ARTICLE pubs.acs.org/Langmuir
Electrophoretic Deposition of Unstable Colloidal Suspensions for Superhydrophobic Surfaces Young Soo Joung and Cullen R. Buie* Department of Mechanical Engineering, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, Massachusetts 02139, United States ABSTRACT: A novel method to fabricate superhydrophobic surfaces using electrophoretic deposition (EPD) is presented. EPD presents a readily scalable, customizable, and potentially low cost surface manufacturing process. Low surface energy materials with high surface roughness are achieved using EPD of unstable hydrophobic SiO2 particle suspensions. The effect of suspension stability on surface roughness is quantitatively explored with optical absorbance measurements (to determine suspension stability) and atomic force microscopy (to measure surface roughness). Varying suspension pH modulates suspension stability. Contrary to most applications of EPD, we show that superhydrophobic surfaces favor mildly unstable suspensions since they result in high surface roughness. Particle agglomerates formed in unstable suspensions lead to highly irregular films after EPD. After only 1 min of EPD, we obtain surfaces with low contact angle hysteresis and static contact angles exceeding 160. We also present a technique to enhance the mechanical durability of the superhydrophobic surfaces by adding a polymeric binder to the suspension prior to EPD.
1. INTRODUCTION Over the past decade, superhydrophobic surfaces have been explored due to their promising applications in diverse areas such as self-cleaning surfaces,1,2 drag reduction,3,4 microfluidics,4,5 heat transfer,6,7 and antiwetting textiles.8,9 To date, dozens of fabrication methods have been investigated to produce superhydrophobicity. Manufacturing demands for superhydrophobic surfaces include process simplicity, low manufacturing cost, environmental compatibility (i.e., nontoxic), scalability, and potential for mass production. With these considerations in mind, we present electrophoretic deposition (EPD) as a potential tool to produce superhydrophobic surfaces. EPD employs electrophoresis of charged particles in dielectric solvents to create dense porous films and structures.10 When a sufficient electric field is supplied to a colloidal suspension, charged particles are attracted to and deposit upon the oppositely charged electrode. Among many applications, EPD has been investigated to fabricate microscale and nanoscale structures.11,12 EPD has also been explored to develop novel electrodes and catalyst layers for electrochemical systems,13,14 since EPD is considered as an effective technique to control porosity, surface area, and density of porous films. EPD is a well-established process, but wettability of structures fabricated with EPD has largely been overlooked. In one particular study, the wettability of thin films produced by EPD with titanate nanotubes was investigated.15 In their work, the surface of the titanate deposition layer was switched from superhydrophilic to superhydrophobic after a surface modification with 1H,1H,2H,2H-perfluorooctyltriethoxysilane. The porous structure of the titanate deposition layer was considered the r 2011 American Chemical Society
primary factor to produce superhydrophobicity, but EPD itself was not investigated as a tool to control wettability. Recently, Ogihara et al. demonstrated the possibility of using EPD to fabricate superhydrophobic surfaces.16 Several hydrophobic particles including carbon black, activated carbon, vapor-grown carbon nanofibers, titanium dioxide, beta-type copper phthalocyanine, and phthalocyanine green were used to produce superhydrophobic surfaces. However, they did not explore mechanisms to control wettability with EPD or address the relatively weak adhesion of EPD surfaces. The purpose of this work is to illuminate how EPD can be utilized to control surface roughness and achieve superhydrophobicity. Previous EPD studies have used suspension stability, electric field, and deposition time as variables to control surface roughness of deposited films for applications in medicine and ceramics, but not wettability.17,18 This work is a significant extension beyond the studies of Lai et al.15 and Ogihara et al.16 as we elucidate the role of suspension stability and deposition time to enhance surface roughness for the purposes of antiwetting. The effect of colloid stability on surface wettability is explored experimentally, resulting in superhydrophobic surfaces with static contact angles exceeding 160.
