- Email: [email protected]
Pulsed laser deposition of alumina coating for corrosion protection against liquid uranium
Materials Chemistry and Physics 143 (2014) 1446e1451
Contents lists available at ScienceDirect
Materials Chemistry and Physics journal homepage: www.elsevier.com/locate/matchemphys
Pulsed laser deposition of alumina coating for corrosion protection against liquid uranium A.K. Singh a, Santu Kaity b, Kulwant Singh c, J. Thomas a, T.R.G. Kutty b, Sucharita Sinha a, * a
Laser & Plasma Technology Division, Bhabha Atomic Research Centre, Mumbai, India Radiometallurgy Division, Bhabha Atomic Research Centre, Mumbai, India c Fusion Reactor Materials Section, Bhabha Atomic Research Centre, Mumbai, India b
h i g h l i g h t s Alumina films deposited on stainless steel via pulsed laser deposition (PLD) technique. PLD alumina films investigated as potential corrosion protective coatings. Deposited coatings have been characterized in terms of their microstructure and crystalline phase. Corrosion resistance of coatings against liquid uranium was tested. Results suggest PLD alumina films have a promising potential for containment of molten uranium.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 February 2013 Received in revised form 24 October 2013 Accepted 29 November 2013
Alumina coatings find wide applications as tribological coatings and as corrosion protective coatings for structural materials against chemical attack. We have investigated alumina coatings deposited on Stainless Steel (SS) substrates via pulsed laser deposition (PLD) technique. Characterization tests performed on these coatings including their compatibility with liquid uranium suggests alumina to be a potential candidate as a coating material for handling and containment of liquid uranium. We present here results of our detailed parametric study including dependence of average mass removal rate on laser fluence and ablation geometry and average deposition efficiency during PLD. These measurements provide vital inputs facilitating proper choice of process parameters for PLD runs. Deposited coatings have been characterized in terms of their microstructure, surface profile, adhesion to substrate, crystalline phase and corrosion resistance against liquid uranium. Our PLD based alumina coatings have shown a high degree of compaction and excellent corrosion resistance to molten uranium even upto a temperature of 1165 C. Ó 2013 Elsevier B.V. All rights reserved.
Keywords: Coatings Ceramics Thin films Corrosion
1. Introduction Among various physical vapor deposition techniques, Pulsed Laser Deposition (PLD) is a simple and convenient method for deposition of films and coatings [1]. In this technique, focused laser radiation is used to vaporize material from a target and this ablated vapor is then deposited on a desired substrate [2]. The evaporation process of the target can be either thermal or non-thermal determined largely by the laser conditions. PLD technique has numerous advantages over other conventional techniques of film deposition. This technique enables deposition of coatings having precise stoichiometry and crystalline purity, offering at the same time,
* Corresponding author. Tel.: þ91 22 25595352; fax: þ91 22 25505151. E-mail addresses: [email protected], [email protected] (S. Sinha). 0254-0584/$ e see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.matchemphys.2013.11.063
flexibility and control over the deposition rate [1,2]. Properties such as surface morphology, crystalline phase and density of defects in deposited film can be tailored by controlling PLD process parameters such as, laser wavelength, fluence, pulse duration and pulse repetition rate, distance between substrate and target and the environment of deposition [3,4]. In addition, PLD technique allows facile process automation and is also environment friendly [5]. In PLD process congruent removal of material from the target leads to preservation of stoichiometry in the deposited film. Thus PLD is widely used for deposition of chemically inert and complex oxide films with accurate stoichiometry [1,6]. However, PLD technique does have its limitations too. Narrow angular spread and inhomogeneous flux of species within the ablated plume restricts scaling up of PLD technique for deposition of large area uniform coatings [7]. A technical difficulty with PLD arises on account of presence of macroscopic particulates (size w few micron) in the deposited film
A.K. Singh et al. / Materials Chemistry and Physics 143 (2014) 1446e1451
due to splashing effect [1,7]. Despite these drawbacks researchers have developed techniques to deposit large area coatings (up to 200 mm diameter substrates) having reduced density of macroscopic particulates [8]. Refractory materials, organic compounds, metals, dielectrics, semiconductors and superconductors, deposited as thin films over a wide range of substrates serve many applications. Thin film dielectric coatings are deposited on glass substrate for optical components such as optical wave-guides, reflecting and antireflecting coatings [9]. Various coatings on structural components provide protection against harsh environmental conditions. Ceramic coatings used in automobile industry lead to economical and ecological benefits through reduced weight of the components and enhanced service life of various parts of the vehicle [10]. Also, in nuclear technology corrosion resistant coatings are necessary for containment and handling of highly corrosive liquids such as molten uranium, actinide wastes, and liquid lithium blankets in fusion reactors [11,12]. Alumina is an excellent ceramic material, which is employed in the form of coating for a wide range of applications owing to its properties such as high hot hardness [13,14], high mechanical strength [13], high melting point (2053 C) [15] and good oxidation resistance [13,16]. Alumina also shows very good corrosion resistance against chemical attack [17]. Many polymorphic phases of alumina exist [18]. Of these polymorphs only a-alumina is thermodynamically stable above 1050 C and it is also the densest amongst all polymorphs [19]. Therefore, coatings grown of polymorphs other than a-alumina catastrophically fail to provide corrosion protection at temperatures above 1050 C as irreversible conversion of the unstable polymorphs to the densest a-alumina occurs at these elevated temperatures resulting in formation of cracks and pores in the coating. These cracks and pores formed in the coating due to phase transformation allow penetration of the corrosive chemicals to the structural material. Hence for high temperature (above 1050 C) corrosion protection application, alumina coating having pure a-phase is essential. In the study reported here, coatings of a-alumina have been deposited on surface roughened stainless steel (SS) substrates, employing PLD technique at room temperature and ambient atmospheric pressure. The deposited coatings have been characterized in terms of their surface profile, surface microstructure, adhesion to substrate, crystalline phase and corrosion resistance against molten uranium. Parametric investigation of PLD process including dependence of mass removal rate on laser fluence and ablation geometry, variation of ablation depth with laser fluence and average deposition efficiency (ratio of the mass deposited on substrate to mass ablated from target during PLD) during the PLD runs have also been studied. Results of our investigation enable proper choice of PLD process parameters facilitating deposition of alumina coatings demonstrating excellent corrosion protection capability. 2. Experiment
