Conductivity Measurement of Polydiacetylene Thin Films by Double

Sep 28, 2004 - The conductivity of polydiacetylene thin films has been evaluated in the region below 20 μm using a newly constructed independently dr...
0 downloads 12 Views 175KB Size
J. Phys. Chem. B 2004, 108, 16353-16356

16353

Conductivity Measurement of Polydiacetylene Thin Films by Double-Tip Scanning Tunneling Microscopy Kazuhiro Takami,*,† Jun Mizuno,† Megumi Akai-kasaya,†,‡ Akira Saito,†,‡,§ Masakazu Aono,†,‡,| and Yuji Kuwahara†,‡,§ Department of Material and Life Science, Osaka UniVersity, 2-1 Yamada-oka, Suita, Osaka 565-0871, Japan, Nanoscale Quantum Conductor Array Project, ICORP, Japan Science and Technology Agency (JST), 4-1-8 Honmachi, Kawaguchi, Saitama 332-0112, Japan, Harima Institute, The Institute of Physical and Chemical Research (RIKEN), 1-1-1 Kokuto, Mikazuki, Hyogo 679-5148, Japan, and Nanomaterials Laboratory, National Institute of Material Science (NIMS), 3-13 Sakura, Tsukuba, Ibaraki 305-0003, Japan ReceiVed: June 29, 2004

The conductivity of polydiacetylene thin films has been evaluated in the region below 20 µm using a newly constructed independently driven double-tip scanning tunneling microscope. The polydiacetylene thin films fabricated by means of the Langmuir-Blodgett method showed the formation of islands with sizes of 2-20 µm with the polydiacetylene backbone extended in one direction in each island. It was indicated that the resistance of the polydiacetylene thin films is proportional to the tip-tip distances, suggesting one-dimensional conduction along the polydiacetylene backbones, which was clearly different from that of two-dimensionally uniform conductive thin films of poly(3-octylthiophene). The conductivity of the polydiacetylene thin film was estimated to be (3-5) × 10-6 S/cm, which is 5 orders of magnitude higher than that in the previous report.

Introduction The nanoarchitecture of organic molecules, particularly organic molecular layers on solid surfaces, is highly attractive in view of future applications of nanotechnology. Polydiacetylene (PDA) is one of the candidate materials for molecular wires for the interconnection of devices, because it is a fully π-conjugated conducting polymer1 that overcomes the size constraint imposed by silicon-based technology. Figure 1a shows a schematic structure of PDA (the monomer used is 10,12nonacosadiynoic acid). PDA is obtained by polymerizing diacetylene compounds with appropriate external stimulation, such as heat or ultraviolet (UV) light, in the solid and liquid states.2 For the thin films, the polymerization of the self-assembled PDA layer, such as that fabricated by the Langmuir-Blodgett (LB) method,3-6 has been reported. Recently, we succeeded in controlling the fabrication of a linear PDA wire by using a scanning tunneling microscope (STM) probe tip on a self-assembled monomolecular layer, and the density of states of individual polymers clearly reveals the theoretically predicted π-band and band edge singularities of a one-dimensional polymer.7-10 The conductivity of PDA has been measured in crystalline and thin films. Up to now, it has been considered that PDA without doping is an insulator.11 For example, Day and Lando12,13 reported that the conductivity of the PDA thin film is estimated to be about 10-11 S/cm. The obtained values of * To whom correspondence should be addressed. E-mail: takami@ ss.prec.eng.osaka-u.ac.jp. † Osaka University. ‡ Japan Science and Technology Agency (JST). § Harima Institute, The Institute of Physical and Chemical Research (RIKEN). | Nanomaterials Laboratory, National Institute of Material Science (NIMS).

