NANO LETTERS
Collecting Electrospun Nanofibers with Patterned Electrodes
2005 Vol. 5, No. 5 913-916
Dan Li,† Gong Ouyang,‡ Jesse T. McCann,† and Younan Xia*,† Department of Chemistry and Department of Electrical Engineering, UniVersity of Washington, Seattle, Washington 98195 Received March 4, 2005; Revised Manuscript Received March 29, 2005
ABSTRACT Electrospinning is a simple, versatile, and useful technique for fabricating nanofibers from a rich variety of functional materials. The nanofibers are usually collected as nonwoven mats, in which the fibers are randomly oriented. We have recently demonstrated that the nanofibers can be uniaxially aligned by introducing an insulating gap into the conductive collector. To elucidate the mechanism of alignment, we have systematically studied the effect of the area and geometric shape of the insulating gap on the deposition of fibers. By modeling the electrostatic forces acting on the fiber, it was established that the fibers tended to be oriented along a direction such that the net torque of electrostatic forces applied to the two ends of a discrete segment of the fiber were minimized. By varying the design of electrode pattern, it was possible to control both alignment and assembly of the electrospun nanofibers.
Electrospinning has been extensively explored for fabricating long nanofibers.1,2 It has been demonstrated that a variety of materials such as polymers, ceramics, carbons, and composites can all be electrospun as uniform fibers. In addition to solid fibers, this technique is also capable of generating fibers with a core/shell or hollow structure.3 The as-spun fibers are usually collected as nonwoven mats, which have already found use in applications that include reinforcement of composite materials, ultrafiltration, tissue engineering, catalysis, as well as fabrication of solar cells, sensors, batteries, and many other types of devices.1,4 Most recently, several groups further demonstrated that electrospun nanofibers could be collected as uniaxially aligned arrays by using specially designed collectors.5,6 For example, we have shown that electrospun nanofibers could be readily prepared as wellaligned arrays by using a collector consisting of two conductive electrodes separated by an air gap.6a When the paired electrodes were patterned on a highly insulating substrate, it was also possible to collect electrospun fibers as uniaxially aligned arrays across the insulating region.6b However, it is not clear how the electrospun fibers will behave when the geometric shape and feature size of the insulating region are varied. Here we present a systematic study on this matter. It is clear that patterning of electrodes with complex designs allows one to collect and assemble nanofibers into well-controlled architectures. Controlling the orientation of nanofibers (as well as their stacking into hierarchical structures) is important for many applications, * Corresponding author. E-mail:
[email protected]. † Department of Chemistry. ‡ Department of Electrical Engineering. 10.1021/nl0504235 CCC: $30.25 Published on Web 04/09/2005
© 2005 American Chemical Society
including fiber reinforcement and fabrication of nanoscale fluidic, electronic, and photonic devices. The electrospinning setup for our experiments is similar to the one we have used previously.6 Poly(vinyl pyrrolidone) (PVP, Aldrich, Mw ≈ 1 300 000) was used to demonstrate this concept. In a typical procedure, PVP was dissolved in a mixture of ethanol and water (8:1.5 by volume) to prepare a 6 wt % solution, which was then loaded to a plastic syringe equipped with a stainless steel needle. The needle was connected to a high-voltage power supply (ES30P-5W, Gamma High Voltage Research Inc., Ormond Beach, FL). The solution was continuously supplied using a syringe pump at a rate of 0.2 mL/h. The voltage used for electrospinning was 6 kV, and the collection distance was 9 cm. A typical collection time was 20 s. Quartz wafers with a diameter of 75 mm were used as the collecting substrates. The spinning process was conducted in a lab-made glove box, in which the relative humidity was controlled at the level of ∼20%. Gold electrodes of different designs and feature sizes were patterned in the centers of these quartz wafers by thermal evaporation through physical masks. Optical micrographs were recorded using a Leica microscope equipped with a digital camera. First we examined how the PVP nanofibers were aligned on the insulating (quartz) regions. Figure 1 shows four typical electrode patterns and the corresponding optical images taken in the transmission mode. In this mode, only the fibers deposited on the quartz regions were visible. It was found that the orientation of the fibers is strongly dependent on the design of the electrode. For the circular hole pattern (Figure 1A), the fibers were randomly oriented. For the
Figure 1. Transmission optical micrographs showing gold electrodes (dark area) of various shapes patterned on insulating, quartz substrates. Only fibers deposited on the insulating regions (bright areas) are visible. Note that only a half of the rectangle is shown in (D). The fibers on the other part of the electrode have a similar configuration.
