The influence of impurities on the formation of protective aluminium oxides on RuAl thin films
6
0
0
Texto completo
(2) 166. M.A. Guitar et al. / Journal of Alloys and Compounds 594 (2014) 165–170. can later segregate at grain boundaries. These impurities if added to the oxide layer could affect the transport rates of oxygen and/ or aluminium, altering the Al2O3 growth process [17]. Oxidation of RuAl thin films deposited onto stainless steel (SS) has been previously studied [3] isothermally at 750 °C and 900 °C, resulting in diffusion of Fe and Cr from the substrate into the RuAl film. Several works [18–21] have studied the influence of the Fe and Cr content in the formation of a-Al2O3 from one of its transient oxides. Most of the studies found that the presence of Cr in some alumina-forming alloys and intermetallic compounds allowed the growth of a-Al2O3 instead of transient oxides. Although it is clear that the oxidation behaviour in intermetallic compounds is affected by the presence of reactive elements and/or impurities, there is still a lack of information on the aforementioned system regarding oxidation kinetics and stress. The present work studied the oxidation behaviour of RuAl thin films synthesized by physical vapour deposition (PVD) onto SS and Al2O3 substrates. Austenitic SS was selected due to its technical relevance, since it is usually used as a moulding material in the manufacturing of precision lenses [22,23]. Impurities from this substrate can diffuse through the RuAl upper layer, although Al2O3 substrates avoid the diffusion and thus, the influence of the impurity content on the oxidation behaviour of RuAl thin films could be evaluated. The oxidation kinetic was analysed using an Arrhenius model, whereas microstructural and stress analyses were performed using scanning transmission electron microscopy (STEM) and X-ray diffraction (XRD). 2. Experimental methods The RuAl thin films were grown on AISI 316L SS and Al2O3 substrates by PVD. Ru–Al multilayers were deposited using AC/DC magnetron sputtering and then vacuum annealed at 600 °C during 1 h in order to form single-phase RuAl, as described in more detail elsewhere [24]. The RuAl thin films studied were off-stoichiometric compounds, and the composition before oxidation (Table 1) was measured by energy dispersive X-ray spectroscopy (EDX). It is worth noting that a certain amount of Fe and Cr impurities were present in the RuAl on SS. After thin film synthesis, isothermal oxidation of the RuAl was carried out in air at 900 °C for times ranging from 10 to 60 minutes. Two groups of samples were analysed: (a) RuAl on SS (RuAl/SS) and (b) RuAl on Al2O3 (RuAl/Al2O3), both with a chemical composition shifted towards the Ru-rich side (Table 1). The activation energy for oxidation was calculated by oxidising the samples for 20 min between 750 °C and 1000 °C, with a 50 °C step. The samples were placed in the furnace at the oxidation temperature, maintained the desired time and removed to cool down to room temperature. No special atmosphere was used. The microstructure of the samples was analysed in a dual-beam focused ion beam/scanning electron microscopy (FIB/SEM) workstation (FEI Helios NanoLab 600), which also allowed STEM imaging (30 kV) and EDX. The STEM specimens were prepared by FIB milling. The oxide phase formed on the RuAl thin films was studied by XRD using Cu Ka radiation (k = 0.1542 nm) at 40 kV and 40 mA in a PANalytical X’Pert Pro MPD diffractometer. The diffraction geometry was h–2h, and the incident and reflected optics consisted of parallel beams. Stress measurements were performed by X-ray stress analysis (XSA) through the sin2 w method using Cr Ka radiation (k = 0.2294 nm) at 40 kV and 40 mA in a PANalytical X’Pert Pro MRD. This radiation allows less penetration depth than Cu Ka in RuAl and, therefore the information volume is more surface sensitive. The residual stress of RuAl layers after growth and before oxidation was estimated multiple-{hkl} sin2 w stress analysis from GI-XRD measurements at room temperature [25], with help of the PANalytical X’Pert Stress Plus software v2.1.. 