Abstract
Honeycomb porous polystyrene (PS) films with an aspect ratio of pore depth to pore diameter at approximately 1.0 were fabricated using the breath figure (BF) method. Two modes of water droplet coalescence in the pore growth were observed in real-time by optical microscopy. Pore size significantly increases with the increase in humidity and the decrease in substrate temperature. The porous pattern could emerge even at room temperature under high humidity of 80%. Boiling point and solvent density significantly influence the pore distribution and pore depth. Chloroform and tetrahydrofuran achieve more uniform hexagonal patterns than benzene and dichloromethane. Subsequently, to obtain nanometer porous PS film, the fast-evaporation BF process was designed by regulating the gradient substrate temperature and evaporation time, and porous mesoscopic PS film was obtained. The minimum pore diameter and corresponding pore depth are about 120 nm and 27 nm, respectively. Finally, the fast-evaporation BF process was applied to the honeycomb film formation of photovoltaic polymer poly(3-hexylthiophene) (P3HT), and the heat-resistant polymers polysulfone (PSF) and polyimide (PI).
Export citation and abstract BibTeX RIS
1. Introduction
Ordered porous structures, even with pore sizes in the order of micrometers, are of interest for applications in microarrays of biomolecules, patterning templates, superhydrophobic surfaces, sensing and photoelectronic materials, microcontainers or reactors, sustained drug release and separation membranes [1–4]. Owing to the honeycomb structure working as an adhesive site for the cells, honeycomb films have been widely used as biological stimulation by regulating the cell response, or as a scaffold for cell growth with pore sizes from 200 nm to 7 μm [5].
A variety of methods have been developed to create micro- or nano-scale porous films, including direct writing of polymer patterns [6], soft-lithographic methods [7], the one-step titration-gel method [8], electric-field-induced patterning of block copolymers and a combination of breath figures (BFs) with phase separation [9]. In particular, the preparation of honeycomb-patterned films by the BF method has received great attention [10] since the first study reported in 1994 by François et al [11].
The BF method is a self-assembling physicochemical procedure; evaporative cooling leads to self-assembled arrays of condensed water droplets [12]. After the water droplets evaporate, a porous structure with a hexagonal array or square array emerges. The most rigid structure is the hexagonal pattern because of its lowest free energy [11]. The interfaces between the solution and substrate, the solution and water droplets, and the film surface and air play an important role in determining both the array and function of the honeycomb films [13].
Due to the simple, inexpensive and robust pattern formation in the BF process, this method has been intensively applied to prepare various honeycomb films. At the initial stage of the development of the BF method, a variety of amphiphilic polyion complexes, amphiphilic covalent polymers and inorganic/organic hybrid materials were applied and widely studied by Shimomura's and Nishida groups [14, 15]. Up until now, the material has expanded to homopolymers or blend polymers, such as polystyrene (PS), polystyrene-polyfluorene block copolymer, poly (styrene-butyl acrylate-acrylic acid), poly (methyl methacrylate) (PMMA), cellulose acetate butyrate (CAB), polystyrene-polyparaphenylene (PS-PPP) [16], poly(3,4-ethylenedioxythiophene) poly(styrene-sulfonate) (PEDOT:PSS) [17], protein arrays [18], graphene oxide incorporated poly(epsilon-caprolactone) [19] and ultrathin gold nanowire/honeycomb macroporous pattern blend films [20].
Nishikawa et al deposited a dilute solution of an amphiphilic polymer on the surface of water, and obtained a self-standing porous mesh with a controlled film area and thickness on the water's surface [21]. It has been found that even the superhydrophobic polymers, such as PS, can also achieve honeycomb patterns. To obtain the ideal pore size and pore distribution, some researchers modified the traditional BF process by constructing an alcohol–water mixed vapor atmosphere to replace the water vapor [22], by expanding the casting method from drop-casting to dip-casting and spin-coating [23] or via thermal annealing to tune the morphology and hydrophobic properties of porous block copolymer films [24]. It was reported that humidity, viscosity and the behavior of phase separation played an important role in the formation of ordered patterns [25]. Almost all the solvents with lower boiling points than water could be used in the BF method, even water-miscible solvents.
