- Email: [email protected]
Formation of carbon composite coatings by plasma spraying
Vacuum xxx (2015) 1e6
Contents lists available at ScienceDirect
Vacuum journal homepage: www.elsevier.com/locate/vacuum
Formation of carbon composite coatings by plasma spraying nas Kavaliauskas a, Romualdas Ke_ Liutauras Marcinauskas a, b, *, Zydr u zelis a a b
Lithuanian Energy Institute, Plasma Processing Laboratory, Breslaujos str. 3, LT-44403 Kaunas, Lithuania Department of Physics, Kaunas University of Technology, Studentu˛ str. 50, LT-51368 Kaunas, Lithuania
a r t i c l e i n f o
a b s t r a c t
Article history: Received 3 November 2014 Received in revised form 15 April 2015 Accepted 17 April 2015 Available online xxx
Carbonesiliconesulfur composite coatings were deposited on stainless steel by atmospheric plasma spraying. The effect of spraying power on the coating microstructure and microhardness was investigated. The surface morphology and elemental composition of as-sprayed coatings were examined by scanning electron microscopy and energy dispersive X-ray spectroscopy. The increase of torch power leads to a lower oxidation degree and reduces the silicon concentration in the coatings. The SEM crosssection measurements indicated that the thicknesses of the coatings were in the range of 10e15 mm. The X-ray diffraction measurements indicated an increase of crystalline graphite phase with the increase of power. The results indicated that the coating roughness decreased, while the microhardness and specific surface area increased with increased power. The highest microhardness value of 314 HV was obtained. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Plasma spraying Microhardness Graphite Silicon Surface morphology
1. Introduction Recently much attention has been focused on the deposition and investigation properties of various carbon composites or carbon containing coatings. The unique properties such as chemical inertness, low friction coefficient, high hardness, high specific surface area of carbon composites allows use of these coatings in tribological applications, or, in the energy field as energy storage devices [1e4]. Sulfur-doped, carbons are attractive functional materials for application in fuel cells, energy conversion/storage devices and as an anode materials in Li-ion batteries [4e6]. A number of carbonesulfur nanocomposites for LieS cathodes have been proposed [4,6]. It was demonstrated that the C/S composites may contain up to 90 wt.% of sulfur depending on the method of preparation [4]. Meanwhile, the addition of silicon carbide can increase the fracture toughness and wear resistance of metal oxides or metal coatings or change the thermal conductivity of these coatings [7e11]. Plasma spraying is a very attractive technique for the formation of the carbon-containing composites using various material powders including even high melting temperature materials such as
* Corresponding author. Lithuanian Energy Institute, Plasma Processing Laboratory, Breslaujos str. 3, LT-44403 Kaunas, Lithuania. Tel.: þ370 86111703; fax: þ370 37 351271. E-mail addresses: [email protected], [email protected] (L. Marcinauskas).
graphite, tungsten, silicon carbide, titanium oxide and so on [9e15]. The surface morphologies and properties of the plasma sprayed coatings depend on the process parameters, including torch power, powder feed rate, gas flow rate, spraying distance and particle size [10e21]. The torch power is a crucial parameter, because it influences the velocity and temperature of the plasma flow and, in consequence, powder particles directly. The deposition efficiency, interface bonding and mechanical properties of sprayed coatings increases with the increase of particle velocity and temperature [18e25]. It was demonstrated that the as-sprayed coatings show elemental composition different from the initial feedstock powders [7,8,13]. Variation of plasma gun power allows changing and controlling the plasma jet temperature and, consequently, to adjust the melting degree of the feedstock powders [14,25]. Therefore, it is generally considered that the fully melted particles will lead to production of dense with low pore volume coatings [14]. The nature and elemental composition of the initial powders are directly related to the melting temperature [7,13,14]. Therefore it is very important to obtain the optimal spraying process parameters for individual feedstock powders. The investigations concerning the formation of carbonesiliconesulfur composite coatings from flakelike shape powders by plasma spraying is hard to find in the scientific literature. As a result it is very important to determine appropriate process parameters which will lead to the deposition of coatings with dense structure and high hardness values. In this study, the effect of torch input power on the surface morphologies, roughness, pore size, pore distribution, and
http://dx.doi.org/10.1016/j.vacuum.2015.04.028 0042-207X/© 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Marcinauskas L, et al., Formation of carbon composite coatings by plasma spraying, Vacuum (2015), http:// dx.doi.org/10.1016/j.vacuum.2015.04.028
2
L. Marcinauskas et al. / Vacuum xxx (2015) 1e6
microhardness of sprayed carbonesiliconesulfur composite coatings was investigated. 2. Experimental details The coatings were deposited on steel substrates at atmospheric pressure using a direct current plasma torch. The plasma torch used in this experiment was developed at the Lithuanian Energy Institute [26e28]. Substrates with the dimensions of 10 40 1 mm were made from stainless steel. The steel substrates were polished and chemically cleaned by acetone before starting the deposition process. The steel substrates were placed on the water-cooled sample holder. The coatings were sprayed at a torch power of 17.1 kW, 19.4 kW, and 21.0 kW, respectively. The spraying distance was kept at 35 mm. Argon was used as both the primary gas (flow rate of 7.5 g/s) and the powder carrier gas (flow rate of 0.59 g/s). The experiments were performed using a cylindrical shape (0.15 m of length) reactor with the powder injection place (located 0.10 m from the exit) which was situated at the exit of the anode. The mean temperature of the plasma jet at the exhaust of the plasma torch was calculated from the heat balance corresponding to plasma enthalpy:
Hf ¼ M
ðPT Pw Þ þ Ho G
(1)
where Hf is the plasma enthalpy (kJ mol1), PT is torch power (kW), PW is power used for water heating (kW), G is plasma mass flow rate (kg s1), M argon gas atomic mass (kg mol1), and H0 is the enthalpy under standard conditions (kJ mol1). The mean velocity of the plasma jet at the exhaust of the plasma torch was calculated from Eq. (2):
v¼
4MGT pd2 p
(2)
where T is mean temperature of the plasma (K), d e diameter of the exhaust nozzle (m) and p e pressure (m). The average temperature and velocity of the plasma flow at the different torch powers is given in Table 1. Detailed description of the methodology of the plasma temperature and velocity calculation is given in Ref. [29]. The graphite powders with the elemental composition of carbon (C) e 95.9 at%, oxygen (O) e 2.9 at%, silicon (Si) e 1.0 at% and sulfur (S) e 0.2 at% were used as sprayed powders. The graphite powders used in the spraying process were of a non-regular flake-like shape with various sizes from 20 mm to 150 mm, and thickness of ~10 mm (Fig. 1a). The specific surface area of these powders was 1.84 m2/g. The feedstock powders were grinded and sieved using a 150 mm size grid and were dried for 8 h at ~350 K before starting the deposition process. The feedstock powders were injected into the reactor nozzle at a distance of 100 mm from the exit. The deposition duration was 60 s. The plasma torch was moving in the x-axis direction forward and back during the deposition. The surface morphology of the sprayed coatings and feedstock powders was analyzed by scanning electron microscopy (SEM) using a JEOL JSMe5600. The elemental composition of deposited coatings and powders was obtained by energy dispersive X-ray
Table 1 Plasma spraying parameters for depositing composite coatings. Power, kW
