- Email: [email protected]
Novel low voltage and solution processable organic thin film transistors based on water dispersed polymer semiconductor nanoparticulates
Journal of Colloid and Interface Science 401 (2013) 65–69
Contents lists available at SciVerse ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Novel low voltage and solution processable organic thin film transistors based on water dispersed polymer semiconductor nanoparticulates Darmawati Darwis a,b, Daniel Elkington a, Syahrul Ulum a,b, Glenn Bryant a, Warwick Belcher a, Paul Dastoor a, Xiaojing Zhou a,⇑ a b
Centre for Organic Electronic, University of Newcastle, Callaghan, NSW 2308, Australia Department of Physics, Faculty of Science, Tadulako University, Kampus Bumi Tadulako Tondo kota Palu 94118, Indonesia
a r t i c l e
i n f o
Article history: Received 4 February 2013 Accepted 29 March 2013 Available online 8 April 2013 Keywords: Polymer nanoparticulates OTFT P3HT Surfactant SDS
a b s t r a c t Two novel organic thin film transistor structures that combine a hygroscopic insulator with the use of water-dispersed polymer nanoparticles as the active layer are presented. In the first device structure, the semiconducting layer was fabricated from a nanoparticulate suspension of poly-(3-hexylthiophene) prepared through a mini-emulsion process using sodium dodecyl sulfate as the surfactant whereas a surfactant-free precipitation method has been used for the second device structure. In both cases, fully solution processable transistors have been fabricated in a top gate configuration with hygroscopic poly(4-vinylphenol) as the dielectric layer. Both device structures operate at low voltages (0 to 4 V) but exhibit contrasting output characteristics. A systematic study is presented on the effect of surfactant on the synthesis of semiconducting nanoparticles, the formation of thin nanoparticulate films and, consequently, on device performance. Ó 2013 Elsevier Inc. All rights reserved.
1. Introduction Electronic products based on polymer and other organic materials are emerging rapidly, and encompass items such as: organic light emitting diode (OLED) displays, integrated circuits, organic photovoltaic cells and transistors [1–6]. Organic transistors are key components in many of these organic electronic devices. The main driver for this rapidly increasing interest in organic thin film transistors (OTFTs) is their ease of fabrication. In particular, the deposition and patterning of all layers in an OTFT is possible at low temperature using a combination of solution-based techniques, which makes them ideally suited for the realization of low-cost, large-area electronics on flexible substrates. As such, the inexpensive and simple processing capabilities of organic electronics means that they have the potential to replace conventional devices in many applications [4]. A key advantage of OTFT technology is that conventional printing techniques can be implemented to provide rapid mass production of devices. However, a challenge for printing OTFTs is that the organic semiconducting solutions that are used generally involve chlorinated solvents which are highly toxic and volatile [7,8]. In addition, the fabrication of high performance OTFTs requires optimization of polymer crystallinity in the active layer, which in turn is highly dependent upon the choice of solvents, deposition
⇑ Corresponding author. E-mail address: [email protected] (X. Zhou). 0021-9797/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2013.03.052
techniques and post-deposition treatments such as annealing [9]. Thus, from a fabrication point of view, the development of alternative more environmentally compatible semiconducting solutions capable of producing crystalline semiconducting polymer layers would be highly attractive. Recently, methods have been reported to prepare semiconducting polymer films through pre-formed polymer nanoparticulate dispersions. These nanoparticles can be prepared either by a precipitation approach [10,11] or a surfactant-assisted mini-emulsion method [12–15]. Previous studies have shown that it is possible to form stable suspensions with a tunable nanoparticle size using both methods, thus offering a new pathway to fabricate OTFT active layers with controlled polymer domain size and thin film morphology [10–16]. Furthermore, the polymer chains can be highly crystalline within each particle [10,11]. Many applications for organic transistors also require low voltage operation driven by the need for devices with low power consumption [17]. One approach to low-voltage OTFT operation is to use hygroscopic materials as the dielectric layer [18–20]. Poly(4vinylphenol) (PVP) is one such hygroscopic polymer and the moisture adsorbed by it plays a central role in modulating drain current under low gate voltage values [21]. The current modulation mechanism in these devices involves chemical doping of the semiconducting channel stemming from ion diffusion at the dielectricsemiconductor interface [20]. These devices operate at a few volts and have maximum output currents (ION) comparable to, or even higher than, many conventional OFETs operating at many tens of volts [11].
66
D. Darwis et al. / Journal of Colloid and Interface Science 401 (2013) 65–69
Significantly, the use of a polymer top-gate architecture in these transistors allows for a fully solution-processable device, which can ultimately be printed on flexible substrates [22]. A challenging aspect of fabricating this architecture with nanoparticulate materials is that it requires a pinhole-free semiconductor layer. We have shown that using surfactant-containing nanoparticles (NPs) we can obtain high-quality active layers and subsequently fabricate lowvoltage OTFT devices with a PEDOT:PSS top gate electrode [23]. In this work, we compare NP OTFTs fabricated with and without surfactant to probe the mechanism of operation in these devices. The NP thin film morphologies, chemical and electronic properties and performance of OTFTs fabricated from these two types of nanoparticles are characterized. We show that the performance of the surfactant-containing OTFTs is dictated by counter-ions associated with SDS in the film. By removing the surfactant from the active layer, substantial improvements in the performance of these low-voltage all-solution processable OTFTs are obtained. This novel device structure offers the twin advantages of low-voltage operation combined with the elimination of harmful solvents from the device fabrication process.
20 nm
0 Fig. 1. AFM images of the Type I (surfactant-containing) P3HT nanoparticle (NP) thin films: (a) unannealed NP film (rms roughness of 4.2 nm) and (b) NP film after annealing for 5 min at 140 °C (rms roughness of 3.5 nm), the scale bars are 50 nm.
probe microscopy system in non-contact mode. X-ray photoelectron spectroscopy (XPS) analysis was performed using a multi-technique ultrahigh-vacuum imaging XPS system (Thermo VG Scientific ESCALab 250) equipped with a hemispherical analyzer (of 150 mm mean radius) and a monochromatic Al KR (1486.60 eV) X-ray source. An ultraviolet–visible (UV–Vis) absorption spectrophotometer (Varian Cary 6000i) was used to measure the absorption of the nanoparticle suspensions and the films.