To date, most work focuses on the fabrication of monolayer or multilayer porous films in the order of micrometers [26, 27]. Here, we modified the traditional BF process to a fast-evaporation BF process to decrease the pore size from micrometers to nanometers by regulating the gradient substrate temperature (Ts). The modified BF process is simpler than the chemical procedure [28], sol-gel technique [29] and block copolymer application, which require additional etching to remove the template [30]. In this work, firstly, diverse characteristics of porous films and real-time observation of the water droplet coalescence were carried out. Then, the effects of substrate temperature, relative humidity (RH) and PS solvent on the pore size were investigated. Finally, the modified process was applied for fabricating honeycomb heat-resistant and photovoltaic polymer films.
2. Experimental
All the reagents were of analytical grade. PS (Mw = 20 800 g mol−1), polysulfone (PSF, Mw = 22 000 g mol−1) and polyimide (PI, Mw = 65 000 g mol−1) were purchased from Sigma-Aldrich. Poly(3-hexylthiophene) (P3HT, Mw = 50 000 g mol−1) was acquired from Merck Co., Germany. Chloroform, tetrahydrofuran (THF), benzene and dichloromethane were purchased from Wako Chemical Co., Japan. All commercially available reagents were used without further purification.
All the preparation of the micrometer and nanometer porous films was carried out in a sealed chamber in which a temperature sensor, a Ts controller and a moisture meter were included. Moist air was pushed into the chamber to control the RH. Next, 1.0 wt% PS was dissolved in solvents, and 40 μl PS solution was drop-cast on a clean glass slide. The Ts of glass slides could be well controlled from room temperature (Tr) -10 °C to Tr +15 °C, in real-time. With regards to the fast-evaporation BF process, the Ts was firstly kept at Tr or a lower temperature for a short time. Then, the Ts was raised quickly, which resulted in fast drying of the film to stop the coalescence of water droplets. The porous structure was observed by optical microscopy (E600FN, Nikon, Japan), polarizing microscopy (BH2 BHSP, Olympus, Japan), atomic force microscopy (AFM, CP-II, VEECO, USA) and field-emission scanning electron microscopy (FESEM, S-4800, Hitachi, Japan). Pore diameter and pore depth were detected by AFM and polarizing microscopy, respectively.
3. Results and discussion
3.1. Real-time observation of pore formation and water droplet growth
The 1.0 wt% PS dissolved in chloroform was cast on a glass slide in moist air at the sealed chamber. Evaporative cooling and the generation of an ordered array of BFs led to the formation of hexagonally packed water droplets that are preserved on the final, solid polymer film surface. A schematic diagram of the formation procedure for the porous film is depicted in figure 1(a). Figure 1(b) shows the large-scale image and the laser diffraction image of honeycomb PS film, which is similar to the reported laser pattern [31]. It was reported that the completion of the chloroform evaporation is accompanied by a sudden decrease in the scattering intensity, originating from the change in the refractive index after the chloroform evaporation. The pore depth measured by polarizing microscopy is about 2.5 μm (figure 1(b)).
Figure 1. (a) A schematic diagram of the formation procedure for porous film. (b) A large-scale image, laser diffraction image and the polarizing microscope image of the porous PS film prepared at 65% RH with a Ts of 9.1 °C (Tr = 13.1 °C).
Download figure:
Standard image High-resolution imageSteyer et al suggested that the growth of BFs on liquids evolves through three stages [32]. In the initial stage, there is no strong interaction among the water droplets, and the diameter (D) of the droplets increases with time (t) by a power law of D ∼ t1/3. In the crossover stage, the water droplets are separated by a thin liquid film, and the whole surface has maximal coverage with low dispersity of the droplet diameters. The last stage is coalescence dominated by high and constant surface coverage. Because of the convective currents arising from the evaporation as well as from the air flow across the surface, the water droplets usually pack into a hexagonal array. Based on Steyer's model, we added two stages, including the solvent evaporation and water cluster evaporation, as depicted in figure 2(a).
Figure 2. (a) A model for the formation of honeycomb structured polymer film. (b) Real-time observations of the coalescence of water droplets in the hexagonal matrix (mode I), and of the coalescence of adjacent water droplets (mode II), both at 65% RH. Frames were captured at 60, 120, 180 and 240 s after drop-casting.