1447
the region of 18e20 mJ, pulse duration of 6 ns (FWHM) at 10 Hz pulse repetition rate and a wavelength of 532 nm has been used to ablate the target. Evaporation of material from the target occurs largely normal to the target surface. Therefore, the target was irradiated with the laser beam incident at an angle of 45 . In this configuration, spatial overlap between the incident laser beam and the vapor plume could be reduced, hence restricting attenuation of the incident laser beam by the plume [20,21]. The substrate and the target surface were maintained parallel to each other, separation between the two being w3 mm. A schematic of our experimental setup is shown in Fig. 1. PLD technique for film deposition typically employs high vacuum conditions. However, our studies involved PLD carried out in atmospheric air. Not being restricted by a vacuum chamber meant greater experimental flexibility and technical advantages when coating substrates having complex geometrical shapes [22]. However, PLD in air results in restricted plume length on account of collisions between the vaporized particles and ambient atmosphere which consequently limits the maximum allowable separation between substrate and target. Separation between target and substrate being restricted lead to critical demands on precise scanning movement of the substrate ensuring at the same time, that the incident laser beam does not get blocked during scanning of the substrate. Therefore, substrate could be moved only over a distance of w3 mm in horizontal direction although movement in vertical direction had no such restrictions. In the present work, typically an area of w0.15 cm2 was coated at a time. By sequentially coating a portion of the substrate and then rotating the substrate to access a fresh uncoated region, SS substrates having diameter of 5e 6 mm could be completely PLD coated with alumina in our setup. For different ablation configurations average mass removal rate as a function of laser fluence was systematically investigated and average deposition efficiency (ratio of the mass deposited on substrate to mass ablated from the target) averaged over 18,000 laser pulses was measured using a micro-balance with measuring accuracy of 10 mg. Microstructure of the deposited coatings and their crystalline phases were investigated using a Scanning Electron Microscope (SEM) and X-ray Diffraction (XRD) technique, respectively. Adhesion of the coating to the substrate was tested by adhesive tape test and scratch tests employing a diamond indenter of 200 mm diameter. Corrosion resistance of the deposited coating against molten uranium has been characterized using Differential Thermal Analysis (DTA) technique [23]. DTA signal, which is a measure of temperature difference between a reference sample and the test sample when subjected to identical thermal cycles, was recorded. Observed peaks and dips in DTA signal provide signatures of phase transitions and chemical reactions in the test specimen. A laser fluence of 11 J cm2 has been typically employed for ablation of Al2O3 target and subsequent deposition of these ejected species on surface roughened SS substrates. M1 Nd: YAG Laser X
For our PLD studies sintered alumina pellets (diameter 10 mm and thickness 10 mm) have been used as targets and surface roughened SS discs as substrates. Surface roughening via sandblasting of substrate improves mechanical adhesion of the coating with substrate surface. Both, the target and substrate were mounted on individual translational stages in order to have independent translation of both the target and the substrate maintaining, at the same time, a constant separation between the two. In this manner, deposition of coatings of uniform thickness over large area of the substrate was achieved. A focused beam of Nd:YAG nanosecond (ns) laser delivering laser pulses typically having pulse energy in
Y
Z
Target
M2 L
Plume Substrate Fig. 1. Schematic of the PLD setup with folding mirrors M1, M2 and convex lens (L) having focal length of 50 cm. Target and substrate both have been translated in the Xe Y plane while rotation is about z-axis.
1448
A.K. Singh et al. / Materials Chemistry and Physics 143 (2014) 1446e1451
3. Results and discussion In order to arrive at an optimum laser fluence for PLD of alumina, we first measured the average mass removal rate per pulse from alumina target as a function of incident laser fluence. Three different target surface configurations were studied: (a) laser beam incident at 45 with respect to scanned target surface, (b) normal incidence of laser beam on stationary target surface, (c) laser beam at normal incidence on scanned target surface. Average mass removal per pulse for each laser fluence value has been estimated by measuring the cumulative mass removed from the target on irradiation with a total of 18,000 laser pulses. Fig. 2 shows average mass removal rate for varying incident laser fluence for the three different configurations described above. From Fig. 2 it is evident that mass removal rate is highest when laser beam is incident at 45 to the target surface and the target surface is also translated using a stepper motor controlled translational stage. With the laser beam incident at 45 to target surface, spatial overlap of laser beam and the generated plume of the ablated species is minimized thus reducing resultant attenuation and distortion of the incident laser beam by the vapor plume. Scanning of the target also reduces formation of craters on the surface owing to material removal through laser ablation. Crater formation needs to be avoided since it alters the direction of the ablation plume and also leads to increased scattering of the laser beam from the target surface. This results in reduced coupling of laser energy into the target. During our investigation the target was moved at a speed of 0.02 cm s1 sequentially in X and Y directions. Measured dependence of ablation depth of the crater formed on target surface when irradiated with 600 laser pulses with increasing incident laser fluence is shown in Fig. 3. For these measurements the target surface was maintained static and normal with respect to incident laser beam. Ablation depth has been measured using a surface profiler (Model No: Taylor Hobson, Form Talysurf Series 2). Fig. 3 clearly shows that the ablation depth steadily increases with increasing laser fluence till a maximum crater depth is reached. Beyond this value of laser fluence ablation depth shows a decreasing trend with further increase in laser fluence. Uneven surface and deep crater formation results in grazing incidence of laser beam on target. This not only reduces the effective laser fluence experienced by the target but also enhances loss of laser energy through multiple scattering on crater walls. Based on our observations on ablation rate from alumina target a laser fluence value of 11 J cm2 was chosen for our PLD runs. At this fluence level, ablation plume w3 mm long, were typically
Fig. 2. Average mass removal per pulse as a function of incident laser fluence under different ablation geometries: (a) laser beam incident at 45 with respect to scanned target surface (b) normal incidence of laser beam on stationary target surface (c) normal incidence of laser beam on scanned target surface.