3. Results and discussion 3.1. Microstructural analysis Despite the off-stoichiometry (Table 1), the oxidised films were within the RuAl phase domain and no indication of Ru-rich secondary phases was detected by XRD. The RuAl microstructure of both samples types, RuAl/SS and RuAl/Al2O3, after 50 min of oxidation is shown in the STEM micrographs of Fig. 1. A regular oxide layer formed in the surface of the sample, which was dense and compact, and covered the whole sample. During oxidation, the growth. Table 1 Chemical composition of the RuAl film deposited onto SS and Al2O3 substrates, before oxidation. The error in the measurement corresponds to the EDX instrument, which is around 2%.. Ru/Al Ru Al Fe Cr. RuAl/SS. RuAl/Al2O3. 1.12 51 at.% 45 at.% 3 at.% 1 at.%. 1.12 53 at.% 47 at.% – –. of the a-Al2O3 scale was accompanied by the formation of a new Al-depleted phase beneath the oxide scale, due to the relative narrow existence range of the single-phase RuAl [26], which in this case was identified as d-Ru (hcp), as observed in the diffractogram of Fig. 2. No evidence of the formation of transient alumina has been detected, neither for the different oxidation times studied in the present work nor for lower temperatures (750 °C) [3]. Intermetallic compounds usually oxidise selectively, forming a protective surface oxide and a layer depleted of the most reactive element. In the case of single-phase RuAl [2,27], Al is the most reactive element and thus diffuses towards the surface to react with oxygen, forming an Al2O3 scale at the metal/oxygen interface. Simultaneously, a Ru-rich layer is formed beneath due to the depletion of Al. The thickness of this Al-depleted zone and the degree of Al-depletion determine the stability of the oxide scale during prolonged oxidation. The long-term stability requires a continuous flux of solute to the intermetallic/oxide interface. As showed in Fig. 1, the Al-depleted layer of both, RuAl/Al2O3 and RuAl/SS samples was compact and continuous, but differed in thickness. Moreover, the presence of impurities (Fe and Cr) helped the formation of the larger grains observed in the RuAl/SS sample in both, the Al-depleted zone and the remaining RuAl phase [24]. As shown in Fig. 1, the oxide layer formed on the RuAl/SS samples that oxidised for 60 min had the same morphology as in those oxidised for 50 min; however, the microstructure of the RuAl/Al2O3 sample showed a completely different microstructure. In the region close to the surface, only the a-Al2O3 scale was observed and the RuAl layer had transformed completely into a mixture of both Al2O3 (white areas) and Ru-rich layers (dark areas). Transmission electron microscopy EDX showed that the Al-depleted zone beneath the oxide layer in the oxidised RuAl/SS sample contained 20 at.% Cr and 62 at.% Fe, and the remaining RuAl phase 10 at.% Cr and 17 at.% Fe after oxidising 60 min at 900 °C. The high concentration of Cr and Fe detected at the Al-depleted layer of the RuAl/SS sample was the result of the diffusion of elements from the substrate. These elements were usually found in the form of solid solution or located at the grain boundaries. The presence of impurities might be detrimental for the oxidation resistance of an alloy and could also affect the transport of the aluminium to the intermetallic/oxide and/or to the oxide/gas interface. In the same manner, the oxidation kinetics could also be [3]. The new phase formed beneath the protective oxide scale influenced its subsequent stability, as well as its adherence [28]. Stresses generated at the oxide/Al-depleted zone due to the oxide growth and the d-Ru formation are supposed to enhance the formation of cracks in the d-Ru, especially at the thin zones of the layer [27,29]. The mechanisms that produce the discontinuous formation of the Al2O3 layer, as observed in the RuAl/Al2O3 sample oxidised for 60 min, are not completely clear, but they might include repeated cracking and healing of the oxide scale during the oxidation process. Cracks or pores in the Al2O3 and d-Ru layer act as diffusion paths for oxygen, which is allowed to reach a new zone of fresh RuAl material. The oxygen–aluminium reaction results in the growth of an Al2O3 layer and simultaneously a new Al-depleted.
