Influence of Nanotips on the Hydrophilicity of Metallic Nanorod Surfaces

PRL 100, 256102 (2008)

week ending 27 JUNE 2008

PHYSICAL REVIEW LETTERS

Influence of Nanotips on the Hydrophilicity of Metallic Nanorod Surfaces D.-X. Ye* and T.-M. Lu Department of Physics, Applied Physics and Astronomy, Rensselaer Polytechnic Institute, Troy, New York 12180-3590, USA

T. Karabacak Department of Applied Science, University of Arkansas at Little Rock, Little Rock, Arkansas 72204, USA (Received 9 November 2007; published 24 June 2008) The hydrophilicity of vertically aligned metal nanorods with sharp nanotips were investigated experimentally. Ruthenium and platinum nanorod arrays were deposited on flat silicon substrates using oblique angle sputter deposition. We show that the effects of nanotips on nanorods should be considered in the ‘‘hemiwicking’’ model for hydrophilic metallic samples. With the influence of nanotips, we successfully explained the experimental contact angles of water sessile drops on metallic nanorod surfaces. Our experiments confirm that the shape of the nanorods is an important parameter in determining the hydrophilicity of the nanostructured surfaces. DOI: 10.1103/PhysRevLett.100.256102

PACS numbers: 68.08.Bc

The study of the liquid-solid interface thermodynamics, including the energetics, chemistry, and structure of interfaces, with a focus on wetting phenomena, still attracts considerable attention [1,2]. In particular, the contact angle of a connected drop of liquid resting on a planar solid surface has been related to the interfacial energies of the surface [1]. The prediction of contact angles can be obtained from the consideration of a thermodynamic equilibrium between the three phases on the surface: the solid phase of the substrate (S), the liquid phase of the droplet (L), and the vapor phase of the ambient (V). We denote the interfacial energy of the solid-vapor interface as SV , the interfacial energy of the solid-liquid interface as SL , and the interfacial energy of the liquid-vapor interface as LV . Under thermodynamic equilibrium, minimization of the total interfacial energy of this three-phase system yields the classical Young law, cos0 SV SL =LV , for the contact angle 0 [1– 4]. The equilibrium contact angle 0 determined by Young’s law is experimentally observable only in an ideal thermodynamic system under particular conditions such as the homogeneity and flatness of the bounding surfaces [5]. In practice, however, the bounding surfaces may be rough and may contain more than one element. Therefore, the equilibrium contact angle measured on a real surface often deviates from the prediction of Young’s law. There are two models, Wenzel’s model [6] and the Cassie-Baxter model [7], that have been developed to describe the contact angle on a real surface. Recently, superhydrophobic surfaces with a measured contact angle > 150 have been realized by several researchers using different methods including material synthesis and surface patterning [8]. However, limited effort has been made to achieve the superhydrophilic state of solid surfaces with nanoscale or microscale roughness [9– 12]. Martines et al. fabricated superhydrophilic SiO2 pillars with different tip curvature using the combination of e-beam patterning and plasma dry etching processes [9]. 0031-9007=08=100(25)=256102(4)

Fan et al. reported the work of fabricating vertical aligned Si nanorods using oblique angle electron evaporation technique with substrate rotation [10]. There was no documented study of the water wettability of metallic films with nanoscale or microscale roughness, although there are many important applications of such a film in hydrogen production and storage, surface catalyst, and heat transfer, etc. In this Letter, we studied the water wettability of metallic nanorod surfaces created by oblique angle sputter deposition, which is similar to the technique used by Fan et al. [10]. In our experiments, Ru and Pt cathodes (about 12.7 cm in diameter) and Si substrates were housed in a high vacuum dc magnetron sputtering system. The system was maintained at an Ar pressure of 2.0 mTorr during the deposition. The sputtering power used was 200 W. As such, the deposition rates of Ru and Pt nanorods were measured to be 4:0 0:2 and 9:2 0:1 nm= min , respectively. The Si substrates were mounted on a stepper motor with a distance as large as 15 cm measured from the center of the cathode. The rotation speed was set to 30 revolutions per minute for all the experiments. The deposition flux arrived at the Si substrates at an angle of 85 with respect to the substrate surface normal. With this oblique incident flux and the substrate rotation, the deposited materials on Si substrates formed a porous columnar structure in the nanometer scale, or ‘‘nanorods,’’ due to the shadowing effect and limited surface diffusion. The individual metallic nanorods fabricated using our method have been shown to possess a single-crystal structure [13]. A clear pyramidal-shaped nanotip was also formed on the top end of the nanorod as depicted by scanning electron microscopy (SEM) [13]. For example, in Fig. 1, we presented the high resolution SEM micrographs of selected Ru and Pt nanorods with a nanotip. A wide range of nanorod heights were covered for water wettability studies, from a few nanometers to about 500 nm for Ru nanorod arrays and about 800 nm for Pt nanorod