Fig. 3. Variation of ablation depth with incident laser fluence.
generated. The time averaged length of the visible plume was estimated via viewing using suitable optical filters and attenuators to avoid saturation. Although longer plume lengths are preferred since they allow experimental flexibility while positioning the target and substrate for PLD, ablation at laser flux levels far above ablation threshold often cause explosive type of ejection of large fragments from the target. Deposition of such massive fragments needs to be avoided as it degrades the quality of the deposited film. Thus, optimization of laser fluence employed for PLD not only helps in achieving high deposition rate, but also minimizes presence of macro particulates in the coating hence improving the quality of deposited film. Alumina coatings having a maximum thickness w75 mm have been deposited on surface roughened SS substrate by using PLD technique. Average efficiency for coating deposition was estimated to be w3% for a cumulative 7 h PLD run when average mass removal per pulse was measured to be w205 ng/pulse. The drop in average mass removal rate measured for extended PLD runs (7 h) in comparison to short duration runs (half hour) shown in Fig. 2, is largely on account of increased surface degradation of target during the extended PLD runs. Adhesion of the coating with substrate was tested using adhesive tape test. In this test a strip of adhesive tape was applied to the coated surface and rapidly pulled off. No peeling of the coating occurred when the adhesive tape was pulled off indicating good mechanical adhesion of the deposited coating with the substrate. Thickness of the coating was measured by taking a transverse cut through the coated surface and subsequently
Fig. 4. SEM image of a typical transverse segment cut through PLD alumina coated SS substrate.
A.K. Singh et al. / Materials Chemistry and Physics 143 (2014) 1446e1451
1449
Fig. 5. Surface profile of the deposited alumina coating.
Fig. 6. SEM image showing surface microstructure of the alumina coating at two different levels of magnification.
viewing the cross section under a Scanning Electron Microscope (SEM) (Fig. 4). The cross-sectional SEM image appears to indicate an interface layer weaker than the rest of the coating. This, however, is an artefact arising during transverse cutting of the coated sample when some of the material at the interface got dislodged. Profile in Fig. 5 shows a typical surface profiler scan taken across the coated surface. Typical roughness of the coating ranged between w10 and 15 mm. Fig. 5 confirms the maximum coating thickness to be w75 mm, as was also estimated from the SEM images (Fig. 4). Fig. 6a and b depicts top view of the coated sample showing surface micro-structure of the PLD deposited alumina coating at different magnification levels. It is observed from these images that the deposited coating has a high degree of compaction and homogeneity and only a few open pores are seen of size w2 mm on the coating top surface. These pores are much smaller than those present in plasma spray coatings in which pores of w10 mm are commonly observed [24,25]. Absence of residual large size surface pores in these PLD alumina coatings suggests that such coatings can serve as effective corrosion protective coatings against molten metal [26,27]. Crystalline phase of the PLD alumina coating was investigated by X-ray diffraction technique (XRD) employing Cu-Ka radiation. Fig. 7 shows the results of XRD for a typical deposited alumina coating. While observed peaks at 2q angles equal to 43.56 and 35.12 correspond to (113) and (104) plane orientations for aalumina phase (JCPDS file no. 42-1468), respectively [28], a small peak corresponding to k-alumina phase at 2q equal to 28.76 [14,29] was also evident (JCPDS file no. 04-0878). Additional diffraction peaks observed belong to the stainless steel substrate. These XRD results suggest that the PLD deposited coatings
consisted predominantly of a-alumina phase with a minor fraction of k-alumina. Crystalline a-alumina has high hardness and excellent wear resistance and is thus the most desirable phase of alumina in coatings for tribological applications [13]. Our observations demonstrate the unique advantage associated with PLD technique which enables deposition of a-alumina coating even when the substrate is held at room temperature, whereas, all the other film deposition techniques require higher substrate temperatures, typically >760 C [14] for growth of a-alumina phase. High kinetic energy of species ejected during the process of laser ablation stimulates growth of a-alumina in such PLD coatings. When these energetic particles reach the substrate surface and transfer their kinetic energy to the substrate growth of a-alumina phase is
Fig. 7. XRD spectra of the alumina coating deposited by PLD technique.
1450
A.K. Singh et al. / Materials Chemistry and Physics 143 (2014) 1446e1451
Fig. 8. Variation of penetration depth with applied normal force and position on the coated surface during scratch test.
Fig. 9. Dependence of frictional force and friction coefficient as a function of applied force on the indenter during scratch test.
facilitated. Our observations are also in agreement with the earlier reports [14] where energy deposited through energetic neutral and ion bombardment assisted growth of dense k-alumina phase during conventional magnetron sputtering technique. A quantitative measure of extent of adhesion of PLD alumina coating on substrate was obtained by performing scratch test (Model No.: RST S/N: 27-00497) on samples. Results of the scratch test are shown in Figs. 8 and 9. In Fig. 8 is shown variation of penetration depth of the micro-indenter (radius: 200 mm) with
(b)
DTA signal (µV)
(a)
increasing applied normal force at various positions on the coating surface during the scratch test. Although, penetration depth does not show a monotonically increasing trend with increased applied normal force on indenter, a few random local maxima were observed. The extent of variation in penetration depth shown in Fig. 8 is observed to be of the same order as the surface roughness of the deposited coating measured by surface profiler (Fig. 5). Therefore, the observed variation in penetration depth (Fig. 8) is most likely a reflection of the surface roughness of the deposited
Fig. 10. Results of DTA test to study corrosion resistance of (a) PLD alumina coated SS substrate and (b) uncoated SS substrate against molten uranium.