I, A
U, V
Velocity, m/s
Temperature, K
17.1 19.4 22.0
180 200 220
95 97 100
540 ± 15 600 ± 15 670 ± 15
3180 ± 50 3520 ± 50 3940 ± 50
spectroscopy (EDS). The surface roughness measurements were investigated using a stylus profiler (Ambios XP-200). The measurements were done 5 times; the length of one measurement was 2 mm. The coating structure was analyzed by X-ray diffraction (XRD) (DRON-UM1 with standard BraggeBrentano focusing geometry) in a 10e100 range using the CuKa (l ¼ 0.154059 nm) radiation. The bonding structure of sprayed coatings was investigated by Raman scattering (RS) spectroscopy (“Jobin Yvon” company). RS measurements were done using a Nd:YAG laser (532.3 nm, 50 mW, spot size 0.32 mm, in the 1000e2000 cm1 range) as an excitation source. The RS spectra were fitted by Gaussian-shape lines in the spectral range of (1200e1800) cm1. The specific surface area of the coatings was measured by the BETmethod in a KELVIN 1042 sorptometer. The microhardness measurements were done using a CSM Micro Scratch Tester with Vickers indenter. The microhardness was measured on the polished coating surface at room temperature by applying a load of 10 mN. The speed of the loading and unloading was 200 mN/min and the pause between the loading and unloading rate was 10 s. 10e12 hardness measurements were taken randomly and the arithmetic mean of measurements was taken as the microhardness of the coatings. Beside the hardness values the elastic deformation work of indentation (We), plastic deformation work of indentation (Wp), and total mechanical work of indentation (Wt) were given from the force e penetration depth curves for loading and unloading. The elasticity (elastic part of indentation work) was calculated using Eq. (3) [30]:
ε¼
We $100% ¼ Wt
1
Wp $100% Wt
(3)
3. Results and discussion Fig. 1 shows the surface morphologies of plasma-deposited composite coatings at different torch powers. The surface of a coating formed at the lowest power is composed from partially melted and solidified flake-like particles (Fig. 1b and e). The number of partially molten particles in the coating decreases with increasing torch power. The ball-like shape particles with size of 1e10 mm appear on the surface at higher power (Fig. 1c, d and f). The appearance of pores and non-regular shape particles could be found in all deposited coatings. The existence of sphere-like shape particles on the surface indicates the higher melting degree of the initial feedstock powders. It is well known that pores in sprayed coatings are generally derived from poorly stacked flat particles and that the state of the flat particles is highly dependent on the melted extent of the feedstock powders and the velocity of the plasma jet [21]. The SEM cross-section measurements indicated that the thicknesses of the coatings were in the range of 10e15 mm. The EDS measurements demonstrated that the oxygen content in the assprayed coatings decreased from 44.4 at% to 41.8 at% with increasing power from 17.1 kW to 22.0 kW. The carbon content increased from 31.8 at% to 38.5 at%. It should be noted that, despite the low concentration of Si and S in the initial powders, the sprayed coatings contained quite a high fraction of these elements. The silicon concentration decreased from ~16 at% to ~12 at%, while the sulfur content did not change (~8 at%) in coatings with the increased power values. In order to evaluate the torch power influence on the surface morphologies surface roughness measurements were performed. The surface roughness investigations indicated (see Fig. 2) that coatings sprayed at the lowest power had the highest surface roughness (Ra ¼ 700 nm, Rq ¼ 1000 nm). The surface roughness
Please cite this article in press as: Marcinauskas L, et al., Formation of carbon composite coatings by plasma spraying, Vacuum (2015), http:// dx.doi.org/10.1016/j.vacuum.2015.04.028
L. Marcinauskas et al. / Vacuum xxx (2015) 1e6
3
Fig. 1. Surface micrographs of powders (a) and as-sprayed coatings prepared at different powers: (b, e) 17.1 kW, c) 19.4 kW and (d, f) 22.0 kW.
decreased to 590 nm and 840 nm when 19.4 kW power was used. The further increase of the torch power has almost no influence on the surface roughness values (Fig. 2). As can be seen the surface roughness values and oxygen content decreased with increasing torch power, which implies that the deposition efficiency of composite coatings increased.
Fig. 2. The surface roughness of coatings deposited at different spraying power.
The existence of significant fractions of Si and S in the sprayed coatings indicates that the temperature in the plasma flow was not enough to completely melt the feedstock powders. The increase of Si and S concentrations in the coatings compared to those in the powders is related to the large difference in melting temperatures between carbon, silicon (~1800 K) and sulfur (~400 K) and to the poor wettability of the graphite. Lopez et al. [7] demonstrated that the SiC incorporation in the coating matrix with respect to that the feedstock powder was only 10%. Meanwhile the deposition efficiency of the process in terms of vol.% SiC was not higher than 20%. The increase of the oxygen in the coatings related to the fact that the vaporized carbon reacted with the entrapped or surrounding air and formed graphite oxide dust during the spray process. The lowest (~540 m/s) plasma flow velocity was obtained at 17.1 kW. Thus the particle flight duration in the argon plasma was the longest and the probability of the particle to oxidize increased. The low amount of carbon in the composites could also be due to formation of carbon dioxide (CO2) or carbon monoxide (CO). N.F Fahim [8] demonstrated that the carbon reacted with oxygen and escaped as CO2 and CO during the formation of W/SiC composites. It should be mentioned that there were many small ball-like particles with micrometer sizes on the coating surface. The existence of such structures is related to splash debris from the impacting graphite
Please cite this article in press as: Marcinauskas L, et al., Formation of carbon composite coatings by plasma spraying, Vacuum (2015), http:// dx.doi.org/10.1016/j.vacuum.2015.04.028
4
L. Marcinauskas et al. / Vacuum xxx (2015) 1e6
splats due to their low viscosity [31]. The coating obtained at a low power of 17.1 kW contains many non-regular shape fragments on the surface and the lowest carbon concentration. This is because such plasma temperature is inadequate to completely melt the feedstock powders. The large flake-like particles probably impact on the substrate in the solid state and rebound off the surface of the coatings. However the plasma temperature is quite high to completely melt silicon which is incorporated in initial graphite powders. When the torch power increases, the temperature increases and favors an increase in the degree of melting of the powders and velocity of plasma flow. As a result the amount of unmelted and partially-melted particles decreases and a higher fraction of well melted powders can easily form flat splats. In addition flat splats can integrate tightly. The lower surface roughness values obtained at higher power indicate the improvement of splat bonding. Fig. 3a shows the dependence of surface area and pore volume in plasma sprayed coatings using different torch powers. As can be seen, the specific surface area of the coatings increased from 11.8 m2/g to 14.5 m2/g as the spraying power increased from 17.1 kW to 22.0 kW. Meanwhile, the pore volume increased almost twice from ~19.6 mm3/g to ~33.4 mm3/g with the increased power. It should be noted that the pores surface area was ~4.7 m2/g, 9.4 m2/ g and 7.4 m2/g for the coatings prepared at torch power of 17.1 kW, 19.4 kW and 22.0 kW, respectively. The pore diameter influence on the surface area and pore volume values are demonstrated in Fig. 3b. The measurements demonstrated that the smaller size pores appeared on the surface with the increased input power. It
Fig. 3. The surface area and pore volume dependence on spraying power (a) and pore distribution on the surface and pore volume values (b).
