Multilayered diamond mechanical seal rings under biodiesel lubrication and the full sealing conditions of pressurized water

Wear 384-385 (2017) 178–184

Contents lists available at ScienceDirect

Wear journal homepage: www.elsevier.com/locate/wear

Multilayered diamond mechanical seal rings under biodiesel lubrication and the full sealing conditions of pressurized water M. Shabani a, J.M. Carrapichano b, F.J. Oliveira a, R.F. Silva a,n a b

CICECO, Dept. of Materials & Ceramic Engineering, University of Aveiro, 3810-193 Aveiro, Portugal Mechanical Eng. Dept., Coimbra Superior Eng. Inst., 3040-228 Coimbra, Portugal

art ic l e i nf o

a b s t r a c t

Article history: Received 2 September 2016 Received in revised form 12 January 2017 Accepted 13 January 2017

A mechanical sealing tribosystem based on multilayer microcrystalline/nanocrystalline diamond (MCD/ NCD) coatings is proposed. The tribological behaviour was investigated in ring–on–ring self–mated planar contact configuration. Tests were carried out under biodiesel lubrication and pressurized water (2 bar) sealing conditions. A 3D optical confocal profilometer was used to obtain for the first time the wear coefficients of seal rings, revealing a value of k¼6.3  10  10 mm3/N.m for biodiesel lubrication that is two orders of magnitude lower than for reciprocating sliding ball-on-flat experiments. The advantage of the multilayer strategy was demonstrated in the sealing of pressurized water, where full sealing conditions were achieved within the P  V range of 0.75–5.5 MPa ms  1. For these coatings, full delamination is prevented due to structural discontinuity at the MCD/NCD interfaces whereas for a MCD monolayer coating, premature failure takes place and sealing is not reached. & 2017 Elsevier B.V. All rights reserved.

Keywords: CVD diamond Multilayers Mechanical seal rings biodiesel

1. Introduction Mechanical face seals are critical parts of equipment and machinery across all human activities related to transport of fluids, wherever leakage control and safety issues regarding facilities, environment and people are at stake. The materials used for the seal itself, the moving surfaces where the sealing is performed, span a large range both in their nature and their working principles. There are basically two types of seal combinations: i) one of the materials is sacrificial and will wear over time, with scheduled replacements every few months (or years) of operation; ii) both counteracting materials are hard, can be made of the same or different materials, and they will have expectedly longer running lives. Common to both types of interfaces is the need of very low friction coefficients, minimization of wear rates and a maximum reliability. Additional criteria that will affect the choice of materials are e.g. enhanced corrosion resistance, low weight, ability to operate at higher temperatures or the need to endure dry cycles. From the standpoint of the materials, this means a mixture of thermo-mechanical and surface or chemical properties that seldom are satisfied by a single material. Diamond is an exception, being extremely hard and corrosion resistant, possessing the highest thermal conductivity and a very high thermal shock n

Corresponding author. E-mail addresses: [email protected] (M. Shabani), [email protected] (J.M. Carrapichano), fi[email protected] (F.J. Oliveira), [email protected] (R.F. Silva). http://dx.doi.org/10.1016/j.wear.2017.01.058 0043-1648/& 2017 Elsevier B.V. All rights reserved.

resistance [1,2]. In the present work, the hot filament chemical vapour deposition (HFCVD) technique was used to produce multilayer diamond coatings, alternating microcrystalline diamond (MCD) and nanocrystalline diamond (NCD) to a total of ten layers. Silicon nitride ceramic seal rings coated with this multilayer system and with a monolayer MCD coating were tested under biodiesel lubrication and water sealing in self-mated hard contacts. Biodiesel has become an important ecological alternative to petrol. Pumping is extensively used in the biodiesel industry for raw materials, processing and fluid transfer, where the pump rotors and seals are frequently subjected to hot and aggressive conditions and the transfer pumps may become clogged with wax and fatty deposits. The testing with water, the most widely pumped fluid, serves to validate this tribosystem. The strategy of the multilayer coating is to deposit a first MCD layer, due to its superior adhesion to the Si3N4 ceramic [3] and finish with a top NCD layer. This smooth NCD layer [4–7] prevents the high starting friction coefficients that would happen if the rougher MCD was the top surface, thus decreasing the probability of coating failure. Also, the successive intercalation of NCD layers between the MCD ones has twofold advantages: i) it interrupts the columnar growth mode of MCD [8]; ii) it delays the propagation of cracks that would otherwise easily propagate in the growth direction of MCD, thus improving the fracture toughness of the coating [9,10]. The extremely low wear rates of diamond coated rings prevented so far the evaluation of the wear coefficient. In this work, a

