- Email: [email protected]
Transition of graphene oxide-coated fiber bundles from insulator to conductor by chemical reduction
Synthetic Metals 204 (2015) 90–94
Contents lists available at ScienceDirect
Synthetic Metals journal homepage: www.elsevier.com/locate/synmet
Transition of graphene oxide-coated fiber bundles from insulator to conductor by chemical reduction Dong Han Ha a, * , Suyong Jung a , Ho-Jong Kim a , Daehee Kim a , Wan-Joong Kim b , Sam Nyung Yi c , Yongseok Jun d, Yong Ju Yun d, ** a
Center for Quantum Measurement Science, Korea Research Institute of Standards and Science, Daejeon 305-340, Republic of Korea IT Convergence Technology Research Laboratory, Electronics and Telecommunications Research Institute, Daejeon 302-350, Republic of Korea Department of Nano Semiconductor Engineering, Korea Maritime University, Busan 606-791, Republic of Korea d Department of Materials Chemistry and Engineering, Konkuk University, Seoul 143-701, Republic of Korea b c
A R T I C L E I N F O
A B S T R A C T
Article history: Received 25 February 2015 Accepted 19 March 2015 Available online 28 March 2015
We fabricate graphene oxide (GO)-coated fiber bundles via electrostatic deposition of GO sheets on polyester/cotton fiber bundles and investigate the changes in their properties by chemical reduction in a hydroiodic acid solution. The structural properties of GO sheets change rapidly upon exposure to reduction agents but remain under a high defect density regime after the reduction. The electrical resistance of the fiber bundles decreases logarithmically for the first minute of the reduction. In contrast to the structural property, a longer chemical reduction over tens of minutes deteriorates the electrical property. The effect of chemical reduction on the electrical conductance of GO-coated fiber bundles is discussed based on the competition of various factors, such as the number of conducting pathways and the carrier density, among others. ã 2015 Elsevier B.V. All rights reserved.
Keywords: Graphene oxide Polymer Chemical reduction Raman spectroscopy Electrical conductance
1. Introduction Graphene oxide (GO) has attracted significant attention as an intermediate material for the mass production of graphene because it can be easily converted into conducting reduced graphene oxide (RGO) through the post-reduction process. Chemically exfoliated mono- or few-layered GO sheets are derived by treating bulk graphite with strong oxidizing agents in aqueous solutions [1–4]. New types of textiles fabricated by coating natural or synthetic textiles with GO sheets and post-treating are expected to advance the development of flexible and stretchable smart materials for wearable electronic devices and healthcare fabrics with antibacterial activity at a low cost and on a large scale [5–11]. GO is electrically insulating because of the significant number of attached oxygenated functional groups (hydroxyl and epoxy groups, among others), which induces a highly disordered lattice structure. During the post-reduction process, which converts insulating GO into conducting p-type RGO, the structural and electrical properties of GO change [4,12–16]. Boer et al. have reported a correlation between the apparent surface roughness
* Corresponding author. Fax: +82 42 868 5953. ** Corresponding author. Fax: +82 2 444 3490. E-mail addresses: [email protected] (D.H. Ha), [email protected] (Y.J. Yun). http://dx.doi.org/10.1016/j.synthmet.2015.03.018 0379-6779/ ã 2015 Elsevier B.V. All rights reserved.
and the level of deoxygenation of GO sheets by employing scanning tunneling microscopy [17]. However, the detailed transition behavior of GO from insulator to conductor remains unknown. To date, many researchers have continued the reduction process over an extended period (>30 min) to prepare RGO with high electrical conductivity [2–6,16–19]. A longer reduction time might be an obstacle to developing a continuous and rapid manufacturing process for the mass production of RGO-based materials. It is believed that the RGO sheet is composed of nanometer-sized conducting crystalline sp2 domains interspersed with insulating oxidized regions on which oxygenated functional groups are densely attached [14,15,19], and that the charge transport mechanism through the RGO sheet generally follows the variable range hopping (VRH) model [19–21]. A detailed understanding of the structural and electrical properties of RGO-coated textiles is required to develop flexible and stretchable smart materials suitable for individual requirements. In this study, we fabricate RGO-coated fiber bundles, and analyze their as-fabricated structural and electrical property changes as a function of post-chemical reduction time using micro-Raman spectroscopy and electrical transport measurements. Our measurements show that the property changes occur rapidly upon exposure to the reduction agents (hydroiodic acid). However, the electrical characteristic shows different behaviors from those in the Raman spectroscopy studies upon reduction
D.H. Ha et al. / Synthetic Metals 204 (2015) 90–94
because the resistance further increases, even after the structural variation becomes saturated. 2. Experimental We wrap commercially available polyester fiber bundles (Hana Spinning Co., polyester, 100%) with negatively charged GO sheets via electrostatic deposition using positively charged bovine serum albumin (BSA) molecules as adhesive agents in aqueous solutions [22,23]. Polyester is a flexible synthetic polymer that is extensively employed in daily life (apparel and home furnishings) and in industry (safety belts, bottles, and insulating tapes). We also obtained GO-coated cotton fiber bundles (Fujix Ltd., cotton, 100%) using the same method. The GO sheets are prepared from natural graphite by using a modified Hummers and Offeman method [1,8]. The deposition of the GO sheets on the fibers continues until the attractive forces between the GO sheets and the BSA molecules become too weak to attract additional GO sheets; thus, a few layers of GO might be deposited (Fig. 1). The charges on the GO sheets and BSA molecules depend on the preparation conditions, especially the pH value. Our samples are prepared in aqueous solutions with a pH value from 3 to 4. The GO layers are post-reduced by immersing the GO-coated fiber bundles in a solution of 2.0 ml of hydroiodic acid (57 wt% in H2O) and 5.0 ml of acetic acid (>99.7%) at 40 C. Subsequently, the samples are rinsed with a saturated sodium bicarbonate (NaHCO3) solution and then with distilled water and, finally, dried at room temperature. The reduction changes the color of the GO-coated fibers from light brown to black. Details on the fabrication of the RGO-coated fiber bundles are described elsewhere [8]. The properties of the as-fabricated RGO-coated fiber bundles are investigated using micro-Raman spectroscopy and electrical measurements. A confocal micro-Raman experiment is performed under ambient conditions in backscattering geometry. A laser line of 514.5 nm (2.41 eV) is focused on the RGO sheets attached to the fibers using a 50 objective lens, and the scattered light is analyzed using a Horiba Jobin Yvon spectrometer (LabRAM HR) equipped with a cooled charge-coupled device. The electrical conductance is analyzed by measuring the current versus voltage (I–V) relationships in the low bias-voltage regime in a vacuum using a probe station (LakeShore, CRX-4K).
