Monolayer to Multilayer Nanostructural Growth Transition in N-Type

NANO LETTERS

Monolayer to Multilayer Nanostructural Growth Transition in N-Type Oligothiophenes on Au(111) and Implications for Organic Field-Effect Transistor Performance

2006 Vol. 6, No. 11 2447-2455

Geetha R Dholakia* and M. Meyyappan Center for Nanotechnology, NASA Ames Research Center, Moffett Field, California 94035-1000

Antonio Facchetti* and Tobin J. Marks* Department of Chemistry and the Materials Research Center, Northwestern UniVersity, EVanston, Illinois 60208-3113 Received July 7, 2006; Revised Manuscript Received September 12, 2006

ABSTRACT The evolution in growth morphology and molecular orientation of n-type semiconducting r,ω-diperfluorohexyl-quaterthiophene (DFH-4T) on Au(111) is investigated by scanning tunneling microscopy and scanning tunneling spectroscopy as the film thickness is increased from one monolayer to multilayers. Monolayer-thick DFH-4T films are amorphous and morphologically featureless with a large pit density, whereas multilayer films exhibit drastically different terraced structures consisting of overlapping platelets. Large changes in DFH-4T molecular orientation are observed on transitioning from two to four monolayers. Parallel electrical characterization of top-versus-bottom contact configuration DFH-4T FETs with Au source/drain electrodes reveals greatly different mobilities (µTOP ) 1.1 ± 0.2 10-2 cm2V-1s-1 versus µBOTTOM ) 2.3 ± 0.5 10-5 cm2V-1s-1) and contact resistances (RC-TOP ) 4−12 MΩcm vs RC-BOTTOM > 1 GΩcm). This study provides important information on the organic semiconductor-source\drain electrode interfaces and explains why top-contact OFET devices typically have superior performance. By direct visualization, it demonstrates that the DFH-4T film growth transition from monolayer to multilayer on Au is accompanied by dramatic morphology and molecular orientation changes, starting from an amorphous, pitted, and disordered monolayer, to crystalline and smooth bi/tetralayers but with the molecules reoriented by 90°. These chemisorption-derived inhomogenities at the contact−molecule interface and the large monolayer f multilayer f bulk microstructural changes are in accord with the large bottom-contact device resistance and poor OFET performance.

Introduction. Electronics based on organic semiconductors offer great potential in applications requiring mechanical flexibility and large area coverage, and offer the added attraction of low fabrication costs.1 The broad potential application range includes low-end display/identification/data storage devices fabricated with circuits based on organic field-effect transistors (OFETs).2 Such technologies require organic semiconductor deposition on a variety of substrates, where the semiconductor may exhibit very different film growth mechanisms and microstructures.1,2 This aspect is particularly critical for OFETs where, depending on the device structure, the organic semiconductor must be deposited on materials as diverse as conducting source/drain * Corresponding authors. E-mail: [email protected]; [email protected]; [email protected]. 10.1021/nl061566+ CCC: $33.50 Published on Web 10/24/2006

© 2006 American Chemical Society

contacts (typically metals such as Au, but also conducting polymers) and/or insulators (typically oxides such as SiO2 or Al2O3, but also polymers or self-assembled mono(multi)layers).3 Although the carrier mobilities achieved for organic semiconductor-based FETs have increased dramatically over the past several years,1,4 the mobility of a given semiconductor depends strongly on factors such as film texture, molecular orbital energy alignment/shift with the contacts, and film morphology, which in turn strongly depend on film growth mechanism and interfacial interactions with the underlying substrate.5 Unlike band structure-governed transport in conventional inorganic semiconductors,6 hopping transport predominates in most organic semiconductors, modulated by π-π overlap between neighboring molecular/ polymeric fragments and structural reorganization (lattice deformation) processes. For example, differences in π-π

