- Email: [email protected]
Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation
ARTICLE IN PRESS
JID: JTICE
[m5G;December 5, 2015;13:25]
Journal of the Taiwan Institute of Chemical Engineers 000 (2015) 1–7
Contents lists available at ScienceDirect
Journal of the Taiwan Institute of Chemical Engineers journal homepage: www.elsevier.com/locate/jtice
Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation Xianhua Li, Kunyue Teng, Jie Shi, Wei Wang, Zhiwei Xu∗, Hui Deng, Hanming Lv, Fengyan Li State Key Laboratory of Separation Membranes and Membrane Processes, School of Textiles, Tianjin Polytechnic University, Tianjin 300387, PR China
a r t i c l e
i n f o
Article history: Received 25 June 2015 Revised 3 November 2015 Accepted 16 November 2015 Available online xxx Keywords: Electrospinning Nanoporous fibers Phase separation Water bath Humidity
a b s t r a c t A novel polylactic acid (PLLA) fiber membrane with nanoporous structures has been successfully developed via thermal-nonsolvent induced a phase separation technique by electrospinning a single solvent system of PLLA/dichloromethane. A water bath receiver was necessary for preparing the nanoporous fiber membranes in this paper. The results showed that the nanoporous fiber membranes could be obtained within a wide range of humidity. Compared with non-porous PLLA fiber membranes, the nanoporous fiber membranes displayed significantly improved specific surface area (improved about 20%) and rejection ratio (about 3 times) toward methylene blue. Besides, the PLLA nanoporous fiber membranes showed a water flux up to 4836.6 L m−2 h−1 , increased by 25% than that of the non-porous fiber membranes. In addition, the oil adsorption capacity of the PLLA nanoporous fiber membranes could reach 26.8 g g−1 . The new developed PLLA nanoporous fiber membranes via thermal-nonsolvent induced phase separation technique would demonstrate an impressive prospect for oil adsorption and organic dye filtration. © 2015 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
1. Introduction Electrospinning is a straightforward way to prepare micro-/ nanometer scale and high porosity fibers [1,2]. Since the electrospun fibers have many advantages, such as good repeatability, various morphological modifications, and three-dimensional porous structures [3,4], electrospinning progress has attracted a great deal of attention. Therefore, these nanofibers/nanowebs are quite applicable in many areas including filtration [5], biotechnology [6], etc. Some interesting structures, such as core/shell structure [7], porous structure [8] and hollow structure [9], have emerged in recent years, and attracted widely attention due to the unique properties and functionalities. A lot of applications may be favored if the fibers are endowed with rough or porous structures [10]. In fact, widespread attention has been paid to the electrospinning technique in increasing the specific surface area and porosity of the fiber membranes [11,12]. Up to now, there were four methods to prepare the porous fiber membranes in literature: (1) using high/low boiling point solvent mixtures [13], (2) using solvent/nonsovent mixtures [14,15], (3) post-processing electrospun composite nanofibers by selectively removing one of the components [16,17], (4) electrospinning of polymers in a high humid environment [18]. Currently, porous fiber membranes have been used for various fields including tissue
∗
Corresponding author. Tel./fax: + 86 22 83955231. E-mail address: [email protected] (Z. Xu).
engineering [19], carriers in drug delivery system [20], and electrode application [21]. In recent years, a lot of methods have been published to prepare electrospun porous fibers. Cao and his coworkers [22] have prepared polylactic acid ultrafine fibers with mesopores of 30–150 nm by electrospinning. The fiber diameters and morphologies could be controlled by adjusting the composition ratio of the dichloromethane/dimethyformamide solvent mixtures. A facile method to fabricate porous fiber membranes by electrospinning a ternary system of nonsolvent/solvent/poly(L-lactic acid) was presented by Qi and his coworkers [23]. Nanoporous polyacrylonitrile ultrafine fibers were prepared by Li and his workers [17]. Post-processing electrospun polyacrylonitrile composite nanofibers by selectively removing polyvinylpyrrolidone was the main method to prepare porous fibers. The specific surface area of the porous polyacrylonitrile ultrafine fiber membranes was more than 70 m2 g−1 . The porosity of electrospinning porous fibers obtained from polystyrene/dimethyformamide solution at 60% relative humidity was characterized by H.Fashandi et al. [24]. It was obvious that the high humid environment was the main factor to prepare polystyrene porous fibers. Kim et al. [25] fabricated the electrospun polymer nonwoven mats with porous surface morphologies by varying the collector temperature. Though these fore-mentioned methods have been used widely, there are a few factors to be considered in the process of preparation, such as dispersion of binary polymers in the same solvent, compatibility between binary solvents, and the control of the relative humidity.
http://dx.doi.org/10.1016/j.jtice.2015.11.012 1876-1070/© 2015 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
Please cite this article as: X. Li et al., Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation, Journal of the Taiwan Institute of Chemical Engineers (2015), http://dx.doi.org/10.1016/j.jtice.2015.11.012
ARTICLE IN PRESS
JID: JTICE 2
[m5G;December 5, 2015;13:25]
X. Li et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2015) 1–7
Fig. 1. The process of electrospinning and the principle of nanopore formation. Table 1 The contrast of preparation methods. Solvent
Humidity or temperature
Receiver
Mainly application
Ref.
Double solvent Double solvent Double solvent Single solvent Single solvent
Not mentioned 49%, room temperature 40%, room temperature 25%, 40ºC 20–70%, room temperature
Roller Roller Roller Roller Water bath
Drug carrier Tissue engineering Tissue engineering Tissue engineering Filtration and oil adsorption
[26] [27] [23] [25] This work
In this study, without regard to complicate factors in the process of preparation, polylactic acid fibers with nanoporous structures were obtained by electrospinning a single solvent system of PLLA/dichloromethane. Based on the mechanism of thermalnonsolvent induced phase separation, a water bath was used to obtain the nanoporous fiber membranes within a wide range of humidity. The resulting membranes were subject to characterizations including scanning electron microscope (SEM), nanopore sizes and interconnected pore size distributions, specific surface area measurements, contact angle tests, oil adsorption tests, permeation flux and rejection measurements. The influence of humidity during electrospinning on the morphologies and properties of membranes was investigated. 2. Experiment 2.1. Materials Polylactic acid (PLLA, Mw = 110,000) was purchased from Ningbo global biological material Co.Ltd. Dichloromethane (CH2 Cl2 , AR, 99.5%) was supplied by Tianjin Fuyu Chemical Co. Ltd. All the chemicals were analytical grade. 2.2. Preparation of electrospun PLLA nanoporous fiber membranes The PLLA granules were dissolved in dichloromethane solvent and the concentration was 9 wt%, and then the PLLA solution was mixed uniformly by magnetic stirring for 6 h. The setup of electrospinning process was composed of a high voltage power supply, an injector, a water bath (receiver) and a humidifier. The obtained PLLA spinning solution was placed in the injector with a capillary tip (inner diameter=0.67 mm). The anode of the high voltage power supply was clamped to an injector needle tip, and the cathode was connected to a water bath. The electrospun fibers were collected in the water bath. The applied voltage was 20 kV, the tipto-collector distance was 15 cm, and the flow rate of the spinning solution was 2 ml h−1 . All of the electrospinning operations were
performed under room temperature. After electrospinning, the resulting membranes were subject to characterizations without any treatments except for drying for 24 h. Fig. 1 showed the process of electrospinning and the principle of nanopore formation. It was obvious that the change of ambient humidity was due to the cooperation between humidifier and water bath. And the variation of humidity was read by hygrometer. In this paper, the influence of humidity during electrospinning on the morphologies and properties of membranes was investigated. Hence, according to the reading of hygrometer, the change of ambient humidity was achieved by adjusting the humidifier. In order to identify all various kinds of membranes easily, the membranes which were prepared at the humidity of 30%, 40%, 50% and 60% were named as membrane-30, membrane-40, membrane-50 and membrane-60 correspondingly. In addition, the non-porous fiber membranes were obtained at the same electrospinning conditions except for a roller receiver. Compared with other preparation methods, the advantages of our work were shown in Table 1. 2.3. Morphology analysis The surface morphologies of electrospun PLLA nanoporous fiber membranes were investigated using SEM (Quanta 200,Holland) instruments. Prior to the observation, all the samples were coated with gold. After SEM tests, the nanopore sizes on the single fiber were measured by Image-Pro Plus software on the images of SEM. And each final nanopore diameter was obtained by averaging over more than fifty nanopores of different fibers. The interconnected pore size distributions of the membranes were obtained by capillary flow porometry measurements [28], based on the wet/dry flow method. The membranes were immersed in wetting liquid (Porefil, provided by PMI) with a surface tension of 16 dyn/cm for 24 h at room temperature. A wet gas was performed using a solvent filled sample that was followed by a dry gas (sample without wetting liquid) for obtaining flow rate curves as a function of applied pressures. The differential pressure of the gas at which flow through a pore occurs yields the throat (most constricted) diameter
