 3410 nm, respectively. When the temperature is increased to 500 °C, the maximum depth of the VN coating attains the highest value of 1109 nm, but that of the VN–VN/Ag coating achieves its highest value of 3740 nm. Meanwhile, the widths of the wear track of the VN and VN–VN/Ag coatings are about 364 μm and 106 μm, respectively, at room temperature. When the temperature was increased to 300 °C, the values for both coatings were about 445 μm and 709 μm. At 500 °C, the highest values of 488 μm and 865 μm were attained for the VN coating and VN–VN/Ag coating, respectively. As can be seen, the maximum depth and width of the wear tracks of both coatings show an ascending trend as the temperature rises. Furthermore, the maximum depth and width of the VN–VN/Ag coating are about 17% and 71% lower at room temperature, but about 70% and 44% higher at 500 °C, than those of the VN coating, indicating that it has a better wear resistance at lower temperature but a poorer one at a medium high temperature. Presumably this is because metallic Ag had a lubricating action at RT, but it appears that it showed some level of plastic deformation at higher temperatures. However, at 700 °C, both coating samples show the generation of a liquid phase, i.e., V2O5 for the monolayer VN coating and AgVO3 for the multilayer VN–VN/Ag structure, so the depth and width of the wear tracks are not suitable to be displayed in Fig. 7.


image file: c6ra02370c-f7.tif
Fig. 7 Sectional profiles of the wear track on (a) VN and (b) multilayer VN–VN/Ag coatings.

Based on the results of the cross-sectional profile of the wear track, the wear rates of the VN and multilayer VN–VN/Ag coatings at room temperature, 300 °C, or 500 °C in ambient air are shown in Fig. 8. At room temperature, the wear rate is at its minimal value, of about 3.561 × 10−6 mm3 N−1 m−1 and 1.881 × 10−6 mm3 N−1 m−1 for VN and multilayer VN–VN/Ag coatings, respectively. With the rise of temperature, the wear rates of both the coatings show an ascending trend until 500 °C is reached, where the coatings show the highest values of about 1.079 × 10−5 mm3 N−1 m−1 and 6.107 × 10−5 mm3 N−1 m−1, for VN and VN–VN/Ag coatings, respectively. The initial increase of the wear rate can be attributed to the removal of the original material from the sliding surface because of mechanical wear. However, the further increase is produced by surface oxidation and a tribochemical reaction under the combined action of high temperature exposure to air with sliding contact. In addition, the wear rate of the multilayer VN–VN/Ag coating is lower than that of the VN coating at RT, but higher at 300 °C and 500 °C. Although the wear rates of both coatings by mechanical wear should be lower in high temperature environments than at room temperature because of the production of the Magnéli phase and the graphitization lubricating the film, for VN and VN–VN/Ag coatings, respectively, on the friction interface, it might be higher in practice when the tribochemical reaction dominates the wear mechanism. On the whole, temperature can have a diametrically opposite effect on the anti-friction and anti-wear properties of the coatings by either contributing towards high wear by accelerating the formation of oxides or binary metal oxides, or by reducing the friction coefficient by forming an extremely smooth lubricating film.


image file: c6ra02370c-f8.tif
Fig. 8 The wear rates of the VN and multilayer VN–VN/Ag coatings at room temperature, 300 °C, or 500 °C in ambient air.

For further analysis of the wear mechanism of the VN and multilayer VN–VN/Ag coatings sliding against Al2O3 balls in ambient air, the wear scars of the counterpart at different friction temperatures are shown in Fig. 9. At room temperature, for the VN coating as shown in Fig. 9a, and for the VN–VN/Ag coating as shown in Fig. 9b, the wear scars on the Al2O3 balls are all thin and unconspicuous, and the transferred wear debris from the both coating samples is not much. When the temperature rises to 300 °C and 500 °C, the transferred wear debris on the surface of the counterpart apparently increases in amount, and the wear scar presents as an obvious circle (see Fig. 9c–f). When the temperature is further increased to 700 °C, the amount of the transferred debris is augmented significantly, and the appearance displays a distinct blocky structure with a larger thickness of the transfer layer, especially for the multilayer VN–VN/Ag coating (see Fig. 9g and h). Furthermore, because all the counterparts have the same diameter, a large/small wear scar indicates severe/mild wear to some extent. The diameters of the wear scar on the Al2O3 balls are shown in Fig. 10. At room temperature, the diameter of the wear scar for the VN coating is about 347 μm. After silver is doped into the VN coating, the diameter of the wear scar reduces to 184 μm. With an increase of temperature, the diameters show an ascending trend until 700 °C is reached except for a slight decrease for the VN coating at 300 °C. At 700 °C, it reaches the highest values of about 435 μm and 877 μm for VN and multilayer VN–VN/Ag coatings, respectively. It is significant that the diameter of the wear scar for the VN–VN/Ag coating is lower than that of the VN coating at room temperature, but higher at mid-high and high temperatures.