should be mention that the largest pores (~25 nm) have the highest pore volumes (Fig. 3b). The specific surface area mainly is determined by the micro-pore and meso-pore quantity in the coatings. The amount of pores in the specific surface area of the coating was ~40%, ~78% and ~51% for 17.1 kW, 19.4 kW and 22.0 kW power, respectively. Thus, the decrease in the surface roughness was compensated by the increased amount of pores. P. Ctibor et al. [32] demonstrated that the higher the torch power the lower surface roughness and porosity will be. However, it will result in the production of a higher amount of pores with smaller size. The higher degree of melting and higher velocity of the powders particles lead to a more uniformly and steady distribution of the splats. Another feature of the higher torch power coated splats is that cracks and holes with lower dimensions are formed on solidification [33]. The suggestion would be that the crack and hole areas may serve as the source of meso-pores. Fig. 4 demonstrates the XRD patterns of the steel substrate and composite coatings sprayed at different temperatures. The existence of the peaks related to the steel substrate was obtained for all deposited coatings. The diffraction peaks at ~26.6 , ~44.6 , and 54.7 corresponds to (002), (100), and (004) planes of graphite, respectively [34,35]. The intensity of the peak assigned to (100) graphite increased with increased input power. The appearance of the low intensity peak at ~12.4 is attributed to (001) graphite oxide [35]. The existence of graphite oxide peak indicates that part of graphite was transformed into graphite oxide during plasma spraying. It should be mentioned that the XRD patterns do not indicate formation of SiC, or any other crystalline phase related to Si or S elements. This indicates that the silicon and sulfur are in an amorphous phase. The Raman spectra of the film sprayed at 17.1 kW power has two separated D (1361 cm1) and G (1592 cm1) peaks. The peak at ~1361 cm1 is assigned to the disordered graphite structure (Dband), while the peak at 1592 cm1 (G-band) corresponds to a splitting of the E2g stretching mode of graphite and reflects the structural intensity of the sp2-hybridized carbon atoms [35]. The relative intensity of the D-band and G-band (ID/IG) ratio is 0.25. The shapes of the spectra of coatings deposited at 19.4 kW and 22.0 kW, were almost identical as those for films deposited at 17.1 kW power (Fig. 5). The increase of power influenced the slight increase of the ID/IG ratio, the broadening of the D peak (DD), the narrowing of the G band (DG) and the shift of the D band to higher values (Fig. 5). The ID/IG ratio slightly increased from 0.25 to 0.29 with increasing torch
Fig. 4. XRD patterns of the sprayed coatings and steel substrate.
Please cite this article in press as: Marcinauskas L, et al., Formation of carbon composite coatings by plasma spraying, Vacuum (2015), http:// dx.doi.org/10.1016/j.vacuum.2015.04.028
L. Marcinauskas et al. / Vacuum xxx (2015) 1e6
Fig. 5. Raman spectra of coatings formed at different powers.
power from 17.1 kW to 22.0 kW. The higher intensity ratio of Dpeak to G-peak (ID/IG) shows the higher extent of defects in carbon coatings [35,36]. The narrowing of the G band (from 140 cm1 to 118 cm1) indicates less distortion in the carbon coatings. It was demonstrated that, by thermally oxidizing the graphite flake, the intensity of the G-peak decreases, G-band widening, and the intensity of D-band increases [36,37]. Such result could be due to higher plasma temperatures and lower oxygen content in the coating. The coating sprayed at the lowest power has the highest content of oxygen. Thus the fraction of sp2 C]O bonds in the coatings would be higher and will result in an increase in concentration of defects in the structure. The changes in microhardness values of the carbon composite coatings sprayed at different plasma torch power are shown in Fig. 6a. The lowest hardness value (~120 HV) was obtained for the lowest torch power. As the spraying power was increased to 19.4 kW, the hardness values of the sprayed coating was increased up to 243 HV and was similar to that for the steel substrate (~255 HV). The coating sprayed at the highest power demonstrated the highest microhardness values (314 HV). The elasticity measurements indicated that, despite the low hardness value of coating prepared at 17.1 kW, the elasticity was similar to that of the steel substrate (Fig. 6 b). The highest elasticity (33%) was observed for the coating deposited at 19.4 kW. The hardness values are mainly affected by the microstructure and porosity of the sprayed coatings. At low power the presence of unmelted particles and pores decreases the hardness value mainly due to weakness of cohesion bonds of accumulated splats due to lower particle temperature [23]. As the power increases the plasma flow temperature increases, powder particles get more energy from the plasma and, as a result, the majority of particles are melted and accumulate in the dense structure on the steel substrate. Other authors also indicated
5
Fig. 6. Effect of torch power on the microhardness (a) and elasticity (b) of coatings.
that the dense and low porosity coatings prepared by plasma spraying were harder [18,19,21,23]. It is well known that the microstructure of the as-sprayed coatings is strongly dependent on the interaction of impacting particles with the underlying surface [16]. The flattening degree, defined as the ratio of splat diameter to impacting droplet diameter, is one important parameter used to reflect this interaction. The flattening degree increases with the rise of particle impact velocity [14]. As the torch power increased from 17.1 kW to 22.0 kW, the plasma flow velocity increased from 540 m/s to 670 m/s. Naturally, elevating torch power favors to increase the in-flight particle velocity and herewith flattening degree. Thus, the effective bonding area between the individual splats would be enlarged and the number of pores due to the incomplete inter-splat contact would be decreased. The velocity of plasma increases with the increase in power, which results from an improved thermal pinching effect on the plasma jet [25]. Therefore, an increase in plasma velocity increases particle velocity and results in a slight decrease in dwelling time of particles in the plasma flow. The fact that the hardness values increased and surface roughness decreased can be attributed that the improvement of particle heating can compensate for the shortening of dwelling time of particles in the plasma. The decreased surface roughness values at higher powers are the result of the increased degree of smoothing and higher particle melting. 4. Conclusions Carbonesiliconesulfur composite coatings were prepared using atmospheric pressure plasma spraying at a various torch powers from the flake-like graphite powders. As the plasma flow temperature raises the number of larger particles decreases in the coatings
Please cite this article in press as: Marcinauskas L, et al., Formation of carbon composite coatings by plasma spraying, Vacuum (2015), http:// dx.doi.org/10.1016/j.vacuum.2015.04.028
6
L. Marcinauskas et al. / Vacuum xxx (2015) 1e6
