pubs.acs.org/Langmuir © 2010 American Chemical Society
Comparing Contact Angle Measurements and Surface Tension Assessments of Solid Surfaces Dory Cwikel,† Qi Zhao,‡ Chen Liu,‡ Xueju Su,‡ and Abraham Marmur*,† †
Department of Chemical Engineering, Technion - Israel Institute of Technology, 32000 Haifa, Israel, and ‡ Department of Mechanical Engineering, University of Dundee, Dundee DD1 4HN, United Kingdom Received May 19, 2010. Revised Manuscript Received July 25, 2010
Four types of contact angles (receding, most stable, advancing, and “static”) were measured by two independent laboratories for a large number of solid surfaces, spanning a large range of surface tensions. It is shown that the most stable contact angle, which is theoretically required for calculating the Young contact angle, is a practical, useful tool for wettability characterization of solid surfaces. In addition, it is shown that the experimentally measured most stable contact angle may not always be approximated by an average angle calculated from the advancing and receding contact angles. The “static” CA is shown in many cases to be very different from the most stable one. The measured contact angles were used for calculating the surface tensions of the solid samples by five methods. Meaningful differences exist among the surface tensions calculated using four previously known methods (Owens-Wendt, Wu, acid-base, and equation of state). A recently developed, Gibbsian-based correlation between interfacial tensions and individual surface tensions was used to calculate the surface tensions of the solid surfaces from the most stable contact angle of water. This calculation yielded in most cases higher values than calculated with the other four methods. On the basis of some low surface energy samples, the higher values appear to be justified.
Introduction The assessment of surface tension of solid surfaces has been of much fundamental and practical interest for many decades.1-6 Since existing techniques for surface tension measurement rely on surface deformations, the surface tension of a solid cannot yet be directly measured. Its assessment must be done indirectly, mostly relying on contact angle (CA) measurements and their interpretation. Nonetheless, meaningful measurements of CAs on real, nonideal surfaces are not as simple as they seemingly appear to be.4-6 The CA of a liquid on a solid surface in air that is required for assessing the surface tension of the solid is given by the Young equation σ s - σ sl cos θY ¼ ð1Þ σl Here, θY is the Young CA, σs and σl are the surface tensions of the solid and liquid, respectively, and σsl is the solid-liquid interfacial tension. For macroscopic drops, for which the effect of line tension is negligible, the Young CA is an excellent approximation of the ideal CA that prevails on ideal solid surfaces and characterizes their intrinsic wettability. If the values of σl and the Young CA are known, eq 1 remains with two unknowns: σs and σsl. In order to solve this equation for σs, one has to add a second equation that correlates σsl with σs and σl. However, in most practical cases the Young CA cannot be directly measured. As is well-known, on real surfaces that may be rough or chemically heterogeneous, there exists a measurable *Fax: 972-4-829-3088. E-mail:
[email protected]. (1) Johnson, R. E. Jr.; Dettre, R. H. In Surface and Colloid Science, Matijevic, E., Ed.; Wiley-Interscience; Weinheim, 1969; Vol 2, pp 85-153. (2) Neumann, A. W. In Wetting, Spreading and Adhesion, Padday, J. F., Ed.; Academic Press: London, 1978; pp 3-35. (3) Neumann, A. W.; Good, R. J. In Surface and Colloid Science, Good, R. J., Stromberg, R. R., Eds.; Plenum Press: New York, 1979; Vol 11, pp 31-91. (4) Marmur, A. Soft Matter 2006, 2, 12–17. (5) Marmur, A. Annu. Rev. Mater. Res. 2009, 39, 473–489. (6) Marmur, A. In Contact Angle, Wettability, and Adhesion, Mittal, K., Ed.; Koninklijke Brill NV: Leiden, 2009; Vol 6, pp 3-18.