stacking orientation with respect to the FET dielectric surface may result in several orders of magnitude differences in mobility.2,7 Furthermore, grain boundary and defect densities in organic films act as scattering centers and charge traps, strongly reducing the mobility. For all of these reasons, optimizing film growth parameters to minimize chargetrapping sites within the semiconductor films deposited on various surfaces is crucial to OFET performance. Thus, it is essential to understand the mechanism of organic film growth on metallic and dielectric surfaces. There have been several studies on the effects of growth conditions, especially temperature and dielectric surface functionalization, on OFET semiconductor film morphology and microstructure,8 including topographic imaging of monolayer to multilayer evolution, focusing on prototype molecules such as pentacene and sexithiophene on SiO2.9 However, very few studies have investigated organic semiconductor growth on substrates used as OFET metal contacts, relevant to bottom-contact devices, and none have detailed the film growth characteristics of n-channel materials or related such results to OFET function. The motivation here is to understand how morphological features of organic semiconducting films on metal contacts evolve on the nanoscale level as film thickness increases from one monolayer to multilayers to typical OFET thicknesses. We seek correlations between film nanostructural features such as orientation, molecular alignment on metal contacts, the presence of structural defects, grain boundaries, and other scattering centers with OFET charge transport characteristics. The material chosen is a fluorocarbon-substituted quaterthiophene (R,ω-diperfluorohexyl-4T), DFH-4T, ideal for this investigation because it exhibits excellent n-type transport in top-contact TFTs10 but, as for many n-type fluorocarbonsubstituted semiconductors,11 poor charge transport in bottom-contact devices (vide infra). The nature of

the organic semiconductor-source/drain electrode interface and changes in film morphology affecting electron injection and transport in the transition from a monolayer to multilayers are probed by scanning tunneling microscopy (STM). We also present tunneling spectroscopic studies on these films and estimate the band gap from the tunneling conductance. In parallel, we also analyze charge transport and contact resistance characteristics of both DFH-4T-based topand bottom-contact FETs. The results provide direct and generalizable information as to why DFH-4T bottom contact devices do not perform optimally versus their top-contact counterparts. Experimental Details. DFH-4T synthesis and purification are described elsewhere.10 OFETs were fabricated by evaporating DFH-4T (0.2-0.4 Ås-1, p ≈ 10-6 Torr, 50 nm) onto p+-Si/SiO2 (300 nm oxide) substrates at 70 °C and onto Au(111) for STM measurements. For bottom- and top-contact devices, Au source and drain electrodes (50 nm) were vapordeposited through a shadow mask before and after DFH-4T 2448

deposition, respectively. Mobilities (µ) were calculated in the saturation regime from: µsat ) (2IDSL)/[WCox(VG - Vth)2] where IDS is the source-drain saturation current; W (0.55.0 mm) and L (25-300 µm) are the channel width and length, respectively; Cox (10 nF/cm2) is the oxide capacitance, VG is the gate voltage, and Vth is the threshold voltage. OTFT measurements were carried out in air using a Keithley 6430 subfemtoammeter and Keithley 2400 source meter, operated by a local Labview program and GPIB communication. Triaxial and/or coaxial shielding was incorporated into the Signaton probe station to minimize noise. Nanostructural evolution in four DFH-4T films of thicknesses 3, 5, 10, and 12 nm grown on flame-annealed Au(111)/mica substrates was studied by scanning tunneling microscopy (Oxford-STM) using cut Pt/Ir tips, with the bias voltage applied to the sample. All of the measurements were performed in vacuum (∼10-6 Torr). Topographic images were obtained in the constant current mode with a bias of 1.0 V and a typical set point tunnel current of 10-100 pA, which ensures a large tunnel junction impedance of 10010 GΩ, and avoids destructive tip-surface interactions. The density of states (DOS) of DFH-4T\Au(111) was probed by scanning tunneling spectroscopy (STS). Typical STS I-V curves were obtained in vacuum with a set point bias of 0.50 V and a set point current of 50 pA, corresponding to a junction impedance of 10 GΩ. Each I-V spectrum is recorded as an average of 7 individual curves. AFM images were obtained using a JEOL-5200 Scanning Probe Microscope with Si cantilevers in the tapping mode using WinSPM Software. Thin films of organic semiconductors were analyzed by standard wide-angle θ-2θ X-ray film diffractometry (WAXRD) using monochromated Cu KR radiation. Results and Discussion. The DFH-4T crystal structure illustrating molecular orientation and packing is shown in Figure 1. This will be useful in understanding the oligothiophene film growth process to be discussed below. We first discuss the properties of top- and, for the first time, bottomcontact FETs having the following structures (Figure 2A): (i) Top-contact: p+-Si (gate)/300 nm SiO2 (dielectric)/30 nm DFH-4T/50 nm Au (source, drain); (ii) Bottom-contact: p+Si (gate)/300 nm SiO2 (dielectric)/50 nm Au (source, drain)/ 30 nm DFH-4T. Representative OFET transfer plots are shown in Figure 2B. From the plots in Figure 2B and C, the field-effect mobility (µ), current on-off ratio (Ion/Ioff), threshold voltage (VT), and turn-on voltage (Von) are extracted. The measured electron mobilities for the top- (µTOP) and bottom-contact (µBOTTOM) devices are 1.1 ( 0.2 10-2 cm2V-1s-1 and 1.0 ( 3.5 10-5 cm2V-1s-1, respectively. The other FET parameters are: (Ion/Ioff)TOP ) 4 ( 2 107; (Ion/ Ioff)BOTTOM ) 7 ( 25 103; (VT)TOP ) 64 ( 7 V; (VT)BOTTOM ) 66 ( 11 V; (Von)TOP ) 20 V; (Von)BOTTOM ) 20-30 V. These data underscore the superior performance metrics of top-contact devices. Note that the difference between VT (∼65 V) and Von (∼20 V) for optimum samples is identical for both device structures (∆V ≈ 45 V), suggesting similar dielectric-semiconductor interfacial trap densities.12 Therefore, the current output variations must reflect different electron injection efficiencies from the Au contacts and not, Nano Lett., Vol. 6, No. 11, 2006