surfaces. The surface roughness values and coatings oxidation degree depend on the torch power values. The surface roughness values decreases with the increase of torch power. The oxidation degree of as-sprayed coatings becomes lower with the increase of plasma flow velocity (torch power). The specific surface area and pore volume of coatings increases as the torch power increases. The increase in plasma temperature and velocity lead to an increase in crystalline graphite phase. As the power increases from 17.1 kW to 22.0 kW, the microhardness values of sprayed composite coatings increases from 120 HV to 314 HV. The increase of the microhardness attributed to an increase of crystalline phase and improvement of bonding between the individual splats in the structure of the composite. The highest elasticity was estimated for the coating prepared at 19.4 kW. Acknowledgments The authors thank Dr. D. Mil cius from the Lithuanian Energy Institute for possibility to perform SEM, EDS and surface roughness measurements and Dr. A. Zunda from Aleksandras Stulginskis University for microhardness measurements. References [1] Pandolfo A, Hollenkamp A. Carbon properties and their role in super capacitors. J Power Sources 2006;157:11e27. [2] Zhu T, Li Y, Luo M, Sang S, Wang Q, Zhao L, et al. Microstructure and mechanical properties of MgOeC refractories containing graphite oxide nanosheets (GONs). Ceram Int 2013;39:3017e25. [3] Mu M, Zhou X, Xiao Q, Liang J, Huo X. Preparation and tribological properties of self-lubricating TiO2/graphite composite coating on Ti6Al4V alloy. Appl Surf Sci 2012;258:8570e6. [4] Kicinski W, Szala M, Bystrzejewski M. Sulfur-doped porous carbons: synthesis and applications. Carbon 2014;68:1e32. [5] Yang Z, Nie H, Chen X, Chen X, Huang S. Recent progress in doped carbon nanomaterials as effective cathode catalysts for fuel cell oxygen reduction reaction. J Power Sources 2013;236:238e49. [6] Yang XB, Zhu W, Qin K, Wang HY. Preparation of lamellar carbon matrix for sulfur as cathode material of lithium-sulfur batteries. Electrochim Acta 2014;143:374e82. pez AJ, Torres B, Taltavull C, Rams J. Influence of high velocity oxygen-fuel [7] Lo spraying parameters on the wear resistance of AleSiC composite coatings deposited on ZE41A magnesium alloy. Mater Des 2013;43:144e52. [8] Fahim NF. Thermal conductivity of plasma-sprayed W/SiC composite for hightemperature energy applications. Surf Coat Techn 2008;202:1696e703. [9] Rammo NN, Al-Amery HR, Abdul-Jabbar T, Jaffer HI. Adhesion, hardness and structure of thermal sprayed Al/SiC composite coat on graphite. Surf Coat Tech 2009;203:2891e5. [10] Pan ZY, Wang Y, Li XW, Wang CH. Fabrication and characterization of heat and plasma treated SiC/Al2O3eYSZ feedstocks used for plasma spraying. Vacuum 2012;86:1558e67. [11] Sathiyamoorthy R, Shanmugam K, Balasubramanian V. Dry sliding wear behavior of SiC reinforced titania coating deposited by high velocity oxy fuel spraying. Proc Mater Scien 2014;5:648e55. [12] Chadami F, Ghadami S, Abdollah-Pour H. Structural and oxidation behavior of atmospheric heat treated plasma sprayed WCeCo coatings. Vacuum 2013;94: 64e8. [13] Tului M, Giambi B, Lionetti S, Pulci G, Sarasini F, Valente T. Silicon carbide based plasma sprayed coatings. Surf Coat Tech 2012;207:182e9.
[14] Fauchais P. Understanding plasma spraying. J Phys D Appl Phys 2004;37: 86e108. [15] Cetinkaya S, Eroglu S. Effect of temperature and Ar flow rate on Ti3SiC2 formation from TiSiO4 powders coated with pyrolytic carbon. Int J Refract Met H 2014;43:69e76. [16] Brinkiene K, Kezelis R. Effect of alumina addition on the microstructure of plasma sprayed YSZ. J Eur Ceram Soc 2005;25:2181e4. [17] Niu Y, Zheng W, Liu X, Ji H, Ding C. Influence of powder size on characteristics of air plasma sprayed silicon coatings. Ceram Int 2012;38:5897e905. [18] Wu H, Li H, Lei Q, Fu Q, Ma C, Yao D, et al. Effect of spraying power on microstructure and bonding strength of MoSi2-based coatings prepared by supersonic plasma spraying. Appl Surf Sci 2011;257:5566e70. [19] Wang C, Fan L, Fan J, Zhang D, Wang H. Effect of spraying power on microstructure and properties of supersonic plasma sprayed Al2O3 coating on porous Si3N4 substrate. J Alloys Compd 2013;559:152e7. [20] Marcinauskas L, Valatkevi cius P. The effect of plasma torch power on the microstructure and phase composition of alumina coatings. Mater Sci Pol 2010;28:451e8. [21] An Y, Hou G, Chen J, Zhao X, Liu G, Zhou H, et al. Microstructure and tribological properties of iron-based metallic glass coatings prepared by atmospheric plasma spraying. Vacuum 2014;107:132e40. [22] Morks MF, Kobayashi A. Effect of gun current on the microstructure and crystallinity of plasma sprayed hydroxyapatite coatings. Appl Surf Sci 2007;253:7136e42. [23] Morks MF, Gao Y, Fahim NF, Yingqing FU. Microstructure and hardness properties of cermet coating sprayed by low power plasma. Mater Lett 2006;8:1049e53. [24] Tahara H, Ando Y. Study of titanium nitride deposition by supersonic plasma spraying. Vacuum 2008;83:98e101. [25] Li C, Sun B. Effects of spray parameters on the microstructure and property of Al2O3 coatings sprayed by a low power plasma torch with a novel hollow cathode. Thin Solid Films 2004;450:282e9. esnien _ [26] Brinkiene_ K, C e_ J, Ke_ zelis R, Me_ cius V, Baltusnikas A, Zunda A. Structural and surface analysis of plasma processed zirconia coatings. Mater Sci-Medzg 2006;12:300e4. [27] Ke_ zelis R, Me_ cius V, Pranevi cius LL. High velocity plasma torch design and its flow parameters for ceramics thin film deposition. In: Gylys J, Sven cianas P, Ziug zda V, editors. Heat energy and technologies; 2002. p. 253e8. Kaunas. [28] Brinkiene_ K, Ke_ zelis R, Me cius V, Pranevi cius L. Effect of outlet nozzle design on the YSZ particle in-flight characteristics in plasma jet generated by DC plasma torch. In: Sunden B, Vilemas J, editors. Advances in heat transfer engineering; 2003. p. 629e36. New York. te_ V, Marcinauskas L. [29] Valin cius V, Krusinskaite_ V, Valatkevi cius P, Valin ciu Electrical and thermal characteristics of the linear, sectional dc plasma generator. Plasma Sources Sci Technol 2004;13:199e206. [30] Giannakopoulos AE, Suresh S. Determination of elastoplastic properties by instrumented sharp indentation. Scr Mater 1999;40:1191e8. [31] Zhang C, Guo RQ, Yang Y, Wu Y, Liu L. Influence of the size of spraying powders on the microstructure and corrosion resistance of Fe-based amorphous coating. Electrochim Acta 2011;56:6380e8. [32] Ctibor P, Hrabovvsky M. Plasma sprayed TiO2: the influence of power of an electric supply on particle parameters in the flight and character of sprayed coating. J Eur Ceram Soc 2010;30:3131e6. [33] Yugeswaran S, Selvarajan V, Seo D, Ogawa K. Effect of critical plasma spray parameter on properties of hollow cathode plasma sprayed alumina coatings. Surf Coat Tech 2008;203:129e36. [34] Channu VS, Bobba R, Holze R. Graphite and graphene oxide electrodes for lithium ion batteries. Colloids Surfaces A Physicohem Eng Aspects 2013;436: 245e51. [35] Geng L, Wu S, Zou Y, Jia M, Zhang W, Yan W, et al. Correlation between the microstructures of graphite oxides and their catalytic behaviors in air oxidation of benzyl alcohol. J Colloid Interf Sci 2014;421:71e7. [36] Matsuo Y, Ueda K. Pyrolytic carbon from graphite oxide as a negative electrode of sodium-ion battery. J Power Sources 2014;263:158e62. [37] Li Y, Wang Q, Fan H, Sang S, Li Y, Zhao L. Synthesis of silicon carbide whiskers using reactive graphite as template. Ceram Int 2014;40:1481e8.
