Scholarly article on topic 'Solar Cells with Inkjet Printed Polymer Layers'

Solar Cells with Inkjet Printed Polymer Layers Academic research paper on "Nano-technology"

CC BY-NC-ND
0
0
Share paper
Academic journal
Energy Procedia
OECD Field of science
Keywords
{"Organic solar cells" / "inkjet printing" / P3HT / PEDOT:PSS / grid / "solvent system"}

Abstract of research paper on Nano-technology, author of scientific article — Alexander Lange, Michael Wegener, Bert Fischer, Silvia Janietz, Armin Wedel

Abstract Inkjet printing can be used to deposit large area, functional polymer layers for organic solar cells with limited material waste. In this work, solar cells were produced with inkjet printed polymer:small molecule active layers or inkjet printed hole transport layers. For device active layers, two classes of polymers were printed, semi-crystalline or amorphous, and solar cell performance was examined after different thermal treatments or after deposition from different solvent systems, respectively. Conventional devices were also prepared with inkjet printed hole transport layers as well as with printed grid structures located between the hole transport and active layers. Grid structures are useful because an increased contact area with the active layer could result in more charge extraction. Overall, inkjet printing is well suited to deposit the polymer layers of solar cells with limited performance reduction.

Academic research paper on topic "Solar Cells with Inkjet Printed Polymer Layers"

Available online at www.sciencedirect.com

SciVerse ScienceDirect

Energy Procedía 31 (2012) 150 - 158

E-MRS Spring Meeting 2011, Symposium S: Organic Photovoltaics: Science and Technology

Solar cells with inkjet printed polymer layers

Alexander Lange*a, Michael Wegenera, Bert Fischera, Silvia Janietza

and Armin Wedela

aFraunhofer Institute of Applied Polymer Research, Geiselbergstr. 69, Potsdam, 14476, Germany

Abstract

Inkjet printing can be used to deposit large area, functional polymer layers for organic solar cells with limited material waste. In this work, solar cells were produced with inkjet printed polymer:small molecule active layers or inkjet printed hole transport layers. For device active layers, two classes of polymers were printed, semi-crystalline or amorphous, and solar cell performance was examined after different thermal treatments or after deposition from different solvent systems, respectively. Conventional devices were also prepared with inkjet printed hole transport layers as well as with printed grid structures located between the hole transport and active layers. Grid structures are useful because an increased contact area with the active layer could result in more charge extraction. Overall, inkjet printing is well suited to deposit the polymer layers of solar cells with limited performance reduction.

© 2013 The Authors. Published by Elsevier Ltd.

Selection and peer-review under responsibility of the European Material Research Society (E-MRS) Keywords: Organic solar cells; inkjet printing; P3HT; PEDOT:PSS; grid; solvent system

1. Introduction

Organic solar cells (OSCs) based on polymers have generated considerable interest as an alternative source of energy because of their unique properties such as flexibility [1] and semi-transparency [2]. In addition, solution based preparation processes can be used to deposit the functional layers of these devices in conjunction with high speed and low cost industrial methods such as roll-to-roll printing as previously shown [3]. Many different technologies can be classified as 'solution processing' including, but not limited to, slot die coating, inkjet printing (IP) and spin coating (SC). While spin coating produces

'Corresponding author. Tel.: +49-(0)331-568-1913; fax:+49-(0)331-568-3910. E-mail address: alange@iap.fraunhofer.de

1876-6102 © 2013 The Authors. Published by Elsevier Ltd.

Selection and peer-review under responsibility of the European Material Research Society (E-MRS) doi: 10.1016/j.egypro.2012.11.177

homogeneous films, this method is not well suited for large scale production of OSCs because it is necessary to rotate the substrate at high speeds resulting in a material loss of ~80%. On the other hand, printing and coating can generate films with large areas rather quickly; however, the ink formulations must be well adjusted to generate homogeneous layers. Because of the potential to generate homogeneous films with large areas, printing and coating have received considerable attention in the literature [4].

