Langmuir 2006, 22, 1321-1328
1321
Structure and Nanomechanical Characterization of Electrospun PS/Clay Nanocomposite Fibers Yuan Ji, Bingquan Li, Shouren Ge, Jonathan C. Sokolov, and Miriam H. Rafailovich* Department of Materials Science and Engineering, State UniVersity of New York at Stony Brook, Stony Brook, New York 11794-2275 ReceiVed September 13, 2005. In Final Form: NoVember 16, 2005 Electrospinning has been emerging as one of the most efficient methods to fabricate polymer nanofibers. In this paper, PS/clay nanocomposite fibers with varying diameters were electrospun onto solid substrates. The fiber diameters were adjusted from 4 µm to 150 nm by changing the solution concentration. Scanning electron microscopy (SEM), transmission electron microscopy (TEM), and atomic force microscopy (AFM) were used to characterize the fiber morphology. Shear modulation force microscopy (SMFM) was utilized to investigate the surface nanomechanical properties of electrospun fibers as a function of the fiber diameter and temperature. In the absence of clay, no change in Tg was observed, even though a large increase of shear modulus below the glass transition temperature was found. This effect was postulated to result from the molecular chain alignment during electrospinning. The addition of functionalized clays to the spinning solution produced fibers with a highly aligned montmorillonite layer structure at a clay concentration of 4 wt %. Clay agglomerates were observed at higher concentrations. The existence of clay further enhanced the shear modulus of fibers and increased the glass transition temperature by nearly 20 °C.
Introduction Polymer nanocomposites (PNCs) have attracted much attention during the past decade because of their unique performance in mechanical, thermal, electrical, and barrier properties.1 Through nanoscale engineering, we can combine the flexibility of polymers with the high strength and high modulus of inorganic nanoparticle components to produce a new array of materials.2 Most current work in the field of polymer nanocomposites focuses on the bulk (3D) and thin-film (2D) properties,3-6 while very little has been reported on the properties of fibers or one-dimensional (1D) geometries.7-9 Recently, electrospinning has become popular as an effective method to generate continuous 1D structures. The popularity of this technique is mostly due to its simplicity and high efficiency for production of macroscopic quantities of nanoscale structures. The technique of electrospinning dates back almost 70 years ago with the first patent being issued to Formhals in 1934.10 However, very few papers appeared on this subject until the early 1990s when interest in this method was revived.11 A typical electrospinning setup consists of three parts: a syringe with needle, a high-voltage power supply, and a metal target as * To whom correspondence should be addressed. E-mail: mrafailovich@ notes.cc.sunysb.edu. (1) Vaia, R. A.; Krishnamoorti, R. Polymer Nanocomposites: Synthesis, Characterization, and Modeling; The American Chemical Society: Washington, DC, 2002. (2) Vaia, R. A.; Giannelis, E. P. MRS Bulletin 2001, May, 394-401. (3) Shi, H.; Lan, T.; Pinnavaia, T. J. Chem. Mater. 1996, 8, 1584-1587. (4) Pinnavaia, T. J.; Beall, G. Polymer Clay Nanocomposites; Wiley Press: New York, 2001; Chapter 1. (5) Zhang, W. H.; Fu, B. X.; Seo, Y.; Schrag, E.; Hsiao, B.; Mather, P.; Yang, N. L.; Xu, D.; Ade, H.; Rafailovich, M. H.; Sokolov, J. Macromolecules 2002, 35, 8029-8038. (6) Zhang, W. H.; Ge, S.; Wang, Y.; Rafailovich, M. H.; Dhez, O.; Winesett, D. A.; Ade, H.; Shafi, K.; Ulman, A.; Popovitz-Biro, R.; Tenne, R.; Sokolov, J. Polymer 2003, 44, 2109-2115. (7) Ko, F.; Gogotsi, Y.; Ali, A.; Naguib, N.; Ye, H.; Yang, G. L.; Li, C.; Willis, P. AdV. Mater. 2003, 15, 1161-1165. (8) Sen, R.; Zhao, B.; Perea, D.; Itkis, M. E.; Hu, H.; Love, J.; Bekyarova, E.; Haddon, R. C. Nanoletters 2004, 4, 459-464. (9) Madhugiri, S.; Dalton, A.; Gutierrez, J.; Ferraris, J. P.; Balkus, K. J. J. Am. Chem. Soc. 2003, 125, 14531-14538. (10) Formhals, A. U.S. Patent 1,975,504, 1934. (11) Reneker, D. H.; Chun, I. Nanotechnology 1996, 7, 216-223.
