- Email: [email protected]
Process optimization for inkjet printing of triisopropylsilylethynyl pentacene with single-solvent solutions
Thin Solid Films 578 (2015) 11–19
Contents lists available at ScienceDirect
Thin Solid Films journal homepage: www.elsevier.com/locate/tsf
Process optimization for inkjet printing of triisopropylsilylethynyl pentacene with single-solvent solutions Xianghua Wang a,⁎, Miao Yuan a,b, Xianfeng Xiong a, Mengjie Chen a, Mengzhi Qin a,b, Longzhen Qiu a, Hongbo Lu a, Guobing Zhang a, Guoqiang Lv a, Anthony H.W. Choi c a Key Lab of Special Display Technology, Ministry of Education, National Engineering Lab of Special Display Technology, National Key Lab of Advanced Display Technology, Academy of Opto-Electronic Technology, Hefei University of Technology, Hefei 230009, China b School of Electronic Science & Applied Physics, Hefei University of Technology, Hefei 230009, China c Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, China
a r t i c l e
i n f o
Article history: Received 28 August 2014 Received in revised form 29 January 2015 Accepted 2 February 2015 Available online 9 February 2015 Keywords: Inkjet printing Coffee-ring effect Surface energy Organic thin-film transistor Dispersion force Self-assembly
a b s t r a c t Inkjet printing of 6,13-bis(triisopropylsilylethynyl) pentacene (TIPS-PEN), a small molecule organic semiconductor, is performed on two types of substrates. Hydrophilic SiO2 substrates prepared by a combination of surface treatments lead to either a smaller size or a coffee-ring profile of the single-drop film. A hydrophobic surface with dominant dispersive component of surface energy such as that of a spin-coated poly(4-vinylphenol) film favors profile formation with uniform thickness of the printed semiconductor owing to the strong dispersion force between the semiconductor molecules and the hydrophobic surface of the substrate. With a hydrophobic dielectric as the substrate and via a properly selected solvent, high quality TIPS-PEN films were printed at a very low substrate temperature of 35 °C. Saturated field-effect mobility measured with top-contact thin-film transistor structure shows a narrow distribution and a maximum of 0.78 cm2V−1 s−1, which confirmed the film growth on the hydrophobic substrate with increased crystal coverage and continuity under the optimized process condition. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Solution processing of soluble conjugated molecules at low cost is a key technology to fabricate organic thin-film transistors (OTFTs) for “printed electronics”. Molecular design and synthesis of organic semiconductors (OSCs) are indispensable to achieve high field-effect mobility, and numerous soluble OSCs such as side-chain-substituted acenes [1] and triethylsilylethynyl anthradithiophene [2] have been reported. The optimization of process conditions [3] or ink compositions [4] has been presented as an alternative approach toward high performance OTFTs. As the intrinsic field-effect mobility of conjugated molecules, especially that of small molecules, approaches 1 cm2V−1 s−1 and higher values, their performances are becoming comparable or even superior to those of hydrogenated amorphous silicon (a-Si:H). Therefore, much attention is now devoted to material processing in solution forms. In the report by Kim et al., high quality one-dimensional (1D) singlecrystalline 6,13-bis(triisopropylsilylethynyl) pentacene (TIPS-PEN) microribbon field-effect transistor was fabricated by the solventexchange method and exhibited mobilities as high as 1.42 cm2V−1 s−1 [5]. A similar microribbon structure was also prepared by single-drop ⁎ Corresponding author. E-mail address: [email protected] (X. Wang).
http://dx.doi.org/10.1016/j.tsf.2015.02.004 0040-6090/© 2015 Elsevier B.V. All rights reserved.
casting on an inclined substrate [6]. High mobility bottom-contact OTFTs based on drop-cast TIPS-PEN via high boiling point solvent have also been reported [7]. More recently, devices with even higher fieldeffect mobilities of 4.6 cm2V−1 s−1 were obtained by the solution shearing method at optimized shear speed [8], owing to the increased lattice strain. Therefore, a more advanced solution shearing method was proposed to achieve better control of the solution-printed film [9]. Another systematic work on process development was conducted using dip-coating method [10]. Besides the most studied TIPS-PEN, research on solution processing of other high mobility smallmolecule semiconductors such as 2,7-dialkyl [1] benzothieno[3,2-b] [1] benzothiophenes has been reported. On the other hand, effects of the gate dielectric on the semiconductor processing and OTFT performance have been identified as another factor of paramount importance [11]. The introduction of ultrathin dielectric layer, such as the selfassembled monolayer molecular gate dielectric [12], would enable ultralow-power organic complementary circuits [13], however, nanometer-scale interfacial heterogeneity on the dielectric layer [14] would lead to less continuous film growth of the overlying semiconductor. Inkjet-printed single-drop films typically possess a radial pattern of crystalline orientation [15], which would cause dramatic deviceto-device differences in electrical conductivity between the source and drain electrodes of OTFTs. A promising approach for solution-
12
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
processing of small molecule OSCs is to blend them with a polymer [16] and to print broad-area films instead of single-drop films for better uniformity in film morphology as well as improved repeatability as required for mass production. In this case, vertical phase separation [17] is essential to achieve high mobility charge transport [18] as well as improved reproducibility [19] or device-to-device uniformity [20]. Nevertheless, the highly viscoelastic polymers would sometimes make it difficult for stable jetting of the ink. Another issue is the problematic film thickness uniformity arising from the coffee-ring effect. Usage of binary solvent was reported to eliminate the coffeering stain in a single-drop deposition [15], however, the effectiveness seems insignificant when printing large-area films with overlapping drop assignments. Summing up all previous findings, manipulation of the substrate's surface property in conjunction with proper solvent selection seems to be a more promising solution for inkjet printing of uniform films. In this work, inkjet printing of TIPS-PEN is performed on substrates with fine-tuned surface energies that have been obtained either by surface treatment, including molecule self-assembling and UV-ozone cleaning, or by applying a polymeric insulator that is insoluble in the ink solution, with the target of fabricating high quality large-area semiconductor films with uniform thicknesses and morphologies. The quality of the crystalline film and their interface with the insulator is studied through OTFTs in a bottom-gate/top-contact architecture. Instead of printing a blending of the OSC and a polymer, a singlesolvent solution of TIPS-PEN is printed with the intension of figuring out the primary interactions between the solution and the substrate, together with their effects on film processing via the solution. In addition to the surface properties of the dielectric substrate, dependence of the electrical properties of the OTFTs on other factors including solvent selection and process parameters for inkjet printing of large-area films, such as drop spacing and line spacing, are discussed for overlapping-drop-assignment printing. 2. Experimental details The inkjet printing experiments are performed with a Dimatix DMP3000 printer. TIPS-PEN (TCI Co. Ltd., Tokyo, Japan) is dissolved in orthodichlorobenzene (o-DCB, for Sections 3.1–3.4) or 1,2,3,4tetrahydronaphthalene (tetralin, in Section 3.5) and injected into a cartridge through a 0.45 μm filter. The inkjet printing is operated with a 10 pL printhead (DMC-11610) at 1 kHz jetting frequency for a tradeoff between high efficiency and stable jetting. The printer operates in a temperature-controlled ambient of 25 °C with humidity maintained at 35%. The electrical properties of the semiconductor films are evaluated with a bottom-gate/top-contact thin-film transistor (TFT) structure as shown in the schematic diagram of Fig. 1(a). An optimized 3segment waveform as shown in Fig. 1(b) is used to fire droplets stably at a velocity of 2.5 m/s. The surface energies of the dielectric substrates are modulated by surface treatment with either self-assembled monolayer (SAM) of organosilane, UV-ozone cleaning (PSD PRO-UV10T), or spin-coating of an insulating polymer layer (MIDAS System SPIN1200D). The surface energy is estimated according to the contact angles measured at room temperature with an OCA15 video-based automatic contact angle measuring instrument (Data Physics). Deionized water and methylene iodide are used as the test liquids. More details were described in a previous report [21]. Film morphology of TIPS-PEN is characterized with optical microscope (Leica DM2500M equipped with polarizer and analyzer) and atomic force microscope (AFM, Bruker Multimode). The cross section profile is obtained with an Ambios XP-200 profilometer. Device characteristics were measured at room temperature in ambient air using a Keithley 4200 semiconductor parameter analyzer. Field-effect mobility is extracted from the transfer curves collected in the saturation region with a Vds of −60 V. For surface modulation over a wide range, the hydrophilic SiO2/Si wafer is firstly treated with a SAM layer of 1H,1H,2H,2H-
Fig. 1. (a) Schematic diagram of a bottom-gate/top-contact OTFT with an inkjet-printed TIPS-PEN as the channel semiconductor and (b) the 3-segment jetting waveform developed for stable jetting.
perfluorodecyltrichlorosilane (FDTS) to obtain a hydrophobic surface with very small surface energy, both the dispersive and the polar components. The surface treatment is performed in FDTS vapor for 2 h within a vacuumed container. 2-Phenylethyltrichlorosilane (PETS) is used as the organosilane molecule for SAM treatment. Prior to the surface treatment, the wafer has been cleaned in a piranha solution (70 vol.% H2SO4 + 30 vol.% H2O2) for 40 min at 90 °C and rinsed with copious amounts of deionized water. The sample is then soaked for 20 min in a 10 mM PETS solution in anhydrous toluene. In the final step, the substrate is washed with toluene and ethanol in sequence and blown dry with nitrogen. The surface of the sample is hydrophobic with a total surface energy of 44.79 mNm−1 (Dispersion component: 39.05 mNm−1, Polar component: 5.74 mNm−1). Another hydrophobic substrate is prepared by spin coating a second dielectric layer of poly-4-vinylphenol (PVP) on top of the SiO2 substrate. Poly(melamine-coformaldehyde) methylated as a cross-linker was mixed with PVP at a weight ratio of 4:6 and dissolved in propylene glycol 1-monomethyl ether 2-acetate (C6H12O3, PGMEA). The total concentration of the solution is 10%. The solution is spin-coated at a revolution rate of 4000 rpm, forming a cross-linked PVP of 130 nm thickness after 90 min of vacuum bake at 180 °C. The surface energy of the PVP(dispersion component: 38.81 mNm−1, polar component: 8.63 mNm− 1) is very close to that of the PETS-SAM-treated SiO2 substrate owing to the phenyl groups on the surfaces of both substrates. The overall capacitance of the PVP/SiO2 dielectric stack is 8.5 nFcm−2 as measured at 1 MHz.