A.K. Singh et al. / Materials Chemistry and Physics 143 (2014) 1446e1451
coating rather than a measure of actual penetration of the indenter into the coating material due to the applied normal force. Fig. 9 shows frictional coefficient of the coating as a function of applied normal force on the indenter during scratch test. Peaks in friction coefficient of the coating were observed for values of normal force around 3.8 N, 9 N, 13 N and 28.5 N. However, these peaks could not be correlated to simultaneous occurrences of increased penetration of indenter into the coating (Fig. 8). Thus, we conclude that these peaks in frictional force arise on account of the residual surface roughness of the deposited coating rather than initiation of a scratch generated by the micro-indenter. Acoustic emission signal was also recorded during scratch test. However, observed peaks in the acoustic emission signal provided no indication of development of cracks in the PLD coating. For instance, acoustic signal recorded at a load of around 3.8 N appeared to be a local phenomenon owing to the presence of a dent in the coating itself prior to scratch test rather than an associated damage or crack produced by the micro-indenter. These conclusions were also confirmed by observing the coating under a microscope during the scratch test. Hence no meaningful and conclusive information could be extracted from the acoustic signals recorded during scratch test. In order to establish the corrosion resistance potential of the deposited alumina coating against molten uranium, DTA tests were performed. In this test solid uranium was held on the test sample (alumina coated SS substrate) and subjected to a thermal cycle while housed inside a furnace. The test sample and a reference sample were both taken through the heating cycle in which their temperature was steadily raised to 1165 C, melting point of uranium being 1132 C. Thereafter, temperature of the samples was maintained at 1165 C for 2 h thereby allowing molten uranium to react with the test sample. Subsequently, the temperature of the furnace was gradually reduced and brought back to room temperature. During this entire thermal cycle DTA signals were recorded as shown in Fig. 10(a). No peak or dip, which are indicative of occurrence of exothermic and endothermic reactions were observed during this DTA run. This implies that molten uranium does not react with SS substrate when coated with alumina even when held at an elevated temperature of 1165 C for a period of 2 h. On the other hand, similar tests conducted on uncoated SS substrates showed many peaks in the DTA signal (Fig. 10(b)) signifying chemical reactions between liquid uranium and SS. Our results thus, suggest that PLD alumina coating over SS substrate does have a promising potential to serve as a good corrosion protective layer against molten uranium and other such reactive reagents. 4. Conclusion Alumina coatings have been successfully deposited on surface roughened SS substrates by PLD technique in atmospheric air environment with w3% average deposition efficiency. The deposited coatings were observed to predominantly contain a-alumina phase which is the most stable and dense phase of alumina, preferred for a wide variety of applications. These alumina coatings
1451
have shown very good compaction and homogeneity, adhesion to substrate and excellent corrosion resistance against molten uranium at 1165 C. Our observations suggest that such PLD alumina coatings have a promising potential to effectively serve as high temperature corrosion resistant coatings. Acknowledgments The authors would like to thank Dr. P.V.A. Padmanabhan, L&PTD, BARC, for providing sintered alumina pellets and surface roughened SS substrates, Mr. Lalchand, PED, BARC for surface profiling of the deposited alumina coatings and Mr. P. T. Rao, PMD, BARC for providing micro balance. References [1] Jeffrey T. Cheung, in: D.B. Chrisey, G.K. Hubler (Eds.), Pulsed Laser Deposition of Thin Films, Wiley Interscience, New York, 1994, p. 1. [2] Michael N.R. Ashfold, Frederick Claeyssens, Gareth M. Fuge, Simon J. Henley, Chem. Soc. Rev. 33 (2004) 23. [3] Li-Chyng Chen, in: D.B. Chrisey, G.K. Hubler (Eds.), Pulsed Laser Deposition of Thin Films, Wiley Interscience, New York, 1994, p. 167. [4] Li Guan, Duan Ming Zhang, Xu Li, ZhiHua Li, Nucl. Instrum. Methods Phys. Res. B 266 (2008) 57. [5] J.H. Clark, G.B. Donaldson, IEEE Trans. Appl. Supercond. 5 (1995) 1661. [6] H.M. Christen, G. Eres, J. Phys. Condens. Matter 20 (2008) 264005. [7] T.J. Jackson, S.B. Palmer, J. Phys. D Appl. Phys. 27 (1994) 1581. [8] J. Greer, in: R. Eason (Ed.), Pulsed Laser Deposition of Thin Films: Applicationsled Growth of Functional Materials, Wiley Interscience, New York, 2007, p. 281. [9] P.H. Lissberger, Rep. Prog. Phys. 33 (1970) 197. [10] P. Louda, J. Achieve. Mater. Manuf. Eng. 24 (2007) 51. [11] Takayuki Terai, J. Nucl. Mater. 248 (1997) 153. [12] Jeffry J. Haslam, Joseph C. Farmer, Robert W. Hopper, Keith R. Wilfinger, Metall. Mater. Trans. A 36A (2005) 1085. [13] Aditya Aryasomayajula, N.X. Randall, M.H. Gordon, D. Bhat, Thin Solid Films 517 (2008) 819. [14] Jochen M. Schneider, William D. Sproul, Allan Matthews, Surf. Coat. Technol. 94-95 (1997) 179. [15] J. Hlavac, Pure Appl. Chem. 54 (1982) 68l. [16] A. Kobayashi, G. Shanmugavelayutham, S. Yano, Solid State Phenom. 127 (2007) 313. [17] N. Hegazy, M. Shoeib, Sh. Abdel-Samea, H. Abdel-Kader, in: 13th International Conference on Aerospace Science and Aviation Technology, 2009. ASAT-13MS-14. [18] P. Souza Santos, H. Souza Santos, S.P. Toledo, Mater. Res. 3 (2000) 104. [19] S. Cava, S.M. Tebcherani, I.A. Souza, S.A. Pianaro, C.A. Paskocimas, E. Longo, J.A. Varela, Mater. Chem. Phys. 103 (2007) 394. [20] A. Mele, A. Giardini Guidoni, R. Kelly, A. Miotello, S. Orlando, R. Teghil, C. Flamini, Nucl. Instrum. Methods Phys. Res. B 116 (1996) 257. [21] D.B. Chrisey, J.S. Horwitz, R.E. Leuchtner, Thin Solid Films 206 (1991) 111. [22] V.I. Konov, T.V. Kononenko, E.N. Loubnin, F. Dausinger, D. Breitling, Appl. Phys. A 79 (2004) 931. [23] Michael E. Brown, Introduction to Thermal Analysis: Techniques and Applications, second ed., Springer, 2001. [24] Yourong Liu, Traugott E. . Fischer, Andrew Dent, Surf. Coat. Technol. 167 (2003) 68. [25] Huang Chen, Xiaming Zhou, Chuanxian Ding, J. Eur. Ceram. Soc. 23 (2003) 1449. [26] Sucharita Sinha, T.R.G. Kutty, P.V.A. Padmanabhan, K.G.K. Warrier, J. Laser Appl. 21 (3) (2009) 149. [27] A.K. Singh, T.R.G. Kutty, Sucharita Sinha, J. Nucl. Mater. 420 (2012) 374. [28] Frank R. Feret, Daniel Roy, Clermont Boulanger, Spectrochim. Acta Part B 55 (2000) 1051. [29] L. Favaro, A. Boumaza, P. Roy, J. Ledion, G. Sattonnay, J.B. Brubach, A.M. Huntz, R. Tetot, J. Solid State Chem. 183 (2010) 901.
