Introduction

For over a decade, several methods to obtain superhydrophobic surfaces have been developed due to great industrial interest in reducing flow resistance and fluidic drag [1], in repairable superomniphobic surfaces [2], in applications such as optical antireflection [3], corrosion protection [4], liquid transportation [5], biochemical separation [6], antibiofouling paints for boats, production of food packaging, inhibition of ice or snow adhesion [7, 8], and self-cleaning surface properties [9]. Although easy-dewetting surfaces can be obtained in flat/smooth surfaces even with low contact angle [10], superhydrophobic surfaces with self-cleaning properties are usually characterized by exhibiting a very high Contact Angle (CA) and low Contact Angular Hysteresis [11].

The hydrophobic properties of these surfaces depend on their chemistry and roughness [12]. In order to increase surface roughness, there are two possibilities: the first one is thorough its generation on the substrate, such as by plasma etching [13], chemical etching [14] or embossing [15]; the second is thorough incorporation of a coating on the substrate, such as by electrospinning [1618], chemical vapor deposition [1923], lithography [24, 25], dip coating [26, 27], anodic aluminum oxide [9, 28], sol–gel processing [29, 30] or physical vapor deposition (PVD) [15, 31]. Hierarchical structures, i.e., with micro- and nanostructure, have been reported for stability of the superhydrophobic condition [24].

The nanostructure allows the surface to sustain the highest pressure in nature, so as to maintain a robust Cassie state [32, 33]. At the same time, the microstructure significantly reduces the contact area, thereby largely removing adhesion between solid and fluid at the microscopic level [34]. On the other hand, tuning the surfaces in terms of low surface energy is favorable to super hydrophobic surfaces. In this context, a variety of coatings, such as siloxanes [35], fluoroalkylsilane [36], fluoroalkylphosphate [37], and polytetrafluoroethylene (PTFE) [38], have been used.

Many methods to obtain nanostructure and reduce the surface free energy are not easily scaled up, as they request very stringent conditions of preparation and are often obtained at high cost. In this sense, PTFE deposition by PVD methods is a good choice to provide nanostructure and reduce the surface free energy.

Previous reports are present in the literature on the properties of PTFE-sputtered films [3941]. Particularly regarding chemical composition, wettability, and morphology, RF-sputtered PTFE coatings were grown under several bias voltages and gas compositions on silicon wafers [42, 43]. Thermally evaporated PTFE films have also been reported on several substrates, such as Aluminum [9], glass [44] intermetallic oxide alloys [45], micro- and nano-crystalline diamond-like carbon [46, 47] to promote low surface energy and surface nano-roughness. High-energy electron irradiation was used for the chemical modification of bulk PTFE [4852].

Comparison of film structures and properties obtained by electron beam (e-beam) and by RF sputtering deposition of several materials has been reported [5356], but not of PTFE, whose deposition by e-beam was not previously attempted to our best knowledge.

In this work, we report that PTFE deposited by e-beam, in one step, can significantly improve the hydrophobicity of a flat surface, by simultaneous incorporation of low surface energy and nanostructure. After a comparative study of the refractive index, hydrophobicity, surface chemistry, and nanostructure of PTFE-like coatings deposited by e-beam and by RF Sputtering, we pursue superhydrophobic behavior on a Si substrate.

Experimental

Materials

Silicon Prime (100) wafers were supplied by Wacker Chemitronic, polished on one side and doped as p-type. These were sliced into rectangular, flat Si substrates (25 × 25 × 0.6 mm3), which were cleaned in a piranha solution (H2SO4:H2O2 at 4:1) for 10 min at 120° C. After that, the Si substrates were rinsed in deionized water and then immersed in ammonium hydroxide solution (NH4OH:H2O2:H2O 1:1:4) for 10 min at 70–80 °C, followed by rinsing with copious amounts of deionized water and drying under nitrogen flow.

