Diamond and Related Materials 12 (2003) 1524–1531
Electrical transport properties of conjugated polymer onto self-assembled aligned carbon nanotubes L. Valentinia, I. Armentanoa, P. Santillib, J.M. Kennya,*, L. Lozzic, S. Santuccic a
Materials Engineering Center, Universita` di Perugia, Loc Pentima Bass 21, Terni 05100, Italy b Laboratorio SERMS, Universita¯ di Perugia, Terni 05100, Italy c Dipartimento di Fisica-Unita` INFM, Universita` dell’Aquila, Coppito (AQ) 67010, Italy Received 30 December 2002; received in revised form 12 May 2003; accepted 21 May 2003
Abstract We present a simple fabrication method for a micrometer-scale conducting network made of self-assembled aligned carbon nanotubes (CNTs) thin films deposited by pulsed plasma enhanced chemical vapour deposition technique. Once the nanotubes have been grown, spin coating has been used to prepare a CNTsyconjugated polymer composite. The electrical transport properties of the hybrid system formed by the conjugated polymer (poly(3-octylthiophene)–P3OT) and CNTs have been then characterised without any further processing. At room temperature the aligned nanotubes show an ohmic behaviour, while the current–voltage characteristic curve indicates a surface potential modification of the system CNTsyconjugated polymer leading to a semiconductor behaviour. Raman spectroscopy is successfully applied to demonstrate that in the composite film the changes in the electrical resistance can be explained in terms of intercalation of the polymer matrix and interaction of P3OT with the nanotubes. 䊚 2003 Elsevier Science B.V. All rights reserved. Keywords: Carbon nanotubes; Conjugated polymer; Electrical properties; Raman spectroscopy
1. Introduction The unique electronic and mechanical properties of nanotubes w1x have promised much potential for a vast range of applications, including quantum wires w2x tips for scanning probe microscopy w3x and molecular diodes w4x. Recently, much attention has been given to the use of nanotubes in composite materials, to harness their exceptional mechanical w5x and electronic w6,7x properties. A wide range of host materials has been used, including polymers w8,9x, ceramics w10x and metals w11x. Most recently, the research w12x has focused on composites of electronically active conjugated polymers and carbon nanotubes, which demonstrate a number of advantages. Conjugated polymers show potential for electronic device applications w13x with the incorporation of carbon nanotubes promising to greatly enhance transport properties in these systems w14x. This is thought to be a key issue for the realisation of viable *Corresponding Author. Tel.: q39-744-492-939; fax: q39-744492-925. E-mail address:
[email protected] (J.M. Kenny).
devices such as organic light-emitting diodes (OLEDS) and solar cells. Furthermore, incorporation of nanotubes should also increase the mechanical properties of composite materials w15x and, by increasing their thermal conductivity, improve their environmental stability w12x. In addition, during composite formation the problems of nanotube processing are also addressed due to the not so easy dispersion of nanotubes in the polymer matrix. While conjugated polymers are known to wet both nanotubes and graphitic impurities, only the polymer coated nanotubes form a stable dispersion. The graphitic impurities sediment out of the blend requiring a subsequent purification of the composite by decantation. For these reasons, in most cases building reliable electrical circuits involves difficult and time-consuming manipulation of nanotubesypolymer composites w16– 18x. These are non-compatible with the connection of a large assembly of nanotubes on the same circuit. In order to become technology suitable for VLSI, one should be able to grow CNTs on specific locations in a single batch process. Our idea is to perform the connection of the CNT during the synthesis itself, thus sup-
0925-9635/03/$ - see front matter 䊚 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0925-9635(03)00185-7
L. Valentini et al. / Diamond and Related Materials 12 (2003) 1524–1531
1525
Fig. 1. (a) SEM image of Si3N4 ySi substrate with platinum interdigital electrodes; (b) schematic diagram of CNTs linking pre-patterned platinum contacts in a resistor geometry.