Until recently, OSC research has been dominated by devices with bulk heterojunctions consisting of poly(3-hexyl thiophene) (P3HT) and phenyl-C6i butyric acid methyl ester (PC6jBM). The semi-crystalline polymer P3HT and PC6jBM phase separate on a nanometer scale and excited electron hole pairs generated upon illumination are split at interfaces between the two materials due to proper energy level alignment. Once free charges have been generated, networks of P3HT and PC6jBM transport the charges to the device electrodes. The P3HT:PC6jBM system is well investigated and power conversion efficiencies (nPCE) of between 3 and 5% have been reported [5]. Despite the progress which has been achieved with P3HT:PC6jBM, this system is not ideal for OSC applications because P3HT's absorbance maximum occurs at ~550 nm and incident light with wavelengths of greater than ~650 nm is not absorbed. Also, P3HT is a semi-crystalline material, which means that thermal treatments are necessary to induce crystallization and to increase performance [5]. For large-scale production of OSCs, thermal treatments are undesired because of an increased energy input during production. Because of these two significant factors regarding P3HT, polymers with novel chemical structures have been developed [6, 7]. These new polymers should not only function well as electron donors in organic solar cells but also absorb as much light as possible from the sun and thermal treatments should not be necessary to achieve high device performance.

In this study, inkjet printing was used to deposit either the active layers or the hole transport layers (HTL), which is commonly known as the passive layer, of organic solar cells. First, the well known P3HT:PC6jBM system was deposited with inkjet printing and compared to devices with spin coated layers from the same materials. Despite several articles which investigated inkjet printed P3HT:PC6jBM, little has been reported about the behavior of printed films upon annealing [9-11]. We found similar device performance after complete device annealing (post-annealing) for solar cells with printed or spin coated P3HT:PC61BM layers. Second, a novel, amorphous polymer consisting of ter-polyfluorene with a small amount of triphenylamine (0.05mol%) and dithienyl-2,3-bis-[4-octyloxy-phenyl]-quinoxaline ( 0.475mol%) units in the polymer backbone was combined with PC61BM and deposited with inkjet printing from chlorinated or unchlorinated solvent systems [12]. The chemical structure of this polymer which will be referred to as PFDTBTP is shown in Figure 1.

Figure 1. The chemical structure of PFDTBTP.

Because of the chemical structure of PFDTBTP, different solvents can be used when spin coating or printing including chlorobenzene, trichlorobenzene, anisole or tetralin. In addition, good device performance was achieved for devices with PFDTBTP:PC61BM without extensive annealing. And finally, a grid structure based on poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) was inkjet printed in a traditional solar cell structure before deposition of the active layer. Optimized grids offer advantages over traditional PEDOT:PSS films such as increased charge extraction due to a greater contact area between the HTL and active layer. Furthermore, complicated stamping procedures which have been explored in the literature are avoided when using inkjet to deposit PEDOT:PSS grids [13, 14].

2. Experimental

Commercially available indium tin oxide (ITO) substrates were rinsed with isopropanol and dried in N2. The hole transport layer consisted of PEDOT:PSS (AI4083, Heraeus Clevios) which was spin coated in air and annealed in N2 at 180°C for 15 minutes. The preparation of the device active layers with spin coating or inkjet printing is described below. The cathode structure was thermal evaporated at 10-6 mbar and the device active area was 0.16 or 0.595 cm2. For P3HT:PC6jBM devices, the cathode consisted of 120 nm aluminum. For PFDTBTP:PC6jBM devices, a LiF buffer layer (0.6 nm) was used with 120 nm of aluminum. Current density-voltage characteristics were measured in the dark and under illumination with a K.H. Steuernagel solar simulator (AM 1.5, 1000 W/m2). The intensity of the light source was adjusted with a silicon reference cell from Fraunhofer ISE which was calibrated with a KG3 filter. Adjusting the light source intensity with a KG3 calibrated silicon reference cell reduces the mismatch factor for P3HT:PC61BM devices to close to one [15]. Individual device measurements were not corrected for spectral mismatch. Ultra violet-visible (UV-Vis) spectroscopy measurements were done with a Perkin Elmer Lambda 950 spectrometer.