Figure 1. Structural models of (a) polydiacetylene, (b) polyacetylene, and (c) poly(3-octylthiophene). Blue, white, red, and yellow circles represent carbon, hydrogen, oxygen, and sulfur atoms, respectively. Bond configurations are also shown.

the conductivity of PDA in the three-dimensional bulk structure were less than 10-10 S/cm in other experiments.14,15 In these measurements, however, PDA crystalline and thin films were considered to consist of small crystallites oriented randomly and the electrode intervals were wider than the size of the crystallite. In Day and Lando’s case, for example, the sizes of the PDA crystallites were as large as 100 µm, but the distance between electrodes was 150 µm; thus, it is difficult to say whether they measured the precise conductivity of PDA. On the contrary, polyacetylene (schematically shown in Figure 1b), whose structural/electronic properties are quite similar to those of PDA, has entangled backbones and is a conducting polymer with a

10.1021/jp047177q CCC: $27.50 © 2004 American Chemical Society Published on Web 09/28/2004

16354 J. Phys. Chem. B, Vol. 108, No. 42, 2004

Takami et al.

Figure 2. (a) Typical π-A isotherm of 10,12-nonacosadiynoic acid and (b) typical transfer ratio of the thin films. The red dotted line in (a) is the transition pressure to the substrate mica surface.

conductivity of about 10-6 S/cm for nondoped samples,16,17 which is 5 orders of magnitude higher than that of PDA. To evaluate the conductivity of PDA more precisely, measurement of highly ordered PDA thin films without domain boundaries is essential. When the measuring direction is parallel to the PDA backbone within one highly ordered crystallite, a high conductivity is assumed to be implied. In this study, we have measured the conductivity of polydiacetylene thin films in a micrometer-scale region using the newly constructed independently driven double-tip scanning tunneling microscope (DT-STM). Experimental Method We employed 10,12-nonacosadiynoic acid (CH3(CH2)15CtCCtC(CH2)8COOH, MW ) 430.71, Tokyo Kasei Co. Ltd.) as the diacetylene compound and used it after filtering. For comparison, poly(3-octylthiophene) (H-(C12H18S)n-H, POT; Polysciences Inc.) was used as received to make twodimensional uniform conductive thin films. The structure of POT is schematically shown in Figure 1c. The substrate was natural mica (Nilaco Co. Ltd.). All other chemicals were used as received without further purification. The 10,12-nonacosadiynoic acid thin films were prepared by the conventional Langmuir-Blodgett method.18 The diacetylene monomers were dissolved in a solvent of chloroform with a concentration of 0.25 mg/mL. The solution was spread on an ultrapure water subsurface in a commercial Langmuir trough (KSV Mini trough2, KSV, Finland). After being allowed to stand for 30 min to evaporate the solvent, the film was compressed typically at 18 mN/m and then transferred to the freshly cleaved mica substrate by the lifting method. The typical surface pressure-area (π-A) isotherm of the 10,12-nonacosadiynoic acid and the typical transfer ratio of the thin films are shown in Figure 2. In the transfer, only an even number of layers were transferred to the substrate mica, so we made a threelayer LB film. The thin film was then irradiated with UV light (dominant wavelengths of 194 and 256 nm; Sen special light, 12.7 mW/cm2) for 7 min. After UV irradiation, visible absorption spectra for the films were obtained to confirm the “red/ blue” phase of PDA.19 The structural configurations of PDA thin films were evaluated by atomic force microscopy (AFM; JSTM-4200A, JEOL). For the AFM measurement, a silicon cantilever was used in the ac mode with a resonance frequency of 140 kHz. The POT thin films were prepared by spin casting a polymer/chloroform solution onto freshly cleaved mica substrates at a rotation speed of 2000 rpm. The thickness of the films was confirmed to be about 200 nm by AFM. The conductivity measurements were carried out with the DTSTM. The DT-STM consists of two STM units that can move independently. We employed Pt/Ir tips (Materials Analytical Services Co. Ltd.) whose radius of curvature is less than 50

Figure 3. AFM image of the blue-phase PDA thin film. The crosssectional height profile is indicated at the top. The arrangement of PDA is also shown schematically.