the counterelectrode. The polymer solution is loaded in the syringe, and a high-voltage power supply is applied between the tip of the needle and the counterelectrode to generate a strong electrical field. The pendant polymer droplet on the tip of the needle is deformed under the electrical force from a hemispherical shape to a conical shape, referred to as Taylor Cone.12 When the electrical force reaches a critical value to overcome the surface tension of the polymer solution, a thin polymer jet is initiated from the tip of the needle and travels toward the counterelectrode. The charged polymer jet is elongated and undergoes a bending instability region where it whips swiftly in the air by the electric field.13 The solvent evaporates during this period; the elongated thin polymer jet solidifies and finally deposits on the counterelectrode as randomly oriented fiber mats. In comparison to conventional fibers, electrospun fibers have some unique features, such as a high surface-to-volume ratio, a nanoporous structure, and an extremely long length. These properties make them suitable for a variety of applications in the fields of filtration, composite reinforcement, and biomedical structural element.14,15 Electrospinning is a rapid process that includes the stretch of the polymer jet and the evaporation of the solvent within milliseconds. As a result, large shear forces are induced by the electrical field on the fibers. Once the solvent evaporates, the polymer chains are vitrified and cannot relax back to their equilibrium conformations.16 The large shear force and rapid solidification of electrospun fibers may result in molecular chain orientation along the direction of the fiber axis, as well as the alignment of anisotropic nanofillers in the polymer matrix. For example, Dror and Zussman fabricated electrospun nanocomposite fibers of poly(ethylene oxide), in which multiwall carbon nanotubes (MWCNTs) were embedded.17,18 They found that PEO (12) Taylor, G. Proc. R. Soc. London, Ser. A 1964, 280, 383. (13) Reneker, D. H.; Yarin, A. L.; Fong, H.; Koombhongse, S. J. Appl. Phys. 2000, 87, 9, 4531-4547. (14) Huang, Z. M.; Zhang, Y. Z.; Kotaki, M.; Ranakrishna, S. Compos. Sci. Technol. 2003, 63, 2223-2253. (15) Frenot, A.; Chronakis, I. S. Curr. Opin. Colloid Interface Sci. 2003, 8, 64-75. (16) Li, D.; Xia, Y. AdV. Mater. 2004, 16, 1151-1170. (17) Dror, Y.; Salalha, W.; Khalfin, R. L.; Cohen, Y.; Yarin, A. L.; Zussman, E. Langmuir 2003, 19, 7012-7020.
10.1021/la0525022 CCC: $33.50 © 2006 American Chemical Society Published on Web 12/30/2005
1322 Langmuir, Vol. 22, No. 3, 2006
Ji et al.