Contents lists available at ScienceDirect
Materials Chemistry and Physics journal homepage: www.elsevier.com/locate/matchemphys
Pulsed laser deposition of alumina coating for corrosion protection against liquid uranium A.K. Singh a, Santu Kaity b, Kulwant Singh c, J. Thomas a, T.R.G. Kutty b, Sucharita Sinha a, * a
Laser & Plasma Technology Division, Bhabha Atomic Research Centre, Mumbai, India Radiometallurgy Division, Bhabha Atomic Research Centre, Mumbai, India c Fusion Reactor Materials Section, Bhabha Atomic Research Centre, Mumbai, India b
h i g h l i g h t s Alumina films deposited on stainless steel via pulsed laser deposition (PLD) technique. PLD alumina films investigated as potential corrosion protective coatings. Deposited coatings have been characterized in terms of their microstructure and crystalline phase. Corrosion resistance of coatings against liquid uranium was tested. Results suggest PLD alumina films have a promising potential for containment of molten uranium.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 February 2013 Received in revised form 24 October 2013 Accepted 29 November 2013
Alumina coatings find wide applications as tribological coatings and as corrosion protective coatings for structural materials against chemical attack. We have investigated alumina coatings deposited on Stainless Steel (SS) substrates via pulsed laser deposition (PLD) technique. Characterization tests performed on these coatings including their compatibility with liquid uranium suggests alumina to be a potential candidate as a coating material for handling and containment of liquid uranium. We present here results of our detailed parametric study including dependence of average mass removal rate on laser fluence and ablation geometry and average deposition efficiency during PLD. These measurements provide vital inputs facilitating proper choice of process parameters for PLD runs. Deposited coatings have been characterized in terms of their microstructure, surface profile, adhesion to substrate, crystalline phase and corrosion resistance against liquid uranium. Our PLD based alumina coatings have shown a high degree of compaction and excellent corrosion resistance to molten uranium even upto a temperature of 1165 C. Ó 2013 Elsevier B.V. All rights reserved.
Keywords: Coatings Ceramics Thin films Corrosion
1. Introduction Among various physical vapor deposition techniques, Pulsed Laser Deposition (PLD) is a simple and convenient method for deposition of films and coatings [1]. In this technique, focused laser radiation is used to vaporize material from a target and this ablated vapor is then deposited on a desired substrate [2]. The evaporation process of the target can be either thermal or non-thermal determined largely by the laser conditions. PLD technique has numerous advantages over other conventional techniques of film deposition. This technique enables deposition of coatings having precise stoichiometry and crystalline purity, offering at the same time,
* Corresponding author. Tel.: þ91 22 25595352; fax: þ91 22 25505151. E-mail addresses: [email protected], [email protected] (S. Sinha). 0254-0584/$ e see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.matchemphys.2013.11.063
flexibility and control over the deposition rate [1,2]. Properties such as surface morphology, crystalline phase and density of defects in deposited film can be tailored by controlling PLD process parameters such as, laser wavelength, fluence, pulse duration and pulse repetition rate, distance between substrate and target and the environment of deposition [3,4]. In addition, PLD technique allows facile process automation and is also environment friendly [5]. In PLD process congruent removal of material from the target leads to preservation of stoichiometry in the deposited film. Thus PLD is widely used for deposition of chemically inert and complex oxide films with accurate stoichiometry [1,6]. However, PLD technique does have its limitations too. Narrow angular spread and inhomogeneous flux of species within the ablated plume restricts scaling up of PLD technique for deposition of large area uniform coatings [7]. A technical difficulty with PLD arises on account of presence of macroscopic particulates (size w few micron) in the deposited film
A.K. Singh et al. / Materials Chemistry and Physics 143 (2014) 1446e1451
due to splashing effect [1,7]. Despite these drawbacks researchers have developed techniques to deposit large area coatings (up to 200 mm diameter substrates) having reduced density of macroscopic particulates [8]. Refractory materials, organic compounds, metals, dielectrics, semiconductors and superconductors, deposited as thin films over a wide range of substrates serve many applications. Thin film dielectric coatings are deposited on glass substrate for optical components such as optical wave-guides, reflecting and antireflecting coatings [9]. Various coatings on structural components provide protection against harsh environmental conditions. Ceramic coatings used in automobile industry lead to economical and ecological benefits through reduced weight of the components and enhanced service life of various parts of the vehicle [10]. Also, in nuclear technology corrosion resistant coatings are necessary for containment and handling of highly corrosive liquids such as molten uranium, actinide wastes, and liquid lithium blankets in fusion reactors [11,12]. Alumina is an excellent ceramic material, which is employed in the form of coating for a wide range of applications owing to its properties such as high hot hardness [13,14], high mechanical strength [13], high melting point (2053 C) [15] and good oxidation resistance [13,16]. Alumina also shows very good corrosion resistance against chemical attack [17]. Many polymorphic phases of alumina exist [18]. Of these polymorphs only a-alumina is thermodynamically stable above 1050 C and it is also the densest amongst all polymorphs [19]. Therefore, coatings grown of polymorphs other than a-alumina catastrophically fail to provide corrosion protection at temperatures above 1050 C as irreversible conversion of the unstable polymorphs to the densest a-alumina occurs at these elevated temperatures resulting in formation of cracks and pores in the coating. These cracks and pores formed in the coating due to phase transformation allow penetration of the corrosive chemicals to the structural material. Hence for high temperature (above 1050 C) corrosion protection application, alumina coating having pure a-phase is essential. In the study reported here, coatings of a-alumina have been deposited on surface roughened stainless steel (SS) substrates, employing PLD technique at room temperature and ambient atmospheric pressure. The deposited coatings have been characterized in terms of their surface profile, surface microstructure, adhesion to substrate, crystalline phase and corrosion resistance against molten uranium. Parametric investigation of PLD process including dependence of mass removal rate on laser fluence and ablation geometry, variation of ablation depth with laser fluence and average deposition efficiency (ratio of the mass deposited on substrate to mass ablated from target during PLD) during the PLD runs have also been studied. Results of our investigation enable proper choice of PLD process parameters facilitating deposition of alumina coatings demonstrating excellent corrosion protection capability. 2. Experiment