Polytetrafluoroethylene (PTFE) from Dupont was cleaned by ultrasound in isopropyl alcohol before deposition by physical vapor deposition (PVD), RF sputtering and e-beam, methods. Ethanol (98.5 %), H2SO4 (98 %), H2O2 (98.5 %), and NH4OH (98 %) were supplied by Nuclear Brazil.

Physical vapor deposition methods

Electron beam bombardment (e-beam)

Teflon-like films were deposited by e-beam, using a PTFE target, on silicon substrates by a 3 kW Thermionics RCR0304-BD Electron Beam source. The e-beam system was pre-programmed to operate at a base pressure of about 4.0 × 10−6 Torr and work pressure of about 2.4 × 10−4 Torr. During deposition, the substrates remained fixed on a rotary sample system, at a distance of 40 cm from the target.

RF magnetron sputtering

Radio Frequency, Magnetron Plasma Sputtering of PTFE was carried out with a Leybold Univex 450B Sputtering System, using Ar sputtering gas plasma on a PTFE 7.5 cm diameter target, under RF (13.56 MHz) power. The system was pre-programmed to operate at a base pressure of about 4.0 × 10−6 Torr and work pressure of about 1.0 × 10−4 Torr. During the evaporation process, the samples were fixed on a rotary sample system at a distance of 15 cm from the PTFE target.

Chemical etching

1 % wt of KOH in water solution was used to etch the Silicon surface, which was left for 40 min in the solution at 80 °C. Pyramidal structures were formed, following the (1, 1, 1) Si cleavage plane.

Deposition rates

Film thicknesses and deposition rates were monitored, in real time, by a quartz crystal microbalance during both PVD processes. Comparable deposition rates were established, with regard to later comparison of the resulting films, as shown in Table 1.

Table 1 PVD deposition parameters during PTFE-like film production

Surface characterization

Scanning electron microscopy (SEM) and atomic force microscopy (AFM)

The morphology of the PTFE nanostructure was imaged by scanning electron microscopy (JEOL JSM 6060), using an electron acceleration of 15 kV, and by Atomic Force Microscopy (Dimension 3100 DIVM Group).

Surface Roughness and the Wetting Parameter κB, that relates wetting with surface topography [5760], were evaluated using the PSD (Power Spectral Density) analysis.

X-ray photoelectron spectroscopy (XPS)

XPS spectra were obtained in a conventional electron spectrometer (Omicron) equipped with a high performance hemispherical energy analyzer, with a seven-channeltron detector and using Al Kα radiation as the excitation source. Surveys of spectra were recorded, with pass energy of 50 eV, whereas selected atomic signals were acquired with 10 eV of pass energy. In the spectra, the position of the C–C/C–H was specified and other peaks of different carbon environments were fixed, relative to this peak, set at 285.0 eV. The C 1 s envelope was analyzed and peak-fitted after the subtraction of a Shirley background, using Gaussian-Lorenzian peak shapes, obtained from the Casa XPS software package. Sputtering of Si substrates was carried out with a standard ion gun, using Ar ions as primary particles.

Contact angle measurement

Water contact angle was measured by the dynamic sessile drop method using a Krüss DS-30 goniometer. Droplets of deionized water were placed, with a motor-driven syringe, onto the sample surface. Each equilibrium contact angle was determined from a tangent line, appropriately positioned on the contour of the droplet with respect to the surface plane. The advancing and receding contact angles were measured on both sides of the drop for each sample. The reported values are an average of more than 20 measurements, performed on at least three different locations of each sample surface, using a computer-controlled device (provided by Krüss).

Refractive index measurement

Thickness and refractive index of PTFE-like films were measured with the Spectroscopic Ellipsometer SOPRA GES-5E, which uses one rotating polarizer, one fixed analyzer and the Hadamard transform method. All measurements were performed at six different micro-spots on the sample, at a light beam angle of incidence of 68o and in the wavelength range of 350–750 nm.