Table 1. Solvent Properties of Solutions for Electrospinning Tb solvent (°C) THF DMF
Tm (°C)
66 -108 153 -61
Cp / Pv/25 °C η/20 °C (J g-1 γ/20 °C (kPa) (mPa s) 25 °C K-1) (mN m-1) 21.6 2.7
0.468 0.920
7.6 36.7
1.72 2.06
26.4 37.1
crystals were highly aligned along the fiber axis during the electrospinning process. The MWCNTs were embedded in the nanofibers as individual elements, mostly aligned along the fiber axis. Fong et al. generated electrospun fibers of nylon6montmorillonite nanocomposites and also found that the electrospinning process resulted in highly aligned montmorillonite layers and nylon 6 crystallites.19 However, in most current work, the mechanical properties of electrospun fibers are obtained only for the nonwoven fiber mats rather than for the individual fiber because of the small diameters of electrospun fibers. The bulk mechanical properties depend strongly upon the orientation of the fibers within the mat and the cross-sectional area of the mat, which is very hard to control during electrospinning. Therefore, it is of fundamental importance to study the mechanical properties of a single electrospun fiber and compare it with the corresponding bulk behavior. In our study, we used shear modulation force microscopy (SMFM), which was reported by Ge et al.,20 to investigate the surface nanomechanical property of a single electrospun fiber where standard bulk measurements are not possible. Our goal in this work is to fabricate a series of electrospun PS/clay composite fibers with controllable diameters from micro- to nanometer scale and study the surface nanomechanical properties as a function of the fiber diameter and temperature using the SMFM method. The results were then compared with those of other complementary characterization techniques. Atomic force microscopy (AFM) and scanning electron microscopy (SEM) were utilized to study the morphology of electrospun fibers. Transmission electron microscopy (TEM) was employed to determine the orientation of clay inside the fibers. Experimental Procedures Materials. To prepare electrospun PS/clay composite fibers, polystyrene (MW ) 280 000; Sigma Aldrich, Inc.) with different weight concentrations was dissolved in a mixed solvent of tetrahydrafuran (THF) (HPLC; EM Science) and dimethylformamide (DMF) (certified A.C.S.; Fisher Scientific) with a weight ratio of 50:50. The solvent properties were listed in Table 1. Organically modified MMT clays (cloisite 6A, Southern Clay Products, Inc.) were dissolved in the mixed solvent separately and then sonicated in a Branson 3510 sonicator for 3 h to form a yellow slurry. Both polystyrene and cloisite 6A solutions were subsequently mixed together to form PS/clay blends solutions with 0, 1, 4, and 8 wt % clay concentrations, respectively. The PS/clay solutions were then sonicated for another 3 h to obtain homogeneous solutions for electrospinning. Sample Preparation. The experimental set up for electrospinning is shown in Figure 1. Polymer solutions were loaded in a 5 mL glass syringe (Popper and Sons, Inc.) attached to a syringe pump (KDS200, KD Scientific, Inc.). The syringe pump was controlled by a computer to provide a steady flow of the polymer solution during electrospinning. A high-voltage power supply (Gamma High Voltage Research, 0-30 kV) was employed to generate a high potential to a 25-gauge blunt-end syringe needle (0.26 mm i.d., Popper and Sons, Inc.). An aluminum target was horizontally placed 10 cm (18) Salalha, W.; Dror, Y.; Khalfin, R. L.; Cohen, Y.; Yarin, A. L.; Zussman, E. Langumir 2004, 20, 9852-9855. (19) Fong, H.; Liu, W. D.; Wang, C. S.; Vaia, R. A. Polymer 2002, 43, 775780. (20) Ge, S. R.; Pu, Y.; Zhang, W.; Sokolov, J. S.; Rafailovich, M. H. Phys. ReV. Lett. 2000, 85, 2340-2343.
