Article pubs.acs.org/Langmuir
Surface-Modification of Poly(dimethylsiloxane) Membrane with SelfAssembled Monolayers for Alcohol Permselective Pervaporation Jie Li,† Shulan Ji,*,† Guojun Zhang,*,† and Hongxia Guo‡ †
Center for Membrane Technology, College of Environmental and Energy Engineering, Beijing University of Technology, Beijing 100124, P. R. China ‡ College of Material Science and Engineering, Beijing University of Technology, Beijing 100124, P. R. China ABSTRACT: The use of self-assembled monolayers (SAMs) has recently been recognized as an effective way to tailor the surface properties of films used in various applications. However, application of SAMs in the preparation of separation membranes remains unexplored. In the present study, surface-modified poly(dimethylsiloxane) (PDMS) membranes were prepared using SAMs to fabricate a membrane for use in pervaporation separation of ethanol/water mixtures. A cross-linked PDMS/polysulfone (PSf) composite membrane was transformed by introducing hydroxyl functionalities on the PDMS surface through a UV/ozone conversion process. (Tridecafluoroctyl)triethoxysilane was allowed to be adsorbed on the resulting Si−OH substrate to increase the hydrophobicity of the membrane. Results from Fourier transform infrared spectroscopy, scanning electron microscopy, energy-dispersive X-ray spectrometry, atomic force microscopy, and contact angle analyses suggest that the fluoroalkylsilane monolayer was successfully formed on the modified PDMS/PSf membrane treated by 60 min UV/ozone exposure. The newly SAM-modified membrane exhibited a separation factor of 13.1 and a permeate flux of 412.9 g/(m2 h), which are higher than those obtained from PDMS membranes.
■
INTRODUCTION The global energy crisis has prompted the increased production of fuel ethanol.1,2 Ethanol is a type of biofuel, which is widely recognized as a green alternative to fossil fuels. At present, about 90% of fuel ethanol is produced by biological fermentation. However, this process is impeded by the relatively high ethanol concentration produced during the fermentation process. The ethanol content could reach only 1%−15% in the fermentation broth.3 Therefore, in situ removal of ethanol from the fermentation broth may afford a more efficient way of producing ethanol. In recent decades, pervaporation (PV) has been widely studied, as it is a process with good potential for recovering bioethanol from biomass fermentation.4 Membranes for PV are typically fabricated from hydrophobic material, the most common of which is poly(dimethylsiloxane) (PDMS). However, the separation factor and flux of pure PDMS membranes used for ethanol/water mixtures are still unsatisfactory for practical applications.5 In particular, the selectivity of PDMS membranes still needs further improvement. PV is governed by a solution−diffusion mechanism, a process in which the affinity adsorption between membrane surface and feed solution plays a very important role. Moreover, the water repellency is governed by the surface chemical properties of the topmost layer. The contact angle (CA) on PDMS membrane surfaces is about 90°−120°.6 Therefore, the CA on such surfaces may be further improved by increasing their hydrophobicity. In order to improve the PV performance, attempts have been made to modify PDMS in different ways. These methods include graft polymerization, chemical modification of poly© XXXX American Chemical Society
mers, zeolite doping of polymers, and physical (or chemical) modification of formed membranes. For example, Chang et al. prepared composite membranes of silicone/PVDF by curing a copolymer of polysiloxane and phosphate ester. The porous PVDF substrate was previously plasma-grafted with a thin layer of silicone-compatible material. This multilayer membrane afforded high PV performance (separation factor of 31) and a permeation rate of 0.9 kg/(m2 h) at a 10 wt % ethanol feed concentration at room temperature.7 Field et al. modified crosslinked PDMS membranes by sequential introduction of two different side arm functional groups, −(CH2)3OC2H5 and −(CH2)3NMe2. Membranes containing 20% ethyl ether (in AEE membranes) and 10% dimethylamino (in AMI membranes) groups produced a high cresol flux (30% higher than that achievable using either AEE or AMI modified membranes) and a high separation factor (α = 50.4).8 Kashiwagi et al. prepared highly permselective membranes by surface treatment of thin poly(dimethylsiloxane) membranes with silane compounds containing octadecyl groups. An ethanol permselectivity of 18.0 and a flux of 0.015 kg/(cm2 h) were obtained with these membranes.9 Clearly, modification of the PDMS membrane is significant for the improvement of PV performance. Considering that the surface energy of CF3 (6 mJ/m2) is lower than that of other functional groups, the behavior of fluorosilane molecules may tailor the wetting properties of these membranes. Received: March 16, 2013 Revised: April 23, 2013
A
dx.doi.org/10.1021/la400930y | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
Figure 1. Schematic diagram describing the preparation of the SAM-modified PDMS/PSf membrane.
