Contact Angle Hysteresis: Study by Dynamic Cycling Contact Angle

Jul 8, 2004 - King's College Road, Toronto, Ontario, Canada M5S 3G8 ... these dynamic cycling contact angle measurements are attributed mainly to ...
0 downloads 0 Views 153KB Size
Langmuir 2004, 20, 6685-6691

6685

Contact Angle Hysteresis: Study by Dynamic Cycling Contact Angle Measurements and Variable Angle Spectroscopic Ellipsometry on Polyimide A. Hennig,†,‡ K.-J. Eichhorn,† U. Staudinger,† K. Sahre,† M. Rogalli,‡ M. Stamm,† A. W. Neumann,§ and K. Grundke*,† Leibniz Institute of Polymer Research, Hohe Strasse 6, 01069 Dresden, Germany, Infineon Technologies AG (Germany), Koenigsbrueckerstrasse 180, D-01099 Dresden, Germany, and Department of Mechanical and Industrial Engineering, University of Toronto, King’s College Road, Toronto, Ontario, Canada M5S 3G8 Received January 5, 2004. In Final Form: May 7, 2004 The phenomenon of contact angle hysteresis was studied on smooth films of polyimide, a polymer type used in the microelectronic industry, by dynamic cycling contact angle measurements based on axisymmetric drop shape analysis-profile in combination with variable angle spectroscopic ellipsometry (VASE). It was found that both advancing and receding contact angles became smaller with increasing the number of cycles and are, therefore, not a property of the dry solid alone. The changes of the wetting behavior during these dynamic cycling contact angle measurements are attributed mainly to swelling and/or liquid retention. To reveal the water-induced changes of the polymer film, the polyimide surface was studied before and after the contact with a water droplet by VASE. Both the experimental ellipsometric spectrum for ∆ and that for Ψ as well as the corresponding simulations show characteristic shifts due to the contact with water. The so-called effective medium approximation was applied to recover information about the thickness and effective optical constants of the polymer layer from the ellipsometrically measured values of ∆ and Ψ. On the basis of these results, the swelling and retention behavior of the polyimide films in contact with water droplets were discussed.

Introduction Polyimides are thermally stable polymers that exhibit excellent chemical resistance, good mechanical properties, and good dielectric properties. Therefore, they are used in microelectronic devices, in aircraft, and in space applications. Polyimides are also of interest as dielectric materials for microelectronic and electronic packaging applications. In microelectronic applications, they are typically applied as protective coatings, interlayer dielectrics, or passivation layers. For these applications, also a higher adhesiveness for reliability under humid and temperature cycling conditions needs to be developed. In this context, it is well-known that moisture uptake and swelling of polymers can lead to significant reliability problems. In previous studies, the swelling behavior of polyimides was studied in atmospheres of defined humidity.1-3 Here, we are interested in the effects of swelling and/or water retention when a polyimide surface is in contact with a water droplet. The purpose of this work was to study the effect of liquid sorption and swelling on contact angle hysteresis at smooth polyimide surfaces. It is of fundamental but also practical interest to know why contact * Author to whom correspondence should be addressed: Dr. Karina Grundke, Department of Polymer Interfaces, Leibniz Institute of Polymer Research, Hohe Strasse 6, 01069 Dresden, Germany. Phone: ++49 351 4658-475. Fax: ++49 351 4658-284. E-mail: [email protected]. † Leibniz Institute of Polymer Research. ‡ Infineon Technologies AG. § University of Toronto. (1) Buchhold, R.; Nakladal, A.; Gerlach, G.; Sahre, K.; Mu¨ller, M.; Eichhorn, K.-J. J. Electrochem. Soc. 1998, 145, 4012-4018. (2) Buchhold, R.; Nakladal, A.; Gerlach, G.; Sahre, K.; Eichhorn, K.-J. Thin Solid Films 1998, 312, 238-245. (3) Buchhold, R.; Nakladal, A.; Gerlach, G.; Herold, M.; Gauglitz, G. Sahre, K.; Eichhorn, K.-J. Thin Solid Films 1999, 350, 178-185.