pressing the critical postsynthesis manipulation. Once the nanotubes have been grown, spin coating, drop casting and printing technologies can be used to prepare large area of CNTsypolymer composites. For that purpose, in this paper, we report the preparation and characterisation of micrometer-scale electrical circuitry based on a self-assembled aligned carbon nanotube network. The electronic interaction between a conjugated polymer and nanotubes is then characterised by I–V spectroscopy. A change of the electrical behaviour of the nanotubes by the polymer addition is explained in terms of polymer–nanotube interaction. 2. Experimental details Carbon nanotube thin films were grown using a radiofrequency pulsed plasma enhanced chemical vapour deposition (r.f. PECVD) system. A thin film (3 nm) of Ni catalyst was deposited onto Si3N4 ySi substrates
provided with platinum interdigital electrodes. The serpentine resistor geometry upon which CNTs were grown is reported in Fig. 1. With the plasma discharge in the off position prior to initiating the deposition, the substrates were heated to 650 8C and held at this temperature for 15 min to induce the clusters formation of the catalyst layer before the activation of the CNTs plasma deposition. Substrates were positioned on a heated cathode capable of reaching a maximum temperature of 850 8C and connected to the radiofrequency (RF) power supply. For pulsed-PECVD runs, a peak RF power of 100 W was applied during on-time excitation. In our experiment, we used an on-time excitation of 0.1 s with a duty cycle, defined as a fraction of the total time during which the power was applied, fixed at 18%. The CNTs depositions were carried out with deposition pressure and temperature fixed at 53 Pa and 570 8C, respectively. The total precursor (CH4) gas flow rate was kept
1526
L. Valentini et al. / Diamond and Related Materials 12 (2003) 1524–1531
constant at 40 sccm. The film deposition was performed with a RF bias voltage fixed at y150 V. A poly(3-octylthiophene)(P3OT) organic film was then deposited by drop and spin coating from a 10 mgy ml chloroform solution onto the planar resistor coated with CNTs. Raman scattering spectra were recorded by a Jobin Yvon micro-Raman LabRam system in a backscattering geometry. A 632.8 nm (1.96 eV) He–Ne laser was used as the light source and the power of the laser was adjusted by optical filters. By using a 100= objective lens, the illuminated spot on the sample surface was focused to approximately 2 mm in diameter. The resolution of Raman spectra was better than 1 cmy1. The scanning electron microscopy investigation was performed on a field emission scanning electron microscope LEO 1530 operated at 1 and 5 kV. A transmission electron microscope (TEM) was used to determine the structure of the CNTs. To characterise the temperature dependence of the electrical transport property, CNTs and polymerynanotubes composite deposited onto the planar resistor were held in a leybold cold-head cryodyne refrigerator and the temperature was controlled by 1901 temperature controller. The current–voltage (I–V) characteristic curves were measured with a precision semiconductor parameter analyser (Keithley 236 source measure unit). 3. Results and discussion High-resolution field emission SEM images of CNTs deposited with pulsed plasma are plotted in Fig. 2. Fig. 2a shows the separation zone between the Pt electrode and the CNT film. The reason why CNTs did not grow on the Pt, as shown in Fig. 2a, was due to the lack of fragmentation of the Ni film on the platinum electrode, reasonably due to a diffusion of Ni atoms within platinum during the annealing process at 650 8C, leading eventually to the formation of a binary Ni–Pt alloy. The inset of Fig. 2a shows the top of the aligned CNTs. From Fig. 2b, a very good alignment of the nanotubes with a metal cap predominantly at the top of them is observed. The top view SEM image in Fig. 2c shows that spin coating of the P3OT produces a composite film with individual CNTs well embedded, while still aligned, in the polymer matrix. The presence of the metal cap (bright spots in Fig. 2a and b) indicate a growth mechanism of CNTs similar to that reported in Ref. w19x. In particular, the carbonaceous gas decomposes on the surface of the catalyst particle; the carbon diffuses across the Ni and then, the carbon precipitates outwardly. As carbon atoms are further supplied, a carbon-metal eutectic alloy can be formed, decreasing the melting temperature of the alloy. The formation of carbon–Ni eutectic enhances the dif-
Fig. 2. SEM photomicrograph of as-grown CNTs: (a) top view with the Pt electrode region and the as-grown structure of CNTs on a SiySi3N4 substrate; the inset is a tilted SEM image showing the top of the aligned carbon nanotubes. (b) Magnification of the as-grown structure of CNTs on SiySi3N4. (c) Top view after polymer addition by spin coating.
fusion of carbon in the metal alloy, initiating carbon aggregations, which act as a nucleation seed for nanotube growth.