(3) M.A. Guitar et al. / Journal of Alloys and Compounds 594 (2014) 165–170. 167. Fig. 1. STEM micrographs of the oxidised RuAl films using both Al2O3 and SS substrates. The samples were oxidised at 900 °C for 50 min and 60 min.. Fig. 2. XRD (Cu radiation) pattern for the (left) RuAl/SS and (right) RuAl/Al2O3 samples. The diffractograms of the uncoated SS and Al2O3 substrates and for the unoxidised RuAl films are also shown in order to identify the signals from the substrates.. zone is formed, giving origin to a multilayer morphology, as observed and discussed by Bellina et al. [27] for bulk-RuAl and by Chou in IrAl [14]. The long-term oxidation behaviour of a high-temperature alloy is determined not only by the effectiveness of the oxide scale as a reactant transport barrier, but also by its resistance to mechanical failure. The protective Al2O3 scale formed at high temperature is an effective transport barrier for oxygen and metal cations, but is also susceptible to cracking and spalling during thermal cycling or during cooling from working conditions down to room temperature. The addition of small amounts of reactive elements reduces mechanical failures in thermally growth oxides [19,30], although the mechanisms of the oxidation enhancement or the oxide stability are still unclear. 3.2. Oxidation kinetics The growth kinetics of the a-Al2O3 was determined by measuring the thicknesses of the oxide scale for several oxidation times. using STEM images. Fig. 3 shows the scale thickness as a function of time for samples RuAl/SS and RuAl/Al2O3 oxidised at 900 °C. The point corresponding to the RuAl/Al2O3 sample oxidised for 3600 s was not included in the oxidation rate calculation due to its complex microstructure (Fig. 1) from which it was difficult to determine the thickness. The oxide thickness (x) as a function of time (t) was fitted using the linear rate law x = k t + C, whose adjusted parameters k and C are given in Table 2. The linear growth behaviour has been observed [31,32] at the beginning of the scale formation for thicknesses in the order of 104 cm, which matches the current scale measurements. Based on the theory of the scale kinetics [33,34], at this stage of oxide thickness, there is no thermal equilibrium between metal–oxide and oxide–gas interfaces, and thus phase-boundary processes is the rate-determining steps of the reaction, instead of ion diffusion through the scale, as in the case of the parabolic rate law. Since these phenomena can take place simultaneously, linear and parabolic laws might be confused at the first stages of oxidation. Indeed, the Al2O3 formation onto RuAl films has been related to a.
(4) 168. M.A. Guitar et al. / Journal of Alloys and Compounds 594 (2014) 165–170. Fig. 3. Oxide scale thickness as a function of time for samples RuAl/SS and RuAl/ Al2O3 oxidised at 900 °C. Linear fit was performed on both samples, whose adjusted parameters are resumed in Table 2.. Table 2 Linear fitting parameters obtained from the oxide scale thickness data as a function of time (Fig. 3). Sample. Fitting parameters. RuAl/SS RuAl/Al2O3. k (109 cm/s). C (106 cm). 4.9 ± 0.5 4.8 ± 0.3. 2.6 ± 1.2 2.1 ± 0.6. Fig. 4. Arrhenius plot of a-Al2O3 formation on RuAl thin films deposited onto SS and Al2O3 substrates for the temperature ranges 750 °C–1000 °C and 750 °C–950 °C, respectively.. growth mechanism, reduction in the oxidation rate and modification of the scale microstructure [30,35]. Furthermore, compressive stresses in the RuAl film tended to decrease the outward diffusion of solute elements, thus affecting the activation energy for oxidation [42]. Therefore, it was possible to conclude that Fe and Cr impurities from the SS substrate modified the kinetics of a-Al2O3 formation, acting as energy barriers for ion diffusion through the scale. 