Figure 1. Crystal structure of DFH-4T with views along the b (A) and c (B) axes.

because DFH-4T films were deposited in parallel, from variations in DFH-4T film quality or interface with the dielectric. For full device characterization, the specific contact resistance for top-contact FETs (RC-TOP) was obtained from the y intercept (for a given VG) of the measured total device specific resistance, R, as a function of channel length.13 From Figure 2D, RC-TOP ) 4-12 × 106 Ωcm (VG ) 60-100 V) and decreases as VG is increased. These values can be compared to those reported for p-type pentacene (2 × 103 Ωcm)14 and thiophene-acenes (2-3 × 104 Ωcm),15 and n-type PDCDI-C5-Au (7 × 104 Ωcm)16 with Au electrodes (for molecular structures, see Figure S1 in the Supporting Information). For bottom-contact FETs, R cannot be measured accurately in the linear regime (VSD < VG) because of the very low currents. However, from reasonable assumptions,17 RC-BOTTOM is estimated to be >1 × 109 Ωcm. This value is much greater than that found for p-type semiconductors such as pentacene (∼2-40 × 103 Ωcm),14b,18 6T (∼1 × 105 Ωcm),19 P3HT-Au (5-20 × 105 Ωcm),20 and F8T2 (1 × 107 Ωcm) with Au contacts. To our knowledge, no RC-BOTTOM measurements have been reported for n-type organic semiconductors. These observations raise intriguing questions about the origin of such large bottom-contact resistances, the large top-versus-bottom contact resistance differences and, therefore, the OFET performance differences (∆R > 103 Ωcm, ∆µ > 103 cm2V-1s-1) for fluorinated semiconductors, and why other small molecules such as pentacene behave so differently (∆R ≈ 0-20 Ωcm, ∆µ ≈ 0-5 cm2V-1s-1).20 STM is an ideal tool for addressing these questions. In the following sections, the morphological features of the DFH-4T films on Au(111) are presented in comparable scan ranges in the order of increasing film thickness, beginning Nano Lett., Vol. 6, No. 11, 2006