Please cite this article as: X. Li et al., Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation, Journal of the Taiwan Institute of Chemical Engineers (2015), http://dx.doi.org/10.1016/j.jtice.2015.11.012
ARTICLE IN PRESS
JID: JTICE
[m5G;December 5, 2015;13:25]
X. Li et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2015) 1–7
along the pore. Because of the low pressure applied during the process, the porous structures of fibrous membranes are not distorted. All porometry tests were performed at room temperature. The overall porosity (ε ) was determined by the gravimetric method [29], as defined in the following Eq. (1):
ε=
ω1 −ω2 × 100% A × l × dw
(1)
where ω1 was the weight of the wet membrane (g), ω2 was the weight of the dry membrane (g), A was the membrane effective area (m2 ), dw was the water density (0.998 g cm−3 ) and l was the membrane thickness (nm). The specific surface areas of samples were analyzed by physical adsorption of gases (N2 at 77 K) in volumetric adsorption systems (NOVA 2000e,Quantachrome Ins). All samples were degassed at 60ºC (PLLA fibers hardly endure high temperature) for 4 h prior to the adsorption measurement. In addition, the five point method was used to calculate the value of the specific surface areas. And each final specific surface area was obtained by averaging over more than five specific surface areas. 2.4. Contact angle tests The water and oil contact angles of as-prepared membranes were calculated from a digital video image of the drop on the membranes using an image processing program (JYSP-180 Contact Angle Analizer). Pure water and soybean oil were chosen for the test. In general, each final contact angle was obtained by averaging over more than five contact angle values of different spots. 2.5. The permeation flux and rejection measurements Permeation flux and rejection were evaluated for investigating the filtration performance of as-prepared membranes [30]. A dead-end filtration experimental setup was used to measure the pure water flux with the effective membrane area of 19.3 cm2 . Measuring protocol was depicted as follows: for the first 10 min, the membranes were compacted at 0.1 MPa to get a stead flux; then the flux was recorded at 0.1 MPa every 5 min, and at least 5 readings were collected to obtained an average value. The flux was obtained by the following Eq. (2):
Q = V/(A × T )
(2)
where Q was permeation flux of membranes for pure water (L m−2 h−1 ), V was the volume of permeated pure water (L), T was the permeation time and A was the effective area of the membranes (m2 ). Organic dyes were often used for the target contaminant to measure the rejection ratio of membranes. The rejection tests were carried out with an aqueous solution of methylene blue [31] (molecular weight = 373.9). After the measurement of pure water flux, pure water was changed to methylene blue solution. The final rejection ratio was obtained by the different concentration of permeated and feed solutions [32]. The rejection ratio (R,%) was computed by the following Eq. (3):
R=
Cp 1− Cf
× 100%
(3)
where Cp (mg L−1 ) and Cf (mg L−1 ) were the concentrations of permeation and feed solutions, respectively. The dye concentration of initial feed solution was 6 mg L−1 , and the dye concentrations of permeation solutions were detected by a UV-spectrophotometer (Shimadzu UV2450, Japan). 2.6. Oil adsorption tests Soybean oil was chosen for the test. 0.05 g membrane was placed on the soybean oil. With high buoyancy, the sample floated on the
3
surface and kept sucking the soybean oil initially. After a period of time, the wet membrane was taken out and drained for 5 min until no residual oil droplet was left on the surface, and then weighed [33]. Oil adsorption capacity of membranes was determined by the following Eq. (4):
q = (mw − m0 )/m0
(4)
where q was the adsorption capacity (g g−1 ), m0
was the initial weight of the membranes (g), and mw was the weight of the wet membranes after 5 min drainage (g). 3. Results and discussion 3.1. Morphologies of electrospun PLLA nanoporous fiber membranes The differences of membranes were compared in Fig. 2, as observed by SEM. All the membranes presented typical nanoporous structures regardless of humidity. However, the morphologies of nanoporous fibers were affected by the relative humidity of environment. For membrane-40, the nanopores compared with the other membranes could be viewed visually from the SEM pictures. Based on Image-Pro Plus analysis in Table 2, the changes of the nanopore diameters at various relative humidity were in accord with the SEM pictures of fibers. It was indicated that low or high humidity was not favorable for nanopore formation. Low humidity could not play the role of water vapor sufficiently. When the humidity was increasing constantly, dichloromethane would volatilize rapidly. High humidity might affect technology of spinning, and then intervene nanopore formation. To investigate the mechanism of nanopore formation [34], a typical schematic diagram of polymer-solvent-nonsolvent system was shown in Fig. 1. In the process of electrospinning, dichloromethane could evaporate rapidly because of its low boiling point. Due to rapid solvent evaporation, the fibrous surface temperature decreased significantly. The phenomenon of rapid solvent evaporation might attribute to the thermally induced phase separation [33]. Thermally induced phase separation would cause polymer-rich region to form fiber frame and solvent-rich region to form nanopores. In fact, thermally induced phase separation was just part of reasons for nanopore formation. Another important reason was nonsolvent induced phase separation. When the temperature of fiber surface cooled down, moisture in the air might be condensed and grew in the form of droplets. And then many water droplets would be left on the surface of fibers. Moreover, water was the nonsolvent for PLLA. When the fibers which loaded a large number of water vapor solidified, the nanopores would be formed due to the evaporation of water vapor in the PLLA fibers [35]. Whereas, the regulation of water vapor in the air by adjusting the humidifier hardly meet requirements. To solve this problem, the fibers were received by a water bath. The water bath not only supplied quantities of water vapor, but also dropped the environment temperature in a degree. Therefore, the water bath was crucial for the nanopore formation. In practice, the process of nanopore formation can be interpreted as the exchange of dichloromethane and water into fibers (Fig. 1). The overall porosity information of the membranes was presented in Table 2. The results of the porosity measurements revealed that all the nanoporous fiber membranes possessed a good porosity in the range of 60-70%, while the non-porous fiber membranes had a porosity of 48.5%. It was obvious that there were two kinds of pores in the nanoporous fiber membranes, i.e. the nanopores on the single fiber and the interconnected pores that the fibers made from. Hence, compared with non-porous fiber membranes, the porosity of the nanoporous fiber membranes should be improved due to the nanoporous structures on the single fiber. From Fig. 3, it was clear that the interconnected pore diameter of nanoporous and non-porous membranes was 2-4 μm and 3-5 μm, respectively, indicating the
Please cite this article as: X. Li et al., Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation, Journal of the Taiwan Institute of Chemical Engineers (2015), http://dx.doi.org/10.1016/j.jtice.2015.11.012
ARTICLE IN PRESS
JID: JTICE 4
[m5G;December 5, 2015;13:25]
X. Li et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2015) 1–7
Fig. 2. SEM images of electrospun PLLA fibers: (a) membrane-30, (b) membrane-40, (c) membrane-50, (d) membrane-60. The insets showed the high magnification images, correspondingly. Table 2 The nanopore diameter, porosity and specific surface area of different membranes. Sample
Nanopore diameter (nm)
Porosity (%)
Specific surface area (m2 g−1 )
Membrane-30 Membrane-40 Membrane-50 Membrane-60 Non-porous membranes
187.6 ± 1.1 265.7 ± 0.8 176.3 ± 0.3 199.7 ± 1.4 –
62.7 ± 0.9 69.1 ± 0.4 61.3 ± 1.1 65. 7 ± 0.3 48.5 ± 0.7
182.2 ± 5.3 200.3 ± 3.6 191.5 ± 1.8 173.2 ± 2.1 152.5 ± 3.1
negligible difference in the interconnected pore diameter distributions of all membranes. Therefore, it could be considered that the interconnected pores in different membranes had the same influence on the performance of membranes to some extent. This phenomenon could not be shown in Fig. 2 because the images of SEM revealed only the random and local morphologies of membranes. According to the above analysis, it was shown that the nanoporous fiber membranes were all endowed with advantageous porous structures, which undoubtedly played a positive role in promoting membrane permeability. The specific surface areas of the as-prepared PLLA nanoporous fiber membranes were analyzed by isothermal N2 adsorption measurements [36]. The results showed that all the membranes with
nanoporous structures had higher specific surface areas than nonporous fiber membranes though the humidity changed. Fig. 4 showed the isotherm of the relative pressure (P/P0 ) ranging from 0.05 to 0.35 via five point method. The specific surface areas of the electrospun PLLA nanoporous fiber membranes at different relative humidity were listed in Table 2. Based on the values of specific surface areas, the changes of the specific surface areas at various relative humidity were in accord with the morphologies of fibers. Particularly, when the relative humidity was 40%, the specific surface area reached 200.3 m2 g−1 . Compared with the specific surface area of non-porous fiber membranes, it increased about 20%. The high specific surface areas could be attributed to the construction of nanopores. It was worth mentioning that the high specific surface area could facilitate
Please cite this article as: X. Li et al., Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation, Journal of the Taiwan Institute of Chemical Engineers (2015), http://dx.doi.org/10.1016/j.jtice.2015.11.012
JID: JTICE
ARTICLE IN PRESS
[m5G;December 5, 2015;13:25]
X. Li et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2015) 1–7
5
Fig. 3. The interconnected pore diameter distributions of different membranes. Fig. 5. Contact angles of as-prepared membranes.
