Comparison of Laser-Ablation Techniques for Microstructuring Gravure-Printed PEDOT:PSS and Sputtered ITO on PET Substrates

Transcription

1 Comparison of Laser-blation Techniques for Microstructuring Gravure-Printed PEDOT:PSS and Sputtered ITO on PET Substrates Jukka Hast (1), Ilkka Vanttaja (2), Raimo Penttilä (2), Petri Laakso (2), Markku Känsäkoski (1) (1) Printed Functional Solutions, Oulu, (2) Laser Processing, Lappeenranta, VTT Technical Research Center of Finland, Finland nne-laure Seiler, Isabelle Chartier LITEN/DTNM /Printed Components Laboratory, CE Commissariat à l Energie tomique et aux Energies lternatives, Grenoble, France Petko Petkov, Georgi Lalev Manufacturing Engineering Centre (MEC), Cardiff University, Cardiff, UK Petronella Norberg Printed Electronics, creo, Norrköping, Sweden Thomas laudeck, Reinhard R. aumann Print and Media Technology, Chemnitz University of Technology, Chemnitz, Germany bstract In this paper we summarize the European research collaboration on the topic Integration of laser ablation in roll-to-roll manufacturing running from 2008 to Embedded into a case study for the manufacturing of anode structures for optoelectronic devices (OLED, OPV) on flexible substrates, we present our findings on the printing and laser process windows for the following cases: (i) structuring of gravure-printed PEDOT:PSS and (ii) structuring of sputtered ITO, each on flexible PET foil. The experiments were carried out with help of the following laser systems and plants available in the contributing institutes: (iii) Gaussian-profile ns-, ps- and fs- solid-state laser systems (focused laser beam) and (iv) a flat-profile ns excimer laser system for mask ablation. In this paper we present the results for case (i) structured by both (iii) and (iv) and for case (ii) structured by (iii). The remaining case (ii) structured by (iv) is found as a separate paper at this conference. 1. Introduction In 2008, as shown independently by the German company 3d-micromac and VTT Technical Research Center of Finland, laser micromachining has matured towards a technique with full roll-to-roll compatibility [1, 2]. In that viewpoint, this technology has gained interest for manufacturing of devices of organic and large-area electronics. However, the principles of laser micromachining are based on particular light-matter interaction and rely on both the deposited layer and the substrate beneath. Hence, they are subject for detailed empiric analysis for particular application cases. substrate was Toray s OPET (Lumirror), 125 μm thick. The printing cylinder layout is shown in Figure 1. It contains four different sections having different cylinder parameters as shown in Table 1 (upper section). 2. Gravure Printing of PEDOT:PSS and Characterization of the Printing Quality Gravure printing trials were done using VTT s ROKO R2R pilot production facility located in Oulu, Finland, shown in Figure 1. For the printing trials, modified Clevios PEDOT PHC99 from H.C. Starck was used. The used PET. Figure 1. () VTT's ROKO pilot production facility for gravure printing of PEDOT:PSS, () Gravure cylinder layout used in PEDOT:PSS printing trials. The numbers on the left refer to the rea ID used througout this Section. 1