91
images show that the GO coating on the fibers is inhomogeneous (Fig. 1). We collect Raman signals from the RGO sheets that are partially detached from the fibers to exclude the strong Raman signals from the underlying fibers. In cases of arbitrary points of RGO-coated polyester/cotton fibers, the Raman signals from the underlying polyester/cotton are too strong to overshadow the subtle changes in the Raman spectra of RGO. Raman spectra from the underlying fibers show a polarization dependence on the fiber direction. However, the Raman spectra of the RGO sheets show no polarization dependence. (Fig. A.1, Supplementary material). Fig. 2 shows a series of Raman spectra of RGO as a function of reduction time. The G peak at approximately 1580 cm1 is the stretching mode of the C—C bond of the sp2 structure in the rings and chains, while the D peak at approximately 1350 cm1, related to the breathing of the sp2 carbon rings, is only Raman active in the presence of point defects. In addition, the D0 peak, shown as a shoulder of the G peak at approximately 1620 cm1, is also related to defects [24–26]. A strong D peak and broad G peak compared with those of pristine graphene imply that there are many structural disorders, such as vacancies and corrugations, on the RGO sheets. Note that the intensity ratio of the D peak to the G peak (I(D)/I(G)) increases by reduction, indicating that our samples are in a highly defective state, even after the chemical reduction in a hydroiodic acid solution is completed, as explained in later.
3. Results and discussion RGO-coated fiber bundles are composed of hundreds of individual microfibers with diameters of approximately 10 mm. The optical microscope and scanning electron microscope (SEM)
Fig. 2. Change of the Raman spectrum of the RGO sheets deposited on polyester fibers as a function of chemical reduction time. All of the Raman spectra are normalized to have the same D peak intensity. The incident laser line is 514.5 nm.
Fig. 1. (a) An optical microscope image of the RGO-coated polyester fiber bundle. The inset shows an enlarged image of an individual polyester fiber. (b) SEM image showing the RGO sheet incompletely wrapping a fiber. The dotted circle indicates the part of the RGO sheet from which the as-fabricated Raman spectrum is obtained, selected to exclude the strong effects of the underlying fiber.
92
D.H. Ha et al. / Synthetic Metals 204 (2015) 90–94
Fig. 3. Changes in (a) peak position, (b) FWHM and (c) intensity ratio of each Raman peak of the RGO sheets deposited on polyester fibers versus chemical reduction time. (d) I (D)/I(G) versus I(D0 )/I(G). The intensity is defined as the height of a peak. The letters in (d) denote the reduction time for each point.
Fig. 3 summarizes the Raman spectroscopy studies on RGO sheets deposited on polyester fibers and shows the position, full width at half maximum (FWHM), and intensity change for each Raman peak as a function of reduction time. We extract the Raman parameters after averaging the signals from three to five different sample locations. The peaks are fitted with Lorentzian functions, and the intensity is defined as the height of the peak. As shown in Fig. 3, the Raman signals change abruptly during the first minute of the reduction process, and further treatment does not cause noticeable changes. For approximately one minute, the positions of the G and D0 peaks are blue shifted by 4–5 cm1, but the D peak does not move. The FWHMs of all of the peaks decreased remarkably, causing the D0 peak to become distinct as a shoulder of the G peak (Fig. 2). The FWHMs of the D peak decrease by nearly half after the beginning of the reduction process, indicating that a
significant amount of the oxygenated functional groups are removed. Once the oxygenated groups are removed, the area of the pristine sp2-bonded planar structure expands, and the structural homogeneity of RGO improves, which results in a narrowing of the G peak. Along with the position and FWHM of the G peak, the intensity ratios of the D and D0 peaks to the G peak, I(D)/I(G) and I(D0 )/I(G), are also good indicators for identifying the nature of the defects on graphene. The increase in the I(D)/I(G) ratio by chemical or thermal reduction accompanying the restoration of the pristine sp2-bonded planar carbon structure was a paradox in earlier studies on RGO [3,4,7,18]. However, studies on graphene subjected to plasma treatment or ion bombardment have shown that the I(D)/I(G) ratio increases, and the position of the G peak is blue shifted as the defect density initially increases. We named this low defect density
Fig. 4. (a) Temperature dependence of normalized conductance of the GO-coated polyester fiber bundles after reducing for 15 s and 30 min. (b) Resistance of the RGO-coated polyester fiber bundles at room temperature under vacuum as a function of the reduction time (tr). The fiber bundle for tr = 0 min (not included in the figure) shows insulating behavior. The inset is an optical microscope image of the RGO-coated polyester fiber bundles for the electrical measurement. The fiber bundles are attached to Au electrodes using 4 N indium, and the resistance is measured using the two-probe configurations between the inner electrodes.