P3HT (Honeywell) was combined with PC6jBM (Solenne, 99.5% purity) in a weight to weight ratio of 1:1. Chlorobenzene was used for spin coating and a for inkjet printing; however, an inkjet solvent system based on chloro- and trichlorobenzene was found to improve the stability of the ink in the print head as previously described [8-11]. Despite the limited stability of printing from pure chlorobenzene, this solvent was used for inkjet because it allows for a more direct comparison to devices with spin coated active layers. P3HT:PC6jBM active layers were spin coated or printed in N2 and pre-annealed in N2 at 100°C for 10 min. Printed active layers were dried immediately after printing at 100°C until the film underwent the fluid/solid transition which is indicated by a color change. After initial measurements, completed devices were post-annealed at 150°C, 10 minutes.

PFDTBTP was synthesized via a Suzuki C-C-cross-coupling reaction which started from the diboronic ester of the 9,9-dioctylfluorene (0.5 mol%) and 4,4,-Dibromo-4"-methyl-triphenylamine (0.025 mol%) as one comonomer and 5,8-Di (4,4'dibromothienyl)- 2,3-bis-[4-octyloxy-phenyl]-quinoxaline (0.475 mol%) as the donor-acceptor molecule. Weight average and number average molecular weights of 24,100 and 11,500 g/mol were achieved, respectively. The polymer shows good solubility in environmentally friendly organic solvents like toluene, THF or anisole. For solar cell devices, PFDTBTP was combined with PC6jBM in a weight-to-weight ratio of 1:2. Spin coated active layers were prepared from chlorobenzene and inkjet printed PFDTBTP:PC61BM films were deposited from a mixture of chloro- and trichlorobenzene (90 wt%/10 wt%) or from anisole and tetralin (90 wt%/10 wt%). From this point on for devices with PFDTBTP, the solvent systems for layers spin coated from chlorobenzene and printed from chloro-/trichlorobenzene or printed from anisole/tetralin will be referred to as chlorinated or unchlorinated, respectively. Spin coating and inkjet printing were done in N2 and the films were heated at 80°C for 5 minutes in N2 before evaporation of the cathode. Printed films from chloro-/trichlorobenzene were dried at 100°C and layers from anisole/tetralin were dried at 70°C for 5 minutes followed by the 80°C temperature treatment.

3. Results and Discussion

3.1. Devices with inkjet printed P3HT:PC61BM layers

Organic solar cells with inkjet printed P3HT:PC6jBM active layers have been demonstrated in the literature where the importance of the solvent system was explored in addition to the degree of regioregularity of the polymer [8-11]. Optimal device performance was obtained when using a solvent system based on mesitylene and dichlorobenzene when compared to pure tetralin [8-10]. However, other studies have shown that P3HT forms aggregate structures in mesitylene and that chlorinated ink solvents based on chloro- and trichlorobenzene are well suited for inkjet applications [11]. For printing formulations, solvents with low boiling points are not well suited because they evaporate and the ink

clogs the print head nozzles. Additionally, a combination of a low and a high boiling point solvent has been shown to prevent the well document coffee stain effect [16]. The coffee stain effect occurs because of different rates evaporation in the bulk and at the edges of a printed object. For solar cells with spin coated active layers, chlorinated solvents such as chlorobenzene or dichlorobenzene are commonly used [5]. In order to more directly compare inkjet printing and spin coating, an ink formulation was prepared with only chlorobenzene. Despite the stability problems of this formulation in the print head, devices were prepared and measured under illumination as shown in Figure 2. When printing with pure chlorobenzene, the print head was wiped several times with a towel moistened with chloroform in order to unclog the print head nozzles.

Voltage (V)

Figure 2. Current density versus voltage for solar cells with inkjet printed (IP) or spin coated (SC) P3HT:PC61BM active layers after pre-annealing and after pre- and post-annealing.