M. Shabani et al. / Wear 384-385 (2017) 178–184

relevant and innovative approach of the research is the use of 3D optical profilometry to measure the volume loss after sliding, a feat unreachable by conventional methods such as weighing or AFM analysis.

2. Materials and methods Seal rings were prepared in house by ball milling a mixture of 89.3 wt.% Si3N4 (HC Starck M11), 3.7 wt. % Al2O3 (Alcoa 116SG) and 7.0 wt. % Y2O3 (HC Starck C), drying, pressing and sintering under N2 atmosphere. The sintered silicon nitride rings were machined and flat lapped to the following final dimensions in mm: 43  33  8.5 and 46  31  7.5 (external diameter  internal diameter  thickness). All rings were plasma etched by CF4 during 10 min and ultrasonically seeded with diamond in a nano-diamond slurry during one hour. Diamond coating was accomplished in a home-made semi-industrial HFCVD (hot filament chemical vapour deposition) apparatus with a deposition area as large as 20  30 cm2, using H2/CH4 gas mixtures. Two types of diamond coatings were grown on the Si3N4 seal faces: a tenfold multilayer with alternating layers of micro and nanocrystalline diamond (MCD/NCD) and, for comparison, a monolayer microcrystalline diamond (MCD), both with total thickness of about 10 μm. The deposition conditions of MCD/ NCD multilayer coating are given in Table 1. The monolayer was obtained with the same parameters of MCD but for a total deposition time of 10 h. A rotary tribometer (TE–92 Plint) was used for the tribological tests of the diamond coated Si3N4 mechanical seal faces. The tribological behaviour was investigated in ring–on–ring self–mated planar contact configuration according to ASTM D3702 standard [11]. Fig. 1 shows one pair of rings mounted on the holders to fix on the tribometer. The ring on the left hand side of the photograph is the rotational ring (narrower), while the larger stationary ring (larger width) is on the right hand side. Tests were carried out in biodiesel and in pressurized water (2 bar) conditions, Table 2. The biodiesel is a commercial soybean oil (77%), palm olein (22%) and rapeseed oil (1%) derived biodiesel, with a kinematic viscosity of 4.3 mm2/s at 40 °C. It has an ester content of 98.9 wt%, a cetane number of 51.4 and a maximum of 6 mg/kg of organic and inorganic contaminants. A series of preliminary experiments of short duration of about 2.5 h (Screening in Table 2) was conducted for achieving the stable working conditions under biodiesel lubrication. After these, a single run test with 16 h duration was performed in stable conditions regarding the absence of vibrations or increase in temperature (Stable conditions in Table 2). The coefficient of friction coefficient (COF) values were taken from the instantaneous torque appraised from a load cell. The sealing tests were done with pressurized water (Pwater ¼ 2 bar) and not with biodiesel due to safety issues related to excessive heating of the latter. Test conditions that guarantee full sealing are given in Table 2 (Minimum P  V condition and P  V limit condition). Field emission scanning electron microscopy (FE–SEM) was carried out using a Hitachi SU–70 system for surface morphology characterization of the diamond coated Si3N4 ceramic components before and after the tribological testing. Surface roughness, topographic analysis and quantification of the wear volume was done using a 3D optical profilometer (Sensofar S-neox).

179

Table 1 HFCVD conditions for multilayer diamond deposition on Si3N4 ceramic seal rings (5 MCD þ5 NCD alternating layers).