angle hysteresis even on well-prepared solid surfaces is observed. Young’s equation implies a single, unique contact angle:

γlv cos θY ) γsv - γsl

(1)

where γlv, γsv, and γsl are the interfacial tensions at the liquid-vapor, solid-vapor, and solid-liquid interfaces and θY is the Young contact angle. However, in real systems, instead of one unique contact angle, a range of contact angles exists and each angle in that range can give rise to a mechanically stable liquid meniscus. This discrepancy between theory and practice is believed to be due to deviations of the solid surfaces from ideal behavior. While the derivation of Young’s equation assumes that the solid surface in contact with the liquid is smooth, homogeneous, isotropic, insoluble, and generally nonreactive and nondeformable, real surfaces are often rough or chemically heterogeneous to some extent. In addition, swelling and reorientation of functional surface groups may occur in contact with the liquid. As a result, contact angles on real surfaces exhibit hysteresis. Simple models4,5 were developed showing that multiple metastable equilibrium states exist on rough and/or heterogeneous solid surfaces, thus, providing a possible explanation for contact angle hysteresis. In this context, line tension effects on rough6,7 and heterogeneous7-9 (4) Li, D.; Neumann, A. W. In Applied Surface Thermodynamics; Neumann, A. W., Spelt, J. K., Eds.; Surfactant Science Series 63; Marcel Dekker: New York, 1996; pp 109-168. (5) Grundke, K. In Handbook of Applied Surface and Colloid Chemistry; Holmberg, K., Ed.; John Wiley & Sons: Chichester, 2001; Chapter 7. (6) Amirfazli, A.; Kwok, D. Y.; Gaydos, J.; Neumann, A. W. J. Colloid Interface Sci. 1998, 205, 1-11. (7) Marmur, A. Colloids Surf., A 1998, 136, 81-88. (8) Schwartz, L. W. Langmuir 1998, 14, 3440-3453.

10.1021/la036411l CCC: $27.50 © 2004 American Chemical Society Published on Web 07/08/2004

6686

Langmuir, Vol. 20, No. 16, 2004

surfaces were also considered. However, it is unlikely that these models include all possible causes of contact angle hysteresis, because the latter also occurs on solid surfaces that are extremely smooth and homogeneous, such as polymer films produced by film casting or spin coating. For instance, Garbassi et al.10 reported a kinetic hysteresis caused by time-dependent solid-liquid interactions. While Fadeev and McCarthy11 attributed contact angle hysteresis in part to roughness of the surface in molecular dimensions, they also found evidence for molecular mobility and packing as contributing factors. They found that the contact angle hysteresis of chemically grafted monolayers of trialkylsilanes is a function of alkyl group structure and bonding density. Yasuda et al.12 and Wang et al.13 attributed contact angle hysteresis to an overall change of the surface state. According to their concept, the surface state of a material adjusts to the conditions of the surrounding medium. If the surrounding medium is changed, the surface state of the solid will change to attain an equilibrium with the new surrounding medium to minimize the interfacial tension under new conditions. Their studies indicated that polymeric surfaces change the surface configuration by short-range motion such as rotational motion of chain segments at the surface. Sedev et al.14,15 attributed contact angle hysteresis to liquid penetration and surface swelling. A dependence of the chain length of three alkanes was found. Hysteresis increased with decreasing chain length of the liquid in contact with a hydrophobic polymer surface. Recent work by Lam et al.16,17 also suggests that contact angle hysteresis phenomena on polymeric surfaces are mainly due to liquid penetration and surface swelling, or at least liquid retention, even on very hydrophobic polymer surfaces. In agreement with the findings of Sedev et al.,14,15 they concluded from their study that contact angle hysteresis was found to decrease with increasing chain length of the liquid molecules or by using bulky and quasispherical molecules and, hence, depends on the size of liquid molecules. Cycling contact angle measurements of two homologous series of n-alkanes and 1-alcohols on fluorocarbon-coated surfaces provided more information on the mechanism of contact angle hysteresis.17 The measured advancing and receding contact angles at different contact radii and, hence, also different liquid/ solid contact time became smaller with an increasing number of cycles and are, therefore, not a property of the dry solid alone. These results were interpreted by an increasing solid surface modification with longer solidliquid contact time. However, no direct evidence of this surface modification could be provided. Therefore, we used dynamic cycling contact angle measurements based on axisymmetric drop shape analysis-profile (ADSA-P) in combination with variable angle spectroscopic ellipsometry (VASE). The latter technique allows to quantify the (9) Gaydos, J.; Neumann, A. W. In Applied Surface Thermodynamics; Neumann, A. W., Spelt, J. K., Eds.; Dekker: New York, 1996; pp 169238. (10) Garbassi, F., Morra, M., Occhiello, E., Eds. Polymer Surfaces From Physics to Technology; Chichester, 1996; Chapter 4. (11) Fadeev, A. Y.; McCarthy, T. J. Langmuir 1999, 15, 3759-3766. (12) Yasuda, T.; Miyama, M.; Yasuda, H. Langmuir 1992, 8, 14251430. (13) Wang, J.-H.; Claesson, P. M.; Parker, J. L.; Yasuda, H. Langmuir 1994, 10, 3887-3897. (14) Sedev, R. V.; Petrov, J. G.; Neumann, A. W. J. Colloid Interface Sci. 1996, 180, 36-42. (15) Sedev, R. V.; Budziak, C. J.; Petrov, J. G.; Neumann, A. W. J. Colloid Interface Sci. 1993, 159, 392-399. (16) Lam, C. N. C.; Kim, N.; Hui, D.; Kwok, D. Y.; Hair, M. L.; Neumann, A. W. Colloids Surf., A 2001, 189, 265-278. (17) Lam, C. N. C.; Ko, R. H. Y.; Yu, L. M. Y.; Ng, A.; Li, D.; Hair, M. L.; Neumann, A. W. J. Colloid Interface Sci. 2001, 243, 208-218.