Figure 1. Schematic of the electrospinning apparatus. The polymer solution was loaded in the syringe. A high voltage of 10 kV was applied between the needle and the metal screen. The flow rate of solution was computer-controlled. Table 2. Electrospinning Conditions polystyrene concentration (wt %)
cloisite 6A concentration (wt %)
voltage (kV)
working distance (cm)
flow rate (µL/min)
20 15 10 7.5 5
0, 1, 4, 8 0, 1, 4, 8 0, 1, 4, 8 0, 1, 4, 8 0, 1, 4, 8
10 10 10 10 10
10 10 10 10 10
50 50 20 20 20
away from the tip of the needle as the collector. The spinning conditions of each solution were listed in Table 2. Electrospun fibers were collected on cleaned aluminum foils or silicon wafers as randomly oriented fiber mats on the aluminum target and then annealed in a vacuum of 10-3 Torr at 65 °C for 24 h to remove the residual solvent. Characterization. The morphology of electrospun PS/clay composite fibers was characterized using SEM (LEO1550, LEO, Germany) at 10 kV acceleration voltage and 8 mm working distance. All fiber samples were sputter-coated with gold to improve the surface conductivity. At least 100 fibers were picked randomly from each SEM picture to calculate the average diameter of electrospun fibers using Image Tool (The University of Texas Health Science Center in San Antonio). Surface topography images of electrospun fibers were obtained using a Veeco/DI Dimension 3000 SPM in the contact mode with a Si3N4 tip. A TEM (JEM1200ex, JEOL, Inc.) with a LaB6 filament at 120 kV was employed to investigate the interior structure of electrospun fibers. TEM samples were obtained by directly electrospinning the fibers onto a copper sample grid. SMFM. The principle of the SMFM method was described in an earlier paper,20 and the experimental set up is shown in Figure 2. Glass transition temperature (Tg) and relative surface modulus measurements were made using a Veeco/DI Dimension 3000 AFM located in a sealed glovebox, which was purged with dry nitrogen. We used the same etched silicon cantilever (Digital Instrument) with a spring constant of ∼0.1 N/m for all measurements. A sinusoidal drive signal with a frequency of 1400 Hz was applied to the x-piezo controlling the cantilever, inducing a small oscillatory motion of the tip parallel to the sample surface. A drive frequency of 1400 Hz was chosen for the measurements because it lies in the flattest region of the response curve of the cantilever. We used a drive signal amplitude of 15 mV, which corresponds to an x-piezo displacement of 3 nm. A normal load force of 25 nN was applied to maintain tip contact with the sample. The lateral deflection of the cantilever was detected through a position-sensitive diode and was fed into a dual-phase lock-in amplifier set at the tip modulation frequency. The electrospun fiber samples were collected on cleaned silicon wafers (300 µm thick, (100) orientation, Wafer World, Inc.) and mounted on a heating/ cooling stage (MMR Technology), which was calibrated by determining the melting point of naphthalene (353.3 K) and indium (429.7 K) crystals. As a bulk control sample, polystyrene was also dissolved in toluene with a concentration of 130 mg/mL and spun cast at 2500 rpm onto HF-etched silicon wafers for Tg and relative surface modulus measurements. The film thickness measured using
Electrospun PS/Clay Nanocomposite Fibers
Langmuir, Vol. 22, No. 3, 2006 1323
Figure 2. Principle of the shear modulation force microscopy.
Figure 4. Average fiber diameter of electrospun fibers as a function of the solution concentration with different clay contents.
Figure 3. SEM images of electrospun PS fibers with 4 wt % cloisite 6A at different PS concentrations: (a) 20 wt %, (b) 15 wt %, (c) 10 wt %, (d) 7.5 wt %, and (e) 5 wt %. an ellipsometer was 1 µm. Samples were then annealed in a vacuum of 10-3 Torr at 120 °C for 24 h to remove the residual solvent. The shear modulus and phase angle of bulk PS samples at different temperatures were measured using a rheometric mechanical spectrometer (RMS-605E, Rheometrics, Inc., NJ) in parallel-plate geometry.