3.3. Stress analysis. parabolic growth kinetics elsewhere [3]. By comparing the parameters of Table 2, it is possible to conclude that the oxidation kinetics was not significantly affected by the presence of the impurities from the SS substrate. Nevertheless, a slightly higher oxidation rate was observed in the RuAl/SS samples, in accordance with the oxidation behaviour of aluminides in the presence of reactive elements [35,36]. The parameter C represents the initial scale thickness, which corresponds to the oxide layer after annealing at 600 °C and before the oxidation experiment. The activation energy for oxidation (Q) was estimated by measuring the growth rate of the oxide scale (G = Dx/Dt) at different temperatures. It was determined from the slope of the log G versus 1/T plot of Fig. 4 and is based on the Arrhenius model shown in Eq. (1) [37], where G0 is a constant, R is the universal gas constant and T is the temperature in degrees Kelvin. The activation energy of a-Al2O3 on RuAl/Al2O3 (no impurities) was 90 ± 12 kJ/mol, whereas on RuAl/SS (Cr and Fe impurities) was 116 ± 8 kJ/mol.. InðGÞ ¼ InðG0 Þ þ. Q RT. ð1Þ. The activation energy of aluminium oxide on RuAl/SS was 30% more than RuAl/Al2O3 (fit error of 7%), indicating that the presence of impurities influenced the kinetic behaviour. Besides the previous work [3], no available data of activation energy for oxidation were found in the literature for RuAl bulk, or thin films. The values correlated with those of Pettit and Wagner [31] for iron oxidation (92 kJ/mol) in the same temperature range and in absence of impurities. Variations of activation energies and oxidation rates have been observed for other intermetallic compounds as a result of reactive elements and/or impurities content, mainly due to the agglomeration of foreign atoms at the grain boundaries [19,35,38–40]. Particularly, Zhong et al. [41] observed a decrease in the activation energy by adding AlN into NiAl films. Other expected effects of reactive impurities are changes in the oxide. As discussed previously, a-Al2O3 is susceptible to cracking and spalling due to thermal stresses generated during the oxide growth originating from CTE mismatches with temperature changes. Several mechanisms have been proposed to explain the origin of growth stresses [43]. In order to quantify the effect of thermal stress due to the scale–alloy CTE mismatch and due to the intrinsic growth stress at grain boundaries, residual stresses were measured using XSA. The sin2 w plot method was applied using in the biaxial stress approximation. The stress generated in the a-Al2O3 phase d d was determined by measuring the strain ew ¼ wd0 0 in the crystal lattice for different w directions perpendicular to the interplanar spacing (dw) of the selected {hkl} planes, as presented in Eq. (2), where S1 and ½S2 are the X-ray elastic constants. The stress-free lattice spacing (d0) was obtained for the w position where the strain was null [44]. In the sin2 w method, the stress r was calculated from the slope of the dw vs sin2 w plot whereas S1 and ½S2 were calculated from the Cij constants [45] for a-Al2O3 using the Reuss approximation.. ew ¼. dw d0 1 hkl 2 ¼ S2 sin w þ 2Shkl 1 2 d0. ð2Þ. Eq. (3) gives the theoretical estimation of thermal stresses, where DT is the temperature difference between the treatment and room temperatures, whereas Da is the difference between the linear CTEs of the substrate and the film (Table 3).. eth ¼. Z. T. ½as af dT ffi ðDaÞðDTÞ. ð3Þ. To. Compressive stress was found in RuAl deposited onto SS (1.61 ± 0.04 GPa), whereas tensile stress was found for the Al2O3 substrate (0.51 ± 0.08 GPa). On the other hand, thermal stresses estimated using the CTE mismatch equation (Eq. (3)) had values of 1.9 GPa and 0.2 GPa for the RuAl/SS and RuAl/Al2O3 samples, respectively. Both experimental and theoretical values.