L. Valentini et al. / Diamond and Related Materials 12 (2003) 1524–1531
1527
Fig. 3. TEM photomicrograph of CNTs. Large inclusions of Ni at the nanotube tip are shown.
The tubular and defective structure of the nanotubes was verified using TEM microscopy, as shown in Fig. 3. The nanotubes are 150–200 nm long with the Ni catalyst particle at the top. According to previous research w20x, the above results seem to confirm that the carbon nanotubes final structure and diameter depend on the diameter of the Ni nanoparticles which were formed during the annealing of the metal layer at 650 8C. Raman scattering is a powerful technique to probe the structure-property relationship in both carbon nanotubes and conjugated polymer. In Fig. 4b the high frequency part of the Raman spectrum of the CNTs is compared to that of the composite for an excitation of 632.8 nm. The two main features in the Raman spectra are the D and G peaks at approximately 1350 cmy1 and 1600 cmy1, respectively. The G band corresponds to the symmetric E2 g vibrational mode in graphite-like materials, while the appearance of the strong D line can be associated to the turbostratic structure of carbon sheets in the tubes, namely the finite size (nanometer order) of the crystalline domains and the high fraction of defects w21,22x. Thus, the large amount of defects (Fig. 3) on the surface of the tubes explain the enhancement of the D line at 1348 cmy1. The bands in the range 1550–1650 cmy1 cannot be clearly interpreted due to the multitude of nanotube modes calculated in this range w23x. From Fig. 4, it is evident that the shift to higher frequencies of the G peak position accompanied by a reduction of its width when the polymer is incorporated onto the nanotubes. With introduction of P3OT, the feature at 1574 cmy1 is up shifted to approximately 1600 cmy1 with a full width at half maximum that decreases from 48 cmy1 to approximately 28 cmy1. Moreover, the peak observed at 1574 cmy1 for CNTs increases in intensity in the
Fig. 4. High frequency Raman spectra of CNTs and P3OTyCNTs composite.
1528
L. Valentini et al. / Diamond and Related Materials 12 (2003) 1524–1531
Fig. 5. I–V curves of (a) P3OT and polymeryCNTs composite; (b) CNTs array prepared using pulsed PECVD.
composite indicating a dilution effect of the carbon nanotubes when blended with P3OT. Similar modifications have been observed in spectra of composite systems containing nanotubes and PMMA w24x. These results suggest that the conjugated polymer intercalates between nanotubes decreasing the interaction between individual tubes. This interpretation also explains the observed narrowing of the G line feature. Quantitative measurements of electrical resistance in a current direction perpendicular to the tube axis are obtained according to the schematic diagram shown in Fig. 1b. Most studies on electronic properties of CNTs rely on measuring the I–V curve of individual CNTs or CNT bundles one at a time after they have been laid or grown horizontally on the surface and connected with metal electrodes w25,26x. For our conducting network made of self-assembled vertically aligned carbon nanotube arrays, I–V curves of hundreds of individual CNTs can be investigated quickly. Fig. 5a shows a typical I–V curve for the CNTs array prepared using pulsed PECVD. These CNTs are highly conductive. In particular the I–V curve obtained is essentially a straight line
within the scan range of "0.5 V giving a resistance of 250 V. For comparison, I–V measurements on P3OT and on the hybrid system of the conjugated polymer and CNTs are shown in Fig. 5b. I–V measurement on the P3OT shows a gap of 1 V confirming the much lower conductivity of the polymer than the nanotubes array. As a result of polymer addition on CNTs, the I–V curve evidences a reduction of the gap to 0.1 V indicating an electric interaction between CNTs and the organic polymer. In fact, according to the explanation of the field emission properties on a SWNT–polymer composite proposed by Alexandrou et al. w27x, here we propose that the reduction of the gap observed in the composite is due to the existence of a junction between the cap of semi-metal CNTs, the semiconducting matrix (P3OT) and vacuum. At the junction between the CNTs and the polymer, the surface potential undergoes to a strong step due to the difference in work function between the two materials (4.8 eV for CNTs and 2.8 eV for P3OT w28,29x). The surface irregularity, as shown by SEM analysis in Fig. 2c, will modify the potentials lines immediately above the two materials in vacuum, in the vicinity of the junction. The very rapid change in potential at the junction between the two materials results in an electron tunnelling from the conducting channel (CNTs) into the matrix (P3OT). From