Fine electrically-conductive patterns of silver nanoparticles ink have been laser printed using the laser-induced forward transfer (LIFT) technique. LIFT is a technique that offers the possibility of printing patterns with high spatial resolution from a wide range of materials in solid or liquid state. Influence of drying the ink film, previous to its transfer, on the printed droplet morphology is discussed. The laser pulse energy and donor-receiver substrate separation were systematically varied and their effects on the transferred droplets were analyzed. The use of an intermediate titanium dynamic release layer was also investigated and demonstrated the possibility of a better control of both the size and shape of the printed patterns. Conditions have been determined for printing flat-top droplets with sharp edges. 21 µm width silver lines with 80 nm thickness have been printed with a smooth convex profile. Electrical resistivities of the transferred patterns are only 5 times higher than the bulk silver.
© 2011 OSA
Lines or patterns are directly generated by depositing individual droplets at desired locations. There have been many studies on inkjet-printed droplets and lines using various conductive metal nanoparticles-based inks with narrow width and consistent morphology. A drawback of the liquid printing appeared with the observation of the so-called “coffee stain” effect. This effect is due to a differential evaporation of solvent resulting in a dual-peaked height profile of the droplet. Physical models  of the spreading and drying of droplets have been developed to investigate this effect [2–8]. Combined actions of evaporation rate difference and contact line are found to be responsible. The solvent composition also plays an important role. The use of a high boiling point solvent with low surface tension such as ethylene glycol have been seen to enable the formation of uniform deposits of silver nanoparticles due to surface tension gradient-induced inward Marangoni flow, which can compensate outward convective flow . A diminution of the effect has also been demonstrated by mixing two solvents such as high-boiling-point and low-boiling-point solvents [10, 11]. During the rapid evaporation of the more volatile solvent, the less volatile solvent is gradually concentrated and, because of its slow evaporation, allows to maintain a consistent concentration on the whole drop. Substrate temperature has also been demonstrated to be crucial for the stability of lines in affecting the drying of liquid droplets [12–15]. Conventional approaches for printed electronics include screen and ink-jet printing, both, which facilitate the low cost manufacturing of large-area, flexible devices. While attractive for many applications, both techniques are limited in processable parameters, in particular on variation of the ink viscosity. Direct-laser writing based on the deposition of inks is a compelling alternative as it allows transferring rheological materials with a wide range of viscosity.
This work is focused on the printing of silver nanoparticle ink under ambient condition using laser-induced forward transfer (LIFT) technique . LIFT was initially conceived to deposit simple inorganic solid films that were transferred in a process of vaporization and further recondensation. The extension of the concept to liquid films practically eliminated any restriction concerning the materials which could be deposited with it . In such cases, the material is propelled by the laser pulse to the substrate without any phase change, which allows depositing complex materials without degrading their properties. The LIFT acts as a microprinting technique, in a similar way to inkjet technology, but with the advantages of printing large set of materials with different properties and with potentially higher resolution.
The system consists of the laser transfer of material from a donor substrate placed in close proximity to a receiving substrate. We discuss the printed structure evolution when varying the ink viscosity by drying under air at room temperature, leading to an increase in the concentration of metallic content in the ink. The electrical properties of printed lines on silicon substrates are characterized. LIFT has been shown particularly attractive for the micro deposition of a broad variety of materials [18–21], especially liquid solutions [22–24]. Laser printing of silver nanoparticles ink lines has already been achieved , but profiles of these lines displayed height non-uniformity. Profilometer scans revealed a two-peak behavior reflecting the “coffee stain” effect. This behavior was reduced using a second-pass cw laser to cure the deposited lines and resulted in a smoother line height profile than the oven-curing lines. A previous study demonstrated the printing of lines of ink composed of silver nanoparticles (~20 wt %) in ethanol solvent for the fabrication of electrodes for transistors and these droplets displayed this two-peak behavior as well . Another approach is to print pastes containing nanoparticles. Well-defined lines and patterns have been transferred with only 5 µm width using pastes [27, 28]. Lines made with juxtaposed deposits presented a high uniformity and the boundaries across them were barely visible, except when there was excessive overlap between adjacent deposits.
In this study, we investigate the influence of the metallic nanopaticles content in the ink on the transfer characteristics and demonstrate the possibility of the direct laser printing of silver nanoparticles ink to generate flat-top droplets and lines with uniform smooth convex profile.