Results and Discussion Fabrication of Electrospun PS/Clay Fibers with Controllable Diameter. Electrospinning is a complex process, and the morphology of electrospun fibers can be affected by either the intrinsic properties of the solution such as the concentration, surface tension, and conductivity or the processing parameters such as the voltage, stock solution feed rate, and working distance.21 To study the surface nanomechanical properties of electrospun fibers as a function of the fiber diameter, we need to control the fiber diameter distribution by changing the parameters mentioned above. Deitzel et al.22 evaluated systematically the effects of parameters controlling the electrospinning process and found that the solution concentration most strongly (21) Megelski, S.; Stephens, J. S.; Chase, D. B.; Rabolt, J. F. Macromolecules 2002, 35, 8456-8466. (22) Deitzel, J. M.; Kleinmeyer, J.; Harris, D.; Tan, B. Polymer 2001, 42, 261-272.
affects the fiber size. The fiber diameter increases with an increasing solution concentration according to a power law relationship. Therefore, in our study, we fixed other parameters and chose the solution concentration as the major variable controlling the fiber diameter. The electrospinning conditions are shown in Table 2. As the solution concentration decreased, the average fiber diameter decreased, while the bead density increased dramatically and the electrospun PS fiber showed a beads-on-string morphology. Fong et al.23 found that the formation of beads is strongly influenced by the viscoelasticity of the solution, the charge density carried by the jet, and the surface tension of the solution. Higher viscosity and higher net charge density favor the formation of fibers without beads, while higher surface tension increases the bead density. To decrease the bead density, which normally occurs at a low solution concentration, we chose a mixed solvent system of THF and DMF with a volume ratio of 50:50, which decreased the diameter of electrospun fibers at low solution concentrations while still keeping the uniform shape of fibers.24 As can be seen from the solvent properties in Table 1, THF has a lower surface tension and boiling temperature compared to DMF. Therefore, the addition of THF into DMF decreases the surface tension of the mixed solvent while enhances the volatility because of the low boiling temperature of THF, which makes the solvent easier to evaporate. Both of these effects favor the formation of smooth fibers without beads. Figure 3 shows the SEM images of PS fibers with 4 wt % cloisite 6A at different polystyrene concentrations. Using Image Tool, the average fiber diameters at each solution concentration were calculated and plotted in Figure 4. We found that the average (23) Fong, H.; Chun, I.; Reneker, D. H. Polymer 1999, 40, 4585-4592. (24) Lee, K. H.; Kim, H. Y.; La, Y. M.; Lee, D. R.; Sung, N. H. J. Polym. Sci., Part B: Polym. Phys. 2002, 40, 2259-2268.
1324 Langmuir, Vol. 22, No. 3, 2006
Ji et al.
Figure 5. AFM images of surface topography and corresponding cross-sectional profiles of 15 wt % electrospun PS fibers at different clay concentrations (1 × 1 µm): (a) without clay, (c) with 1 wt % clay, (e) with 4 wt % clay, and (g) with 8 wt % clay.
fiber diameter decreased from 4 µm to 150 nm as the solution concentration decreased from 20 to 5 wt %, and there were no beads that occurred in electrospun fibers even when the solution concentration was lower than 10 wt %. The absence of beads even at a fairly low solution concentration results from the utility of the mixed solvent mentioned above and the existence of clay inside the solution, which may increase the solution viscosity and favor the formation of smooth fibers. Surface Morphology of Electrospun PS/Clay Fibers. The surface morphology of electrospun fibers was investigated using AFM, which is shown in Figure 5. A typically nanoporous morphology, which arises from evaporative cooling of the polymer solution as the polymer jet travels the distance from the spinneret to the metal target was observed on the surface of electrospun
PS fibers.21,25-26 As can be seen from Figure 5a, all nanopores were stretched along the fiber axis direction to form an elliptical shape instead of a spherical shape, which is due to the high shear flow during electrospinning. A corresponding cross-sectional profile in Figure 5b shows that the nanopores have an average diameter of 110 nm and an average depth of 50 nm. As the clay concentration increased to 1 wt %, the nanopores coalesced with each other and were still stretched along the fiber axis direction to form a wrinkle-like morphology. When the clay concentration kept increasing from 4 to 8 wt %, the surface of electrospun (25) Casper, C. L.; Stephens, J. S.; Tassi, N. G.; Chase, D. B.; Rabolt, J. F. Macromolecules 2004, 37, 573-578. (26) Bognitzki, M.; Czado, W.; Frese, T.; Schaper, A.; Hellwig, M.; Steinhart, M.; Greiner, A.; Wendorff, J. H. AdV. Mater. 2002, 13, 70-72.