triangular and square hole patterns (Figures 1B and 1C), the fibers tended to be concentrated at the vertices and appeared to be aligned with their long axes perpendicular to the bisector of the vertex. Only a few fibers were collected in the middle of the insulating region. In the case of a rectangular hole pattern (Figure 1D), the orientation of fibers collected near the vertices is similar to the triangular and square patterns, while the fibers deposited outside of the diagonal corners were uniaxially aligned with their long axis perpendicular to the longer sides of the rectangle. This result is consistent with the observation when a pair of electrodes separated by an air gap or a stripe of insulator was employed as the collector.6 We then evaluated electrodes patterned with insulating regions of different areas. We took images in the dark-field mode to reveal the nanofibers deposited on both the quartz and gold regions. As shown in Figure 2, unlike the fibers collected in the middle of a conductive substrate where they are randomly oriented, the fibers collected at the edge of an electrode appear to have preferred orientations that are similar to those of the fibers nearest to them on the insulating areas. Their configurations are dependent on both the shape and area of the insulating regions. As shown in Figure 2A-F, the fibers collected at the edge of an electrode with a larger insulating area are straighter than those on an electrode with a smaller insulating area. Furthermore, the degree of orientation became greater as the area of the insulating region was increased. As shown in Figure 2B, where the diameter (Φ) of the insulating circle was 8 mm, the fibers at the edges were well-aligned along the circumference of the circle. In contrast, the fibers collected at the edge of a smaller circle (Figure 2A, Φ ) 0.8 mm) were highly disordered. Fewer fibers appeared in the center of the insulating area as the area was increased. We have also tested patterned electrodes for PVP nanofibers with different diameters (e.g., by varying electrospinning 914
Figure 2. Dark-field optical micrographs showing the orientation of fibers deposited on the edges of gold electrodes with different insulating areas. The insets are images taken in the transmission mode, showing the areas from which the dark-field images were captured. The sizes of these insulating areas are: (A) Φ ) 0.8 mm; (B) Φ ) 8 mm; (C) L ) 0.8 mm; (D) L ) 8 mm; and (E, F) L1 ) 0. 8 mm, L2 ) 6 mm. Here Φ and L represent the diameter of the circle and the side length of the triangle or rectangle. The orientation of fibers collected on the edges of a square electrode is similar to the case of rectangular electrode, therefore not shown in this figure.