As shown in Figure 2, greater npce differences were found for devices with inkjet printed or spin coated P3HT:PC61BM layers after pre-annealing. However, the performance differences become smaller upon post-annealing. Table 1 shows a summary of the open circuit voltage, Voc, short circuit current density, Jsc, fill factor, FF and npce for the two devices shown in Figure 2. It can be seen that after only pre-annealing, the Jsc and FF are considerably smaller for the device with a printed active layer. These differences could be related to the degree of organization within the two different films.

Table 1. Summary of the performance of devices with inkjet printed or spin coated P3HT:PC61BM active layers after pre-annealing and after pre- and post-annealing._

Preparation (Thickness) Annealing Voc (V) Jsc (mA/cm2) FF (%) nPCE (%) Rser es (ft*cm2) Rshunt (ft*cm2)

Inkjet printing (~120 nm) pre 0.310 4.7 30.3 0.4 35 207

Inkjet printing (~120 nm) pre+post 0.505 8.3 44.2 1.8 12 724

Spin coating (110 nm) pre 0.385 7.1 44.0 1.2 26 1240

Spin coating (110 nm) pre+post 0.535 7.7 47.6 2.0 11 1859

UV-Vis spectroscopy was used to examine inkjet printed P3HT:PC61BM films as shown in Figure 3. Films with different thicknesses were prepared and pre-annealed at 100°C for 10 minutes before measurement. The spectra in Figure 3 show a red shift for the thinnest printed film (26 nm) with respect to thicker films (129 or 303 nm). This trend indicates that despite the same thermal treatments used for printed P3HT:PC6jBM films, the degree of organization within the films is thickness dependent. It has been shown that intensities of the peaks and shoulders within the absorbance spectra for P3HT:PC61BM at ~550 and ~606 nm are related to the degree of organization within the films [17]. Furthermore, the thickness dependent absorbance properties could also explain the differences between the two devices from Figure 2. Despite relatively similar active layer thicknesses (120 nm for spin coated as compared to

110 nm for inkjet printed P3HT:PC6jBM) and the same pre-annealing temperature, the degree of organization within the two films is not the same. Upon further annealing, these differences become less apparent because the polymer within the printed film can further crystallize which results in better performance. The exact cause of the blue shift absorbance spectra for thicker, inkjet printed P3HT:PC61BM films is still under investigation.

Wavelength (nm)

Figure 3. Normalized absorbance versus wavelength for inkjet printed P3HT:PC61BM films with different thicknesses that were pre-annealed at 100°C for 10 min.

3.2. Devices with inkjet printed PFDTBTP:PC61BM layers

In addition to the classical system used for OSCs, polymers with novel structures can also be deposited with inkjet. Solar cells were prepared with inkjet printed PFDTBTP:PC61BM active layers from chlorinated (chloro-/trichlorobenzene) or from unchlorinated (anisole/tetralin) solvent systems and their performance was measured as shown in Figure 4. Additionally, spin coating was used to deposit films from chlorobenzene as a reference for the devices with printed PFDTBTP:PC61BM layers.

Figure 4. Current density versus voltage for solar cells with spin coated (SC) or inkjet printed (IP) PFDTBTP:PC61BM layers where the inkjet printed layers were prepared from chlorinated (chloro-/trichlorobenzene) or unchlorinated (anisole/tetralin) solvent systems.

Upon inkjet printing PFDTBTP:PC6jBM from a chlorinated system, a nPCE of 3.0% was measured as shown in Figure 4. A considerable change in the Jsc and the FF was found upon changing the solvent

system to anisole/tetralin. Furthermore, the series (Rseri£s) and shunt (Rshunt) resistance values were larger and smaller, respectively, for the device which was printed from the unchlorinated system. An increase in Rseries indicates a greater resistance within the active layer of the device in addition to larger contact resistances between the different layers. On the other hand, a decrease in Rseries correlates to leakage within the active layer of the device which could be caused by incomplete layer formation. As for devices with P3HT:PC61BM active layers, Rshunt is smaller for devices with printed active layers. This indicates mild leakage which could be due to inhomogeneous regions within the printed film because of print head nozzles which no longer properly functioned properly. Overall, the performance values in Table 2 indicate that PFDTBTP:PC61BM can be inkjet printed from different solvent systems with a small decrease in performance. Furthermore, performance values for devices with printed layers compared well to devices with spin coated PFDTBTP:PC61BM films where a nPCE of over 3.0% was found.