a practical application point of view, the selective location of aligned CNTs between the platinum electrodes allows us to measure the electrical resistivity of the tubes perpendicular to the tube axis. Generally, the resistivity measurements are performed on CNT slurries coated on an insulating substrate. To our knowledge few works exist on the resitivity measurements of self assembled carbon nanotubes perpendicularly aligned and sticking to an insulating substrate w30x. Heer and co-workers w31x reported that the electrical resistivities of aligned CNTs are anisotropic, being smaller along the tubes than normal to them, because of corresponding differences in electrical transport. Chauvet et al. w32x observed magnetic anisotropies of the aligned CNTs as well. Recently, electrical, thermal and structural anisotropies of magnetically aligned single-wall CNT films have been obtained by the Fischer group w33,34x. However, all of the above aligned CNT films were not as grown but they were either transferred on a Teflon surface w31,32x or deposited on a nylon filter membrane, w33,34x which resulted in the deviation of the experimental results. Moreover, these CNT films are imperfectly aligned with some entanglement or curvature of individual tubes (ropes). Due to the disadvantageous factors, the observed results are not real reflections of the intrinsic properties of those partly aligned CNT films. The scaled resistivity of our CNTs shows (Fig. 6) a non-linear power law with temperature. Qualitatively, the nanotube material can be thought of as being
L. Valentini et al. / Diamond and Related Materials 12 (2003) 1524–1531
1529
Fig. 6. Resistance vs. temperature for CNTs, P3OT and P3OTyCNTs composite.
composed of conductive rods w35x. Hence, the high resistivity of the material indicates that strong scattering occurs at the tube boundaries as a result of intertube energy barriers, so that the tube–tube contacts act as static defects, limiting the mean free path of the electrons. The resistivity (or resistance) behaviour then reflects the mean-free-path perpendicular to the tubes. The electrons may localise on the individual tubes and intertube electron transport is thermally activated, requir-
ing electrons to thermally hop across intertube energy barriers w36x. Since the tube–tube interactions that deviate from a one-dimensional system and represent additional three-dimensional effects, may play a role as non-interacting Fermi liquids w37,38x, we separate the linear term from Luttinger liquids. In this approach, the resistivity is expressed by adding both linear and nonlinear power laws, r(T)saTyaqbT where a/y1. This prediction concurs greatly with the measured
1530
L. Valentini et al. / Diamond and Related Materials 12 (2003) 1524–1531
Fig. 7. TEM photomicrograph of P3OTyCNTs composite.
values, as shown in the inset of Fig. 6. The transport phenomena from the mat of aligned nanotubes can be understood using the following model. The network of CNTs is connected to CNT–CNT junctions. The crossjunctions between the CNTs or ropes importantly act as a gate for the carriers to move in the mat. Therefore, our film can be represented as a network of effective resistors, in which the effective series resistors is formed by the resistance of the nanotube body (Luttinger liquid) w39x, the resistance of the junctions (Fermi liquid) w38x and the tunnelling resistance due to tube–tube electronic coupling w40x. For the semiconducting nanotubes, we take an average energy gap (Eg) of 0.5 eV, therefore, the dependence of the CNT film resistance at temperature kT-F (i.e. )100 K) can be accounted by using usual expressions for the resistivity of systems containing Fermi and Luttinger liquids conductors. Considering that a fraction of approximately 2y3 of the tubes is semiconducting, while the remaining 1y3 shows a metallic behaviour w41x, the coupling tunnelling resistance can be modelled by considering the coupling of two uni-dimensional wave-guides separated by a tunnelling barrier w42x FfEgy2. In order to explain our observations regarding the strong change in the film resisitivity when the polymer is added onto the tubes (Fig. 6), we suggest that the most important mechanism involved is the effect on the tunnelling resistance between tubes. The current flow in our samples is strongly influenced from the tunnelling within the nanotube separation. All nanotubes share the same graphene structure, hence their work function is expected to be nearly the same, and the Fermi level of the metallic tubes is expected to align the midgap of the semiconducting energy gap. The insertion of the polymer between CNTs (Fig. 7) modifies the density of states introducing an impurity-like level near the onset of the valence band of the nanotube.