(5) M.A. Guitar et al. / Journal of Alloys and Compounds 594 (2014) 165–170. 169. Table 3 Coefficient of thermal expansion corresponding to the film and substrate materials. Element. CTE (K1). RuAl Al2O3 SS Ru. 6–8 106 5–6 106 16 106 5.1–9.6 106. agreed within 30% of accuracy. These stresses were generated during cooling after deposition, while intrinsic growth stresses were generated during formation of the single-phase RuAl. The sin2 w plot method using the reflections {214} and {300} of a-Al2O3 was performed to measure residual stresses of the oxide scale on the RuAl/Al2O3 sample, whereas {202} and {211} were used to measure on the RuAl/SS sample, due to peak convolution of the phase formed at the Al-depleted zone. Fig. 5 shows the reflections of the a-Al2O3 phase measured on the RuAl/Al2O3 samples before and after oxidation, using the same experimental conditions. Since the same XRD measurement time and optical parameters were used to measure on RuAl/Al2O3 samples before and after oxidation, the increase in the peak intensities of reflections {214} and {300} is due to the presence of the a-Al2O3 scale at the surface of the sample. The averaged strains of the the RuAl/Al2O3 unoxidised and oxidised samples, and for the oxidised RuAl/SS samples are shown in Fig. 6. Since the strain in the Al2O3 substrate was found to be within the measurement error, it be was considered negligible and did not affect the values obtained in the oxidized samples. On the other hand, the oxide scale grown on top of the RuAl/ Al2O3 sample showed tensile stresses (51.1 ± 10.5 MPa), whereas that formed at the RuAl/SS surface had compressive stresses (1277 ± 107 MPa). Stresses generated during oxidation can be relaxed by several mechanisms, among others, by cracking and/or plastic deformation [33,43]. Oxide cracking occurs when the oxides acquire tensile stresses [33], which has severe consequences for the material stability, because they can expose fresh material to the oxidising atmosphere, as is the case of the RuAl/Al2O3 samples oxidised for 60 minutes (Fig. 1). If the elastic strain energy stored in the scale exceeds the fracture resistance of the interface due to compressive stresses, spalling of the oxide might occur [33]. However, if the metal beneath the scale can deform plastically, the stresses can be relaxed by simultaneous deformation of the scale and alloy. Fig. 6. Sin2 w plots measured in the a-Al2O3 scale. The curves show the strain average weighted by the multiplicity factor of each reflection [47].. Fig. 5. XRD (Cr radiation) reflections corresponding to the RuAl/Al2O3 samples before and after oxidation. The increase of the peak intensities is due to the presence of the a-Al2O3 scale at the surface of the sample.. Based on the microstructural evidence of Fig. 1 and the measured stress values, this is the most probable mechanism acting in the oxidised RuAl/SS samples. The new alloy formed in these samples at the Al-depleted zones (containing high quantities of Fe and Cr) increases the ductility of the underlayer metal, avoiding spalling of the scale. Moreover, it is not yet clear if the Al-depleted zone still retains the hcp structure of the d-Ru phase, or if it acquires some of the structures of the Fe–Cr alloys. Indeed, due to the presence of Fe and Cr, this zone could also have a bcc and/or hcp structure according to the Cr–Ru and Fe–Ru alloy phase diagrams [46]. To the best of our knowledge, alloys and/or compounds containing Fe/Cr/Ru are not available in the literature; therefore mechanical and thermal properties cannot be compared to those of single d-Ru. For these reasons, further studies are needed to identify this phase and to understand its influence on the stability of the protective Al2O3 scale, with the aim of tailoring thermal and mechanical properties of the protective layer..