3. Results 3.1. Inkjet printing of isolating dots The diameter and size uniformity depends heavily on the physical interaction between the substrate and the droplet of the ink solution. The ink solution and the pure solvent used in this experiment had a surface energy lower than that of the untreated silicon dioxide substrate and the UV-ozone-treated samples. Therefore, complete wetting on these substrates is observed with the contact angle measurement
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
system using a 1 μL testing droplet. When printing with a 10 pL droplet to produce a single-drop film of micron-scale diameter, the drying process is much faster due to the enhanced specific surface area. Evaporation-induced flow within the sessile drop is therefore strengthened. Contact line pinning during the drying process of an ink drop can be described with a generally accepted model [22]. As a critical value of solute concentration is required for nucleation, the intensified evaporation-induced flow will speed up solute diffusion to the edge and result in earlier pinning (i.e., larger size of the solidified film). This is verified through the observation of increasing diameter of the single-drop film when the substrate is heated up. It is further observed that the critical value of concentration is negatively related with the surface energy of the dielectric substrate. Solvent-cleaning of the SiO2/Si wafer with acetone would reduce uniformity in surface property at the micron scale [21]. Single-drop films printed on such a heterogeneous surface are smaller in diameter, which was ascribed to distortion of contact line pinning due to the micron scale perturbation on the substrate. The aforementioned model is established on interactions between the solvent and the substrate. For inkjet printing of picoliter volumes, the interactions between the solute molecules and the substrate may play a vital role. This effect was investigated with surface-modified SiO2/Si wafers, whereby the surface energy was modulated either with an UV-ozone clean or with a SAM layer followed by UV-ozone cleaning. Surface silanization with a SAM layer of FDTS effectively reduces the surface energy of SiO2. While the FDTS-SAM treatment reduces the
13
dispersion component as well as the polar component of the surface energy, surface cleaning such as UV-ozone cleaning was found to enhance more effectively the polar contribution of surface energy. The combination of a FDTS-SAM treatment and UV-ozone cleaning was proven to be effective in tuning the diameter and profile of the printed single-drop films. For comparison, two sets of surface treatments are performed and their effects on surface energy are measured. One is solely treated with UV-ozone cleaning and the other is pre-treated with a FDTS-SAM treatment. Fig. 2(a) and (b) respectively shows their surface energy in a 2-dimensional coordinate space (polar component and dispersive component). Dependence of surface energies, including their two components, on the treatment time of UV-ozone cleaning is plotted in Fig. 2(c) and (d) respectively for the two sets of surface treatments. As shown in Fig. 2(d), a prior FDTS-SAM treatment significantly improves the sensitivity of surface energy on the treatment time of UV-ozone cleaning. Therefore, the combination of FDTS-SAM treatment and UV-ozone cleaning provide more effective modulation of the surface energy and the film morphology printed thereon. For substrates treated with UV-ozone cleaning alone, the dispersive component and the polar component are of comparable magnitude. For the FDTS-treated sample, the two components deviate from each other after 6 min of UV-ozone cleaning. The strongest contrast is noticed after 8 min of UV-ozone cleaning, when the polarity is maximized to 0.71 for the FDTS-treated sample. Nevertheless, the total surface energies, of both sets of substrates, saturate at almost identical level of 75 mNm−1. Based on these experimental findings, a comparative study on inkjet printing is performed on two SiO2/Si samples, one of
Fig. 2. Surface energy of SiO2 plotted in coordinates of polar component versus dispersive component (a, b), and their dependence on the treatment time of UV-ozone cleaning (c, d). Two sets of surface treatments are performed on SiO2 for comparison: (a, c) solely treated with UV-ozone cleaning, (b, d) pre-treated with a FDTS-SAM treatment before UV-ozone cleaning.
14
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
which is solely surface-treated with 8 min of UV-ozone cleaning (S1) while the other is grown with a FDTS-SAM followed by the same treatment of 8-min UV-ozone cleaning (S2). Among the isolating dot arrays printed with a 2 wt.% TIPS-PEN solution in o-DCB at a substrate temperature of 45 °C, the best size uniformity was achieved with substrate S2, to which a combined surface treatment of 1 h FDTS-SAM followed by an 8-min UV-ozone clean had been applied. Fig. 3(a–c) and (e–g) shows the optical microphotographs of single-drop films printed on S1 and S2 respectively. As shown in Fig. 3(e), besides the improved size uniformity, the singledrop films were composed of 1–3 spherulitic domains and were obviously smaller in size compared to those on S1. The superior uniformity in film size and geometry on S2 indicated a uniform surface property after the combined surface treatment. Difference in singledrop film diameter of S1 with that of S2 cannot be explained with the previous model that ascribes contact-line pinning to solventevaporation. As the same ink solution is employed for the printing, the large film size on S1 could be explained by the higher dispersion contribution of surface energy, which not only contributed to liquid wetting but also to nucleation at a lower solute concentration. The mechanism will be discussed in the discussion section. The single layer film thickness as shown in Fig. 3(d) is higher than 100 nm for the single-drop films printed on S2, while the film thickness is merely ~15 nm in the central region of the film printed on S1. The coffee-ring effect is still apparent with the isolated dot films, especially those of larger size. The coffee-ring profile seems to be strengthened when the contact-line was pinned at larger diameters, which can result from higher surface energy or elevated temperature of the substrate. Fig. 3(g) shows the cross-polarized microphotograph of a dual layer dot film, which has been produced by 2 identical printing cycles for the generation of thicker films. However, the thickness has
been increased at the cost of aggravated coffee staining. A quantitative description of the coffee-ring profile in terms of thickness homogeneity can be defined as the ratio between the minimum and the maximum thickness. The profile ratio is apparently increased from 15% as on the UV-ozone cleaned substrate S1 to 68% as on S2, while the value for the dual-layer film on S2 is reduced to 34%. 3.2. Inkjet printing of isolating lines and large-area films Single-line films were printed with a drop spacing of 20 μm. Fig. 4(a) shows the microphotographs of the single-line film printed on S1 (left) and S2 (right) respectively. A wider line film printed with 2 runs along the y-direction is also printed on S2. The inset diagram shows the printing parameters, drop spacing (DS) and line spacing (LS), defined for the pattern. The upper diagram of Fig. 4(b) shows the crosssection profile of the single-line films (intersected by a plane normal to the y-direction), showing a concave profile in the middle due to coffee-ring effect. The lower diagram shows the cross-section profile across a wider line printed with two runs along the y-direction shifted by a line spacing (LS) of 160 μm in the x-direction, which is obviously unsymmetrical, being thicker on one side than the other. The unsymmetrical profile is also discernable on its microphotographs as colored contours are clearly seen as shown in Fig. 4(c,d). Correspondingly, the optical microphotograph of a 4-line broad area film in Fig. 4(e,f) shows contrasting morphology change from the left in the form of crystal flakes to the right side of the film showing a spherulite or amorphous morphology. The coffee-ring effect of the single-line film is still evident in the cross-section profile along the x-direction short axis and the sequentially printed coalesced lines induce a right-to-left solute diffusion. Interestingly, the profile ratios of the single-line films printed on S1 and S2 are comparable as shown in Fig. 4(b). Different from the
Fig. 3. Optical (a) unpolarized and (b, c, e–f) cross-polarized images of single-drop films. Films in image (a–c) are printed on a SiO2/Si substrate (S1) treated with 8 min of UV-ozone cleaning, and (e–g) taken from another SiO2/Si substrate (S2) treated with FDTS-SAM followed by 8 min of UV-ozone cleaning. (g) A representative microphotograph of a 2-layer isolating dot printed with 2 shots on the same site at a time interval of 10 s. (d) Profiles of the dot films printed on S1 (d, upper) and S2 (d, middle and lower).
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
15
Fig. 4. Microphotographs of single-line, overlapping 2-line and multiple-line films of TIPS-pentacene printed on SiO2. (a) Image taken with the fiducial camera of the printer (left: singleline film on S1; right: single-line and 2-line films on S2.) The inset diagram shows the printing parameters, DS and LS, as defined for the pattern. (b) Profiles of the single-line films and the 2-line film as shown in (a). Cross-polarized (c, e) and unpolarized (d, f) microphotographs of the films printed on the SiO2 that is surface modified with a FDTS-SAM treatment followed by 8 min (c, d) and 10 min (e, f) of UV-ozone cleaning respectively. Films in (c, d) and (e, f) are printed with 2 and 4 overlapping lines respectively.
printing of isolated dots, the fluorinated silica substrate with additional UV-ozone cleaning does not reduce the coffee-ring effect when printing a single line, while the printing of large-area films with multiple-line overlapping-drop assignment leads to a decreasing thickness from the left starting line to the right ending line. Empirically, the poor thickness uniformity can be explained by the enhanced coffee-ring effect as the radius of curvature at the contact-line is remarkably reduced as the droplet is spread over a large area and deformed from the spherical shape by gravity.
3.3. Inkjet-printed OTFTs and effects of printing parameters An n-type silicon wafer with 300 nm thermal SiO2 (capacitance = 10.8 nFcm−2) is used as the substrate for the fabrication of OTFTs. The substrate is fluorinated with FDTS followed by 8 min of UV-ozone cleaning. A single nozzle is used for the printing of isolating single-line films or large-area films with overlapping-drop assignment. The print head performs a forward single-trip run along the y-direction and producing a single line with a specified DS that is smaller than the single-drop diameter. At the ending of the forward trip, the print head stops jetting and returns to its starting position on standby or move a step distance along the x-direction according to the specified LS. Films of broader widths were printed by repeating the round trip shifted by LS smaller than the single-line width. With the substrate temperature set at 45 °C, printing of the semiconductor is conducted at various drop spacings (20, 30, 40 μm) and line spacings (60, 80, 100, 150, 180 μm) to determine the optimal spacing parameters for the best device performance. In Fig. 5(a), the dependence of field-effect mobility is plotted against average semiconductor film thickness. Below 200 nm, the field-effect mobility increases with film thickness, saturating for thicker semiconductor films. In addition to film thickness, the effect of the drop-assignment pattern on device performance was evident. Fig. 5(b) plots the on/off ratio of the OTFTs on spacing ratio (LS/DS).
Fig. 5. Saturated field-effect mobility of inkjet-printed TIPS-PEN via an o-DCB solution on SiO2 substrates treated with FDTS-SAM and 10 min of UV-ozone cleaning. (a) Dependence of mobility along x- and y-direction on film thickness; (b) dependence of on/off ratio for field-effect charge transport along x- and y-direction on the ratio of LS/DS.