Langmuir 2010, 26(19), 15289–15294
range of metastable CAs, termed the hysteresis range.1-3,7-9 The highest and lowest CAs in this range are the advancing and receding CAs, respectively, and the difference between them may be quite substantial. Measuring the hysteresis range is, therefore, essential for identifying chemical and physical heterogeneities on solid surfaces. In order to assess the ideal (Young) CA, it is necessary first to identify the most stable CA that is associated with the absolute minimum in the Gibbs energy of the system.4-6 This is so, because the most stable CA is the only measurable CA that can be theoretically linked with the Young CA under proper conditions. The most stable CA can be assessed by direct measurement using a vibrating system10-12 or approximately estimated by proper averaging of the advancing and receding CAs.11,12 Thus, a potential additional advantage of measuring these CAs is the ability to estimate the most stable CA. For smooth, chemically heterogeneous surfaces, the most stable CA is, by the Cassie equation,13,14 the weighted average of the local Young CAs on the surface. In this case, the most stable CA should be used in eq 1 to assess the average surface tension of the solid. For rough, chemically uniform surfaces, the Wenzel equation should be used to calculate the Young CA.15,16 In both of these cases, the assessment of the Young CA from the most stable one is meaningful only if the drop is sufficiently large compared with the scale of roughness or chemical heterogeneity.14,16,17 In many cases in practice, the “static” CA has been measured; in (7) Shanahan, M. E. R. Surf. Interface Anal. 1991, 17, 489–495. (8) di Meglio, J. M. Europhys. Lett. 1992, 17, 607–612. (9) Marmur, A. Adv. Colloid Interface Sci. 1994, 50, 121–141. (10) Meiron, T. S.; Marmur, A.; Saguy, I. S. J. Colloid Interface Sci. 2004, 274, 637–644. (11) Andrieu, C.; Sykes, C.; Brochard, F. Langmuir 1994, 10, 2077–2080. (12) Decker, E. L.; Garoff, S. Langmuir 1996, 12, 2100–2110. (13) Cassie, A. B. D. Disc. Faraday Soc. 1948, 3, 11–16. (14) Brandon, S.; Haimovich, N.; Yeger, E.; Marmur, A. J. Colloid Interface Sci. 2003, 263, 237–243. (15) Wenzel, R. N. J. Ind. Eng. Chem. 1948, 28, 988–994. (16) Wolansky, G.; Marmur, A. Colloids Surf., A 1999, 156, 381–388. (17) Marmur, A.; Bittoun, E. Langmuir 2009, 25, 1277–1281.
Published on Web 09/03/2010
DOI: 10.1021/la1020252
15289
Article
Cwikel et al.
this case, the drop is placed on the solid surface and measured as is. This CA is the least defined of the measurable CAs, since the drop may end up at any metastable position between the advancing and receding CAs. The “static” CA depends on the initial kinetic energy of the drop and the dynamics of the vibrations it experiences after landing on the surface.5,6 Thus, when measurements are done in a given lab, under exactly the same conditions, the results may be reproducible. However, under different conditions (different drop volume, height of needle, measuring instrument, lab procedure, etc.) the results may be quite different. The adjective “static” has been used in the literature although a better choice would have been “as is,” since actually, all measured CAs are static (even if they are measured by a slow flow, quasi-static process). As mentioned above, to calculate the surface tension of the solid from eq 1 σsl needs to be correlated with σs and σl. Until recently, two main types of correlations have been employed. The first type considers surface tension to consist of nonpolar (dispersion) and various polar contributions.18-22 Thus, Owens and Wendt20 suggested the following form 1=2 1=2 - 2 σps σ pl σsl ¼ σs þ σl - 2 σ ds σ dl
ð2Þ
where the superscript d indicates the nonpolar contribution to the surface tension and the polar contribution is σp σ - σd. Wu21 Suggested a different equation that reads 4σ p σ p 4σd σ d σ sl ¼ σ s þ σ l - d s l d - p s l p σs þ σl σs þ σl
ð3Þ
The latest version of this line of thought is the Lifshitz-van der Waals/acid-base approach, initiated by van Oss, Chaudhuri, and Good.22,23 In this approach, the polar component, σAB, is split into an acceptor (Lewis acid) surface parameter, σþ, and a donor (Lewis base) surface parameter, σ-, in such a way that σAB = 2(σþσ-)1/2. On the basis of this assumption, the following correlation was proposed
ð4Þ