1447
the region of 18e20 mJ, pulse duration of 6 ns (FWHM) at 10 Hz pulse repetition rate and a wavelength of 532 nm has been used to ablate the target. Evaporation of material from the target occurs largely normal to the target surface. Therefore, the target was irradiated with the laser beam incident at an angle of 45 . In this configuration, spatial overlap between the incident laser beam and the vapor plume could be reduced, hence restricting attenuation of the incident laser beam by the plume [20,21]. The substrate and the target surface were maintained parallel to each other, separation between the two being w3 mm. A schematic of our experimental setup is shown in Fig. 1. PLD technique for film deposition typically employs high vacuum conditions. However, our studies involved PLD carried out in atmospheric air. Not being restricted by a vacuum chamber meant greater experimental flexibility and technical advantages when coating substrates having complex geometrical shapes [22]. However, PLD in air results in restricted plume length on account of collisions between the vaporized particles and ambient atmosphere which consequently limits the maximum allowable separation between substrate and target. Separation between target and substrate being restricted lead to critical demands on precise scanning movement of the substrate ensuring at the same time, that the incident laser beam does not get blocked during scanning of the substrate. Therefore, substrate could be moved only over a distance of w3 mm in horizontal direction although movement in vertical direction had no such restrictions. In the present work, typically an area of w0.15 cm2 was coated at a time. By sequentially coating a portion of the substrate and then rotating the substrate to access a fresh uncoated region, SS substrates having diameter of 5e 6 mm could be completely PLD coated with alumina in our setup. For different ablation configurations average mass removal rate as a function of laser fluence was systematically investigated and average deposition efficiency (ratio of the mass deposited on substrate to mass ablated from the target) averaged over 18,000 laser pulses was measured using a micro-balance with measuring accuracy of 10 mg. Microstructure of the deposited coatings and their crystalline phases were investigated using a Scanning Electron Microscope (SEM) and X-ray Diffraction (XRD) technique, respectively. Adhesion of the coating to the substrate was tested by adhesive tape test and scratch tests employing a diamond indenter of 200 mm diameter. Corrosion resistance of the deposited coating against molten uranium has been characterized using Differential Thermal Analysis (DTA) technique [23]. DTA signal, which is a measure of temperature difference between a reference sample and the test sample when subjected to identical thermal cycles, was recorded. Observed peaks and dips in DTA signal provide signatures of phase transitions and chemical reactions in the test specimen. A laser fluence of 11 J cm2 has been typically employed for ablation of Al2O3 target and subsequent deposition of these ejected species on surface roughened SS substrates. M1 Nd: YAG Laser X
For our PLD studies sintered alumina pellets (diameter 10 mm and thickness 10 mm) have been used as targets and surface roughened SS discs as substrates. Surface roughening via sandblasting of substrate improves mechanical adhesion of the coating with substrate surface. Both, the target and substrate were mounted on individual translational stages in order to have independent translation of both the target and the substrate maintaining, at the same time, a constant separation between the two. In this manner, deposition of coatings of uniform thickness over large area of the substrate was achieved. A focused beam of Nd:YAG nanosecond (ns) laser delivering laser pulses typically having pulse energy in
Y
Z
Target
M2 L
Plume Substrate Fig. 1. Schematic of the PLD setup with folding mirrors M1, M2 and convex lens (L) having focal length of 50 cm. Target and substrate both have been translated in the Xe Y plane while rotation is about z-axis.
1448
A.K. Singh et al. / Materials Chemistry and Physics 143 (2014) 1446e1451
3. Results and discussion In order to arrive at an optimum laser fluence for PLD of alumina, we first measured the average mass removal rate per pulse from alumina target as a function of incident laser fluence. Three different target surface configurations were studied: (a) laser beam incident at 45 with respect to scanned target surface, (b) normal incidence of laser beam on stationary target surface, (c) laser beam at normal incidence on scanned target surface. Average mass removal per pulse for each laser fluence value has been estimated by measuring the cumulative mass removed from the target on irradiation with a total of 18,000 laser pulses. Fig. 2 shows average mass removal rate for varying incident laser fluence for the three different configurations described above. From Fig. 2 it is evident that mass removal rate is highest when laser beam is incident at 45 to the target surface and the target surface is also translated using a stepper motor controlled translational stage. With the laser beam incident at 45 to target surface, spatial overlap of laser beam and the generated plume of the ablated species is minimized thus reducing resultant attenuation and distortion of the incident laser beam by the vapor plume. Scanning of the target also reduces formation of craters on the surface owing to material removal through laser ablation. Crater formation needs to be avoided since it alters the direction of the ablation plume and also leads to increased scattering of the laser beam from the target surface. This results in reduced coupling of laser energy into the target. During our investigation the target was moved at a speed of 0.02 cm s1 sequentially in X and Y directions. Measured dependence of ablation depth of the crater formed on target surface when irradiated with 600 laser pulses with increasing incident laser fluence is shown in Fig. 3. For these measurements the target surface was maintained static and normal with respect to incident laser beam. Ablation depth has been measured using a surface profiler (Model No: Taylor Hobson, Form Talysurf Series 2). Fig. 3 clearly shows that the ablation depth steadily increases with increasing laser fluence till a maximum crater depth is reached. Beyond this value of laser fluence ablation depth shows a decreasing trend with further increase in laser fluence. Uneven surface and deep crater formation results in grazing incidence of laser beam on target. This not only reduces the effective laser fluence experienced by the target but also enhances loss of laser energy through multiple scattering on crater walls. Based on our observations on ablation rate from alumina target a laser fluence value of 11 J cm2 was chosen for our PLD runs. At this fluence level, ablation plume w3 mm long, were typically
Fig. 2. Average mass removal per pulse as a function of incident laser fluence under different ablation geometries: (a) laser beam incident at 45 with respect to scanned target surface (b) normal incidence of laser beam on stationary target surface (c) normal incidence of laser beam on scanned target surface.