2. Materials and methods 3. Results and discussion Semiconducting polymeric nanoparticles were prepared in aqueous media via a mini-emulsion process (Type I). P3HT (Mn = 22 K, PDI 2.58, purchased from Lumtec) was dissolved in chloroform at a concentration of 30 mg/mL and then mixed into an aqueous sodium dodecyl sulfate (SDS) solution (42 mg SDS in 2.8 mL MilliQ water). A macroemulsion was then formed by stirring the solution at 1200 rpm for 1 h and this was then sonicated for 2 min to form a mini-emulsion. The mini-emulsion was then gently stirred at 60 °C to evaporate the chloroform. After evaporation of the solvent, the nanoparticle suspension was dialysed to concentrate the samples and remove excess surfactant. Stable polymer dispersions with a polymer particle size of about 60 nm, as measured by a Zetasizer Nano-ZS (Malvern Instruments, UK), were obtained. Semiconducting polymeric nanoparticles were prepared without surfactant using the direct precipitation method (Type II). P3HT was dissolved in 1 mL chloroform (0.25 wt.%) and stirred for 10 min at room temperature. This solution was transferred into a syringe and rapidly dropped into 4 mL anhydrous ethanol while simultaneously stirring the mixture (200 rpm). After injection, stable nanoparticle suspensions were obtained. Both types of suspensions were subsequently used in the fabrication of OTFTs. OTFTs were fabricated on glass substrates with pre-patterned indium-tin-oxide (ITO) source and drain electrodes. The channel length and width was 20 lm and 3 mm respectively. A 100 nm thick P3HT nanoparticle layer was spin-coated (60 s at 2000 rpm) from the suspensions onto the substrate as measured by a KLA Tencor profilometer. A hot plate with a controlled temperature was used to anneal the formed films. For the dielectric layer, PVP (Aldrich) was dissolved in ethanol at a concentration of 80 mg/ mL and then spin-coated (60 s at 2000 rpm) to a thickness of approximately 500 nm. The P3HT/PVP two-layer structures were annealed at 85 °C in air to remove any remaining solvent. Lastly, PEDOT:PSS (Aldrich, pH = 1–2) was drop-cast on the top of gate dielectric layer and dried on a hot plate at 40 °C in air. Two Keithley 2400 source meters were used for the device characterization. All output characteristic measurements were conducted at a scan rate of 0.1 V/s. All fabricated devices were measured in air immediately after drying the PEDOT:PSS layer. The relative humidity (RH) of the laboratory was consistently recorded as 50 ± 10%. Atomic force microscopy (AFM) images of the films formed from these suspensions were measured using a Cypher scanning
Fig. 1a and b show AFM images of the Type I (surfactant-containing) nanoparticle (NP) thin film before and after annealing at 140 °C respectively. A particle size distribution from 30 nm to 100 nm is observed in Fig. 1a in good agreement with transmission electron microscopy (TEM) and dynamic light scattering (DLS) measurements (see Supporting information in Figs. S1 and S2). The AFM images also show that multiple particles stack together; resulting in a continuous film with a thickness of 60 nm as measured by profilometry. Annealing at 140 °C for 5 min results in the coalescing of particles shown in Fig. 1b and, based on the AFM roughness analysis, a slightly smoother film is obtained. The surface roughness (based on RMS) of the film decreases from 4.2 nm to 3.5 nm upon annealing. Fig. 2 compares the UV–Vis spectra of the surfactant-containing NPs in suspension with that of the thin film before and after annealing at 140 °C. Consistent with previous reports, the absorption spectra of P3HT are all red-shifted compared to the absorption spectrum of P3HT in CHCl3 solution [24]. In addition, the red-shifts for solid state NP films are larger than that observed for NPs in suspension and exhibit an increased background signal. The increased red-shift in the UV–Vis spectra for the NP films can be attributed to more crystalline NPs being preferably retained during the spincasting process, whilst the higher background signal is consistent with increased optical scattering due to the particulate nature of the film. Three pronounced absorption peaks are observed for both the P3HT NPs in suspension (510 nm, 535 nm and 585 nm) and for the NP thin films (525 nm, 550 nm and 600 nm). The three adsorption features are attributed to the p–p stacking of the P3HT backbone and the intermolecular p–p stacking interaction and are consistent with the absorption features for nanoparticle P3HT reported by Millstone [11]. The pronounced vibronic shoulders at 550 nm and 600 nm indicate highly ordered polymer chains in the NPs and, given that these shoulders increase in magnitude upon annealing, shows that annealing at 140 °C enhances polymer crystallinity within the NPs themselves. X-ray photoelectron spectroscopy (XPS) was used to investigate the relative SDS:P3HT concentration ratio at the thin film surface before the PVP dielectric layer is deposited. Fig. 3a and b shows the XPS S 2p and C 1s spectra respectively. In Fig. 3a, the peak doublet located at 164.0 eV and 164.5 eV corresponds to the S 2p signal for P3HT [25] while the other doublet at 170.0 eV and 171.2 eV
D. Darwis et al. / Journal of Colloid and Interface Science 401 (2013) 65–69
Fig. 2. UV–Vis spectra of the surfactant-containing nanoparticle in a suspension solution (dashed line), a thin film before (dotted line) and after 140 °C annealing (solid line).
Fig. 3. (a)S 2p, (b) C 1s and (c) Na 1s XPS spectra of the surfactant-containing nanoparticle thin films, unannealed at room temperature (RT, solid line) and annealed at 100 °C (dash dotted line), 140 °C (dashed line) and 160 °C (dotted line).