image file: c6ra02370c-f9.tif
Fig. 9 The SEM images of wear scar on Al2O3 balls matching with VN and multilayer VN–VN/Ag coatings at different friction temperatures.

image file: c6ra02370c-f10.tif
Fig. 10 The diameters of wear scar on Al2O3 balls matching with VN and multilayer VN–VN/Ag coatings at different friction temperatures.

3.4 Phase composition of the worn surface

As mentioned previously, the excellent friction performance of coating samples depends mainly on the new lubricating phases at the high-temperature region. The micro-Raman data used to investigate the phase composition of the worn surface at 500 °C and 700 °C are shown in Fig. 11. For the VN coating (Fig. 11a), it is observed that the Raman spectrum is vibrating at 146 cm−1, 286 cm−1, 406 cm−1, 697 cm−1, 993 cm−1 and so on when the sample is taken from the central area of the worn surface at 500 °C, of which the shape is identical to a V2O5 standard presented by Constable et al.36 As the friction temperature is further increased to 700 °C, the V2O5 peaks appear as well but with an increasing intensity, manifesting a further anabatic oxidation process. The presence of V2O5, a Magnéli phase,37 and its liquid state at 700 °C, is mainly responsible for the decrease of the friction coefficient at elevated temperatures. For the multilayer VN–VN/Ag sample (Fig. 11b), at 500 °C, the Raman peaks at 711 cm−1, 763 cm−1, 921 cm−1 and so on are detected for the first time, and these can be assigned to the formation of Ag3VO4 and AgVO3 compounds. This phenomenon is attributed to the chemical reaction of elemental Ag and vanadium oxide with Ag3VO4 as the product at high temperature exposure to air, as well as the transformation dominated by Ag3VO4 → Ag + liquid AgVO3 at about 450 °C.14 However, the subsequent segregation is suppressed by the friction process, i.e., the Ag3VO4 phase is retained in the tribo-contact area, but the AgVO3 phase presently reappears as the counterpart is removed from the worn surface. So the main lubricant serviced for the tribological test at 500 °C is silver vanadate besides vanadium pentoxide. The lubrication mechanism of vanadium pentoxide is consistent with that for the VN coating, whereas that of Ag3VO4 is because of the layered anti-zinc blende structure that was described in detail by Albrecht et al.,38 where the weak interplanar Ag–O and Ag–Ag bonds can be sheared easily with the application of a load, thus generating the kind of graphitized lubricating film on the friction interface. At 700 °C, the Raman spectrum shows a similar pattern, but with the intensities at 711 cm−1, 921 cm−1, 986 cm−1 becoming weaker and the ones at 761 cm−1, 849 cm−1 and 895 cm−1 becoming stronger. The temperature dependent characteristic of the spectrum intensity is largely related to the transformation between Ag3VO4 and AgVO3, which further proceeds because the AgVO3 is more stable than the Ag3VO4 at high temperatures. So the AgVO3 phase dominates in the compound, whereas the Ag3VO4 phase becomes subordinate as the temperature increases. So at 700 °C, the main lubricant is the liquid AgVO3 phase in addition to vanadium pentoxide, which provides the decrease of the friction coefficient.
image file: c6ra02370c-f11.tif
Fig. 11 The Raman spectra taken from the central area of the worn surface of (a) VN and (b) multilayer VN–VN/Ag coatings at 500 °C and 700 °C.