D.H. Ha et al. / Synthetic Metals 204 (2015) 90–94
regime ‘Stage 1’. As the defect density further increases to a higher defect density regime, the ‘Stage 2’ I(D)/I(G) ratio decreases, and the G peak becomes red shifted [24–26]. Based on our data, we can claim that the RGO sheets remain in ‘Stage 2’, even after a lengthy reduction in a hydroiodic acid solution. Even after the reduction, the widths of every peak remain broad; for example, the FWHMs of the G and D peaks are approximately 50 cm1 and 70 cm1, respectively. Especially, the FWHM of the G peak remains significantly broader than that (10 cm1) of pristine graphene [27]. Although a previous study has suggested that significant restoration of the planar sp2 carbon structure could be achieved by chemical reduction [16], our results indicate that many structural defects remain on the RGO sheets, even after the chemical reduction. Fig. 3(d) shows the relationship between I(D)/I(G) and I (D0 )/I(G), which confirms that the restoration of the planar sp2 carbon structure becomes saturated under a high defect density regime after approximately one minute [24]. We observed similar Raman signals from the RGO-coated cotton fiber bundles. However, many RGO lumps are observed on the surfaces of the cotton fibers, which is consistent with the previous reports [5,9], while the surfaces of the polyester fibers are relatively smoother in the optical microscope images (Fig. A.2, Supplementary material). Our results confirm that the natural and synthetic fibers are easily coated with GO sheets in aqueous solutions, and the chemical reduction of GO in a hydroiodic acid solution becomes saturated under a regime of high defect density. Fig. 4 shows an electrical property change in the RGO-coated polyester fiber bundles versus the reduction time (tr). The starting GO-coated bundle without reduction, not shown in Fig. 4, is insulating. All of the other samples with chemical reduction show ohmic I–V characteristics between VDC = 0.1 V, indicating that conducting pathways are formed by the reduction (Fig. A.3, Supplementary material). The electrical resistance of the RGOcoated fiber bundles shows good homogeneity in which each part of a long RGO-coated fiber bundle has approximately the same resistance per unit length. For both fiber bundles reduced for 15 s and 30 min, the normalized conductance shows similar behavior to that of a multi-layered GO flake annealed at 250 C; linear dependence on the temperature and extrapolation to zero value at zero temperature (Fig. 4a) [21]. It is interesting to note that the electrical resistance shows different behaviors with the reduction time compared with those of the Raman spectroscopy studies. As the reduction time increases, the resistance decreases at a logarithmic rate for the first minute and does not show a noticeable change until it starts increasing for tr 20 min (Fig. 4b). In comparison, the Raman spectrum does not show a noticeable change for tr 1 min, suggesting that the degree of structural disorder is no longer a decisive factor in the resistance change for tr 20 min. The conductance of the RGO-coated fiber bundle increases by more than three orders of magnitude as the reduction process begins (if we assume the conductance increases uniformly for the first 30 s, then the total conductance increases by five orders of magnitude). We extract the conductivity of the RGO sheet by assuming that the current flows through the surfaces of the RGO-coated layers on the polyester microfibers with a 10 mm diameter. The measured conductivity, on the order of 101 S m1 for 1 min tr 20 min, is significantly smaller than that of graphite (105 S m1) [28], indicating that our RGO sheet still remains in ‘Stage 2’, even after the chemical reduction in a hydroiodic acid solution is complete. The RGO sheet consists of nanometer-sized conducting crystalline sp2 domains interspersed with insulating oxidized disordered regions [14,15,19]. The initial GO-coated sample is insulating because a highly disordered region covers the majority of the GO sheets so that no conducting pathway is formed between
93
the source and drain electrode. As the oxygenated functional groups are removed, the area of each crystalline domain increases; in addition, new crystalline domains are formed, which result in the opening of conducting pathways. We attribute the rapid logarithmic decrease in resistance over the first minute to the increase in the number of conducting pathways. The percolation theory predicts that the electrical conductivity s is related to the number of conducting pathways by the power law s ðpÞ / ðp pc Þa near the percolation threshold, where p is the coverage, pc is the critical coverage at which conduction begins, and a depends on the dimensions of the system [22,29,30]. Fig. 4b indicates that most of the conducting pathways between the electrodes are formed for approximately one minute, which is supported by the Raman spectroscopy results. Although the reducing agents remove oxygenated functional groups during the reduction process, they might leave iodine ions (I) on the RGO sheets, which act as scattering centers for the current flow. Moreover, electrons could transfer from the iodine ions to the p-type RGO, causing a decrease in the RGO hole density [18,31]. Therefore, we must consider the conducting pathways, carrier density and mobility, among other factors, to understand the change in the RGO resistance during the chemical reduction process. The number of conducting pathways has the dominant effect on the resistance until the formation of most of the conducting pathways at approximately one minute. Further removal of the remaining oxygenated groups is not active for the next 1 tr 20 min, as summarized in Fig. 3. Therefore, we expect only a small increase in the number of conducting pathways, which is considered to be balanced against the decrease in the charge density and mobility caused by the adsorption of iodine ions on the RGO sheets for 1 tr 20 min. The charge density and mobility become the decisive factors for the resistance for tr 20 min because we no longer expect the formation of additional conducting pathways. Shateri-Khalilabad et al. have fabricated GO-coated cotton fabrics by dipping the fabrics into GO dispersed aqueous solutions. These authors have observed a slight increase (1.2–1.7 times) in the resistance of the GO-coated fabrics reduced for tr 60 min compared with those optimally reduced in Na2S2O4 aqueous solutions and ascribed the resistance increase to the removal of certain RGO sheets from the cotton fabrics during the long reduction process [6]. However, we consider that the amount of resistance increase (5–10 times) for our samples reduced for tr 20 min is too large to explain in terms of the peeling of certain RGO sheets from the polyester fibers. We could observe no evidence for the removal of RGO sheets coated on the fiber bundles using BSA molecules as adhesive agents. Moreover, all of our samples reduced for tr 1 min have the same black color.