corresponds to S 2p signal associated with the SO24 group of the SDS molecule [26]. The relative intensity at 164.0 eV and 170.0 eV from the XPS spectra indicates that a significant amount of SDS remains on the film surface both at room temperature (RT) and after annealing at 100 °C. This observation is consistent with the observed higher binding energy (285.9 eV) for the C 1s peak at RT and 100 °C
67
shown in Fig. 3b. The presence of a charged SDS layer at the surface screens the emitted photoelectrons resulting in the observed chemical shift for the C 1s peak. However, upon annealing to 140 °C, the S 2p peak intensity for P3HT becomes more pronounced whilst that of SDS is diminished, with a corresponding shift of the C 1s signal to a lower binding energy of 285.1 eV. The increase of the P3HT:SDS S 2p XPS intensity ratio indicates that, at the elevated temperatures, the SDS molecules are mobilized and move away from the surface of the P3HT NP film. This observation is consistent with previous work on PFB:F8BT NP materials which also showed that SDS migrates away from the NPs upon annealing [14]. The diminish of Na 1s XPS intensity shown in Fig. 3c upon annealing is co-occurring to that for SDS S 2p, indicating Na+ as the counter ions are still associated the SDS. Transistors incorporating an unannealed surfactant-containing NP thin film and a 140 °C annealed nanoparticle thin film as the semiconducting layer were fabricated and their output characteristics are shown in Fig. 4. The turn-on voltages of the unannealed and annealed transistors are 0.4 V and 0.8 V, respectively, and both devices operate fully between 0 V and 2 V; consistent with the low voltage operation previously reported for P3HT/PVP noncross-linked hygroscopic transistors [18–20]. Large gate leakage currents (of the order of 10% of the drain current (ID)) are also observed for both transistors, indicating that ionic transport is contributing to the turn-on current [21]. As a consequence, a relatively low current modulation ratio of less than 50 is generally obtained. The observation of a high ionic gate leakage current is consistent with the presence of residual ionic species in the active layer, which diffuse to the relevant counter electrode under the influence of the applied voltage and thus add to the on current. XPS spectra in Fig. 3c of the NP films show the presence of high surface concentrations of Na+, which is the counter ion in the SDS surfactant. As such, the deposition of a PVP dielectric layer on top of the NP active layer is likely to result in the creation of a P3HT/PVP interface that is highly doped with Na+ ions. Given the highly mobile nature of this ionic species in organic semiconductors [27] these Na+ ions will diffuse into both the active and dielectric layers and subsequently contribute to both the turn-on drain current and the leakage gate current. In contrast to the Type I NPs, P3HT NP thin films prepared from a surfactant-free suspension in ethanol require multiple spin-cast coatings to form continuous films. Indeed, optimized P3HT NP thin films required up to eight spin-cast coatings with a 3 min 100 °C annealing step in between each spin-cast, as well as a 5 min 140 °C annealing step after the last coating. Fig. 5 shows the AFM images of surfactant-free P3HT NP thin films after different numbers of coating stages and reveals how the morphology of the thin film changes upon repeated coating. From Fig. 5a, we observe that a single spin-cast coating results in isolated P3HT NPs on the substrate surface. Without the surfactant, Type II NPs are highly hydrophobic and repel each other during the spin-casting process resulting in a sparsely coated layer. In addition, the particle sizes and size distribution of the NPs in the film are relatively large and are in the region of 50–300 nm compared to 60 nm for the surfactant-containing NP film. Fig. 5b–d show that subsequent coatings result primarily in a filling in of the vacancies in the layer with little evidence of multiple layer formation. Fig. 5d indicates that eight coatings produce a relatively continuous film, but surface roughness measurements reveal that the Type II P3HT NP film is 10 times rougher (38.5 nm) than the Type I NP film (3.5 nm). It should be noted that films made from Type II nanoparticles are inherently more likely to contain pinhole defects (due to the absence of surfactant in the fabrication process), which in turn can lead to non-working devices. Fig. 6 shows the UV–Vis spectra of the surfactant-free P3HT NP film following the multiple spin casts. In accordance with the observation of the Type I P3HT NP film, three absorption peaks at 525 nm, 550 nm and 600 nm have been observed, indicating a similar conjugated ordering for the polymer chains in the NPs synthe-
68
D. Darwis et al. / Journal of Colloid and Interface Science 401 (2013) 65–69
(a)
(b)
(c)
(d)
Fig. 4. Current–voltage characteristics of thin film transistors fabricated with surfactant-containing nanoparticle: (a) ID VD from an unannealed film at different gate voltages; (b) ID VG from an unannealed film at VD = 1 V; (c) ID VD from an annealed film at different voltages; and (d) ID VG from an annealed film at VD = 1 V.
Fig. 6. UV–Vis of the surfactant-free P3HT nanoparticle thin film after (a) one coating (dash dot), (b) three coatings (dash), (c) six coatings (dot) and (d) eight coatings (solid line). Fig. 5. AFM images of the surfactant-free P3HT nanoparticle thin film surface after (a) one coating, (b) two coatings, (c) four coatings and (d) eight coatings, the scale bars are1 lm.
sized without SDS. The absorption peak locations and their relative intensity also agree with the previous observations [11]. The systematic increase in UV–Vis intensity with increasing numbers of coatings is consistent with each subsequent deposition filling vacancies in the preceding layer. A comparison of the UV–Vis absorption intensity of the P3HT surfactant-free film after eight coatings with that of the single coating surfactant-containing NP film (Fig. 2) indicates that a similar amount of P3HT material is present in both films. Fig. 7a shows the output characteristic of an OTFT fabricated from a surfactant-free P3HT NP thin film. The transistor is turned on at a gate voltage of 2.8 V, which is higher than that for a
surfactant-containing NP thin film. However, a lower leakage current is observed for the Type II P3HT NP devices resulting in an improved device performance (relative to the Type I P3HT NP devices) with a maximum drain current of 35 lA and a current modulation ratio of 200. This performance is comparable with that reported for both spin-coated and reverse gravure printed OTFT devices fabricated from P3HT in chlorinated solvents [28]. As such, it would appear that the removal of surfactant and counter ions from the active layer in the Type II P3HT NP devices reduces the associated ion contribution to the leakage current and hence reduces the offcurrent. In addition, the fact that much lower on currents (less than 1 lA) are observed for surfactant-containing OTFTs relative to those for the surfactant-free devices (a few tens of lA) suggests that the surfactant acts as a charge trap in the channel; thus reducing charge mobility.
D. Darwis et al. / Journal of Colloid and Interface Science 401 (2013) 65–69
(a)
Fig. 7. (a) Output characteristics of ID
VD at different gate voltages and (b) ID
69
(b)
VG (at VD =
The data demonstrates that the P3HT NP based hygroscopic transistors presented here exhibit much higher drain currents and current modulation ratios but much lower operation voltages relative to the NP OFETs prepared using silicon oxide and OTS as the dielectric layer reported by Millstone et al. [11]. This behavior is consistent with dependence of the current modulation on the hygroscopic dielectric, which is known to facilitate low voltage operation [18–21]. We hypothesize that the absence of a clear drain current saturation characteristic in devices fabricated from the surfactant-free NPs is most likely due to the presence of a rough P3HT/PVP interface as indicated by the rough P3HT surface shown in Fig. 5d. Further work is underway to improve the morphology of the surfactant-free NP film and further enhance device performance. 4. Conclusions Surfactant-containing and surfactant-free P3HT NPs have been synthesized through a mini-emulsion process and a precipitation method respectively. The surfactant-containing NPs form a continuous film more readily than the surfactant-free particles. In addition, surfactant can be largely removed from the film surface by annealing the film to 140 °C. Low voltage solution processable OTFTs based on a hygroscopic dielectric can be fabricated using both types of suspensions and their IV characteristics resemble the P3HT bulk film transistor for the same device architecture and dielectric material. Transistors prepared from surfactant-free NPs offer better device performance with a higher drain current and current modulation ratio, due to the elimination of surfactant and counter ions from the active layer and the associated ion contribution to the leakage current. In addition, the much lower source and drain on currents (less than 1 lA) for surfactant-containing OTFTs than that for surfactant-free OTFTs (tens of lA) suggest that the SDS and counter ion species can act as charge traps in the channel. Acknowledgments This research was supported by a Directorate General of Higher Education Indonesia (DIKTI) scholarship and the Commonwealth of Australian Research Council through Discovery Project Funding Scheme. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2013.03.052.