2 Table 1. Gravure cell properties (upper section) and properties of the resulting printed layers of gravure-printed PEDOT:PSS sectors (lower section) consists of unwinder-rewinder units and a Lumera rapidablation laser with high-speed scanner optics. Gravure cylinder rea ID Engraving Mesh [lines/cm] Cell depth [µm] 2-1 X-treme ,2 2-2 Regular ,6 2-3 X-treme ,7 2-4 X-treme ,4 Cell volume [ml/m²] Resulting printed layer of PEDOT:PSS rea ID Engraving Thickness [nm] 2-1 X-treme Regular X-treme X-treme Roughness [nm] Tonal value The processing parameters during the trials were the following: Nip pressure: 2 bar Web speed: m/min 120 C: 20 sec The quality of the printing trials (with respect to layer thickness, roughness) was characterized using a Wyco whitelight optical profilometer. The results are listed in the Table 1 (lower section) including photo-micrographs of each printed PEDOT:PSS sector to illustrate the tonal value indicating the layer homogeneity (or uniformity). The homogeneity of printed areas was found to be positively correlated with printing speed so that the highest possible printing speed 10 m/min was chosen for the further trials of laser processing. Regarding the cell geometry, areas 2-2 and 2-3 were found to be suited best for the modified Clevios PEDOT PHC99. Contrarily, the printed structures of area 2-1 seem to be too large to produce a smooth film and area 2-4 transfers not enough material, producing only a thin film. This finding is clear as area 2-2 contains regular pyramidal cells whereas areas 2-1, 2-3 and 2-4 contain so-called X-treme engraved, irregularly shaped pyramidal cells with a relatively high transfer volume. Engraved areas 2-2 and 2-3 have similar mesh but different engraved structures, yet the materials flow appears to be quite similar. The film uniformity and thickness could be improved by optimizing the PEDOT:PSS formulation, the cylinder engraving and other process parameters. 3. Laser Structuring of Gravure-Printed PEDOT:PSS 3.1. Gaussian-Profile Laser Structuring by Short- Pulse Laser with Roll-to-Roll Web System fter the gravure printing, the PEDOT:PSS samples were sent to VTT, Lappeenranta, Finland for R2R laser processing with a picosecond Nd:YVO4 laser mounted on an in-housebuilt roll-to-roll (R2R) web system. Figure 2 shows the R2R laser processing unit developed at VTT. The system Figure 2. () VTT's R2R laser processing unit, () Layout of the 4x4 electrode structure chosen as test structure for laser processing. The laser processed pattern was a 4x4 electrode structure usually intended for optoelectronic applications (light emitting or photovoltaic devices) which is shown in Figure 2. Processing tests were started with stationary samples and when proper parameters were found the processing on the fly was started. In processing on the fly, the scan head adjusts trajectories so that the web movement is compensated and the geometry is done with same speed as with stationary samples. fter series of preliminary testing, the optimal laser processing parameters were found as the following: 355 nm: 150 mw Spot 355 nm: 20 μm 532 nm: 300 mw Spot 532 nm: 30 µm Pulse repetition rate: 100 khz Web speed: m/min The quality and function of the gravure-printed and laserprocessed thin-films of PEDOT:PSS was analysed by optical and mechanical profilometry. The inset of Figure 3 shows an exemplary profilogram taken for any top-left (*) and topright (*) areas of the samples recorded and evaluated. The profile along the laser processing line (blue curve) in Figure 3 shows a high periodic variation. With Fourier analysis, the distance of the peaks was proven to match well chosen spot size of the laser beam at this example, i. e nm. Together with the profilogram it shows also that the laser spots are separated. cross the laser processing direction (red curve), the line shows a edge height of roughly 1 µm. Furthermore, to prove the electrically insulating functionality, conductivity measurements between pad (*) and pixel (*) were performed using a standard multimeter. The electrical characterization showed that all connections were both conductive and isolated from the ground (i. e. the area 2

3 outside) and that a very good electrical shielding was obtained For the terms with (*), see Figure 2 (red arrows). of etched PEDOT:PSS for one pulse of 94 mj/cm² or for one pulse of 130 mj/cm² is clearly below 500 nm thickness. We suppose here that thickness of the conductive layer is not uniform over the surface. However, detritus is still remaining around corner-shaped forms. It can be concluded that a single pulse of a fluence of 130 mj/cm² etches 500 nm of PEDOT:PSS on PET. Multiple shots are burning the PET. C D Figure 3. Typical morphology profiles of of laser-processed gravure-printed PEDOT:PSS layers. The web speed for printing and laser processing was 10 m/min. The red and blue curves show the profiles _across_ (red) and _along_ (blue) the laser processing line, respectively. The colored image of the mechanical profilogram is shown in the inset Flat-Profile Laser Structuring of Gravure- Printed PEDOT:PSS using Excimer Laser In parallel to the results on Gaussian-profile short-pulse lasers, the samples were sent to CE-Liten, Saint Etienne, France to study the feasibility to etch gravure-printed PEDOT:PSS (in this case, Orgacon EL-350 was used) on PET (supplied by VTT and printed as described in Section 2) with an industrial excimer laser system at a wavelength of 248 nm. For flat-profile excimer laser structuring, a laser machine is projecting through an optical system a specific pattern localised on a <quartz/cr> mask onto a substrate situated on an x-y-translation stage at a specific fluence. The pulse durations were 30 ns at a typical frequency of ablation of 40 Hz. For laser fluences of 34, 54, 74, 94 and 130 mj/cm², single-pulse experiments were carried out. For the highest fluence achievable that date (130 mj/cm²), experiments with 2, 3, 5 and 10 pulses were additionally performed. The resulting patterns were analyzed by optical and mechanical profilometry as shown in Figure 4-D. Electrical measurements (carried out mounting the substrate on a wafer with a 3M sticker showed that electrical insulation is yielded at fluence of 94 mj/cm² or higher. Figure 4E shows the measured mean depth of the structures as a function of the laser fluence and repetition number. Note that the thickness E Figure 4. Optical micrographs of a 500 nm layer of gravure-printed PEDOT:PSS on PET etched by flat-profile ablation using a 248 nm excimer laser. Trench diameter is appr. 250 µm. () pulse energy 74 mj/cm², single shot; parts of PEDOT:PSS layer still remaining; (C) pulse energy 130 mj/cm², single shot. In the case of the higher pulse energy, a complete removal of the PEDOT:PSS layer is clearly seen. (E) Profilometry measurements of 500 nm of gravure-printed PEDOT:PSS on PET after one or many pulses with various fluences. Each color of the bar represents a specific measure. From 5 pulses at 130 mj/cm², the substrate is destroyed, too Comparison between Gaussian- and Flat-Profile Laser Structuring comparative view on the results on Gaussian- and flatprofile laser structuring reveals that the ultimate advantage of using mask projection flat-profile excimer laser systems is the possibility to obtain high-throughput production for the selective ablation of like patterns in sheet-to-sheet or roll-toroll manner. production speed of meters per minute (or khz) can be obtained. Edges and features have a better quality and definition both along processing direction and in the vertical direction of the profile at a improved production throughputs. On the contrary, the Gaussian-profile shortpulse lasers equipped with a roll-to-roll web system allow a selective, individual structuring. 3