Download figure:
Standard image High-resolution imageTo obtain the real-time images of water droplet growth, we photographed the dynamical movement of the rearrangement of water droplets using optical microscopy, as depicted in figure 2(b). The coalescence of hexagonal (up) and neighboring (down) water droplets is exhibited at the left-hand side of figure 2(b). The right-hand side images in figure 2(b) display the growth procedure of water droplets occurring at 65% RH, Ts of 8.4 °C (mode I) and 9.1 °C (mode II), respectively. The pore diameter increases from 1.0 μm to 2.0 μm in the time scale from 60 s to 240 s. Here, the optical microscopic images provide direct proof that there are two different coalescence modes existing in the growth of BFs. When the system reaches an equilibrium condition, the preferred pore arrangement will possess the lowest free energy [33]. In mode I the hexagonal patterned droplets which are repulsed by the lateral capillary force unite towards the center, and finally form a uniform porous film.
In the initial stage of porous pattern formation, the pore distribution in mode I is much homogeneous than that in mode II, probably due to the van der Waals force and Coulomb force, which play a primary role in the initial stage. Van der Waals forces and Coulomb forces in molecular assemblies are short-range forces acting at a molecular level, thus mesoscopic patterning is not expected. On the other hand, surface tension and convection force, which act on a larger length-scale, could construct structures at micrometer scales [15]. Mode I achieves mono-dispersed pores, while mode II results in poly-dispersed pores with scattered smaller droplets surrounding the large core droplet. Finally, the ordered honeycomb pattern forms in the two modes. The Ts difference induces two different modes, implying that the BF process is very sensitive to the Ts.
3.2. Influence of substrate temperature and humidity on pore size
Owing to the high surface tension of the solution surface, in the traditional BF process water droplets easily coalesce so that the final pore size is usually at the micrometer scale. It was reported that the honeycomb hole diameter gradually decreases along the radius of the cast area, and the largest holes are located near the geometric center of the cast area [16, 34]. A schematic diagram of the traditional BF process is shown in figure 3(a). When the Ts was kept at 7 °C under a RH of 85% and Tr of 13 °C, the resultant pore diameter was about 3.0 ∼ 4.0 μm (figures 3(b), (c)), and the pore depth was about 3.0 μm (figure 3(d)). Honeycomb PS films with an aspect ratio of pore depth to pore diameter at approximately 1.0 were obtained.
Figure 3. (a) A schematic diagram of the traditional BF process. An FESEM image (b), and top view (c) and cross-sectional (d) AFM images of the resultant honeycomb porous PS film prepared at a Ts of 7 °C and RH of 85%.
Download figure:
Standard image High-resolution imageMore important influencing factors for pore size are the Ts and RH. The dependence of pore size on Ts and RH in the traditional BF process is investigated. By using the identical process for the Ts control, the effect of RH ranging between 45% and 80% on the pore diameter is investigated. Figure 4 demonstrates the dependence of pore diameter on temperature difference (Tr-Ts) under various RHs. There is a linear correlation between the pore diameter and the temperature difference under all the specific RHs. The critical temperature difference (at which pores begin to appear on the surface) decreases with RH increase, and at a high RH of 80% the porous film could also form even at Tr without substrate cooling. However, at a relatively low RH of 45%, the pores begin to form at a temperature difference of 7.2 °C.
Figure 4. The dependence of pore diameter on the temperature difference between Tr (13 °C) and Ts at various RHs using the identical process as that shown in figure 3. The inset is the optical microscope image of the porous film prepared at RH 80% and Tr.
Download figure:
Standard image High-resolution imageThe effect of humidity on pore-size distribution is also investigated by using a similar control process for Ts as that shown in figure 3. Figure 5 shows the optical microscope images of honeycomb PS films prepared at RHs of 67.0%, 70.9%, 78.7%, 84.3% and 87.0%, respectively. The corresponding average pore diameters are 4.3, 5.3, 6.1, 6.6 and 8.0 μm, respectively. When the RH is over 85%, the pore diameter significantly increases and the pore size becomes heterogeneous, suggesting that the aggregation kinetics of the water droplet are more complex under ultrahigh RH. With regards to the mechanism of humidity effect on pore size, it could be explained from the view point of the evaporation rate, thermodynamic and Young–Laplace equation [35].
Figure 5. Optical microscope images of honeycomb PS films prepared at various RHs and fixed Ts of 8.0 °C (Tr = 13 °C).