2.0 V) for a thin film transistor fabricated with surfactant-free P3HT NP thin film.
References [1] V. Subarmanian, J.M.J. Frechet, P.C. Chang, D.C. Huang, J.B. Lee, S.E. Molesa, A.R. Murphy, D.R. Redinger, S.K. Volkman, Proc. IEEE 93 (2008) 1330. [2] S. Holdcroft, Macromolecules 124 (1991) 4834. [3] A.C. Arias, S.E. Ready, R. Lujan, W.S. Wong, K.E. Paul, Appl. Phys. Lett. 85 (2004) 3304. [4] H. Sirringhaus, Science 290 (2000) 2123. [5] Y. Lan, C.H. Yang, V. Yang, Polym. Int. 59 (2010) 16. [6] R.J. Chesterfield, J.C. McKee, C.R. Newman, P.C. Ewbank, D.A. da Sila Filho, J.-L. Brédas, L.L. Miller, K.R. Mann, C.D. Frisbie, J. Phys. Chem. B 108 (2004) 19281. [7] A.M. Ruder, Ann. N.Y. Acad. Sci. 1076 (2006) 207. [8] E.M. Ward, P.A. Schulte, K. Straif, N.B. Hopf, J.C. Caldwell, T. Carreón, D.M. DeMarini, B.A. Fowler, B.D. Goldstein, K. Hemminki, C.J. Hines, K.H. Pursiainen, E. Kuempel, J. Lewtas, R.M. Lunn, E. Lynge, D.M. McElvenny, H. Muhle, T. Nakajima, L.W. Robertson, N. Rothman, A.M. Ruder, M.K. Schubauer-Berigan, J. Siemiatycki, D. Silverman, M.T. Smith, T. Sorahan, K. Steenland, R.G. Stevens, P. Vineis, S.H. Zahm, L. Zeise, V. Cogliano, Environ. J. Health Perspect. 118 (2010) 1355. [9] S.H. Liu, W.M. Wang, A.L. Briseno, S.C.B. Mannsfeld, Z. Bao, Adv. Mater. 21 (2009) 1217. [10] F. Brustolin, F. Goldoni, E.W. Meijer, N.A.J.M. Somerdijk, Macromolecules 35 (2002) 1054. [11] J.E. Millstone, D.F.J. Kavulak, C.H. Woo, T.W. Holcombe, E.J. Westling, A.L. Briseno, M.F. Toney, J.J. Fréchet, Langmuir 26 (2010) 13056. [12] H. Shimizu, M. Yamada, R. Wada, M. Okabe, Polym. J. 40 (2008) 33. [13] K.B. Burke, A.J. Stapleton, B. Vaughan, X.J. Zhou, A.L.D. Kilcoyne, W.J. Belcher, P.C. Dastoor, Nanotechnology 22 (2011) 265710. [14] A.J. Stapleton, B. Vaughan, B.F. Xue, E. Sesa, K. Burke, X.J. Zhou, G. Bryant, O. Werzer, A. Nelson, A.L.D. Kilcoyne, L. Thomsen, E. Wanless, W.J. Belcher, P.C. Dastoor, Sol. Energy Mater. Sol. Cells 102 (2012) 114. [15] K. Landfester, R. Montenegro, U. Scherf, R. Güntner, U. Asawapirom, S. Patil, D. Neher, T. Kietzke, Adv. Mater. 14 (2002) 651. [16] I.Y. Perevyazko, A. Vollrath, C. Pietsch, S. Schubert, G.M. Pavlov, U.S. Schubert, J. Polym. Sci. 50 (2012) 2906. [17] B.N. Pal, B.M. Dhar, K.C. See, H.E. Katz, Nat. Mater. 8 (2009) 898. [18] H.G.O. Sandberg, T.G. Bäcklund, R. Österbacka, S. Jussila, T. Mäkelä, H. Stubb, Synth. Met. 155 (2005) 662. [19] H.G.O. Sandberg, T.G. Bäcklund, R. Österbacka, H. Stubb, Adv. Mater. 16 (2004) 1112. [20] T.G. Bäcklund, H.G.O. Sanberg, R. Österbacka, H. Stubb, Appl. Phys. Lett. 85 (2004) 3887. [21] D. Elkington, D. Darwis, X.J. Zhou, W.J. Belcher, P.C. Dastoor, Org. Electron. 13 (2012) 153. [22] H. Yan, Z.H. Chen, Y. Zheng, C. Newman, J.R. Quinn, F. Dötz, M. Kastler, A. Facchetti, Nature 457 (2009) 679. [23] D. Darwis, D. Elkington, S. Ulum, A. Stapleton, G. Bryant, X.J. Zhou, W.J. Belcher, P.C. Dastoor, in: The 4th Nanoscience and Nanotechnology Symposium (nns2011): An International Symposium. AIP Conference Proceedings, vol. 1415, 2011, p. 124. [24] V. Shrotriya, J. Ouyang, R.J. Tseng, G. Li, Y. Yang, Chem. Phys. Lett. 411 (2005) 138. [25] H. Hintz, H.J. Egelhaaf, H. Peisert, T. Chassé, Polym. Degrad. Stab. 95 (2010) 818. [26] S. Palaniappan, C.A. Amarnath, Mater. Chem. Phys. 92 (2005) 82. [27] D.B.A. Rep, A.F. Morpurgo, W.G. Sloof, T.M. Klapwijk, J. Appl. Phys. 93 (2003) 2082. [28] N.J. Kaihovirta, D. Tobjörk, T. Mäkelä, R. Österbacka, Adv. Eng. Mater. 10 (2008) 640.