Hennig et al.

effects caused by liquid penetration and retention at the solid surface. Recent spectroscopic ellipsometry studies of polymer/ water interfaces were described by Tang et al.18,19 and Gilchrist et al.20 Tang et al. determined the swelling of thin phosphorylcholine polymer films measured in a liquid cell. Gilchrist investigated the behavior of poly(methyl methacrylate)/water interfaces and the adsorption of nonionic surfactants. Information about the nature of the changes of the solid-liquid interface were extracted from the ellipsometric spectra using difference plots of the quantities ∆ and Ψ, as we do also in this work. An appropriate optical model is necessary to determine parameters such as layer thickness and optical constants from the ellipsometric data. The effective medium approximation (EMA) is a well-known tool for the calculation of effective optical constants of a mixture of two or three materials.21 According to Keddie et al.22 it can be also applied to describe a swollen polymer layer. In that case, the dry polymer is the host material and the solvent is the second material. Mathe et al.23,24 used the MaxwellGarnett EMA to calculate the effective refractive index and thickness of poly(ethylene glycol) films during swelling in humid air. To model the putative polyimide/water composite layer, the Bruggeman EMA was used in the present study. Both EMAs use only sligthly different approximations.21,25 Thus, for most of the applications the differences are very small.26 In summary, it is the aim of this work to quantify the effect of swelling and liquid retention on contact angle hysteresis using dynamic cycling contact angle measurements with water on polyimide surfaces in combination with VASE. Experimental Section Material. In this study, polyimide films applied in the microelectronic industry as negative photosensitive polyimide precursors purchased from the Asahi Kasei Corp. were used. The layer was prepared by spin coating onto 200-mm silicon wafers with a spinning speed of 3500 rpm and soft baking. Then, it was exposed to coherent light using a step-and-repeat projection printing system (stepper). The light source has a wavelength of 365 nm (i line of a Hg arc lamp). The films were developed with cyclopentanone. The samples were hard-baked to dry and prepolymerized at 200 °C for 30 s on a hotplate. The layers were finally cured at 340 and 380 °C for 1 h under a nitrogen atmosphere (O2 content