Langmuir 2007, 23, 1619-1622
1619
The Wetting of Gold and Silicon Nanoscale Arrays Maria-Victoria Meli and R. Bruce Lennox* Department of Chemistry, McGill UniVersity, 801 Sherbrooke Street West, Montreal, Quebec H3A 2K6, Canada ReceiVed August 7, 2006. In Final Form: December 10, 2006 The relative hydrophobicity of surfaces containing highly regular, nanoscale ( θ > 90° is often observed, even when complete wetting is energetically favorable. This is likely the reason why, in a recent paper,26 the application of eq 4 predicted Wenzel wetting, whereas Cassie wetting is observed. The issue of metastability has caused some confusion in understanding what conditions lead to composite wetting and, hence, ultrahydrophobic and superhydrophobic effects. Gentle placement of a drop of water on a surface (for which complete wetting is energetically favored) can result in a contact angle corresponding to the composite wetting situation.40 However, when the same drop is released from a height onto the surface, the energy barrier required to achieve Wenzel wetting is overcome, and complete wetting of the solid interface results. Furthermore, when Cassie wetting is favored, compressiondecompression cycles of a drop can reversibly switch the system from Wenzel to Cassie states.39 This paper describes the application of eqs 1-4 in characterizing the wetting of nanopatterned samples (with both feature height and diameter < 100 nm) obtained from a block copolymer patterning methodology. Wenzel-type wetting is indeed observed for these high areal density nanopatterned surfaces whose features have aspect ratios ranging from 0.2 to 0.7. Experimental Section Nanopatterned Au and Si/SiOx substrates were prepared as per the procedures described elsewhere.31,41 An argon ion-milling instrument with a 1 cm diameter spot size was used for pattern transfer to the substrates (6 keV, 3.5 µA/cm2, samples rocked at 40°/s from 0 to 50° from the surface normal, 60 rpm, MET-ETCH, Gatan, Inc.). Freshly patterned Au substrates were immersed in 1 mM ethanolic thiol (n-C5H11SH and n-C18H37SH) solution for 1 h, rinsed with approximately 250 mL of anhydrous ethanol, and dried under a stream of nitrogen gas. Nanopatterned substrates were compared to a flat, alkylthiolated gold substrate and an ion-milled42 Si/SiOx substrate. Si/SiOx substrates that were not exposed to the argon ion beam were also cleaned and assessed in order to compare them to the ion-milled Si/SiOx sample. Freshly formed templatestripped gold replicas43 were used as flat gold substrates and were immediately placed in 1 mM ethanolic thiol solution for 1 h followed by rinsing and drying as previously described. Clean Si/SiOx substrates were exposed to the Ar+ beam under the same conditions used in the pattern-transfer step in the nanopatterning methodology (6.5 keV, 12 µA, 180 s). Finally, patterned Au nanodot samples were prepared as described elsewhere41 and were cleaned by exposure to UV/ozonolysis conditions44,45 for 1 h, followed by rinsing with water and ethanol and drying under a stream of nitrogen gas. Fabrication of the nanodots by this methodology results in gold (41) Meli, M.-V.; Lennox, R. B. Langmuir 2003, 19, 9097-9100. (42) Substrates were cleaned using a degreasing treatment consisting of soaking them in recently boiled 50:50 ethanol/chloroform solution for 5 min followed by 2 min of sonication. (43) Prepared as per ref 31. (44) Substrates were placed 1 cm away from a Hg pen lamp (4.5 mW/cm2) in ambient air for 20 min. Previous experiments and literature reports (see ref 45 for example) have demonstrated the effectiveness of this method in alkanethiol oxidation and subsequent removal from gold surfaces. (45) Tarlov, M. J.; Burgess, D. R. F., Jr.; Gillen, G. J. Am. Chem. Soc. 1993, 115, 5305-5306.
Letters
Langmuir, Vol. 23, No. 4, 2007 1621
Table 1. AFM-Measured (Average) Values of the Features Shown in Figures 1 and 2a
sample
height (nm)
patterned Si/SiOx 4.0 (0.7) patterned Au 7.8 (0.8) Au-Si-1 13 (2) Au-Si-2
6.5 (0.8)
maximum peak-to-peak diameter spacing (nm) (nm) 56 (7) 42 (6) 55 (9) 27 (3)c 57 (6) 30 (3)c
rb
fb
128 (15) 104 (14) 120 (16)
1.07 0.25 1.10 0.14 1.23 0.06
112 (12)
1.12 0.07
a Standard deviations are in brackets. b Calculated by approximating the shape of the topological features to be cylinders with heights and diameters as given in the table and an areal density of 1010 features/cm2. c Average FEGSEM-measured diameter of Au nanodots.