16
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
The maximum on/off ratio of the OTFT is achieved at spacing ratios of 5 and 3.75 respectively for x- and y-direction of current conduction. 3.4. Effect of substrate's surface property As mentioned in Section 3.1, the size and cross-section profile of single-drop films on SiO2 substrate are determined by the surface energy of the substrate, and it seems to be the dispersion component that determines the diameter. It should be noted that wetting between the substrate and the ink solution is a prerequisite for regular inkjet printing, and therefore, UV-ozone cleaning is an indispensable step to achieve sufficient wettability between the FDTS-SAM-treated sample and the ink solution. However, UV-ozone cleaning invariably produces hydrophilic surface properties. It would be interesting if the surface energy can be hydrophobic with a dispersion component high enough to wet the organic ink solution. To test this feasibility, a SAM layer of PETS or a spin-coated PVP is prepared on SiO2 for hydrophobic surface treatment. A TIPS-PEN ink (2 wt.% in o-DCB) was printed onto the hydrophobic substrate as single-drop and single-line films. The film thickness uniformity is remarkably improved compared to those printed on the hydrophilic substrates. Fig. 6 shows the cross-section profiles of the single-drop films printed on the hydrophobic substrates showing a flat-plateau shape instead of a coffee-ring stain. Improved thickness uniformity is maintained even at elevated substrate temperatures. The average thickness of the single-drop films of around 80 μm in diameter is 30 nm and that of the single-line films is 50 nm. Nevertheless, these films are insufficient in thickness for high mobility charge transport according to the results shown in Fig. 5. Taking into account the flexible nature of the PVP coating as well as its uniform hydrophobic surface energy, the PVP layer is selected as the substrate for the fabrication of OTFT arrays. Inkjet-printed multi-line films with larger thicknesses are employed as the active layer for higher device performance. With PVP as a second dielectric layer, the field-effect mobility of the inkjet-printed multi-line films can reach 0.1 cm2V−1 s−1, but the average mobility is not as high due to the lower performance of samples in the first few columns especially from column No. 2 to around No. 7 of the 30 columns of printed OTFTs. The transfer characteristics and optical micrograph of the best and the worst devices from the same line of the OTFT array are presented in Fig. 7. Compared to normal devices showing the typical self-aligned crystals strips, the semiconductor layer of an inferior device is composed of disordered polycrystalline, and this type of morphology was more frequently observed when printing with
Fig. 6. Cross-section profile of the single-drop film printed via o-DCB on hydrophobic surfaces: (a) PVP-coated SiO2 and (b) PETS-SAM-treated SiO2 substrate.
Fig. 7. Transfer curves under cycle-swept Vg and the cross-polarized microscopy of the best and the worst device within the same line of the printed OTFT arrays on PVP/SiO2 substrate.
an ink that had been aged for a few hours in the cartridge. Solvent compatibility of the black plastic material that constitutes the printhead of the cartridge was subsequently tested; obvious dissolution of the plastic is observed in o-DCB. Intuitively, plastic dissolved in the ink solution constitutes an impurity that would degrade the semiconductor film morphology and its electrical properties.
3.5. Further optimization via solvent selection The issue with the o-DCB solvent suggests that dissolution of the cartridge material by the ink solution should be avoided. Milder solvent such as toluene, anisole, xylene and tetralin are potential candidates. Of these, tetralin is more promising for inkjet printing via tiny droplets because of its high boiling point that favors self-organization of the semiconductor molecules as crystalline films. The multi-line film of TIPS-PEN printed via tetralin on a PVP/SiO2 substrate is polycrystalline with domains in the shape of wide strips (20–100 μm in width) when printed at a substrate temperature of 35 °C. The crystal growth direction runs parallel with the y-direction of the printed pattern, therefore, top-contact/bottom-gate OTFT architecture is fabricated with the channel perpendicular to the y-direction of the printed film. The typical channel width and length are 250 μm and 84 μm respectively for these devices. Devices printed at a lower substrate temperature of 35 °C are generally superior to those printed at a higher temperature of 45 °C. The average mobility of devices printed at 35 °C is enhanced by one order of magnitude to a value of 0.36 cm2V− 1 s− 1 compared with those printed with the o-DCB ink, while the maximum mobility reaches 0.78 cm2V−1 s−1, which is about a 7-fold increase. The threshold voltage is − 21 V as estimated from the transfer curve in Fig. 8. Mobility distribution of these OTFTs is shown in Fig. 9. OTFTs printed at 35 °C Films outperforms those printed a 45 °C owing to the larger film thickness. As shown in Fig. 10, the increasing diameter of the isolating dot film printed via tetralin indicates a decreasing film thickness with increasing substrate temperature. Among the 36 devices printed at 35 °C, 35 devices exhibit typical OTFT characteristics with ~95% of the devices falls within the range of 0.1–0.8 cm2V−1 s−1. Compared to devices printed with an o-DCB ink, the tremendous improvement in mobility as well as the more uniform device performance is at least partially ascribed to the tetralin solvent that possesses a higher surface energy than o-DCB, providing a wider process window for the tuning of single-drop film diameter and film thickness as seen in Fig. 10.
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
17
Fig. 10. Diameters of single-drop films printed via the single-solvent solution of o-DCB or tetralin on PVP-coated substrate with increasing substrate temperature. Fig. 8. Transfer curves under cycle-swept Vg of the highest performance device with the accumulation curve shown in open squares and depletion curve in open circles, inset: the cross-polarized microphotograph of the OTFT.
Film coverage of TIPS-PEN films inkjet-printed for top-contact OTFTs is characterized with AFMs as shown in Fig. 11. Almost 100% material coverage within the domains was observed for samples printed with the 2 wt.% tetralin solution on the PVP/SiO 2 substrate. When a
Fig. 9. Mobility distribution of 54 TIPS-PEN OTFTs fabricated on PVP/SiO2 substrate with the semiconductor printed via a 2 wt.% solution in tetralin. Inkjet printing was conducted with the substrate temperature maintained at 45 °C (18 devices) or 35 °C (36 devices).
2 wt.% o-DCB solution is employed, the crystal ribbons become narrower in width and are not fully grown but the crystal growth direction is well aligned. In contrast, crystal domains are not as well aligned over large areas on an untreated hydrophilic SiO2 substrate. 4. Discussion Film formation starts from the edge of the liquid when the critical concentration required for nucleation at the contact line is reached during the drying process [22]. On a wettable substrate, single-drop film diameter is found to be determined by the dispersion contribution of the surface energy. Herein, a simple model is proposed that qualitatively explains the shape of the deposits and the dependence on process parameters such as substrate temperature. Depending on the magnitude of the dispersion component and the polarity of the surface energy, there are generally two types of drying process after a droplet impact. In the first case, the solid surface is hydrophilic with a relatively high polarity, solvent-substrate interaction dominates and nucleation is relatively difficult until a relatively high concentration is reached locally as the solvent evaporates and the outward flow thereof causes solute accumulation toward the contact line. The coffee-stain profile is typically observed especially at higher substrate temperatures as the evaporation rate is enhanced and the outward flow intensified. In the other case, the solid surface is hydrophobic with a dominative dispersion component of surface energy, interaction between solute molecules and the substrate becomes predominant and pinning happens at a lower solute concentration. As a result, a larger film is deposited with more uniform film thickness, and the cross section profile is less sensitive to the evaporation induced flow at elevated substrate temperatures. The crystallographic morphology of the multiple-line film is studied with out-of-plane XRD showing the typically observed (001) diffraction peaks [23]. The side groups of the TIPS-PEN molecules are attracted by the surface of the substrate presumably via a dispersive force that induced contact-line pinning and self-assembling of the molecules in an “edge-on” orientation. The side group of TIPS-PEN is hydrophobic and is disposed to interact with the substrate through the dispersion force. For high performance devices especially those printed via the tetralin solution, thermal annealing at a higher temperature or vacuum annealing at room temperature does not improve on-current or the saturated field-effect mobility, but a few hours of vacuum annealing would cause a shift in threshold voltage toward zero volts and a reduced hysteresis in the transfer curve. Independence of high mobility to annealing indicates that the system was in a state of thermodynamical equilibrium after drying on the substrate when the semiconductor molecules self-assembled in the “edge-on” orientation. Vacuum
18
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
Fig. 11. AFM images (50 × 50 μm2) of large-area TIPS-PEN films printed with overlapping multiple lines on 3 substrates (Left: via o-DCB on untreated hydrophilic SiO2, middle: via o-DCB on PVP/SiO2, right: via tetralin on PVP/SiO2).