256102-1

© 2008 The American Physical Society

PRL 100, 256102 (2008)



FIG. 1. Scanning electron microscopy (SEM) images of short (a) ruthenium (Ru) nanorods and (b) platinum (Pt) nanorods. The insets are the cross-sectional images of the Ru and Pt nanorods. The tall nanorods in (c) and (d) have a sharp apexlike nanotip. The scale bars of the SEM micrograph are 100 nm.

arrays, as listed in Table I. The nanorods grow with the deposition time and the pyramidal-shaped tips develop and grow as well. We observed that, however, the shape of the tips are similar to each other as they grow; that is, the facets of the tips self-aligned to a specific crystal plane due to the evolution of the single-crystal texture [13]. The water contact angle measurements were carried out following the commonly used sessile drop method. In our experiments, a drop of 5 l deionized water was deposited on the surface of our nanostructured samples. The measured contact angles on the two types of metallic nanorods (Ru and Pt) demonstrated that the contact angles strongly depend on the morphology of the nanostructured surfaces. For short deposition time, the surface resembles a rough sur-

face [Figs. 1(a) and 1(b)], but for long enough deposition time, the surface should be treated as a heterogeneous surface. Thus, the behavior of the water contact angles follows different models. In the case of rough surfaces, as shown in Fig. 2(a), the experimentally measured contact angles can be related to the equilibrium contact angles 0 by Wenzel’s law as cos r cos0 [6], where the surface roughness factor r is defined as the ratio of the total surface area of a rough surface and the corresponding projected area 0 on a horizontal plane, i.e., r =0 . The surface roughness factor r can dramatically change the contact angles on a solid substrate [14,15]. In the case of heterogeneous surfaces and porous surfaces, the approach of Cassie and Baxter has good agreement with experimental results [7,15]. The Cassie-Baxter theory assumes that a heterogeneous surface is composed of patches with different surface energies. The observed contact angle is a weighted average of the equilibrium contact angles of all the patches with their surface area concentrations as the weights. In a heterogeneous system with two components i and j, the fraction of the surface is assumed to be i and j with i j 1. The equilibrium contact angles of ideal flat surfaces Si and Sj are 0i and 0j , respectively. The Cassie-Baxter theory gives the contact angle of the two-component system as [7,15] arccosi cos0i j cos0j :

(1)

In general, a porous surface can be treated as a composite surface either with solid and air for a hydrophobic surface or with solid and liquid for a hydrophilic surface. For a rough hydrophilic surface, 0i 0 and i for the flattop solid part, and 0j 0 for the water part; thus, the contact angle given by Eq. (1) is arccos cos0 1:

(2)

TABLE I. The average height H, diameter a, and separation d of Ru and Pt nanorods measured from the SEM images. Sputter time t (min)

Height H (nm)

10 20 25 30 60 90 120

45:2 2:1 86:6 3:7 105:7 4:5 121:6 6:2 252:5 3:7 370:2 9:6 480:6 9:6

5 8 15 35 60 75 90

48:3 1:6 81:2 3:2 145:0 4:6 327:2 5:7 566:8 9:3 701:6 7:8 820:0 6:9

Diameter a (nm)

Separation d (nm)

16:5 4:3 36:1 5:3 31:1 4:4 46:1 5:9 56:7 13:2 96:0 10:5 114:1 20:7

22:4 3:5 40:4 7:1 36:3 6:8 50:3 5:8 65:5 12:1 129:2 19:3 288:5 40:8

25:2 4:1 32:7 4:8 55:1 5:7 125:8 12:3 145:8 13:6 162:3 15:0 205:1 27:7

30:0 2:8 41:2 3:7 67:5 4:5 166:7 13:5 203:5 13:3 264:9 14:3 350:6 37:2

Ru nanorods

Pt nanorods

256102-2

PRL 100, 256102 (2008)


FIG. 2. Schematics of (a) Wenzel’s law of water contact angle on a rough surface and (b) ’’hemiwicking’’ model of water contact angle on a hydrophilic surface with nanotips. The tips are supposed to be ‘‘dry’’ outside of the sessile drop of water. The contact angles of nanorod samples can be described using Wenzel’s law when the nanorods are short. The contact angles then follow the hemiwicking model as the nanorods grow to longer ones.