Since the substitution of eqs 2, 3, or 4 into eq 1 still leaves more than one unknown, a few liquids have to be used in order to calculate σsl. The choice of the right liquids for this purpose may be critical.24,25 In addition, it is important to realize that, while the dispersion part of this equation is derived from an approximate theory for dispersion interactions, the functional form of the terms related to the polar components has not yet been substantiated by theory.26 (18) Girifalco, L. A.; Good, R. J. J. Phys. Chem. 1957, 61, 904–909. (19) Fowkes, F. M. J. Phys. Chem. 1963, 67, 2538–2541. (20) Owens, D. K.; Wendt, R. C. J. Appl. Polym. Sci. 1969, 13, 1741–1747. (21) Wu, S. J. Polym. Sci., Part C 1971, 34, 19–30. (22) van Oss, C. J.; Good, R. J.; Chaundhury, M. K. Langmuir 1988, 4, 884–891. (23) van Oss, C. J.; Good, R. J.; Chaundhury, M. K. J. Colloid Interface Sci. 1986, 111, 378–390. (24) Della Volpe, C.; Siboni, S. J. Colloid Interface Sci. 1997, 195, 121–136. (25) Shalel-Levanon, S.; Marmur, A. J. Colloid Interface Sci. 2003, 262, 489-499; 268, 272. (26) Marmur, A.; Valal, D. Langmuir 2010, 26, 5568–5575. (27) Neumann, A. W.; Good, R. J.; Hoppe, C. J.; Sejpal, M. J. Colloid Interface Sci. 1974, 49, 291–304.
15290 DOI: 10.1021/la1020252
water (W), H2O diiodomethane (DI), CH2I2 ethylene glycol (EG), C2H6O2
σl
σlLW
σlAB
σlþ
σl-
72.8 50.8 48.0
21.8 50.8 29.0
51.0 0 19.0
25.5 0 1.92
25.5 0 47.0
Another type of a correlation was developed by Neumann and co-workers and named by them “the equation of state (EoS)” approach.27,28 Their fundamental assumption was that σsl is a function of only σs and σl. The correlation based on this approach reads 2 pffiffiffiffiffiffiffiffiffiffiffi ð5Þ σ sl ¼ σ s þ σ l - 2 σs 3 σ l 3 e - βðσl - σs Þ where β is an empirical constant that was found to be 0.000 124 7 (mN/m)-2. Again, while the square-root dependence is derived from an approximate theory for dispersion interactions,26 the exponential correction term has not been yet substantiated by theory. Recently, an entirely different approach has been suggested that is directly derived from the Gibbs theory for surface or interfacial tension.26 This approach leads to a general form for the correlation between interfacial tensions and the corresponding surface tensions Ψðσ sl Þσ sl ¼ ψðσ l Þσ l - ψðσs Þσ s
ð6Þ
where Ψ and ψ are functions to be determined, and it is assumed that σs < σl (so that the CA > 0). Specific functions that were empirically found to best fit liquid-liquid data (for which also the interfacial tension can be directly measured) are adopted here for liquid-solid systems coshðσ ef =σ o Þðσef Þ - m σsl ¼ coshðσ l =σo Þðσ l Þ1 - m - coshðσs =σ o Þðσ s Þ1 - m
ð7Þ where the “effective” interfacial tension, σef, is given by σ ef σ s þ σ l - ησnl σ 1s - n
qffiffiffiffiffi qffiffiffiffiffi2 σds - σdl σ sl ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffi þ þ þ þ þ2 σs σs þ σl σl - σs σl - σs σl
Table 1. Surface Tensions (mN/m), and Surface Tension Components of Test Liquids [7, 8]22,23
ð8Þ
and m, σo, η, and n are empirical constants. For a large variety of liquid-liquid systems, the optimal values of the constants turned out to be m = 0.938 84, σo = 42.121, η = 0.837 55, and n = 0.949 65. The main difficulty in CA measurement and interpretation is the lack of an independent standard. The objective of this manuscript is to somewhat circumvent this difficulty by comparing and discussing, for a large number of samples, measurements of “static,” most stable, receding, and advancing CAs, as well as the above-mentioned approaches to calculating the surface tension of the solid. The solid surfaces that were characterized were produced within the AMBIO project (European Commission’s 6th Framework Program), the goal of which was the development of anti marine-biofouling coatings. Many of the developed coatings underwent CA measurements and surface tension assessment by two independent laboratories. One lab (University of Dundee) performed the measurements and calculations using the “static” CA of three liquids and eqs 2-5 for calculating the surface tension of the sample; the other lab (Technion - Israel Institute of Technology) measured the most stable, receding, and advancing CAs, and used eq 7 to calculate σs. (28) Kwok, D. Y.; Neumann, A. W. Colloid Surf., A 2000, 161, 31–48.