2. Experimental details
The pulsed-laser direct-write system used in this work has been described previously . The laser is a Nd:YAG 50 ps operating at 355 nm at 10 Hz repetition rate. The donor substrate used was a 12 x 12 mm2 UV-transparent quartz substrate (suprasil) which was placed to within a few microns gap from a receiving substrate using a vacuum frame holder. The SunTronic silver nanoparticles ink used in this work was supplied by Sun Chemical. The ink has silver solids content of 20 wt % and nanoparticle sizes around 100 nm which are important for lowering their melting and sintering temperatures. The solvents were ethanediol, ethanol, glycerine and 2-isopropoxyethanol. The viscosity of the ink was 10 to 13 cps. The donor substrates were prepared by spin-coating the ink onto quartz substrates (20 s at 3000 rpm) resulting in a homogeneous thin film of 1.2 to 1.4 μg of ink on the suprasil substrate, with a corresponding 0.24 to 0.28 µg of silver content and film thickness of 5 µm. Then it was carefully placed with the ink layer facing the receiver substrate.
As the silver nanoparticles were in solution, the liquid solvent would act simultaneously as energy absorber and transport vector. We studied the influence of the metallic concentration by decreasing the proportion of solvent, and to keep the energy absorption properties constant we investigated the transfer of the ink using an intermediate UV-absorbing sacrificial layer, also known as dynamic release layer (DRL). The use of absorbing layers of materials such as gold, silver  or titanium  had produced good results in the laser transfer of various materials including sensitive materials . We used a titanium (Ti) film as the DRL. For this case, the donor was prepared by spin-coating the ink on a suprasil substrate coated with 30-nm thick titanium film, which was deposited by thermal evaporation. The donor substrate was placed in a position where the Gaussian laser beam presented a diameter of 20 µm at the interface ink-quartz substrate, or at the surface of Ti layer when DRL is used, and this beam size was kept constant for all the experiences.
The quality of the laser-printed silver structures was first assessed by optical microscopy. The morphology was then characterized by scanning electronic (SEM) and atomic force (AFM) microscopies.
3. Results and discussion
3.1 Laser printing of the ink
We first studied the ink transfer without any DRL. Figure 1 exhibits a laser-printed microarray for different pulse energies at a donor-receiver substrate separation (d) of 50 µm. The best transfer condition is close to the transfer threshold conditions. Thus, pulse energies of 5 µJ (1.5J/cm2), just above the transfer threshold, always produced uniform, well-defined and circular droplets. The quality and shape of droplets depended directly on the laser pulse energy. When the energy was increased (higher than 10 μJ), uniformity was lost and the droplets presented irregular shapes. The variation of the size of the droplets with the laser fluence is displayed in Fig. 2a . The size of droplets increased with the fluence and debris appeared all around the droplets. Similar behavior has already been reported by Dinca et al. for laser printing of protein containing solutions . The electronic absorption spectrum of thin ink layer has an absorption band around 400 nm, giving rise to an absorption depth of 1 µm at 355 nm. The laser energy is transferred to a large volume of the ink. A vapor bubble is generated inside the ink layer, which then expands causing a mechanical ejection of a jet of liquid. When the fluence is increased, the mechanical stress inducing the ejection of the liquid is higher. The vapor bubble reaches the surface with too high energy and do not lead to the formation of a jet . The laser energies used correspond to fluences varying from 1.5J/cm2 to 6J/cm2 as previously observed for liquid printing . However these fluences are higher than those used when silver NPs inks are transferred as paste . That confirms the jet formation mechanism in our conditions.
When printed, the silver nanoparticles are not yet bonded together. The bonding process occurs when the printed structure is heated to 150°C or higher. During thermal annealing, the nanoparticles fuse into an interconnected structure. Thus, the transferred patterns were cured at a temperature of 150°C during 30 min (Fig. 1b). Dots as small as 25 μm diameter were obtained for laser energy of 5 μJ. Figure 3 shows AFM image of a 25 µm dot after curing. The thickness was measured to be 200 nm and the AFM profiles revealed a small coffee stain effect with an average height difference of 15 nm between the edges and center. No splashes or ejected particles were observed. The volume of this dot was calculated to be ~70 µm3. The transferred droplet volume displayed a dependence on the laser pulse energy, as already described in previous studies on laser printing of liquids [22,32,35].