2. THEORY When the surface of a particle in an electrolyte is electrically charged, the particle has an electrophoretic mobility, μ. Under an applied electric field, E, the charged particle moves toward an Received: January 21, 2011 Published: March 01, 2011 4156
dx.doi.org/10.1021/la200286t | Langmuir 2011, 27, 4156–4163
Langmuir
ARTICLE
oppositely charged electrode with the velocity, ν, expressed by v ¼ μE
ð1Þ
The mobility, μ, is a function of zeta potential, ζ, permittivity, ε, and viscosity, η, of the fluid as is shown in Henry’s equation,19 μ¼
2εζ 3η
ð2Þ
This equation assumes spherically shaped particles with small r/λD, where λD-1 is the Debye-H€uckel length and r is particle radius. Particles transported to the electrode agglomerate on the surface of the electrode if the electric field is sufficiently high to induce deposition. In EPD, suspension stability is critical since the morphology of the deposition layer is significantly affected by particle flocculation. A stable suspension results in well-dispersed particles, devoid of serious flocculation. In contrast, fast particle sedimentation is observed in unstable suspensions due to particle agglomeration. Interfacial forces between particles determine suspension stability. Two opposing forces are induced between particles in close proximity. The attraction force is commonly known as the van der Waals force, and the repulsive force is due to the electrical double layer. The net interaction potential, Φnet, is the summation of the attractive potential, ΦΑ, and the repulsive potential, ΦR, between two particles. Assuming spherical particles of identical size, the interaction potential can be expressed as20 Φnet ¼ ΦR þ ΦA ¼ 64kB n¥ TλD
ζ0 2 A k ed=λD 12πd2
ð3Þ
where A is the Hamaker constant, d is the distance between particles, kB is the Boltzmann constant, T is the absolute temperature, and n¥ is the bulk ionic concentration expressed as the number of ions per cubic meter. ζ0 is a function of the surface potential, ψ0, defined as20 ζ0 ¼
ezcψ0 =2kB T - 1 ezcψ0 =2kB T þ 1
ð4Þ
where c is the elementary electric charge and z is the valence number. Surface potential is directly proportional to the zeta potential such that we can consider the electric repulsion a function of the zeta potential. To evaluate suspension stability theoretically, the stability ratio, W, is employed and expressed by21 R ¥ eΦnet ðsÞ=kB T ds 0 kr s2 W ¼ ¼ R ¥ eΦA ðsÞ=kB T ks ds 0 s2
ð5Þ
Here, kr is the rate constant for rapid coagulation, ks is the rate constant for slow coagulation, and s is the ratio of the particle radius to the distance between two-particle centers. Equation 5 assumes that fast coagulation occurs when the attraction force dominates and electric double layer repulsion is negligible. From eqs 2 and 5, it is notable that the zeta potential, which can be determined experimentally, influences both deposition rate and stability. In general, higher zeta potential leads to higher deposition rate and improved stability. The effects of pH,22 ionic concentration,23 surfactants,24 and solvent composition25 on stability have already been investigated. In this work, we chose to vary stability by varying suspension pH at a
specified ionic concentration, since the zeta potential is a strong function of pH.
3. EXPERIMENTAL SECTION For all experiments in this study, hydrophobic SiO2 particles (polydimethylsiloxane (PDMS) coated, average particle diameter 14 nm [PlasmaChem, Berlin, Germany]) were used without additional purification or modification. A mixture of 90% methanol (ACS Reagent, Baker analyzed) and deionized (DI) water (by volume) was used as the solvent. The PDMS coated SiO2 particles are hydrophobic, making dispersion in aqueous solvents impractical. Therefore, nonaqueous solvents were used and methanol was chosen since its refractive index (n = 1.33 at 25 C) is similar to water. (Matching refractive index is critical to obtaining reliable stability ratio measurements via absorbance spectra, as explained below.) We note that our technique to create superhydrophobic surfaces is not limited to PDMS modified hydrophobic SiO2 particles and methanol based solvents as presented here. We have also successfully produced superhydrophobic surfaces with both hydrophobic TiO2 particles and SiO2 particles modified by octylsilane (unpublished results). For our colloid stability, zeta potential, and particle size measurements, we use particle concentrations of 0.1 g/L. Following initial dispersion, potassium nitrate (ACS reagent, g99%, Sigma-Aldrich) was added to the suspension in order to adjust the salt concentration to ∼10-6 M. The purpose of the salt is to vary the gradient of the zeta potential curve.23,26-29 After 5 min of sonication (all sonication was conducted with the amplitude 0.25 μm/mL [Sonicator 400, Qsonica, LLC.]), the SiO2 particles were well dispersed in the suspension. Next, the suspension pH was adjusted with acid (nitric acid, 70% ACS reagent, Sigma-Aldrich) or base (potassium hydroxide, 45 wt % solution in water, Sigma-Aldrich). Ten different pHs (3.0, 4.0, 5.0, 6.0, 7.0, 7.4, 7.6, 7.9, 8.0, and 8.3) were prepared for the measurements. Five minutes after the pH was adjusted, the suspension was sonicated again for 5 min. After the suspension settled for 30 s, zeta potential and particle sizes were measured by dynamic light scattering (Zetasizer Nano ZS, Malvern Instruments, Inc.) and suspension absorbance versus time was recorded with a spectrophotometer (wavelength 450 nm, Spectrophotometer UV-1800, Shimadzu). Each measurement at a given pH was conducted three times with independently prepared suspensions. Suspension stability can be quantitatively evaluated from the change of absorbance with respect to time.26,30 A stable suspension shows moderate change of absorbance due to its slow sedimentation. In contrast, unstable suspensions are more likely to have fast sedimentation attributed to the low net interaction potential between particles, as shown in eq 3. This results in a steep gradient in absorbance versus time. Change of absorbance, R, versus time is directly proportional to the initial rate constant, k, for coagulation when time, t, is small, and can be expressed by26,30 dR kKτ N0 2 λ2 ¼ dt ð1 þ kN0 tÞ2