Langmuir 2010, 26(19), 15289–15294
Cwikel et al.
Article Table 2. Measured Contact Angles
sample
”static” CA° water
”static” CA° diiodomethane
”static” CA° ethylene glycol
most-stable CA° water
advancing CA° water
receding CA° water
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
77.4 ( 1.2 82.5 ( 0.8 85.1 ( 1.0 89.7 ( 0.6 77.4 ( 0.7 89.1 ( 0.3 81.4 ( 1.1 92.6 ( 0.3 77.9 ( 3.1 76.3 ( 0.5 77.5 ( 0.1 85.7 ( 1.1 87.3 ( 0.7 92.9 ( 0.5 99.9 ( 0.4 91.2 ( 0.3 91.7 ( 0.3 95.2 ( 0.2 78.4 ( 0.2 66.2 ( 0.2 80.5 ( 0.5 85.0 ( 0.5 88.8 ( 0.3 90.1 ( 0.4 92.3 ( 0.9 94.3 ( 0.7 90.2 ( 0.6 91.1 ( 0.7 88.4 ( 0.1 87.4 ( 0.0 86.3 ( 0.3 84.5 ( 0.3 83.3 ( 0.3 130.8 ( 1.0 109.2 ( 1.3 101.2 ( 0.4 96.8 ( 1.5 95.7 ( 0.9 92.3 ( 0.7 79.8 ( 0.4 64.2 ( 0.8 61.0 ( 0.7
32.4 ( 0.3 74.7 ( 0.9 71.2 ( 0.4 75.5 ( 0.4 69.1 ( 0.6 85.3 ( 1.9 44.3 ( 0.9 83.5 ( 0.3 39.0 ( 0.2 40.8 ( 0.2 41.7 ( 0.3 45.7 ( 0.6 37.7 ( 1.3 38.6 ( 0.5 68.5 ( 0.8 36.1 ( 0.8 46.7 ( 0.7 83.0 ( 0.9 40.1 ( 0.7 37.0 ( 0.2 66.4 ( 0.4 67.5 ( 0.3 69.5 ( 0.5 69.2 ( 0.7 70.5 ( 0.2 70.4 ( 0.4 70.1 ( 0.8 75.6 ( 1.3 70.1 ( 0.8 68.2 ( 0.2 74.0 ( 0.6 59.7 ( 0.3 67.4 ( 0.3 107.5 ( 0.5 82.9 ( 1.3 68.6 ( 0.6 68.3 ( 0.3 69.8 ( 0.7 54.7 ( 1.1 59.4 ( 0.5 10.7 ( 1.1 12.3 ( 0.6
37.2 ( 0.3 51.5 ( 1.6 74.7 ( 0.6 54.7 ( 1.9 76.2 ( 0.5 81.9 ( 0.3 72.4 ( 1.1 81.5 ( 0.2 65.4 ( 0.2 65.4 ( 0.5 64.3 ( 0.1 69.9 ( 0.1 58.6 ( 0.6 75.6 ( 0.5 78.7 ( 0.4 79.3 ( 0.4 79.8 ( 0.9 81.2 ( 1.4 70.4 ( 0.1 68.4 ( 0.7 62.7 ( 0.4 67.2 ( 0.1 74.2 ( 0.4 73.3 ( 0.3 76.8 ( 0.8 77.4 ( 0.2 75.7 ( 0.3 73.6 ( 0.3 68.6 ( 0.4 69.7 ( 0.0 67.1 ( 1.2 65.5 ( 0.1 65.1 ( 0.1 104.4 ( 0.3 99.5 ( 1.2 93.7 ( 0.7 94.1 ( 1.5 90.9 ( 1.1 79.3 ( 0.7 62.5 ( 0.1 40.6 ( 0.8 43.2 ( 0.5
76.2 ( 0.9 83.3 ( 2.4 84.6 ( 1.5 83. Six (2. 83.7 ( 4.2 92.9 ( 1.3 77.2 ( 1.1 63 ( 1.6 65.6 ( 0.3 65.1 ( 0.6 63.5 ( 0.5 56.7 ( 0.8 76.6 ( 1.2 86.9 ( 1.3 93.3 ( 1.3 73.4 ( 1.2 87.3 ( 0.4 93.7 ( 1.2 68.9 ( 2 60.5 ( 1.7 86.4 ( 3.6 87.3 ( 1.3 91.9 ( 4 90.1 ( 1.1 89.3 ( 2.3 88.4 ( 4.4 90.8 ( 0.8 87.8 ( 0.8 76.6 ( 1.6 90.8 ( 1.5 77.5 ( 2 74.6 ( 2 80.1 ( 1 91.3 ( 0.8 101.8 ( 1.3 70.3 ( 0.9 72.5 ( 0.4 71.1 ( 0.6 98.5 ( 1.0 48.7 ( 1.5 69 ( 3 38.7 ( 1.6
93.8 ( 1.4 98.2 ( 0.3 102.3 ( 0.2 100.7 ( 1.8 101.3 ( 1.2 111.1 ( 0.3 95.8 ( 4.2 93.8 ( 1.3 89.7 ( 0.1 88.2 ( 1.8 88.1 ( 0.8 89.8 ( 0 99.4 ( 0.7 109.9 ( 1.1 117.6 ( 1.1 100.2 ( 1 107.8 ( 0.3 108.9 ( 1.2 85 ( 0.5 76.2 ( 0.3 101.1 ( 0.5 105.7 ( 2 102.1 ( 1.7 103.5 ( 0.9 104 ( 0.5 106.1 ( 0.4 104.4 ( 0.2 90.8 ( 0.1 81.8 ( 3.1 104.6 ( 1.1 89.7 ( 0.5 82.8 ( 0.4 89.3 ( 2.4 130.7 ( 2.4 131.9 ( 0.4 102.9 ( 0.8 117.5 ( 0.3 121.6 ( 1.3 109.9 ( 1.5 52.1 ( 0.4 69 ( 1.5 68.8 ( 2.7