nanoparticles (nanodots) partially embedded in the silicon substrate to an unknown degree. After cleaning, the Au nanodot samples were immersed in 1 mM ethanolic alkanethiol (n-C5H11SH) solution for 1 h followed by rinsing with approximately 100 mL of ethanol (anhydrous) and dried under a stream of nitrogen gas. The resulting surfaces were imaged using tapping-mode AFM (AutoProbe CP, Park Scientific Instruments) using silicon cantilevers (fo ∼ 300 kHz, Asylum Research) to determine the size and shape of the patterned features. Field emission (gun) scanning electron microscopy (FEGSEM) was performed on the gold nanodot samples using a Hitachi S-4700 (1-2 kV, 10 µA, 5 mm working distance, upper detector) in order to measure the diameter of the nanodots. Contact angles were measured using the sessile drop method in a home-built humidity chamber46 and placed on a leveled, temperature-controlled stage. The measurements were made under a relative humidity and temperature of 45(2)% and 25(1) °C, respectively. Advancing (θa) and receding (θr) contact angle measurements were made by placing a ca. 5 µL drop of Ultrapure MilliQ (18 MΩ) water on the surface with a syringe, subsequently adding/removing ∼1 µL quantities to/from the drop, and retracting the needle before measurement. Given the large degree of adhesion and subsequent pinning of the three-phase contact line observed on the patterned Si/SiOx and Au nanodot surfaces, receding angle measurements obtained for these surfaces are likely systematically less than their equilibrium values. This is common for high surface energy substrates yielding contact angles of approximately 20° or less. A video capture image of the drop was obtained using a CCD camera within 5 s of removing the needle. The contact angle was determined by fitting the drop shape to that of a spherical cap (Multiskop, Optrel GbR). Values reported are the average of at least 3 replicate measurements made on multiple drops on two identically prepared surfaces, with the exception of the Au nanodot patterns, which exhibited unique topological features and hence could not be averaged (see Table 1).
Figure 1. 5 × 5 µm AFM images of (A) flat and (B) nanopatterned template-stripped Au “replicas”. Both images are displayed on a height scale of 0-20 nm.
(eq 5)
with the entire surface. Measured and calculated contact angles are summarized in Table 2. Ion-milling significantly lowers the surface energy of the silicon substrate (θa ) 56°) compared to its unmilled precursor (θ ) 24°). Further comparisons between nanopatterned silicon and “smooth” or unpatterned silicon are thus made with the ionmilled silicon counterpart. The effect of the topology introduced using the nanopatterning methodology on the surface wetting can be made by comparing the substrates presenting chemically homogeneous surfaces (i.e., patterned vs smooth surfaces of SiOx) and Au-RSH. Comparing the experimental and calculated contact angles of these surfaces, the measured (advancing) contact angles are consistent with Wenzel (complete) wetting. Additionally, adhesion of the droplet, manifested as surface pinning and hysteresis, decreased as the nanoscale features became increasingly hydrophobic (Si/SiOx < Au-C5SH < Au-C18SH). Furthermore, the chemical and topological heterogeneities introduced by the patterned gold nanodot substrates yield contact
where f′ is the ratio of the Au nanodot to the total surface area. In contrast to eq 3, eq 5 takes surface roughness into account in the calculation of f′ because the drop is assumed to be in contact
(46) Nitrogen gas was circulated five times through a saturated solution of potassium carbonate and sent to the humidity chamber (80 mL) for vapor exchange at a flow rate of ∼50 mL/min for at least 2 h. The humidity chamber was placed on top of a temperature-controlled stage set at 25 °C.
Results and Discussion Figure 1 shows a typical AFM image of the nanopatterned Au surface described above with its flat counterpart (an unpatterned Au surface). Table 1 lists the dimensions of the features measured from representative images such as those depicted in Figures 1 and 2, and the respective values for r and f as defined by eqs 1 and 3. Advancing contact angles measured on the flat substrates were used to estimate the contact angle of the nanopatterned substrates using eqs 1 and 3. By application of eq 2, eq 5 was then used to calculate the Wenzel contact angle (θw) of the Au nanodots on Si, using the measured advancing contact angles for flat Si/ SiOx and Au surfaces (Table 2).