Table 2. Summary of the performance of devices with inkjet printed or spin coated PFDTBTP:PC61BM layers where the printed layers were prepared from chlorinated (chloro-/trichlorobenzene) or unchlorinated (anisole/tetralin) solvent systems. Chlorobenzene was used for spin coating.

Preparation (Thickness) Voc (V) Jsc (mA/cm2) FF (%) VPCE (%) Rser le, (Q'cm2) Rshunt (ft*cm2)

Inkjet Printing, chlorinated (~82 nm) 0.825 6.0 60.0 3.0 7 1240

Inkjet printing, unchlorinated (~105 nm) 0.880 4.3 52.4 2.0 17 1190

Spin coating, chlorinated (88 nm) 0.915 6.4 58.2 3.4 10 1860

In contrast to other photo-active polymer examined in this report (P3HT), PFDTBTP is an amorphous material which was confirmed by differential scanning calorimetry (DSC). Because of this, the impact of thermal treatments on the performance of devices with PFDTBTP:PC61BM was not explored in detail. For the devices considered in this report, only a mild thermal treatment at 80°C for 5 min was used to completely remove all of the high boiling point solvents from the active layer. However, in contrast to P3HT:PC61BM, complete device annealing (post-annealing) was not necessary to achieve comparable performance for devices with printed PFDTBTP:PC61BM layers from the chlorinated system and device with spin coated layers from chlorobenzene. For the unchlorinated solvent system, smaller nPCE values of 2.0% were found. The exact reason for the differences in performance for devices is still under investigation. The surface topographies of spin coated and inkjet printed PFDTBTP:PC61BM films are shown in Figure 5 where distinct surface structures were found for printed films. The root mean square roughness value for the spin coated film was 0.5 nm which compares to 1.1 and 1.4 nm for the two printed films. Phase contrast AFM showed the same surface structures as seen in the topography images. This indicates that the surface features in Figure 5 for the different films could correspond to different materials, either PFDTBTP or PC61BM. More specifically, much larger phases are present for the two printed films. Other studies reported larger surface features for a polymer similar to PFDTBTP upon spin coating at a slower speed [18]. When polymer:small molecules solutions are spun at slower speeds, the drying time increases. This process is similar to printing where a considerably longer drying time was observed with respect to spin coating. The printing process can take up to several minutes followed by the drying whereas spin coated films undergo the fluid/solid transition after ~10 seconds when spin coating from chlorobenzene. In the literature, better solar cell performance was found for devices with smaller surface features or phases [18]. This trend also corresponds to what was found here where the best performance was seen for the spin coated film where no distinct surface features were seen for a scan size of 2.5 ^m.

Figure 5. Surface topographies of spin coated (SC) and inkjet printed (IP) PFDTBTP:PC61BM films where chlorinated and unchlorinated inkjet solvent systems were used. Spin coated films were deposited from chlorobenzene.

3.3. Printed PEDOT:PSS grids

In order to enhance the performance of OSCs, different attempts have been explored in such as stamping PEDOT:PSS in order to increase the degree of charge extraction from the active layer [14]. Additionally, solar cells with imprinted active layers have been used to maximize the contact between the donor and acceptor phases where the size of the imprinted features was varied [13]. Grids could also be used in devices without ITO where a metallic grid followed by a high conductivity PEDOT:PSS layer could function as the device anode. Inkjet printing is a practical method for grid structures because the sizes of printed objects can be defined by the user. In this work, inkjet printing was used to deposit PEDOT:PSS grids on top of printed PEDOT:PSS films in order to increase the contact area between the passive and active layers. Grids with different dimensions (grid 1: 250 ^m wide and 500 ^m line spacing, grid 2: 125 ^m wide and 250 ^m line spacing) were prepared and performance was measured under