Contents lists available at SciVerse ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Novel low voltage and solution processable organic thin film transistors based on water dispersed polymer semiconductor nanoparticulates Darmawati Darwis a,b, Daniel Elkington a, Syahrul Ulum a,b, Glenn Bryant a, Warwick Belcher a, Paul Dastoor a, Xiaojing Zhou a,⇑ a b
Centre for Organic Electronic, University of Newcastle, Callaghan, NSW 2308, Australia Department of Physics, Faculty of Science, Tadulako University, Kampus Bumi Tadulako Tondo kota Palu 94118, Indonesia
a r t i c l e
i n f o
Article history: Received 4 February 2013 Accepted 29 March 2013 Available online 8 April 2013 Keywords: Polymer nanoparticulates OTFT P3HT Surfactant SDS
a b s t r a c t Two novel organic thin film transistor structures that combine a hygroscopic insulator with the use of water-dispersed polymer nanoparticles as the active layer are presented. In the first device structure, the semiconducting layer was fabricated from a nanoparticulate suspension of poly-(3-hexylthiophene) prepared through a mini-emulsion process using sodium dodecyl sulfate as the surfactant whereas a surfactant-free precipitation method has been used for the second device structure. In both cases, fully solution processable transistors have been fabricated in a top gate configuration with hygroscopic poly(4-vinylphenol) as the dielectric layer. Both device structures operate at low voltages (0 to 4 V) but exhibit contrasting output characteristics. A systematic study is presented on the effect of surfactant on the synthesis of semiconducting nanoparticles, the formation of thin nanoparticulate films and, consequently, on device performance. Ó 2013 Elsevier Inc. All rights reserved.
1. Introduction Electronic products based on polymer and other organic materials are emerging rapidly, and encompass items such as: organic light emitting diode (OLED) displays, integrated circuits, organic photovoltaic cells and transistors [1–6]. Organic transistors are key components in many of these organic electronic devices. The main driver for this rapidly increasing interest in organic thin film transistors (OTFTs) is their ease of fabrication. In particular, the deposition and patterning of all layers in an OTFT is possible at low temperature using a combination of solution-based techniques, which makes them ideally suited for the realization of low-cost, large-area electronics on flexible substrates. As such, the inexpensive and simple processing capabilities of organic electronics means that they have the potential to replace conventional devices in many applications [4]. A key advantage of OTFT technology is that conventional printing techniques can be implemented to provide rapid mass production of devices. However, a challenge for printing OTFTs is that the organic semiconducting solutions that are used generally involve chlorinated solvents which are highly toxic and volatile [7,8]. In addition, the fabrication of high performance OTFTs requires optimization of polymer crystallinity in the active layer, which in turn is highly dependent upon the choice of solvents, deposition
⇑ Corresponding author. E-mail address: [email protected] (X. Zhou). 0021-9797/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2013.03.052
techniques and post-deposition treatments such as annealing [9]. Thus, from a fabrication point of view, the development of alternative more environmentally compatible semiconducting solutions capable of producing crystalline semiconducting polymer layers would be highly attractive. Recently, methods have been reported to prepare semiconducting polymer films through pre-formed polymer nanoparticulate dispersions. These nanoparticles can be prepared either by a precipitation approach [10,11] or a surfactant-assisted mini-emulsion method [12–15]. Previous studies have shown that it is possible to form stable suspensions with a tunable nanoparticle size using both methods, thus offering a new pathway to fabricate OTFT active layers with controlled polymer domain size and thin film morphology [10–16]. Furthermore, the polymer chains can be highly crystalline within each particle [10,11]. Many applications for organic transistors also require low voltage operation driven by the need for devices with low power consumption [17]. One approach to low-voltage OTFT operation is to use hygroscopic materials as the dielectric layer [18–20]. Poly(4vinylphenol) (PVP) is one such hygroscopic polymer and the moisture adsorbed by it plays a central role in modulating drain current under low gate voltage values [21]. The current modulation mechanism in these devices involves chemical doping of the semiconducting channel stemming from ion diffusion at the dielectricsemiconductor interface [20]. These devices operate at a few volts and have maximum output currents (ION) comparable to, or even higher than, many conventional OFETs operating at many tens of volts [11].
66
D. Darwis et al. / Journal of Colloid and Interface Science 401 (2013) 65–69
Significantly, the use of a polymer top-gate architecture in these transistors allows for a fully solution-processable device, which can ultimately be printed on flexible substrates [22]. A challenging aspect of fabricating this architecture with nanoparticulate materials is that it requires a pinhole-free semiconductor layer. We have shown that using surfactant-containing nanoparticles (NPs) we can obtain high-quality active layers and subsequently fabricate lowvoltage OTFT devices with a PEDOT:PSS top gate electrode [23]. In this work, we compare NP OTFTs fabricated with and without surfactant to probe the mechanism of operation in these devices. The NP thin film morphologies, chemical and electronic properties and performance of OTFTs fabricated from these two types of nanoparticles are characterized. We show that the performance of the surfactant-containing OTFTs is dictated by counter-ions associated with SDS in the film. By removing the surfactant from the active layer, substantial improvements in the performance of these low-voltage all-solution processable OTFTs are obtained. This novel device structure offers the twin advantages of low-voltage operation combined with the elimination of harmful solvents from the device fabrication process.
20 nm
0 Fig. 1. AFM images of the Type I (surfactant-containing) P3HT nanoparticle (NP) thin films: (a) unannealed NP film (rms roughness of 4.2 nm) and (b) NP film after annealing for 5 min at 140 °C (rms roughness of 3.5 nm), the scale bars are 50 nm.
probe microscopy system in non-contact mode. X-ray photoelectron spectroscopy (XPS) analysis was performed using a multi-technique ultrahigh-vacuum imaging XPS system (Thermo VG Scientific ESCALab 250) equipped with a hemispherical analyzer (of 150 mm mean radius) and a monochromatic Al KR (1486.60 eV) X-ray source. An ultraviolet–visible (UV–Vis) absorption spectrophotometer (Varian Cary 6000i) was used to measure the absorption of the nanoparticle suspensions and the films.