Please cite this article in press as: Marcinauskas L, et al., Formation of carbon composite coatings by plasma spraying, Vacuum (2015), http:// dx.doi.org/10.1016/j.vacuum.2015.04.028
Contents lists available at ScienceDirect
Vacuum journal homepage: www.elsevier.com/locate/vacuum
Formation of carbon composite coatings by plasma spraying nas Kavaliauskas a, Romualdas Ke_ Liutauras Marcinauskas a, b, *, Zydr u zelis a a b
Lithuanian Energy Institute, Plasma Processing Laboratory, Breslaujos str. 3, LT-44403 Kaunas, Lithuania Department of Physics, Kaunas University of Technology, Studentu˛ str. 50, LT-51368 Kaunas, Lithuania
a r t i c l e i n f o
a b s t r a c t
Article history: Received 3 November 2014 Received in revised form 15 April 2015 Accepted 17 April 2015 Available online xxx
Carbonesiliconesulfur composite coatings were deposited on stainless steel by atmospheric plasma spraying. The effect of spraying power on the coating microstructure and microhardness was investigated. The surface morphology and elemental composition of as-sprayed coatings were examined by scanning electron microscopy and energy dispersive X-ray spectroscopy. The increase of torch power leads to a lower oxidation degree and reduces the silicon concentration in the coatings. The SEM crosssection measurements indicated that the thicknesses of the coatings were in the range of 10e15 mm. The X-ray diffraction measurements indicated an increase of crystalline graphite phase with the increase of power. The results indicated that the coating roughness decreased, while the microhardness and specific surface area increased with increased power. The highest microhardness value of 314 HV was obtained. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Plasma spraying Microhardness Graphite Silicon Surface morphology
1. Introduction Recently much attention has been focused on the deposition and investigation properties of various carbon composites or carbon containing coatings. The unique properties such as chemical inertness, low friction coefficient, high hardness, high specific surface area of carbon composites allows use of these coatings in tribological applications, or, in the energy field as energy storage devices [1e4]. Sulfur-doped, carbons are attractive functional materials for application in fuel cells, energy conversion/storage devices and as an anode materials in Li-ion batteries [4e6]. A number of carbonesulfur nanocomposites for LieS cathodes have been proposed [4,6]. It was demonstrated that the C/S composites may contain up to 90 wt.% of sulfur depending on the method of preparation [4]. Meanwhile, the addition of silicon carbide can increase the fracture toughness and wear resistance of metal oxides or metal coatings or change the thermal conductivity of these coatings [7e11]. Plasma spraying is a very attractive technique for the formation of the carbon-containing composites using various material powders including even high melting temperature materials such as
* Corresponding author. Lithuanian Energy Institute, Plasma Processing Laboratory, Breslaujos str. 3, LT-44403 Kaunas, Lithuania. Tel.: þ370 86111703; fax: þ370 37 351271. E-mail addresses: [email protected], [email protected] (L. Marcinauskas).
graphite, tungsten, silicon carbide, titanium oxide and so on [9e15]. The surface morphologies and properties of the plasma sprayed coatings depend on the process parameters, including torch power, powder feed rate, gas flow rate, spraying distance and particle size [10e21]. The torch power is a crucial parameter, because it influences the velocity and temperature of the plasma flow and, in consequence, powder particles directly. The deposition efficiency, interface bonding and mechanical properties of sprayed coatings increases with the increase of particle velocity and temperature [18e25]. It was demonstrated that the as-sprayed coatings show elemental composition different from the initial feedstock powders [7,8,13]. Variation of plasma gun power allows changing and controlling the plasma jet temperature and, consequently, to adjust the melting degree of the feedstock powders [14,25]. Therefore, it is generally considered that the fully melted particles will lead to production of dense with low pore volume coatings [14]. The nature and elemental composition of the initial powders are directly related to the melting temperature [7,13,14]. Therefore it is very important to obtain the optimal spraying process parameters for individual feedstock powders. The investigations concerning the formation of carbonesiliconesulfur composite coatings from flakelike shape powders by plasma spraying is hard to find in the scientific literature. As a result it is very important to determine appropriate process parameters which will lead to the deposition of coatings with dense structure and high hardness values. In this study, the effect of torch input power on the surface morphologies, roughness, pore size, pore distribution, and
http://dx.doi.org/10.1016/j.vacuum.2015.04.028 0042-207X/© 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Marcinauskas L, et al., Formation of carbon composite coatings by plasma spraying, Vacuum (2015), http:// dx.doi.org/10.1016/j.vacuum.2015.04.028
2
L. Marcinauskas et al. / Vacuum xxx (2015) 1e6
microhardness of sprayed carbonesiliconesulfur composite coatings was investigated. 2. Experimental details The coatings were deposited on steel substrates at atmospheric pressure using a direct current plasma torch. The plasma torch used in this experiment was developed at the Lithuanian Energy Institute [26e28]. Substrates with the dimensions of 10 40 1 mm were made from stainless steel. The steel substrates were polished and chemically cleaned by acetone before starting the deposition process. The steel substrates were placed on the water-cooled sample holder. The coatings were sprayed at a torch power of 17.1 kW, 19.4 kW, and 21.0 kW, respectively. The spraying distance was kept at 35 mm. Argon was used as both the primary gas (flow rate of 7.5 g/s) and the powder carrier gas (flow rate of 0.59 g/s). The experiments were performed using a cylindrical shape (0.15 m of length) reactor with the powder injection place (located 0.10 m from the exit) which was situated at the exit of the anode. The mean temperature of the plasma jet at the exhaust of the plasma torch was calculated from the heat balance corresponding to plasma enthalpy:
Hf ¼ M
ðPT Pw Þ þ Ho G
(1)
where Hf is the plasma enthalpy (kJ mol1), PT is torch power (kW), PW is power used for water heating (kW), G is plasma mass flow rate (kg s1), M argon gas atomic mass (kg mol1), and H0 is the enthalpy under standard conditions (kJ mol1). The mean velocity of the plasma jet at the exhaust of the plasma torch was calculated from Eq. (2):
v¼
4MGT pd2 p
(2)
where T is mean temperature of the plasma (K), d e diameter of the exhaust nozzle (m) and p e pressure (m). The average temperature and velocity of the plasma flow at the different torch powers is given in Table 1. Detailed description of the methodology of the plasma temperature and velocity calculation is given in Ref. [29]. The graphite powders with the elemental composition of carbon (C) e 95.9 at%, oxygen (O) e 2.9 at%, silicon (Si) e 1.0 at% and sulfur (S) e 0.2 at% were used as sprayed powders. The graphite powders used in the spraying process were of a non-regular flake-like shape with various sizes from 20 mm to 150 mm, and thickness of ~10 mm (Fig. 1a). The specific surface area of these powders was 1.84 m2/g. The feedstock powders were grinded and sieved using a 150 mm size grid and were dried for 8 h at ~350 K before starting the deposition process. The feedstock powders were injected into the reactor nozzle at a distance of 100 mm from the exit. The deposition duration was 60 s. The plasma torch was moving in the x-axis direction forward and back during the deposition. The surface morphology of the sprayed coatings and feedstock powders was analyzed by scanning electron microscopy (SEM) using a JEOL JSMe5600. The elemental composition of deposited coatings and powders was obtained by energy dispersive X-ray