annealing is beneficial to improve interface quality as most of the residue solvent is removed. In this work, OTFTs of highest performance are printed at a low temperature of 35 °C on PVP/SiO2 using the TIPS-PEN/tetralin solution. The diameter of the single-drop film on PVP still increases with increasing substrate temperature, a trend similar to that observed from the hydrophilic substrates. It should be noted that the single-drop film diameter is heavily dependent on solvent selection. At a substrate temperature of 35 °C, the diameter decreases from around 88 μm to 50 μm (Fig. 10) when using a tetralin as the solvent instead of o-DCB, while the thickness uniformity is maintained. It is supposed that the critical value of concentration required for contact-line pinning on PVP is independent on the initial concentration of the droplet, but critical to the surface energy of the solvent and temperature of the substrate. The critical concentration for nucleation from a tetralin solution is higher than that from an o-DCB medium because the inward surface tension of tetralin is higher. Therefore, delayed contact-line pinning and higher film thickness are achieved when printing with a tetralin solution. Furthermore, the critical concentration is tunable across a wider range by printing at different substrate temperatures, whereby a suitable timing for contact-line pinning becomes possible. Incomplete film coverage within the channel of an OTFT or within a crystal domain as well as insufficient film thickness or uniformity is obviously a primary obstacle to the printing of high-performance devices. Specifically, thickness uniformity is limited when printing on a hydrophilic substrate, while insufficient film thickness or incomplete material coverage is inevitable when printing on a hydrophobic substrate where premature contact-line pinning takes place. With the highest mobility up to 0.8 cm2V−1 s−1 achieved at 35 °C on PVP/SiO2 substrate, further optimization on drop spacing and the stroke length would improve the average mobility of a large-scale device array up to 0.5 cm2V−1 s−1. At a substrate temperature of 35 °C, the mobility averaged on a line was lower for the first line (0.3 cm2V−1 s−1 on average) but the average value can reach 0.5 cm2V− 1 s− 1 for the following lines printed without process idling, which is over 60% of the maximum mobility (0.8 cm2V−1 s−1). The percentage is used as a measurement of performance uniformity, which is comparable to vacuum deposited organic semiconductors and obviously higher than the previously reported values for solution prepared devices including inkjet printing [24] and spin coating [25]. TIPS-PEN OTFTs of similar field-effect mobility can be prepared by inkjet-printing over a spin-coated poly(α-methylstyrene) (PαMS) on SiO2 [26]. The locally dissolved PαMS-coatsing by the ink solution functions as a self-aligned bank, which contributes to a uniform film thickness of the inner plateau and also higher precision and pattern fidelity of the printed film. The improved switch-on characteristics compared to the reference device (without PαMS) are ascribed to the high quality interface with the substrate. Compared to the current
work, the difference in drying behavior of the blend ink formed on PαMS/SiO2 is corroborated by the negative dependence of deposit diameter on substrate temperature. Although the highest mobility is comparable, several differences are noteworthy. Firstly, devices of the highest performance are printed at a higher substrate temperature of 70 °C on PαMS-coated substrates, especially on the central circular region with a thickness as low as ~ 20 nm. Owing to the hydrophobic surface with a sufficiently high dispersion component of surface energy, the PVP layer contributes to improved thickness uniformity across the overall printed areas as well as a high film coverage when TIPS-PEN is printed via tetralin at a low substrate temperature of 35 °C. Secondly, the PVP-coating is intact and therefore a smooth interface with the semiconductor persists, which can reasonably lead to enhanced fieldeffect mobility, although the inferior steepness of the subthreshold slope still indicate a poor interface quality. A potential solution via interfacial nucleation on hydrophobic surfaces has been proposed recently [27]. In addition to surface modification, PVP can be easily spin-coated on almost any substrates providing uniform surface energies over large areas and can be directly employed as the gate dielectric in OTFTs. Therefore, the application of PVP-based process is promising for largescale fabrication of integrated flexible electronics.
5. Conclusions In summary, process optimization are performed for inkjet printing of TIPS-PEN via surface engineering of the substrate, solvent selection as well as optimization of the printing parameters such as drop spacing, line spacing and substrate temperature. The coffee-ring profile is totally eliminated when printing on a hydrophobic surface such as PETStreated SiO2 substrate or on a PVP-coated sample due to the strong dispersion force between the phenyl-groups on the solid surface and the side groups of the semiconductor molecules that contribute to contact-line pinning at lower solute concentrations. An optimized process condition is formulated for picoliter inkjet printing of TIPSPEN that uses a PVP-coating as the substrate and tetralin as the solvent. Dramatic improvement in saturated field-effect mobility is achieved when printing at a low substrate temperature of 35 °C where optimal film morphology with complete film coverage is achieved.
Acknowledgments This work was supported by the National Natural Science Foundation of China (51203039, 61107014, 21204017), National Basic Research Program of China (Grant No. 2012CB723406), Program for New Century Excellent Talents in University (Grant No. NCET-12-0839).
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
References [1] M.M. Payne, S.R. Parkin, J.E. Anthony, C.C. Kuo, T.N. Jackson, Organic field-effect transistors from solution-deposited functionalized acenes with mobilities as high as 1 cm2/V s, J. Am. Chem. Soc. 127 (2005) 4986. [2] K.C. Dickey, J.E. Anthony, Y.-L. Loo, Improving organic thin-film transistor performance through solvent-vapor annealing of solution-processable triethylsilylethynyl anthradithiophene, Adv. Mater. 18 (2006) (1721- +). [3] H. Minemawari, T. Yamada, H. Matsui, J.y. Tsutsumi, S. Haas, R. Chiba, R. Kumai, T. Hasegawa, Inkjet printing of single-crystal films, Nature 475 (2011) 364. [4] S.Y. Cho, J.M. Ko, J. Lim, J.Y. Lee, C. Lee, Inkjet-printed organic thin film transistors based on TIPS pentacene with insulating polymers, J. Mater. Chem. C 1 (2013) 914. [5] D.H. Kim, D.Y. Lee, H.S. Lee, W.H. Lee, Y.H. Kim, J.I. Han, K. Cho, High‐mobility organic transistors based on single‐crystalline microribbons of triisopropylsilylethynyl pentacene via solution‐phase self‐assembly, Adv. Mater. 19 (2007) 678. [6] W.H. Lee, D.H. Kim, Y. Jang, J.H. Cho, M. Hwang, Y.D. Park, Y.H. Kim, J.I. Han, K. Cho, Solution-processable pentacene microcrystal arrays for high performance organic field-effect transistors, Appl. Phys. Lett. 90 (2007) (132106-132106-132103). [7] S.K. Park, T.N. Jackson, J.E. Anthony, D.A. Mourey, High mobility solution processed 6,13-bis(triisopropyl-silylethynyl) pentacene organic thin film transistors, Appl. Phys. Lett. 91 (2007). [8] G. Giri, E. Verploegen, S.C. Mannsfeld, S. Atahan-Evrenk, S.Y. Lee, H.A. Becerril, A. Aspuru-Guzik, M.F. Toney, Z. Bao, Tuning charge transport in solution-sheared organic semiconductors using lattice strain, Nature 480 (2011) 504. [9] Y. Diao, B.C. Tee, G. Giri, J. Xu, H.A. Becerril, R.M. Stoltenberg, T.H. Lee, G. Xue, S.C. Mannsfeld, Z. Bao, Solution coating of large-area organic semiconductor thin films with aligned single-crystalline domains, Nat. Mater. 12 (2013) 665. [10] R.Z. Rogowski, A. Dzwilewski, M. Kemerink, A.A. Darhuber, Solution processing of semiconducting organic molecules for tailored charge transport properties, J. Phys. Chem. C 115 (2011) 11758. [11] X. Sun, C.-a. Di, Y. Liu, Engineering of the dielectric-semiconductor interface in organic field-effect transistors, J. Mater. Chem. 20 (2010) 2599. [12] M. Halik, H. Klauk, U. Zschieschang, G. Schmid, C. Dehm, M. Schutz, S. Maisch, F. Effenberger, M. Brunnbauer, F. Stellacci, Low-voltage organic transistors with an amorphous molecular gate dielectric, Nature 431 (2004) 963. [13] H. Klauk, U. Zschieschang, J. Pflaum, M. Halik, Ultralow-power organic complementary circuits, Nature 445 (2007) 745. [14] X. Sun, Y. Liu, C.-a. Di, Y. Wen, Y. Guo, L. Zhang, Y. Zhao, G. Yu, Interfacial heterogeneity of surface energy in organic field-effect transistors, Adv. Mater. 23 (2011) 1009. [15] J.A. Lim, W.H. Lee, H.S. Lee, J.H. Lee, Y.D. Park, K. Cho, Self-organization of ink-jetprinted triisopropylsilylethynyl pentacene via evaporation-induced flows in a drying droplet, Adv. Funct. Mater. 18 (2008) 229.
19
[16] C. Liu, Y. Li, T. Minari, K. Takimiya, K. Tsukagoshi, Forming semiconductor/dielectric double layers by one-step spin-coating for enhancing the performance of organic field-effect transistors, Org. Electron. 13 (2012) 1146. [17] J. Lee, J.Y. Jung, D.H. Kim, J.-Y. Kim, B.-L. Lee, J.-I. Park, J.W. Chung, J.S. Park, B. Koo, Y.W. Jin, S. Lee, Enhanced electrical stability of organic thin-film transistors with polymer semiconductor-insulator blended active layers, Appl. Phys. Lett. 100 (2012). [18] X. Li, W.T.T. Smaal, C. Kjellander, B. van der Putten, K. Gualandris, E.C.P. Smits, J. Anthony, D.J. Broer, P.W.M. Blom, J. Genoe, G. Gelinck, Charge transport in highperformance ink-jet printed single-droplet organic transistors based on a silylethynyl substituted pentacene/insulating polymer blend, Org. Electron. 12 (2011) 1319. [19] M.-B. Madec, P.J. Smith, A. Malandraki, N. Wang, J.G. Korvink, S.G. Yeates, Enhanced reproducibility of inkjet printed organic thin film transistors based on solution processable polymer-small molecule blends, J. Mater. Chem. 20 (2010) 9155. [20] Y.-H. Kim, J.E. Anthony, S.K. Park, Polymer blended small molecule organic field effect transistors with improved device-to-device uniformity and operational stability, Org. Electron. 13 (2012) 1152. [21] X.H. Wang, X.F. Xiong, L.Z. Qiu, G.Q. Lv, Morphology of inkjet printed 6,13 bis(triisopropylsilylethynyl) pentacene on surface-treated silica, J. Vac. Sci. Technol. B 30 (2012) 021206. [22] J.A. Lim, W.H. Lee, D. Kwak, K. Cho, Evaporation-induced self-organization of inkjet-printed organic semiconductors on surface-modified dielectrics for highperformance organic transistors, Langmuir 25 (2009) 5404. [23] J. Chen, C.K. Tee, M. Shtein, D.C. Martin, J. Anthony, Controlled solution deposition and systematic study of charge-transport anisotropy in single crystal and singlecrystal textured TIPS pentacene thin films, Org. Electron. 10 (2009) 696. [24] Y.H. Kim, B. Yoo, J.E. Anthony, S.K. Park, Controlled deposition of a high‐performance small‐molecule organic single‐crystal transistor array by direct ink‐jet printing, Adv. Mater. 24 (2012) 497. [25] Y. Yuan, G. Giri, A.L. Ayzner, A.P. Zoombelt, S.C.B. Mannsfeld, J. Chen, D. Nordlund, M.F. Toney, J. Huang, Z. Bao, Ultra-high mobility transparent organic thin film transistors grown by an off-centre spin-coating method, Nat. Commun. 5 (2014) 3005. [26] B.K.C. Kjellander, W.T.T. Smaal, J.E. Anthony, G.H. Gelinck, Inkjet printing of TIPSPEN on soluble polymer insulating films: a route to high-performance thin-film transistors, Adv. Mater. 22 (2010) 4612. [27] X.H. Wang, M. Yuan, S.C. Lv, M.Z. Qin, M.J. Chen, L.Z. Qiu, G.B. Zhang, H.B. Lu, Interfacial nucleation behavior of inkjet-printed 6,13 bis(tri-isopropylsilylethynyl) pentacene on dielectric surfaces, J. Appl. Phys. 117 (2015) 024902.