Fig. 4. The isotherm with five point method BET of different membranes.
efficient contact of fibers with organic pollutants, which was conductive to filtration.
Fig. 6. Comparison of pure water flux and methylene blue rejection ratio of the different membranes.
contact angles of membranes [38]. The changes of water contact angles might play an unfavourable role in improving the water flux of membranes [39].
3.2. Hydrophobicities of electrospun PLLA nanoporous fiber membranes Contact angle is an important parameter for measuring surface hydrophilicity or hydrophobicity [37]. In general, larger water contact angle refers to higher hydrophobicity. Compared in Fig. 5 were the contact angle results of various membranes. It could be found that all nanoporous fiber membranes were more hydrophobic and oleophilic than non-porous fiber membranes. The water contact angles of membrane-40 could attain 115°, and the water contact angles of membrane-30, membrane-50 and membrane-60 were all below 105°. In comparison, the soybean oil droplet immediately spread on the membranes. The oil contact angles of membrane-30, membrane50 and membrane-60 were all above 15°. Particularly, the oil contact angle of membrane-40 could reach 12°. As the humidity increased, the water contact angle values of membranes increased and then decreased, and the maximum value was observed for membrane-40. The change trend of oil contact angles were contrary to that of water contact angles. These results indicated that the nanoporous structures of the membranes led to high water contact angle and low oil contact angle, especially compared with the non-porous membranes. It was reported that the surface roughness could affect the
3.3. Permeation flux and rejection of electrospun PLLA nanoporous fiber membranes The data of the pure water flux of different membranes were depicted in Fig. 6. It suggested an obvious trend that all the membranes with nanoporous structures revealed higher water permeation fluxes compared with non-porous fiber membranes. When the relative humidity ranged from 30% to 40%, the pure water flux went up. However, when the relative humidity ranged from 40% to 60%, the pure water flux went down. Expressly, the water flux of membrane-40 reached a peak value of 4836.6 L m−2 h−1 , and increased approximately four times compared with that of flat ultrafiltration membrane [40]. Considering the results in Fig. 6, the improvement of permeability might be due to the morphologies of membranes. The nanopores on the single fiber undoubtedly played a positive role in promoting membrane permeability. Consequentially, though the membranes were hydrophobic, nanoporous fiber membranes still showed good water permeability [41]. The results for the methylene blue rejection ratio of different membranes were also depicted in Fig. 6. As could be seen, all the
Please cite this article as: X. Li et al., Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation, Journal of the Taiwan Institute of Chemical Engineers (2015), http://dx.doi.org/10.1016/j.jtice.2015.11.012
ARTICLE IN PRESS
JID: JTICE 6
[m5G;December 5, 2015;13:25]
X. Li et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2015) 1–7 Table 3 The oil adsorption efficiency of different membranes. Sample Efficiency(g g
−1
)
Membrane-30
Membrane-40
Membrane-50
Membrane-60
Non-porous membranes
16.6 ± 0.3
26.8 ± 0.1
19.7 ± 0.2
20.8 ± 0.4
12.1 ± 0.2
nanoporous fiber membranes presented higher methylene blue rejection ratios than non-porous fiber membranes. Particularly, the methylene blue rejection ratio increased significantly when the relative humidity was 40%. However, when the relative humidity continued increasing, the rejection ratio declined. Additionally, compared with those of non-porous fiber membranes (37.0%), the methylene blue rejection ratios of membrane-30, membrane-40, membrane-50 and membrane-60 were improved by 27.9%, 40.7%, 28.9% and 29.9%, respectively. It might be due to the nanopores on the single fiber [42]. The uniform nanopores of membranes could enhance methylene blue rejection ratios. Moreover, the decline in hydrophobic interaction between methylene blue dye and hydrophobic surface of membranes might be a reason for slight decrease in methylene blue rejection ratios. Thereby, the membrane-40 presented higher rejection ratios than other membranes because of uniform nanopores endowing membranes with excellent performance. 3.4. Oil adsorption efficiency of electrospun PLLA nanoporous fiber membranes According to the results of hydrophobic tests, all the nanoporous fiber membranes were hydrophobic and oleophylic. This kind of electrospun PLLA nanoporous fiber membranes could have a high selectivity of the oil phase for the oil-water mixtures [43]. Therefore, in order to study the oil-absorbing ability of PLLA nanoporous fiber membranes, the oil absorption efficiency was tested. Table 3 showed that all the membranes at different humidity had higher oil absorption than non-porous fiber membranes. As could be seen, when the humidity increased, the oil adsorption efficiency of membranes increased and then decreased, and the maximum value was reached when the humidity was 40%. Additionally, compared with that of non-porous fiber membranes, the oil adsorption efficiency of membrane-30, membrane-40, membrane-50 and membrane-60 was improved by 27.1%, 54.9%, 38.6% and 41.8%, respectively. The variation trend of oil absorption was in accord with the changes of oil contact angles. The high absorption efficiency was related to the nanoporous structures on the single fiber, which could hold oil in them with proper sizes. Therefore, increasing the specific surface areas of PLLA nanoporous fiber membranes can improve the efficiency of oil adsorption. 4. Conclusions A novel PLLA nanoporous fiber membrane was successfully fabricated via the thermal-nonsolvent induced phase separation technique by electrospinning a single solvent system of PLLA/dichloromethane. In addition, a water bath was used to obtain the nanoporous fiber membranes within a wide range of humidity. The following can be highlighted from the experimental study: (1) The PLLA nanoporous fiber membranes exhibited an ideal morphology, i.e. higher specific surface area compared with nonporous fiber membranes, resulting in enhanced oil adsorption properties and separation performances. (2) The PLLA nanoporous fiber membranes exhibited significant improvement in hydrophobicity and water permeability, and the pure water flux of PLLA nanoporous fiber membranes reached 4836.6 L m−2 h−1 in comparasion to 3362.3 L m−2 h−1 for non-porous fiber membranes, implying the improved
