High-Performance, All-Solution-Processed Organic Nanowire

Nov 1, 2011 - Aligned organic NW arrays are patterned via dip coating, and the S/D electrodes are patterned by inkjet printing conducting polymer poly...
7 downloads 0 Views 3MB Size
Download PDF
LETTER pubs.acs.org/Langmuir

High-Performance, All-Solution-Processed Organic Nanowire Transistor Arrays with Inkjet-Printing Patterned Electrodes Nanliu Liu,† Yan Zhou,‡ Na Ai,† Chan Luo,† Junbiao Peng,† Jian Wang,*,† Jian Pei,*,‡ and Yong Cao† †

Institute of Polymer Optoelectronic Materials and Devices, State Key Laboratory of Physics and Chemistry of Luminescence, South China University of Technology, Guangzhou 510640, PR China ‡ Key Laboratory of Bioorganic Chemistry and Molecular Engineering of the Ministry of Education, College of Chemistry, Peking University, Beijing 100871, PR China ABSTRACT: Organic nanowire (NW) transistor arrays with a mobility of as high as 1.26 cm2 3 V 1 3 S 1 are fabricated by combining the dip-coating process to align the NW into arrays with the inkjet printing process to pattern the source/drain electrodes. A narrow gap of ∼20 μm has been obtained by modifying the inkjet process. The all-solution process is proven to be a low-cost, high-yield, simple approach to fabricating highperformance organic NW transistor arrays over a large area.

’ INTRODUCTION Recently, considerable attention has been paid to 1D nano/ microsize organic single-crystal transistors because of their potential applications as building blocks in the field of electronics owing to their high mobility as a result of high crystallinity. The field effect transistor (FET) mobility of organic nanowires (NWs) has exceeded 2 cm2 3 V 1 3 S 1,1 4 which has already met the demand for the transistors to drive the back panels in electrophoretic displays, liquid-crystal displays, and organic lightemitting displays.5,6 However, to put organic NWs in real applications, there are two obstacles that have to be overcome: how to pattern the organic NWs into aligned arrays over large areas and how to make electrodes for such arrays at low temperature with a simple, cost-effective, and scalable process. Surface energy templates, flow orientation, and solvent evaporation-induced methods have been successfully developed to pattern 1D nano- or microstructures.7 14 Inorganic nanostructure arrays with high mobility and multifunctions have been demonstrated by Lieber’s group.15 17 Unfortunately for the aligned organic 1D nanostructures, the transistor mobility of the array devices was found to be 1 order of magnitude lower than that of the single-wire devices because of overlapped NWs between the electrodes and the low contact quality of the insulator/semiconductor and semiconductor/electrode interfaces.18 21 The interface issue comes from the FET fabrication method. Because the organic NWs are mechanically fragile and incompatible with conventional lithography techniques, the most common way to fabricate organic NW FETs is to deposit NWs directly onto the prepatterned source/drain (S/D) electrodes. The whole process involves facile lithography for S/D and multitime surface modifications for both the electrodes and the dielectrics, which leads to large contact resistance generally associated with the bottom contact FET configuration.22,23 It is very desirable to develop a simple process to pattern the S/D r 2011 American Chemical Society

electrodes without affecting the performance of the organic NW arrays. Herein, we present an all-solution processing method for assembling the organic NWs into aligned arrays and completing the S/D electrodes on top of the array. Aligned organic NW arrays are patterned via dip coating, and the S/D electrodes are patterned by inkjet printing conducting polymer poly(3,4ethylenedioxythiophene)-poly(styrene sulfonate) (PEDOT/ PSS) (Figure 1). By combining the dip-coating process with the inkjet process, an FET mobility of as high as 1.26 cm2 3 V 1 3 S 1 was achieved for the organic NW transistor arrays. Not only is this the first organic FET array device with printing patterned electrodes, but also the mobility reaches a level that is comparable to that of the device with evaporated electrodes. Our bottom-up nonphotolithographic all-solution scalable approach has opened a new, simple route for cost-effectively fabricating high-performance organic NW transistor arrays in the applications of microelectronics.