Fig. 12 shows the temperature dependent composition of the lubricating phase on the worn surface of the multilayer VN–VN/Ag coating, which benefits the decrease of friction coefficient in a wide temperature range. Firstly, it can be seen that metallic Ag acts as the main lubricant at room temperature and at a lower temperature of 300 °C, forming a smooth glaze layer on the worn area during wear testing. At a mid-high temperature of about 500 °C, oxidation of materials is unavoidable, thus, the formation of vanadium pentoxide and silver vanadate is responsible for the improvement of the antifriction property, and the Ag3VO4 with a lamellar graphitization structure in particular is the main lubricant. Finally, Ag3VO4 has almost disappeared but the liquid AgVO3 in conjunction with V2O5 serves as the primary lubricant at the elevated temperature of 700 °C. Thus, it can be deduced that the worn surface of the Ag-doped sample is all composed of suitable lubricants at different temperatures. The self-adaptive action of the phase composition on the wear track, explains the protective principle of the multilayer VN–VN/Ag structure, suggesting its potential application as an external coating for the parts exposed to friction, operated in a wide temperature range.


image file: c6ra02370c-f12.tif
Fig. 12 Schematic of the self-adaptive action of a multilayer VN–VN/Ag coating operating in a wide temperature range.

4. Conclusions

The monolayer VN and multilayer VN–VN/Ag coatings were prepared on Inconel 718 and Si (100) substrates using a multi-arc ion plating technique. The effect of Ag-dopant on the microstructure, mechanical properties and wide temperature tribological properties of the VN matrix have been investigated. The main conclusions that can be drawn are:

(1) The intermittent implanting of silver nanoparticles into the VN coating can promote the decrease of average grain size, and the formation of a diffuse structure in the VN/Ag layer.

(2) The developed multilayer structure of the VN–VN/Ag coating shows an enhanced adhesion strength in comparison with the monolayer VN coating, but its microhardness and elastic modulus are decreased with the formation of a high content of soft metallic Ag.

(3) Tribological tests show that the multilayer VN–VN/Ag coating has an excellent adaptive lubricating property from 25 to 700 °C. A favorable friction coefficient of 0.19 is obtained because of the generation of a Magnéli phase and graphitized bimetallic oxide.

(4) The compound of V2O5, Ag3VO4 and AgVO3 is formed to act as the lubricant when the temperature increases to a high level, but severe wear is produced because of oxidation and the tribochemical reaction under exposure to air.

Acknowledgements

The authors gratefully acknowledge the financial support from the National Natural Science Foundation of China (Grant nos 51202261 & 51475449), the National Basic Research Program of China (973 Program) (Grant no. 2013CB632302) and the State Key Laboratory for Mechanical Behavior of Materials Foundation (Grant no. 20141605).