with the 3 nm films. Flame-annealing of the gold surface ensures an atomically smooth Au(111) substrate21 and guarantees that morphology and rms roughness evolution observed by STM are due to inherent variations in film nanostructure arising from the DFH-4T film growth rather than to variations in the underlying substrate. From singlecrystal X-ray diffraction data, we obtain a molecular length of 3.325 nm for DFH-4T. XRD of DFH-4T thin films on Si/SiO2 reveals that the films are highly textured (Figure S2, Supporting Information; d spacing ) 2.93 nm), with the oligothiophene molecules aligned with the long molecular axes along the surface normal (see Figure 1A). Thus, a monolayer of the film should approximately be 3 nm in thickness and the films characterized here by the STM span a range from one monolayer through four multilayers. Figure 3A and B shows the nanostructure in different regions of the 3 nm DFH-4T film in the 2 µm and 1 µm scan ranges. From these images it appears that the organic molecules completely cover the gold surface and the film is featureless except for a number of 15-50 nm diameter pits, observed as dark depressions. The underlying (111) terraces present in the Au substrate intersecting at 120° angles are also visible. No grains or grain boundaries are observed, indicating that the first monolayer of oligomer, except for the presence of pits and defects, completely wets the gold substrate. The rms roughness in the 2 × 2 µm2 scan area is 2.62 nm, almost comparable to the film thickness. Figure 3C and D are images from different regions of a 5-nm-thick film. They demonstrate that the nanostructural changes in the transition from 3 f 5 nm are quite dramatic. Although the 3-nm-thick monolayer is mostly featureless and amorphous, the 5 nm film consisting predominantly of two multilayers exhibits a terraced morphology with distinct circular grains arranged in islands. Although the underlying Au(111) planes are still visible, this film does not exhibit detectable pitting. Importantly, the rms surface roughness falls to 1.5 nm. The width of the terraces typically ranges from 20 to 50 nm, and the grains vary in diameter from 100 to 300 nm. In locations where two or more terraced columns of grains intersect, there are circular 20-30 nm diameter “holes”, similar to grain boundary triple points seen in inorganic thinfilm growth.22 Unlike the terraces observed in the growth of films containing screw dislocations, the grains here in each terraced layer are separated individually into islands and have closed circumferences. The step heights of the terraces are typically 103 cm2 V-1 s-1) for a number of fluorocarbon-substituted oligothiophenes, cyanated perylenes and naphthalenes, and acenes (Facchetti, A.; Brooks, J.; Marks, T. J., manuscript in preparation). (12) ntrap ) [Ci(VT - Von)]/q. Pernstich, K. P.; Haas, S.; Oberhoff, D.; Goldmann, C.; Gundlach, D. J.; Batlogg, B.; Rashid, A. N.; Schitter, G. J. Appl. Phys. 2004, 96, 6431. (13) Luan, S.; Neudeck, G. W. J. Appl. Phys. 1992, 72, 766. (14) (a) Necliudov, P. V.; Shur, M. S.; Gundlach, D. J.; Jackson, T. N. Solid-State Electron. 2003, 47, 259. (b) Hill, I. G. Appl. Phys. Lett. 2005, 87, 163505. (15) Merlo, J. A.; Newman, C. R.; Gerlach, C. P.; Kelley, T. W.; Muyres, D. V.; Fritz, S. E.; Toney, M. F.; Frisbie, C. D. J. Am. Chem. Soc. 2005, 127, 3997. (16) Chesterfield, R. J.; McKeen, J. C.; Newman, C. R.; Frisbie, C. D.; Ewbank, P. C.; Mann, K. R.; Miller, L. L. J. Appl. Phys. 2004, 95, 6396. (17) If the linear regime current magnitude for the top- and bottom-contact devices tracks that in saturation, (Ilinear)BOTTOM e (Ilinear)TOP × 10-3. Assuming that (Rchannel)BOTTOM ) (Rchannel)TOP, it is derived that RC-BOTTOM g (RC-TOP + R) × 103. See also ref 18. (18) Klauk, H.; Schmid, G.; Radlik, W.; Weber, W.; Zhou, L.; Sheraw, C. D.; Nichols, J. A.; Jackson, T. N. Solid-State Electron. 2003, 47, 297. (19) Chwang, A. B.; Frisbie, C. D. J. Phys. Chem. B 2000, 104, 12202. (20) (a) Burgi, L.; Richards, T. J.; Friend, R. H.; Sirringhaus, H. J. Appl. Phys. 2003, 94, 6129. (b) Meijer, E. J.; Gelinck, G. H.; Van Veenendaal, E.; Huisman, B.-H.; De Leeuw, D. M.; Klapwijk, T. M. Appl. Phys. Lett. 2003, 82, 4576. (21) Dholakia, G. R.; Fan, W.; Meyyappan, M. Appl. Phys. A 2005, 80, 1215. (22) Ramaswamy, G.; Raychaudhuri, A. K.; Goswami, J.; Shivashankar, S. A. J. Appl. Phys. 1997, 82, 3797. (23) (a) Ackermann, J.; Videlot, C.; Kassmi, E. A.; Guglielmetti, R.; Fages, F. AdV. Funct. Mater. 2005, 15, 810. (b) Hajlaoui, M. E.; Garnier, F.; Hassine, L.; Kouki, F.; Bouchriha, H. Synth. Met. 2002, 129, 215. (c) Heutz, S.; Ferguson, A. J.; Rumbles, G.; Jones, T. S. Org. Electron. 2002, 3, 119. (d) Bouchoms, I. P. M.; Schoonveld, W. A.; Vrijmoeth, J.; Klapwijk, T. M. Synth. Met. 1999, 104, 175. (e) Schoonveld, W. A.; Stok, R. W.; Weijtmans. J. W.; Vrijmoeth, J.; Wildeman, J.; Klapwijk, T. M. Synth. Met. 1997, 84, 583. (24) (a) Locklin, J.; Bao, Z. Anal. Bioanal. Chem. 2006, 384, 336. (b) Loi, M. A.; Da Como, E.; Dinelli, F.; Murgia, M.; Zamboni, R.; Biscarini, F.; Muccini, M. Nat. Mater. 2005, 4, 81.