2. Materials and methods 3. Results and discussion Semiconducting polymeric nanoparticles were prepared in aqueous media via a mini-emulsion process (Type I). P3HT (Mn = 22 K, PDI 2.58, purchased from Lumtec) was dissolved in chloroform at a concentration of 30 mg/mL and then mixed into an aqueous sodium dodecyl sulfate (SDS) solution (42 mg SDS in 2.8 mL MilliQ water). A macroemulsion was then formed by stirring the solution at 1200 rpm for 1 h and this was then sonicated for 2 min to form a mini-emulsion. The mini-emulsion was then gently stirred at 60 °C to evaporate the chloroform. After evaporation of the solvent, the nanoparticle suspension was dialysed to concentrate the samples and remove excess surfactant. Stable polymer dispersions with a polymer particle size of about 60 nm, as measured by a Zetasizer Nano-ZS (Malvern Instruments, UK), were obtained. Semiconducting polymeric nanoparticles were prepared without surfactant using the direct precipitation method (Type II). P3HT was dissolved in 1 mL chloroform (0.25 wt.%) and stirred for 10 min at room temperature. This solution was transferred into a syringe and rapidly dropped into 4 mL anhydrous ethanol while simultaneously stirring the mixture (200 rpm). After injection, stable nanoparticle suspensions were obtained. Both types of suspensions were subsequently used in the fabrication of OTFTs. OTFTs were fabricated on glass substrates with pre-patterned indium-tin-oxide (ITO) source and drain electrodes. The channel length and width was 20 lm and 3 mm respectively. A 100 nm thick P3HT nanoparticle layer was spin-coated (60 s at 2000 rpm) from the suspensions onto the substrate as measured by a KLA Tencor profilometer. A hot plate with a controlled temperature was used to anneal the formed films. For the dielectric layer, PVP (Aldrich) was dissolved in ethanol at a concentration of 80 mg/ mL and then spin-coated (60 s at 2000 rpm) to a thickness of approximately 500 nm. The P3HT/PVP two-layer structures were annealed at 85 °C in air to remove any remaining solvent. Lastly, PEDOT:PSS (Aldrich, pH = 1–2) was drop-cast on the top of gate dielectric layer and dried on a hot plate at 40 °C in air. Two Keithley 2400 source meters were used for the device characterization. All output characteristic measurements were conducted at a scan rate of 0.1 V/s. All fabricated devices were measured in air immediately after drying the PEDOT:PSS layer. The relative humidity (RH) of the laboratory was consistently recorded as 50 ± 10%. Atomic force microscopy (AFM) images of the films formed from these suspensions were measured using a Cypher scanning
Fig. 1a and b show AFM images of the Type I (surfactant-containing) nanoparticle (NP) thin film before and after annealing at 140 °C respectively. A particle size distribution from 30 nm to 100 nm is observed in Fig. 1a in good agreement with transmission electron microscopy (TEM) and dynamic light scattering (DLS) measurements (see Supporting information in Figs. S1 and S2). The AFM images also show that multiple particles stack together; resulting in a continuous film with a thickness of 60 nm as measured by profilometry. Annealing at 140 °C for 5 min results in the coalescing of particles shown in Fig. 1b and, based on the AFM roughness analysis, a slightly smoother film is obtained. The surface roughness (based on RMS) of the film decreases from 4.2 nm to 3.5 nm upon annealing. Fig. 2 compares the UV–Vis spectra of the surfactant-containing NPs in suspension with that of the thin film before and after annealing at 140 °C. Consistent with previous reports, the absorption spectra of P3HT are all red-shifted compared to the absorption spectrum of P3HT in CHCl3 solution [24]. In addition, the red-shifts for solid state NP films are larger than that observed for NPs in suspension and exhibit an increased background signal. The increased red-shift in the UV–Vis spectra for the NP films can be attributed to more crystalline NPs being preferably retained during the spincasting process, whilst the higher background signal is consistent with increased optical scattering due to the particulate nature of the film. Three pronounced absorption peaks are observed for both the P3HT NPs in suspension (510 nm, 535 nm and 585 nm) and for the NP thin films (525 nm, 550 nm and 600 nm). The three adsorption features are attributed to the p–p stacking of the P3HT backbone and the intermolecular p–p stacking interaction and are consistent with the absorption features for nanoparticle P3HT reported by Millstone [11]. The pronounced vibronic shoulders at 550 nm and 600 nm indicate highly ordered polymer chains in the NPs and, given that these shoulders increase in magnitude upon annealing, shows that annealing at 140 °C enhances polymer crystallinity within the NPs themselves. X-ray photoelectron spectroscopy (XPS) was used to investigate the relative SDS:P3HT concentration ratio at the thin film surface before the PVP dielectric layer is deposited. Fig. 3a and b shows the XPS S 2p and C 1s spectra respectively. In Fig. 3a, the peak doublet located at 164.0 eV and 164.5 eV corresponds to the S 2p signal for P3HT [25] while the other doublet at 170.0 eV and 171.2 eV
D. Darwis et al. / Journal of Colloid and Interface Science 401 (2013) 65–69
Fig. 2. UV–Vis spectra of the surfactant-containing nanoparticle in a suspension solution (dashed line), a thin film before (dotted line) and after 140 °C annealing (solid line).
Fig. 3. (a)S 2p, (b) C 1s and (c) Na 1s XPS spectra of the surfactant-containing nanoparticle thin films, unannealed at room temperature (RT, solid line) and annealed at 100 °C (dash dotted line), 140 °C (dashed line) and 160 °C (dotted line).