parameters such as voltage or injection rate) as well as other organic polymers such as polystyrene, poly(ethylene oxide), and polyacrylonitrile. It was found that the orientation of fibers on both the insulating area and the edge of conductor were orientated in the same fashion, indicating that this phenomenon is independent of the fiber composition and diameter. Likewise, the size and morphology of the nanofibers did not appear to be affected by the geometry of a patterned electrode. When the design of a patterned electrode was fixed, the major factor that influenced the degree of orientation came from the humidity of air in which the electrospinning experiments were carried out. Since water in the air can help discharge as-spun fibers, the collected fibers became less ordered with increasing the relative humidity due to the reduction in charge density on the fibers. In addition to humidity, the collection time also has some effect on the orientation of the fibers. We have previously demonstrated that the deposited nanofibers retained some charges, which could improve the orientation through electrostatic repulsions between the fibers.6b These results indicate that the density of charges on electrospun nanofibers plays the most important role in controlling their alignment and assembly. Electrospinning involves the use of high voltage to produce a jet through electrostatic forces. In an electrospinning Nano Lett., Vol. 5, No. 5, 2005
process, both the initial jet and as-spun fibers are highly charged. The initial direction of a spinning jet is randomized due to the bending instability of the charged jet. After the fiber has been ejected, its motion is mainly controlled by electrostatic forces exerted by the strong external electric field. In addition, the charges on the fiber induce opposite charges on the surface of the collecting electrode, which also help attract the fibers to the electrode. When a continuous, conductive plate is used as the collector, the two types of electrostatic forces acting on the fiber have no preferential direction in the plane of the collector. As we have demonstrated previously, if an insulating gap is introduced into the collector, it will change the structure of the electric field, causing the fibers to align along a preferred direction across the gap. When an insulating gap with a specific shape is introduced into a conductive plate (as shown in Figure 1), the distribution of external electric field becomes much more complicated. To understand the mechanism behind the alignment of electrospun fibers on these patterned electrodes, we calculated the electrostatic forces acting on a charged nanofiber using the classic method of moments (MoM).7 Both the external electric field and the electrostatic attraction from the collecting electrode were considered in our calculations. The charge distribution of the collecting electrode and the spinneret was obtained by discretizing them into tiny triangles (as shown in Figure 3A), and pulse basis functions were used to model the surface charge densities on each triangle. The section of fiber was discretized into smaller segments and a uniform linear charge density was assumed. The voltage applied to the needle and the electrode served as the boundary conditions. The free-space Green’s function was used as the charge interaction kernel. The final formulation is in the form of Pq ) v, where P is the interaction matrix which results from testing procedure on each triangle surface and each segment of fiber, q is the unknown charge density vector, and v is a known voltage vector which is composed of voltages at the geometric centers of the triangles and the segments. Once the charge densities had been derived, the force on each segment of the fiber and the net torque on the fiber could be computed. Figures 3B and 3C illustrate the relative values and direction of electrostatic forces that a fiber experiences when it moves toward the corner of a triangular electrode with different initial directions. If its long axis is perpendicular to the bisector of the vertex, the forces acting on the entire fiber are uniform (Figure 3C) and the fiber will move to the vertex and be collected without any adjustment to its orientation. However, if the moving fiber is not perpendicular to the bisector of the vertex (as shown in Figure 3B), the forces acting on the two ends of the fiber are unequal, thus the two ends of the segment accelerate at different rates. As a result, the moving rates of the ends are continuously adjusted until the long axis of the segment is perpendicular to the bisector of the vertex, and thus the forces acting on the two ends of the segment are equal. Performing this calculation at other positions and using different electrode shapes demonstrates that the fibers tend to be oriented along Nano Lett., Vol. 5, No. 5, 2005
Figure 3. (A) Schematic illustration showing discretization of the electrode and the spinneret for the analysis of electrostatic forces acting on an electrospun nanofiber. (B, C) Electrostatic force analysis of a segment of the charged fiber. When the fiber segment is not perpendicular to the bisector of the corner, the forces on the segment are unequal and tend to act on the fiber such that it will orient itself perpendicular to the bisector of the corner, as shown in (B). When the fiber is perpendicular to the bisector of the corner, the forces acting along the segment of the fiber are the same and the orientation is preserved, as shown in (C).