membrane structures due to the formation of nanoporous structures on the single fiber. (3) For the PLLA nanoporous fiber membranes, there was a 40.7% and 54.9% increase in methylene blue rejection ratio and oil adsorption capacity respectively, compared with nonporous fiber membranes, suggesting that nanoporous structures formed via the thermal-nonsolvent induced phase separation technique endowed membranes with preferable organic dye rejection ratio and oil adsorption capacity. Acknowledgments The work was funded by the National Natural Science Foundation of China (51408416) and the Petrochemical Joint Funds of National Natural Science Fund Committee-China National Petroleum Corporation (U1362108). Reference [1] Lei C, YongYi Y, RuiXia L. Recent advance in manufacture of nano-fibers by electrospinning. Textile Inst 2004;5:1–6. [2] Pant HR, Pokharel P, Joshi MK, Adhikari S, Kim HJ, Park CH, Kim CS. Processing and characterization of electrospun graphene oxide/polyurethane composite nanofibers for stent coating. Chem Eng J 2015;270:336–42. [3] Zheng J, Liu K, Reneker DH, Becker ML. Post-assembly derivatization of electrospun nanofibers via strain-promoted azide alkyne cycloaddition. J Am Chem Soc 2012;134:17274–7. [4] Wu J, Wang N, Zhao Y, Jiang L. Electrospinning of multilevel structured functional micro-/nanofibers and their applications. J Mater Chem A 2013;1:7290–305. [5] Linh NTB, Lee KH, Lee BT. Fabrication of photocatalytic PVA–TiO2 nano-fibrous hybrid membrane using the electro-spinning method. J Mater Sci 2011;46:5615– 20. [6] Liu W, Thomopoulos S, Xia Y. Electrospun nanofibers for regenerative medicine. Adv Healthc Mater 2012;1:10–25. [7] Srivastava Y, Rhodes C, Marquez M, Thorsen T. Electrospinning hollow and core/sheath nanofibers using hydrodynamic fluid focusing. Microfluid Nanofluid 2008;5:455–8. [8] Zhang Q, Li M, Liu J, Long S, Yang J, Wang X. Porous ultrafine fibers via a saltinduced electrospinning method. Colloid Polym Sci 2012;290:793–9. [9] Seyyed J, Saner B, Letofsky I, Yildiz M, Menceloglu YZ. Rational design and direct fabrication of multi-walled hollow electrospun fibers with controllable structure and surface properties. Eur Polym J 2015;62:66–76. [10] Wang Q, Cao Q, Wang X, Jing B, Kuang H, Zhou L. Dual template method to prepare hierarchical porous carbon nanofibers for high-power supercapacitors. J Solid State Electr 2013;17:2731–9. [11] Moon S, Choi J, Farris RJ. Highly porous polyacrylonitrile/polystyrene nanofibers by electrospinning. Fiber Polym 2008;9:276–80. [12] Sun D, Qin G, Lu M, Wei W. Preparation of mesoporous polyacrylonitrile and carbon fibers by electrospinning and supercritical drying. Carbon 2013;63:585–9. [13] Guangming G, Juntao W, Lei J. Novel polyimide materials produced by electrospinning. Prog Chem 2011;23:750–60. [14] Hardick O, Stevens B, Bracewell DG. Nanofibre fabrication in a temperature and humidity controlled environment for improved fibre consistency. J Mater Sci 2011;46:3890–8. [15] Honarbakhsh S, Pourdeyhimi B. Scaffolds for drug delivery, part I: electrospun porous poly(lactic acid) and poly(lactic acid)/poly(ethylene oxide) hybrid scaffolds. J Mater Sci 2010;46:2874–81. [16] Kim ES, Kim SH, Lee CH. Electrospinning of polylactide fibers containing silver nanoparticles. Macromol Res 2010;18:215–21. [17] Li X. Nano-porous ultra-high specific surface ultrafine fibers. Chinese Sci Bull 2004;49:2368–71. [18] Huang L, Bui NN, Manickam SS, McCutcheon RJ. Controlling electrospun nanofiber morphology and mechanical properties using humidity. Polym Phys 2011;49:1734–44. [19] Lannutti J, Reneker D, Ma T, Tomasko D, Farson D. Electrospinning for tissue engineering scaffolds. Mater Sci Eng: C 2007;27:504–9. [20] Zeng J, Aigner A, Czubayko F, Kissel T. Poly(vinyl alcohol) nanofibers by electrospinning as a protein delivery system and the retardation of enzyme release by additional polymer coatings. Biomacromolecules 2005;6:1484–8. [21] Kim CH, Kim BH. Electrochemical behavior of zinc oxide-based porous carbon composite nanofibers as an electrode for electrochemical capacitors. J Electroanal Chem 2014;730:1–9.
Please cite this article as: X. Li et al., Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation, Journal of the Taiwan Institute of Chemical Engineers (2015), http://dx.doi.org/10.1016/j.jtice.2015.11.012
JID: JTICE
ARTICLE IN PRESS
[m5G;December 5, 2015;13:25]
X. Li et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2015) 1–7 [22] Shengguan C, Binghuang H, Haiqing L. Fabrication of nano-porous structured polylactide (PLLA) fibers through electrospinning. Acta Polym Sin 2010;10:1193– 8. [23] Qi Z, Yu H, Chen Y, Zhu M. Highly porous fibers prepared by electrospinning a ternary system of nonsolvent/solvent/poly(l-lactic acid). Mater Lett 2009;63:415– 18. [24] Fashandi H, Karimi M. Characterization of porosity of polystyrene fibers electrospun at humid atmosphere. Thermochim Acta 2012;547:38–46. [25] Chi HK, Yoon HJ, Hak YK, Lee DR. Effect of collector temperature on the porous structure of electrospun fibers. Macromol Res 2006;14:59–65. [26] Park JY, Lee IH. Controlled release of ketoprofen from electrospun porous polylactic acid (PLA) nanofibers. J Polym Res 2010;18:1287–91. [27] Liu R., Liu J., Haiqing L. Fabrication of PLLA/CA composites porous ultrafine fibers. Acta Polym Sin 2013;1312–18. [28] Agarwal C, Pandey AK, Das S, Sharma MK, Pattyn D, Ares P, Goswami A. Neck-size distributions of through-pores in polymer membranes. J Membr Sci 2012;415416:608–15. [29] Li JF, Xu ZL, Yang H. Microporous polyethersulfone membranes prepared under the combined precipitation conditions with non-solvent additives. Polym Advan Technol 2008;19:251–7. [30] Zhang J, Xu Z, Mai W, Min C, Zhou B, Shan M, Li Y, Yang C, Wang Z, Qian X. Improved hydrophilicity, permeability, antifouling and mechanical performance of PVDF composite ultrafiltration membranes tailored by oxidized low-dimensional carbon nanomaterials. J Mater Chem A 2013;1:3101–11. [31] Xia C, Zhao Q, Ye Z, Zhu W. Removal of organic pollutants from mononitrotoluene (MNT) wastewater by reduced pressure distillation. Sep Purif Technol 2013;120:1–5. [32] Lalia BS, Kochkodan V, Hashaikeh R, Hilal N. A review on membrane fabrication: Structure, properties and performance relationship. Desalination 2013;326:77– 95.
7
[33] Wu J, Wang N, Wang L, Dong H, Zhao Y, Jiang L. Electrospun porous structure fibrous film with high oil adsorption capacity. ACS Appl Mater & Inter 2012;4:3207–12. [34] Czado W, Frese T. Nanostructured Fibers via Electrospinning. Adv Mater 2001;13:70–2. [35] Huang L, Bui NN, Manickam SS, McCutcheon JR. Controlling electrospun nanofiber morphology and mechanical properties using humidity. J Polym Sci, Part B: Polym. Phys 2011;49:1734–44. [36] Su C, Tong Y, Zhang M, Zhang Y, Shao C. TiO2 nanoparticles immobilized on polyacrylonitrile nanofibers mats: a flexible and recyclable photocatalyst for phenol degradation. RSC Adv 2013;3:7503–12. [37] Gryta M, Bastrzyk J, Lech D. Evaluation of fouling potential of nanofiltration membranes based on the dynamic contact angle measurements. Pol J Chem Technol 2012;14:97–104. [38] Wang G, Guo Z, Liu W. Interfacial Effects of Superhydrophobic Plant Surfaces: A Review. J Bionic Eng 2014;11:325–45. [39] Mansouri J, Harrisson S, Chen V. Classification, fabrication methods and applications of inorganic hollow spheres. Prog Chem 2011;23:2522–34. [40] Ma J, Zhao Y, Xu Z, Min C, Zhou B, Li Y, Li B, Niu J. Role of oxygen-containing groups on MWCNTs in enhanced separation and permeability performance for PVDF hybrid ultrafiltration membranes. Desalination 2013;320:1–9. [41] Woo SH, Park J, Min BR. Relationship between permeate flux and surface roughness of membranes with similar water contact angle values. Sep Purif Technol 2015;146:187–91. [42] Zhang Y, Su Y, Peng J, Zhao X, Liu J, Zhao J, Jiang Z. Composite nanofiltration membranes prepared by interfacial polymerization with natural material tannic acid and trimesoyl chloride. J Membr Sci 2013;429:235–42. [43] Nuraje N, Khan WS, Lei Y, Ceylan M, Asmatulu R. Superhydrophobic electrospun nanofibers. J Mater Chem A 2013;1:1929–46.