cos θw ) f′ cos θAu + (1 - f′) cos θSi
1622 Langmuir, Vol. 23, No. 4, 2007
Letters Table 2. Measured Advancing Contact Angles (θa), Hysteresis (∆θ), and Calculated Contact Angles (CA) of Nanopatterned and Smooth Surfacesa experimental CA patterned surface smooth surface calculated CAb surface
θa (°)
Si/SiOx Si/SiOx (milled) Au -C5SH Au -C18SH Au-Si-1-C5SH Au-Si-2-C5SH
n/a 23 (2) 83 (1) 108 (2) 48 (3) 55 (3)
∆θ (°) n/a 10c 2 0 20c 23c
θa (°)
∆θ (°)
θw (°)
θcb (°)
56 (3) 24 (3) 83 (2) 102 (2) n/a n/a
3 6c 7 1 n/a n/a
n/a 12 82 103 29-38d 30-35d
n/a 122 89 153 90 89
a Standard deviations are in brackets. b θ and θ were calculated w cb using eqs 1 and 3, respectively. c A large hysteresis resulted from surface pinning of the drop. d Calculated using eq 5 for a nanodot thickness estimated to be between a lower limit of 1 nm and the total AFMmeasured height.
The experimental results obtained for the nanopatterned surfaces measured here generally agree with the complete wetting scenario predicted by the Que´re´ equation (eq 4). Moreover, evaluation of eq 4 suggests that composite, Cassie-type wetting for a surface with θ g θcrit, and a nanoscale topology as described here (typically 160°. The Que´re´ model also predicts that composite wetting of features with the same areal density as here requires feature heights of ∼185 nm if a nanopatterned n-C18SH-SAM surface (contact angle ) 102°) is used and ∼40 nm if a nanopatterned n-perfluoroeicosane film is used.
Conclusions
Figure 2. 3.5 × 3.5 µm AFM images of (A) nanopatterned Si/SiOx and (B) Au nanodots (Au-Si-2). Both images are displayed on a height scale of 0-20 nm.
angles of intermediate values between those predicted by eqs 5 and 3. Although most of the surface is wetted by water, it appears that some incomplete wetting has also occurred, likely due to the presence of metastable states arising through surface pinning of the drop. Overall, the contact angles measured suggest that Wenzel wetting is the preferred state. In general, our measurements are consistent with accounts of the wetting of surfaces with similar degrees of nanoscale roughness.20,22 Wenzel wetting and an increase in hysteresis with sample hydrophilicity is observed. Others have observed Cassie wetting for systems with larger feature heights. For instance, Cassie wetting is observed for polymer nanopillars formed via anodized alumina porous templates, yielding feature diameters and areal densities similar to those studied here, but with heights typically greater than 50 nm.47,48 Several examples of ultrahydrophobic surfaces whose nanostructure aspect ratios are greater than 10 and micron-scale heights have also been reported.49-52 (47) Guo, C.; Feng, L.; Zhai, J.; Wang, G.; Song, Y.; Jiang, L.; Zhu, D. ChemPhysChem 2004, 5, 750-753.
Contact angle measurements were performed on hydrophobic, hydrophilic, and mixed hydrophobic/hydrophilic nanopatterned substrates. The wetting behavior was found to be consistent with the theory of wetting on rough substrates proposed by Wenzel,33 while the latter surfaces exhibited some incomplete wetting. The patterning process used here offers a versatile approach to surface derivatization. Secondary elaboration of the pattern using chemically selective etching/milling processes (such as reactive ion-beam etching) on the Au nanodot patterns may offer aspect ratios (increased height) currently unavailable in the current methodology. Future efforts include the exploration of the wetting of water on these surfaces with systematic changes in both the topology and derivatization of the Au and/or Si/SiOx substrate. A better understanding of nanoscale wetting is expected to be of fundamental importance in the design of functional surfaces with or without intentional nanoscale architecture. Acknowledgment. We thank NSERC Canada and FQRNTRQMP for financial support for this work. M.-V.M. thanks NSERC Canada for financial support for this research. LA0623275 (48) Lee, W.; Jin, M.-K.; Yoo, W.-C.; Lee, J.-K. Langmuir 2004, 20, 76657669. (49) Feng, X.; Feng, L.; Jin, M.; Jiang, L.; Zhu, D. J. Am. Chem. Soc. 2004, 126, 62-63. (50) Feng, L.; Li, S.; Li, H.; Zhai, J.; Song, Y.; Jiang, L.; Zhu, D. Angew. Chem., Int. Ed. 2002, 41, 1221-1223. (51) Feng, L.; Yang, Z.; Zhai, J.; Song, Y.; Liu, B.; Ma, Y.; Yang, Z.; Jiang, L.; Zhu, D. Angew. Chem., Int. Ed. 2003, 42, 4217-4220. (52) Feng, L.; Song, Y.; Zhai, J.; Liu, B.; Xu, J.; Jiang, L.; Zhu, D. Angew. Chem., Int. Ed. 2003, 42, 800-802. (53) Abdelsalam, M. E.; Bartlett, P. N.; Kelf, T.; Baumberg, J. Langmuir 2005, 21, 1753-1757.