a direction such that the net torque of electrostatic forces applied to the two ends of a discrete segment of the fiber are minimized. The electrostatic forces acting on a fiber close to the insulating area can be resolved into two components: one in the plane of the electrodes and the other perpendicular to the plane. The two components affect the orientation of the fiber in different ways. Our calculations indicate that the component perpendicular to the plane is inversely proportional to the area of the insulating region. When a collector with a larger insulating region is employed, the fiber will experience a stronger component in the electrode plane, adjusting their orientation to the preferred direction. In addition, a decrease in the component perpendicular to the plane implies that the fiber will have more time to remain in motion before it contacts the electrode. Therefore there is more time for these forces to act on the fiber and adjust its orientation to a lower-energy state. These results are in agreement with our experimental observations. In principle, if the nanofiber could move toward the electrode at a slow speed, it would be perfectly aligned in the preferred lowerenergy direction. 915
interactions, it may also be well-suited for the manipulation of charged fibers generated using other techniques. Acknowledgment. This work has been supported in part by an AFOSR-MURI grant on smart skin materials awarded to the University of Washington and a research fellowship from the David and Lucile Packard Foundation. Y.X. is a Camille Dreyfus Teacher Scholar (2002) and an Alfred P. Sloan Research Fellow (2000). J.M. thanks the MLSC program at the UW for a student fellowship award. The authors thank Dr. Yuliang Wang for fabricating the electrode patterns. References
Figure 4. Dark-field optical micrographs showing electrodes with complex designs (A, C) and the behaviors of fibers deposited on the conductive stripes (B, D). Note that the fibers in (B) are mostly aligned parallel to the conducting strip. The fibers in (D) behave similarly, though they fan out at the opening of the gap.
Both the experimental and theoretical work clearly indicate that the introduction of an insulating region into a conductive collector will influence the electrostatic forces acting on a charged fiber. With the aid of electrostatic interactions, electrospun nanofibers can be assembled into controllable structures with different configurations by simply changing the design of the collector. Figure 4 shows two additional examples. When an electrode containing two square, insulating regions separated by a thin conductive strip (Figure 4A) was used, most of the fibers collected on the strip were wellaligned along the strip (Figure 4B). This experiment implies that uniaxially aligned fibers can also be directly collected on a conducting surface by controlling the configuration of the electrode. When an annular electrode shown in Figure 4C was used, the fibers in the annular insulating area were oriented with their long axes along the radial direction. It is believed that more complex structures could be achieved by judiciously designing the test pattern for the collector. Assembling one-dimensional nanostructures into ordered architectures has received growing interest in recent years, with the hope that these ordered nanoscale arrays will serve as building blocks to fabricate complex devices and systems. The use of patterned electrodes with different designs and/ or feature sizes as the collectors for electrospun nanofibers provides a convenient approach to the manipulation of nanofibers. One of the most interesting features associated with this approach is that this technique enables direct integration of nanofibers with controllable configurations into an electrode system such that the nanofibers can be fabricated and aligned simultaneously, which will significantly simplify the production of nanofiber-based devices. In addition, since the alignment method used here is based on electrostatic