4. Conclusions We investigate the transition behaviors of GO-coated fiber bundles from insulator to conductor during the chemical reduction. The structural property of the as-coated GO sheets changes rapidly for the first minute but remains under a high defect density regime, even after the completion of chemical reduction in a hydroiodic acid solution. The starting GO-coated bundle without reduction is electrically insulating, but the resistance of the fiber bundle decreases by five orders of magnitude after optimum reduction. The electrical characteristic shows different behaviors from the structural property upon reduction; the resistance further increases, even after the structural variation becomes saturated. We expect that our results are useful for tuning the structural and electrical properties of RGO-coated materials and for facilitating
94
D.H. Ha et al. / Synthetic Metals 204 (2015) 90–94
the development of continuous and rapid manufacturing processes for the mass production of flexible and stretchable smart materials. Acknowledgements This research was supported by the Fusion Research Program for Green Technologies through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning (NRF-2012M3C1A1048861). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j. synthmet.2015.03.018. References [1] W.S. Hummers Jr., R.E. Offeman, J. Am. Chem. Soc. 80 (1958) 1339. [2] S. Gilje, S. Han, M. Wang, K.L. Wang, R.B. Kaner, Nano Lett. 7 (2007) 3394. [3] S. Stankovich, D.A. Dikin, R.D. Piner, K.A. Kohlhaas, A. Kleinhammes, Y. Jia, Y. Wu, S.T. Nguyen, R.S. Ruoff, Carbon 45 (2007) 1558. [4] V.C. Tung, M.J. Allen, Y. Yang, R.B. Kaner, Nat. Nanotechnol. 4 (2009) 25. [5] J. Molina, J. Fernández, A.I. del Río, J. Bonastre, F. Cases, Appl. Surf. Sci. 279 (2013) 46. [6] M. Shateri-Khalilabad, M.E. Yazdanshenas, Carbohydr. Polym. 96 (2013) 190. [7] W.-W. Liu, X.-B. Yan, J.-W. Lang, C. Peng, Q.-J. Xue, J. Mater. Chem. 22 (2012) 17245. [8] Y.J. Yun, W.G. Hong, W.-J. Kim, Y. Jun, B.H. Kim, Adv. Mater. 25 (2013) 5701. [9] K. Krishnamoorthy, U. Navaneethaiyer, R. Mohan, J. Lee, S.-J. Kim, Appl. Nanosci. 2 (2012) 119.
[10] J. Zhao, B. Deng, M. Lv, J. Li, Y. Zhang, H. Jiang, C. Peng, J. Li, J. Shi, Q. Huang, C. Fan, Adv. Healthcare Mater. 2 (2013) 1259. [11] X. Cai, M. Peng, X. Yu, Y. Fu, D. Zou, J. Mater. Chem. C 2 (2014) 1184. [12] S. Mao, H. Pu, J. Chen, RSC Adv. 2 (2012) 2643. [13] D.W. Boukhvalov, M.I. Katsnelson, J. Am. Chem. Soc. 130 (2008) 10697. [14] K. Erickson, R. Erni, Z. Lee, M. Alem, W. Gannett, A. Zettl, Adv. Mater. 22 (2010) 4467. [15] N. Lu, Y. Huang, H.-B. Li, Z. Li, J. Yang, J. Chem. Phys. 133 (2010) 034502. [16] W. Gao, L.B. Alemany, L. Ci, P.M. Ajayan, Nat. Chem. 1 (2009) 403. [17] D. den Boer, J.G. Weis, C.A. Zuniga, S.A. Sydlik, T.M. Swager, Chem. Mater. 26 (2014) 4849. [18] H.-J. Shin, K.K. Kim, A. Benayad, S.-M. Yoon, H.K. Park, I.-S. Jung, M.H. Jin, H.-K. Jeong, J.M. Kim, J.-Y. Choi, Y.H. Lee, Adv. Funct. Mater. 19 (2009) 1987. [19] C. Gómez-Navarro, R.T. Weitz, A.M. Bittner, M. Scolari, A. Mews, M. Burghard, K. Kern, Nano Lett. 7 (2007) 3499. [20] A.B. Kaiser, C. Gómez-Navarro, R.S. Sundaram, M. Burghard, K. Kern, Nano Lett. 9 (2009) 1787. [21] S.J. Baek, W.G. Hong, M. Park, A.B. Kaiser, H.J. Kim, B.H. Kim, Y.W. Park, Synth. Met. 191 (2014) 1. [22] H. Jiang, L. Chen, S. Chai, X. Yao, F. Chen, Q. Fu, Compos. Sci. Technol. 103 (2014) 28. [23] M. Hu, B. Mi, J. Membr. Sci. 469 (2014) 80. [24] A. Eckmann, A. Felton, A. Mishchenko, L. Britnell, R. Krupke, K.S. Novoselov, C. Casiraghi, Nano Lett. 12 (2012) 3925. [25] A.C. Ferrari, J. Robertson, Phil. Trans R. Soc. Lond. A 362 (2004) 2477. [26] M.M. Lucchese, F. Stavale, E.H. Martins Ferreira, C. Vilani, M.V.O. Moutinho, R.B. Capaz, C.A. Achete, A. Jorio, Carbon 48 (2010) 1592. [27] S. Berciaud, S. Ryu, L.E. Brus, T.F. Heinz, Nano Lett. 9 (2009) 346. [28] W.W. Tyler, A.C. Wilson, Phys. Rev. 89 (1953) 870. [29] I. Jung, D.A. Dikin, R.D. Piner, R.S. Ruoff, Nano Lett. 8 (2008) 4283. [30] E. Bekyarova, M.E. Itkis, N. Cabrera, B. Zhao, A. Yu, J. Gao, R.C. Haddon, J. Am. Chem. Soc. 127 (2005) 5990. [31] M. Hirtz, A. Oikonomou, T. Georgiou, H. Fuchs, A. Vijayaraghavan, Nat. Commun. 4 (2013) 2591.