When the interacting polymer is sufficiently high, the CNT Fermi level is shifted to lower energies. Therefore, the barrier offered to charge transport is lowered with respect to the as-deposited nanotubes and the resitivity of the whole CNT layer, as experimentally observed, decreases. 4. Conclusions In summary, we prepared micrometer-scale electrical circuitry based on a self-assembled aligned carbon nanotube array. This simple technique enables us to characterise the electronic properties of the vertical array of CNTs and the hybrid composite system of a conjugated polymer and CNTs. Several interesting effects are observed, including the interaction between the polymer and nanotube, the low resistance of the nanotubes array in the direction perpendicular to the tube axis and the substantial changes in the electrical behaviour of the CNTs due to the irregularity in surface potential due to the existence of the polymer-CNTs junction. Finally, we have demonstrated that our approach could be useful to fabricate large-scale nanoelectronic devices in a vertical geometry such as crossbar-addressed transistor network, field emitters and nanoelectrode arrays. References w1x S. Iijima, Nature 354 (1991) 56. w2x S.J. Tans, M.H. Devoret, H. Dai, A. Thess, R.E. Smalley, Nature 386 (1997) 474. w3x H. Dai, J.H. Hafner, A.G. Rinzler, D.T. Colbert, R.E. Smalley, Nature 384 (1996) 147. w4x S.J. Sander, J. Tans, A.R.M. Verschueren, C. Dekker, Nature 393 (1998) 49. w5x M.M.J. Treacy, T.W. Ebbesen, J.M. Gibson, Nature 381 (1996) 678.
L. Valentini et al. / Diamond and Related Materials 12 (2003) 1524–1531 w6x J.W.G. Wildoer, L.C. Venema, A.G. Rinzler, R.E. Smalley, C. Dekker, Nature 391 (1998) 59. w7x T.W. Odom, J. Huang, P. Kim, C.M. Lieber, Nature 391 (1998) 62. w8x B. Tang, H. Xu, Macromolecules 32 (1999) 2569. w9x Z. Jin, K. Pramoda, G. Xu, S. Goh, Chem. Phys. Lett. 337 (2001) 43. w10x G.L. Hwang, K. Hwang, Mater. Chem. 11 (2001) 1722. w11x S.R. Dong, J.P. Tu, X.B. Zhang, Mater. Sci. Eng. A 313 (2001) 83. w12x S. Curran, P. Ajayan, W. Blau, D.L. Carroll, J.N. Coleman, A.B. Dalton, et al., Adv. Mater. 10 (1998) 1091. w13x R.H. Friend, R.W. Gymer, A.B. Holmes, J.H. Burroughes, R.N. Marks, C. Taliani, et al., Nature 397 (1999) 121. w14x P. Fournet, J.N. Coleman, D.F. O’Brien, B. Lahr, A. Drury, W.J.J. Blau, Appl. Phys. 90 (2001) 969. w15x P.M. Ajayan, L.S. Schadler, C. Giannaris, A. Rubio, Adv. Mater. 12 (2000) 750. w16x A. Bachtold, M. Henny, C. Terrier, C. Strunk, C. Schonenberger, J.P. Salvetat, et al., Appl. Phys. Lett. 73 (1998) 274. w17x D.J. Tans, A.R. Verschueren, C. Dekker, Nature 393 (1998) 49. w18x H. Dai, J. Kong, C. Zhou, N. Franklin, T. Tombler, A. Cassel, et al., J. Phys. Chem. B 103 (1999) 11 246. w19x M. Chhowalla, K.B.K. Teo, C. Ducati, N.L. Rupesinghe, G.A.J. Amaratunga, A.C. Ferrari, et al., J. Appl. Phys. 90 (2001) 5308. w20x L. Valentini, J.M. Kenny, L. Lozzi, S. Santucci, J. Appl. Phys. 92 (2002) 6188. w21x M.S. Dresselhaus, P.C. Eklund, Adv. Phys. 49 (2000) 705. w22x A.M. Rao, A. Jorio, M.A. Pimenta, M.S. dantas, R. Saito, G. Dresselhaus, et al., Phys. Rev. Lett. 84 (2000) 1820. w23x M.S. Dresselhaus, P.C. Eklund, Adv. Phys. 49 (2000) 705. w24x C. Stephan, T.P. Nguyen, L. Lamy de la Chapelle, S. Lefrant, C. Journet, P. Bernier, Synth. Met. 108 (2000) 139.