Contents lists available at ScienceDirect
Thin Solid Films journal homepage: www.elsevier.com/locate/tsf
Process optimization for inkjet printing of triisopropylsilylethynyl pentacene with single-solvent solutions Xianghua Wang a,⁎, Miao Yuan a,b, Xianfeng Xiong a, Mengjie Chen a, Mengzhi Qin a,b, Longzhen Qiu a, Hongbo Lu a, Guobing Zhang a, Guoqiang Lv a, Anthony H.W. Choi c a Key Lab of Special Display Technology, Ministry of Education, National Engineering Lab of Special Display Technology, National Key Lab of Advanced Display Technology, Academy of Opto-Electronic Technology, Hefei University of Technology, Hefei 230009, China b School of Electronic Science & Applied Physics, Hefei University of Technology, Hefei 230009, China c Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, China
a r t i c l e
i n f o
Article history: Received 28 August 2014 Received in revised form 29 January 2015 Accepted 2 February 2015 Available online 9 February 2015 Keywords: Inkjet printing Coffee-ring effect Surface energy Organic thin-film transistor Dispersion force Self-assembly
a b s t r a c t Inkjet printing of 6,13-bis(triisopropylsilylethynyl) pentacene (TIPS-PEN), a small molecule organic semiconductor, is performed on two types of substrates. Hydrophilic SiO2 substrates prepared by a combination of surface treatments lead to either a smaller size or a coffee-ring profile of the single-drop film. A hydrophobic surface with dominant dispersive component of surface energy such as that of a spin-coated poly(4-vinylphenol) film favors profile formation with uniform thickness of the printed semiconductor owing to the strong dispersion force between the semiconductor molecules and the hydrophobic surface of the substrate. With a hydrophobic dielectric as the substrate and via a properly selected solvent, high quality TIPS-PEN films were printed at a very low substrate temperature of 35 °C. Saturated field-effect mobility measured with top-contact thin-film transistor structure shows a narrow distribution and a maximum of 0.78 cm2V−1 s−1, which confirmed the film growth on the hydrophobic substrate with increased crystal coverage and continuity under the optimized process condition. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Solution processing of soluble conjugated molecules at low cost is a key technology to fabricate organic thin-film transistors (OTFTs) for “printed electronics”. Molecular design and synthesis of organic semiconductors (OSCs) are indispensable to achieve high field-effect mobility, and numerous soluble OSCs such as side-chain-substituted acenes [1] and triethylsilylethynyl anthradithiophene [2] have been reported. The optimization of process conditions [3] or ink compositions [4] has been presented as an alternative approach toward high performance OTFTs. As the intrinsic field-effect mobility of conjugated molecules, especially that of small molecules, approaches 1 cm2V−1 s−1 and higher values, their performances are becoming comparable or even superior to those of hydrogenated amorphous silicon (a-Si:H). Therefore, much attention is now devoted to material processing in solution forms. In the report by Kim et al., high quality one-dimensional (1D) singlecrystalline 6,13-bis(triisopropylsilylethynyl) pentacene (TIPS-PEN) microribbon field-effect transistor was fabricated by the solventexchange method and exhibited mobilities as high as 1.42 cm2V−1 s−1 [5]. A similar microribbon structure was also prepared by single-drop ⁎ Corresponding author. E-mail address: [email protected] (X. Wang).
http://dx.doi.org/10.1016/j.tsf.2015.02.004 0040-6090/© 2015 Elsevier B.V. All rights reserved.
casting on an inclined substrate [6]. High mobility bottom-contact OTFTs based on drop-cast TIPS-PEN via high boiling point solvent have also been reported [7]. More recently, devices with even higher fieldeffect mobilities of 4.6 cm2V−1 s−1 were obtained by the solution shearing method at optimized shear speed [8], owing to the increased lattice strain. Therefore, a more advanced solution shearing method was proposed to achieve better control of the solution-printed film [9]. Another systematic work on process development was conducted using dip-coating method [10]. Besides the most studied TIPS-PEN, research on solution processing of other high mobility smallmolecule semiconductors such as 2,7-dialkyl [1] benzothieno[3,2-b] [1] benzothiophenes has been reported. On the other hand, effects of the gate dielectric on the semiconductor processing and OTFT performance have been identified as another factor of paramount importance [11]. The introduction of ultrathin dielectric layer, such as the selfassembled monolayer molecular gate dielectric [12], would enable ultralow-power organic complementary circuits [13], however, nanometer-scale interfacial heterogeneity on the dielectric layer [14] would lead to less continuous film growth of the overlying semiconductor. Inkjet-printed single-drop films typically possess a radial pattern of crystalline orientation [15], which would cause dramatic deviceto-device differences in electrical conductivity between the source and drain electrodes of OTFTs. A promising approach for solution-
12
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
processing of small molecule OSCs is to blend them with a polymer [16] and to print broad-area films instead of single-drop films for better uniformity in film morphology as well as improved repeatability as required for mass production. In this case, vertical phase separation [17] is essential to achieve high mobility charge transport [18] as well as improved reproducibility [19] or device-to-device uniformity [20]. Nevertheless, the highly viscoelastic polymers would sometimes make it difficult for stable jetting of the ink. Another issue is the problematic film thickness uniformity arising from the coffee-ring effect. Usage of binary solvent was reported to eliminate the coffeering stain in a single-drop deposition [15], however, the effectiveness seems insignificant when printing large-area films with overlapping drop assignments. Summing up all previous findings, manipulation of the substrate's surface property in conjunction with proper solvent selection seems to be a more promising solution for inkjet printing of uniform films. In this work, inkjet printing of TIPS-PEN is performed on substrates with fine-tuned surface energies that have been obtained either by surface treatment, including molecule self-assembling and UV-ozone cleaning, or by applying a polymeric insulator that is insoluble in the ink solution, with the target of fabricating high quality large-area semiconductor films with uniform thicknesses and morphologies. The quality of the crystalline film and their interface with the insulator is studied through OTFTs in a bottom-gate/top-contact architecture. Instead of printing a blending of the OSC and a polymer, a singlesolvent solution of TIPS-PEN is printed with the intension of figuring out the primary interactions between the solution and the substrate, together with their effects on film processing via the solution. In addition to the surface properties of the dielectric substrate, dependence of the electrical properties of the OTFTs on other factors including solvent selection and process parameters for inkjet printing of large-area films, such as drop spacing and line spacing, are discussed for overlapping-drop-assignment printing. 2. Experimental details The inkjet printing experiments are performed with a Dimatix DMP3000 printer. TIPS-PEN (TCI Co. Ltd., Tokyo, Japan) is dissolved in orthodichlorobenzene (o-DCB, for Sections 3.1–3.4) or 1,2,3,4tetrahydronaphthalene (tetralin, in Section 3.5) and injected into a cartridge through a 0.45 μm filter. The inkjet printing is operated with a 10 pL printhead (DMC-11610) at 1 kHz jetting frequency for a tradeoff between high efficiency and stable jetting. The printer operates in a temperature-controlled ambient of 25 °C with humidity maintained at 35%. The electrical properties of the semiconductor films are evaluated with a bottom-gate/top-contact thin-film transistor (TFT) structure as shown in the schematic diagram of Fig. 1(a). An optimized 3segment waveform as shown in Fig. 1(b) is used to fire droplets stably at a velocity of 2.5 m/s. The surface energies of the dielectric substrates are modulated by surface treatment with either self-assembled monolayer (SAM) of organosilane, UV-ozone cleaning (PSD PRO-UV10T), or spin-coating of an insulating polymer layer (MIDAS System SPIN1200D). The surface energy is estimated according to the contact angles measured at room temperature with an OCA15 video-based automatic contact angle measuring instrument (Data Physics). Deionized water and methylene iodide are used as the test liquids. More details were described in a previous report [21]. Film morphology of TIPS-PEN is characterized with optical microscope (Leica DM2500M equipped with polarizer and analyzer) and atomic force microscope (AFM, Bruker Multimode). The cross section profile is obtained with an Ambios XP-200 profilometer. Device characteristics were measured at room temperature in ambient air using a Keithley 4200 semiconductor parameter analyzer. Field-effect mobility is extracted from the transfer curves collected in the saturation region with a Vds of −60 V. For surface modulation over a wide range, the hydrophilic SiO2/Si wafer is firstly treated with a SAM layer of 1H,1H,2H,2H-
Fig. 1. (a) Schematic diagram of a bottom-gate/top-contact OTFT with an inkjet-printed TIPS-PEN as the channel semiconductor and (b) the 3-segment jetting waveform developed for stable jetting.
perfluorodecyltrichlorosilane (FDTS) to obtain a hydrophobic surface with very small surface energy, both the dispersive and the polar components. The surface treatment is performed in FDTS vapor for 2 h within a vacuumed container. 2-Phenylethyltrichlorosilane (PETS) is used as the organosilane molecule for SAM treatment. Prior to the surface treatment, the wafer has been cleaned in a piranha solution (70 vol.% H2SO4 + 30 vol.% H2O2) for 40 min at 90 °C and rinsed with copious amounts of deionized water. The sample is then soaked for 20 min in a 10 mM PETS solution in anhydrous toluene. In the final step, the substrate is washed with toluene and ethanol in sequence and blown dry with nitrogen. The surface of the sample is hydrophobic with a total surface energy of 44.79 mNm−1 (Dispersion component: 39.05 mNm−1, Polar component: 5.74 mNm−1). Another hydrophobic substrate is prepared by spin coating a second dielectric layer of poly-4-vinylphenol (PVP) on top of the SiO2 substrate. Poly(melamine-coformaldehyde) methylated as a cross-linker was mixed with PVP at a weight ratio of 4:6 and dissolved in propylene glycol 1-monomethyl ether 2-acetate (C6H12O3, PGMEA). The total concentration of the solution is 10%. The solution is spin-coated at a revolution rate of 4000 rpm, forming a cross-linked PVP of 130 nm thickness after 90 min of vacuum bake at 180 °C. The surface energy of the PVP(dispersion component: 38.81 mNm−1, polar component: 8.63 mNm− 1) is very close to that of the PETS-SAM-treated SiO2 substrate owing to the phenyl groups on the surfaces of both substrates. The overall capacitance of the PVP/SiO2 dielectric stack is 8.5 nFcm−2 as measured at 1 MHz.