Please cite this article as: X. Li et al., Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation, Journal of the Taiwan Institute of Chemical Engineers (2015), http://dx.doi.org/10.1016/j.jtice.2015.11.012
JID: JTICE
[m5G;December 5, 2015;13:25]
Journal of the Taiwan Institute of Chemical Engineers 000 (2015) 1–7
Contents lists available at ScienceDirect
Journal of the Taiwan Institute of Chemical Engineers journal homepage: www.elsevier.com/locate/jtice
Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation Xianhua Li, Kunyue Teng, Jie Shi, Wei Wang, Zhiwei Xu∗, Hui Deng, Hanming Lv, Fengyan Li State Key Laboratory of Separation Membranes and Membrane Processes, School of Textiles, Tianjin Polytechnic University, Tianjin 300387, PR China
a r t i c l e
i n f o
Article history: Received 25 June 2015 Revised 3 November 2015 Accepted 16 November 2015 Available online xxx Keywords: Electrospinning Nanoporous fibers Phase separation Water bath Humidity
a b s t r a c t A novel polylactic acid (PLLA) fiber membrane with nanoporous structures has been successfully developed via thermal-nonsolvent induced a phase separation technique by electrospinning a single solvent system of PLLA/dichloromethane. A water bath receiver was necessary for preparing the nanoporous fiber membranes in this paper. The results showed that the nanoporous fiber membranes could be obtained within a wide range of humidity. Compared with non-porous PLLA fiber membranes, the nanoporous fiber membranes displayed significantly improved specific surface area (improved about 20%) and rejection ratio (about 3 times) toward methylene blue. Besides, the PLLA nanoporous fiber membranes showed a water flux up to 4836.6 L m−2 h−1 , increased by 25% than that of the non-porous fiber membranes. In addition, the oil adsorption capacity of the PLLA nanoporous fiber membranes could reach 26.8 g g−1 . The new developed PLLA nanoporous fiber membranes via thermal-nonsolvent induced phase separation technique would demonstrate an impressive prospect for oil adsorption and organic dye filtration. © 2015 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
1. Introduction Electrospinning is a straightforward way to prepare micro-/ nanometer scale and high porosity fibers [1,2]. Since the electrospun fibers have many advantages, such as good repeatability, various morphological modifications, and three-dimensional porous structures [3,4], electrospinning progress has attracted a great deal of attention. Therefore, these nanofibers/nanowebs are quite applicable in many areas including filtration [5], biotechnology [6], etc. Some interesting structures, such as core/shell structure [7], porous structure [8] and hollow structure [9], have emerged in recent years, and attracted widely attention due to the unique properties and functionalities. A lot of applications may be favored if the fibers are endowed with rough or porous structures [10]. In fact, widespread attention has been paid to the electrospinning technique in increasing the specific surface area and porosity of the fiber membranes [11,12]. Up to now, there were four methods to prepare the porous fiber membranes in literature: (1) using high/low boiling point solvent mixtures [13], (2) using solvent/nonsovent mixtures [14,15], (3) post-processing electrospun composite nanofibers by selectively removing one of the components [16,17], (4) electrospinning of polymers in a high humid environment [18]. Currently, porous fiber membranes have been used for various fields including tissue
∗
Corresponding author. Tel./fax: + 86 22 83955231. E-mail address: [email protected] (Z. Xu).
engineering [19], carriers in drug delivery system [20], and electrode application [21]. In recent years, a lot of methods have been published to prepare electrospun porous fibers. Cao and his coworkers [22] have prepared polylactic acid ultrafine fibers with mesopores of 30–150 nm by electrospinning. The fiber diameters and morphologies could be controlled by adjusting the composition ratio of the dichloromethane/dimethyformamide solvent mixtures. A facile method to fabricate porous fiber membranes by electrospinning a ternary system of nonsolvent/solvent/poly(L-lactic acid) was presented by Qi and his coworkers [23]. Nanoporous polyacrylonitrile ultrafine fibers were prepared by Li and his workers [17]. Post-processing electrospun polyacrylonitrile composite nanofibers by selectively removing polyvinylpyrrolidone was the main method to prepare porous fibers. The specific surface area of the porous polyacrylonitrile ultrafine fiber membranes was more than 70 m2 g−1 . The porosity of electrospinning porous fibers obtained from polystyrene/dimethyformamide solution at 60% relative humidity was characterized by H.Fashandi et al. [24]. It was obvious that the high humid environment was the main factor to prepare polystyrene porous fibers. Kim et al. [25] fabricated the electrospun polymer nonwoven mats with porous surface morphologies by varying the collector temperature. Though these fore-mentioned methods have been used widely, there are a few factors to be considered in the process of preparation, such as dispersion of binary polymers in the same solvent, compatibility between binary solvents, and the control of the relative humidity.
http://dx.doi.org/10.1016/j.jtice.2015.11.012 1876-1070/© 2015 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
Please cite this article as: X. Li et al., Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation, Journal of the Taiwan Institute of Chemical Engineers (2015), http://dx.doi.org/10.1016/j.jtice.2015.11.012
ARTICLE IN PRESS
JID: JTICE 2
[m5G;December 5, 2015;13:25]
X. Li et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2015) 1–7
Fig. 1. The process of electrospinning and the principle of nanopore formation. Table 1 The contrast of preparation methods. Solvent
Humidity or temperature
Receiver
Mainly application
Ref.
Double solvent Double solvent Double solvent Single solvent Single solvent
Not mentioned 49%, room temperature 40%, room temperature 25%, 40ºC 20–70%, room temperature
Roller Roller Roller Roller Water bath
Drug carrier Tissue engineering Tissue engineering Tissue engineering Filtration and oil adsorption
[26] [27] [23] [25] This work
In this study, without regard to complicate factors in the process of preparation, polylactic acid fibers with nanoporous structures were obtained by electrospinning a single solvent system of PLLA/dichloromethane. Based on the mechanism of thermalnonsolvent induced phase separation, a water bath was used to obtain the nanoporous fiber membranes within a wide range of humidity. The resulting membranes were subject to characterizations including scanning electron microscope (SEM), nanopore sizes and interconnected pore size distributions, specific surface area measurements, contact angle tests, oil adsorption tests, permeation flux and rejection measurements. The influence of humidity during electrospinning on the morphologies and properties of membranes was investigated. 2. Experiment 2.1. Materials Polylactic acid (PLLA, Mw = 110,000) was purchased from Ningbo global biological material Co.Ltd. Dichloromethane (CH2 Cl2 , AR, 99.5%) was supplied by Tianjin Fuyu Chemical Co. Ltd. All the chemicals were analytical grade. 2.2. Preparation of electrospun PLLA nanoporous fiber membranes The PLLA granules were dissolved in dichloromethane solvent and the concentration was 9 wt%, and then the PLLA solution was mixed uniformly by magnetic stirring for 6 h. The setup of electrospinning process was composed of a high voltage power supply, an injector, a water bath (receiver) and a humidifier. The obtained PLLA spinning solution was placed in the injector with a capillary tip (inner diameter=0.67 mm). The anode of the high voltage power supply was clamped to an injector needle tip, and the cathode was connected to a water bath. The electrospun fibers were collected in the water bath. The applied voltage was 20 kV, the tipto-collector distance was 15 cm, and the flow rate of the spinning solution was 2 ml h−1 . All of the electrospinning operations were
performed under room temperature. After electrospinning, the resulting membranes were subject to characterizations without any treatments except for drying for 24 h. Fig. 1 showed the process of electrospinning and the principle of nanopore formation. It was obvious that the change of ambient humidity was due to the cooperation between humidifier and water bath. And the variation of humidity was read by hygrometer. In this paper, the influence of humidity during electrospinning on the morphologies and properties of membranes was investigated. Hence, according to the reading of hygrometer, the change of ambient humidity was achieved by adjusting the humidifier. In order to identify all various kinds of membranes easily, the membranes which were prepared at the humidity of 30%, 40%, 50% and 60% were named as membrane-30, membrane-40, membrane-50 and membrane-60 correspondingly. In addition, the non-porous fiber membranes were obtained at the same electrospinning conditions except for a roller receiver. Compared with other preparation methods, the advantages of our work were shown in Table 1. 2.3. Morphology analysis The surface morphologies of electrospun PLLA nanoporous fiber membranes were investigated using SEM (Quanta 200,Holland) instruments. Prior to the observation, all the samples were coated with gold. After SEM tests, the nanopore sizes on the single fiber were measured by Image-Pro Plus software on the images of SEM. And each final nanopore diameter was obtained by averaging over more than fifty nanopores of different fibers. The interconnected pore size distributions of the membranes were obtained by capillary flow porometry measurements [28], based on the wet/dry flow method. The membranes were immersed in wetting liquid (Porefil, provided by PMI) with a surface tension of 16 dyn/cm for 24 h at room temperature. A wet gas was performed using a solvent filled sample that was followed by a dry gas (sample without wetting liquid) for obtaining flow rate curves as a function of applied pressures. The differential pressure of the gas at which flow through a pore occurs yields the throat (most constricted) diameter