Nano Lett., Vol. 6, No. 11, 2006

(25) (a) Hawley, M.; Raistrick, I. D.; Beery, J. G.; Houlton, R. J. Science 1991, 251, 1587. (b) Gerber, C.; Anselmetti, D.;Bednorz, J. G.; Mannhart, J.; Schlom, D. G. Nature 1991, 350, 279. (c) Okawa, N.; Tanaka, H.; Akiyama, R.; Matsumoto, T.; Kawai, T. Solid State Commun. 2000, 114, 601. (d) Ramaswamy, G.; Raychaudhuri, A. K. J. Appl. Phys. 1996, 80, 4519. (26) Dong, Y.; Feenstra, R. M.; Greve, D. W.; Moore, J. C.; Sievert, M. D.; Baski, A. A. Appl. Phys. Lett. 2005, 86, 121914. (27) (a) Kang, J. H.; Zhu, X.-Y. Chem. Mater. 2006, 18, 1318. (b) France, C. B.; Schroeder, P. G.; Forsythe, J. C.; Parkinson, B. A. Langmuir 2003, 19, 1274. (28) Makinen, A. J.; Long, J. P.; Watkins, N. J.; Kafafi, Z. H. J. Phys. Chem. 2005, 109, 5790. (29) Knipp, D.; Street, R. A.; Krusor, B.; Apte, R.; Ho, J. J. Non-Cryst. Solids 2002, 299-302, 1042. (30) Poirier, G. E.; Chem. ReV. 1997, 97, 1117 and references therein. (31) Ulman, A. An Introduction to Ultrathin Organic Films from Langmuir-Blodgett to Self-Assembly; Academic Press: San Diego, CA, 1991. (32) Dholakia, G. R.; Fan, W.; Koehne, J.; Han, J.; Meyyappan, M. Phys. ReV. B 2004, 69, 153402. (33) (a) Hill, I. G.; Schwartz, J.; Kahn, A. Org. Electron. 2000, 1, 5. (b) Elfeninat, F.; Fredriksson, C.; Sacher, E.; Selmani, A. J. Chem. Phys. 1995, 102, 6153. (c) Lachkar, A.; Selmani, A.; Leclerc, E.; Mokhliss, R. Synth. Met. 1994, 66, 209. (34) Baldacchini, C.; Mariani, C.; Betti, M. G. J. Chem. Phys. 2006, 124, 154702/1. (35) Hutchison, G. R.; Ratner, M. A.; Marks, T. J. J. Am. Chem. Soc. 2005, 127, 16866. (36) Lukas, S.; Vollmer, S.; Witte, G.; Wo¨ll, Ch. J. Chem. Phys. 2001, 114, 10123. (37) Zhong, D. Y.; Lin, F.; Chi, L. F.; Wang, Y.; Fuchs, H. Phys. ReV. B 2005, 71, 125336/1. (38) Dishner, M. H.; Hemminger, J. C.; Feher, F. J. Langmuir 1996, 12, 6176. (39) (a) Sako, E. O.; Kondoh, H.; Nakai, I.; Nambu, A.; Nakamura, T.; Ohta, T. Chem. Phys. Lett. 2005, 413, 267. (b) Noh, J.; Ito, E.; Nakajima, K.; Kim, J.; Lee, H.; Hara, M. J. Phys. Chem. B 2002, 106, 7139. (c) Liu, G.; Rodriguez, J. A.; Dvorak, J.; Hrbek, J.; Jirsak, T. Surf. Sci. 2002, 505, 295. (40) (a) Liu, J.-T.; Lu, H.-J. J. Fluorine Chem. 2001, 111, 207. (b) Kurykin, M. A.; Vol’pin, I. M.; German, L. S. J. Fluorine Chem. 1996, 80, 9. (41) Facchetti, A.; Yoon, M.-H.; Stern, C. L.; Hutchison, G. R.; Ratner, M. A.; Marks, T. J. J. Am. Chem. Soc. 2004, 126, 13859. (42) (a) Minari, T.; Miyata, Y.; Terayama, M.; Nemoto, T.; Nishinaga, T.; Komatsu, K.; Isoda, S. Appl. Phys. Lett. 2006, 88, 083514/1. (b) Halik, M.; Klauk, H.; Zschieschang, U.; Schmid, G.; Ponomarenko, S.; Kirchmeyer, S.; Weber, W. AdV. Mater. 2003, 15, 917. (43) Kymissis, I.; Dimitrakopoulos, C. D.; Purushothaman, S. IEEE Trans. Electron DeV. 2001, 48, 1060.

NL061566+

2455