3. Results 3.1. Inkjet printing of isolating dots The diameter and size uniformity depends heavily on the physical interaction between the substrate and the droplet of the ink solution. The ink solution and the pure solvent used in this experiment had a surface energy lower than that of the untreated silicon dioxide substrate and the UV-ozone-treated samples. Therefore, complete wetting on these substrates is observed with the contact angle measurement
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
system using a 1 μL testing droplet. When printing with a 10 pL droplet to produce a single-drop film of micron-scale diameter, the drying process is much faster due to the enhanced specific surface area. Evaporation-induced flow within the sessile drop is therefore strengthened. Contact line pinning during the drying process of an ink drop can be described with a generally accepted model [22]. As a critical value of solute concentration is required for nucleation, the intensified evaporation-induced flow will speed up solute diffusion to the edge and result in earlier pinning (i.e., larger size of the solidified film). This is verified through the observation of increasing diameter of the single-drop film when the substrate is heated up. It is further observed that the critical value of concentration is negatively related with the surface energy of the dielectric substrate. Solvent-cleaning of the SiO2/Si wafer with acetone would reduce uniformity in surface property at the micron scale [21]. Single-drop films printed on such a heterogeneous surface are smaller in diameter, which was ascribed to distortion of contact line pinning due to the micron scale perturbation on the substrate. The aforementioned model is established on interactions between the solvent and the substrate. For inkjet printing of picoliter volumes, the interactions between the solute molecules and the substrate may play a vital role. This effect was investigated with surface-modified SiO2/Si wafers, whereby the surface energy was modulated either with an UV-ozone clean or with a SAM layer followed by UV-ozone cleaning. Surface silanization with a SAM layer of FDTS effectively reduces the surface energy of SiO2. While the FDTS-SAM treatment reduces the
13
dispersion component as well as the polar component of the surface energy, surface cleaning such as UV-ozone cleaning was found to enhance more effectively the polar contribution of surface energy. The combination of a FDTS-SAM treatment and UV-ozone cleaning was proven to be effective in tuning the diameter and profile of the printed single-drop films. For comparison, two sets of surface treatments are performed and their effects on surface energy are measured. One is solely treated with UV-ozone cleaning and the other is pre-treated with a FDTS-SAM treatment. Fig. 2(a) and (b) respectively shows their surface energy in a 2-dimensional coordinate space (polar component and dispersive component). Dependence of surface energies, including their two components, on the treatment time of UV-ozone cleaning is plotted in Fig. 2(c) and (d) respectively for the two sets of surface treatments. As shown in Fig. 2(d), a prior FDTS-SAM treatment significantly improves the sensitivity of surface energy on the treatment time of UV-ozone cleaning. Therefore, the combination of FDTS-SAM treatment and UV-ozone cleaning provide more effective modulation of the surface energy and the film morphology printed thereon. For substrates treated with UV-ozone cleaning alone, the dispersive component and the polar component are of comparable magnitude. For the FDTS-treated sample, the two components deviate from each other after 6 min of UV-ozone cleaning. The strongest contrast is noticed after 8 min of UV-ozone cleaning, when the polarity is maximized to 0.71 for the FDTS-treated sample. Nevertheless, the total surface energies, of both sets of substrates, saturate at almost identical level of 75 mNm−1. Based on these experimental findings, a comparative study on inkjet printing is performed on two SiO2/Si samples, one of
Fig. 2. Surface energy of SiO2 plotted in coordinates of polar component versus dispersive component (a, b), and their dependence on the treatment time of UV-ozone cleaning (c, d). Two sets of surface treatments are performed on SiO2 for comparison: (a, c) solely treated with UV-ozone cleaning, (b, d) pre-treated with a FDTS-SAM treatment before UV-ozone cleaning.
14
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
which is solely surface-treated with 8 min of UV-ozone cleaning (S1) while the other is grown with a FDTS-SAM followed by the same treatment of 8-min UV-ozone cleaning (S2). Among the isolating dot arrays printed with a 2 wt.% TIPS-PEN solution in o-DCB at a substrate temperature of 45 °C, the best size uniformity was achieved with substrate S2, to which a combined surface treatment of 1 h FDTS-SAM followed by an 8-min UV-ozone clean had been applied. Fig. 3(a–c) and (e–g) shows the optical microphotographs of single-drop films printed on S1 and S2 respectively. As shown in Fig. 3(e), besides the improved size uniformity, the singledrop films were composed of 1–3 spherulitic domains and were obviously smaller in size compared to those on S1. The superior uniformity in film size and geometry on S2 indicated a uniform surface property after the combined surface treatment. Difference in singledrop film diameter of S1 with that of S2 cannot be explained with the previous model that ascribes contact-line pinning to solventevaporation. As the same ink solution is employed for the printing, the large film size on S1 could be explained by the higher dispersion contribution of surface energy, which not only contributed to liquid wetting but also to nucleation at a lower solute concentration. The mechanism will be discussed in the discussion section. The single layer film thickness as shown in Fig. 3(d) is higher than 100 nm for the single-drop films printed on S2, while the film thickness is merely ~15 nm in the central region of the film printed on S1. The coffee-ring effect is still apparent with the isolated dot films, especially those of larger size. The coffee-ring profile seems to be strengthened when the contact-line was pinned at larger diameters, which can result from higher surface energy or elevated temperature of the substrate. Fig. 3(g) shows the cross-polarized microphotograph of a dual layer dot film, which has been produced by 2 identical printing cycles for the generation of thicker films. However, the thickness has
been increased at the cost of aggravated coffee staining. A quantitative description of the coffee-ring profile in terms of thickness homogeneity can be defined as the ratio between the minimum and the maximum thickness. The profile ratio is apparently increased from 15% as on the UV-ozone cleaned substrate S1 to 68% as on S2, while the value for the dual-layer film on S2 is reduced to 34%. 3.2. Inkjet printing of isolating lines and large-area films Single-line films were printed with a drop spacing of 20 μm. Fig. 4(a) shows the microphotographs of the single-line film printed on S1 (left) and S2 (right) respectively. A wider line film printed with 2 runs along the y-direction is also printed on S2. The inset diagram shows the printing parameters, drop spacing (DS) and line spacing (LS), defined for the pattern. The upper diagram of Fig. 4(b) shows the crosssection profile of the single-line films (intersected by a plane normal to the y-direction), showing a concave profile in the middle due to coffee-ring effect. The lower diagram shows the cross-section profile across a wider line printed with two runs along the y-direction shifted by a line spacing (LS) of 160 μm in the x-direction, which is obviously unsymmetrical, being thicker on one side than the other. The unsymmetrical profile is also discernable on its microphotographs as colored contours are clearly seen as shown in Fig. 4(c,d). Correspondingly, the optical microphotograph of a 4-line broad area film in Fig. 4(e,f) shows contrasting morphology change from the left in the form of crystal flakes to the right side of the film showing a spherulite or amorphous morphology. The coffee-ring effect of the single-line film is still evident in the cross-section profile along the x-direction short axis and the sequentially printed coalesced lines induce a right-to-left solute diffusion. Interestingly, the profile ratios of the single-line films printed on S1 and S2 are comparable as shown in Fig. 4(b). Different from the
Fig. 3. Optical (a) unpolarized and (b, c, e–f) cross-polarized images of single-drop films. Films in image (a–c) are printed on a SiO2/Si substrate (S1) treated with 8 min of UV-ozone cleaning, and (e–g) taken from another SiO2/Si substrate (S2) treated with FDTS-SAM followed by 8 min of UV-ozone cleaning. (g) A representative microphotograph of a 2-layer isolating dot printed with 2 shots on the same site at a time interval of 10 s. (d) Profiles of the dot films printed on S1 (d, upper) and S2 (d, middle and lower).
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
15
Fig. 4. Microphotographs of single-line, overlapping 2-line and multiple-line films of TIPS-pentacene printed on SiO2. (a) Image taken with the fiducial camera of the printer (left: singleline film on S1; right: single-line and 2-line films on S2.) The inset diagram shows the printing parameters, DS and LS, as defined for the pattern. (b) Profiles of the single-line films and the 2-line film as shown in (a). Cross-polarized (c, e) and unpolarized (d, f) microphotographs of the films printed on the SiO2 that is surface modified with a FDTS-SAM treatment followed by 8 min (c, d) and 10 min (e, f) of UV-ozone cleaning respectively. Films in (c, d) and (e, f) are printed with 2 and 4 overlapping lines respectively.
printing of isolated dots, the fluorinated silica substrate with additional UV-ozone cleaning does not reduce the coffee-ring effect when printing a single line, while the printing of large-area films with multiple-line overlapping-drop assignment leads to a decreasing thickness from the left starting line to the right ending line. Empirically, the poor thickness uniformity can be explained by the enhanced coffee-ring effect as the radius of curvature at the contact-line is remarkably reduced as the droplet is spread over a large area and deformed from the spherical shape by gravity.
3.3. Inkjet-printed OTFTs and effects of printing parameters An n-type silicon wafer with 300 nm thermal SiO2 (capacitance = 10.8 nFcm−2) is used as the substrate for the fabrication of OTFTs. The substrate is fluorinated with FDTS followed by 8 min of UV-ozone cleaning. A single nozzle is used for the printing of isolating single-line films or large-area films with overlapping-drop assignment. The print head performs a forward single-trip run along the y-direction and producing a single line with a specified DS that is smaller than the single-drop diameter. At the ending of the forward trip, the print head stops jetting and returns to its starting position on standby or move a step distance along the x-direction according to the specified LS. Films of broader widths were printed by repeating the round trip shifted by LS smaller than the single-line width. With the substrate temperature set at 45 °C, printing of the semiconductor is conducted at various drop spacings (20, 30, 40 μm) and line spacings (60, 80, 100, 150, 180 μm) to determine the optimal spacing parameters for the best device performance. In Fig. 5(a), the dependence of field-effect mobility is plotted against average semiconductor film thickness. Below 200 nm, the field-effect mobility increases with film thickness, saturating for thicker semiconductor films. In addition to film thickness, the effect of the drop-assignment pattern on device performance was evident. Fig. 5(b) plots the on/off ratio of the OTFTs on spacing ratio (LS/DS).
Fig. 5. Saturated field-effect mobility of inkjet-printed TIPS-PEN via an o-DCB solution on SiO2 substrates treated with FDTS-SAM and 10 min of UV-ozone cleaning. (a) Dependence of mobility along x- and y-direction on film thickness; (b) dependence of on/off ratio for field-effect charge transport along x- and y-direction on the ratio of LS/DS.