Fig. 3. Variation of ablation depth with incident laser fluence.
generated. The time averaged length of the visible plume was estimated via viewing using suitable optical filters and attenuators to avoid saturation. Although longer plume lengths are preferred since they allow experimental flexibility while positioning the target and substrate for PLD, ablation at laser flux levels far above ablation threshold often cause explosive type of ejection of large fragments from the target. Deposition of such massive fragments needs to be avoided as it degrades the quality of the deposited film. Thus, optimization of laser fluence employed for PLD not only helps in achieving high deposition rate, but also minimizes presence of macro particulates in the coating hence improving the quality of deposited film. Alumina coatings having a maximum thickness w75 mm have been deposited on surface roughened SS substrate by using PLD technique. Average efficiency for coating deposition was estimated to be w3% for a cumulative 7 h PLD run when average mass removal per pulse was measured to be w205 ng/pulse. The drop in average mass removal rate measured for extended PLD runs (7 h) in comparison to short duration runs (half hour) shown in Fig. 2, is largely on account of increased surface degradation of target during the extended PLD runs. Adhesion of the coating with substrate was tested using adhesive tape test. In this test a strip of adhesive tape was applied to the coated surface and rapidly pulled off. No peeling of the coating occurred when the adhesive tape was pulled off indicating good mechanical adhesion of the deposited coating with the substrate. Thickness of the coating was measured by taking a transverse cut through the coated surface and subsequently
Fig. 4. SEM image of a typical transverse segment cut through PLD alumina coated SS substrate.
A.K. Singh et al. / Materials Chemistry and Physics 143 (2014) 1446e1451
1449
Fig. 5. Surface profile of the deposited alumina coating.
Fig. 6. SEM image showing surface microstructure of the alumina coating at two different levels of magnification.
viewing the cross section under a Scanning Electron Microscope (SEM) (Fig. 4). The cross-sectional SEM image appears to indicate an interface layer weaker than the rest of the coating. This, however, is an artefact arising during transverse cutting of the coated sample when some of the material at the interface got dislodged. Profile in Fig. 5 shows a typical surface profiler scan taken across the coated surface. Typical roughness of the coating ranged between w10 and 15 mm. Fig. 5 confirms the maximum coating thickness to be w75 mm, as was also estimated from the SEM images (Fig. 4). Fig. 6a and b depicts top view of the coated sample showing surface micro-structure of the PLD deposited alumina coating at different magnification levels. It is observed from these images that the deposited coating has a high degree of compaction and homogeneity and only a few open pores are seen of size w2 mm on the coating top surface. These pores are much smaller than those present in plasma spray coatings in which pores of w10 mm are commonly observed [24,25]. Absence of residual large size surface pores in these PLD alumina coatings suggests that such coatings can serve as effective corrosion protective coatings against molten metal [26,27]. Crystalline phase of the PLD alumina coating was investigated by X-ray diffraction technique (XRD) employing Cu-Ka radiation. Fig. 7 shows the results of XRD for a typical deposited alumina coating. While observed peaks at 2q angles equal to 43.56 and 35.12 correspond to (113) and (104) plane orientations for aalumina phase (JCPDS file no. 42-1468), respectively [28], a small peak corresponding to k-alumina phase at 2q equal to 28.76 [14,29] was also evident (JCPDS file no. 04-0878). Additional diffraction peaks observed belong to the stainless steel substrate. These XRD results suggest that the PLD deposited coatings
consisted predominantly of a-alumina phase with a minor fraction of k-alumina. Crystalline a-alumina has high hardness and excellent wear resistance and is thus the most desirable phase of alumina in coatings for tribological applications [13]. Our observations demonstrate the unique advantage associated with PLD technique which enables deposition of a-alumina coating even when the substrate is held at room temperature, whereas, all the other film deposition techniques require higher substrate temperatures, typically >760 C [14] for growth of a-alumina phase. High kinetic energy of species ejected during the process of laser ablation stimulates growth of a-alumina in such PLD coatings. When these energetic particles reach the substrate surface and transfer their kinetic energy to the substrate growth of a-alumina phase is
Fig. 7. XRD spectra of the alumina coating deposited by PLD technique.
1450
A.K. Singh et al. / Materials Chemistry and Physics 143 (2014) 1446e1451
Fig. 8. Variation of penetration depth with applied normal force and position on the coated surface during scratch test.
Fig. 9. Dependence of frictional force and friction coefficient as a function of applied force on the indenter during scratch test.
facilitated. Our observations are also in agreement with the earlier reports [14] where energy deposited through energetic neutral and ion bombardment assisted growth of dense k-alumina phase during conventional magnetron sputtering technique. A quantitative measure of extent of adhesion of PLD alumina coating on substrate was obtained by performing scratch test (Model No.: RST S/N: 27-00497) on samples. Results of the scratch test are shown in Figs. 8 and 9. In Fig. 8 is shown variation of penetration depth of the micro-indenter (radius: 200 mm) with
(b)
DTA signal (µV)
(a)
increasing applied normal force at various positions on the coating surface during the scratch test. Although, penetration depth does not show a monotonically increasing trend with increased applied normal force on indenter, a few random local maxima were observed. The extent of variation in penetration depth shown in Fig. 8 is observed to be of the same order as the surface roughness of the deposited coating measured by surface profiler (Fig. 5). Therefore, the observed variation in penetration depth (Fig. 8) is most likely a reflection of the surface roughness of the deposited
Fig. 10. Results of DTA test to study corrosion resistance of (a) PLD alumina coated SS substrate and (b) uncoated SS substrate against molten uranium.