53.0 ( 0.1 78.4 ( 0. 8 70.5 ( 0.3 72.5 ( 1.3 54.6 ( 0.8 78.7 ( 1.7 51.5 ( 1.2 47.9 ( 1.3 54 ( 2.9 60 ( 0.6 55.5 ( 0. 8 55.3 ( 0.8 69 ( 0.4 61.7 ( 2. 5 87.7 ( 1.1 60.1 ( 3.8 63.7 ( 0.5 83.1 ( 1.8 48.8 ( 1 29.1 ( 0.3 82.9 ( 0.3 84 ( 0.1 85.4 ( 0.1 88.3 ( 1.3 83.5 ( 0.5 86.5 ( 1.1 83.9 ( 0.2 65.2 ( 2.4 50.9 ( 0.1 80,5 ( 0.9 72.9 ( 0.4 66.6 ( 0.3 56.3 ( 0.7 68.5 ( 0.2 74.5 ( 0.8 65.9 ( 2.3 61.4 ( 1.1 62.8 ( 2.4 73.4 ( 1.8 33.2 ( 0.1 56.6 ( 1.3 47.1 ( 0.1
Experimental Section The samples were prepared by the following companies and laboratories: Netherlands Organisation for Applied Scientific Research, The Netherlands; University of Pisa, Italy; University of Dundee, UK; University of Mons-Hainaut, Belgium; Centro de Tecnologias Electroquimicas, Spain; Polymer Laboratories Ltd., UK; and TEER Coating Ltd., UK. “Static” CAs were obtained using the sessile drop method with a Dataphysics OCA-20 contact angle analyzer. This instrument consists of a CCD video camera with a resolution of 768 576 pixels and up to 50 images per second, multiple dosing/microsyringe units, and a temperature-controlled environmental chamber. The digital drop image was processed by an image analysis system, which calculated both the left and right contact angles from the shape of the drop with an accuracy of (0.1°. Three test liquids were used as a probe for surface free energy calculations: distilled water (W), diiodomethane (DI) (Sigma), and ethylene glycol (EG) (Sigma). The data for surface tension components of the test liquids at 20 °C are given in Table 1.22,23 The surface tension is calculated using commercial software provided by Dataphysics OCA-20, which uses eqs 2-5. A different number of liquids may be required for each correlation. For example, the Wu approach requires a minimum of two test liquids; the surface tension of a solid is then given as the average of three sets of surface tensions, using the liquid pairs W-DI, W-EG, and DI-EG. On the other hand, the equation of state Langmuir 2010, 26(19), 15289–15294
approach requires only one test liquid, so the surface tension is given as the average of three calculations using W, DI, and EG. The receding and advancing CAs of water in air were measured using a Kruss DSA 100 drop analysis system. High-purity water, obtained by ultrafiltration (Elgar UHQ system) of deionized water (reverse osmosis), was used. For the determination of most stable CA4-6 of water in air, the sample was vibrated, using a horizontal drop shaking apparatus. Several sessile water drops (50 to 100 μL) were dispensed on the same sample with the help of a precision microsyringe and then gently shaken for 15 