Table 1 Plasma spraying parameters for depositing composite coatings. Power, kW
I, A
U, V
Velocity, m/s
Temperature, K
17.1 19.4 22.0
180 200 220
95 97 100
540 ± 15 600 ± 15 670 ± 15
3180 ± 50 3520 ± 50 3940 ± 50
spectroscopy (EDS). The surface roughness measurements were investigated using a stylus profiler (Ambios XP-200). The measurements were done 5 times; the length of one measurement was 2 mm. The coating structure was analyzed by X-ray diffraction (XRD) (DRON-UM1 with standard BraggeBrentano focusing geometry) in a 10e100 range using the CuKa (l ¼ 0.154059 nm) radiation. The bonding structure of sprayed coatings was investigated by Raman scattering (RS) spectroscopy (“Jobin Yvon” company). RS measurements were done using a Nd:YAG laser (532.3 nm, 50 mW, spot size 0.32 mm, in the 1000e2000 cm1 range) as an excitation source. The RS spectra were fitted by Gaussian-shape lines in the spectral range of (1200e1800) cm1. The specific surface area of the coatings was measured by the BETmethod in a KELVIN 1042 sorptometer. The microhardness measurements were done using a CSM Micro Scratch Tester with Vickers indenter. The microhardness was measured on the polished coating surface at room temperature by applying a load of 10 mN. The speed of the loading and unloading was 200 mN/min and the pause between the loading and unloading rate was 10 s. 10e12 hardness measurements were taken randomly and the arithmetic mean of measurements was taken as the microhardness of the coatings. Beside the hardness values the elastic deformation work of indentation (We), plastic deformation work of indentation (Wp), and total mechanical work of indentation (Wt) were given from the force e penetration depth curves for loading and unloading. The elasticity (elastic part of indentation work) was calculated using Eq. (3) [30]:
ε¼
We $100% ¼ Wt
1
Wp $100% Wt
(3)
3. Results and discussion Fig. 1 shows the surface morphologies of plasma-deposited composite coatings at different torch powers. The surface of a coating formed at the lowest power is composed from partially melted and solidified flake-like particles (Fig. 1b and e). The number of partially molten particles in the coating decreases with increasing torch power. The ball-like shape particles with size of 1e10 mm appear on the surface at higher power (Fig. 1c, d and f). The appearance of pores and non-regular shape particles could be found in all deposited coatings. The existence of sphere-like shape particles on the surface indicates the higher melting degree of the initial feedstock powders. It is well known that pores in sprayed coatings are generally derived from poorly stacked flat particles and that the state of the flat particles is highly dependent on the melted extent of the feedstock powders and the velocity of the plasma jet [21]. The SEM cross-section measurements indicated that the thicknesses of the coatings were in the range of 10e15 mm. The EDS measurements demonstrated that the oxygen content in the assprayed coatings decreased from 44.4 at% to 41.8 at% with increasing power from 17.1 kW to 22.0 kW. The carbon content increased from 31.8 at% to 38.5 at%. It should be noted that, despite the low concentration of Si and S in the initial powders, the sprayed coatings contained quite a high fraction of these elements. The silicon concentration decreased from ~16 at% to ~12 at%, while the sulfur content did not change (~8 at%) in coatings with the increased power values. In order to evaluate the torch power influence on the surface morphologies surface roughness measurements were performed. The surface roughness investigations indicated (see Fig. 2) that coatings sprayed at the lowest power had the highest surface roughness (Ra ¼ 700 nm, Rq ¼ 1000 nm). The surface roughness
Please cite this article in press as: Marcinauskas L, et al., Formation of carbon composite coatings by plasma spraying, Vacuum (2015), http:// dx.doi.org/10.1016/j.vacuum.2015.04.028
L. Marcinauskas et al. / Vacuum xxx (2015) 1e6
3
Fig. 1. Surface micrographs of powders (a) and as-sprayed coatings prepared at different powers: (b, e) 17.1 kW, c) 19.4 kW and (d, f) 22.0 kW.
decreased to 590 nm and 840 nm when 19.4 kW power was used. The further increase of the torch power has almost no influence on the surface roughness values (Fig. 2). As can be seen the surface roughness values and oxygen content decreased with increasing torch power, which implies that the deposition efficiency of composite coatings increased.
Fig. 2. The surface roughness of coatings deposited at different spraying power.
The existence of significant fractions of Si and S in the sprayed coatings indicates that the temperature in the plasma flow was not enough to completely melt the feedstock powders. The increase of Si and S concentrations in the coatings compared to those in the powders is related to the large difference in melting temperatures between carbon, silicon (~1800 K) and sulfur (~400 K) and to the poor wettability of the graphite. Lopez et al. [7] demonstrated that the SiC incorporation in the coating matrix with respect to that the feedstock powder was only 10%. Meanwhile the deposition efficiency of the process in terms of vol.% SiC was not higher than 20%. The increase of the oxygen in the coatings related to the fact that the vaporized carbon reacted with the entrapped or surrounding air and formed graphite oxide dust during the spray process. The lowest (~540 m/s) plasma flow velocity was obtained at 17.1 kW. Thus the particle flight duration in the argon plasma was the longest and the probability of the particle to oxidize increased. The low amount of carbon in the composites could also be due to formation of carbon dioxide (CO2) or carbon monoxide (CO). N.F Fahim [8] demonstrated that the carbon reacted with oxygen and escaped as CO2 and CO during the formation of W/SiC composites. It should be mentioned that there were many small ball-like particles with micrometer sizes on the coating surface. The existence of such structures is related to splash debris from the impacting graphite
Please cite this article in press as: Marcinauskas L, et al., Formation of carbon composite coatings by plasma spraying, Vacuum (2015), http:// dx.doi.org/10.1016/j.vacuum.2015.04.028
4
L. Marcinauskas et al. / Vacuum xxx (2015) 1e6
splats due to their low viscosity [31]. The coating obtained at a low power of 17.1 kW contains many non-regular shape fragments on the surface and the lowest carbon concentration. This is because such plasma temperature is inadequate to completely melt the feedstock powders. The large flake-like particles probably impact on the substrate in the solid state and rebound off the surface of the coatings. However the plasma temperature is quite high to completely melt silicon which is incorporated in initial graphite powders. When the torch power increases, the temperature increases and favors an increase in the degree of melting of the powders and velocity of plasma flow. As a result the amount of unmelted and partially-melted particles decreases and a higher fraction of well melted powders can easily form flat splats. In addition flat splats can integrate tightly. The lower surface roughness values obtained at higher power indicate the improvement of splat bonding. Fig. 3a shows the dependence of surface area and pore volume in plasma sprayed coatings using different torch powers. As can be seen, the specific surface area of the coatings increased from 11.8 m2/g to 14.5 m2/g as the spraying power increased from 17.1 kW to 22.0 kW. Meanwhile, the pore volume increased almost twice from ~19.6 mm3/g to ~33.4 mm3/g with the increased power. It should be noted that the pores surface area was ~4.7 m2/g, 9.4 m2/ g and 7.4 m2/g for the coatings prepared at torch power of 17.1 kW, 19.4 kW and 22.0 kW, respectively. The pore diameter influence on the surface area and pore volume values are demonstrated in Fig. 3b. The measurements demonstrated that the smaller size pores appeared on the surface with the increased input power. It
Fig. 3. The surface area and pore volume dependence on spraying power (a) and pore distribution on the surface and pore volume values (b).