(6) 170. M.A. Guitar et al. / Journal of Alloys and Compounds 594 (2014) 165–170. 4. Concluding remarks In the present work, the oxidation behaviour of RuAl thin films deposited onto SS and Al2O3 substrates was analysed. Since impurities diffuse only from the SS substrate, their influence on the formation of protective aluminium oxides was evaluated in comparison with the films grown on Al2O3. After kinetic and microstructural analyses, it was concluded that the oxidation behaviour of non-stoichimetric RuAl films was affected by the presence of Fe and Cr impurities. The kinetics analysis showed that the activation energy of aluminium oxidation on RuAl/SS was around 30% more than on RuAl/Al2O3, indicating the influence of Fe (62 at.%) and Cr (20 at.%) impurities on the RuAl oxidation, mainly due to the generation of energy barriers at the grain boundaries. Furthermore, activation energy is increased by compressive stresses, due to a decreasing in the diffusion coefficient of solute elements inside the intermetallic compound. Cracking in the aAl2O3 oxide layer was observed in absence of impurities as a result of tensile stresses generated in the oxide scale during the oxidation. On the contrary, compressive stresses developed in the oxide scale after Fe and Cr diffusion into the RuAl film, enhancing the material stability at high-temperature. Finally, it is concluded that the control of the impurities and reactive elements content allows the potential tailoring of thermo-mechanical properties of protective oxides. Acknowledgements This study was funded within the research Project MU 959/24-1 of the Deutsche Forschungsgemeinschaft (DFG). This work was also supported by the SUMA2 Network Project, 7th Framework Program of the European Commission (IRSES Project No. 318903). The authors would like to thank the EFRE Funds of the European Commission for support of activities within the AME-Lab project. The authors are also grateful to Prof. Seidel, from the Department of Mechatronics, Saarland University, for the use of the magnetron-sputtering device; and to Dipl.-Ing. N. Souza and Dipl.-Ing. C. Pauly for the usefully comments and discussion. M.A. Guitar thanks to the German Academic Exchange Service (DAAD) for the financial support. References [1] F. Mücklich, N. Ilić, K. Woll, Intermetallics 16 (2008) 593–608. [2] F. Soldera, N. Ilic, S. Brännström, I. Barrientos, H.A. Gobran, F. Mücklich, Oxid. Met. 59 (2003) 529–542. [3] M.A. Guitar, F. Mücklich, Oxid. Met. 80 (2013) 423–436.. [4] B. Tryon, T.M. Pollock, M.F.X. Gigliotti, K. Hemker, Scr. Mater. 50 (2003) 9–14. [5] D. Zhong, E. Mateeva, I. Dahan, J.J. Moore, G.G.W. Mustoe, T. Ohno, Surf. Coat. Technol. 133–134 (2000) 8–14. [6] F. Klocke, T. Bergs, K. Georgiadis, H. Sarikaya, F. Wang, Coating Systems for Precision Glass Molding Tools, in: Proc. 7th Int. Conf. Coatings Manuf. Eng., 2008, pp. 209–218. [7] K.J. Ma, H.H. Chien, W.H. Chuan, C.L. Chao, K.C. Hwang, Key Eng. Mater. 364– 366 (2008) 655–661. [8] I. Baker, Mater. Sci. Eng. A 192–193 (1995) 1–13. [9] C. Fu, Y. Ye, M. Yoo, K. Ho, Phys. Rev. B. Condens. Matter 48 (1993) 6712–6715. [10] H.A. Gobran, K.W. Liu, D. Heger, F. Mücklich, Scr. Mater. 