s. The magnitudes of amplitude and frequency of the vibrating plate were usually around 0.2 mm and 13 Hz, respectively. Digital pictures of the drops were taken from above, using a A 034451 Tamron or a Computar MLH-X10 magnifying lens, attached to a high-resolution U-eye digital camera that was aligned perpendicularly above the surface. For calibration of the pixel size, a small cylinder of precisely known diameter and of the same thickness as that of the sample was used. The pictures were processed with the Image-Pro Plus program, version 4.0 (Media Cybernetics). All drops that were not axisymetric or almost axisymetric (ratio of longest to shortest radius of >1.2) were excluded. The contact angles were then calculated by fitting to solutions of the YoungLaplace equation, using the maximum diameter, volume of the drop, and the surface tension of water as input values. All contact angles were measured at 23 °C. Due to the limited number of samples allocated by the project for each type of measurement, questionable results could not be verified by repetition. Therefore, DOI: 10.1021/la1020252
15291
Article
Cwikel et al. Table 3. Calculated Surface Tensions (mN/m) Using Various Correlations
sample
Owens-Wendt from “static” CA
Wu from “static” CA
Van Oss et al. from “static” CA
EoS from “static” CA
Marmur-Valal from ”most stable” CA
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
46.4 29.5 24.7 26.7 27.4 19.7 30.9 19.0 35.0 35.0 35.1 31.7 39.3 30.5 23.4 27.0 26.4 18.7 32.8 38.3 29.8 27.5 24.5 24.7 23.2 23.0 23.8 22.9 25.9 26.3 25.6 30.1 28.4 7.2 13.3 17.2 17.5 18.5 25.8 31.7 49.3 48.9
50.2 34.2 30.5 31.6 32.7 25.3 39.8 24.8 43.3 43.0 42.7 39.2 43.7 40.2 28.0 40.6 36.6 24.4 41.9 45.4 34.9 32.8 30.2 30.2 28.8 28.4 29.5 28.2 31.1 31.6 30.8 35.5 33.6 11.0 19.5 25.7 26.2 26.3 33.6 37.1 57.0 56.8
46.1 28.0 22.2 24.7 22.1 16.7 37.2 17.6 40.1 39.2 38.7 34.4 40.6 40.1 24.0 41.3 35.9 18.0 39.4 40.9 28.4 26.6 25.0 24.3 22.6 23.1 22.8 22.3 25.4 25.5 24.7 29.7 27.3 6.2 16.0 23.6 23.7 22.9 31.5 30.4 49.0 47.8
40.3 31.1 27.5 28.8 30.0 23.7 33.6 22.9 35.9 36.0 35.6 32.6 35.0 32.5 24.0 33.1 30.5 22.3 35.2 39.0 30.8 28.9 26.8 26.6 25.5 24.9 26.2 25.3 27.5 27.9 27.7 30.5 29.6 8.9 17.3 22.6 23.6 23.7 28.5 32.1 45.0 45.3
48.9 44.7 43.9 44.5 44.5 38.4 48.3 55.8 54.6 54.8 55.6 58.8 48.7 42.4 38.2 50.4 42.2 37.9 52.9 57.1 42.9 42.2 39.1 40.3 40.5 41.5 39.8 41.8 48.7 39.8 48.1 49.8 46.6 39.5 31.8 52.1 51 51.7 34.4 62.2 52.8 65.5
samples were eliminated from the present study if the standard deviation was above 4.5° for any of the four measured contact angles or above 2.0 for two of the contact angles.