should be mention that the largest pores (~25 nm) have the highest pore volumes (Fig. 3b). The specific surface area mainly is determined by the micro-pore and meso-pore quantity in the coatings. The amount of pores in the specific surface area of the coating was ~40%, ~78% and ~51% for 17.1 kW, 19.4 kW and 22.0 kW power, respectively. Thus, the decrease in the surface roughness was compensated by the increased amount of pores. P. Ctibor et al. [32] demonstrated that the higher the torch power the lower surface roughness and porosity will be. However, it will result in the production of a higher amount of pores with smaller size. The higher degree of melting and higher velocity of the powders particles lead to a more uniformly and steady distribution of the splats. Another feature of the higher torch power coated splats is that cracks and holes with lower dimensions are formed on solidification [33]. The suggestion would be that the crack and hole areas may serve as the source of meso-pores. Fig. 4 demonstrates the XRD patterns of the steel substrate and composite coatings sprayed at different temperatures. The existence of the peaks related to the steel substrate was obtained for all deposited coatings. The diffraction peaks at ~26.6 , ~44.6 , and 54.7 corresponds to (002), (100), and (004) planes of graphite, respectively [34,35]. The intensity of the peak assigned to (100) graphite increased with increased input power. The appearance of the low intensity peak at ~12.4 is attributed to (001) graphite oxide [35]. The existence of graphite oxide peak indicates that part of graphite was transformed into graphite oxide during plasma spraying. It should be mentioned that the XRD patterns do not indicate formation of SiC, or any other crystalline phase related to Si or S elements. This indicates that the silicon and sulfur are in an amorphous phase. The Raman spectra of the film sprayed at 17.1 kW power has two separated D (1361 cm1) and G (1592 cm1) peaks. The peak at ~1361 cm1 is assigned to the disordered graphite structure (Dband), while the peak at 1592 cm1 (G-band) corresponds to a splitting of the E2g stretching mode of graphite and reflects the structural intensity of the sp2-hybridized carbon atoms [35]. The relative intensity of the D-band and G-band (ID/IG) ratio is 0.25. The shapes of the spectra of coatings deposited at 19.4 kW and 22.0 kW, were almost identical as those for films deposited at 17.1 kW power (Fig. 5). The increase of power influenced the slight increase of the ID/IG ratio, the broadening of the D peak (DD), the narrowing of the G band (DG) and the shift of the D band to higher values (Fig. 5). The ID/IG ratio slightly increased from 0.25 to 0.29 with increasing torch
Fig. 4. XRD patterns of the sprayed coatings and steel substrate.
Please cite this article in press as: Marcinauskas L, et al., Formation of carbon composite coatings by plasma spraying, Vacuum (2015), http:// dx.doi.org/10.1016/j.vacuum.2015.04.028
L. Marcinauskas et al. / Vacuum xxx (2015) 1e6
Fig. 5. Raman spectra of coatings formed at different powers.
power from 17.1 kW to 22.0 kW. The higher intensity ratio of Dpeak to G-peak (ID/IG) shows the higher extent of defects in carbon coatings [35,36]. The narrowing of the G band (from 140 cm1 to 118 cm1) indicates less distortion in the carbon coatings. It was demonstrated that, by thermally oxidizing the graphite flake, the intensity of the G-peak decreases, G-band widening, and the intensity of D-band increases [36,37]. Such result could be due to higher plasma temperatures and lower oxygen content in the coating. The coating sprayed at the lowest power has the highest content of oxygen. Thus the fraction of sp2 C]O bonds in the coatings would be higher and will result in an increase in concentration of defects in the structure. The changes in microhardness values of the carbon composite coatings sprayed at different plasma torch power are shown in Fig. 6a. The lowest hardness value (~120 HV) was obtained for the lowest torch power. As the spraying power was increased to 19.4 kW, the hardness values of the sprayed coating was increased up to 243 HV and was similar to that for the steel substrate (~255 HV). The coating sprayed at the highest power demonstrated the highest microhardness values (314 HV). The elasticity measurements indicated that, despite the low hardness value of coating prepared at 17.1 kW, the elasticity was similar to that of the steel substrate (Fig. 6 b). The highest elasticity (33%) was observed for the coating deposited at 19.4 kW. The hardness values are mainly affected by the microstructure and porosity of the sprayed coatings. At low power the presence of unmelted particles and pores decreases the hardness value mainly due to weakness of cohesion bonds of accumulated splats due to lower particle temperature [23]. As the power increases the plasma flow temperature increases, powder particles get more energy from the plasma and, as a result, the majority of particles are melted and accumulate in the dense structure on the steel substrate. Other authors also indicated
5
Fig. 6. Effect of torch power on the microhardness (a) and elasticity (b) of coatings.
that the dense and low porosity coatings prepared by plasma spraying were harder [18,19,21,23]. It is well known that the microstructure of the as-sprayed coatings is strongly dependent on the interaction of impacting particles with the underlying surface [16]. The flattening degree, defined as the ratio of splat diameter to impacting droplet diameter, is one important parameter used to reflect this interaction. The flattening degree increases with the rise of particle impact velocity [14]. As the torch power increased from 17.1 kW to 22.0 kW, the plasma flow velocity increased from 540 m/s to 670 m/s. Naturally, elevating torch power favors to increase the in-flight particle velocity and herewith flattening degree. Thus, the effective bonding area between the individual splats would be enlarged and the number of pores due to the incomplete inter-splat contact would be decreased. The velocity of plasma increases with the increase in power, which results from an improved thermal pinching effect on the plasma jet [25]. Therefore, an increase in plasma velocity increases particle velocity and results in a slight decrease in dwelling time of particles in the plasma flow. The fact that the hardness values increased and surface roughness decreased can be attributed that the improvement of particle heating can compensate for the shortening of dwelling time of particles in the plasma. The decreased surface roughness values at higher powers are the result of the increased degree of smoothing and higher particle melting. 4. Conclusions Carbonesiliconesulfur composite coatings were prepared using atmospheric pressure plasma spraying at a various torch powers from the flake-like graphite powders. As the plasma flow temperature raises the number of larger particles decreases in the coatings
Please cite this article in press as: Marcinauskas L, et al., Formation of carbon composite coatings by plasma spraying, Vacuum (2015), http:// dx.doi.org/10.1016/j.vacuum.2015.04.028
6
L. Marcinauskas et al. / Vacuum xxx (2015) 1e6
surfaces. The surface roughness values and coatings oxidation degree depend on the torch power values. The surface roughness values decreases with the increase of torch power. The oxidation degree of as-sprayed coatings becomes lower with the increase of plasma flow velocity (torch power). The specific surface area and pore volume of coatings increases as the torch power increases. The increase in plasma temperature and velocity lead to an increase in crystalline graphite phase. As the power increases from 17.1 kW to 22.0 kW, the microhardness values of sprayed composite coatings increases from 120 HV to 314 HV. The increase of the microhardness attributed to an increase of crystalline phase and improvement of bonding between the individual splats in the structure of the composite. The highest elasticity was estimated for the coating prepared at 19.4 kW. Acknowledgments The authors thank Dr. D. Mil cius from the Lithuanian Energy Institute for possibility to perform SEM, EDS and surface roughness measurements and Dr. A. Zunda from Aleksandras Stulginskis University for microhardness measurements. References [1] Pandolfo A, Hollenkamp A. Carbon properties and their role in super capacitors. J Power Sources 2006;157:11e27. [2] Zhu T, Li Y, Luo M, Sang S, Wang Q, Zhao L, et al. Microstructure and mechanical properties of MgOeC refractories containing graphite oxide nanosheets (GONs). Ceram Int 2013;39:3017e25. [3] Mu M, Zhou X, Xiao Q, Liang J, Huo X. Preparation and tribological properties of self-lubricating TiO2/graphite composite coating on Ti6Al4V alloy. Appl Surf Sci 2012;258:8570e6. [4] Kicinski W, Szala M, Bystrzejewski M. Sulfur-doped porous carbons: synthesis and applications. Carbon 2014;68:1e32. [5] Yang Z, Nie H, Chen X, Chen X, Huang S. Recent progress in doped carbon nanomaterials as effective cathode catalysts for fuel cell oxygen reduction reaction. J Power Sources 2013;236:238e49. [6] Yang XB, Zhu W, Qin K, Wang HY. Preparation of lamellar carbon matrix for sulfur as cathode material of lithium-sulfur batteries. Electrochim Acta 2014;143:374e82. pez AJ, Torres B, Taltavull C, Rams J. Influence of high velocity oxygen-fuel [7] Lo spraying parameters on the wear resistance of AleSiC composite coatings deposited on ZE41A magnesium alloy. Mater Des 2013;43:144e52. [8] Fahim NF. Thermal conductivity of plasma-sprayed W/SiC composite for hightemperature energy applications. Surf Coat Techn 2008;202:1696e703. [9] Rammo NN, Al-Amery HR, Abdul-Jabbar T, Jaffer HI. Adhesion, hardness and structure of thermal sprayed Al/SiC composite coat on graphite. Surf Coat Tech 2009;203:2891e5. [10] Pan ZY, Wang Y, Li XW, Wang CH. Fabrication and characterization of heat and plasma treated SiC/Al2O3eYSZ feedstocks used for plasma spraying. Vacuum 2012;86:1558e67. [11] Sathiyamoorthy R, Shanmugam K, Balasubramanian V. Dry sliding wear behavior of SiC reinforced titania coating deposited by high velocity oxy fuel spraying. Proc Mater Scien 2014;5:648e55. [12] Chadami F, Ghadami S, Abdollah-Pour H. Structural and oxidation behavior of atmospheric heat treated plasma sprayed WCeCo coatings. Vacuum 2013;94: 64e8. [13] Tului M, Giambi B, Lionetti S, Pulci G, Sarasini F, Valente T. Silicon carbide based plasma sprayed coatings. Surf Coat Tech 2012;207:182e9.