nm. The experimental procedure was as follows. First, the two tips were brought to within 50 µm of each other on the substrate, using a coarse driving stage (Micro piezo slide MS5, Omicron), monitored with a charge-coupled device (CCD) viewing system, and then touched to the thin film. The conditions of tip contact with the substrate were judged from the CCD image of the relative positions of the tip and its shadow. The contact area can be estimated to be about 2 µm in diameter from the contact traces for POT thin films and 1 µm in diameter for PDA thin films by analogy with the case of POT. A current was obtained by applying voltage to one tip, which was then moved toward the other tip gradually to protect the thin film between the two tips. All sample preparations and measurements were performed at room temperature under ambient conditions. Results and Discussion Figure 3 shows a typical AFM image of PDA thin films after 7 min of UV irradiation. It is well-known that the PDA structure changes from the blue phase to the red phase depending on the duration of UV light irradiation.19 On the basis of the results of visible absorption spectroscopy, the sample shown in Figure 2 is attributed to the “blue-phase” thin film. The obtained AFM image is in good agreement with that in the previous study.20 From the AFM image in Figure 3, one can see three regions

Conductivity Measurement of PDA Thin Films

J. Phys. Chem. B, Vol. 108, No. 42, 2004 16355

Figure 5. Schematic diagrams of (a) one-dimensional and (b) twodimensional current flow. Figure 4. Variation of the resistance between two tips for the bluephase PDA thin films as a function of the tip-tip distance. The detection limit of the DT-STM is about 0.1 pA (>8000 GΩ), as shown by the dotted line. The resistance of diacetylene monomer thin films was above the detection limit.

with different heights. The cross-sectional profile across A-B is also shown above the image. The lowest layer is ascribed to the substrate mica surface because of its high flatness. The middle layer is 2.5 nm higher than the lowest layer. This layer was also seen in the diacetylene thin film before UV irradiation, which indicates that it is a nonpolymerized diacetylene thin film. Since the molecular length of 10,12-nonacosadiynoic acid is about 3.8 nm, the axis of the molecule in this image was tilted 50° from the substrate normal. The top layer of this image is 7.5 nm higher than the mica substrate, which corresponds to three layers of molecules. The observed top layers were about 2-20 µm in size. In the top layer, bright lines along the same direction can be seen, as shown in Figure 3, and they are attributed to the PDA backbones. The directions of all PDA backbones are the same in one domain. In addition, there are two kinds of breaks on the terrace of the top layer: one is a prominent line perpendicular to the backbones, and the other is a deep ditch parallel to the backbones. The perpendicular and parallel breaks might be joints of PDA backbones and gaps of molecules due to shrinkage induced by polymerization, respectively. It is noted that a red-phase thin film is created on increasing the UV irradiation time. The AFM image of the redphase thin film indicated that there were many cracks inside the island and the surface was not smooth. We evaluated the conductivity of the blue-phase PDA thin films using the DT-STM. The conductance/resistance was obtained from the current with fixed applied voltage between the two tips when the tip-tip distance was changed under 20 µm. For all data hereafter, the voltage between the two tips was kept at 1 V. When the voltage was gradually increased from 1 V, the current suddenly dropped at around 1.5-2.0 V. This might be caused by the destruction of the polymerized chain between the two tips. The tip-tip distance was identified to be 1 µm when the two tips came into contact with each other. Figure 4 shows the variation of the resistances of the blue-phase PDA thin films depending on the tip-tip distance. Before UV irradiation, the observed currents were below the detection limit of about 0.1 pA (>8000 GΩ, shown as a dotted line in Figure 4) for all tip-tip distances and no I-V dependence was found. In the cases of blue-phase PDA thin films, it was found that the resistance was proportional to the tip-tip distance, as shown in Figure 4. It is well-known that the relationship of resistance vs distance between two point electrodes differs with the dimensions of the objects; that is, when the conduction between the two point contacts has two- and one-dimensional characteristics, the resistance vs distance obeys logarithmic and linear functions, respectively. The scheme of one- and two-dimensional current flows is schematically shown in Figure 5.