Download figure:
Standard image High-resolution image3.3. Influence of fast-evaporation process on pore size
The smallest theoretical size of a water droplet at Tr (≈25 °C) has been reported to be 10 nm [36]. However, until now the minimum pore size using the BF approach is about 80 nm by using a dilute solution (0.2 wt%) with dip-coating [23]. The formation of nanometer pores is limited by the coalescence of water droplets. Pore size is mainly influenced by the nucleation and growth of water droplets. The water droplet size is dictated by a wide range of variables, including the Ts, the evaporation rate of the solvent and the precipitation of the polymer, etc. It is noteworthy that the important factor is the surface thermodynamics. The complexity of the process with its manifold restricts us to make a rough prediction of the pore size by developing empirical relationships.
The video in figure 2(b) shows the droplet-coalescence procedure in the traditional BF process. However, Knobler and co-workers showed that the droplet size was more uniform without coalescence [37]. Nishikawa and co-workers illustrated that the pore size depended upon the evaporation time of the polymer solution [21]. To obtain nanometer scale porous film, we designed a fast-evaporation process by controlling the Ts and evaporation time to impede the coalescence of water droplets.
Figure 6 illustrates four processes for Ts control at RH 85% and Tr 13 °C. The red point on the y axis represents the Tr. The detailed Ts and control time are different in the four processes. For processes A–C, the resultant average pore diameters are 120 nm, 270 nm and 120 nm, respectively. The pore depths are 27 nm, 60 nm and 27 nm, respectively. Process C achieves a similar pore diameter and depth to process A, but much sparser pore distribution.
Figure 6. Diagrams of diverse processes (from A to D) for Ts control at RH = 85% and Tr = 13 °C with chloroform as the solvent (the red point on the y axis represents the Tr). The inset is the AFM image of the obtained porous film.
Download figure:
Standard image High-resolution imageProcess D incorporates three temperature steps to obtain a small pore diameter and a dense pore distribution. Nevertheless, the overall pore diameter and pore depth are 2.0 μm and 1.0 μm with an interpore distance of 5.0 μm. The larger pore size probably originates from the longer evaporation time (40 s) than processes A–C (20 s). It could be concluded that pore diameter and distribution can be adjusted by controlling the Ts and evaporation times of water droplets. A low Ts and short evaporation time could balance the pore diameter and distribution. Among the four processes, process A always achieves porous film with dense sub-micrometer pores at a relatively high RH. In the experiment on the solvent effect, process A was used.
3.4. Influence of solvent on porous pattern
Four organic solvents, including chloroform, THF, benzene and dichloromethane, were used to prepare 1.0 wt% PS solution to study the effect of solvent. The density offset between the organic solvent and water will make the water droplet float or sink into the PS solution, thus obtaining various pore diameters and pore depths. When a water-miscible solvent is used, the pore diameter will be larger than that using a water-immiscible solvent at the same solvent density (ρsolvent). Srinivasarao et al investigated the effect of ρsolvent on two-dimensional and three-dimensional (3D) ordered arrays in porous films, and found that solvents with a smaller ρsolvent than ρwater would induce a 3D array because the water droplets sank into the solution [27]. Limaye et al studied the influence of the solvent nature on the time resolution of the average water droplet diameter [38].
The pore diameter, pore depth and aspect ratio (depth/diameter) of these porous films are summarized in table 1. The pore depths from THF and benzene are larger than from chloroform due to their lower ρsolvent. In addition, the boiling point (BP) of the solvent influences the pore distribution on the surface. The BP of chloroform and THF is similar; therefore, the pore distribution is identical. Moreover, because THF is water-miscible, the pore diameter from THF is 300 nm, which is larger than that (120 nm) from chloroform under identical conditions. These results prove that the ρsolvent and BP significantly influence the pore depth and pore distribution, respectively. A too high or too low BP would result in sparse distribution of pores. Overall, chloroform is the optimal solvent for obtaining ideal porous PS films.