3.2 Laser printing with titanium DRL
The minimum energy required to print uniform droplets with circular shape was 5μJ (1.5J/cm2). Dots with 25 µm diameter were obtained at laser energy of 5 µJ using Ti as DRL as shown in Fig. 4a and b . On increasing laser energy, the size of the dots increased but no morphological changes were observed (Fig. 2a). The transferred patterns kept their circular shape, reaching a size of 56 µm at energy of 20 µJ (Fig. 4c). Very little or no debris was associated with the dots. A precise control of the droplet diameter was achieved through the variation of the laser pulse energy without a loss in the quality of the droplet morphology. A linear dependence of the droplets volume on the laser fluence had already been demonstrated by Duocastella et al. . During the LIFT process, the laser pulse ablates the titanium layer. A confined plasma is created and its expansion leads to the generation of a cavitation bubble in the liquid. The expansion of that bubble provides the impulsion required to propel the adjacent liquid film in a thin and stable jet resulting in the formation of a droplet. The dimensions of the cavitation bubble increase with laser energy and by this way the amount of liquid pushed by the vapor increases, resulting in the transfer of bigger droplets. The use of the DRL allows keeping uniform and circular droplets when the energy increases. There is no direct laser energy absorption by the ink itself, and that strongly reduces the thermal and jetting effects. AFM characterizations of the printed droplets revealed the same “stain” effect as observed previously with the no-DRL printed droplets after curing.
3.3 Influence of the film to substrate spacing
The distance between the donor film and the receiver substrate is another critical parameter that affects the morphology of the printed droplets. The donor film to receiver substrate spacing was varied from 50 μm to 400 μm and Fig. 5 presents arrays of droplets printed with and without Ti DRL at 5 µJ.
Without Ti-DRL, donor-receiver distances higher than 100 µm lead to splashy droplets with irregular shapes and small adjacent satellites. The size of the printed-pattern also increased together with the distance, as shown in Fig. 2b.
With Ti-DRL, circular droplets with uniform surface were obtained reproducibly up to the maximum donor-receiver separation of 400 µm and only small isolated satellites were observed around few droplets at higher separations. As shown in Fig. 2b, the Ti-DRL improved the possibility of laser printing of good quality droplets even at large separations between donor/receiver substrates.
On investigating the donor substrates circular holes were observed on the ablated areas. The removal of the metallic film suggests that the dots might contain titanium residues.
3.4 Influence of the metallic concentration
In summary, all these considerations indicate that the optimum working conditions for laser printing of silver nanoparticles ink correspond to laser pulse energies slightly higher than the transfer threshold. Uniform quality circular droplets without any splashing or satellites are printed reproducibly at these fluence values. A further improvement in quality was observed with Ti-DRL and good control of droplet dimensions was achieved through the laser pulse energy. As the experiments were performed under ambient condition in air, the properties of the ink were affected by the evaporation of the solvents. When the solvents evaporated, the concentration of the metallic content in the droplet increased.
We investigated the ink printing as a function of the concentration of the silver nanoparticles contained in the solution by drying. Donor substrates with thin film of silver nanoparticles ink were dried in air and ambient condition. Arrays of microdroplets were laser-printed every hour. After 4 hours, a loss of 30% in the mass of ink was measured corresponding to a silver content of 30 wt. Laser pulse energies just above the transfer threshold generate droplets with well-defined circular shape. AFM investigations after curing 20 min at 150°C revealed that all the printed droplets had flat top hat shapes, as shown in Fig. 6 . The diameter of the dots was measured to be 21 µm and 30 µm for laser energies of 5 µJ and 10 µJ respectively. The “stain” effect completely disappeared and homogeneous droplets with smooth top profile with a very low roughness were obtained. The sintering process did not affect the dot diameters. As earlier, when droplets were printed without any DRL, laser energy increased the dimensions of the pattern (Fig. 2a) with rapid loss of the circular shape and the spatial resolution. Splashes appeared over a large area around printed droplets at higher fluences. With Ti-DRL, very little or no debris was observed at higher fluences (Fig. 2a), and the size of the pattern was better controlled. From 15 µJ, the circular shape was lost and splashes appeared all around. At 20 µJ, the size of the transferred pattern was 100 µm and the edges were irregular, corresponding to approximately twice of the size of a transferred droplet without drying. The thickness of a 27 µm diameter dot was measured to be 320 nm after curing at 150°C during 20 min and the average volume was calculated to be 210 μm3.