Contents lists available at ScienceDirect
Synthetic Metals journal homepage: www.elsevier.com/locate/synmet
Transition of graphene oxide-coated fiber bundles from insulator to conductor by chemical reduction Dong Han Ha a, * , Suyong Jung a , Ho-Jong Kim a , Daehee Kim a , Wan-Joong Kim b , Sam Nyung Yi c , Yongseok Jun d, Yong Ju Yun d, ** a
Center for Quantum Measurement Science, Korea Research Institute of Standards and Science, Daejeon 305-340, Republic of Korea IT Convergence Technology Research Laboratory, Electronics and Telecommunications Research Institute, Daejeon 302-350, Republic of Korea Department of Nano Semiconductor Engineering, Korea Maritime University, Busan 606-791, Republic of Korea d Department of Materials Chemistry and Engineering, Konkuk University, Seoul 143-701, Republic of Korea b c
A R T I C L E I N F O
A B S T R A C T
Article history: Received 25 February 2015 Accepted 19 March 2015 Available online 28 March 2015
We fabricate graphene oxide (GO)-coated fiber bundles via electrostatic deposition of GO sheets on polyester/cotton fiber bundles and investigate the changes in their properties by chemical reduction in a hydroiodic acid solution. The structural properties of GO sheets change rapidly upon exposure to reduction agents but remain under a high defect density regime after the reduction. The electrical resistance of the fiber bundles decreases logarithmically for the first minute of the reduction. In contrast to the structural property, a longer chemical reduction over tens of minutes deteriorates the electrical property. The effect of chemical reduction on the electrical conductance of GO-coated fiber bundles is discussed based on the competition of various factors, such as the number of conducting pathways and the carrier density, among others. ã 2015 Elsevier B.V. All rights reserved.
Keywords: Graphene oxide Polymer Chemical reduction Raman spectroscopy Electrical conductance
1. Introduction Graphene oxide (GO) has attracted significant attention as an intermediate material for the mass production of graphene because it can be easily converted into conducting reduced graphene oxide (RGO) through the post-reduction process. Chemically exfoliated mono- or few-layered GO sheets are derived by treating bulk graphite with strong oxidizing agents in aqueous solutions [1–4]. New types of textiles fabricated by coating natural or synthetic textiles with GO sheets and post-treating are expected to advance the development of flexible and stretchable smart materials for wearable electronic devices and healthcare fabrics with antibacterial activity at a low cost and on a large scale [5–11]. GO is electrically insulating because of the significant number of attached oxygenated functional groups (hydroxyl and epoxy groups, among others), which induces a highly disordered lattice structure. During the post-reduction process, which converts insulating GO into conducting p-type RGO, the structural and electrical properties of GO change [4,12–16]. Boer et al. have reported a correlation between the apparent surface roughness
* Corresponding author. Fax: +82 42 868 5953. ** Corresponding author. Fax: +82 2 444 3490. E-mail addresses: [email protected] (D.H. Ha), [email protected] (Y.J. Yun). http://dx.doi.org/10.1016/j.synthmet.2015.03.018 0379-6779/ ã 2015 Elsevier B.V. All rights reserved.
and the level of deoxygenation of GO sheets by employing scanning tunneling microscopy [17]. However, the detailed transition behavior of GO from insulator to conductor remains unknown. To date, many researchers have continued the reduction process over an extended period (>30 min) to prepare RGO with high electrical conductivity [2–6,16–19]. A longer reduction time might be an obstacle to developing a continuous and rapid manufacturing process for the mass production of RGO-based materials. It is believed that the RGO sheet is composed of nanometer-sized conducting crystalline sp2 domains interspersed with insulating oxidized regions on which oxygenated functional groups are densely attached [14,15,19], and that the charge transport mechanism through the RGO sheet generally follows the variable range hopping (VRH) model [19–21]. A detailed understanding of the structural and electrical properties of RGO-coated textiles is required to develop flexible and stretchable smart materials suitable for individual requirements. In this study, we fabricate RGO-coated fiber bundles, and analyze their as-fabricated structural and electrical property changes as a function of post-chemical reduction time using micro-Raman spectroscopy and electrical transport measurements. Our measurements show that the property changes occur rapidly upon exposure to the reduction agents (hydroiodic acid). However, the electrical characteristic shows different behaviors from those in the Raman spectroscopy studies upon reduction
D.H. Ha et al. / Synthetic Metals 204 (2015) 90–94
because the resistance further increases, even after the structural variation becomes saturated. 2. Experimental We wrap commercially available polyester fiber bundles (Hana Spinning Co., polyester, 100%) with negatively charged GO sheets via electrostatic deposition using positively charged bovine serum albumin (BSA) molecules as adhesive agents in aqueous solutions [22,23]. Polyester is a flexible synthetic polymer that is extensively employed in daily life (apparel and home furnishings) and in industry (safety belts, bottles, and insulating tapes). We also obtained GO-coated cotton fiber bundles (Fujix Ltd., cotton, 100%) using the same method. The GO sheets are prepared from natural graphite by using a modified Hummers and Offeman method [1,8]. The deposition of the GO sheets on the fibers continues until the attractive forces between the GO sheets and the BSA molecules become too weak to attract additional GO sheets; thus, a few layers of GO might be deposited (Fig. 1). The charges on the GO sheets and BSA molecules depend on the preparation conditions, especially the pH value. Our samples are prepared in aqueous solutions with a pH value from 3 to 4. The GO layers are post-reduced by immersing the GO-coated fiber bundles in a solution of 2.0 ml of hydroiodic acid (57 wt% in H2O) and 5.0 ml of acetic acid (>99.7%) at 40 C. Subsequently, the samples are rinsed with a saturated sodium bicarbonate (NaHCO3) solution and then with distilled water and, finally, dried at room temperature. The reduction changes the color of the GO-coated fibers from light brown to black. Details on the fabrication of the RGO-coated fiber bundles are described elsewhere [8]. The properties of the as-fabricated RGO-coated fiber bundles are investigated using micro-Raman spectroscopy and electrical measurements. A confocal micro-Raman experiment is performed under ambient conditions in backscattering geometry. A laser line of 514.5 nm (2.41 eV) is focused on the RGO sheets attached to the fibers using a 50 objective lens, and the scattered light is analyzed using a Horiba Jobin Yvon spectrometer (LabRAM HR) equipped with a cooled charge-coupled device. The electrical conductance is analyzed by measuring the current versus voltage (I–V) relationships in the low bias-voltage regime in a vacuum using a probe station (LakeShore, CRX-4K).