Equation (2) can also be derived from the minimization of the surface energy of the composite interface [16]. The contact angles can be changed by the arrangement of the surface fraction of the solid, as suggested by the CassieBaxter model. It is necessary to point out that Eq. (2) is valid only for a smooth composite interface; modified equations can be obtained for a geometrically clear defined interface [9,15,17]. In our present work, the metallic nanorod arrays with sharp tips were randomly arranged on the Si substrates with a characteristic average separation d of adjacent nanorods [18]. Therefore, unlike the conventional flattop composite ‘‘hemiwicking’’ model [described by Eq. (2)], our nanorod’s water wicking behavior is schematically shown in Fig. 2(b). Here we assume that the surface of the tips is smooth and the pyramidal surface remains dry. The total pyramidal surface can be estimated from the SEM images of the samples. We measured the average of diameter a and separation d of the nanorods. The tilted angle of the facets of the nanotips was measured with respect to a horizontal plane from the cross-sectional SEM images. The angle is about 44.5 for Ru nanorods and about 35.3 for Pt nanorods. We approximated the individual nanorod as a pillar with a pyramidal-shaped apex and obtained a2 =d2 and r 1= cos. Thus, Eq. (2) should be modified to give 2 a cos0 1 1 : (3) arccos d2 cos In Fig. 3, we show the side view shape of water drops deposited on Ru nanorod arrays with different heights of the nanorods. The contact angle of a water drop on a smooth Ru film was determined to be 0 65:9 1:5 in our experiment. From the CCD images, the contact angle decreased dramatically from 0 to about 30 with short Ru nanorods; the contact angle decreased slowly thereafter. The same behavior of the change of contact angles on Pt nanorod arrays has been observed in our experiment. The contact angles were predicted a priori using Wenzel’s law and Eq. (3) with the parameters of the nanorods listed in Table I.


FIG. 3 (color online). Images of water sessile drops sit on Ru nanorod samples with different rod heights. A drop of 5 l deionized water was deposited on each samples for the measurement of contact angles. From (a) to (d), the height of Ru nanorods increases while the measured contact angle of the drops decreases. From (a) to (d), the nanorod height is about 45.2, 121.6, 252.5, and 370.2 nm, respectively.

The measured and predicted contact angles of the Ru and Pt nanorod samples using Wenzel’s law and the hybrid model are shown in Figs. 4(a) and 4(b), respectively. In the first region in Fig. 4(a), the Ru nanorod arrays started to grow from 0 to less than 100 nm in height and demonstrated a Wenzel-type wetting behavior; i.e., the measured contact angles can be predicted using Wenzel’s law. From the cross-sectional SEM images, the tips of the Ru nanorods did not develop clearly yet at this early growth stage. The roughness factor r was thus estimated as r 1 aH=d2 , where H is the average height of the nanorods, by assuming the shape of nanorods to be a cylinder with a flat top. Similar behavior of the Pt nanorods was observed and presented in the first region in Fig. 4(b). In this part, the height of nanorods has an effect on the surface roughness factor, as well as the diameter and the separation of the nanorods. The growth of the diameter and separation of nanorods are coherently related through the shadowing effect. Therefore, the ratio a=d is roughly constant. When the nanorods reach a critical height, Wenzel’s equation is no longer valid. Wenzel’s equation predicts a complete wetting state of the samples with relatively short nanorods, about 100 nm for Ru samples and about 200 nm for Pt samples. However, our measurements show a great difference from this scenario. In our case, the decrease of the contact angles is gradual after Wenzel’s region in the curves in Figs. 4(a) and 4(b). That is, the wetting mechanism changes at a certain height. The second part of the curves in Fig. 4(a) shows a nonflat shape that is attributed to the influence of the development of the nanotips on top of each nanorod. We could calculate and predict this part of the contact angles using the abovementioned hybrid model of hemiwicking as given by Eq. (3). The parameters for the calculation were listed in Table I and no adjustable parameter was used. The calcu-

256102-3

PRL 100, 256102 (2008)



This work is financially supported by National Science Foundation under Grant No. NIRT-0506738. We thank Meral Reyhan, Milan Begliarbekov, and Li Li for their assistance in the contact angle measurements. We also thank Dr. Gwo-Ching Wang of Rensselaer Polytechnic Institute for valuable suggestions.