916
(1) Recent reviews: (a) Reneker, D. H.; Chun, I. Nanotechnology 1996, 7, 216. (b) Li, D.; Xia, Y. AdV. Mater. 2004, 16, 1151. (2) See, for example, (a) Srinivasan, G.; Reneker, D. H. Polym. Int. 1995, 36, 195. (b) MacDiarmid, A. G.; Jones, W. E.; Norris, I. D.; Gao, J.; Johnson, A. T.; Pinto, N. J.; Hone, J.; Han, B.; Ko, F. K.; Okuzaki, H.; Llaguno, M. Synth. Met. 2001, 119, 27. (c) Bognitzki, M.; Czado, W.; Frese, T.; Schaper, A.; Hellwig, M.; Steinhart, M.; Greiner, A.; Wendorff, J. H. AdV. Mater. 2001, 13, 70. (d) Larsen, G.; VelardeOrtiz, R.; Minchow, K.; Barrero, A.; Loscertales, I. G. J. Am. Chem. Soc. 2003, 125, 1154. (e) Ko, F.; Gogotsi, Y.; Ali, A.; Naguib, N.; Ye, H.; Yang, G. L.; Li, C.; Willis, P. AdV. Mater. 2003, 14, 1161. (f) Li, D.; Xia, Y. Nano Lett. 2003, 3, 555. (g) Li, D.; Herricks, T.; Xia, Y. Appl. Phys. Lett. 2003, 83, 4586. (h) Dai, H.; Gong, J.; Kim, H.; Lee, D. Nanotechnology 2002, 13, 674. (i) Chun, I.; Reneker, D. H.; Fong, H.; Fang, X.; Deitzel, J.; Tan, N. B.; Kearns, K. J. AdV. Mater. 2003, 31, 37. (3) (a) Li, D.; Xia, Y. Nano Lett. 2004, 4, 933. (b) Li, D.; Babel, A.; Jenekhe, S. A.; Xia, Y. AdV. Mater. 2004, 16, 2062. (c) Li, D.; McCann, J. T.; Xia, Y. Small 2005, 1, 83. (d) McCann, J. T.; Li, D.; Xia, Y. J. Mater. Chem. 2005, 15, 735. (e) Loscertales, I. G.; Barrero, A.; Marquez, M.; Spretz, R.; Velarde-Ortiz, R.; Larsen, G. J. Am. Chem. Soc. 2004, 126, 5376. (f) Sun, Z.; Zussman, E.; Yarin, A. L.; Wendorff, J. H.; Greiner, A. AdV. Mater. 2003, 15, 1929. (g) Yu, J. H.; Fridrikh, S. V.; Rutledge, G. C. AdV. Mater. 2004, 16, 1562. (h) Zhang, Y. Z.; Huang, Z. M.; Xu, X. J.; Lim, C. T.; Ramakrishna, S. Chem. Mater. 2004, 16, 3406. (4) (a) Pawlowski, K. J.; Belvin, H. L.; Raney, D. L.; Su, J.; Harrison, J. S.; Siochi, E. J. Polymer 2003, 44, 1309. (b) Gibson, P.; SchreuderGibson, H.; Rivin, D. Colloids Surf., A 2001, 187-188, 469. (c) Wang, X.; Drew, C.; Lee, S.-H.; Senecal, K. J.; Kumar, J.; Samuelson, L. A. Nano Lett. 2002, 2, 1273. (d) Wnek, G. E.; Carr, M. E.; Simpson, D. G.; Bowlin, G. L. Nano Lett. 2003, 3, 213. (e) Bergshoef, M. M.; Vancso, G. J. AdV. Mater. 1999, 11, 1362. (f) Li, W.-J.; Laurencin, C. T.; Caterson, E. J.; Tuan, R. S.; Ko, F. K. J. Bio. Mater. Res. 2002, 60, 613. (g) Jia, H.; Zhu, G.; Vugrinovich, B.; Kataphinan, W.; Reneker, D. H.; Wang, P. Biotechnol. Prog. 2002, 18, 1027. (h) Kim, C.; Yang, K. S. Appl. Phys. Lett. 2003, 83, 1216. (i) Choi, S. W.; Jo, S. M.; Lee, W. S.; Kim, Y.-R. AdV. Mater. 2003, 15, 2027. (5) (a) Theron, A.; Zussman, E.; Yarin, A. L. Nanotechnology 2001, 12, 384. (b) Dersch, R.; Liu, T.; Schaper, A. K.; Greiner, A.; Wendorff, J. H. J. Polym. Sci., Part A 2003, 41, 545. (c) Huang, Z. M.; Zhang, Y. Z.; Kotaki, M.; Ramakrishna, S. Compos. Sci. Technol. 2003, 63, 2223. (d) Kameoka, J.; Craighead, H. G. Appl. Phys. Lett. 2003, 83, 371. (e) Sundaray, B.; Subramanian, V.; Natarajan, T. S.; Xiang, R. Z.; Chang, C. C.; Fann, W. S. Appl. Phys. Lett. 2004, 84, 1222. (f) Katta, P.; Alessandro, M.; Ramsier, R. D.; Chase, G. G. Nano Lett. 2004, 4, 2215. (6) (a) Li, D.; Wang, Y.; Xia, Y. Nano Lett. 2003, 3, 1167. (b) Li, D.; Wang, Y.; Xia, Y. AdV. Mater. 2004, 16, 361. (7) Harrington, R. F. Field Computation by Moment Methods; Macmillan: New York, 1968.
NL0504235
Nano Lett., Vol. 5, No. 5, 2005
