Preparation and fluid drag reduction properties of superhydrophobic paper-based films comprising carbon nanotubes and fluoropolymers

2.1 Materials

Gases for preparing CNTs, including nitrogen (N₂), hydrogen (H₂), ammonia (NH₃) and acetylene (C₂H₂), were obtained from Zhenjiang Oxygen Supply Co., Ltd (Zhenjiang, Jiangsu Province, China). 2-Bromoisobutyryl bromide was purchased from Tokyo Chemical Industry Co., Ltd (Tokyo, Japan). Concentrated sulfuric acid and nitric acid were supplied by Dongguan Shuangde Chemical Co., Ltd (Dongguan, Guangdong Province, China). Thionyl dichloride was purchased from Shanghai Baoman Biological Technology Co., Ltd (Shanghai, China). Tetrahydrofuran was purchased from Shenzhen Dongmao Chemical Reagent Co., Ltd (Shenzhen, Guangdong Province, China). Ethylene glycol was the product of Jinan Tianjiang Chemical Co., Ltd (Jinan, Shandong Province, China). Triethylamine was supplied by Nanjing Yedi Chemical Co., Ltd (Nanjing, Jiangsu Province, China). N,N,N′,N″,N″-pentamethyldiethylenetriamine was obtained from Dongguan Hechen Trading Co., Ltd (Dongguan, Guangdong Province, China). The n-butyl acrylate and 2,2,3,4,4,4-hexafluorobutyl acrylate were obtained from Nanjing Baotai Chemical Co., Ltd (Nanjing, Jiangsu Province, China) and Chengdu Huaxia Chemical Reagent Co., Ltd (Chengdu, Sichuan Province, China), respectively. The unprocessed base paper and chrome paper were obtained from Shandong Chenming Paper Group Co., Ltd (Weifang, Shandong Province, China).

2.2 Synthesis of CNTs on nickel films by chemical vapor deposition

Nickel (Ni) films were deposited on silicon substrates by K575X Peltier Cooled High Resolution Sputtering Coater (QUORUM/EMITECH, UK) with base pressure <10^-5 mbar. Then they were heated to 500 °C in N₂ atmosphere at chemical vapor deposition (CVD) reactor (CVD(Z)-06/60/3, Risine Heatek, China), and then the N₂ gas valve was closed while H₂ flowed into the reaction chamber; the reaction chamber was continuously heated up to 900°C in H₂ atmosphere. Next, the H₂ gas valve was closed while NH₃ was introduced to the reaction chamber for 10 min at 900°C . Finally, the NH₃ gas valve was closed and the reaction chamber was cooled down to room temperature under N₂ atmosphere.

The samples above were heated following the same procedure before NH₃ was introduced to the reaction chamber. After that, the H₂ gas valve was closed while NH₃ and C₂H₂ from different valves were introduced into the reaction chamber. Finally, both gas valves were closed and the reaction chamber was cooled down to room temperature under N₂ atmosphere.

2.3 Functionalization of CNTs

CNTs (1.92 g) were treated with a mixture of concentrated sulfuric acid (75 ml) and nitric acid (25 ml) at reflux temperature for 6 h. Then the samples were rinsed with distilled water and filtered. After that, they were dried at 60°C for 48 h. Then, these samples were reacted with thionyl dichloride (20 ml) at 70°C for 24 h. After the reaction was completed, tetrahydrofuran was employed to rinse the samples thoroughly. After that, they were reacted with ehylene glycol (19 ml) at 120°C for 24 h. Tetrahydrofuran was also employed to rinse the reaction products thoroughly. Thus, the CNTs with hydroxyl groups were obtained through these reactions. Then, the CNTs with hydroxyl groups were mixed with tetrahydrofuran (27 ml) and triethylamine (0.9 ml) under N₂ atmosphere, anhydrous, and ice bath conditions. Half an hour later, tetrahydrofuran solution (37 ml) containing 2-bromoisobutyryl bromide (1.02 g) was dropped in the system and agitated for 72 h. These samples were rinsed with tetrahydrofuran and filtered again. Finally, they were dried at 40°C for 6 h.

2.4 Surface graft polymerization on CNTs

The above-mentioned modified CNTs (0.98 g) were mixed with copper bromide (0.32 g) and anisole (21 ml) under ultrasonic environment for 30 min. Second, N,N,N′,N″,N″-pentamethyldiethylenetriamine (0.15 g), a mixture of n-butyl acrylate, and 2,2,3,4,4,4-hexafluorobutyl acrylate with a volume of 57 ml were introduced into the system while purging with nitrogen gas. After that, they were heated to 70°C. The viscosity of the system was closely observed during the reaction. They were treated with ice water until the system became sticky and thick. A small amount of tetrahydrofuran was employed to rinse the reaction products again. In the end, unprocessed base paper was soaked in the final reaction product and dried at room temperature.