91
images show that the GO coating on the fibers is inhomogeneous (Fig. 1). We collect Raman signals from the RGO sheets that are partially detached from the fibers to exclude the strong Raman signals from the underlying fibers. In cases of arbitrary points of RGO-coated polyester/cotton fibers, the Raman signals from the underlying polyester/cotton are too strong to overshadow the subtle changes in the Raman spectra of RGO. Raman spectra from the underlying fibers show a polarization dependence on the fiber direction. However, the Raman spectra of the RGO sheets show no polarization dependence. (Fig. A.1, Supplementary material). Fig. 2 shows a series of Raman spectra of RGO as a function of reduction time. The G peak at approximately 1580 cm1 is the stretching mode of the C—C bond of the sp2 structure in the rings and chains, while the D peak at approximately 1350 cm1, related to the breathing of the sp2 carbon rings, is only Raman active in the presence of point defects. In addition, the D0 peak, shown as a shoulder of the G peak at approximately 1620 cm1, is also related to defects [24–26]. A strong D peak and broad G peak compared with those of pristine graphene imply that there are many structural disorders, such as vacancies and corrugations, on the RGO sheets. Note that the intensity ratio of the D peak to the G peak (I(D)/I(G)) increases by reduction, indicating that our samples are in a highly defective state, even after the chemical reduction in a hydroiodic acid solution is completed, as explained in later.
3. Results and discussion RGO-coated fiber bundles are composed of hundreds of individual microfibers with diameters of approximately 10 mm. The optical microscope and scanning electron microscope (SEM)
Fig. 2. Change of the Raman spectrum of the RGO sheets deposited on polyester fibers as a function of chemical reduction time. All of the Raman spectra are normalized to have the same D peak intensity. The incident laser line is 514.5 nm.
Fig. 1. (a) An optical microscope image of the RGO-coated polyester fiber bundle. The inset shows an enlarged image of an individual polyester fiber. (b) SEM image showing the RGO sheet incompletely wrapping a fiber. The dotted circle indicates the part of the RGO sheet from which the as-fabricated Raman spectrum is obtained, selected to exclude the strong effects of the underlying fiber.
92
D.H. Ha et al. / Synthetic Metals 204 (2015) 90–94
Fig. 3. Changes in (a) peak position, (b) FWHM and (c) intensity ratio of each Raman peak of the RGO sheets deposited on polyester fibers versus chemical reduction time. (d) I (D)/I(G) versus I(D0 )/I(G). The intensity is defined as the height of a peak. The letters in (d) denote the reduction time for each point.
Fig. 3 summarizes the Raman spectroscopy studies on RGO sheets deposited on polyester fibers and shows the position, full width at half maximum (FWHM), and intensity change for each Raman peak as a function of reduction time. We extract the Raman parameters after averaging the signals from three to five different sample locations. The peaks are fitted with Lorentzian functions, and the intensity is defined as the height of the peak. As shown in Fig. 3, the Raman signals change abruptly during the first minute of the reduction process, and further treatment does not cause noticeable changes. For approximately one minute, the positions of the G and D0 peaks are blue shifted by 4–5 cm1, but the D peak does not move. The FWHMs of all of the peaks decreased remarkably, causing the D0 peak to become distinct as a shoulder of the G peak (Fig. 2). The FWHMs of the D peak decrease by nearly half after the beginning of the reduction process, indicating that a
significant amount of the oxygenated functional groups are removed. Once the oxygenated groups are removed, the area of the pristine sp2-bonded planar structure expands, and the structural homogeneity of RGO improves, which results in a narrowing of the G peak. Along with the position and FWHM of the G peak, the intensity ratios of the D and D0 peaks to the G peak, I(D)/I(G) and I(D0 )/I(G), are also good indicators for identifying the nature of the defects on graphene. The increase in the I(D)/I(G) ratio by chemical or thermal reduction accompanying the restoration of the pristine sp2-bonded planar carbon structure was a paradox in earlier studies on RGO [3,4,7,18]. However, studies on graphene subjected to plasma treatment or ion bombardment have shown that the I(D)/I(G) ratio increases, and the position of the G peak is blue shifted as the defect density initially increases. We named this low defect density
Fig. 4. (a) Temperature dependence of normalized conductance of the GO-coated polyester fiber bundles after reducing for 15 s and 30 min. (b) Resistance of the RGO-coated polyester fiber bundles at room temperature under vacuum as a function of the reduction time (tr). The fiber bundle for tr = 0 min (not included in the figure) shows insulating behavior. The inset is an optical microscope image of the RGO-coated polyester fiber bundles for the electrical measurement. The fiber bundles are attached to Au electrodes using 4 N indium, and the resistance is measured using the two-probe configurations between the inner electrodes.