Please cite this article as: X. Li et al., Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation, Journal of the Taiwan Institute of Chemical Engineers (2015), http://dx.doi.org/10.1016/j.jtice.2015.11.012
ARTICLE IN PRESS
JID: JTICE
[m5G;December 5, 2015;13:25]
X. Li et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2015) 1–7
along the pore. Because of the low pressure applied during the process, the porous structures of fibrous membranes are not distorted. All porometry tests were performed at room temperature. The overall porosity (ε ) was determined by the gravimetric method [29], as defined in the following Eq. (1):
ε=
ω1 −ω2 × 100% A × l × dw
(1)
where ω1 was the weight of the wet membrane (g), ω2 was the weight of the dry membrane (g), A was the membrane effective area (m2 ), dw was the water density (0.998 g cm−3 ) and l was the membrane thickness (nm). The specific surface areas of samples were analyzed by physical adsorption of gases (N2 at 77 K) in volumetric adsorption systems (NOVA 2000e,Quantachrome Ins). All samples were degassed at 60ºC (PLLA fibers hardly endure high temperature) for 4 h prior to the adsorption measurement. In addition, the five point method was used to calculate the value of the specific surface areas. And each final specific surface area was obtained by averaging over more than five specific surface areas. 2.4. Contact angle tests The water and oil contact angles of as-prepared membranes were calculated from a digital video image of the drop on the membranes using an image processing program (JYSP-180 Contact Angle Analizer). Pure water and soybean oil were chosen for the test. In general, each final contact angle was obtained by averaging over more than five contact angle values of different spots. 2.5. The permeation flux and rejection measurements Permeation flux and rejection were evaluated for investigating the filtration performance of as-prepared membranes [30]. A dead-end filtration experimental setup was used to measure the pure water flux with the effective membrane area of 19.3 cm2 . Measuring protocol was depicted as follows: for the first 10 min, the membranes were compacted at 0.1 MPa to get a stead flux; then the flux was recorded at 0.1 MPa every 5 min, and at least 5 readings were collected to obtained an average value. The flux was obtained by the following Eq. (2):
Q = V/(A × T )
(2)
where Q was permeation flux of membranes for pure water (L m−2 h−1 ), V was the volume of permeated pure water (L), T was the permeation time and A was the effective area of the membranes (m2 ). Organic dyes were often used for the target contaminant to measure the rejection ratio of membranes. The rejection tests were carried out with an aqueous solution of methylene blue [31] (molecular weight = 373.9). After the measurement of pure water flux, pure water was changed to methylene blue solution. The final rejection ratio was obtained by the different concentration of permeated and feed solutions [32]. The rejection ratio (R,%) was computed by the following Eq. (3):
R=
Cp 1− Cf
× 100%
(3)
where Cp (mg L−1 ) and Cf (mg L−1 ) were the concentrations of permeation and feed solutions, respectively. The dye concentration of initial feed solution was 6 mg L−1 , and the dye concentrations of permeation solutions were detected by a UV-spectrophotometer (Shimadzu UV2450, Japan). 2.6. Oil adsorption tests Soybean oil was chosen for the test. 0.05 g membrane was placed on the soybean oil. With high buoyancy, the sample floated on the
3
surface and kept sucking the soybean oil initially. After a period of time, the wet membrane was taken out and drained for 5 min until no residual oil droplet was left on the surface, and then weighed [33]. Oil adsorption capacity of membranes was determined by the following Eq. (4):
q = (mw − m0 )/m0
(4)
where q was the adsorption capacity (g g−1 ), m0
was the initial weight of the membranes (g), and mw was the weight of the wet membranes after 5 min drainage (g). 3. Results and discussion 3.1. Morphologies of electrospun PLLA nanoporous fiber membranes The differences of membranes were compared in Fig. 2, as observed by SEM. All the membranes presented typical nanoporous structures regardless of humidity. However, the morphologies of nanoporous fibers were affected by the relative humidity of environment. For membrane-40, the nanopores compared with the other membranes could be viewed visually from the SEM pictures. Based on Image-Pro Plus analysis in Table 2, the changes of the nanopore diameters at various relative humidity were in accord with the SEM pictures of fibers. It was indicated that low or high humidity was not favorable for nanopore formation. Low humidity could not play the role of water vapor sufficiently. When the humidity was increasing constantly, dichloromethane would volatilize rapidly. High humidity might affect technology of spinning, and then intervene nanopore formation. To investigate the mechanism of nanopore formation [34], a typical schematic diagram of polymer-solvent-nonsolvent system was shown in Fig. 1. In the process of electrospinning, dichloromethane could evaporate rapidly because of its low boiling point. Due to rapid solvent evaporation, the fibrous surface temperature decreased significantly. The phenomenon of rapid solvent evaporation might attribute to the thermally induced phase separation [33]. Thermally induced phase separation would cause polymer-rich region to form fiber frame and solvent-rich region to form nanopores. In fact, thermally induced phase separation was just part of reasons for nanopore formation. Another important reason was nonsolvent induced phase separation. When the temperature of fiber surface cooled down, moisture in the air might be condensed and grew in the form of droplets. And then many water droplets would be left on the surface of fibers. Moreover, water was the nonsolvent for PLLA. When the fibers which loaded a large number of water vapor solidified, the nanopores would be formed due to the evaporation of water vapor in the PLLA fibers [35]. Whereas, the regulation of water vapor in the air by adjusting the humidifier hardly meet requirements. To solve this problem, the fibers were received by a water bath. The water bath not only supplied quantities of water vapor, but also dropped the environment temperature in a degree. Therefore, the water bath was crucial for the nanopore formation. In practice, the process of nanopore formation can be interpreted as the exchange of dichloromethane and water into fibers (Fig. 1). The overall porosity information of the membranes was presented in Table 2. The results of the porosity measurements revealed that all the nanoporous fiber membranes possessed a good porosity in the range of 60-70%, while the non-porous fiber membranes had a porosity of 48.5%. It was obvious that there were two kinds of pores in the nanoporous fiber membranes, i.e. the nanopores on the single fiber and the interconnected pores that the fibers made from. Hence, compared with non-porous fiber membranes, the porosity of the nanoporous fiber membranes should be improved due to the nanoporous structures on the single fiber. From Fig. 3, it was clear that the interconnected pore diameter of nanoporous and non-porous membranes was 2-4 μm and 3-5 μm, respectively, indicating the
Please cite this article as: X. Li et al., Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation, Journal of the Taiwan Institute of Chemical Engineers (2015), http://dx.doi.org/10.1016/j.jtice.2015.11.012
ARTICLE IN PRESS
JID: JTICE 4
[m5G;December 5, 2015;13:25]
X. Li et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2015) 1–7
Fig. 2. SEM images of electrospun PLLA fibers: (a) membrane-30, (b) membrane-40, (c) membrane-50, (d) membrane-60. The insets showed the high magnification images, correspondingly. Table 2 The nanopore diameter, porosity and specific surface area of different membranes. Sample
Nanopore diameter (nm)
Porosity (%)
Specific surface area (m2 g−1 )
Membrane-30 Membrane-40 Membrane-50 Membrane-60 Non-porous membranes
187.6 ± 1.1 265.7 ± 0.8 176.3 ± 0.3 199.7 ± 1.4 –
62.7 ± 0.9 69.1 ± 0.4 61.3 ± 1.1 65. 7 ± 0.3 48.5 ± 0.7
182.2 ± 5.3 200.3 ± 3.6 191.5 ± 1.8 173.2 ± 2.1 152.5 ± 3.1
negligible difference in the interconnected pore diameter distributions of all membranes. Therefore, it could be considered that the interconnected pores in different membranes had the same influence on the performance of membranes to some extent. This phenomenon could not be shown in Fig. 2 because the images of SEM revealed only the random and local morphologies of membranes. According to the above analysis, it was shown that the nanoporous fiber membranes were all endowed with advantageous porous structures, which undoubtedly played a positive role in promoting membrane permeability. The specific surface areas of the as-prepared PLLA nanoporous fiber membranes were analyzed by isothermal N2 adsorption measurements [36]. The results showed that all the membranes with
nanoporous structures had higher specific surface areas than nonporous fiber membranes though the humidity changed. Fig. 4 showed the isotherm of the relative pressure (P/P0 ) ranging from 0.05 to 0.35 via five point method. The specific surface areas of the electrospun PLLA nanoporous fiber membranes at different relative humidity were listed in Table 2. Based on the values of specific surface areas, the changes of the specific surface areas at various relative humidity were in accord with the morphologies of fibers. Particularly, when the relative humidity was 40%, the specific surface area reached 200.3 m2 g−1 . Compared with the specific surface area of non-porous fiber membranes, it increased about 20%. The high specific surface areas could be attributed to the construction of nanopores. It was worth mentioning that the high specific surface area could facilitate
Please cite this article as: X. Li et al., Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation, Journal of the Taiwan Institute of Chemical Engineers (2015), http://dx.doi.org/10.1016/j.jtice.2015.11.012
JID: JTICE
ARTICLE IN PRESS
[m5G;December 5, 2015;13:25]
X. Li et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2015) 1–7
5
Fig. 3. The interconnected pore diameter distributions of different membranes. Fig. 5. Contact angles of as-prepared membranes.