Printing using Ti-DRL helped to control both the size and shape of the printed pattern, even on increasing the donor-receiver substrates separation, as shown in Fig. 2b. This behavior was observed until 50% of loss of ink mass, corresponding to silver content of ~40 wt % and ink lifetime of approximately two days. Beyond ~50 wt % of silver content in the ink, poor spatial resolution of the droplets was observed.
On printing two successive droplets with an overlap of ~3 µm they fused together resulting into a straight line with a very homogeneous surface and thickness, shown in Fig. 7 for two droplets of 27 µm diameter as printed with a center to center separation of 51 µm at laser energy of 8 µJ. AFM analysis revealed that this two-droplets line had uniform sharp edges and smooth flat top profile and a thickness of 360 nm. The average roughness was 11 nm with a peak-to-valley ratio of 37 nm. Further decreasing the drop spacing to an overlapping of 10 µm allows eliminating the scalloping and lead to a smooth, straight line. The laser-printed silver nanoparticles ink is still in a fluid form when it reaches the surface of the receiver substrate that allows the droplets to readily fuse with each other. In this condition, lines were laser transferred at laser energy of 5 µJ. Figure 8a presents AFM analysis of a printed silver line with a width of 18 ± 3 µm. The line presents smooth convex profiles without the coffee stain effect, similar to the profile of a single flat top droplet. The line had an average thickness of only 80 nm, a lower thickness than a 21 µm single droplet. The average roughness was 1 nm with a peak-to-valley ratio of 5 nm. Two printed lines separated by 20 µm are presented in Fig. 8b. As expected, AFM characterization of the channel revealed a clean area, without debris, satellites or splashes, over the 40 µm scan length area and was representative of the whole channel. Straight lines with length from 45 µm to 2 mm were laser printed. Figure 9a shows a series of 200 µm length lines with different separation distances on polyethylene terephthalate substrate demonstrating a good quality of the lines, even on rough surfaces. The minimum line pitch observed depends on the receiver substrate properties (wettability, roughness), but a separation of 5 μm has always been reached. Figures 9b and 9c present the realization of 2-dimensional patterns with interconnected silver lines. Printing these ink lines in liquid phase and with high viscosity, allows keeping the width and the thickness of the line uniform, even at the crossing points. As the design of the patterns is defined by the motion of the receiver substrate, there is no limitation for printing any shape (circular, grid …), expect the line pitch.
Characterizations of printed lines were performed with 4-probe measurements. Curing at a low temperature of 150 °C during 20 min was found sufficient to reach the lowest resistivity value. For lines of 21-μm width, 80-nm thickness and lengths from 50 µm to 2 mm, an average resistivity of 9 ± 1 μΩ·cm was measured, which is ≈5 times of bulk silver (1.59 µΩ.cm). The obtained AgNP resistivity was close to what has been obtained already for AgNP lines deposited by laser printing (20 μΩ.cm)  and ink-jet printing (17 μΩ.cm for an annealing at 130°C) , and is of the same order as for the laser-printed silver paste (5.5 μΩ.cm) . The laser printed silver lines could serve as conducting lines for various electronic applications.
20 μm width and 80 nm high silver lines with very smooth and uniform profile have been transferred using LIFT technique. The study showed that circular and uniform droplets are obtained with good reproducibility at laser pulse energies slightly above the transfer threshold. The droplet diameter can be precisely controlled through the variation of the laser pulse energy, without loss in the quality of the droplet morphology. Its ability to print both low and high viscosity ink demonstrates its versatility as a direct-write technique. The smallest droplet diameter obtained is ~20 μm and that exhibited a flat-top profile. Ti-DRL improved the transfer characteristics. This DRL provided a high degree of flexibility and a wider range of donor-receiver distances where transfer can be carried out without loss of resolution. Electrical resistivities of the laser-printed lines in the range of 4-5 times bulk silver were obtained.
This work has been carried out within the FP7 European project e-LIFT (project n°247868 –call FP7-ICT-2009-4) and the DGE-FUI project I2FLEX.
References and links
1. M. Ikegawa and H. Azuma, “Droplet behaviors on substrates in thin film formation using ink-jet printing,” Trans. Jpn. Soc. Mech. Eng. Ser. B 47(3), 490–496 (2004).