’ RESULTS AND DISCUSSION To demonstrate the all-solution process, organic compound 1 (Figure 1, inset) was synthesized and used. The π π stacking between the large aromatic planes and the van der Waals interactions between the long alkyl chains enable the molecules to self-assemble into 1D nano- or microstructures.24,25 We have demonstrated a hole mobility of 2.1 cm2 3 V 1 3 S 1 on a singlewire FET device by evaporating asymmetric gold electrodes.26 The drop-on-demand injecting technique is superior to the vacuum evaporation process in terms of cost, yield, and its noncontact maskless approach.27,28 Printing PEDOT/PSS as Received: August 24, 2011 Revised: October 20, 2011 Published: November 01, 2011 14710

dx.doi.org/10.1021/la2033324 | Langmuir 2011, 27, 14710–14715

Langmuir

Figure 1. Schematic illustration of inkjet printing S/D electrodes onto organic NWs. (Inset) Chemical structure of compound 1.

Figure 2. (a, c, e) Schematic illustration of the S/D electrodes and channel length by inkjet printing and (b, d, f) the actual photographs. (a, b) One drop jetting for each electrode; (c, d) three drops jetting for each electrode; and (e, f) three drops jetting for each electrode with a 2 ms delay. The substrate was heated to 40 °C during inkjet printing. Ds is the maximum radius of droplet spreading. Rd is the final radius after droplets are dried. Dp is the distance between S and D jetting locations. (Insets of b and d) Three-dimensional photographs of the printed electrodes.

S/D electrodes on thin film organic FET devices has been proven to be very successful by using a self-aligned printing (SAP) technique.29 33 A submicrometer gap was achieved through surface modification. However, no attempt to print electrodes has ever been tried on any NW FET devices, including organic and inorganic. The challenge of printing electrodes on NW FET devices is that a narrow channel length (the gap between S and D) is difficult to obtain. During the inkjet printing process, upon

LETTER

striking the substrate, the PEDOT/PSS droplet spreads first as driven by the kinetic energy, then shrinks because of the surface tension, and finally dries to form a stable shape.34 To use the SAP method, the first printed electrode has to be modified to become hydrophobic either with CF4 plasma or additional surfactant so that the second printed hydrophilic electrode will be naturally repelled and dewetted away, leading to a narrow gap. Unfortunately, the organic NWs are easily damaged by the CF4 plasma, and when surfactants were added, printed PEDOT/PSS would spread along the NW, resulting in connected electrodes. In our experiment, high-conductivity PEDOT/PSS PH500 (conductivity > 1 S/cm) was selected. The substrate was heated to 40 °C to facilitate solvent evaporation and force the droplet to dry fast. A 40-μm-diameter PH500 droplet was ejected from a 30-μm-diameter nozzle. The distance between the drops’ jetting positions (Dp) must be more than 80 μm, for at values less than that the drops could not separate because the maximum radius of droplet spreading (Ds) was about 35 μm (the additional 10 μm was needed for error tolerance). After each droplet dried, the radius of the droplet (Rd) shrank to 10 μm, which left the S/D electrode gap at about 60 μm (Figure 2a,b). To reduce the channel length while keeping Dp the same, we printed three drops at each S/D position. Because the spreading of the droplet was mainly driven by the kinetic energy, which was same for all of the drops, the successive drop would not spread beyond the spreading boundary of the former drop. As a result, the maximum spreading radius of three overlapped drops was still about 35 μm. During the subsequent process of drying, because the volume of the materials was tripled, the radius of the driven droplet increased to 25 μm. As clearly shown in the 3D photographs (insets in Figure 2b,d), although the surface profiles of the one-drop electrode and the three-drop electrode are almost identical, the radius of the three-drop electrode is much larger than that of the one-drop electrode, thereby making a 30 μm gap as illustrated in Figure 2c,d. As shown above, the channel length is mainly determined by Ds and Rd. Dp should be larger than 2Ds to avoid electrodes shorting. If we waited for the first electrode to dry before we printed the second electrode, then Dp as well as the gap could be reduced. As demonstrated in Figure 2e,f, we delayed 2 ms between each drop jetting and shortened Dp to 70 μm. The final channel length decreased to 20 μm. Further narrowing the electrode gap is currently under investigation in our laboratory, and the performance could likely be increased further. All of the single-wire transistors with inkjet-printed PEDOT as the S/D electrodes show typical p-channel FET characteristics with an average mobility of 0.29 cm2 3 V 1 3 S 1, whereas the highest hole mobility reaches 0.94 cm2 3 V 1 3 S 1 (Figure 3a,b). The threshold is 4 V, and the on/off ratio is ∼104 for the highest-performance device with a width to length ratio of 0.0125. The performance is on par with that of the devices with evaporated gold electrodes.3 The experiments demonstrate that patterning the electrodes with inkjet-printing-conductive polymer is a low-cost, high-yield approach to fabricate high-performance organic NW FETs over a large area. To assemble the organic NWs into aligned arrays, a dipcoating process was used (Figure 4a). During the process, the evaporation rate of solvent should be adjusted to obtain the best aligned NW arrays.35 If the evaporation rate was fast, then the capillary flow would carry many NWs into the contact line within a short time and those NWs would be immediately washed away by the dewetting force before being aligned. Hexane is the best 14711