1531
w25x S.J. Tans, A.R.M. Verschueren, C. Dekker, Nature (London) 393 (1998) 49. w26x M.S. Fuhrer, J. Nygard, L. Shih, M. Forero, Y.-G. Yoon, M.S.C. Mazzoni, et al., Science 288 (2000) 494. w27x I. Alexandrou, E. Kymakis, G.A.J. Amaratunga, Appl. Phys. Lett. 80 (2002) 1435. w28x O. Groning, O.M. Kuttel, C. Emmenegger, P. Groning, L. Schlapbach, J. Vac. Sci. Technol. B 18 (2000) 665. w29x K. Yoshino, M. Onoda, Y. Manda, M. Yokoyama, Jpn. J. Appl. Phys., Part 1 27 (1988) 1606. w30x X. Wang, Y. Liu, G. Yu, C. Xu, J. Zhang, D. Zhu, J. Phys. Chem. B 105 (2001) 9422. w31x W.A. De Heer, W.S. Bacsa, A. Chatelain, T. Gerfin, R. Humphrey-Baker, L. Forro, et al., Science 268 (1995) 845. w32x O. Chauvet, L. Forro, W. Bacsa, D. Ugarte, B. Doudin, W.A. De Heer, Phys. Rev. B 52 (1995) R6963. w33x J. Hone, M.C. Llaguno, N.M. Nemes, A.T. Johnson, J.E. Fischer, D.A. Walters, et al., Appl. Phys. Lett. 77 (2000) 666. w34x B.W. Smith, Z. Benes, D.E. Luzzi, J.E. Fischer, D.A. Walters, M.J. Casavant, et al., Appl. Phys. Lett. 77 (2000) 663. w35x A. Thess, R. Lee, P. Nikolaev, H. Dai, P. Petit, J. Robert, et al., Science 273 (1996) 483. w36x L. Langer, L. Stockman, J.P. Heremans, V. Bayot, C.H. Olk, C.V. Haesendonck, et al., J. Mater. Res. 9 (1994) 927. w37x G.U. Sumanasekera, C.K.W. Adu, S. Fang, P.C. Eklund, Phys. Rev. Lett. 85 (2000) 1096. w38x D.J. Bae, K.S. Kim, Y.S. Park, E.K. Suh, K.H. An, J. Moon, et al., Phys. Rev. B 64 (2001) 2 33401. w39x C. Kane, L. Balents, M.P.A. Fisher, Phys. Rev. Lett. 79 (1997) 5096. w40x H. Stahl, J. Appenzeller, R. Martel, Ph. Avorious, B. Lengeler, Phys. Rev. Lett. 85 (2000) 5186. w41x R. Saito, M. Fujita, G. Dresselhaus, M.S. Dresselhaus, Appl. Phys. Lett. 60 (1992) 2204. w42x C.C. Eugster, J.A. Alamo, Phys. Rev. Lett. 67 (1991) 3586.