Filament temperature (°C) Substrate temperature (°C) CH4/H2 flow ratio Total gas flow (ml min–1) Total pressure (mbar) Deposition time per layer (min) Total deposition time (h)

MCD layers

NCD layers

2260 850 0.02 1800 75 60 5

2330 750 0.04 900 25 84 7

Fig. 1. Example of multilayer diamond coated seal rings mounted on the tribometer fixture. Table 2 Test conditions for tribological testing in ring–on–ring self–mated planar contact configuration. F – normal applied load; V – linear velocity; P – nominal contact pressure; t – testing time; L – sliding distance. Type of test

Test objective

V F (m s  1) (N)

P (MPa)

Biodiesel lubrication

Screening

0.5–1.2

Stable conditions Minimum PV condition P  V limit condition

0.5

150– 900 350

0.25–1.50 up to 396 0.59 28.8

up to 220 16

0.5

350– 1000

0.59–1.68 2.8–9

1.5–5

3.9

350– 900

0.59–1.50 295

21

Sealing of pressurized water

L (km)

t (h)

3. Results and discussion 3.1. Morphology and surface roughness of the diamond films The typical morphology of the top surfaces of the diamond coated Si3N4 ceramic mechanical seal rings are illustrated for the monolayer MCD diamond in Fig. 2a and the multilayer MCD/NCD coatings in Fig. 2b. The MCD coating is coarse–grained, having pyramid–like shape diamond crystals, contrarily to the multilayer coated seal rings that have a top layer of nanocrystalline diamond. Details about the crystallinity and phase composition of the MCD and NCD types of diamond are given elsewhere [12]. As illustrated in the above SEM micrographs, the monolayer MCD coating has a much higher surface roughness than the multilayer coatings that have a top NCD layer. The bar chart in Fig. 3 compares the average values of the surface roughness (areal root mean square parameter, Sq) of these two types of coatings as deposited, and also contains the values of the worn surfaces as discussed below in Section 3.2. Since it is a relevant feature in the as-deposited surfaces, it should be mentioned that there is some lack of uniformity between the borders and the centre of the ring

180

M. Shabani et al. / Wear 384-385 (2017) 178–184

Fig. 2. SEM micrographs of the top surfaces of: a) monolayer MCD diamond coated seal ring; b) multilayer MCD/NCD coated seal ring. Insets correspond to higher magnification.

Fig. 3. Areal surface roughness parameter (Sq) at the Centre and near the Edge of the seal rings of the as-deposited coatings and after biodiesel lubricated tests (Worn).

surfaces in both the monolayer and multilayer coated ones. This is most likely due to the edge effect during the CVD deposition that results in larger growth rates in the border regions [13] that almost doubles the surface roughness. In the case of the multilayer coating, two factors contribute to the much lower Sq values: i) the intercalation of the NCD layers prevents the excessive growth of columnar MCD crystals; ii) NCD has an intrinsic surface roughness as low as 16 nm when grown over polished substrates [14], and when used as a top layer it also contributes to the decrease of Sq relatively to monolayer MCD with the same thickness. 3.2. Tribological behaviour under biodiesel lubrication SEM analysis and 3D optical profilometry were used to characterize the morphology and the topography of the seal ring surfaces after the complete set of tests under biodiesel lubrication. Comparing to the corresponding pristine surface in Fig. 2a, the monolayer MCD coated rings show a plateau/valley morphology, characteristic of MCD abraded surfaces [15], Fig. 4a. Sliding between the rough microcrystalline diamond surfaces starts with the