Results and discussion

Figure 1 shows a comparative summary of the PTFE-like film properties, deposited on polished Si (100) substrates, by RF Sputtering and by e-beam. Contact angles and refractive indices are shown, in Fig. 1a, for films deposited by RF argon plasma sputtering at three different powers (100, 150, and 200 W) and, in Fig. 1b, for films deposited by e-beam at three different currents (5, 10, and 20 mA).

Fig. 1
figure 1

Contact Angle and Refractive Index at 633 nm of PTFE-like films deposited on polished Si flat substrates by: a RF Sputtering at different Ar ion-beam energies, and b e-beam at different currents. Lines were drawn as an aid to the eye

Comparatively, as shown in Fig. 1 and Table 2, PTFE-like films deposited by RF Sputtering (Fig. 1a), present higher refractive indices than those obtained by e-beam (Fig. 1b). The refractive index of the RF-Sputtered films increased with deposition power (Fig. 1a) as expected, since higher Ar plasma energies promote molecular mobility during deposition, and thus higher film packing densities. Also sputtered molecules arrive at random angles on a near substrate, whereas a more collimated molecular flux arrives at longer distances during e-beam deposition. In this case, larger voids and thus smaller refractive indices can be produced, as shown in Fig. 1b and Table 2.

Table 2 Properties of PTFE-like films on Si, deposited by RF sputtering and e-beam processes

As reported in the literature, the contact angle of PTFE-like films, deposited by RF Sputtering, depends on Ar ion-beam energy [40, 41]. Basically, when large moieties of CF2 and CF3 groups are generated, the surface energy is reduced and thus the contact angle increases. Also the contact angle of PTFE-like films, obtained by e-beam (Fig. 1b and Table 2), increases with electron beam current.

Additionally, PTFE-like films obtained by e-beam can present higher contact angles than those obtained by RF Sputtering (see Fig. 1 and Table 2). These results can be originated from two factors: (1) the distinct surface chemical compositions of the treated samples, and (2) their different roughnesses, due to the presence of nanostructure.

XPS and SEM measurements were carried out to study in detail the surface chemical composition and morphology of both PVD methods. Figure 2 shows the C 1 s spectra of the PTFE-like films obtained by e-beam and by RF Sputtering. Both PVD methods produced a very different surface chemical composition of the prepared coatings. The C 1 s envelope of the e-beam method resembles a PTFE-like film [61], but with a value of the high-energy peak moved to even higher energies than a typical –CF2– signal in pure PTFE (292.53 eV). Therefore, the C 1 s signal position of 294.0 eV (Fig. 2a) can be assigned to a –CF3 functionality [15, 42, 43, 62]. On the contrary, RF Sputtering led to a high degree of PTFE target fragmentation and recombination of the desorbed species on the surface of the Si substrate. Several C 1 s fluorinated chemical environments overlap in the deconvolution as shown in Fig. 2b. Oxidized functionalities, such as C=O or COO, can be included in the fittings as presented in Fig. 2b due to overlapping in binding energies. Partial oxidation of the surface was not observed in the XPS survey spectrum and therefore the contribution of functionalities such as C=O or COO can be ruled out. The larger moieties of CF3 groups showed in Fig. 2a, as compared to Fig. 2b (RF Sputtering), may explain the higher contact angle of the films deposited by e-beam, since higher fluorine concentration at the surface given by the -CF3 group would decrease even more the surface free energy compared to the RF Sputtering coating.

Fig. 2
figure 2

XPS wide scan spectra of PTFE-like films deposited by: a e-beam with current of 20 mA and b RF Sputtering with power of 100 W

The second factor that contributed to higher contact angles from the e-beam deposited films, compared to those from the RF-sputtered films, is the higher roughness of the former, which is evident in SEM Fig. 3 and AFM Fig. 3.

Fig. 3
figure 3

SEM images of PTFE-like films deposited by: a e-beam on flat Si with current of 20 mA, b RF Sputtering on flat Si with power of 100 W and c e-beam on etched Si with current of 20 mA

From the different AFM scan areas, the isotropic PSD functions can be combined into a single PSD function [60, 63] that contains the spectral representation of the morphology in a wide spectral range, as shown in Fig. 5. Here the PSD function related with the e-beam deposited films presents a peak around 10 μm−1. This peak represents the micro- and nanostructure, with dimensions between 65 and 120 nm, presented in Fig. 4a.