Results and Discussion Table 2 lists the measured “static”, most stable, advancing, and receding CAs of the samples. The “static” CA was measured for three different liquids to enable calculations based on eqs 2-5, as described above. The numbers following the plus/minus sign identify the standard deviations. Each result for the “static” CA is based on 9 measurements. Each result for the receding, most stable, or advancing CA is based on 10-12 measurements. Table 3 lists the surface tensions, calculated by five different methods: eqs 2-5 and eq 7. Obviously, since each method deals in a different way with the question of surface tension components, the only common denominator for comparison can be the total value of the surface tension. Table 4 lists the generic chemistries of the solid surfaces. First, the various CAs are compared and discussed. Figure 1 shows four different water CAs for all samples: receding, most stable, advancing, and “static”. Since this paper focuses on comparing the various CAs measured for each sample, the specific identity of the sample is of little interest; the important point is that the samples span over a sufficiently wide CA range. It is 15292 DOI: 10.1021/la1020252
clearly seen in Figure 1 that the most stable CA is within the hysteresis range (between the advancing and receding CAs), as expected4-6 (except for one case, for which the measurements could not be repeated for lack of additional samples). The same observation applies to the “static” CA; however, for many samples it is clearly different from the most stable CA (differences much higher than the standard deviations). In many of the cases studied here, the “static” contact angle is closer to the advancing CA than the most stable CA. This seems reasonable, since the placement of the drop on the surface is basically an advancing process. As may be expected, all CAs have similar values when the CA hysteresis range is small. The observation of an appreciable hysteresis range for some of the samples indicates that the surfaces are not ideal. The samples were neither roughened after the coating nor designed to be rough to the extent that the roughness ratio is meaningfully higher than 1. On the other hand, chemical heterogeneities, even in the nanoscale, may cause CA hysteresis.29 Therefore, the source of hysteresis is most probably chemical heterogeneity, and the most stable CA appears to represent the average Young CA for these samples. (29) Bittoun, E.; Marmur, A.; Ostblom, M.; Ederth, T.; Liedberg, B. Langmuir 2009, 25, 12374–12379.
Langmuir 2010, 26(19), 15289–15294
Cwikel et al.
Article Table 4. Generic Chemistries of Samples
sample
chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
Titanium Fluorosilane on titanium deposited on glass Fluorosilane on modifed titanium Fluorosilane on electropolished titanium Fluorosilane on “pickled” titanium Polytrifluoroethyoxyphosphazene/polytrifluoroactafluoropentoxyphosphazene blend on silicone elastomer Poly(bispropylcarboxylatophenoxy)phosphazene spin coated on glass Blend of poly(bispropylcarboxylatophenoxy) phosphazene and polybistrifluoroethoxyphosphazene, spin-coated on glass Poly(bispropylcarboxylatophenoxy)phosphazene spin coated on Ni/Cr coated glass Blend of poly(bispropylcarboxylatophenoxy)phosphazene and Pluronic F108 spin coated on Ni/Cr coated glass Blend of poly(bispropylcarboxylatophenoxy)phosphazene and Pluronic F108 spin coated on Ni/Cr coated glass Blend of poly(bispropylcarboxylatophenoxy)phosphazene and Pluronic F108 spin coated on Ni/Cr coated glass Blend of poly(bispropylcarboxylatophenoxy)phosphazene and poly(bistrifluoroethoxy)phosphazene, spin coated on to glass. Blend of poly(bispropylcarboxylatophenoxy)phosphazene and poly(bistrifluoroethoxy)phosphazene, spin-coated on glass Silicone elastomer: Silastic-T2 (Dow Corning) Blend of poly(bispropylcarboxylatophenoxy)phosphazene and poly(bistrifluoroethoxy)phosphazene, spin coated on Ni/Cr coated glass. Blend of poly(bispropylcarboxylatophenoxy)phosphazene and poly(bistrifluoroethoxy)phosphazene, spin coated on 80/20 Ni/Cr coated glass. Blend of poly(bispropylcarboxylatophenoxy)phosphazene and poly(bistrifluoroethoxy)phosphazene, spin coated on Ni/Cr coated