truncation and levelling-off of the vertices and edges of the salient pyramid–shaped diamond crystals. The main wear mechanism is fine scale abrasion resulting in self-polishing of the mating surfaces. For these coatings, the areal surface roughness parameter Sq of the worn ring decreased to one third of the initial value, from 300 nm to 120 nm, Fig. 3. After ring–on–ring sliding for almost 200 h, Table 1, the surface roughness values of the worn areas are still somewhat high when compared to a fully polished surface. It should be stressed that this abraded morphology is found in the outer edge of the smaller, rotational rings, as can be clearly seen in the right hand side of the topographic image in Fig. 5a. This is due to the edge effect during CVD coating, as mentioned in Section 3.1. Only this protuberant edge contacts directly with the larger stationary rings making the worn circular scar. The tracks of the tested multilayer coated surfaces are also a result of self-polishing action, Fig. 4b, showing extensive regions of smooth mesa-tops. For these coatings, Sq is reduced by a factor of nine in the central region of the ring, Fig. 3, from 170 nm to 20 nm, characteristic of a polished surface. This much smaller value is a result of the lower starting surface roughness when comparing to the monolayer MCD coated rings. The consequence of the edge effect is also observed for the multilayer coating, as illustrated by the topographic image of Fig. 5b. Figs. 6 and 7 present 3D surface images together with an example of a topographic profile across the wear tracks of the monolayer and multilayer coated rings, respectively. The MCD coated ring, Fig. 6, depicts a deeper profile that the multilayer MCD/NCD ring, Fig. 7. This is due to the more pronounce edge effect during the growth of a monolayer of microcrystalline diamond. It is worth mentioning that under the test conditions, the maximum depth of the track on the multilayer coated ring, Fig. 7, is smaller than the thickness of the top NCD layer (around 1 mm). The quantification of the wear coefficient of diamond coated seal rings is seldom done due to the limitations of conventional techniques (weighing, AFM) for evaluating the very low wear rates of this type of tribological system and of large dimension

Fig. 4. SEM micrographs of the worn surfaces after sliding under biodiesel lubrication: a) monolayer MCD diamond coated seal ring; b) multilayer MCD/NCD coated seal ring.

M. Shabani et al. / Wear 384-385 (2017) 178–184

181

Fig. 5. Topographic images obtained by 3D optical profilometry of at the outer edge of the rotational rings after sliding under biodiesel lubrication: a) monolayer MCD diamond coated seal ring; b) multilayer MCD/NCD coated seal ring.

specimens. With the newly acquired capability of the 3D optical profilometers with high vertical resolution (down to 3 nm in confocal mode and sub-nanometer for white light interferometry), this is now a feasible high precision, task. The wear coefficient (k) values were evaluated on the static rings where the volume loss is easily assessed since there is always an original surface surrounding it, making possible a direct precise evaluation, unlike the case of the case of the narrower rotational rings where there is no reference surface to evaluate the volume loss at the edge. For the calculation of k, four 3D images as those shown in Figs. 6 and 7 were obtained for each circular wear track at angle shifts of 90°, with a size of 350  1320 mm2 (or 350  1950 mm2) using a 50  objective in confocal mode. Each image was stitched from six to nine single images acquired sequentially across the wear tracks. The software SensoMap was used to directly obtain the volume of the trenches produced by removal of diamond from the surface of the stationary rings. For this, a specific function of SensoMap (Volume of Hole) is used, where the contour of the worn valley in each image is firstly delineated and then the software computes the entire volume of the hole below the surface defined by the contour line. The average value of the four areas (in mm3/mm2) was then computed across the entire projected area of the circular worn track (mm2), thus yielding the wear volume. The wear coefficients were finally calculated dividing the average of these volumes by the total sum of F  L products from all the screening runs and the stable conditions as given in Table 2. The calculated wear coefficient values are given in Table 3. It is worth mentioning that this calculation is done here for the first time in diamond coated mechanical seal rings. The wear coefficient values obtained for biodiesel lubrication in reciprocating sliding ball-on-flat experiments performed in an earlier work were k ¼1.5  10  7 mm3/N.m for monolayer MCD and k ¼5  10  8 mm3/N.m for multilayer coatings [16]. The seal rings have much lower wear rates, by about two to three orders of magnitude which can be explained by the difference in the linear speeds and contact pressures:

i) speeds are in the 0.5–1.2 m.s  1 range (Table 2) in the seal ring characterization while for the reciprocating tests they are in the 0.012–0.016 m.s  1 range [16]; ii) the apparent contact pressures calculated taking into account the real width of the wear track fall in the range 1.6 – 9.5 MPa while the apparent contact pressures in the reciprocating tests go from the initial Hertzian contact stress values (3  5.8 GPa) down to 0.2 – 1.2 GPa, when considering the final diameter of the wear scars of the balls [16]. In the ring-on-ring tests, the combined effect of a much larger speed with a much lower contact pressure, translates into a V/P ratio of the Stribeck parameter that is at least five orders of magnitude higher than that of the reciprocating ball-on-flat tests. As a consequence, during the operation of the mechanical seal rings system, lubrication is effective for preserving the internal part of the rings while in the ball-on-flat reciprocating tests, boundary lubrication or even no lubrication conditions are established, increasing the wear coefficient. Regarding the friction behaviour, during the screening stage of ring-on-ring testing, in each individual run the typical evolution of the coefficient of friction (COF) can be seen in the graphs of Fig. 8. These curves illustrate that there is an initial maximum peak coming from the interlocking and fragmentation of contacting asperities of the pristine diamond surfaces. After that, there is a steep decrease and the stationary regime is rapidly reached, where the self–polishing action described above dominates the friction behaviour. The COF values at the steady state regime are slightly lower for the monolayer MCD coated rings (  0.07, Fig. 8a) than for the multilayer coated ones ( 0.12, Fig. 8b). This behaviour was also observed in reciprocating ball-on-flat tests being attributed to the lower ability of the NCD top surface to retain the lubricant, unlike the monolayer MCD where it may remain in the valleys between the polished microcrystalline diamond grains [16]. Kelly et al. argued that the micro-asperities can generate hydrodynamic lift and thus there is a lubricant film between the bearing surfaces supporting the load [17].

Fig. 6. 3D image and topographic profile of the wear track of a monolayer MCD coated ring after testing under biodiesel lubrication.

182

M. Shabani et al. / Wear 384-385 (2017) 178–184

Fig. 7. 3D image and topographic profile of the wear track of the multilayer coated ring after testing under biodiesel lubrication. Table 3 Wear coefficient values of the stationary rings tested under biodiesel lubrication and in the sealing of pressurized water. Coating

Fluid

k (mm3/N.m)

Monolayer MCD Multilayer MCD/NCD Multilayer MCD/NCD

Biodiesel lubrication Biodiesel lubrication Water sealing

5.3  10  10 6.3  10  10 5.6  10  10

Despite these low values of the COF and of the wear coefficients, the biodiesel temperature increases steadily for values above 50 °C, so it was not possible to conduct sealing tests of biodiesel due to safety issues. This occurred because the tests were done under biodiesel lubrication, at ambient pressure, without cooling or circulation of the test fluid, contrarily to real operating conditions, where the fluid is continuously renewed. 3.3. Performance of diamond coated mechanical seal rings in sealing of water Ring-on-ring tests were performed in pressurized water lubrication to assess the sealing performance of the monolayer microcrystalline diamond and of the micro/nanocrystalline multilayer diamond coated Si3N4 mechanical seal rings. In a first set of experiments with the objective of determining the minimum P  V sealing conditions, the angular speed was kept constant at

250 rpm (V¼0.5 m s–1) and increasing nominal pressures were applied, Table 2. For the monolayer MCD coated pair of seal rings, the minimum P  V value for sealing is 0.84 MPa ms–1 while for the multilayer MCD/NCD coated pair, the value is slightly lower, P  V ¼0.75 MPa ms–1. The evolution of the friction coefficient of runs performed under these sealing conditions is similar to the observed under biodiesel lubrication, Fig. 8, but with pressurized water the steady state COF values in both tribosystems are smaller (r0.05) than those reached under biodiesel lubrication. The determination of the P  V limit condition is the most important aspect of the sealing systems as they relate directly to practical applications. For this, tests were performed at the maximum angular speed of the tribometer (2000 RPM), corresponding to a linear velocity of V ¼3.9 m s–1. For the monolayer MCD coatings at the lower load used (350 N), corresponding to P  V ¼2.2 MPa ms–1, vibrations were felt on the sealing system, and post-test analysis confirmed the delamination and spalling–off the diamond coating, as is clear from the SEM micrograph in Fig. 9a. The breakdown of the diamond coating in mechanical seals is attributed to the seizure of diamond–coated seal faces when boundary lubrication conditions are established and solid-solid contact takes place [18–22]. Hollman et al. reported partial spallation of a self-mated MCD diamond coated system after 75 km of running, although without water leakage, under the P  V conditions of the tests (8.6 MPa ms  1) [23]. For the multilayer MCD/NCD coatings the P  V limit condition