FIG. 4. Measured contact angles on (a) Ru nanorods and (b) Pt nanorods with varying nanorod height H. The heights of the nanorods were measured from the SEM cross-sectional images of corresponding samples. The contact angles cannot be predicted by a single model. Specifically, the experimental data of contact angles were predicted using Wenzel’s law for short nanorods and hemiwicking model for long nanorods.

lated results show an excellent agreement with the measured contact angles. In Eq. (3), there is no parameter height H and it has no effect on the change of contact angles. Instead, the change of the contact angles depends on the average diameter a and separation d of the nanorods, as well as the shape of the growing nanotips. We also presented the predicted curve in Fig. 4(b) for the Pt nanorod samples. Finally, a small contact angle was measured for large heights in our experiment, indicating a possible superhydrophilic state of the Ru and Pt nanorod arrays. However, this state is retarded by the development of the nanotip of metallic nanorods. A significantly tall nanorod has to be grown in order to achieve the complete wetting of metallic nanorod surfaces. In conclusion, we have demonstrated that Ru and Pt nanostructured surfaces can be wetted by water if the surface is engineered to consist of tall nanorod arrays. The measured contact angles on these surfaces depend greatly on the morphology of the nanorods. Even the pyramidshaped tips of the nanorods play an important role in the wettability of the metallic nanorod arrays. However, one should be aware that not all metallic surfaces can be tailored to be hydrophilic by growing nanorod arrays [19].

*[email protected] [1] P. G. de Gennes, Rev. Mod. Phys. 57, 827 (1985). [2] A. W. Adamson and A. P. Gast, Physical Chemistry of Surfaces (Wiley-Interscience, New York, 1997); Contact Angle, Wettability and Adhesion, edited by K. L. Mittal (VSP, Utrecht, The Netherlands, 1993). [3] T. Young, Philos. Trans. R. Soc. London 95, 65 (1805). [4] G. Alberti and A. DeSimone, Proc. R. Soc. A 461, 79 (2005). [5] R. Finn, Equilibrium Capillary Surfaces (Springer-Verlag, New York, 1986). [6] R. N. Wenzel, Ind. Eng. Chem. 28, 988 (1936); J. Phys. Colloid Chem. 53, 1466 (1949). [7] A. B. D. Cassie and S. Baxter, Trans. Faraday Soc. 40, 546 (1944). [8] Z. Wang, L. Ci, P. M. Ajayan, and N. Koratkar, Appl. Phys. Lett. 90, 143117 (2007); D. Que´re´, Rep. Prog. Phys. 68, 2495 (2005), and the references therein. [9] E. Martines, K. Seunarine, H. Morgan, N. Gadegaard, C. D. W. Wilkinson, and M. O. Riehle, Nano Lett. 5, 2097 (2005). [10] J.-G. Fan, X.-J. Tang, and Y.-P. Zhao, Nanotechnology 15, 501 (2004). [11] G. McHale, N. J. Shirtcliffe, S. Aqil, C. C. Perry, and M. I. Newton, Phys. Rev. Lett. 93, 036102 (2004). [12] J. Bico, C. Tordeux, and D. Que´re´, Europhys. Lett. 55, 214 (2001); J. Bico, U. Thiele, and D. Que´re´, Colloids Surf. A 206, 41 (2002). [13] T. Karabacak, A. Mallikarjunan, J. P. Singh, D.-X. Ye, G.-C. Wang, and T.-M. Lu, Appl. Phys. Lett. 83, 3096 (2003); F. Tang, T. Karabacak, P. Morrow, C. Gaire, G.-C. Wang, and T.-M. Lu, Phys. Rev. B 72, 165402 (2005); P. Morrow, F. Tang, T. Karabacak, P.-I. Wang, D.-X. Ye, G.-C. Wang, and T.-M. Lu, J. Vac. Sci. Technol. A 24, 235 (2006). [14] T. Onda, S. Shibuichi, N. Satoh, and K. Tsujii, Langmuir 12, 2125 (1996). [15] D. Que´re´, Physica (Amsterdam) 313A, 32 (2002). [16] J. Bico, C. Marzolin, and D. Que´re´, Europhys. Lett. 47, 220 (1999). [17] M. Nosonovsky and B. Bhushan, Microsystem Technologies 11, 535 (2005). [18] F. Tang, T. Karabacak, L. Li, M. Pelliccione, G.-C. Wang, and T.-M. Lu, J. Vac. Sci. Technol. A 25, 160 (2007); C. Yang, U. Tartaglino, and B. N. J. Persson, Phys. Rev. Lett. 97, 116103 (2006). [19] M. E. Abdelsalam, P. N. Bartlett, T. Kelf, and J. Baumberg, Langmuir 21, 1753 (2005); N. J. Shirtcliffe, G. McHale, M. I. Newton, and C. C. Perry, Langmuir 21, 937 (2005).

256102-4