References

  1. C. Donnet and A. Erdemir, Tribol. Lett., 2004, 17, 389–397 CrossRef CAS.
  2. W. Wang, Surf. Coat. Technol., 2004, 177, 12–17 Search PubMed.
  3. C. Dellcorte, Tribol. Trans., 2000, 43, 257–262 CrossRef.
  4. Q. Y. Li, Z. B. Feng, L. H. Liu, J. Sun, Y. T. Qu, F. H. Li and M. Z. An, RSC Adv., 2015, 5, 12025–12033 RSC.
  5. S. Zhu, Q. Bi, M. Niu, J. Yang and W. Liu, Wear, 2012, 274, 423–434 CrossRef.
  6. S. M. Aouadi, H. Gao, A. Martini, T. W. Scharf and C. Muratore, Surf. Coat. Technol., 2014, 257, 266–277 CrossRef CAS.
  7. M. J. Cui, J. B. Pu, J. Liang, L. P. Wang, G. G. Zhang and Q. J. Xue, RSC Adv., 2015, 5, 104829–104840 RSC.
  8. A. A. Voevodin, C. Muratore and S. M. Aouadi, Surf. Coat. Technol., 2014, 257, 247–265 CrossRef CAS.
  9. R. H. Zhang, Z. B. Lu and L. P. Wang, RSC Adv., 2015, 5, 32690–32697 RSC.
  10. T. M. Gong, P. P. Yao, Y. L. Xiao, K. Y. Fan, H. Q. Tan, Z. Y. Zhang, L. Zhao, H. B. Zhou and M. W. Deng, Wear, 2015, 328–329, 270–276 CrossRef CAS.
  11. G. X. Xie, D. Guo and J. B. Luo, Tribol. Int., 2015, 84, 22–35 CrossRef.
  12. C. Muratore and A. A. Voevodin, Annu. Rev. Mater. Res., 2009, 39, 297–324 CrossRef CAS.
  13. D. Stone, J. Liu, D. P. Singh, C. Muratore, A. A. Voevodin, S. Mishra, C. Rebholz, Q. Ge and S. M. Aouadi, Scr. Mater., 2010, 62, 735–738 CrossRef CAS.
  14. S. M. Aouadi, D. P. Singh, D. S. Stone, K. Polychronopoulou, F. Nahif, C. Rebholz, C. Muratore and A. A. Voevodin, Acta Mater., 2010, 58, 5326–5331 CrossRef CAS.
  15. S. M. Aouadi, B. Luster, P. Kohli, C. Muratore and A. A. Voevodin, Surf. Coat. Technol., 2009, 204, 962–968 CrossRef CAS.
  16. J. N. Kim, S. Park, T. Kim and J. J. Lee, Thin Solid Films, 2011, 519, 6876–6880 CrossRef CAS.
  17. T. Suszko, W. Gulbiński and J. Jagielski, Surf. Coat. Technol., 2006, 200, 6288–6292 CrossRef CAS.
  18. R. R. Chromik, C. C. Baker, A. A. Voevodin and K. J. Wahl, Wear, 2007, 262, 1239–1252 CrossRef CAS.
  19. D. S. Stone, J. Migas, A. Martini, T. Smith, C. Muratore, A. A. Voevodin and S. M. Aouadi, Surf. Coat. Technol., 2012, 206, 4316–4321 CrossRef CAS.
  20. K. J. Wahl, L. E. Seitzman, R. N. Bolster, I. L. Singer and M. B. Peterson, Surf. Coat. Technol., 1997, 89, 245–251 CrossRef CAS.
  21. J. J. Hu, C. Muratore and A. A. Voevodin, Compos. Sci. Technol., 2007, 67, 336–347 CrossRef CAS.
  22. C. P. Mulligan, T. A. Blanchet and D. Gall, Wear, 2010, 269, 125–131 CrossRef CAS.
  23. P. A. Papi, C. P. Mulligan and D. Gall, Thin Solid Films, 2012, 524, 211–217 CrossRef CAS.
  24. C. P. Mulligan, P. A. Papi and D. Gall, Thin Solid Films, 2012, 520, 6774–6779 CrossRef CAS.
  25. P. Wieciński, J. Smolik, H. Garbacz and K. J. Kurzydłowski, Vacuum, 2014, 107, 277–283 CrossRef.
  26. E. Martınez, J. Romero, A. Lousa and J. Esteve, Surf. Coat. Technol., 2003, 163–164, 571–577 CrossRef.
  27. W. C. Oliver and G. M. Pharr, J. Mater. Res., 1992, 7, 1564–1583 CrossRef CAS.
  28. Y. Wang, L. Wang, J. Li, J. Chen and Q. Xue, Tribol. Int., 2013, 60, 147–155 CrossRef CAS.
  29. A. W. Burton, K. Ong, T. Rea and I. Y. Chan, Microporous Mesoporous Mater., 2009, 117, 75–90 CrossRef CAS.
  30. R. Antunes, A. Rodas, N. Lima, O. Higa and I. Costa, Surf. Coat. Technol., 2010, 205, 2074–2081 CrossRef CAS.
  31. N. A. Sakharova, J. V. Fernandes, M. C. Oliveira and J. M. Antunes, J. Mater. Sci., 2010, 45, 3812–3823 CrossRef CAS.
  32. T. L. Oberle, J. Met., 1951, 3, 438–439 Search PubMed.
  33. Y. W. Ye, Y. X. Wang, H. Chen, J. L. Li, S. G. Zhou and Q. J. Xue, Surf. Coat. Technol., 2015, 270, 305–313 CrossRef CAS.
  34. B. Bhushan, Diamond Relat. Mater., 1999, 8, 1985–2015 CrossRef CAS.
  35. International Chemical Safety Cards, ICSC: 0596.
  36. C. P. Constable, J. Yarwood, P. Hovsepian, L. A. Donohue, D. B. Lewis and W.-D. Münz, J. Vac. Sci. Technol., A, 2000, 18, 1681–1689 CAS.
  37. Z. Zhou, W. M. Rainforth, Q. Luo, P. Eh. Hovsepian, J. J. Ojeda and M. E. Romero-Gonzalez, Acta Mater., 2010, 58, 2912–2925 CrossRef CAS.
  38. T. A. Albrecht, C. L. Stern and K. R. Poeppelmeier, Inorg. Chem., 2007, 46, 1704–1708 CrossRef CAS PubMed.
This journal is © The Royal Society of Chemistry 2016
Click here to see how this site uses Cookies. View our privacy policy here.