Fig. 8. Typical evolution of the coefficient of friction (COF) versus sliding distance (L) of seal rings under biodiesel lubrication coated with: a) monolayer MCD diamond; b) multilayer MCD/NCD diamond.

M. Shabani et al. / Wear 384-385 (2017) 178–184

183

Fig. 9. Microstructural and topographic features of rings after testing under 2 bar pressurized water above the P  V maximum conditions: a) monolayer MCD coated rotational ring; b, c, d) multilayer MCD/NCD coated stationary ring.

was achieved at a much higher load of 850 N, corresponding to P  V ¼5.5 MPa ms–1. At this P  V, the coefficient of friction starts showing instabilities and increases to values higher than those of the minimum P  V sealing condition. After testing it was possible to infer the wear coefficient of the multilayer coated rings, as done for the biodiesel lubrication, from topographic images similar to those of Fig. 9b. The result, k¼ 5.6  10  10 mm3/N.m in Table 3, obtained after almost 300 km of sliding, confirms the extremely low wear rates of this tribological system. Also, the k value estimate in water sealing is two orders of magnitude lower than the values found in the literature for ball-on-flat reciprocating experiments [24]. Lubrication is thus effective, unlike in the ball-onflat contacts where boundary lubrication or even no lubrication conditions are established. The better mechanical behaviour of the multilayer coatings is due to the action of the MCD/NCD interfaces in deflecting cracks, thus acting as ‘‘energy sinks’’ to further propagation [25]. For values above the P  V limit condition the coating becomes mechanically deteriorated as observed for the monolayer MCD coated rings, but with important differences. The wear depth crosses more than one layer of diamond, since it can be as large as 2 mm, Fig. 9b. Full delamination of the multilayer coating from the substrate does now occur catastrophically as for the MCD monolayer, as can be seen from the SEM micrograph in Fig. 9c and the topographic image of Fig. 9d. The coating fails across interfaces between MCD and NCD rather than directly from the substrate, wearing first the top layers. This is a distinctive advantage in real life applications where absence of lubrication or extemporaneous high loads could result in mechanical damage to such coatings. Both the minimum P  V sealing and the P  V limit conditions obtained in this work are very similar to the results of other studies with CVD diamond coated seal rings but produced by microwave assisted CVD (MPCVD) [18,19]. In such works, the full

sealing P  V range was 0.8–3.1 MPa ms  1, for a MCD monolayer [18] and 0.5–4.8 MPa ms  1 for nanocrystalline diamond [19]. An anticipated advantage of the present multilayer system, with a full sealing P  V range of 0.75–5.5 MPa ms  1, is the structural discontinuity at the interfaces between the two types of diamond to prevent full delamination of the coating in the cases where increased friction coefficients may occur in service. In such an occurrence, only the top layer would delaminate, thus preserving the integrity of the diamond/substrate interface that ultimately determines the component life. For engineering purposes, since self-polishing of diamond leads to complete flattening of the surface, the seal rings will greatly benefit from texturing of the substrate surface prior to coating [26], in order to guarantee the beneficial effects of lubricant pocketing and debris collection [27]. Also, the edge effect will need to be eliminated either at the growth stage or by post-polishing the coated seal rings, so that the entire ring area is effectively in contact with its counter face.