[14] Fauchais P. Understanding plasma spraying. J Phys D Appl Phys 2004;37: 86e108. [15] Cetinkaya S, Eroglu S. Effect of temperature and Ar flow rate on Ti3SiC2 formation from TiSiO4 powders coated with pyrolytic carbon. Int J Refract Met H 2014;43:69e76. [16] Brinkiene K, Kezelis R. Effect of alumina addition on the microstructure of plasma sprayed YSZ. J Eur Ceram Soc 2005;25:2181e4. [17] Niu Y, Zheng W, Liu X, Ji H, Ding C. Influence of powder size on characteristics of air plasma sprayed silicon coatings. Ceram Int 2012;38:5897e905. [18] Wu H, Li H, Lei Q, Fu Q, Ma C, Yao D, et al. Effect of spraying power on microstructure and bonding strength of MoSi2-based coatings prepared by supersonic plasma spraying. Appl Surf Sci 2011;257:5566e70. [19] Wang C, Fan L, Fan J, Zhang D, Wang H. Effect of spraying power on microstructure and properties of supersonic plasma sprayed Al2O3 coating on porous Si3N4 substrate. J Alloys Compd 2013;559:152e7. [20] Marcinauskas L, Valatkevi cius P. The effect of plasma torch power on the microstructure and phase composition of alumina coatings. Mater Sci Pol 2010;28:451e8. [21] An Y, Hou G, Chen J, Zhao X, Liu G, Zhou H, et al. Microstructure and tribological properties of iron-based metallic glass coatings prepared by atmospheric plasma spraying. Vacuum 2014;107:132e40. [22] Morks MF, Kobayashi A. Effect of gun current on the microstructure and crystallinity of plasma sprayed hydroxyapatite coatings. Appl Surf Sci 2007;253:7136e42. [23] Morks MF, Gao Y, Fahim NF, Yingqing FU. Microstructure and hardness properties of cermet coating sprayed by low power plasma. Mater Lett 2006;8:1049e53. [24] Tahara H, Ando Y. Study of titanium nitride deposition by supersonic plasma spraying. Vacuum 2008;83:98e101. [25] Li C, Sun B. Effects of spray parameters on the microstructure and property of Al2O3 coatings sprayed by a low power plasma torch with a novel hollow cathode. Thin Solid Films 2004;450:282e9. esnien _ [26] Brinkiene_ K, C e_ J, Ke_ zelis R, Me_ cius V, Baltusnikas A, Zunda A. Structural and surface analysis of plasma processed zirconia coatings. Mater Sci-Medzg 2006;12:300e4. [27] Ke_ zelis R, Me_ cius V, Pranevi cius LL. High velocity plasma torch design and its flow parameters for ceramics thin film deposition. In: Gylys J, Sven cianas P, Ziug zda V, editors. Heat energy and technologies; 2002. p. 253e8. Kaunas. [28] Brinkiene_ K, Ke_ zelis R, Me cius V, Pranevi cius L. Effect of outlet nozzle design on the YSZ particle in-flight characteristics in plasma jet generated by DC plasma torch. In: Sunden B, Vilemas J, editors. Advances in heat transfer engineering; 2003. p. 629e36. New York. te_ V, Marcinauskas L. [29] Valin cius V, Krusinskaite_ V, Valatkevi cius P, Valin ciu Electrical and thermal characteristics of the linear, sectional dc plasma generator. Plasma Sources Sci Technol 2004;13:199e206. [30] Giannakopoulos AE, Suresh S. Determination of elastoplastic properties by instrumented sharp indentation. Scr Mater 1999;40:1191e8. [31] Zhang C, Guo RQ, Yang Y, Wu Y, Liu L. Influence of the size of spraying powders on the microstructure and corrosion resistance of Fe-based amorphous coating. Electrochim Acta 2011;56:6380e8. [32] Ctibor P, Hrabovvsky M. Plasma sprayed TiO2: the influence of power of an electric supply on particle parameters in the flight and character of sprayed coating. J Eur Ceram Soc 2010;30:3131e6. [33] Yugeswaran S, Selvarajan V, Seo D, Ogawa K. Effect of critical plasma spray parameter on properties of hollow cathode plasma sprayed alumina coatings. Surf Coat Tech 2008;203:129e36. [34] Channu VS, Bobba R, Holze R. Graphite and graphene oxide electrodes for lithium ion batteries. Colloids Surfaces A Physicohem Eng Aspects 2013;436: 245e51. [35] Geng L, Wu S, Zou Y, Jia M, Zhang W, Yan W, et al. Correlation between the microstructures of graphite oxides and their catalytic behaviors in air oxidation of benzyl alcohol. J Colloid Interf Sci 2014;421:71e7. [36] Matsuo Y, Ueda K. Pyrolytic carbon from graphite oxide as a negative electrode of sodium-ion battery. J Power Sources 2014;263:158e62. [37] Li Y, Wang Q, Fan H, Sang S, Li Y, Zhao L. Synthesis of silicon carbide whiskers using reactive graphite as template. Ceram Int 2014;40:1481e8.
Please cite this article in press as: Marcinauskas L, et al., Formation of carbon composite coatings by plasma spraying, Vacuum (2015), http:// dx.doi.org/10.1016/j.vacuum.2015.04.028