2.5 Surface etching of films

The samples were put on top of the watch glass with tetrahydrofuran. They were heated to 55°C for 12 h to etch the surface. Then the samples were taken out and dried at room temperature.

2.6 Measurements

The morphologies of the as-obtained samples were examined using scanning electron microscopy (SEM; JEOL JEM-7001F, Japan), atomic force microscope (AFM; EXPLORER VEECO, USA), and transmission electron microscope (TEM, JEOL JEM-2100, Japan). The chemical structures of the synthesized samples were characterized by Fourier transform infrared spectra (FTIR, NICOLET NEXUS 470, USA). The wetting properties were evaluated using contact angle meter (CA, DATAPHYSICS, Germany). These films were tested to obtain fluid drag reduction properties using rheometer (AR-G2, TA, USA). The thicknesses of water and glycerol column were 300 μm and 500 μm, respectively.

3.1 Growth of CNTs

Figures 1A,B and 2 show the effect of NH₃ pretreatment on the formation of catalyst nanoparticles from homogeneous nickel film. It was found that etching and annealing effect played an important role in forming catalyst nanoparticles. The NH₃ pretreatment reduced the metalparticle size as well as the particle distribution density. The measured particle density was about 9×10⁹/cm². The size observed by AFM ranged from 20 to 40 nm. Root-mean-square roughness was measured about 2.4 nm.

Figure 1:

SEM images of (A) untreated substrate, (B) treated substrate, and (C) the grown CNTs.

Figure 2:

AFM image of formed nanoparticles after pretreatment process, which is consistent with Figure 1B.

The carbon source gas was adsorbed by catalyst nanoparticles at the initial stage. Subsequently, carbon atoms formed by decomposition of C₂H₂ were diffused into the inner of metal particles. When supersaturation of carbon in the metal was achieved, the carbon atoms were extruded from the particles and growth of the CNTs occurred [13]. The diameter of CNTs was controlled by the size of catalyst particles. As shown in Figure 1C, the upper region of CNTs was better aligned than the bottom area. This was due to the formed metastable carbon pushing the catalyst particle upward during the growth of CNTs, since there existed a weak interaction between the metal and the substrate. Moreover, the catalyst particles were gradually passivated [14]. Therefore, the influence of Van der Waals forces above was more obvious than that in the bottom region. In general, CNTs were grown vertically aligned on silicon substrates.

3.2 Analysis of chemical structures

The chemical structure changes during the reaction are shown in Figure 3. The absorption band of unmodified CNTs was scarcely observed while strong absorption peak at 1900 cm^-1 corresponding to a stretching vibration of C=O bond in Figure 3B and C was visible. It was a good indication that 2-bromoisobutyryl bromide reacted with CNTs. In Figure 3C, the characteristic absorption peaks occurring at 1330 cm^-1 and 1380 cm^-1were attributed to stretching vibration of C-F bond. Therefore, the FTIR spectra revealed that fluoropolymers had effectively participated in the polymerization on the surface of CNTs.

Figure 3:

FTIR spectra of (A) unmodified CNTs, (B) functionalized CNTs, and (C) CNTs grafted with fluoropolymers.

3.3 Contact angles and surface energies

Distilled water and ethylene glycol were used as liquid probes to characterize the wettability of films. As shown in Figure 4, the mass ratio of n-butyl acrylate to 2,2,3,4,4,4-hexafluorobutyl acrylate has great influence on hydrophobic properties of films. The hydrophobic groups on the film were inadequate when fluorinated acrylate monomer was supplied deficiently. However, excessive dosage could decrease the reaction activity and conversion rate. The optimal proportion was 3:1. As shown in Figure 5, the surface energy and its polar and dispersive part were calculated from contact angle data according to the Owens-Wendt-Rabel-Kaelble method [15]. It indicated that polar components of surface energy remained small regardless of the monomer feed ratio. Especially the polar component of surface energy was only 0.014 mJ.m^-2when the mass ratio reached 1:1. But the reverse was the large value of dispersive components. The variation of the dispersive component of surface energy was basically stable within a given range while the numerical value was up to 56.3 mJ.m^-2 when the mass ratio reached 6:1.

Figure 4:

Effect of monomer ratio on the hydrophobicity of the samples.

Figure 5:

Effect of monomer ratio on the surface energies of the samples.