Table 1. Pore diameter, pore depth and aspect ratio (depth/diameter) of the porous PS film using various solvents.
| Solvent | chloroform | THF | benzene | dichloromethane |
|---|---|---|---|---|
| Density (g ml−1) | 1.498 | 0.886 | 0.879 | 1.326 |
| Boiling point (°C) | 61 | 66 | 80 | 40 |
| Pore diameter (nm) | 120 | 300 | 1000 | 100 |
| Pore depth (nm) | 27 | 60 | 80 | 11 |
| Aspect ratio (depth/diameter) | 0.22 | 0.20 | 0.08 | 0.11 |
3.5. Application of the fast-evaporation processes to honeycomb heat-resistant and photovoltaic polymer films
Heat-resistant plastics are a light, versatile alternative to metal, ceramics and older-generation polymers. The glass transition temperature (Tg) is a key index to evaluate the thermal properties of heat-resistant polymers, such as PI, PSF and polyether ether ketone. General PI shows the best heat resistance with a Tg over 250 °C, and the diamantine-based PIs possess a high Tg ranging from 281 °C to 379 °C [39]. Porous PI films possess lower dielectric properties [40], and are thus used as carbon molecular sieving membranes [41]. The homogeneous and inhomogeneous porous pattern of PI was fabricated by etching out the SiO2 microspheres from a polyion complexsor [42].
PSF is an amorphous high performance thermoplastic with a high Tg of ∼190 °C, and is a potential substitute for perfluorosulfonic acid membranes used in polymer electrolyte fuel cells. The hollow core/pore shell structure of porous PSF microspheres was applied to remove oil from water [43]. Porous PSF coatings enhanced drug delivery [44]. A hydrophilic porous PSF membrane was prepared using amphiphilic cellulose [45]. P3HT has been widely applied in organic photovoltaic (OPV) cells. Huang et al fabricated inhomogeneous distributed porous P3HT at the micrometer scale using a freeze-dry method and applied it to OPVs [46]. Aryal et al prepared nano-porous P3HT by nanoimprint lithography [47]. Azmer et al applied an electro-spraying technique to obtain VOPcPhO:P3HT microstructures with nano-porous surface morphology for humidity sensors [48].
The water-driven BF method is much simpler than the above-mentioned templating and lithographic techniques. There have been no reports on the application of the BF method for the fabrication of porous PI, PSF and P3HT films. Here, we apply the fast-evaporation processes to prepare honeycomb porous films by drop-casting 1.0 wt% PSF, 1.0 wt% P3HT and 0.3 wt% PI chloroform solution. The respective RH, Tr and the Ts control processes are demonstrated in figure 7. The resultant average pore diameter and depth of the PSF film are 1.5 μm and 1.5 μm (figure 7(a)), respectively. The pore distribution on porous PI film is not uniform because of the relatively low solubility of PI in chloroform. However, the obtained 3D porous structure is ordered and almost hollow, as shown in the cross-section FESEM image (figure 7(b)), and the pore diameter and depth are 500 nm and 1.0 μm, respectively. The porous P3HT film was fabricated with different control processes for the Ts, and the resultant hollow porous film shows a network pattern with the pore diameter and depth of 1.0∼5.0 μm and 0.8∼1.0 μm, respectively, together with a high aspect ratio of pore diameter to interpore distance. These results demonstrate that the fast-evaporation processes could achieve an ordered porous structure for PI, PSF and P3HT. Diverse pore morphology and pore distribution of these polymer films is probably ascribed to varying solubility as well as viscosity in chloroform [49].
Figure 7. FESEM and AFM images at low and high magnification and cross-sections of porous PSF (a), PI (b) and P3HT (c) films fabricated by the fast-evaporation processes.
Download figure:
Standard image High-resolution image
4. Conclusions
In summary, we fabricated ordered honeycomb PS film in a sealed chamber under precisely controlled RH and Ts. Humidity and Ts both play an important role in the pore size. The modified BF process involves the re-design of both the Ts thermal processing and the evaporation time. The obtained pore diameter and depth at nanometer scale demonstrate the feasibility of the approach. The BP and ρsolvent significantly influence the pore distribution and pore depth, respectively. It is demonstrated that the fast-evaporation processes could achieve an ordered porous structure for PI, PSF and P3HT with diverse morphology and pore distribution. The modified BF process shows great potential in the fabrication of a variety of porous functional films.
Acknowledgments
This work was sponsored by the National Natural Science Foundation of China (61504077, 61674101), the Science and Technology Commission of Shanghai Municipality (17DZ2291500), the Program of Shanghai Academic/Technology Research Leader (18XD1424400), the 'Shuguang Program' supported by Shanghai Education Development Foundation and Shanghai Municipal Education Commission.