The UVO conversion method was employed to create Si−OH layers on the PDMS/PSf composite membranes. These Si−OH layers react with (tridecafluoroctyl)triethoxysilane molecules. In comparison with other surface modification methods, such as chemical vapor deposition and plasma systems, UVO utilizes relatively inexpensive equipment.17−19 Successful fabrication of the SAM-modified membranes was confirmed by attenuated total reflection Fourier transform infrared (FTIR) spectroscopy, energy-dispersive X-ray (EDX) spectrometry, and CA measurements. Scanning electron microscopy (SEM) and atomic force microscopy (AFM) were also conducted to further understand the microtopographical changes on the membrane surfaces. Separation measurements were carried out to determine the PV performance.
Over the past decades, the deposition of low-surface-energy semifluorinated (SF) molecules has been applied to increase the hydrophobicity of polymer substrates.10−15 For example, Sermon et al. prepared SiO2(1−x) TiO2x coatings doped with perfluoroalkylsilane (PFAS).10 The PFAS−sol−gel interaction caused coating restructuring and, therefore, changed the wettability of the SiO2(1−x)TiO2x coating at a modest temperature. Cao et al. fabricated superhydrophobic surfaces on common industrial steels (Cu alloys and Ti alloys) by utilizing wet chemical etchants and surface coating with fluoroalkylsilane.11 The surfaces showed stable superhydrophobicity in many corrosive solutions. The fabrication procedure is timesaving, inexpensive, and fairly easy to carry out. Genzer et al. tailored elastomeric surfaces using mechanically assembled nonolayers, structures that are fabricated by combining selfassembly of surface-grafting SF molecules with mechanical manipulation of the grafting points in the underlying elastic surface. This method produced monolayers with superior nonwetting and barrier properties.13 Among the various available surface coatings and modification techniques, selfassembled monolayers (SAMs) formed from organosilanes are promising candidates as hydrophobic coatings because of their high bonding strength, low surface energy, and high thermal stability.16 The wettability and free energy of SAMs can be easily controlled by altering the terminal group from completely hydrophilic (e.g., −OH or −COOH) to very hydrophobic (e.g., −CH3 or −CF3). For example, the attachment points for organosilanes can simply be generated after UV/ozone (UVO) photo-oxidation; thus, patterned surfaces can be prepared by deposition of SF molecules. SAMs on the solid−liquid interface have been studied extensively and widely applied to the modification of solid surfaces. However, there are few reports on the preparation of SF monolayers on porous substrates and the application of SF monolayers to separation of mixtures. The purpose of this study was to prepare a more hydrophobic, alcohol-permselective PV membrane via the application of SAMs. A schematic diagram describing the preparation of the SAM-modified poly(dimethylsiloxane)/ polysulfone (PDMS/PSf) membrane is shown in Figure 1.
■
EXPERIMENTAL SECTION
Materials. PDMS with a viscosity of 2550 Pa·s was purchased from China Bluestar Chengrand Chemical Co. Ltd. (China). Tetraethyl silicate (TEOS) and ethanol were obtained from Beijing Chemical Company (China). Dibutyltin dilaurate, n-heptane, tert-butanol, and isopropanol were supplied by Tianjin Fuchen Chemical Reagents Company (China). (Tridecafluoroctyl)triethoxysilane was obtained from Degussa (Germany). In our experiments, all reagents were of analytical grade and were used without further purification. Preparation of the PDMS/PSf Composite Membrane. The PSf substrate was soaked for 24 h using pure water. The pores of the supports were filled with water to prevent the penetration of the coating polymer solution into the pores; subsequently, the pores were allowed to dry for 20 min to remove excess water on the support surface before dip-coating the polymer solution.20 PDMS was dissolved in n-heptane to form a 10 wt % PDMS solution. After the polymer solutions were stirred at room temperature for 1 h, the crosslinking agent TEOS and the catalyst dibutyltin dilaurate were added to the polymer solution (WPDMS/Wn‑heptane/WTEOS/Wdibutyltin dilaurate = 1:10:0.1:0.005),20 and the resulting mixture was continuously stirred for 3 h. Air bubbles trapped in the polymer solution were removed by degassing at 100 Pa for 10 min. The polymer solution was dip-coated on the surface of the PSf support for 1 min. Following removal of the solvent at room temperature for 12 h and subsequent curing in the oven at 80 °C for 12 h, the PDMS/PSf composite membrane was finally fabricated. B
dx.doi.org/10.1021/la400930y | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