Electrospun PS/Clay Nanocomposite Fibers
Langmuir, Vol. 22, No. 3, 2006 1325
fibers became rougher and formed a ridge-like morphology with an average depth of 150 nm. Surface Nanomechanical Property. The principle of SMFM was discussed in some early papers,20,27-28 and the tip-surface interaction is shown in Figure 2. In our measurements, a small fixed sinusoidal drive signal, ∆Xp, was applied to the x-piezo in a direction perpendicular to the fast scan direction, inducing a small oscillatory motion of the cantilever tip parallel to the sample surface. At the same time, a normal force was applied to maintain tip contact with the sample. The lateral deflection amplitude of the cantilever, ∆X, which is proportional to the torsion of the cantilever ∆Xlever, can be measured as a function of the temperature using a lock-in amplifier, tuned to the drive frequency. A full theory of operation would include an account of the elastic behavior of the tip and cantilever, tip-polymer, and polymer viscoelastic response. Such a theory is not currently available, and the precise experimental characterization of the tip and tip-polymer interface is difficult and incomplete. However, we may make some rough estimates based on simple elastic and viscous models. The indentation of an elastic substrate by a hard spherical tip of radius R, h, is given in the Hertz model by29
[ ]
L h ) D 1/2 R
2/3
Figure 6. Lateral deflection (∆X) of the PS bulk thin-film sample as a function of the drive amplitude at different temperatures.
(1)
Because
D)
3 1 - ν2 4 E
(2)
E ) 2(1 + ν)G
(3)
Figure 7. Log ∆X versus log G at different temperatures. The solid line is a linear regression.
∆X ∼ G-0.66
We have
h)
(
)
31-ν L 8 G R1/2
2/3
(4)
where R is the tip radius, L is the load applied to the tip, E and G are Young’s modulus and shear modulus, respectively, and ν is the Poisson’s ratio. From eq 4, we know that h is proportional to G-2/3. However, at this point, we can only conclude qualitatively that the increase in the tip penetration, h, as the polymer softens leads to a larger lateral deflection, ∆X. The lateral deflection, ∆X, plotted as a function of the drive amplitude at different temperatures on the surface of a spun-cast PS bulk thin-film sample, is shown in Figure 6. We found that ∆X was proportional to the drive amplitude within the temperature range from 295 to 410 K. The lateral deflection ∆X increased with an increasing temperature, which indicates that the surface modulus decreases. To compare our results with the bulk measurements data, we need to find the quantitative relationship between measured bulk modulus (G) and their surface mechanical response (∆X). The slope of the surface mechanical response curve in Figure 6 as a function of the measured bulk shear modulus at three different temperatures was plotted in Figure 7. The relationship is linear with a fitted slope of -0.66, and then for the electrospun PS/clay fiber, we have (27) Pu, Y.; Ge, S. R.; Rafailovich, M. H.; Sokolov, J. S. Langumir 2001, 17, 5865-5871. (28) Zhang, Y.; Ge, S.; Rafailovich, M. H.; Sokolov, J. C.; Colby, R. H. Polymer 2003, 44, 3327-3332. (29) Johnson, K. L. Contact Mechanics; Cambridge University Press: Cambridge, MA, 1985.