2. R. D. Deegan, O. Bakajin, T. F. Dupont, G. Huber, S. R. Nagel, and T. A. Witten, “Capillary flow as the cause of ring stains from dried liquid drop,” Nature 389(6653), 827–829 (1997). [CrossRef]
3. R. D. Deegan, O. Bakajin, T. F. Dupont, G. Huber, S. R. Nagel, and T. A. Witten, “Contact line deposits in an evaporating drop,” Phys. Rev. E Stat. Phys. Plasmas Fluids Relat. Interdiscip. Topics 62(11 Pt B), 756–765 (2000). [CrossRef] [PubMed]
6. R. A. Street, W. S. Wong, S. E. Ready, M. L. Chabinyc, A. C. Arias, S. Limb, A. Salleo, and R. Lujan, “Jet printing flexible displays,” Mater. Today 9(4), 32–37 (2006). [CrossRef]
7. J. Perelaer, B. J. de Gans, and U. S. Schubert, “Ink-jet printing and microwave sintering of conductive silver tracks,” Adv. Mater. (Deerfield Beach Fla.) 18(16), 2101–2104 (2006). [CrossRef]
8. B. J. Kang and J. H. Oh, “Geometrical characterization of inkjet-printed conductive lines of nanosilver suspensions on a polymer substrate,” Thin Solid Films 518(10), 2890–2896 (2010). [CrossRef]
9. D. Kim, S. Jeong, B. K. Park, and J. Moon, “Direct writing of silver conductive patterns: Improvement of film morphology and conductance by controlling solvent compositions,” Appl. Phys. Lett. 89(26), 264101 (2006). [CrossRef]
12. J. Chung, S. Ko, C. P. Grigoropoulos, N. R. Bieri, C. Dockendorf, and D. Poulikakos, “Damage-free low temperature pulsed laser printing of gold nanoinks on polymer,” J. Heat Transfer 127(7), 724 (2005). [CrossRef]
13. S. Molesa, D. R. Redinger, D. C. Huang, and V. Subramanian, “High-quality inkjet-printed multilevel Interconnects and inductive components on plastic for ultra-low-cost RFID applications,” MRS. Symp. Proc., 769, H8.3.1 (2003).
14. A. M. J. van den Berg, A. W. M. de Laat, P. J. Smith, J. Perelaer, and U. S. Schubert, “Geometric control of inkjet printed features using a gelating polymer,” J. Mater. Chem. 17(7), 677 (2007). [CrossRef]
16. J. Bohandy, B. F. Kim, and F. J. Adrian, “Metal-deposition from a supported metal-film using an excimer laser,” J. Appl. Phys. 60(4), 1538–1539 (1986). [CrossRef]
17. C. B. Arnold, P. Serra, and A. Piqué, “Laser direct-write techniques for printing of complex materials,” MRS Bull. 32(01), 23–32 (2007). [CrossRef]
18. R. Fardel, M. Nagel, F. Nüesch, T. Lippert, and A. Wokaun, “Fabrication of organic light-emitting diode pixels by laser-assisted forward transfer,” Appl. Phys. Lett. 91(6), 061103 (2007). [CrossRef]
19. A. Palla-Papavlu, V. Dinca, I. Paraico, A. Moldovan, J. Shaw-Stewart, C. W. Schneider, E. Kovacs, T. Lippert, and M. Dinescu, “Microfabrication of polystyrene microbead arrays by laser induced forward transfer,” J. Appl. Phys. 108(3), 033111 (2010). [CrossRef]
20. K. S. Kaur, R. Fardel, T. C. May-Smith, M. Nagel, D. P. Banks, C. Grivas, T. Lippert, and R. W. Eason, “Shadowgraphic studies of triazene assisted laser-induced forward transfer of ceramic thin films,” J. Appl. Phys. 105(11), 113119 (2009). [CrossRef]
21. C. Boutopoulos, C. Pandis, K. Giannakopoulos, P. Pissis, and I. Zergioti, “Polymer/carbon nanotube composite patterns via laser induced forward transfer,” Appl. Phys. Lett. 96(4), 041104 (2010). [CrossRef]
22. M. Colina, M. Duocastella, J. M. Fernández-Pradas, P. Serra, and J. L. Morenza, “Laser-induced forward transfer of liquids: Study of the droplet ejection process,” J. Appl. Phys. 99(8), 084909 (2006). [CrossRef]