Figure 6. Angular dependence of resistance on moving the position of one tip around the other tip while maintaining the tip-tip distance of 5 µm. The schematic diagram of the experimental configuration is also shown.

The linear characteristics of the resistance obtained for the PDA thin films strongly suggest one-dimensional conduction along the PDA backbone. Note that, in the case of red-phase PDA thin films, the currents were below the detection limit and no I-V dependence was found, similar to the case of the nonUV-irradiated films. In addition, an angular dependence of the resistance was observed on moving the position of one tip around the other tip while maintaining the tip-tip distance of 5 µm. An abrupt increase of the current in one direction was observed in some cases, as shown in Figure 6, although it is quite difficult to achieve reproducibility. This result suggested that a significant current increase occurred when the direction of the two tips matched that of the PDA backbones. Notice that the distance dependence was employed first, because the angle dependence was employed in the 5 µm of tip-tip distance. If the angle dependence is carried out first, the film might be destroyed and the distance dependence cannot be measured precisely. Figure 7 shows the results of the resistance measurements of regiorandom POT thin films as functions of the tip-tip distance. Regiorandom POT thin films are regarded as two-dimensional uniform conductive films,21 and a logarithmic function was clearly observed. These results for POT thin films were clearly different from those for PDA thin films. For PDA thin films, the value of the resistance at L ) 1 µm is considered to be the contact resistance between each tip electrode and the PDA thin film, and was estimated to be about 103 GΩ. Concerning these relatively high values, the condition of tip contact with the film includes the destruction of the PDA thin film at the contact point, and there may exist not chemical contact but wide tunneling gaps between the tip and the film. The conductivity of the PDA thin films can be estimated from σ ) L/(RS), where L, R, and S are the tip-tip distance, resis-

16356 J. Phys. Chem. B, Vol. 108, No. 42, 2004

Takami et al. induce a significant conductivity improvement and enhance the possibility of using PDA for a molecular wire, so that PDA might be a promising candidate for the material of an organic molecular wire. Conclusion

Figure 7. Variation of the resistance between two tips for poly(3-octylthiophene) thin films as a function of the tip-tip distance.

We evaluated the conductivity of PDA thin films on a micrometer scale, using an independently driven DT-STM. The AFM observation revealed the PDA backbones to be in the same direction in one domain, and the typical domain size to be 2-20 µm. The resistance of the PDA thin films is proportional to the tip-tip distances, and the conduction along the PDA backbones is one-dimensional. The conductivity of PDA thin films was estimated to be about (3-5) × 10-6 S/cm, which is comparable to that of polyacetylene. We believe that a more precise conductivity of PDA thin films was evaluated using the DTSTM compared with previous data.12 The DT-STM is a very effective tool for measurement on a micrometer scale. Significant progress can be expected in nanometer-scale measurement using this DT-STM. Also, PDA is a promising material for conductive molecular wires in molecular electronics because its conductivity is comparable to that of other conducting polymers. Acknowledgment. This work is supported by a Grant-in Aid for Scientific Research (A) (No. 11305007) from the Ministry of Education, Culture, Sports, Science and Technology, Japan, and partly supported by the ICORP program, Japan Science and Technology Agency. References and Notes

Figure 8. Nondoped conductivities of specified materials. Day and Lando’s data,12 our data for the PDA thin films, and the value for polyacetylene16 are also indicated. Abbreviations: TTF, tetrathiafulvalene; TCNQ, tetracyanoquinodimethane; PPy, polypyrrole; PET, poly(ethylene terephthalate); PE, polyethylene.