Fig. 4. The isotherm with five point method BET of different membranes.
efficient contact of fibers with organic pollutants, which was conductive to filtration.
Fig. 6. Comparison of pure water flux and methylene blue rejection ratio of the different membranes.
contact angles of membranes [38]. The changes of water contact angles might play an unfavourable role in improving the water flux of membranes [39].
3.2. Hydrophobicities of electrospun PLLA nanoporous fiber membranes Contact angle is an important parameter for measuring surface hydrophilicity or hydrophobicity [37]. In general, larger water contact angle refers to higher hydrophobicity. Compared in Fig. 5 were the contact angle results of various membranes. It could be found that all nanoporous fiber membranes were more hydrophobic and oleophilic than non-porous fiber membranes. The water contact angles of membrane-40 could attain 115°, and the water contact angles of membrane-30, membrane-50 and membrane-60 were all below 105°. In comparison, the soybean oil droplet immediately spread on the membranes. The oil contact angles of membrane-30, membrane50 and membrane-60 were all above 15°. Particularly, the oil contact angle of membrane-40 could reach 12°. As the humidity increased, the water contact angle values of membranes increased and then decreased, and the maximum value was observed for membrane-40. The change trend of oil contact angles were contrary to that of water contact angles. These results indicated that the nanoporous structures of the membranes led to high water contact angle and low oil contact angle, especially compared with the non-porous membranes. It was reported that the surface roughness could affect the
3.3. Permeation flux and rejection of electrospun PLLA nanoporous fiber membranes The data of the pure water flux of different membranes were depicted in Fig. 6. It suggested an obvious trend that all the membranes with nanoporous structures revealed higher water permeation fluxes compared with non-porous fiber membranes. When the relative humidity ranged from 30% to 40%, the pure water flux went up. However, when the relative humidity ranged from 40% to 60%, the pure water flux went down. Expressly, the water flux of membrane-40 reached a peak value of 4836.6 L m−2 h−1 , and increased approximately four times compared with that of flat ultrafiltration membrane [40]. Considering the results in Fig. 6, the improvement of permeability might be due to the morphologies of membranes. The nanopores on the single fiber undoubtedly played a positive role in promoting membrane permeability. Consequentially, though the membranes were hydrophobic, nanoporous fiber membranes still showed good water permeability [41]. The results for the methylene blue rejection ratio of different membranes were also depicted in Fig. 6. As could be seen, all the
Please cite this article as: X. Li et al., Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation, Journal of the Taiwan Institute of Chemical Engineers (2015), http://dx.doi.org/10.1016/j.jtice.2015.11.012
ARTICLE IN PRESS
JID: JTICE 6
[m5G;December 5, 2015;13:25]
X. Li et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2015) 1–7 Table 3 The oil adsorption efficiency of different membranes. Sample Efficiency(g g
−1
)
Membrane-30
Membrane-40
Membrane-50
Membrane-60
Non-porous membranes
16.6 ± 0.3
26.8 ± 0.1
19.7 ± 0.2
20.8 ± 0.4
12.1 ± 0.2
nanoporous fiber membranes presented higher methylene blue rejection ratios than non-porous fiber membranes. Particularly, the methylene blue rejection ratio increased significantly when the relative humidity was 40%. However, when the relative humidity continued increasing, the rejection ratio declined. Additionally, compared with those of non-porous fiber membranes (37.0%), the methylene blue rejection ratios of membrane-30, membrane-40, membrane-50 and membrane-60 were improved by 27.9%, 40.7%, 28.9% and 29.9%, respectively. It might be due to the nanopores on the single fiber [42]. The uniform nanopores of membranes could enhance methylene blue rejection ratios. Moreover, the decline in hydrophobic interaction between methylene blue dye and hydrophobic surface of membranes might be a reason for slight decrease in methylene blue rejection ratios. Thereby, the membrane-40 presented higher rejection ratios than other membranes because of uniform nanopores endowing membranes with excellent performance. 3.4. Oil adsorption efficiency of electrospun PLLA nanoporous fiber membranes According to the results of hydrophobic tests, all the nanoporous fiber membranes were hydrophobic and oleophylic. This kind of electrospun PLLA nanoporous fiber membranes could have a high selectivity of the oil phase for the oil-water mixtures [43]. Therefore, in order to study the oil-absorbing ability of PLLA nanoporous fiber membranes, the oil absorption efficiency was tested. Table 3 showed that all the membranes at different humidity had higher oil absorption than non-porous fiber membranes. As could be seen, when the humidity increased, the oil adsorption efficiency of membranes increased and then decreased, and the maximum value was reached when the humidity was 40%. Additionally, compared with that of non-porous fiber membranes, the oil adsorption efficiency of membrane-30, membrane-40, membrane-50 and membrane-60 was improved by 27.1%, 54.9%, 38.6% and 41.8%, respectively. The variation trend of oil absorption was in accord with the changes of oil contact angles. The high absorption efficiency was related to the nanoporous structures on the single fiber, which could hold oil in them with proper sizes. Therefore, increasing the specific surface areas of PLLA nanoporous fiber membranes can improve the efficiency of oil adsorption. 4. Conclusions A novel PLLA nanoporous fiber membrane was successfully fabricated via the thermal-nonsolvent induced phase separation technique by electrospinning a single solvent system of PLLA/dichloromethane. In addition, a water bath was used to obtain the nanoporous fiber membranes within a wide range of humidity. The following can be highlighted from the experimental study: (1) The PLLA nanoporous fiber membranes exhibited an ideal morphology, i.e. higher specific surface area compared with nonporous fiber membranes, resulting in enhanced oil adsorption properties and separation performances. (2) The PLLA nanoporous fiber membranes exhibited significant improvement in hydrophobicity and water permeability, and the pure water flux of PLLA nanoporous fiber membranes reached 4836.6 L m−2 h−1 in comparasion to 3362.3 L m−2 h−1 for non-porous fiber membranes, implying the improved