Figure 2. FTIR spectra of PDMS, PDMS−UVO, and PDMS−UVO−F membranes. (a) Comparisons of PDMS(i), PDMS−UVO(ii), and PDMS− UVO−F(iii) membranes. (b) Changes in the CH3 (1446, 1412, 1258, and 800 cm−1) and Si−O−Si (930 and 1200 cm−1) on the surface of PDMS membranes. (c) Changes in the OH (around 3300 cm−1) and CH3 (2906 cm−1) on the surface of PDMS. SAM-Modified PDMS/PSf Membrane. The SAM-modified membranes were fabricated by depositing CF3(CF2)5(CH2)2Si(OCH2CH3)3 onto PDMS−UVO substrates prepared by exposing PDMS network films to UVO for a period of time. The UVO treatment of the PDMS/PSf surface was carried out at a distance of 20 cm from the UV source for 10−120 min at ambient conditions in a commercial UV chamber (Beijing Institute of Opto-Electronic Technology, China). The UV intensity on the sample surface was 4 mW/cm2 at λ = 254 nm, based on measurement by a UV power meter (Photoelectric Instrument Factory of Beijing Normal University). Ozone was generated in situ from atmospheric oxygen during UV exposure. The CF3(CF2)5CH2CH2Si(OCH2CH3)3 in ethanol solution (1.0 wt %) was hydrolyzed by the addition of a 3-fold molar excess of water at room temperature. After UVO treatment, the oxidized PDMS membranes (hereafter known as PDMSOH) were immediately treated with the mixture of hydrolyzed silane solution and were allowed to react for 1 h at room temperature. In this step, the adsorbed SF molecules react with the silanol groups present on PDMSOH to yield terminal −CF3 groups (the resulting membrane is referred to as PDMSCF3). After removal from the solution, the membranes were washed with ethanol and then vacuum-dried at 40 °C for 1 h. The temperature was found to be an important factor in obtaining highquality SAMs. The drying temperature has been set to 40 °C because higher temperatures significantly increased the degradation rate of the membranes.21 Membrane Characterization. The FTIR spectra of the PDMS/ PSf membrane, UVO-exposed PDMS/PSf membrane, and SAMmodified PDMS/PSf membrane were obtained using a Vertex-70 spectrophotometer (Bruker, Germany). Contact angles of water and
ethanol on the membranes were measured by using a contact angle analyzer (DSA 100, Germany). Prior to measurements, water and ethanol were applied as droplets onto the membrane surfaces. The average value of 10 measurements at different positions of the sample was adopted as the CA value. The membrane surfaces were observed under a scanning electron microscope (Hitachi-4300, Japan), and AFM images were acquired on a Pico ScanTM 2500 (USA) to characterize their morphology. The chemical compositions of the membranes were analyzed using an EDX spectrometer. The surface compositions of the membranes were measured by X-ray photoelectron spectroscopy (XPS), using a Thermo Fisher Scientific ESCALAB 250 K-Alpha X-ray photoelectron spectrometer equipped with a monochromatic Al Kα X-ray source (1486.6 eV). To determine the thickness of the SAMs, the films were monitored with a variableangle spectroscopic ellipsometer (Shanghai Bright Enterprise Development Co., Ltd., China). Ellipsometric data were collected over a wavelength range of 300−1100 nm in 10 nm increments and incidence angles of 65°, 70°, and 75°. As a controlled experiment, single-sidepolished silicon wafers were cut into small pieces (≈2 × 2 cm) and then subjected to UVO treatment for 60 min. This treatment produces a high concentration of the surface −OH groups at the silica surfaces; these groups serve as attachment points for the fluoroalkylsilane molecules.22 Afterward, the silicon wafers were immersed in the hydrolyzed (tridecafluoroctyl)triethoxysilane solution at room temperature for 1 h. The apparent total film thickness was determined from the collected data by fitting a Cauchy model for single wavelengthdependent refractive index. Similarly, the film thickness of SAMmodified PDMS was also measured by fitting a two-layer Cauchy model, which was based on the refraction index and thickness values of C
dx.doi.org/10.1021/la400930y | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
silicon-based SAMs obtained from the above controlled experiments. All measurements were carried out at room temperature. Pervaporation Experiments. The PV performance of the membrane was evaluated using a PV apparatus fabricated in our laboratory.23−27 The selected ethanol concentration in the feed solution was 5 wt %. The experiments were conducted at a feed temperature of 60 °C. The downstream pressure was maintained at around 100 Pa. The PV performance was measured three times by using three membranes under the same fabrication and modification conditions; the average of the three trials was used as a data point. The permeate vapor was trapped in liquid nitrogen and subsequently analyzed by gas chromatography (GC-14C, Shimadzu, Japan). The permeate flux (J) was determined according to the following equation:
W J= At
Table 2. Changes of Atomic Concentration Determined by XPS of SAM Membranes with Different UVO Times C (%) O (%) Si (%) F (%) O/Si O/C
(1)
YEtOH/YH2O XEtOH /X H2O
(2)
where YEtOH, YH2O, XEtOH and XH2O are the weight ratios of ethanol and water in the permeation and feed sides, respectively.