glass. Poly(bispropylcarboxylatophenoxy)phosphazene, spin coated on glass. Blend of poly(bispropylcarboxylatophenoxy)phosphazene and poly(4-vinylpyridiniumhexyl bromide), spin coated on glass. SiOx-like coating deposited on glass from hexamethylsiloxane precursor by plasma-assisted CVD. SiOx-like coating deposited on glass from hexamethylsiloxane precursor by plasma-assisted CVD. SiOx-like coating deposited on glass from hexamethylsiloxane precursor by plasma-assisted CVD. SiOx-like coating deposited on glass from hexamethylsiloxane precursor by plasma-assisted CVD. SiOx-like coating deposited on glass from hexamethylsiloxane precursor by plasma-assisted CVD. SiOx-like coating deposited on glass from hexamethylsiloxane precursor by plasma-assisted CVD. SiOx-like coating deposited on glass from hexamethylsiloxane precursor by plasma-assisted CVD. Hybrid sol-gel coating Hybrid sol-gel coating Hybrid sol-gel coating with double-modified large platelets Hybrid sol-gel coating with organically modified fibers Hybrid sol-gel coating with large platelets with fluor edge groups Hybrid sol-gel coating with large platelets aliphatic modified with fluor edge groups Fluorinated siloxane copolymer blend Fluorinated siloxane copolymer blend Amphiphilic fluorinated block copolymer on SEBS base Amphiphilic fluorinated block copolymer on SEBS base Amphiphilic fluorinated block copolymer on SEBS base Poly(styrene-b-(ethylene-co-butylene)-b-styrene) (SEBS) triblock thermoplastic elastomer (Kraton G1652M) DLC coating modified with N and Si DLC coating modified with N DLC coating modified with Si
Figure 1. Comparison among the advancing (9), receding (b), “static” (Δ), and most stable (solid equality line) contact angles for all solid surface samples.
Figure 2 shows a test of the empirical suggestion that the most stable CA can be approximated by averaging the cosines of the advancing and receding CAs11 (another suggestion12 is that the averaging of the CAs themselves should be done; however, the differences between the two averages for most of the present data are negligible). It can be concluded from this figure that the average is in some cases close to the most stable CA but very different in others. Thus, this average may serve as a substitute for the most stable CA when the value of the latter is not available; however, a large measure of uncertainty may be associated with this assumption. Langmuir 2010, 26(19), 15289–15294
Figure 2. Comparison between the average contact angles (2) (calculated from the averages of the cosines of the advancing and receding contact angles) and the most stable contact angles (solid equality line) for all solid surface samples.
Next, the surface tensions of the solid sample surfaces, as calculated from the measured CAs using the various methods, are presented and discussed. Figure 3 compares the surface tension values calculated by the four methods described in eqs 2-5 (Ownes and Wendt, Wu, van Oss, and the EoS correlations) with those calculated from eqs 7 and 8 (the recently developed Marmur and Valal correlation). Two main conclusions may be drawn from this figure. First, it is clear that there exist very meaningful differences among the surface tensions calculated by the various methods. Second, while there is no clear regularity as to which of DOI: 10.1021/la1020252
15293
Article
Figure 3. Comparison of the surface tensions of the solid surface samples as calculated from the Ownes and Wendt ()), Wu (0), van Oss et al. (Δ), and equation of state () correlations, using the “static” contact angles, with the surface tension calculated from the Marmur and Valal correlation (solid equality line), using the most stable contact angle.
eqs 2-5 yields lower or higher surface tension values, it is clear that eq 7 predicts in most cases higher values than the others do. Since there is no absolute solid surface tension standard to compare with, definitive conclusions cannot be reached yet. However, some indication can be deduced, for example, from the surface with the highest most stable CA (sample no. 35), which is a fluorinated siloxane copolymer blend. The most stable CA is 101.8°, and eq 7 predicts a value of 31.8 mN/m. This seems to be more realistic than the low values (