(5)
Therefore, we have
( )
∆X0 Gt ∼ G0 ∆Xt
1.51
(6)
where ∆X0 is the measured lateral deflection and G0 is the measured shear modulus of the PS bulk thin-film sample. Using the scaling relationship, we can apply the SMFM method to study the relative surface modulus as a function of the fiber diameter and temperature for electrospun PS/clay fibers where standard bulk measurements are not possible. The lateral deflection ∆X and corresponding relative surface modulus value of electrospun PS fibers as a function of the fiber diameter at 295 K were plotted in Figure 8. As the fiber diameter decreased, ∆X values decreased, which indicates that the surface modulus increases with a decreasing fiber diameter. At a fiber diameter of 4 µm, the relative surface modulus was only 0.22 times larger than the bulk thin-film value. When the fiber diameter decreased from 4 µm to 410 nm, the relative surface modulus increased to 3.7 times larger than the bulk value. We then measured the change of the relative surface modulus as a function of the temperature at different fiber diameters and plotted them in Figure 9. ∆X values increased with an increasing temperature, which indicates that the fiber surface become softer. Meanwhile, the corresponding relative surface modulus values of all fibers dropped, while fibers with small diameters still maintained higher relative modulus values than those with large diameters. As the temperature increased above 375 K, the relative surface modulus values of all fibers approached to 1, which
1326 Langmuir, Vol. 22, No. 3, 2006
Figure 8. Lateral deflection ∆X and the corresponding relative modulus G of electrospun PS fibers as a function of the fiber diameter at 295 K.
Figure 9. Lateral deflection ∆X and relative modulus G of electrospun PS fibers as a function of the temperature at different fiber diameters.
corresponds to the bulk thin-film value. Tg measurements using SMFM showed that Tg was around 377 K and did not change up to a fiber diameter of 620 nm, as can be seen in Figure 10. The large increase of relative surface modulus values as the fiber diameter decreases may be explained by the shear-induced molecular chain alignment during electrospinning. This type of molecular chain alignment was first reported by Vancso et al.30 in electrospun PEO fibers, as was explained by the large shear forces during electrospinning. In our experiments, a high voltage of 10 kV was applied between the tip of the needle and the metal target, which generated an electrical field of around 105 V/m at a distance of 10 cm. A large shear force was created during (30) Jaeger, R.; Schonherr, H.; Vancso, G. J. Macromolecules 1996, 29, 76347636.
Ji et al.
Figure 10. Surface Tg of electrospun PS fibers at different diameters.
electrospinning, which stretched out the pendant polymer droplet on the tip and formed a stable jet. The large shear force, coupled with the rapid solidification of the thin polymer jet, prevented the molecular chains from relaxing back to their equilibrium conformations and resulted in the molecular chain alignment along the fiber axis direction.31-33 Because fibers with small diameters received a higher stress than those with large diameters, the degree of molecular chain orientation in electrospun fibers with small diameters was higher than fibers with large diameters, which explains the relative surface modulus difference between them. As the temperature increased above Tg, the oriented molecular chains relaxed to their equilibrium state and relative modulus values were restored to the bulk value. Further experiments using a polarized extended X-ray absorption fine structure (EXAFS) method are currently in progress, which will confirm this model by measuring the degree of orientation relative to the fiber axis direction. Clay Effects on the Surface Nanomechanical Property. The effects of clay on the surface nanomechanical properties of electrospun PS fibers were also investigated using the SMFM technique. The lateral deflection ∆X and the corresponding relative surface modulus value as a function of the fiber diameter with different clay concentrations at 295 K were plotted in Figure 11. We found that relative surface modulus values increased with a decreasing fiber diameter, which is in good agreement with the data of fibers without clay. At a fiber diameter of 400 nm, the relative surface modulus values increased from 3.5 to 4.5 when the clay concentration increased from 1 to 8 wt %. This value is larger than the value of electrospun pure PS fibers with the same diameter, which indicates that clay has a reinforcement effect on the fibers. As the fiber diameter increased from 400 nm to 4 µm, relative surface modulus values of all fibers decreased, while fibers with 4 wt % clay still maintained a larger relative modulus value than those with 1 and 8 wt % clay. TEM pictures in Figure 12 show the clay distribution inside electrospun fibers. Clay was wellexfoliated inside the fibers at a 4 wt % concentration, as can be seen in Figure 12b. The dark lines, with a thickness of around 1 nm, represent the montmorillonite layers, which were aligned along the fiber axis as the arrow shows. The same structure was also reported in electrospun PA/clay nanocomposite fibers.19 As the clay concentration increased to 8 wt %, they formed agglomerates inside the fibers, which can be seen in Figure 12c. This may be due to the inhomogeneous distribution of clay within (31) Stephens, J. S.; Chase, D. B.; Rabolt, J. F. Macromolecules 2004, 37, 877-881. (32) Pedicini, A.; Farris, R. J. Polymer 2003, 44, 6857-6862. (33) Fennessey, S. F.; Farris, R. J. Polymer 2004, 45, 4217-4225.