■
RESULTS AND DISCUSSION Surface Characterization of SAM Membranes. The method for fabrication of SAMs is based on combining (i) the Table 1. Changes of Atomic Concentration Determined by EDX of SAM Membranes with Different UVO Times C (%) O (%) Si (%) F (%) O/Si O/C
0 min
10 min
30 min
60 min
90 min
120 min
39.77 25.36 34.86
32.21 22.45 43.35 01.99 0.52 0.70
28.48 29.96 38.98 02.59 0.77 1.05
27.00 33.61 37.73 01.66 0.89 1.24
23.97 29.47 45.29 01.27 0.65 1.23
30.22 29.30 39.01 01.48 0.75 0.97
0.72 0.64
10 min
30 min
60 min
90 min
120 min
59.98 22.03 16.18
53.38 22.24 17.55 5.67 1.27 0.42
49.24 23.10 18.00 7.22 1.28 0.47
47.53 24.19 18.78 7.84 1.29 0.51
45.86 21.69 27.17 5.28 0.80 0.47
43.4 20.04 24.41 12.15 0.82 0.46
1.36 0.37
of hydroxyl groups. Investigators have also reported ellipsometric measurements on UVO-modified PDMS and found that, under irradiation, a uniform silica-like layer with a thickness of around 20−30 nm was created.29,30 The reactions involved in the modification of PDMS by UV radiation (at 185 and 254 nm in air) and the fabrication of SAMs are presented as Figure 1. The chemical changes in the PDMS surface after UVO exposure and self-assembly were studied by FTIR spectroscopy (Figure 2). The reference spectrum of the cross-linked PDMS (Figure 2a) is in accordance with previous published data.31 Bands originating from the deformation vibrations (δ(CH3)) are located at 1446, 1412, and 1258 cm−1. The asymmetric strethching vibrations of the Si−O−Si group (υas(SiOSi)) appear between 930 and 1200 cm−1. Bands due to asymmetric stretching (υas(CH3)) and rocking (ρ(CH3)) of the methyl groups are centered at 2906 cm−1 and at around 800 cm−1. An overall decrease in intensity of the PDMS absorption bands was observed, indicating a steady degradation of the siloxane polymer. In addition, absorption bands around 3300 cm−1 initially increased in intensity. Surface composition analysis of PDMS membranes and SAMs after various periods of UVO exposure was also carried out by EDX and XPS. The probing depths of EDX and XPS were shorter than 2 μm and 10 nm, respectively. These two techniques could probe the composition difference between the upper and deeper layers of the surface. The atomic composition obtained by EDX is shown in Table 1. The silicon/carbon (Si/ C) and oxygen/carbon (O/Si) ratios increased with exposure times but decreased after 60 min of UVO exposure. The change in the Si/C and O/C ratios may be attributed to a reduction in the amount of carbon and the increase or decrease in the amount of oxygen within the converted surface film. This behavior indicates that a photo-oxidative reaction involving atomic oxygen generated in situ removed the methyl groups of the polysiloxane precursors. The organic portions were transformed into low-molecular-weight species, including carbon dioxide, water, and small amounts of volatile carbonaceous compounds (Scheme 1).29 This process left a surface film
where W is the weight of the liquid collected in the cold traps, A is the effective area of the membrane, and t is the certain time for the PV. The selectivity of the composite membranes is expressed as a separation factor α, which is defined as αHEtOH = 2O
0 min
grafting reaction between SF trichlorosilanes and hydroxyl functionalities present on the silica surfaces and (ii) the creation of HO−PDMS groups. To accomplish the latter, the surface −OH functionalities are created by UVO exposure. Ozone is generated in situ from atmospheric oxygen by exposure to 184.7 nm UV light. Ozone absorbs 254.7 nm UV light and subsequently photodissociates into molecular oxygen and atomic oxygen.28 Atomic oxygen and ozone may react strongly with carbon atoms and, thereby, lead to the formation Scheme 1. Mechanism of the Oxidation of PDMS
D
dx.doi.org/10.1021/la400930y | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
Figure 3. SEM images of the flat sheet SAM-modified membranes: (a) top surface of untreated PDMS/PSf membrane; (b) top surface after UVO treatment with 10 min and SF self-assembly; (c) top surface after UVO treatment for 30 min and SF self-assembly; (d) top surface after UVO treatment for 60 min and SF self-assembly; (e) top surface after UVO treatment for 90 min and SF self-assembly; (f) top surface after UVO treatment for 120 min and SF self-assembly.
consisting primarily of SiOx and OH. The XPS results in Table 2 illustrate a change in trend of atomic compositions that is consistent with the results in Table 1. Additionally, the change in the concentration of the F-containing moieties in Table 2 was drastic. However, the XPS measurements were only of the order of 10 nm; thus, the SAM formed was