ð6Þ
Here N0 is the initial number of particles, λ is the wavelength of incident light, and Kτ is a proportional constant. The experimental stability ratio can be acquired by dividing the maximum rate of change of absorbance by the rate at a particular pH, as in the following equation, dR dt kr ¼ max ð7Þ Wabs ¼ dR ks dt pH As an alternative, changes in particle size can be employed using dynamic light scattering (DLS) measurements to assess stability.23,31 4157
dx.doi.org/10.1021/la200286t |Langmuir 2011, 27, 4156–4163
Langmuir
ARTICLE
Figure 1. Schematic illustration of the EPD cell. The cell consists of two titanium electrodes with a 15 mm of gap, a DC power supply (10 V applied), and a suspension consisting of hydrophobic SiO2 particles dispersed in a mixture of 90% methanol and 10% DI water by volume. The average hydrodynamic particle diameter, Dh, can be calculated by the Stokes-Einstein equation, Dh ¼
kB T 3πηD
ð8Þ
In eq 8, D is the average translational diffusion coefficient of colloidal particles in dilute suspension, which can be determined by the temporal evolution of intensity fluctuations in dynamic light scattering measurements. The change of particle size, Dh, with respect to time is a function of the initial aggregation rate constant k and the initial particle concentration, C0, when time, t, is small, as dDh ¼ βkC0 dt
ð9Þ
where, β is a constant that depends on scattering angle and material properties of particles. Noting the presence of coagulation rate in eq 9, the experimental stability ratio can be expressed in terms of the ratio of the fast coagulation rate to the slow coagulation rate, Wsize ¼
kr ðdDh =dtÞmax ¼ ks ðdDh =dtÞpH
ð10Þ
If the initial particle size, Dh,i, and the maximum particle size, Dh,max, are known at a specific time t, the stability ratio based on size, Wsize, can be calculated as Wsize ¼
Dh;max - Dh;i Dh;pH - Dh;i
ð11Þ
The size based stability ratio, Wsize, calculated by eq 11 should be comparable with the stability ratio, Wabs, obtained through the absorbance measurement as in eq 7. Notably, the initial particle size can be estimated by comparing both stability ratios. The suspensions used for EPD were prepared using the procedure indicated above except that the SiO2 particle concentration was increased to 1 g/L. Figure 1 shows a schematic of the EPD system. Two titanium plates (purity g 99.6%, annealed, Goodfellow Corporation), with identical dimensions of thickness (0.5 mm), width (10 mm), and length (50 mm), were used as working and counter electrodes. In each experiment, 10 V was applied between two electrodes separated by 15 mm. The EPD process was conducted at ambient temperature and without mechanical stirring. Deposition times were 30, 60, 90, 120, and 180 s. After completing EPD, the sample was dried in ambient air. To improve deposit adhesion, in some cases, two part conductive epoxy (CW2400, Chemtronics) was added into the suspension. After the 1 g/L SiO2 suspension was prepared, 10 g/L of each epoxy was separately mixed into the suspension via 5 min of sonication. After mixing both epoxies, the EPD process was conducted immediately. A stirrer (7 7 in. ceramic top plate, 1 0.5 in. Teflon magnetic stirring bar, VWR) was used with a rotating speed of 1000 rpm during the EPD process in order to maintain well dispersed epoxy in the suspension. The
Figure 2. Characterization of PDMS coated SiO2 particles as a function of pH. (a) Average zeta potential showing the IEP at pH 3. (b) Average agglomerate diameter, indicating smaller agglomerates at high pH. The error bars shown in (a) and (b) represent two standard deviations of three measurements. electric potential was 30 V/cm, and the deposition time was 10 min. Other conditions for the EPD process were the same as those for the nonepoxy suspensions. The morphology of deposited films was characterized with a scanning electron microscope (SEM, JEOL 6320FV field-emission high-resolution SEM). Surface roughness was measured by atomic force microscopy (AFM, DI Nanoscope) at three distinct points on each sample. The area scanned for AFM measurements was 40 40 μm, and the average rootmean-square (rms) surface roughness was calculated using commercial software provided by the AFM manufacturer. A goniometer (Kyowa, DM-CE1) was used to dispense and image 3 μL drops of DI water at four different points on each sample. Static contact angles (CAs) were calculated using the tangential curve-fitting method.
4. RESULTS AND DISCUSSION The first suspension property measured was zeta potential, since it is the key characteristic of a colloidal suspension. Zeta potential measurements were conducted at 10 pH values: 3.0, 4.0, 5.0, 6.0, 7.0, 7.4, 7.6, 7.9, 8.0, and 8.3. Figure 2a shows zeta potential as a function of suspension pH. The isoelectric point (IEP) is roughly pH 3 for the PDMS coated SiO2 particles, and zeta potential generally decreases with increasing pH. From eqs 3-5, one would expect that particles in lower pH suspensions (
