1 Laboratoire Analyse et Environnement, UMR 85 87, C.N.R.S., 2 Rue H.
Dunant, 94 320 THIAIS (France )
2 Laboratoire de Physique des Gaz et des Plasmas, UMR 85 78 Bat. 210, Universit� Paris Sud 91 405 ORSAY Cedex (France)
Keywords protective coating , HMDSO precursor , microwave plasma , electrochemical impedance spectroscopy.
Table 1 shows weight, thickness and density for plasma-polymerised coating from different HMDSO/O2 mixtures and at several time deposition. One may notice the short deposition time values due to the used of a microwave plasma reactor. The density values are obtained with precision near to 20%. At low HMDSO percentage, the highest density is obtained : it is close to the one of thermal silica (2,22 g.cm-3). For higher HMDSO content in the feed, the densities seem to be constant with a value around 1,5 g.cm-3.
Table 1 : Weight, thickness and density of several plasma-polymerised coating obtained from different HMDSO/O2 ratio mixtures, deposited on iron samples.
|Deposition time (min)||12||12||12||12||6||30||45|
|Coating weight (mg)||0,3||0.3||0.25||0,2||0,2||0,6||1,1|
The structure and bondings in the deposited films have been studied by means of IR spectroscopy (Figure 1). According to the litterature [11-13], the absortion bands can be assigned as follow : 790-800 cm-1 (Si-C stretching and CH3 rocking in Si-(CH3)2), 835 cm-1 (Si-C stretching and CH3 rocking in Si-(CH3)3), 880 cm-1 (Si-(CH3)2 rocking and bending), 1000-1100 cm-1 (Si-O asymmetric stretching), 1260 cm-1 (CH3 symmetric bending in Si-(CH3)X), a weak band at 1350 cm-1 (-CH2- scissoring and wagging vibrations in disilymethylene Si-CH2-Si), 1408 cm-1 ( CH3 symmetric bending in Si-(CH3)X), 1458 cm-1 ((CH)X asymmetric bending), 1710 cm-1 (C=O groups), 2900-2960 cm-1 (CHX symmetric and asymmetric stretching), 2100-2250 cm-1 (Si-H stretching vibration). With increasing O2 in the HMDSO/O2 mixture, the relative intensities of the absortion band at 1260 cm-1 (characteristic of Si-(CH3)X) is seen to decrease and the one of the band at 1000-1100 cm-1 (Si-O) increase. We consider that this is because of a decrease in the relative amount of organic content and to the high degree of crosslinking in the coating. This view is supported by measurement of the film density. The highest density is obtained at high O2 content in the feed. One may conclude that plasma-polymerised coatings obtained from HMDSO/O2 could be considered as polymer like or silica like as a function of O2 content in the feed.
Figure 1 : Infrared spectra of plasma polymerized coating deposited from various HMDSO/O2 ratios : (a) HMDSO/O2 100/0 ; (b) HMDSO/O2 80/20 ; (c) HMDSO/O2 50/50 ; (d) HMDSO/O2 20/80.
The RAIR spectra of coatings, obtained from several HMDSO/O2 mixture after exposure of the iron coated samples to NaCl for eight days, didn't show any change in the films. This fact demonstrates the chemical stability of these plasma-polymerised coatings. Corrosion protection properties by EIS The frequency dependence of the electrochemical impedance of a polymer coated iron sample usually can be modeled by an equivalent circuit shown in Figure 2 [14 - 17]. This model will be tested for the thin HMDSO plasma-polymerised coating. The non-conducting organic coating appears as a capacitor Cc called coating capacitance. Ionically conducting paths either due to the presence of defects or pores, or due to the slight degree of solvatation of water and ions by the coating, produce finite resistance Rpor (pore resistance), that short the coating. In series with this resistive element is a parallel resistance Rpol (polarisation resistance) and the double layer capacitance Cdl, representing the corrosion process at the electrolyte-saturated coating/metal interface. Finally, in series with the entire network representing the coated surface is Rsol : the electrolyte resistance due to ohmic drop within the electrolyte which is negligible in the case of coatings immersed in 0,1 mol.L-1 NaCl.