corresponds to S 2p signal associated with the SO24 group of the SDS molecule [26]. The relative intensity at 164.0 eV and 170.0 eV from the XPS spectra indicates that a significant amount of SDS remains on the film surface both at room temperature (RT) and after annealing at 100 °C. This observation is consistent with the observed higher binding energy (285.9 eV) for the C 1s peak at RT and 100 °C
67
shown in Fig. 3b. The presence of a charged SDS layer at the surface screens the emitted photoelectrons resulting in the observed chemical shift for the C 1s peak. However, upon annealing to 140 °C, the S 2p peak intensity for P3HT becomes more pronounced whilst that of SDS is diminished, with a corresponding shift of the C 1s signal to a lower binding energy of 285.1 eV. The increase of the P3HT:SDS S 2p XPS intensity ratio indicates that, at the elevated temperatures, the SDS molecules are mobilized and move away from the surface of the P3HT NP film. This observation is consistent with previous work on PFB:F8BT NP materials which also showed that SDS migrates away from the NPs upon annealing [14]. The diminish of Na 1s XPS intensity shown in Fig. 3c upon annealing is co-occurring to that for SDS S 2p, indicating Na+ as the counter ions are still associated the SDS. Transistors incorporating an unannealed surfactant-containing NP thin film and a 140 °C annealed nanoparticle thin film as the semiconducting layer were fabricated and their output characteristics are shown in Fig. 4. The turn-on voltages of the unannealed and annealed transistors are 0.4 V and 0.8 V, respectively, and both devices operate fully between 0 V and 2 V; consistent with the low voltage operation previously reported for P3HT/PVP noncross-linked hygroscopic transistors [18–20]. Large gate leakage currents (of the order of 10% of the drain current (ID)) are also observed for both transistors, indicating that ionic transport is contributing to the turn-on current [21]. As a consequence, a relatively low current modulation ratio of less than 50 is generally obtained. The observation of a high ionic gate leakage current is consistent with the presence of residual ionic species in the active layer, which diffuse to the relevant counter electrode under the influence of the applied voltage and thus add to the on current. XPS spectra in Fig. 3c of the NP films show the presence of high surface concentrations of Na+, which is the counter ion in the SDS surfactant. As such, the deposition of a PVP dielectric layer on top of the NP active layer is likely to result in the creation of a P3HT/PVP interface that is highly doped with Na+ ions. Given the highly mobile nature of this ionic species in organic semiconductors [27] these Na+ ions will diffuse into both the active and dielectric layers and subsequently contribute to both the turn-on drain current and the leakage gate current. In contrast to the Type I NPs, P3HT NP thin films prepared from a surfactant-free suspension in ethanol require multiple spin-cast coatings to form continuous films. Indeed, optimized P3HT NP thin films required up to eight spin-cast coatings with a 3 min 100 °C annealing step in between each spin-cast, as well as a 5 min 140 °C annealing step after the last coating. Fig. 5 shows the AFM images of surfactant-free P3HT NP thin films after different numbers of coating stages and reveals how the morphology of the thin film changes upon repeated coating. From Fig. 5a, we observe that a single spin-cast coating results in isolated P3HT NPs on the substrate surface. Without the surfactant, Type II NPs are highly hydrophobic and repel each other during the spin-casting process resulting in a sparsely coated layer. In addition, the particle sizes and size distribution of the NPs in the film are relatively large and are in the region of 50–300 nm compared to 60 nm for the surfactant-containing NP film. Fig. 5b–d show that subsequent coatings result primarily in a filling in of the vacancies in the layer with little evidence of multiple layer formation. Fig. 5d indicates that eight coatings produce a relatively continuous film, but surface roughness measurements reveal that the Type II P3HT NP film is 10 times rougher (38.5 nm) than the Type I NP film (3.5 nm). It should be noted that films made from Type II nanoparticles are inherently more likely to contain pinhole defects (due to the absence of surfactant in the fabrication process), which in turn can lead to non-working devices. Fig. 6 shows the UV–Vis spectra of the surfactant-free P3HT NP film following the multiple spin casts. In accordance with the observation of the Type I P3HT NP film, three absorption peaks at 525 nm, 550 nm and 600 nm have been observed, indicating a similar conjugated ordering for the polymer chains in the NPs synthe-
68
D. Darwis et al. / Journal of Colloid and Interface Science 401 (2013) 65–69
(a)
(b)
(c)
(d)
Fig. 4. Current–voltage characteristics of thin film transistors fabricated with surfactant-containing nanoparticle: (a) ID VD from an unannealed film at different gate voltages; (b) ID VG from an unannealed film at VD = 1 V; (c) ID VD from an annealed film at different voltages; and (d) ID VG from an annealed film at VD = 1 V.
Fig. 6. UV–Vis of the surfactant-free P3HT nanoparticle thin film after (a) one coating (dash dot), (b) three coatings (dash), (c) six coatings (dot) and (d) eight coatings (solid line). Fig. 5. AFM images of the surfactant-free P3HT nanoparticle thin film surface after (a) one coating, (b) two coatings, (c) four coatings and (d) eight coatings, the scale bars are1 lm.
sized without SDS. The absorption peak locations and their relative intensity also agree with the previous observations [11]. The systematic increase in UV–Vis intensity with increasing numbers of coatings is consistent with each subsequent deposition filling vacancies in the preceding layer. A comparison of the UV–Vis absorption intensity of the P3HT surfactant-free film after eight coatings with that of the single coating surfactant-containing NP film (Fig. 2) indicates that a similar amount of P3HT material is present in both films. Fig. 7a shows the output characteristic of an OTFT fabricated from a surfactant-free P3HT NP thin film. The transistor is turned on at a gate voltage of 2.8 V, which is higher than that for a
surfactant-containing NP thin film. However, a lower leakage current is observed for the Type II P3HT NP devices resulting in an improved device performance (relative to the Type I P3HT NP devices) with a maximum drain current of 35 lA and a current modulation ratio of 200. This performance is comparable with that reported for both spin-coated and reverse gravure printed OTFT devices fabricated from P3HT in chlorinated solvents [28]. As such, it would appear that the removal of surfactant and counter ions from the active layer in the Type II P3HT NP devices reduces the associated ion contribution to the leakage current and hence reduces the offcurrent. In addition, the fact that much lower on currents (less than 1 lA) are observed for surfactant-containing OTFTs relative to those for the surfactant-free devices (a few tens of lA) suggests that the surfactant acts as a charge trap in the channel; thus reducing charge mobility.
D. Darwis et al. / Journal of Colloid and Interface Science 401 (2013) 65–69
(a)
Fig. 7. (a) Output characteristics of ID
VD at different gate voltages and (b) ID
69
(b)
VG (at VD =
The data demonstrates that the P3HT NP based hygroscopic transistors presented here exhibit much higher drain currents and current modulation ratios but much lower operation voltages relative to the NP OFETs prepared using silicon oxide and OTS as the dielectric layer reported by Millstone et al. [11]. This behavior is consistent with dependence of the current modulation on the hygroscopic dielectric, which is known to facilitate low voltage operation [18–21]. We hypothesize that the absence of a clear drain current saturation characteristic in devices fabricated from the surfactant-free NPs is most likely due to the presence of a rough P3HT/PVP interface as indicated by the rough P3HT surface shown in Fig. 5d. Further work is underway to improve the morphology of the surfactant-free NP film and further enhance device performance. 4. Conclusions Surfactant-containing and surfactant-free P3HT NPs have been synthesized through a mini-emulsion process and a precipitation method respectively. The surfactant-containing NPs form a continuous film more readily than the surfactant-free particles. In addition, surfactant can be largely removed from the film surface by annealing the film to 140 °C. Low voltage solution processable OTFTs based on a hygroscopic dielectric can be fabricated using both types of suspensions and their IV characteristics resemble the P3HT bulk film transistor for the same device architecture and dielectric material. Transistors prepared from surfactant-free NPs offer better device performance with a higher drain current and current modulation ratio, due to the elimination of surfactant and counter ions from the active layer and the associated ion contribution to the leakage current. In addition, the much lower source and drain on currents (less than 1 lA) for surfactant-containing OTFTs than that for surfactant-free OTFTs (tens of lA) suggest that the SDS and counter ion species can act as charge traps in the channel. Acknowledgments This research was supported by a Directorate General of Higher Education Indonesia (DIKTI) scholarship and the Commonwealth of Australian Research Council through Discovery Project Funding Scheme. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2013.03.052.