dx.doi.org/10.1021/la2033324 |Langmuir 2011, 27, 14710–14715

Langmuir

Figure 3. Best performances of the single-crystal NW and NW array devices. (a) Output curve and (b) transfer curve of the single-wire NW FET. (c) Output curve and (d) transfer curve of the NW array FET device.

solvent for fabricating single-crystal NWs.26 However, hexane has a low boiling point of 65 °C, which causes the dip-coated NWs to be aligned in a disorderly manner as shown in Figure 4c,e. To slow down the fast evaporation, we covered the vessel

LETTER

Figure 4. Schematic illustrations of the dip-coating process with (a) fast solvent evaporation and (b) controlled solvent evaporation. Photographs of dip-coated NW arrays: (c, d) the pulling speed is 25 μm/s, and the sticking time is 10 s; (e, f) the pulling speed is 25 μm/s, and the sticking time is 30 s. (c, e) Under fast solvent evaporation and (d, f) under slow solvent evaporation. Photographs of the patterned device matrix with inkjet-printed electrodes: g is from e, and h is from f. i is a zoom-in image of h.

with two pieces of glass as lids, therefore controlling the evaporation rate by precisely adjusting the gap between the two lids (Figure 4b). Under such a setup, the parallel organic NW 14712

dx.doi.org/10.1021/la2033324 |Langmuir 2011, 27, 14710–14715

Langmuir

LETTER

Figure 6. Photograph of amorphous materials formed during spin coating.

Figure 5. Histogram of the NW angular distribution of 40 NWs. (a, b) The pulling speed is 25 μm/s, and the sticking time is 10 s; (c, d) the pulling speed is 25 μm/s, and the sticking time is 30 s. (a, c) Under fast evaporation and (b, d) under slow evaporation.