Fig. 4
figure 4

AFM topography with 200-20,000 nm Scan Range area windows square side, where Root Mean Square (rms) roughnesses are shown for PTFE-like films deposited by: a e-beam with current of 20 mA and b RF Sputtering with power of 100 W

From the PSD functions related to the e-beam and sputtering processes in Fig. 5, roughnesses and wetting parameters, whose values correspond to the areas below the PSD functions, can be calculated directly [5760] and are shown in Table 3.

Fig. 5
figure 5

PSD functions in a wide frequency range for PTFE-like films deposited by: (red) e-beam with current of 20 mA and (black) RF Sputtering with a power of 100 W (Color figure online)

Table 3 Roughness and κ B wetting parameter obtained from the PSD functions of PTFE-like films

The sub-micrometer roughness produced by the e-beam, shown in Table 3, contributes to increasing the contact angle to values higher than those obtained by RF Sputtering. According to the necessary criterion for superhydrophobicity, κ B > 0.4 [58, 60], the κ B value for the e-beam deposition is close to the threshold for the superhydrophobic state. Also, the κ B value for RF Sputtering in Table 3 explains why by this method a superhydrophobic state could not be obtained, even with the existence of a fluorinated coating (see Fig. 2b).

Due to the two major factors, low surface energy and high roughness, PTFE-like films deposited by e-beam showed a superhydrophobic static behavior. Good adhesion of all the PTFE-like films on the Si (100) surfaces was verified by the Scotch tape test. However, their Contact Angular Hysteresis was high and thus full superhydrophobicity was not achieved. For this, before PTFE deposition, the Silicon substrate was subjected to chemical etching.

The resulting micro-pyramidal structures, shown in Fig. 3c, with the PTFE-like coating nanostructure, lead to a high contact angle (163 ± 2°) and very small hysteresis (<3°), unlike those coated by RF sputtering (154 ± 2° and 40°, respectively), which demonstrate that a full superhydrophobic state can be achieved by e-beam deposition of PTFE on silicon.

The Thermionics e-beam system presents a maximum power of 3KW, whereas the RF sputtering unit can reach a voltage of 3 kV. Since also the Ar+ mass is five orders of magnitude larger than that of the electron, the former presents a much larger momentum than the latter. Electron bombardment is essentially a thermal process, whereas momentum transfer dominates in ion bombardment.

This implies that sputtering leads to a higher degree of PTFE target fragmentation and recombination of the desorbed species on the Si surface, as observed by XPS in Fig. 2, and also that the higher Ar plasma momenta promote molecular mobility during deposition, and thus higher film packing densities, resulting in higher refractive indices (Fig. 1; Table 2) and smaller roughnesses, as shown by SEM (Fig. 3) and by AFM (Fig. 4), as well as (via PSD, Fig. 5) in a smaller κ B Wetting Parameter (Table 3).

Conclusions

A comparative study between PTFE-like coatings deposited by e-beam and by RF Sputtering, under similar deposition rates on flat Silicon substrates, was presented, based on SEM, XPS, Sessile drop, and Ellipsometric measurements. As discussed, Sputtering produces higher film packing densities, and thus higher refractive indices, with lower roughnesses. These, combined with the larger moieties of CF2 and CF3 groups that decrease surface free energy on the e-beam deposited films, result in higher contact angles, which can be linearly adjusted with applied current in the studied range (5–20 mA), up to 159 + 2°, although still with high hysteresis. Further micro-structuring of the pre-coated Si surface by chemical etching allows an even higher contact angle (165 + 2°) with very low hysteresis (<3°). These first results indicate that one-step e-beam deposition is a promising production tool with present Si microelectronics technology, such as for digital microfluidic or lab-on-a-chip devices and for Si solar cells.