16
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
The maximum on/off ratio of the OTFT is achieved at spacing ratios of 5 and 3.75 respectively for x- and y-direction of current conduction. 3.4. Effect of substrate's surface property As mentioned in Section 3.1, the size and cross-section profile of single-drop films on SiO2 substrate are determined by the surface energy of the substrate, and it seems to be the dispersion component that determines the diameter. It should be noted that wetting between the substrate and the ink solution is a prerequisite for regular inkjet printing, and therefore, UV-ozone cleaning is an indispensable step to achieve sufficient wettability between the FDTS-SAM-treated sample and the ink solution. However, UV-ozone cleaning invariably produces hydrophilic surface properties. It would be interesting if the surface energy can be hydrophobic with a dispersion component high enough to wet the organic ink solution. To test this feasibility, a SAM layer of PETS or a spin-coated PVP is prepared on SiO2 for hydrophobic surface treatment. A TIPS-PEN ink (2 wt.% in o-DCB) was printed onto the hydrophobic substrate as single-drop and single-line films. The film thickness uniformity is remarkably improved compared to those printed on the hydrophilic substrates. Fig. 6 shows the cross-section profiles of the single-drop films printed on the hydrophobic substrates showing a flat-plateau shape instead of a coffee-ring stain. Improved thickness uniformity is maintained even at elevated substrate temperatures. The average thickness of the single-drop films of around 80 μm in diameter is 30 nm and that of the single-line films is 50 nm. Nevertheless, these films are insufficient in thickness for high mobility charge transport according to the results shown in Fig. 5. Taking into account the flexible nature of the PVP coating as well as its uniform hydrophobic surface energy, the PVP layer is selected as the substrate for the fabrication of OTFT arrays. Inkjet-printed multi-line films with larger thicknesses are employed as the active layer for higher device performance. With PVP as a second dielectric layer, the field-effect mobility of the inkjet-printed multi-line films can reach 0.1 cm2V−1 s−1, but the average mobility is not as high due to the lower performance of samples in the first few columns especially from column No. 2 to around No. 7 of the 30 columns of printed OTFTs. The transfer characteristics and optical micrograph of the best and the worst devices from the same line of the OTFT array are presented in Fig. 7. Compared to normal devices showing the typical self-aligned crystals strips, the semiconductor layer of an inferior device is composed of disordered polycrystalline, and this type of morphology was more frequently observed when printing with
Fig. 6. Cross-section profile of the single-drop film printed via o-DCB on hydrophobic surfaces: (a) PVP-coated SiO2 and (b) PETS-SAM-treated SiO2 substrate.
Fig. 7. Transfer curves under cycle-swept Vg and the cross-polarized microscopy of the best and the worst device within the same line of the printed OTFT arrays on PVP/SiO2 substrate.
an ink that had been aged for a few hours in the cartridge. Solvent compatibility of the black plastic material that constitutes the printhead of the cartridge was subsequently tested; obvious dissolution of the plastic is observed in o-DCB. Intuitively, plastic dissolved in the ink solution constitutes an impurity that would degrade the semiconductor film morphology and its electrical properties.
3.5. Further optimization via solvent selection The issue with the o-DCB solvent suggests that dissolution of the cartridge material by the ink solution should be avoided. Milder solvent such as toluene, anisole, xylene and tetralin are potential candidates. Of these, tetralin is more promising for inkjet printing via tiny droplets because of its high boiling point that favors self-organization of the semiconductor molecules as crystalline films. The multi-line film of TIPS-PEN printed via tetralin on a PVP/SiO2 substrate is polycrystalline with domains in the shape of wide strips (20–100 μm in width) when printed at a substrate temperature of 35 °C. The crystal growth direction runs parallel with the y-direction of the printed pattern, therefore, top-contact/bottom-gate OTFT architecture is fabricated with the channel perpendicular to the y-direction of the printed film. The typical channel width and length are 250 μm and 84 μm respectively for these devices. Devices printed at a lower substrate temperature of 35 °C are generally superior to those printed at a higher temperature of 45 °C. The average mobility of devices printed at 35 °C is enhanced by one order of magnitude to a value of 0.36 cm2V− 1 s− 1 compared with those printed with the o-DCB ink, while the maximum mobility reaches 0.78 cm2V−1 s−1, which is about a 7-fold increase. The threshold voltage is − 21 V as estimated from the transfer curve in Fig. 8. Mobility distribution of these OTFTs is shown in Fig. 9. OTFTs printed at 35 °C Films outperforms those printed a 45 °C owing to the larger film thickness. As shown in Fig. 10, the increasing diameter of the isolating dot film printed via tetralin indicates a decreasing film thickness with increasing substrate temperature. Among the 36 devices printed at 35 °C, 35 devices exhibit typical OTFT characteristics with ~95% of the devices falls within the range of 0.1–0.8 cm2V−1 s−1. Compared to devices printed with an o-DCB ink, the tremendous improvement in mobility as well as the more uniform device performance is at least partially ascribed to the tetralin solvent that possesses a higher surface energy than o-DCB, providing a wider process window for the tuning of single-drop film diameter and film thickness as seen in Fig. 10.
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
17
Fig. 10. Diameters of single-drop films printed via the single-solvent solution of o-DCB or tetralin on PVP-coated substrate with increasing substrate temperature. Fig. 8. Transfer curves under cycle-swept Vg of the highest performance device with the accumulation curve shown in open squares and depletion curve in open circles, inset: the cross-polarized microphotograph of the OTFT.
Film coverage of TIPS-PEN films inkjet-printed for top-contact OTFTs is characterized with AFMs as shown in Fig. 11. Almost 100% material coverage within the domains was observed for samples printed with the 2 wt.% tetralin solution on the PVP/SiO 2 substrate. When a
Fig. 9. Mobility distribution of 54 TIPS-PEN OTFTs fabricated on PVP/SiO2 substrate with the semiconductor printed via a 2 wt.% solution in tetralin. Inkjet printing was conducted with the substrate temperature maintained at 45 °C (18 devices) or 35 °C (36 devices).
2 wt.% o-DCB solution is employed, the crystal ribbons become narrower in width and are not fully grown but the crystal growth direction is well aligned. In contrast, crystal domains are not as well aligned over large areas on an untreated hydrophilic SiO2 substrate. 4. Discussion Film formation starts from the edge of the liquid when the critical concentration required for nucleation at the contact line is reached during the drying process [22]. On a wettable substrate, single-drop film diameter is found to be determined by the dispersion contribution of the surface energy. Herein, a simple model is proposed that qualitatively explains the shape of the deposits and the dependence on process parameters such as substrate temperature. Depending on the magnitude of the dispersion component and the polarity of the surface energy, there are generally two types of drying process after a droplet impact. In the first case, the solid surface is hydrophilic with a relatively high polarity, solvent-substrate interaction dominates and nucleation is relatively difficult until a relatively high concentration is reached locally as the solvent evaporates and the outward flow thereof causes solute accumulation toward the contact line. The coffee-stain profile is typically observed especially at higher substrate temperatures as the evaporation rate is enhanced and the outward flow intensified. In the other case, the solid surface is hydrophobic with a dominative dispersion component of surface energy, interaction between solute molecules and the substrate becomes predominant and pinning happens at a lower solute concentration. As a result, a larger film is deposited with more uniform film thickness, and the cross section profile is less sensitive to the evaporation induced flow at elevated substrate temperatures. The crystallographic morphology of the multiple-line film is studied with out-of-plane XRD showing the typically observed (001) diffraction peaks [23]. The side groups of the TIPS-PEN molecules are attracted by the surface of the substrate presumably via a dispersive force that induced contact-line pinning and self-assembling of the molecules in an “edge-on” orientation. The side group of TIPS-PEN is hydrophobic and is disposed to interact with the substrate through the dispersion force. For high performance devices especially those printed via the tetralin solution, thermal annealing at a higher temperature or vacuum annealing at room temperature does not improve on-current or the saturated field-effect mobility, but a few hours of vacuum annealing would cause a shift in threshold voltage toward zero volts and a reduced hysteresis in the transfer curve. Independence of high mobility to annealing indicates that the system was in a state of thermodynamical equilibrium after drying on the substrate when the semiconductor molecules self-assembled in the “edge-on” orientation. Vacuum
18
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
Fig. 11. AFM images (50 × 50 μm2) of large-area TIPS-PEN films printed with overlapping multiple lines on 3 substrates (Left: via o-DCB on untreated hydrophilic SiO2, middle: via o-DCB on PVP/SiO2, right: via tetralin on PVP/SiO2).
annealing is beneficial to improve interface quality as most of the residue solvent is removed. In this work, OTFTs of highest performance are printed at a low temperature of 35 °C on PVP/SiO2 using the TIPS-PEN/tetralin solution. The diameter of the single-drop film on PVP still increases with increasing substrate temperature, a trend similar to that observed from the hydrophilic substrates. It should be noted that the single-drop film diameter is heavily dependent on solvent selection. At a substrate temperature of 35 °C, the diameter decreases from around 88 μm to 50 μm (Fig. 10) when using a tetralin as the solvent instead of o-DCB, while the thickness uniformity is maintained. It is supposed that the critical value of concentration required for contact-line pinning on PVP is independent on the initial concentration of the droplet, but critical to the surface energy of the solvent and temperature of the substrate. The critical concentration for nucleation from a tetralin solution is higher than that from an o-DCB medium because the inward surface tension of tetralin is higher. Therefore, delayed contact-line pinning and higher film thickness are achieved when printing with a tetralin solution. Furthermore, the critical concentration is tunable across a wider range by printing at different substrate temperatures, whereby a suitable timing for contact-line pinning becomes possible. Incomplete film coverage within the channel of an OTFT or within a crystal domain as well as insufficient film thickness or uniformity is obviously a primary obstacle to the printing of high-performance devices. Specifically, thickness uniformity is limited when printing on a hydrophilic substrate, while insufficient film thickness or incomplete material coverage is inevitable when printing on a hydrophobic substrate where premature contact-line pinning takes place. With the highest mobility up to 0.8 cm2V−1 s−1 achieved at 35 °C on PVP/SiO2 substrate, further optimization on drop spacing and the stroke length would improve the average mobility of a large-scale device array up to 0.5 cm2V−1 s−1. At a substrate temperature of 35 °C, the mobility averaged on a line was lower for the first line (0.3 cm2V−1 s−1 on average) but the average value can reach 0.5 cm2V− 1 s− 1 for the following lines printed without process idling, which is over 60% of the maximum mobility (0.8 cm2V−1 s−1). The percentage is used as a measurement of performance uniformity, which is comparable to vacuum deposited organic semiconductors and obviously higher than the previously reported values for solution prepared devices including inkjet printing [24] and spin coating [25]. TIPS-PEN OTFTs of similar field-effect mobility can be prepared by inkjet-printing over a spin-coated poly(α-methylstyrene) (PαMS) on SiO2 [26]. The locally dissolved PαMS-coatsing by the ink solution functions as a self-aligned bank, which contributes to a uniform film thickness of the inner plateau and also higher precision and pattern fidelity of the printed film. The improved switch-on characteristics compared to the reference device (without PαMS) are ascribed to the high quality interface with the substrate. Compared to the current
work, the difference in drying behavior of the blend ink formed on PαMS/SiO2 is corroborated by the negative dependence of deposit diameter on substrate temperature. Although the highest mobility is comparable, several differences are noteworthy. Firstly, devices of the highest performance are printed at a higher substrate temperature of 70 °C on PαMS-coated substrates, especially on the central circular region with a thickness as low as ~ 20 nm. Owing to the hydrophobic surface with a sufficiently high dispersion component of surface energy, the PVP layer contributes to improved thickness uniformity across the overall printed areas as well as a high film coverage when TIPS-PEN is printed via tetralin at a low substrate temperature of 35 °C. Secondly, the PVP-coating is intact and therefore a smooth interface with the semiconductor persists, which can reasonably lead to enhanced fieldeffect mobility, although the inferior steepness of the subthreshold slope still indicate a poor interface quality. A potential solution via interfacial nucleation on hydrophobic surfaces has been proposed recently [27]. In addition to surface modification, PVP can be easily spin-coated on almost any substrates providing uniform surface energies over large areas and can be directly employed as the gate dielectric in OTFTs. Therefore, the application of PVP-based process is promising for largescale fabrication of integrated flexible electronics.