arrays were successfully obtained. As shown in Figure 4d,f, the NW arrays with a low solvent evaporation rate are packed more densely and more orderly than those assembled with a high solvent evaporation rate. The distance between neighboring NWs is more than 50 μm, when the sticking time is 10 s under fast solvent evaporation (Figure 4c), whereas it is shortened to about 30 μm under slow solvent evaporation (Figure 4d). The density of the array is around 10/100 μm when the sticking time increases to 30 s with a high evaporation rate (Figure 4e), whereas it becomes 15/100 μm under a low evaporation rate (Figure 4f). Figure 5 shows the distribution of the NW alignment angles with respect to the pulling direction for 40 NWs under different processing conditions. The NW angular distribution becomes narrower as the sticking time increases and the rate of solvent evaporation decreases. After the organic NW arrays were dip coated, the PEDOT/ PSS S/D electrode matrix was patterned following the printing procedure on the single-wire NW device described earlier (i.e., 3 drops per electrode with an S/D distance of 80 μm). The channel length is therefore about 30 μm as shown in Figure 4g,h. Figure 4i shows a 30  50 electrode array over a 2 mm  3.5 mm area. The size of the organic NW transistor arrays is decided by the density of the printed electrodes. There are a total of 750 independent transistors formed by the 30  50 electrodes. Before being tested, the device arrays were baked in a vacuum oven at 70 °C for 12 h. All of the devices in arrays exhibit p-channel transistor characteristics with an average hole mobility of 0.51 cm2 3 V 1 3 S 1. The best device with a channel width to length ratio of 0.02 has a mobility of as high as 1.26 cm2 3 V 1 3 S 1, which is even higher than that of the best single-wire NW device. The threshold is 2.6 V, and the on/off ratio reaches 105 for the best array device shown in Figure 3c,d. The better performance in terms of mobility, threshold, and on/off ratio achieved in the array devices compared to that in the single-wire NW devices can be attributed to the better contact quality of the nanowire/electrode and dielectric/nanowire interfaces due to the natural solvent evaporation realized in the dipcoating process. By carefully examining the substrate surface after spin coating the NW suspension under a microscope, we found some amorphous stains around the NWs (Figure 6). It is suspected that the stains come from the dissolved compound 1 molecules in the suspension. Those amorphous residues become contaminants at the interfaces between the organic semiconductor and the dielectric layer and between the electrodes and the semiconductor, which damages the organic FET performance.26 In the array devices, no such amorphous stains have been observed. The slow natural solvent evaporation during the 14713

dx.doi.org/10.1021/la2033324 |Langmuir 2011, 27, 14710–14715

Langmuir dip-coating process may prevent the compound 1 molecules from precipitating onto the substrate. In conclusion, we have developed an all-solution process for achieving high-performance organic NW array transistors by dip coating NWs into arrays and inkjet printing conductive polymers to pattern the S/D electrodes. Our achievement demonstrates the first organic NW FET array devices obtained by printing patterned electrodes. Through the modification of the inkjet process, a channel length of ∼20 μm is obtained. When the solvent evaporation rate is controlled, the NW arrays become aligned in a more orderly manner. The state of the art mobility of 1.26 cm2 3 V 1 3 S 1 of the array devices not only is on par with that of the device with evaporated metal electrodes but also is better than that of the single-wire NW device. The all-solution process is proven to be a low-cost, high-yield, simple approach to fabricate high-performance organic NW FET arrays over a large area.

’ EXPERIMENTAL SECTION Materials. Following an earlier publication,26 we dissolved 2 mg of

compound 1 in 1 mL of THF. The solution was heated to 50 °C for 5 min and was filtered through a 0.45 μm filter. The clear solution was kept under 25 °C for about 1 week to allow the formation of the 1D nano/ microbelts. The suspensions were then dispersed into three times the volume of hexane. All of the solvents were purchased from Aldrich, Inc. and Acros Organics without further purification. Device Fabrication and Characterization. The SiO2-coated Si substrates (1 cm  1.5 cm) were cleaned in an ultrasonic bath sequentially with acetone, isopropanol, detergent, deionized water, and isopropanol for at least 10 min in each step. Prior to PMMA coating, the substrates were subjected to oxygen plasma for 15 min. The 1,2-dichlorobenzene PMMA solution was spin coated onto SiO2. Finally, the substrates were annealed in a vacuum oven at 70 °C for 3 h. To make the single-wire NW device, the NW suspension was spin coated onto a PMMA-coated SiO2/Si substrate. During the dip-coating process, the substrate’s movement was provided by a high-precision linear motor stage with the minimum incremental motion of 2.5 μm (TSA400-B) from Zolix Instruments Co. Ltd. The angles of NWs with respect to the pulling direction were calculated with an E ruler. PEDOT printing was carried out with an inkjet printer (Jetlab II) from Microfab Technologies, Inc. The printed electrodes’ surface morphologies were obtained from a Dektak 150 surface profiler from Veeco Instruments, Inc. The 3D photographs were recorded with an NT 9300 surface profiler from Veeco Instruments, Inc. The optical microscopy images were taken with a Nikon Eclipse E600 POL with a DXM1200F digital camera. The transistor characteristics of the devices were measured using an Agilent 4155C semiconductor parameter analyzer connected to a Cascade manual probe station. All of the processing and tests were carried out under an ambient atmosphere.