Figure 2 : The equivalent circuit used to model coated electrode
The initial coating capacitance Cc is the lowest for silica-like coating (D) which has the thickest coating layer (Figure 3(a)). Coatings A, B and C have the same thickness but sample A has a lower Cc value This fact could be correlated to the excited species present in the plasma during deposition . The low atomic oxygen concentration in the plasma provide to a coating less crosslinked. And hence this coating possesses the lower barrier properties. The increase of Cc in the first day of exposure is considered due to water uptake by the coating. In fact, the coating capacitance increases with increasing water uptake . The volume fraction V of electrolyte absorbed by the coating can be generally determined from the experimental values of Cc  as follow: v = log(Cc(t)/Cc(0))/log80 (1) where Cc(t) is the coating capacitance at time t and Cc(0) is the initial coating capacitance. The application of equation (1) allow to high volumique fraction of water penetrated close to 1. These values have not been confirmed by weight mesearements. In fact, the thickness of the plasma polymerised coating is very low comparatively to the classical organic protective coating and consequently, Rpor is so low and Cc cannot be determined with sufficient accuracy to determine water uptake. Figure 3(b) shows that the pore resistance Rpor decreased the most for coating A. At the end of exposure, Rpor was the lowest for sample A but the Rpor values of B, C and D coatings were similar. The increase of Cdl (Figure 3(c)) can be considered as evidence that the area at which delamination was increasing. This increase occured first and was the largest for sample A. The initial Cdl value of the silica-like coating (D) is the lowest but it doesn't vary very much with time exposure. The decrease of Rpol (Figure 3(d)), which suggests an increase of the delaminated area at the metal/coating interface is the largest for sample A. This analysis of EIS data for this four coatings systems shows qualitatively that the polymer-like coating A suffer the higher degradation and corrosion at the metal/coating interface during exposure to NaCl for eight days while B, C and D samples seem to have similar protective properties and show good performance relatively to the low thicknesses of all the coatings. Several parameters have been related to the delaminated area . The polarization resistance has been used to estimate the delaminated area because Rpol is obtained with most accuracy. Assuming that Ad is equal to the corroding area, one obtains the following relation ship : Ad = R0pol/Rpol (2) The specific polarization resistance, R0pol is associated with the charge transfer behavior of the metal substrate and can be estimated using the linear polarization of an uncoated sample. The value of R0pol is assumed to be constant. Ad has been calculated for the plasma-polymerised coating obtained from the HMDSO/O2 80/20 mixture. The results are reported in table 2 and is coherent with the corroding area observed. The delaminated area seems to stabilize around 0.015 cm2 after 4 days of immersion in NaCl.
Table 2 : Delaminated areas calculated from Equation (2) for the plasma-polymerised coating obtained from HMDSO/O2 80/20 mixture as a function of exposure time
|Exposure time (days)||0||0.5||1||1.5||2||3||4||7||8|
Figure 3 : Time dependence of Cc (a), Rpor (b), Cdl (c) and Rpol (d) for plasma-polymerised coating on iron substrates for various HMDSO/O2 ratios : (circle) HMDSO/O2 100/0 ; (square) HMDSO/O2 80/20 ; (cross) HMDSO/O2 50/50 ; (bold circle) HMDSO/O2 20/80
 K. D. Conners, W. J. van Ooij, S. J. Clarson and A. Sabata, J. Appl. Polym. Sci. : Appl. Polym. Symp., 54 (1994) 167.
 E. Sacher, J. E. Klemberg-Sapieha, H. P. Schreiber and M. R. Wertheimer, J. Appl. Polym. Sci. : Appl. Polym. Symp., 38 (1984) 163.
 W. J. van Ooij, A. Sabata and I. H. Loh, Pro. Eng. Symp., (1994) 253.
 W. J. van Ooij and N. Tang, Polym. Mater. Sci. Eng., (1996) 155.
 J. S. Tonge, T. H. Lane, P. A. Giwa Agbomeirele and H.M. Klimisch, Electrochem. Soc., INC, "Advances in Corrosion Protection by Organic Coatings", 13 (1995) 151
 W. J. van Ooij, A. Sabata, D. B. Zeik, C. E. Taylor, F. J. Boerio and S. J. Clarson, Journal of Testing and Evaluation, JTEVA, 23 (1) (1995) 33.
 T. F. Wang, T.J. Lin, D. J. Yang, J. A. Antonelli and H.K. Yasuda, Progress in Organic Coatings, 28 (1996) 291.
 M. Statmann, R. Feser and A. Leng, Electrochimica Acta, 39 (8/9) (1994) 1207.
 C. Vautrin-Ul, C. Boisse-Laporte, N. Benissad, A. Chausse, P. Leprince and R. Messina, to be published in Progress in Organic Coatings
 N. Bennissad, C. Boisse-Laporte, C. Vallee and A. Goullet, to be published in Surf. Coat. Tech.(1999).
 A.M. Wrobel, M.R. Wertheimer, J. Dib and H.P. Schreiber, J. Macromol. Sci. -Chem., A14 (3) (1980) 321.
 N. Inagaki and M. Taki, J. Appl. Polym. Sci., 27 (1982) 4337.
 F. Mansfeld, Journal of Applied Electrochemistry, 25 (1995) 187.
 J. M. McIntyre and H.Q. Pham, Progress in Organic Coatings, 27 (1996) 201.
 P.L. Bonora, F. Deflorian and L. Fedrizzi, Electrochimica Acta, 41 (7/8) (1996) 1073.
 M. Kendig, S. Jeanjaquet, R. Brown and F. Thomas, J. Coat. Technol., 68 (863) (1996) 39.
 D. M. Brasher and A.H. Kingsbury, J. Appl. Chem., 4 (1954) 62.