4 4. Laser Structuring of ITO on PET using Gaussian-Profile Short-Pulse Laser Systems Striving for the optoelectronic applications of our process in the fabrication of anode structures, it turned out that in case of laser treatment the roles of both the substrates and the transparent conductive oxide coatings such as ITO have to be classified as well. For the experiments, standard ITO-coated PET (100 nm thick ITO layer, supplied by VTT) was used as a substrate Femtosecond laser systems For the femtosecond regime of pulse durations, VTT incorporated a Quantronics Integra C2.0 system based on Ti- Sapphire which can deliver 800 nm pulses of 130 fs duration at a maximum pulse energy of 2 mj situated in Lappeenranta, Finland. The laser can be tuned to operate from 1 Hz to 2 khz, however, the shortest pulses and best operational conditions were achieved at the operational frequency of 1 khz which was kept for the experiments. The laser was equipped with a f100 optics and the scanning speed was set to 10 mm/s. ssuming a Gaussian intensity profile, the investigated power ranges were 2 mw, 4 mw and 6 mw. The surface profiles of each sample was analyzed using Wyco optical profilometer. The 3D images of the optical profilograms are shown in Figure 5-C. electrically isolated. further optimization of the processing condition for this laser can be found by tuning other laser parameters and beam shape. Electrical shielding of the patterned structures was tested using a multimeter. It was found out that samples where processing power was 4 mw and 6 mw were electrically isolated, whereas sample 2 mw was in short circuit. dditionally, an ltechna tlsc laser was used, operating at 1030 nm with a pulse duration of 300 fs and a pulse repetition ratio adjustable between 1 and 350 khz. The samples were processed at 100 khz pulse rate and at a pulse density of 1000 pulses/mm (corresponding to 100 mm/s) and 5000 pulses/mm (corresponding to 500 mm/s). The results are shown in Figure 5D Picosecond laser systems For the picosecond regime, tests experiments were conducted on the Lumera laser system at a wavelength of 355 nm, varying the speed and power in order to eliminate the burrs on the top edge of the ablated material. few process windows showed very promising results with virtually no burrs formed after ablation. The evaluation is currently in progress. For picosecond processing, new results with acceptable edge quality have been achieved by optimizing the laser parameters in a new way. Same optimizing analogy worked also with a femtosecond laser. These results will be published elsewhere soon. C D Figure 5-C. Optical profilograms of abladed ITO on PET using a Quantronics Integra C2.0 fs laser system for different power ranges: () 2 mw, () 4 mw and (C) 6 mw. Figure 5D: Optical profilogram of abladed ITO on PET using an ltechna tlsc fs laser system. See text for details. s a conclusion, the best processing quality can be obtained when using the intermediate laser power of 4 mw. The value 6 mw is too high and the processing quality is quire rough. The value 2 mw is too low and patterned shapes are not 5. Summary In summary, this European research collaboration on the topic "Integration of laser ablation in roll-to-roll manufacturing" yielded comparative results on ablation of both gravure-printed PEDOT:PSS and sputtered ITO (100 nm) on flexible PET substrates. It has shown that structuring PEDOT:PSS and ITO with laser ablation is feasible with both flat-profile UV excimer laser machine and regular short-pulse laser systems combined with a roll-to-roill web system. Depending on the used drawings, production speeds of meters per minute could be obtained. 6. cknowledgements This project is a result of a research collaboration between VTT Technical Research Center of Finland, CE-Liten Commissariat à l Energie tomique et aux Energies lternatives (France), Cardiff University (U. K.), creo (Sweden) and Chemnitz University of Technology (Germany) within the EU-FP7-Network of Excellence PolyNet, funded by the European Community's Seventh Framework Programme (FP7/ ) under grant agreement n References [1] M. Clair et al. Laser Machining of Thin Solar Cells, Proceedings of the 2nd International Symposium on Laser Microfabrication (ISL), November 2008, Chemnitz, Germany. [2] J. Hast et al. Possibilities of Ultrafast Pico and Femtosecond Laser Systems in Roll-to-Roll Manufacturing Processes for Printed Intelligence, ibid. 4

5 5