’ AUTHOR INFORMATION Corresponding Author

*(J.W.) Phone: 86-20-8711-4525. Fax: 86-20-8711-0606. E-mail: [email protected]. (J.P.) Phone: 86-10-6275-8745. E-mail: [email protected].

’ ACKNOWLEDGMENT We thank the Ministry of Science and Technology (973 Program 2009CB623604, 2009CB930604, and 863 Program 2008AA03A311), the National Natural Science Foundation of

LETTER

China (61076116), and the Fundamental Research Funds for the Central Universities (20092M0111) for their financial support.

’ REFERENCES (1) Briseno, A. L.; Mannsfeld, S. C. B.; Ling, M.; Liu, S.; Tseng, R. J.; Reese, C.; Roberts, M. E.; Yang, Y.; Wudl, F.; Bao, Z. Nature 2006, 444, 913. (2) Podzorov, V.; Menard, E.; Borissov, A.; Kiryukhin, V.; Rogers, J. A.; Gershenson, M. E. Phys. Rev. Lett. 2004, 93, 086602. (3) Zhang, L.; Tan, L.; Wang, Z.; Hu, W.; Zhu, D. Chem. Mater. 2009, 21, 1993. (4) Zhang, Y.; Dong, H.; Tang, Q.; He, Y.; Hu, W. J. Mater. Chem. 2010, 20, 7029. (5) Daniel, J.; Arias, A. C.; Russo, B.; Krusor, B. SID Int. Symp. Dig. Tech. Pap. 2009, 11, 660. (6) Souk, J. H.; Roh, N. S. SID Int. Symp. Dig. Tech. Pap. 2009, 11, 622. (7) Huang, J.; Fan, R.; Connor, S.; Yang, P. Angew. Chem., Int. Ed. 2007, 46, 2414. (8) Liu, S.; Wang, W. C. M.; Mannsfeld, S. C. B.; Locklin, J.; Erk, P.; Gomez, M.; Richter, F.; Bao, Z. Langmuir 2007, 23, 7428. (9) Mannsfeld, S. C. B.; Sharei, A.; Liu, S.; Roberts, M. E.; McCulloch, I.; Heeney, M.; Bao, Z. Adv. Mater. 2008, 20, 4044. (10) Zhang, C.; Zhang, X.; Zhang, X.; Fan, X.; Jie, J.; Chang, J.; Lee, C. S.; Zhang, W.; Lee, S. T. Adv. Mater. 2008, 20, 1716. (11) Briseno, A. L.; Mannsfeld, S. C. B.; Jenekhe, S. A.; Bao, Z.; Xia, Y. Mater. Today 2008, 11, 38. (12) Zhang, C.; Zhang, X.; Zhang, X.; Ou, X.; Zhang, W.; Jie, J.; Chang, J.; Lee, C. S.; Lee, S. T. Adv. Mater. 2009, 21, 4172. (13) Liu, S.; Wang, W. C. M.; Briseno, A. L.; Mannsfeld, S. C. E.; Bao, Z. Adv. Mater. 