Voltage (V)

Figure 6. Current density versus voltage for solar cells with spin coated P3HT:PC61BM active layers and inkjet printed PEDOT:PSS grids. For the devices considered here, a conventional structure with ITO was used followed by a printed PEDOT:PSS film and then a printed PEDOT: PSS grid. For grid 1, the grid lines were 250 ^m wide and the spacing between each grid line was 500 ^m. For grid 2, the grid lines were 125 ^m wide and the spacing between each line was 250 ^m.

No significant improvement in device performance was found for devices with inkjet printed PEDOT:PSS grids for the two styles of grids examined in this work as indicated in Figure 6 and Table 3. The heights of the two grids were roughly 30 nm. However, an exact height is difficult to determine because the grid lines consisted of individual droplets from the print head nozzles which resulted rough surface profiles. The ratio between the height of the grid and the width of the grid lines was calculated to

be 0.00024 for the 125 ^m grid and 0.00012 for the 250 ^m grid for an assumed grid height of 30 nm. These values compare to ratios of 0.038 and 0.014 which were estimated from the literature where a considerable increase in Jsc was found for devices with stamped PEDOT:PSS films [14]. In order to increase this ratio, the height of the grids must be increased or the width of the grid lines must be decreased. Further reducing the width of the grid lines is difficult because this is limited by the diameter of the print head nozzles which directly controls the droplet size and printed feature size. Grids with widths that were smaller than 125 ^m were not successfully printed. Therefore, the heights of the grids must be increased in order to see a corresponding impact on performance. For example, a grid height of 200 nm would result in a height/width ratio of 0.0016 for the 125 ^m grid. This ratio is not as high as that reported in the literature; however, a larger effect during measurement should be seen with respect to the two grids examined in this report. Despite similar npce values for the three devices shown in Table 3, a clear trend is found for Rseries for the three devices. Specifically, Rseries is considerably smaller for the devices with a grid structure. This trend indicates that the fewer resistive losses occur for devices with grids which could be related to better charge extraction from the active layer.

Table 3. Summary of the performance of devices with spin coated P3HT:PC6iBM active layers and inkjet printed PEDOT:PSS layers with or without a grid structure._

HTL layer Voc (V) Jsc (mA/cm2) FF (%) VpCE (%) Rser, le, (n*cm2) Rshunt (ft*cm2)

IP PEDOT:PSS 0.525 7.2 47.5 1.8 63 850

IP PEDOT:PSS+Grid 1 0.515 7.4 43.8 1.7 22 567

IP PEDOT:PSS+Grid 2 0.520 7.6 51.1 2.0 11 700

4. Conclusions

Inkjet printing can be used to deposit various layers of OSCs including active layers based on P3HT or on novel polymers such as PFDTBTP combined with PC6iBM in addition to the passive layer consisting of PEDOT:PSS. As shown in this report, with proper processing optimizations, such as thermal treatments and the solvent system used for inkjet printing, comparable device performance can be obtained for solar cells with printed or spin coated active layers. Finally, structures which are not accessible with spin coating such as grids can be generated quite easily with inkjet printing. Overall, inkjet printing is a versatile tool and a viable alternative to spin coating because it could be used to generate large area films on a semi-industrial scale.

Acknowledgements

The authors would like to thank Eileen Katholing and Lica Pabel (Fraunhofer IAP, Germany) for the synthesis of PFDTBTP and Steffi KreiBl (Fraunhofer IAP, Germany) for solar cell preparation and measurement. Additionally, Prof. Dr. Dieter Neher (University of Potsdam, Germany) is acknowledged for fruitful discussions. Funding was provided by the German Federal Ministry of Education and Research (BMBF) project I3NI03I7.