tance, and cross-sectional area of the contact area, respectively. The cross-sectional area obtained from the estimated PDA film thickness and the diameter of the contact areas was about 7.5 × 10-3 µm2. The conductivity of the PDA thin films was estimated to be about (3-5) × 10-6 S/cm in our experiments. It is surprising that this value of the conductivity of the PDA film is about 5 orders of magnitude higher than that in the previous report12 and is almost the same as that of nondoped polyacetylene.16,17 It is strongly suggested that the conductivity in this work reflects the intrinsic conductivity of pristine PDA thin films because the tip-tip distance is small enough to detect the conductivity of the highly oriented PDA thin films without domain boundaries. The conductivity of PDA thin films obtained in our experiment was compared with those of other materials, as shown in Figure 8. Day and Lando’s data are also included. The conductivity of the PDA thin film in our experiment is in the same range as those of semiconductors such as Si and Ge. On the contrary, conductivities of other typical organic molecular materials are less than 10-10 S/cm, except for tetrathiafulvalenetetracyanoquinodimethane (TTF-TCNQ). Moreover, doping an impurity such as iodine in the case of polyacetylene would

(1) Nalwa, H. S., Ed. Handbook of Organic ConductiVe Molecules and Polymers; John Wiley: New York, 1998. (2) See, for example, Ogawa, T. Prog. Polym. Sci. 1995, 20, 943. (3) Lio, A.; Reichert, A.; Nagy, J. O.; Salmeron, M.; Charych, D. H. J. Vac. Sci. Technol. 1996, B14, 1481. (4) Nezu, S.; Lando, J. B. J. Polym. Sci. 1995, A33, 2455. (5) Batchelder, D. N.; Evans, S. D.; Freeman, T. L.; Ha¨ussling, L.; Ringsdorf, H.; Wolf, H. J. Am. Chem. Soc. 1994, 116, 1050. (6) Grim, P. C. M.; Feyter, S. D.; Gesquie`re, A.; Vanoppen, P.; Ru¨cker, M.; Valiyaveettil, S.; Moessner, G.; Mu¨llen, K.; Schryver, F. C. D. Angew. Chem., Int. Ed. Engl. 1997, 36, 2601. (7) Okawa, Y.; Aono, M. Nature 2001, 409, 683. (8) Okawa, Y.; Aono, M. RIKEN ReV. 2001, 37, 3. (9) Akai-Kasaya., M.; Shimizu, K.; Watanabe, Y.; Saito, A.; Aono, M.; Kuwahara, Y. Phys. ReV. Lett. 2003, 91, 255501. (10) Kuwahara, Y.; Akai-kasaya, M.; Aono, M. Oyo Butsuri 2003, 72, 1291 (in Japanese). (11) Baughman, R. H.; Chance, R. R. Ann. N. Y. Acad. Sci. 1978, 313, 705. (12) Day, D. R.; Lando, J. B. J. Appl. Polym. Sci. 1981, 26, 1605. (13) Day, D.; Lando, J. B. Macromolecules 1980, 13, 1478. (14) Shimada, S.; Masaki, A.; Matsuda, H.; Okada, S.; Nakanishi, H. Mol. Cryst. Liq. Cryst. 1998, 315, 83. (15) Sakamoto, M.; Wasserman, B.; Dresselhaus, M. S.; Wnek, G. E.; Elman, B. S.; Sandman, D. J. J. Appl. Phys. 1986, 60, 2788. (16) Shirakawa, H.; Ito, T.; Ikeda, S. Makromol. Chem. 1978, 179, 1565. (17) Hatano, M.; Kambara, S.; Okamoto, S. J. Polym. Sci. 1961, 51, S26. (18) Blodgett, K. B. J. Am. Chem. Soc. 1935, 57, 1007. (19) See, for example, Tachibana, H.; Yamanaka, Y.; Sakai, H.; Abe, M.; Matsumoto, M. Langmuir 2000, 16, 2975. (20) Carpick, R. W.; Sasaki, D. Y.; Burns A. R. Langmuir 2000, 16, 1270. (21) Bolognesi, A.; Botta, C.; Mercogliano, C.; Marinelli, M.; Porzio, W.; Angiolini, L.; Salatelli, E. Polym. AdV. Technol. 2003, 14, 537.