A.K. Singh et al. / Materials Chemistry and Physics 143 (2014) 1446e1451
coating rather than a measure of actual penetration of the indenter into the coating material due to the applied normal force. Fig. 9 shows frictional coefficient of the coating as a function of applied normal force on the indenter during scratch test. Peaks in friction coefficient of the coating were observed for values of normal force around 3.8 N, 9 N, 13 N and 28.5 N. However, these peaks could not be correlated to simultaneous occurrences of increased penetration of indenter into the coating (Fig. 8). Thus, we conclude that these peaks in frictional force arise on account of the residual surface roughness of the deposited coating rather than initiation of a scratch generated by the micro-indenter. Acoustic emission signal was also recorded during scratch test. However, observed peaks in the acoustic emission signal provided no indication of development of cracks in the PLD coating. For instance, acoustic signal recorded at a load of around 3.8 N appeared to be a local phenomenon owing to the presence of a dent in the coating itself prior to scratch test rather than an associated damage or crack produced by the micro-indenter. These conclusions were also confirmed by observing the coating under a microscope during the scratch test. Hence no meaningful and conclusive information could be extracted from the acoustic signals recorded during scratch test. In order to establish the corrosion resistance potential of the deposited alumina coating against molten uranium, DTA tests were performed. In this test solid uranium was held on the test sample (alumina coated SS substrate) and subjected to a thermal cycle while housed inside a furnace. The test sample and a reference sample were both taken through the heating cycle in which their temperature was steadily raised to 1165 C, melting point of uranium being 1132 C. Thereafter, temperature of the samples was maintained at 1165 C for 2 h thereby allowing molten uranium to react with the test sample. Subsequently, the temperature of the furnace was gradually reduced and brought back to room temperature. During this entire thermal cycle DTA signals were recorded as shown in Fig. 10(a). No peak or dip, which are indicative of occurrence of exothermic and endothermic reactions were observed during this DTA run. This implies that molten uranium does not react with SS substrate when coated with alumina even when held at an elevated temperature of 1165 C for a period of 2 h. On the other hand, similar tests conducted on uncoated SS substrates showed many peaks in the DTA signal (Fig. 10(b)) signifying chemical reactions between liquid uranium and SS. Our results thus, suggest that PLD alumina coating over SS substrate does have a promising potential to serve as a good corrosion protective layer against molten uranium and other such reactive reagents. 4. Conclusion Alumina coatings have been successfully deposited on surface roughened SS substrates by PLD technique in atmospheric air environment with w3% average deposition efficiency. The deposited coatings were observed to predominantly contain a-alumina phase which is the most stable and dense phase of alumina, preferred for a wide variety of applications. These alumina coatings
1451
have shown very good compaction and homogeneity, adhesion to substrate and excellent corrosion resistance against molten uranium at 1165 C. Our observations suggest that such PLD alumina coatings have a promising potential to effectively serve as high temperature corrosion resistant coatings. Acknowledgments The authors would like to thank Dr. P.V.A. Padmanabhan, L&PTD, BARC, for providing sintered alumina pellets and surface roughened SS substrates, Mr. Lalchand, PED, BARC for surface profiling of the deposited alumina coatings and Mr. P. T. Rao, PMD, BARC for providing micro balance. References [1] Jeffrey T. Cheung, in: D.B. Chrisey, G.K. Hubler (Eds.), Pulsed Laser Deposition of Thin Films, Wiley Interscience, New York, 1994, p. 1. [2] Michael N.R. Ashfold, Frederick Claeyssens, Gareth M. Fuge, Simon J. Henley, Chem. Soc. Rev. 33 (2004) 23. [3] Li-Chyng Chen, in: D.B. Chrisey, G.K. Hubler (Eds.), Pulsed Laser Deposition of Thin Films, Wiley Interscience, New York, 1994, p. 167. [4] Li Guan, Duan Ming Zhang, Xu Li, ZhiHua Li, Nucl. Instrum. Methods Phys. Res. B 266 (2008) 57. [5] J.H. Clark, G.B. Donaldson, IEEE Trans. Appl. Supercond. 5 (1995) 1661. [6] H.M. Christen, G. Eres, J. Phys. Condens. Matter 20 (2008) 264005. [7] T.J. Jackson, S.B. Palmer, J. Phys. D Appl. Phys. 27 (1994) 1581. [8] J. Greer, in: R. Eason (Ed.), Pulsed Laser Deposition of Thin Films: Applicationsled Growth of Functional Materials, Wiley Interscience, New York, 2007, p. 281. [9] P.H. Lissberger, Rep. Prog. Phys. 33 (1970) 197. [10] P. Louda, J. Achieve. Mater. Manuf. Eng. 24 (2007) 51. [11] Takayuki Terai, J. Nucl. Mater. 248 (1997) 153. [12] Jeffry J. Haslam, Joseph C. Farmer, Robert W. Hopper, Keith R. Wilfinger, Metall. Mater. Trans. A 36A (2005) 1085. [13] Aditya Aryasomayajula, N.X. Randall, M.H. Gordon, D. Bhat, Thin Solid Films 517 (2008) 819. [14] Jochen M. Schneider, William D. Sproul, Allan Matthews, Surf. Coat. Technol. 94-95 (1997) 179. [15] J. Hlavac, Pure Appl. Chem. 54 (1982) 68l. [16] A. Kobayashi, G. Shanmugavelayutham, S. Yano, Solid State Phenom. 127 (2007) 313. [17] N. Hegazy, M. Shoeib, Sh. Abdel-Samea, H. Abdel-Kader, in: 13th International Conference on Aerospace Science and Aviation Technology, 2009. ASAT-13MS-14. [18] P. Souza Santos, H. Souza Santos, S.P. Toledo, Mater. Res. 3 (2000) 104. [19] S. Cava, S.M. Tebcherani, I.A. Souza, S.A. Pianaro, C.A. Paskocimas, E. Longo, J.A. Varela, Mater. Chem. Phys. 103 (2007) 394. [20] A. Mele, A. Giardini Guidoni, R. Kelly, A. Miotello, S. Orlando, R. Teghil, C. Flamini, Nucl. Instrum. Methods Phys. Res. B 116 (1996) 257. [21] D.B. Chrisey, J.S. Horwitz, R.E. Leuchtner, Thin Solid Films 206 (1991) 111. [22] V.I. Konov, T.V. Kononenko, E.N. Loubnin, F. Dausinger, D. Breitling, Appl. Phys. A 79 (2004) 931. [23] Michael E. Brown, Introduction to Thermal Analysis: Techniques and Applications, second ed., Springer, 2001. [24] Yourong Liu, Traugott E. . Fischer, Andrew Dent, Surf. Coat. Technol. 167 (2003) 68. [25] Huang Chen, Xiaming Zhou, Chuanxian Ding, J. Eur. Ceram. Soc. 23 (2003) 1449. [26] Sucharita Sinha, T.R.G. Kutty, P.V.A. Padmanabhan, K.G.K. Warrier, J. Laser Appl. 21 (3) (2009) 149. [27] A.K. Singh, T.R.G. Kutty, Sucharita Sinha, J. Nucl. Mater. 420 (2012) 374. [28] Frank R. Feret, Daniel Roy, Clermont Boulanger, Spectrochim. Acta Part B 55 (2000) 1051. [29] L. Favaro, A. Boumaza, P. Roy, J. Ledion, G. Sattonnay, J.B. Brubach, A.M. Huntz, R. Tetot, J. Solid State Chem. 183 (2010) 901.