4. Conclusions For multilayer MCD/NCD coated mechanical seal rings tested under biodiesel lubrication the surface roughness endures a larger reduction of Sq to values equivalent to those of polished surfaces ( 20 nm). The seal rings present a wear coefficient value (k ¼6.3  10  10 mm3/N.m) that is lower by two orders of magnitude when comparing to the values obtained in reciprocating sliding ball-on-flat experiments. Under biodiesel lubrication the coefficient of friction of the multilayer coating, COF 0.12, is slightly higher than the one measured for the monolayer MCD seal ring pairs ( 0.07). This is probably due to the lower ability of the NCD top surface to retain

184

M. Shabani et al. / Wear 384-385 (2017) 178–184

the lubricant, unlike in MCD where it may remain in the valleys between the polished microcrystalline diamond grains. For both types of coatings, each individual run of the total sliding distance had to be stopped due to the increase of the biodiesel temperature to values above 50 °C. The multilayer MCD/NCD coated tribosystem allowed full sealing of pressurized water in the P  V range of 0.75–5.5 MPa ms  1, contrarily to the monolayer MCD seal rings that fail before sealing conditions are reached. The k value estimated in water sealing (5.6  10  10 mm3/N.m) is similar to the value obtained under biodiesel lubrication. An anticipated advantage of the proposed multilayer system are the structural discontinuities at the interfaces between the MCD and NCD layers that facilitate a layer-by-layer failure mode rather than full delamination of the coating.

[8]

[9]

[10]

[11]

[12]

[13]

[14]

Acknowledgments [15]

This work was developed within the scope of project POCI-01– 0247-FEDER-006318 co-financed by FEDER through the POCI program and project CICECO-Aveiro Institute of Materials, POCI01–0145-FEDER-007679 (FCT Ref. UID/CTM/50011/2013), financed by national funds through the FCT/MEC and when appropriate cofinanced by FEDER under the PT2020 Partnership Agreement.

References [1] S. Tong Lee, Z. Lin, X. Jiang, CVD diamond films: nucleation and growth, Mater. Sci. Eng. R: Rep. 25 (4) (1999) 123–154. [2] P.W. May, Diamond thin films: a 21st century material, Philosophical Transactions of the Royal Society of London, Series A: mathematical, Phys. Eng. Sci. 358 (1766) (2000) 473–495. [3] M. Vojs, M. Vesely, R. Redhammer, J. Janik, M. Kadlecikova, T. Danis, M. Marton, M. Michalka, P. Sutta, Double bias HF CVD multilayer diamond films on WC–Co cutting tools, Diam. Relat. Mater. 14 (2005) 613–616. [4] H.J. Lee, H. Li, H. Jeon, W.S. Lee, Some novel aspects of nanocrystalline diamond nucleation and growth by direct current plasma assisted chemical vapor deposition, Diam. Relat. Mater. 19 (11) (2010) 1393–1400. [5] D. Zhou, D.M. Gruen, L.C. Qin, T.G. McCauley, A.R. Krauss, Control of diamond film microstructure by Ar additions to CH4/H2 microwave plasmas, J. Appl. Phys. 84 (4) (1998) 1981–1989. [6] A. Erdemir, G.R. Fenske, A.R. Krauss, D.M. Gruen, T. McCauley, R.T. Csencsits, Tribological properties of nanocrystalline diamond films, Surf. Coat. Technol. 120–121 (1999) 565–572. [7] A.R. Krauss, O. Auciello, D.M. Gruen, A. Jayatissa, A. Sumant, J. Tucek, D. C. Mancini, N. Moldovan, A. Erdemir, D. Ersoy, M.N. Gardos, H.G. Busmann, E. M. Meyer, M.Q. Ding, Ultrananocrystalline diamond thin films for MEMS and

[16]

[17] [18]

[19]

[20]

[21] [22]

[23] [24]

[25]