membrane structures due to the formation of nanoporous structures on the single fiber. (3) For the PLLA nanoporous fiber membranes, there was a 40.7% and 54.9% increase in methylene blue rejection ratio and oil adsorption capacity respectively, compared with nonporous fiber membranes, suggesting that nanoporous structures formed via the thermal-nonsolvent induced phase separation technique endowed membranes with preferable organic dye rejection ratio and oil adsorption capacity. Acknowledgments The work was funded by the National Natural Science Foundation of China (51408416) and the Petrochemical Joint Funds of National Natural Science Fund Committee-China National Petroleum Corporation (U1362108). Reference [1] Lei C, YongYi Y, RuiXia L. Recent advance in manufacture of nano-fibers by electrospinning. Textile Inst 2004;5:1–6. [2] Pant HR, Pokharel P, Joshi MK, Adhikari S, Kim HJ, Park CH, Kim CS. Processing and characterization of electrospun graphene oxide/polyurethane composite nanofibers for stent coating. Chem Eng J 2015;270:336–42. [3] Zheng J, Liu K, Reneker DH, Becker ML. Post-assembly derivatization of electrospun nanofibers via strain-promoted azide alkyne cycloaddition. J Am Chem Soc 2012;134:17274–7. [4] Wu J, Wang N, Zhao Y, Jiang L. Electrospinning of multilevel structured functional micro-/nanofibers and their applications. J Mater Chem A 2013;1:7290–305. [5] Linh NTB, Lee KH, Lee BT. Fabrication of photocatalytic PVA–TiO2 nano-fibrous hybrid membrane using the electro-spinning method. J Mater Sci 2011;46:5615– 20. [6] Liu W, Thomopoulos S, Xia Y. Electrospun nanofibers for regenerative medicine. Adv Healthc Mater 2012;1:10–25. [7] Srivastava Y, Rhodes C, Marquez M, Thorsen T. Electrospinning hollow and core/sheath nanofibers using hydrodynamic fluid focusing. Microfluid Nanofluid 2008;5:455–8. [8] Zhang Q, Li M, Liu J, Long S, Yang J, Wang X. Porous ultrafine fibers via a saltinduced electrospinning method. Colloid Polym Sci 2012;290:793–9. [9] Seyyed J, Saner B, Letofsky I, Yildiz M, Menceloglu YZ. Rational design and direct fabrication of multi-walled hollow electrospun fibers with controllable structure and surface properties. Eur Polym J 2015;62:66–76. [10] Wang Q, Cao Q, Wang X, Jing B, Kuang H, Zhou L. Dual template method to prepare hierarchical porous carbon nanofibers for high-power supercapacitors. J Solid State Electr 2013;17:2731–9. [11] Moon S, Choi J, Farris RJ. Highly porous polyacrylonitrile/polystyrene nanofibers by electrospinning. Fiber Polym 2008;9:276–80. [12] Sun D, Qin G, Lu M, Wei W. Preparation of mesoporous polyacrylonitrile and carbon fibers by electrospinning and supercritical drying. Carbon 2013;63:585–9. [13] Guangming G, Juntao W, Lei J. Novel polyimide materials produced by electrospinning. Prog Chem 2011;23:750–60. [14] Hardick O, Stevens B, Bracewell DG. Nanofibre fabrication in a temperature and humidity controlled environment for improved fibre consistency. J Mater Sci 2011;46:3890–8. [15] Honarbakhsh S, Pourdeyhimi B. Scaffolds for drug delivery, part I: electrospun porous poly(lactic acid) and poly(lactic acid)/poly(ethylene oxide) hybrid scaffolds. J Mater Sci 2010;46:2874–81. [16] Kim ES, Kim SH, Lee CH. Electrospinning of polylactide fibers containing silver nanoparticles. Macromol Res 2010;18:215–21. [17] Li X. Nano-porous ultra-high specific surface ultrafine fibers. Chinese Sci Bull 2004;49:2368–71. [18] Huang L, Bui NN, Manickam SS, McCutcheon RJ. Controlling electrospun nanofiber morphology and mechanical properties using humidity. Polym Phys 2011;49:1734–44. [19] Lannutti J, Reneker D, Ma T, Tomasko D, Farson D. Electrospinning for tissue engineering scaffolds. Mater Sci Eng: C 2007;27:504–9. [20] Zeng J, Aigner A, Czubayko F, Kissel T. Poly(vinyl alcohol) nanofibers by electrospinning as a protein delivery system and the retardation of enzyme release by additional polymer coatings. Biomacromolecules 2005;6:1484–8. [21] Kim CH, Kim BH. Electrochemical behavior of zinc oxide-based porous carbon composite nanofibers as an electrode for electrochemical capacitors. J Electroanal Chem 2014;730:1–9.
Please cite this article as: X. Li et al., Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation, Journal of the Taiwan Institute of Chemical Engineers (2015), http://dx.doi.org/10.1016/j.jtice.2015.11.012
JID: JTICE
ARTICLE IN PRESS
[m5G;December 5, 2015;13:25]
X. Li et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2015) 1–7 [22] Shengguan C, Binghuang H, Haiqing L. Fabrication of nano-porous structured polylactide (PLLA) fibers through electrospinning. Acta Polym Sin 2010;10:1193– 8. [23] Qi Z, Yu H, Chen Y, Zhu M. Highly porous fibers prepared by electrospinning a ternary system of nonsolvent/solvent/poly(l-lactic acid). Mater Lett 2009;63:415– 18. [24] Fashandi H, Karimi M. Characterization of porosity of polystyrene fibers electrospun at humid atmosphere. Thermochim Acta 2012;547:38–46. [25] Chi HK, Yoon HJ, Hak YK, Lee DR. Effect of collector temperature on the porous structure of electrospun fibers. Macromol Res 2006;14:59–65. [26] Park JY, Lee IH. Controlled release of ketoprofen from electrospun porous polylactic acid (PLA) nanofibers. J Polym Res 2010;18:1287–91. [27] Liu R., Liu J., Haiqing L. Fabrication of PLLA/CA composites porous ultrafine fibers. Acta Polym Sin 2013;1312–18. [28] Agarwal C, Pandey AK, Das S, Sharma MK, Pattyn D, Ares P, Goswami A. Neck-size distributions of through-pores in polymer membranes. J Membr Sci 2012;415416:608–15. [29] Li JF, Xu ZL, Yang H. Microporous polyethersulfone membranes prepared under the combined precipitation conditions with non-solvent additives. Polym Advan Technol 2008;19:251–7. [30] Zhang J, Xu Z, Mai W, Min C, Zhou B, Shan M, Li Y, Yang C, Wang Z, Qian X. Improved hydrophilicity, permeability, antifouling and mechanical performance of PVDF composite ultrafiltration membranes tailored by oxidized low-dimensional carbon nanomaterials. J Mater Chem A 2013;1:3101–11. [31] Xia C, Zhao Q, Ye Z, Zhu W. Removal of organic pollutants from mononitrotoluene (MNT) wastewater by reduced pressure distillation. Sep Purif Technol 2013;120:1–5. [32] Lalia BS, Kochkodan V, Hashaikeh R, Hilal N. A review on membrane fabrication: Structure, properties and performance relationship. Desalination 2013;326:77– 95.
7
[33] Wu J, Wang N, Wang L, Dong H, Zhao Y, Jiang L. Electrospun porous structure fibrous film with high oil adsorption capacity. ACS Appl Mater & Inter 2012;4:3207–12. [34] Czado W, Frese T. Nanostructured Fibers via Electrospinning. Adv Mater 2001;13:70–2. [35] Huang L, Bui NN, Manickam SS, McCutcheon JR. Controlling electrospun nanofiber morphology and mechanical properties using humidity. J Polym Sci, Part B: Polym. Phys 2011;49:1734–44. [36] Su C, Tong Y, Zhang M, Zhang Y, Shao C. TiO2 nanoparticles immobilized on polyacrylonitrile nanofibers mats: a flexible and recyclable photocatalyst for phenol degradation. RSC Adv 2013;3:7503–12. [37] Gryta M, Bastrzyk J, Lech D. Evaluation of fouling potential of nanofiltration membranes based on the dynamic contact angle measurements. Pol J Chem Technol 2012;14:97–104. [38] Wang G, Guo Z, Liu W. Interfacial Effects of Superhydrophobic Plant Surfaces: A Review. J Bionic Eng 2014;11:325–45. [39] Mansouri J, Harrisson S, Chen V. Classification, fabrication methods and applications of inorganic hollow spheres. Prog Chem 2011;23:2522–34. [40] Ma J, Zhao Y, Xu Z, Min C, Zhou B, Li Y, Li B, Niu J. Role of oxygen-containing groups on MWCNTs in enhanced separation and permeability performance for PVDF hybrid ultrafiltration membranes. Desalination 2013;320:1–9. [41] Woo SH, Park J, Min BR. Relationship between permeate flux and surface roughness of membranes with similar water contact angle values. Sep Purif Technol 2015;146:187–91. [42] Zhang Y, Su Y, Peng J, Zhao X, Liu J, Zhao J, Jiang Z. Composite nanofiltration membranes prepared by interfacial polymerization with natural material tannic acid and trimesoyl chloride. J Membr Sci 2013;429:235–42. [43] Nuraje N, Khan WS, Lei Y, Ceylan M, Asmatulu R. Superhydrophobic electrospun nanofibers. J Mater Chem A 2013;1:1929–46.
Please cite this article as: X. Li et al., Electrospun preparation of polylactic acid nanoporous fiber membranes via thermal-nonsolvent induced phase separation, Journal of the Taiwan Institute of Chemical Engineers (2015), http://dx.doi.org/10.1016/j.jtice.2015.11.012