5. Conclusions In summary, process optimization are performed for inkjet printing of TIPS-PEN via surface engineering of the substrate, solvent selection as well as optimization of the printing parameters such as drop spacing, line spacing and substrate temperature. The coffee-ring profile is totally eliminated when printing on a hydrophobic surface such as PETStreated SiO2 substrate or on a PVP-coated sample due to the strong dispersion force between the phenyl-groups on the solid surface and the side groups of the semiconductor molecules that contribute to contact-line pinning at lower solute concentrations. An optimized process condition is formulated for picoliter inkjet printing of TIPSPEN that uses a PVP-coating as the substrate and tetralin as the solvent. Dramatic improvement in saturated field-effect mobility is achieved when printing at a low substrate temperature of 35 °C where optimal film morphology with complete film coverage is achieved.
Acknowledgments This work was supported by the National Natural Science Foundation of China (51203039, 61107014, 21204017), National Basic Research Program of China (Grant No. 2012CB723406), Program for New Century Excellent Talents in University (Grant No. NCET-12-0839).
X. Wang et al. / Thin Solid Films 578 (2015) 11–19
References [1] M.M. Payne, S.R. Parkin, J.E. Anthony, C.C. Kuo, T.N. Jackson, Organic field-effect transistors from solution-deposited functionalized acenes with mobilities as high as 1 cm2/V s, J. Am. Chem. Soc. 127 (2005) 4986. [2] K.C. Dickey, J.E. Anthony, Y.-L. Loo, Improving organic thin-film transistor performance through solvent-vapor annealing of solution-processable triethylsilylethynyl anthradithiophene, Adv. Mater. 18 (2006) (1721- +). [3] H. Minemawari, T. Yamada, H. Matsui, J.y. Tsutsumi, S. Haas, R. Chiba, R. Kumai, T. Hasegawa, Inkjet printing of single-crystal films, Nature 475 (2011) 364. [4] S.Y. Cho, J.M. Ko, J. Lim, J.Y. Lee, C. Lee, Inkjet-printed organic thin film transistors based on TIPS pentacene with insulating polymers, J. Mater. Chem. C 1 (2013) 914. [5] D.H. Kim, D.Y. Lee, H.S. Lee, W.H. Lee, Y.H. Kim, J.I. Han, K. Cho, High‐mobility organic transistors based on single‐crystalline microribbons of triisopropylsilylethynyl pentacene via solution‐phase self‐assembly, Adv. Mater. 19 (2007) 678. [6] W.H. Lee, D.H. Kim, Y. Jang, J.H. Cho, M. Hwang, Y.D. Park, Y.H. Kim, J.I. Han, K. Cho, Solution-processable pentacene microcrystal arrays for high performance organic field-effect transistors, Appl. Phys. Lett. 90 (2007) (132106-132106-132103). [7] S.K. Park, T.N. Jackson, J.E. Anthony, D.A. Mourey, High mobility solution processed 6,13-bis(triisopropyl-silylethynyl) pentacene organic thin film transistors, Appl. Phys. Lett. 91 (2007). [8] G. Giri, E. Verploegen, S.C. Mannsfeld, S. Atahan-Evrenk, S.Y. Lee, H.A. Becerril, A. Aspuru-Guzik, M.F. Toney, Z. Bao, Tuning charge transport in solution-sheared organic semiconductors using lattice strain, Nature 480 (2011) 504. [9] Y. Diao, B.C. Tee, G. Giri, J. Xu, H.A. Becerril, R.M. Stoltenberg, T.H. Lee, G. Xue, S.C. Mannsfeld, Z. Bao, Solution coating of large-area organic semiconductor thin films with aligned single-crystalline domains, Nat. Mater. 12 (2013) 665. [10] R.Z. Rogowski, A. Dzwilewski, M. Kemerink, A.A. Darhuber, Solution processing of semiconducting organic molecules for tailored charge transport properties, J. Phys. Chem. C 115 (2011) 11758. [11] X. Sun, C.-a. Di, Y. Liu, Engineering of the dielectric-semiconductor interface in organic field-effect transistors, J. Mater. Chem. 20 (2010) 2599. [12] M. Halik, H. Klauk, U. Zschieschang, G. Schmid, C. Dehm, M. Schutz, S. Maisch, F. Effenberger, M. Brunnbauer, F. Stellacci, Low-voltage organic transistors with an amorphous molecular gate dielectric, Nature 431 (2004) 963. [13] H. Klauk, U. Zschieschang, J. Pflaum, M. Halik, Ultralow-power organic complementary circuits, Nature 445 (2007) 745. [14] X. Sun, Y. Liu, C.-a. Di, Y. Wen, Y. Guo, L. Zhang, Y. Zhao, G. Yu, Interfacial heterogeneity of surface energy in organic field-effect transistors, Adv. Mater. 23 (2011) 1009. [15] J.A. Lim, W.H. Lee, H.S. Lee, J.H. Lee, Y.D. Park, K. Cho, Self-organization of ink-jetprinted triisopropylsilylethynyl pentacene via evaporation-induced flows in a drying droplet, Adv. Funct. Mater. 18 (2008) 229.
19
[16] C. Liu, Y. Li, T. Minari, K. Takimiya, K. Tsukagoshi, Forming semiconductor/dielectric double layers by one-step spin-coating for enhancing the performance of organic field-effect transistors, Org. Electron. 13 (2012) 1146. [17] J. Lee, J.Y. Jung, D.H. Kim, J.-Y. Kim, B.-L. Lee, J.-I. Park, J.W. Chung, J.S. Park, B. Koo, Y.W. Jin, S. Lee, Enhanced electrical stability of organic thin-film transistors with polymer semiconductor-insulator blended active layers, Appl. Phys. Lett. 100 (2012). [18] X. Li, W.T.T. Smaal, C. Kjellander, B. van der Putten, K. Gualandris, E.C.P. Smits, J. Anthony, D.J. Broer, P.W.M. Blom, J. Genoe, G. Gelinck, Charge transport in highperformance ink-jet printed single-droplet organic transistors based on a silylethynyl substituted pentacene/insulating polymer blend, Org. Electron. 12 (2011) 1319. [19] M.-B. Madec, P.J. Smith, A. Malandraki, N. Wang, J.G. Korvink, S.G. Yeates, Enhanced reproducibility of inkjet printed organic thin film transistors based on solution processable polymer-small molecule blends, J. Mater. Chem. 20 (2010) 9155. [20] Y.-H. Kim, J.E. Anthony, S.K. Park, Polymer blended small molecule organic field effect transistors with improved device-to-device uniformity and operational stability, Org. Electron. 13 (2012) 1152. [21] X.H. Wang, X.F. Xiong, L.Z. Qiu, G.Q. Lv, Morphology of inkjet printed 6,13 bis(triisopropylsilylethynyl) pentacene on surface-treated silica, J. Vac. Sci. Technol. B 30 (2012) 021206. [22] J.A. Lim, W.H. Lee, D. Kwak, K. Cho, Evaporation-induced self-organization of inkjet-printed organic semiconductors on surface-modified dielectrics for highperformance organic transistors, Langmuir 25 (2009) 5404. [23] J. Chen, C.K. Tee, M. Shtein, D.C. Martin, J. Anthony, Controlled solution deposition and systematic study of charge-transport anisotropy in single crystal and singlecrystal textured TIPS pentacene thin films, Org. Electron. 10 (2009) 696. [24] Y.H. Kim, B. Yoo, J.E. Anthony, S.K. Park, Controlled deposition of a high‐performance small‐molecule organic single‐crystal transistor array by direct ink‐jet printing, Adv. Mater. 24 (2012) 497. [25] Y. Yuan, G. Giri, A.L. Ayzner, A.P. Zoombelt, S.C.B. Mannsfeld, J. Chen, D. Nordlund, M.F. Toney, J. Huang, Z. Bao, Ultra-high mobility transparent organic thin film transistors grown by an off-centre spin-coating method, Nat. Commun. 5 (2014) 3005. [26] B.K.C. Kjellander, W.T.T. Smaal, J.E. Anthony, G.H. Gelinck, Inkjet printing of TIPSPEN on soluble polymer insulating films: a route to high-performance thin-film transistors, Adv. Mater. 22 (2010) 4612. [27] X.H. Wang, M. Yuan, S.C. Lv, M.Z. Qin, M.J. Chen, L.Z. Qiu, G.B. Zhang, H.B. Lu, Interfacial nucleation behavior of inkjet-printed 6,13 bis(tri-isopropylsilylethynyl) pentacene on dielectric surfaces, J. Appl. Phys. 117 (2015) 024902.