3.4. Analysis of microstructures

Figure 6 shows the surface morphologies and water contact angles of uncoated, coated, and etched samples. Fibers and their adjacent holes were visible in Figure 6A. The average width of fibers was about 30 μm. Fibers were closely intertwined with each other, which increased the hydrophilic property of uncoated paper with a water contact angle of only 26.1°. The water contact angle had ballooned to 140.0° after a continuous film was formed on the surface of paper, as shown in Figure 6B. The surface was completely covered with hydrophobic film and internal holes disappeared. The water-friendly ends were directed inward and water-shy ends were arranged along the membrane’s exterior seam to improve the water resistance of the film. CNTs were embedded in a polymer matrix with the growth of film and it had hardly been exposed to the surface. Judging from this point, the outer surface of the film was composed mainly of fluoropolymers. Fluorinated groups arranged closely on the outer surface. The small atomic radius, high electronegativity, and low polarizability of fluorine atoms lead to excellent water resistance of the film. As can be seen from Figure 6C, a micro/nanocomposite structure similar to that of lotus leaf [16] was obtained when the polymer matrix was partly etched. The CNTs with low surface energy were ultimately exposed outside to form thin hydrophobic layers with papillary structures, and water contact angle subsequently increased to 157.7°. In this case, although some fluorine atoms were removed, structures similar to lotus leaf could still significantly improve the water resistance of the film. The paper fibers were completely covered with superhydrophobic film, and water had little chance to infiltrate the internal fibers. In essence, the formation of superhydrophobic surface was cooperation between the inherent low surface energy of the polymer matrix and the appearance of micro/nano structures.

Figure 6:

SEM images of (A) uncoated paper, (B) coated paper, (C) etched sample, and (D) water contact angles on their surfaces.

Figure 7 shows typical TEM images of CNTs before and after the reaction. Figure 7A shows the dense CNTs with diameters around 20–30 nm. The CNTs tended to agglomerate together, thereby making it difficult to disperse them. The walls outside were smooth. The presence of catalytic nanoparticles in inner tubes suggested the weak binding force between catalyst particle and the substrate. It revealed that its growth followed tip growth mode [17]. As shown in Figure 7B, the polymer layers were attached to the surface of CNTs at one end, and the long chains floated in water, repelling each other to prevent aggregation with the assistance of surface groups. The CNTs with rough tube walls were almost completely wrapped by polymeric outer layers.

Figure 7:

TEM images of (A) unmodified CNTs and (B) CNTs grafted with fluoropolymers.

3.5 The forming process of superhydrophobic films

Nickel films were deposited on silicon substrates using the magnetron sputtering method. The NH₃ pretreatment changed the deposited film to small particles, which was propitious to the growth of aligned CNTs. Carbon atoms formed by decomposition of C₂H₂ were diffused into the inner of metal particles. Meanwhile, CNTs grew with the precipitation of carbon atoms. Afterward, CNTs were modified to obtain functional ends, which served as initiators for ATRP. Monomers were subsequently grafted on the surface of CNTs. The paper fibers were completely covered with the composite film, which greatly improved the hydrophobic property of paper. The surface roughness was significantly increased through vapor etching treatment. Meanwhile, CNTs were exposed to form micro/nano and papillary structures similar to that of the lotus leaf and eventually forming superhydrophobic surface. The formation of superhydrophobic film resulted from the low surface energy of film forming materials and unique surface structures.

3.6 Fluid drag reduction on superhydrophobic surface

Distilled water and glycerol were employed to investigate the surface rheological properties of chrome paper, polyethylene packaging film with a water contact angle of 132.7°, and superhydrophobic film comprising CNTs and fluoropolymers. As can be seen from Figure 8, fluid shear stress on the superhydrophobic film was biggest among the three samples at the same shear rate. Polyethylene packaging film ranked the second and chrome paper third. The gap of shear stress widened dramatically with the increase in shear rate. Hence, the velocity change of the fluid along the normal direction was the same as that of shear stress according to Newton’s law of viscosity. It indicated that superhydrophobic film offered the least resistance to liquid flow, followed by polyethylene packaging film. Chrome paper was still at the bottom rung.

Figure 8:

Surface rheological properties of chrome paper, polyethylene packaging film and superhydrophobic film comprising CNTs and fluoropolymers using (A) distilled water and (B) glycerol.

The fluid resistance was reduced owing to the existence of bubbles or air layers on the surface [18], [19], [20]. A no-slip boundary condition is assumed for the surface of chrome paper. It is calculated as

τslipτno-slip|c=11+(δ/h) [21],

where τ_slip is the shear stress on the superhydrophobic film, τ_no-slip is the shear stress on the chrome paper, c stands for Couette flow, h is the thickness of water or glycerol column, and δ is effective slip length. Figure 9 shows the changes in slip length at different shear rates. The slip length increased sharply with the increase in the shear rate at the initial stage but decreased rapidly and then tended to be stable. Glycerol had a greater density than water did, which reduced the velocity gradient in a fluid flow. Consequently, the slip length increased rapidly. The internal surface was constantly infiltrated by liquid flow with the increasing shear rate, which resulted in the transition from Cassie state to Wenzel state on the solid-liquid interface [22]. Accordingly, the slip length decreased and then fluctuated in a tight range with the increasing shear rate.

Figure 9:

The slip length on superhydrophobic film comprising CNTs and fluoropolymers using (A) distilled water and (B) glycerol.