[26] [27]

moving mechanical assembly devices, Diam. Relat. Mater. 10 (2001) 1952–1961. S. Takeuchi, S. Oda, M. Murakawa, Synthesis of multilayer diamond film and evaluation of its mechanical properties, Thin Solid Films 398–399 (2001) 238–243. X. Lu, Q. Yang, C. Xiao, Synthesis and field electron emission characteristics of diamond multilayer films grown by graphite etching, J. Phys. D: Appl. Phys. 40 (2007) 4010–4014. J.L. Luo, X.T. Ying, P.N. Wang, L.Y. Chen, Growth and optical properties of nanocrystalline/microcrystalline diamond multilayer film, J. Korean Phys. Soc. 46 (2005) 224–228. ASTM D3702 - 94, Standard Test Method for Wear Rate and Coefficient of Friction of Materials in Self-Lubricated Rubbing Contact Using a Thrust Washer Testing Machine, 2014. F.A. Almeida, M. Amaral, F.J. Oliveira, A.J.S. Fernandes, R.F. Silva, Nano to micrometric HFCVD diamond adhesion strength to Si3N4, Vacuum 81 (2007) 1443–1447. F.A. Almeida, M. Amaral, F.J. Oliveira, R.F. Silva, Machining behavior of silicon nitride tools coated with micro–, submicro– and nanometric HFCVD diamond crystallite sizes, Diam. Relat. Mater. 15 (2006) 2029–2034. M. Amaral, A.J.S. Fernandes, M. Vila, F.J. Oliveira, R.F. Silva, Growth rate improvements in the hot-filament CVD deposition of nanocrystalline diamond, Diam. Relat. Mater. 15 (2006) 1822–1827. C.S. Abreu, E. Salgueiredo, F.J. Oliveira, A.J.S. Fernandes, R.F. Silva, J.R. Gomes, CVD diamond water lubricated tribosystems for high load planar sliding, Wear 265 (2008) 1023–1028. F.A. Almeida, M.M. Maru, M. Shabani, F.J. Oliveira, R.F. Silva, C.A. Achete, Enhancing the tribological performance under biodiesel lubrication using CVD diamond coated parts, Wear 302 (2013) 1370–1377. P.J. Kelly, R.D. Arnell, M.D. Hudson, A.E.J. Wilson, G. Jones, Enhanced mechanical seal performance through CVD diamond, Vacuum 61 (2001) 61–74. M.A. Tome, A.J.S. Fernandes, F.J. Oliveira, R.F. Silva, J.M. Carrapichano, High performance sealing with CVD diamond self–mated rings, Diam. Relat. Mater. 14 (2005) 617–621. F. Mubarok, J.M. Carrapichano, F.A. Almeida, A.J.S. Fernandes, R.F. Silva, Enhanced sealing performance with CVD nanocrystalline diamond films in self– mated mechanical seals, Diam. Relat. Mater. 17 (2008) 1132–1136. A.V. Sumant, A.R. Krauss, D.M. Gruen, O. Auciello, Ultrananocrystalline diamond film as a wear resistant and protective coating for mechanical seal applications, Tribol. Trans. 48 (2005) 24–31. G.A. Jones, On the tribological behavior of mechanical seal face materials in dry line contact Part I. mechanical carbon, Wear 256 (2004) 415–432. G.A. Jones, On the tribological behavior of mechanical seal face materials in dry line contact Part II: bulk ceramics, diamond and diamond–like carbon films, Wear 256 (2004) 433–455. P. Hollman, H. Björkman, A. Alahelisten, S. Hogmark, Diamond coatings applied to mechanical face seals, Surf. Coat. Technol. 105 (1998) 169–174. C.S. Abreu, M. Amaral, F.J. Oliveira, J.R. Gomes, R.F. Silva, HFCVD nanocrystalline diamond coatings for tribo–applications in the presence of water, Diam. Relat. Mater. 18 (2009) 271–275. E. Salgueiredo, F.A. Almeida, M. Amaral, M.A. Neto, F.J. Oliveira, R.F. Silva, A multilayer approach for enhancing the erosive wear resistance of CVD diamond coatings, Wear 297 (2013) 1064–1073. I. Etsion, Y. Kligerman, G. Halperin, Analytical and Experimental Investigation of Laser-Textured Mechanical Seal Faces, Tribol. Trans. 42 (1999) 511–516. I. Etsion, State of the Art in Laser Surface Texturing, J. Tribol. 127 (2005) 248–253.