2009, 21, 1217. (14) Liu, N.; Zhou, Y.; Wang, L.; Peng, J.; Wang, J.; Pei, J.; Cao, Y. Langmuir 2009, 25, 665. (15) Duan, X.; Huang, Y.; Lieber, C. M. Nano Lett. 2002, 2, 487. (16) Lu, W.; Lieber, C. M. J. Phys. D: Appl. Phys. 2006, 39, 387. (17) Whang, D.; Jin, S.; Wu, Y.; Lieber, C. M. Nano Lett. 2003, 3, 1255. (18) Liu, S.; Tok, J. B. H.; Locklin, J.; Bao, Z. Small 2006, 2, 1448. (19) Briseno, A. L.; Mannsfeld, S. C. B.; Lu, X.; Xiong, Y.; Jenekhe, S. A.; Bao, Z.; Xia, Y. Nano Lett. 2007, 7, 668. (20) Oh, J. H.; Lee, H. W.; Mannsfeld, S.; Stoltenberg, R. M.; Jung, E.; Jin, Y.; Kim, J. M.; Yoo, J. B.; Bao, Z. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 6065. (21) Tong, Y.; Tang, Q.; Lemke, H. T.; Moth-Poulsen, K.; Westerlund, F.; Hammershoj, P.; Bechgaard, K.; Hu, W.; Bjornholm, T. Langmuir 2010, 26, 1130. (22) Roichman, Y.; Tessler, N. Appl. Phys. Lett. 2002, 80, 151. (23) Ji, H.; Hu, J.; Tang, Q.; Hu, W.; Song, W.; Wan, L. Adv. Mater. 2006, 18, 2753. (24) Niu, Q.; Zhou, Y.; Wang, L.; Luo, C.; Luo, J.; Peng, J.; Cao, Y.; Pei, J.; Wang, J. Langmuir 2010, 26, 5213. (25) Niu, Q.; Zhou, Y.; Wang, L.; Peng, J.; Wang, J.; Pei, J.; Cao, Y. Adv. Mater. 2008, 20, 964. (26) Zhou, Y.; Lei, T.; Wang, L.; Pei, J.; Cao, Y.; Wang, J. Adv. Mater. 2010, 22, 1484. (27) Zhao, N.; Chiesa, M.; Sirringhausa, H.; Li, Y.; Wu, Y J. Appl. Phys. 2007, 10, 064513. (28) Ha, M.; Xia, Y.; Green, A. A.; Zhang, W.; Renn, M. J.; Kim, C. H.; Hersam, M. C.; Frisbie, C. D. ACS Nano 2010, 4, 4388. (29) Sele, C. W.; von Werne, T.; Friend, R. H.; Sirringhaus, H. Adv. Mater. 2005, 17, 997. (30) Piliego, C.; Mazzeo, M.; Cortese, B.; Cingolani, R.; Gigli, G. Org. Electron. 2008, 9, 40. (31) Hubler, A. C.; Schmidt, G. C.; Kempa, H.; Reuter, K.; Hambsch, M.; Bellmann, M. Org. Electron. 2011, 12, 419. 14714

dx.doi.org/10.1021/la2033324 |Langmuir 2011, 27, 14710–14715

Langmuir

LETTER

(32) Caironi, M.; Gili, E.; Sakanoue, T.; Cheng, X.; Sirringhaus, H. ACS Nano 2010, 4, 1451. (33) Tseng, H. Y.; Subramanian, V. Org. Electron. 2011, 12, 249. (34) Ikegawa, M.; Azuma, H. JSME Int. J., Ser. B 2004, 47, 490. (35) Deegan, R. D.; Bakajin, O.; Dupont, T. F.; Huber, G.; Nagel, S. R.; Witten, T. A. Nature 1997, 389, 827.

14715

dx.doi.org/10.1021/la2033324 |Langmuir 2011, 27, 14710–14715