References

[1] Bedeloglu A, Demir A, Bozkurt Y, Sariciftci N. A flexible textile structure based on polymeric photovoltaics using transparent cathode. Syn Met 2009;159:2043-2048.

[2] Schmidt H, Flügge H, Winkler T, Bülow T, Riedl T, Kowalsky W, Appl Phys Lett 2009;94: 243302.

[3] Krebs F, Gevorgyan S, Alstrup J. A roll-to-roll process to flexible polymer solar cells: model studies, manufacture and operational stability studies. J Mater Chem 2009;19:5442-5451.

[4] Krebs F. Fabrication and processing of polymer solar cells: A review of printing and coating techniques. Sol Energy Mater Sol Cells 2009;93: 394-412.

[5] Dennler G, Scharber M, Brabec C. Polymer-Fullerene Bulk-Heterojunction Solar Cells. AdvMater 2009;21:1323-1338.

[6] Soci C, Hwang IW, Moses D, Zhu Z, Waller D. Gaudiana R, Brabec C, Heeger A. Photoconductivity of a Low-Bandgap Conjugated Polymer. AdvFunctMater 2007;17:632-636.

[7] Janietz S, Krueger H, Schleiermacher HF, Würfel U, Niggermann M. Tailoring of low bandgap poylmer and its performance analysis in organic solar cells. Macromol Chem Phys 2009;210:1493-1503.

[8] Hoth C, Choulis SA, Schilinsky P, Brabec CJ. High Photovoltaic Performance of Inkjet Printed Polymer:Fullerene Blends. Adv Mater 2007;19:3973-3978.

[9] Hoth CN, Choulis SA, Schilinsky P, Brabec CJ. On the effect of poly(3-hexylthiophene) regioregularity on inkjet printed organic solar cells. J Mater Chem 2009;19:5398-5404.

[10] Hoth CN, Schilinsky P, Choulis SA, Brabec CJ. Printing Highly Efficient Organic Solar Cells. Nano Lett 2008;8:2806-2813.

[11] Lange A, Wegener M, Boeffel C, Fischer B, Wedel A, Neher D. A new approach to the solvent system for inkjet-printed P3HT:PCBM solar cells and its use in devices with printed passive and active layers. Sol Energy Mater Sol Cells 2010;94:1816-1821.

[12] Janietz S. Synthesis of poly[9,9-dioctylfluorenyl-2,7-dyil-co-(10,12-bis(thiophen-2-y)-3,6-dioxooctyl-11-thia-9,13-diaza-cyclopenta[b]triphenylene], unpublished results, Fraunhofer IAP.

[13] Kim MS, Kim JS, Cho JC, Shtein M, Guo LJ, Kim J. Flexible conjugated polymer photovoltaic cells with controlled heterojunctions fabricated using nanoimprint lithography. Appl Phys Lett 2007;90:123113.

[14] Emah J, Curry R, Silva S. Low cost patterning of poly(3,4-ethylenedioxythiophene) poly(styrenesulfonate) films to increase organic photovoltaic device efficiency. Appl Phys Lett 2008;3:103301.

[15] Shrotriya V, Li G, Yao Y, Moriarty T, Emery K, Yang Y. Accurate Measurement and Characterization of Organic Solar Cells.

Adv Funct Mater 2006;16:2016-2023.

[16] Tekin E. Thin Film Libraries of Functional Polymers and Materials Prepared by Inkjet Printing. Eindhoven University of Technology (15.10.2007), PhD thesis.

[17] Keawprajak A, Piyakulawat P, Klamchuen A, Iamraksa P, Asawapirom U. Influence of crystallizable solvent on the morphology and performance of P3HT:PCBM bulk-heterojunction solar cells. Sor Energy Mater Sol Cells 2010;94:531-536.

[18] Lindgren L, Zhang F, Andersson M, Barrau S, Hellström S, Mammo W, Perzon E, Inganäs O, Andersson M. Synthesis, Characterization and Devices of a Series of Alternating Copolymers for Solar Cells. Chem Mater 2009;21:3491-3502.