49 (2003) 1097–1102. [11] A. Katsman, H.J. Grabke, L. Levin, Oxid. Met. 46 (1996) 313–331. [12] N. Ilić, F. Soldera, F. Mücklich, Intermetallics 13 (2005) 444–453. [13] F. Soldera, N. Ilić, N. Manent Conesa, I. Barrientos, F. Mücklich, Intermetallics 13 (2005) 101–107. [14] T. Chou, J. Mater. Res. 5 (1990) 378–384. [15] H.A. Gobran, D. Heger, F. Mücklich, Zeitschrift Für Met. 96 (2005) 792–798. [16] I.M. Wolff, J. Miner. Met. Mater. Soc. 49 (1997) 34–39. [17] P. Hou, J. Am. Ceram. Soc. 86 (2003) 660–668. [18] R. Klumpes, C. Maree, E. Schramm, J. Wit, Mater. Corros. 47 (1996) 619–624. [19] V. Tolpygo, H. Grabke, Scr. Mater. 38 (1997) 123–129. [20] N. Takahashi, T. Mizoguchi, T. Nakagawa, T. Tohei, I. Sakaguchi, A. Kuwabara, et al., Phys. Rev. B 82 (2010) 1–5. [21] G. Bye, G. Simpkin, J. Am. Ceram. Soc. 57 (1974) 367–371. [22] D. Zhong, G.G.W. Mustoe, J.J. Moore, J. Disam, Surf. Coat. Technol. 146–147 (2001) 312–317. [23] A.Y. Yi, A. Jain, J. Am. Ceram. Soc. 88 (2005) 579–586. [24] M.A. Guitar, K. Woll, E. Ramos-Moore, F. Mücklich, Thin Solid Films 527 (2013) 1–8. [25] M. Birkholz, Thin Film Analysis by X-Ray Scattering, Wiley-VCH Verlag, GmbH, 2006. [26] F. Mücklich, N. Ilić, Intermetallics 13 (2005) 5–21. [27] P.J. Bellina, A. Catanoiu, F.M. Morales, M. Rühle, J. Mater. Res. 21 (2006) 276– 286. [28] G. Meier, F. Pettit, Mater. Sci. Eng. A 153 (1992) 548–560. [29] D.R. Clarke, Acta Mater. 51 (2003) 1393–1407. [30] B.A. Pint, J. Am. Ceram. Soc. 86 (2003) 686–695. [31] F. Pettit, J. Wagner, Acta Metall. 12 (1964) 35–40. [32] F. Pettit, R. Yinger, J. Wagner, Acta Metall. 8 (1960) 617–623. [33] N. Birks, G.H. Meier, F.S. Pettit, 2nd ed., Cambridge University Press, 2006. [34] K.R. Lawless, Rep. Prog. Phys. 37 (1974) 231–316. [35] C. Xu, W. Gao, H. Gong, Intermetallics 8 (2000) 769–779. [36] J. Smialek, J. Doychak, D. Gaydosh, Oxid. Met. 34 (1990) 259–275. [37] G.H. Meier, Mater. Corros. 47 (1996) 595–618. [38] G.C. Rybicki, J.L. Smialek, Oxid. Met. 31 (1989) 275–304. [39] G. Liu, Y. Zhu, K. Mimura, M. Isshiki, Open Corros. J. 3 (2010) 92–98. [40] H. Grabke, M. Brumm, B. Wagemann, Mater. Corros. 47 (1996) 675–677. [41] D. Zhong, J.J. Moore, E. Sutter, B. Mishra, Surf. Coat. Technol. 200 (2005) 1236– 1241. [42] A. Huntz, J. Mater. Sci. Lett. 8 (1999) 1981–1984. [43] J. Stringer, Corros. Sci. 1 (1970) 513–543. [44] U. Welzel, J. Ligot, P. Lamparter, A.C. Vermeulen, E.J. Mittemeijer, J. Appl. Crystallogr. 38 (2005) 1–29. [45] S.-L. Shang, H. Zhang, Y. Wang, Z.-K. Liu, J. Phys.: Condens. Matter 22 (2010) 1– 8. [46] ASM Handbook, Alloy Phase Diagrams (ASM Handbook), vol.3, ASM International, 1992. [47] Y. Waseda, E. Matsubara, K. Shinoda, X-Ray Diffraction Crystallography, Springer-Verlag, 2011..
(7)
Figure
![Fig. 6. Sin 2 w plots measured in the a -Al 2 O 3 scale. The curves show the strain average weighted by the multiplicity factor of each reflection [47].](https://thumb-us.123doks.com/thumbv2/123dok_es/7294768.446018/5.892.98.398.841.1080/measured-curves-strain-average-weighted-multiplicity-factor-reflection.webp)
Documento similar