2.0 V) for a thin film transistor fabricated with surfactant-free P3HT NP thin film.
References [1] V. Subarmanian, J.M.J. Frechet, P.C. Chang, D.C. Huang, J.B. Lee, S.E. Molesa, A.R. Murphy, D.R. Redinger, S.K. Volkman, Proc. IEEE 93 (2008) 1330. [2] S. Holdcroft, Macromolecules 124 (1991) 4834. [3] A.C. Arias, S.E. Ready, R. Lujan, W.S. Wong, K.E. Paul, Appl. Phys. Lett. 85 (2004) 3304. [4] H. Sirringhaus, Science 290 (2000) 2123. [5] Y. Lan, C.H. Yang, V. Yang, Polym. Int. 59 (2010) 16. [6] R.J. Chesterfield, J.C. McKee, C.R. Newman, P.C. Ewbank, D.A. da Sila Filho, J.-L. Brédas, L.L. Miller, K.R. Mann, C.D. Frisbie, J. Phys. Chem. B 108 (2004) 19281. [7] A.M. Ruder, Ann. N.Y. Acad. Sci. 1076 (2006) 207. [8] E.M. Ward, P.A. Schulte, K. Straif, N.B. Hopf, J.C. Caldwell, T. Carreón, D.M. DeMarini, B.A. Fowler, B.D. Goldstein, K. Hemminki, C.J. Hines, K.H. Pursiainen, E. Kuempel, J. Lewtas, R.M. Lunn, E. Lynge, D.M. McElvenny, H. Muhle, T. Nakajima, L.W. Robertson, N. Rothman, A.M. Ruder, M.K. Schubauer-Berigan, J. Siemiatycki, D. Silverman, M.T. Smith, T. Sorahan, K. Steenland, R.G. Stevens, P. Vineis, S.H. Zahm, L. Zeise, V. Cogliano, Environ. J. Health Perspect. 118 (2010) 1355. [9] S.H. Liu, W.M. Wang, A.L. Briseno, S.C.B. Mannsfeld, Z. Bao, Adv. Mater. 21 (2009) 1217. [10] F. Brustolin, F. Goldoni, E.W. Meijer, N.A.J.M. Somerdijk, Macromolecules 35 (2002) 1054. [11] J.E. Millstone, D.F.J. Kavulak, C.H. Woo, T.W. Holcombe, E.J. Westling, A.L. Briseno, M.F. Toney, J.J. Fréchet, Langmuir 26 (2010) 13056. [12] H. Shimizu, M. Yamada, R. Wada, M. Okabe, Polym. J. 40 (2008) 33. [13] K.B. Burke, A.J. Stapleton, B. Vaughan, X.J. Zhou, A.L.D. Kilcoyne, W.J. Belcher, P.C. Dastoor, Nanotechnology 22 (2011) 265710. [14] A.J. Stapleton, B. Vaughan, B.F. Xue, E. Sesa, K. Burke, X.J. Zhou, G. Bryant, O. Werzer, A. Nelson, A.L.D. Kilcoyne, L. Thomsen, E. Wanless, W.J. Belcher, P.C. Dastoor, Sol. Energy Mater. Sol. Cells 102 (2012) 114. [15] K. Landfester, R. Montenegro, U. Scherf, R. Güntner, U. Asawapirom, S. Patil, D. Neher, T. Kietzke, Adv. Mater. 14 (2002) 651. [16] I.Y. Perevyazko, A. Vollrath, C. Pietsch, S. Schubert, G.M. Pavlov, U.S. Schubert, J. Polym. Sci. 50 (2012) 2906. [17] B.N. Pal, B.M. Dhar, K.C. See, H.E. Katz, Nat. Mater. 8 (2009) 898. [18] H.G.O. Sandberg, T.G. Bäcklund, R. Österbacka, S. Jussila, T. Mäkelä, H. Stubb, Synth. Met. 155 (2005) 662. [19] H.G.O. Sandberg, T.G. Bäcklund, R. Österbacka, H. Stubb, Adv. Mater. 16 (2004) 1112. [20] T.G. Bäcklund, H.G.O. Sanberg, R. Österbacka, H. Stubb, Appl. Phys. Lett. 85 (2004) 3887. [21] D. Elkington, D. Darwis, X.J. Zhou, W.J. Belcher, P.C. Dastoor, Org. Electron. 13 (2012) 153. [22] H. Yan, Z.H. Chen, Y. Zheng, C. Newman, J.R. Quinn, F. Dötz, M. Kastler, A. Facchetti, Nature 457 (2009) 679. [23] D. Darwis, D. Elkington, S. Ulum, A. Stapleton, G. Bryant, X.J. Zhou, W.J. Belcher, P.C. Dastoor, in: The 4th Nanoscience and Nanotechnology Symposium (nns2011): An International Symposium. AIP Conference Proceedings, vol. 1415, 2011, p. 124. [24] V. Shrotriya, J. Ouyang, R.J. Tseng, G. Li, Y. Yang, Chem. Phys. Lett. 411 (2005) 138. [25] H. Hintz, H.J. Egelhaaf, H. Peisert, T. Chassé, Polym. Degrad. Stab. 95 (2010) 818. [26] S. Palaniappan, C.A. Amarnath, Mater. Chem. Phys. 92 (2005) 82. [27] D.B.A. Rep, A.F. Morpurgo, W.G. Sloof, T.M. Klapwijk, J. Appl. Phys. 93 (2003) 2082. [28] N.J. Kaihovirta, D. Tobjörk, T. Mäkelä, R. Österbacka, Adv. Eng. Mater. 10 (2008) 640.