23. J. M. Fernández-Pradas, M. Colina, P. Serra, J. Dominguez, and J. L. Morenza, “Laser-induced forward transfer of biomolecules,” Thin Solid Films 453–454, 27–30 (2004). [CrossRef]
24. M. Duocastella, A. Patrascioiu, J. M. Fernández-Pradas, J. L. Morenza, and P. Serra, “Film-free laser forward printing of transparent and weakly absorbing liquids,” Opt. Express 18(21), 21815–21825 (2010). [CrossRef] [PubMed]
25. R. C. Y. Auyeung, H. Kim, S. A. Mathews, and A. Piqué, “Laser direct-write of metallic nanoparticle inks,” J. Laser Micro/Nanoeng. 2(1), 1–21 (2007). [CrossRef]
26. L. Rapp, A. K. Diallo, A. P. Alloncle, C. Videlot-Ackermann, F. Fages, and P. Delaporte, “Pulsed-laser printing of organic thin-film transistors,” Appl. Phys. Lett. 95(17), 171109 (2009). [CrossRef]
27. R. C. Y. Auyeung, H. Kim, A. J. Birnbaum, M. Zalalutdinov, S. A. Mathews, and A. Piqué, “Laser decal transfer of freestanding microcantilevers and microbridges,” Appl. Phys., A Mater. Sci. Process. 97(3), 513–519 (2009). [CrossRef]
28. R. C. Y. Auyeung, H. Kim, N. A. Charipar, A. J. Birnbaum, S. A. Mathews, and A. Piqué, “Laser forward transfer based on a spatial light modulator,” Appl. Phys., A Mater. Sci. Process. 102(1), 21–26 (2011). [CrossRef]
29. L. Rapp, C. Cibert, A. P. Alloncle, and P. Delaporte, “Characterization of organic material micro-structures transferred by laser in nanosecond and picosecond regimes,” Appl. Surf. Sci. 255(10), 5439–5443 (2009). [CrossRef]
30. P. Serra, M. Colina, J. M. Fernández-Pradas, L. Sevilla, and J. L. Morenza, “Preparation of functional DNA microarrays through laser-induced forward transfer,” Appl. Phys. Lett. 85(9), 1639 (2004). [CrossRef]
31. V. Dinca, M. Farsari, D. Kafetzopoulos, A. Popescu, M. Dinescu, and C. Fotakis, “Patterning parameters for biomolecules microarrays constructed with nanosecond and femtosecond UV lasers,” Thin Solid Films 516(18), 6504–6511 (2008). [CrossRef]
32. M. Duocastella, A. Patrascioiu, V. Dinca, J. M. Fernández-Pradas, J. L. Morenza, and P. Serra, “Study of liquid deposition during laser printing of liquids,” Appl. Surf. Sci. 257(12), 5255–5258 (2011). [CrossRef]
33. M. Duocastella, J. M. Fernández-Pradas, J. L. Morenza, and P. Serra, “Time-resolved imaging of the laser forward transfer of liquids,” J. Appl. Phys. 106(8), 084907 (2009). [CrossRef]
34. J. Wang, R. C. Auyeung, H. Kim, N. A. Charipar, and A. Piqué, “Three-dimensional printing of interconnects by laser direct-write of silver nanopastes,” Adv. Mater. (Deerfield Beach Fla.) 22(40), 4462–4466 (2010). [CrossRef] [PubMed]
35. P. Serra, J. M. Fernández-Pradas, M. Colina, M. Duocastella, J. Domininguez, and J. L. Morenza, “Laser-induced forward transfer: a direct-writing technique for biosensors preparation,” J. Laser Micro/Nanoeng. 1(3), 236–242 (2006). [CrossRef]
36. H. Kim, R. C. Y. Auyeung, S. H. Lee, A. L. Huston, and A. Piqué, “Laser forward transfer of silver electrodes for organic thin-film transistors,” Appl. Phys., A Mater. Sci. Process. 96(2), 441–445 (2009). [CrossRef]
37. M. Barret, S. Sanaur, and P. Collot, “Inkjet-printed low-voltage organic thin-film transistors: towards low-cost flexible electronics,” Mater. Res. Soc. Symp. Proc. 1003 (2007).