D.H. Ha et al. / Synthetic Metals 204 (2015) 90–94
regime ‘Stage 1’. As the defect density further increases to a higher defect density regime, the ‘Stage 2’ I(D)/I(G) ratio decreases, and the G peak becomes red shifted [24–26]. Based on our data, we can claim that the RGO sheets remain in ‘Stage 2’, even after a lengthy reduction in a hydroiodic acid solution. Even after the reduction, the widths of every peak remain broad; for example, the FWHMs of the G and D peaks are approximately 50 cm1 and 70 cm1, respectively. Especially, the FWHM of the G peak remains significantly broader than that (10 cm1) of pristine graphene [27]. Although a previous study has suggested that significant restoration of the planar sp2 carbon structure could be achieved by chemical reduction [16], our results indicate that many structural defects remain on the RGO sheets, even after the chemical reduction. Fig. 3(d) shows the relationship between I(D)/I(G) and I (D0 )/I(G), which confirms that the restoration of the planar sp2 carbon structure becomes saturated under a high defect density regime after approximately one minute [24]. We observed similar Raman signals from the RGO-coated cotton fiber bundles. However, many RGO lumps are observed on the surfaces of the cotton fibers, which is consistent with the previous reports [5,9], while the surfaces of the polyester fibers are relatively smoother in the optical microscope images (Fig. A.2, Supplementary material). Our results confirm that the natural and synthetic fibers are easily coated with GO sheets in aqueous solutions, and the chemical reduction of GO in a hydroiodic acid solution becomes saturated under a regime of high defect density. Fig. 4 shows an electrical property change in the RGO-coated polyester fiber bundles versus the reduction time (tr). The starting GO-coated bundle without reduction, not shown in Fig. 4, is insulating. All of the other samples with chemical reduction show ohmic I–V characteristics between VDC = 0.1 V, indicating that conducting pathways are formed by the reduction (Fig. A.3, Supplementary material). The electrical resistance of the RGOcoated fiber bundles shows good homogeneity in which each part of a long RGO-coated fiber bundle has approximately the same resistance per unit length. For both fiber bundles reduced for 15 s and 30 min, the normalized conductance shows similar behavior to that of a multi-layered GO flake annealed at 250 C; linear dependence on the temperature and extrapolation to zero value at zero temperature (Fig. 4a) [21]. It is interesting to note that the electrical resistance shows different behaviors with the reduction time compared with those of the Raman spectroscopy studies. As the reduction time increases, the resistance decreases at a logarithmic rate for the first minute and does not show a noticeable change until it starts increasing for tr 20 min (Fig. 4b). In comparison, the Raman spectrum does not show a noticeable change for tr 1 min, suggesting that the degree of structural disorder is no longer a decisive factor in the resistance change for tr 20 min. The conductance of the RGO-coated fiber bundle increases by more than three orders of magnitude as the reduction process begins (if we assume the conductance increases uniformly for the first 30 s, then the total conductance increases by five orders of magnitude). We extract the conductivity of the RGO sheet by assuming that the current flows through the surfaces of the RGO-coated layers on the polyester microfibers with a 10 mm diameter. The measured conductivity, on the order of 101 S m1 for 1 min tr 20 min, is significantly smaller than that of graphite (105 S m1) [28], indicating that our RGO sheet still remains in ‘Stage 2’, even after the chemical reduction in a hydroiodic acid solution is complete. The RGO sheet consists of nanometer-sized conducting crystalline sp2 domains interspersed with insulating oxidized disordered regions [14,15,19]. The initial GO-coated sample is insulating because a highly disordered region covers the majority of the GO sheets so that no conducting pathway is formed between
93
the source and drain electrode. As the oxygenated functional groups are removed, the area of each crystalline domain increases; in addition, new crystalline domains are formed, which result in the opening of conducting pathways. We attribute the rapid logarithmic decrease in resistance over the first minute to the increase in the number of conducting pathways. The percolation theory predicts that the electrical conductivity s is related to the number of conducting pathways by the power law s ðpÞ / ðp pc Þa near the percolation threshold, where p is the coverage, pc is the critical coverage at which conduction begins, and a depends on the dimensions of the system [22,29,30]. Fig. 4b indicates that most of the conducting pathways between the electrodes are formed for approximately one minute, which is supported by the Raman spectroscopy results. Although the reducing agents remove oxygenated functional groups during the reduction process, they might leave iodine ions (I) on the RGO sheets, which act as scattering centers for the current flow. Moreover, electrons could transfer from the iodine ions to the p-type RGO, causing a decrease in the RGO hole density [18,31]. Therefore, we must consider the conducting pathways, carrier density and mobility, among other factors, to understand the change in the RGO resistance during the chemical reduction process. The number of conducting pathways has the dominant effect on the resistance until the formation of most of the conducting pathways at approximately one minute. Further removal of the remaining oxygenated groups is not active for the next 1 tr 20 min, as summarized in Fig. 3. Therefore, we expect only a small increase in the number of conducting pathways, which is considered to be balanced against the decrease in the charge density and mobility caused by the adsorption of iodine ions on the RGO sheets for 1 tr 20 min. The charge density and mobility become the decisive factors for the resistance for tr 20 min because we no longer expect the formation of additional conducting pathways. Shateri-Khalilabad et al. have fabricated GO-coated cotton fabrics by dipping the fabrics into GO dispersed aqueous solutions. These authors have observed a slight increase (1.2–1.7 times) in the resistance of the GO-coated fabrics reduced for tr 60 min compared with those optimally reduced in Na2S2O4 aqueous solutions and ascribed the resistance increase to the removal of certain RGO sheets from the cotton fabrics during the long reduction process [6]. However, we consider that the amount of resistance increase (5–10 times) for our samples reduced for tr 20 min is too large to explain in terms of the peeling of certain RGO sheets from the polyester fibers. We could observe no evidence for the removal of RGO sheets coated on the fiber bundles using BSA molecules as adhesive agents. Moreover, all of our samples reduced for tr 1 min have the same black color.