Electrospun PS/Clay Nanocomposite Fibers
Langmuir, Vol. 22, No. 3, 2006 1327
Figure 11. Lateral deflection ∆X and corresponding relative modulus values of electrospun PS/clay fibers as a function of the diameter at 295 K.
the spinning solution at high concentrations. The inhomogeneous distribution of clay explains the difference of relative modulus values at different clay concentrations. Clay has a better reinforcement effect at a concentration of 4 wt % because of the well-exfoliated montmorillonite layer structure inside the fiber; therefore, the relative modulus values of electrospun fibers at 4 wt % clay were larger than the values of fibers at other clay concentrations especially when the fiber diameter was larger than 1 µm. Figure 13 shows the surface glass transition temperature measurements as a function of fiber diameter at a 4 wt % clay concentration. A drastic increase of Tg, nearly 20 K at a fiber diameter of 700 nm, was observed in our measurement. The large increase of Tg can be also explained as a result of the well-exfoliated clay layer structure inside the electrospun fibers, which greatly enhanced the thermal property of fibers. Relative modulus measurements as a function of the temperature show that relative modulus values of all fibers at different clay concentrations were restored to the bulk value as we increased the temperature above Tg (data not shown here). Therefore, the increase of shear modulus values in electrospun PS/clay fibers below the glass transition temperature could be a combination effect of molecular chain orientation and clay exfoliation. The molecular chain orientation is the leading factor, while clay reinforcement is minor because of the inhomogeneous distribution of clay inside electrospun fibers.
Conclusions Electrospun PS/clay nanocomposite fibers were successfully fabricated with diameters ranging from 4 µm to 150 nm. The surface morphology of electrospun fibers changed from a typical nanoporous structure (without clay) to a ridge-like structure (with
Figure 12. TEM images of clay distribution inside electrospun PS fibers: (a) 1 wt % cloisite 6A, (b) 4 wt % cloisite 6A, and (c) 8 wt % cloisite 6A.
Figure 13. Tg as a function of the fiber diameter at a 4 wt % clay concentration.
clay). Surface relative modulus measurements using SMFM showed a large increase of shear modulus below the glass transition temperature, suggesting molecular chain orientation within the electrospun fibers. As the temperature increased above Tg, the oriented molecular chains relaxed to their equilibrium
1328 Langmuir, Vol. 22, No. 3, 2006
state and the relative modulus of all fibers was restored to the bulk value. A highly aligned montmorillonite layer structure inside the electrospun fibers was observed at a clay concentration of 4 wt %, which results from the high shear flow during electrospinning. The increase of the shear modulus in electrospun PS/clay fibers could be explained as a combination effect of molecular chain orientation and clay exfoliation. This effect may help fabricate high-strength and high-modulus electrospun
Ji et al.
polymer nanocomposite fibers utilized in the automotive and aerospace industry. Acknowledgment. Support from the NSF-MRSEC program is gratefully acknowledged. The authors also thank Dr. Jim Quinn and Mr. Gregory Rodumen for their expertise in SEM and TEM. LA0525022