4. Conclusions We investigate the transition behaviors of GO-coated fiber bundles from insulator to conductor during the chemical reduction. The structural property of the as-coated GO sheets changes rapidly for the first minute but remains under a high defect density regime, even after the completion of chemical reduction in a hydroiodic acid solution. The starting GO-coated bundle without reduction is electrically insulating, but the resistance of the fiber bundle decreases by five orders of magnitude after optimum reduction. The electrical characteristic shows different behaviors from the structural property upon reduction; the resistance further increases, even after the structural variation becomes saturated. We expect that our results are useful for tuning the structural and electrical properties of RGO-coated materials and for facilitating
94
D.H. Ha et al. / Synthetic Metals 204 (2015) 90–94
the development of continuous and rapid manufacturing processes for the mass production of flexible and stretchable smart materials. Acknowledgements This research was supported by the Fusion Research Program for Green Technologies through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning (NRF-2012M3C1A1048861). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j. synthmet.2015.03.018. References [1] W.S. Hummers Jr., R.E. Offeman, J. Am. Chem. Soc. 80 (1958) 1339. [2] S. Gilje, S. Han, M. Wang, K.L. Wang, R.B. Kaner, Nano Lett. 7 (2007) 3394. [3] S. Stankovich, D.A. Dikin, R.D. Piner, K.A. Kohlhaas, A. Kleinhammes, Y. Jia, Y. Wu, S.T. Nguyen, R.S. Ruoff, Carbon 45 (2007) 1558. [4] V.C. Tung, M.J. Allen, Y. Yang, R.B. Kaner, Nat. Nanotechnol. 4 (2009) 25. [5] J. Molina, J. Fernández, A.I. del Río, J. Bonastre, F. Cases, Appl. Surf. Sci. 279 (2013) 46. [6] M. Shateri-Khalilabad, M.E. Yazdanshenas, Carbohydr. Polym. 96 (2013) 190. [7] W.-W. Liu, X.-B. Yan, J.-W. Lang, C. Peng, Q.-J. Xue, J. Mater. Chem. 22 (2012) 17245. [8] Y.J. Yun, W.G. Hong, W.-J. Kim, Y. Jun, B.H. Kim, Adv. Mater. 25 (2013) 5701. [9] K. Krishnamoorthy, U. Navaneethaiyer, R. Mohan, J. Lee, S.-J. Kim, Appl. Nanosci. 2 (2012) 119.
[10] J. Zhao, B. Deng, M. Lv, J. Li, Y. Zhang, H. Jiang, C. Peng, J. Li, J. Shi, Q. Huang, C. Fan, Adv. Healthcare Mater. 2 (2013) 1259. [11] X. Cai, M. Peng, X. Yu, Y. Fu, D. Zou, J. Mater. Chem. C 2 (2014) 1184. [12] S. Mao, H. Pu, J. Chen, RSC Adv. 2 (2012) 2643. [13] D.W. Boukhvalov, M.I. Katsnelson, J. Am. Chem. Soc. 130 (2008) 10697. [14] K. Erickson, R. Erni, Z. Lee, M. Alem, W. Gannett, A. Zettl, Adv. Mater. 22 (2010) 4467. [15] N. Lu, Y. Huang, H.-B. Li, Z. Li, J. Yang, J. Chem. Phys. 133 (2010) 034502. [16] W. Gao, L.B. Alemany, L. Ci, P.M. Ajayan, Nat. Chem. 1 (2009) 403. [17] D. den Boer, J.G. Weis, C.A. Zuniga, S.A. Sydlik, T.M. Swager, Chem. Mater. 26 (2014) 4849. [18] H.-J. Shin, K.K. Kim, A. Benayad, S.-M. Yoon, H.K. Park, I.-S. Jung, M.H. Jin, H.-K. Jeong, J.M. Kim, J.-Y. Choi, Y.H. Lee, Adv. Funct. Mater. 19 (2009) 1987. [19] C. Gómez-Navarro, R.T. Weitz, A.M. Bittner, M. Scolari, A. Mews, M. Burghard, K. Kern, Nano Lett. 7 (2007) 3499. [20] A.B. Kaiser, C. Gómez-Navarro, R.S. Sundaram, M. Burghard, K. Kern, Nano Lett. 9 (2009) 1787. [21] S.J. Baek, W.G. Hong, M. Park, A.B. Kaiser, H.J. Kim, B.H. Kim, Y.W. Park, Synth. Met. 191 (2014) 1. [22] H. Jiang, L. Chen, S. Chai, X. Yao, F. Chen, Q. Fu, Compos. Sci. Technol. 103 (2014) 28. [23] M. Hu, B. Mi, J. Membr. Sci. 469 (2014) 80. [24] A. Eckmann, A. Felton, A. Mishchenko, L. Britnell, R. Krupke, K.S. Novoselov, C. Casiraghi, Nano Lett. 12 (2012) 3925. [25] A.C. Ferrari, J. Robertson, Phil. Trans R. Soc. Lond. A 362 (2004) 2477. [26] M.M. Lucchese, F. Stavale, E.H. Martins Ferreira, C. Vilani, M.V.O. Moutinho, R.B. Capaz, C.A. Achete, A. Jorio, Carbon 48 (2010) 1592. [27] S. Berciaud, S. Ryu, L.E. Brus, T.F. Heinz, Nano Lett. 9 (2009) 346. [28] W.W. Tyler, A.C. Wilson, Phys. Rev. 89 (1953) 870. [29] I. Jung, D.A. Dikin, R.D. Piner, R.S. Ruoff, Nano Lett. 8 (2008) 4283. [30] E. Bekyarova, M.E. Itkis, N. Cabrera, B. Zhao, A. Yu, J. Gao, R.C. Haddon, J. Am. Chem. Soc. 127 (2005) 5990. [31] M. Hirtz, A. Oikonomou, T. Georgiou, H. Fuchs, A. Vijayaraghavan, Nat. Commun. 4 (2013) 2591.