Graphene Oxide:SingleWalled Carbon ... - Semantic Scholar
www.advenergymat.de
www.MaterialsViews.com
Communication
Graphene Oxide:Single-Walled Carbon Nanotube-Based Interfacial Layer for All-Solution-Processed Multijunction Solar Cells in Both Regular and Inverted Geometries Vincent C. Tung, Jaemyung Kim, and Jiaxing Huang* Over the past decade, advances in material synthesis and interfacial engineering have produced more efficient polymer solar cells with improved absorption of the solar irradiation and charge carrier transport that have led to continuously increased power conversion efficiencies (PCE), recently surpassing 8%.[1] In parallel to the material design and engineering in the active layer, novel organic photovoltaic architectures have emerged. A representative example is tandem photovoltaics in which subcells are stacked along the optical path to increase optical absorption.[2–5] When the subcells are connected in series, the ideal open-circuit voltage (VOC) is equivalent to the sum of the VOCs of the individual subcells. PCE enhancement is most significant when high-efficiency polymers with complementary absorption profiles are stacked. Central to the construction of tandem devices is the interconnect layer that spatially separates but electrically connects the two subcells. Ideal candidates for interconnect layers require (a) high optical transparency over the entire solar spectrum; (b) sufficient vertical conductivity; (c) complete coverage to avoid intermixing of the two subcells; (d) low surface roughness to avoid interrupting the deposition of neighboring subcells; and (e) appropriate energetics to facilitate efficient charge recombination. For solution-processable interconnect materials, there are additional challenges such as orthogonal solubility, chemical and mechanical stability, and so on. Thus far, the conducting polymer poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) is predominantly used as a component in solution-processable interconnect layers.[3–7] However, low transmittance of PEDOT:PSS in the red to near infrared (NIR) region results in a loss of available solar flux, which could adversely affect the overall photocurrent. In addition, PEDOT:PSS could also react with some common interfacial materials such as metal oxides, thus degrading the quality of interfaces over time.[6] Recently, graphene oxide (GO) was found to be capable of replacing PEDOT:PSS as a hole transporting layer in polymer solar cells.[8,9] Since GO is an insulating soft carbon sheet[10,11] with thickness of around 1 nm and lateral dimension extends readily into tens of micrometers, an effective GO modifying layer needs to have as small thickness
Dr. V. C. Tung, J. Kim, Prof. J. Huang Department of Materials Science and Engineering Northwestern University 2220 Campus Drive, Evanston, IL 60208, USA E-mail: [email protected]
DOI: 10.1002/aenm.201100595
Adv. Energy Mater. 2012, 2, 299–303
as possible (e.g., 1–2 monolayers) while achieving full coverage. In earlier works, we revealed that GO is amphiphilic and can act as surfactant to disperse insoluble materials such as carbon nanotubes in water.[12–14] Later we discovered that adding a small amount of single-walled carbon nanotubes (SWCNTs) into GO does not affect its solution processability but improves the vertical resistance of the resulting thin film (see Figure S1, Supporting Information). This relaxes the spincoating conditions and allows the use of thicker (i.e., easier to make) 3–4 nm GO thin films. Thus, GO:SWCNTs can replace PEDOT:PSS as the anode-modifying layer to construct highperformance poly(3-hexylthiophene):[6,6]-phenyl C61 butyric acid methyl ester (P3HT:PCBM) bulk heterojunction solar cells with efficiency up to 4.1%.[15] Recently, we reported the first use of graphene-based sheets in the interconnect layer for polymer tandem solar cell.[7] In that study, two serially connected subcells were glued together using a sticky interconnect based on GO and PEDOT:PSS composites, in which the polymer became more rigid and conductive by interacting with GO. Here, we show that GO:SWCNTs thin films can also replace PEDOT:PSS as a foundation of the interconnect layer for constructing serially connected P3HT:PCBM tandem devices. Both regular and inverted tandem cells are fabricated, producing significantly increased VOCs reaching 84% and 80% of the sum of the subcell VOCs, respectively, suggesting successful serial connection of subcells. Although polymers with complementary absorptions are needed to achieve higher PCE of tandem solar cells, to prevent the complexity introduced by optimization of each subcell and to facilitate the evaluation of the interconnect layer, here tandem architectures with both subcells made of the prototypical P3HT:PCBM materials are constructed[7] because they have been developed to a rather mature stage with reliable output characteristics.[16] The key index of successful tandem structure is the value of VOC, which ideally should be the combination of the values of the individual subcells. Figure 1 illustrates the step-by-step procedure of constructing a regular tandem cell using GO:SWCNTs as the hole transporting layer for individual subcells as well as in the interconnect. GO:SWCNT stock solution (1:0.2 weight ratio with GO concentration of 0.15 wt%) was deposited on oxygen plasma treated indium tin oxide (ITO) substrates and then annealed at 120 °C to remove the residual solvent. Such prepared GO:SWCNTs thin film is only a few nanometers thick, but the conductance is already on the same order of magnitude as that of PEDOT:PSS thin film, which is usually tens of nanometers thick (Figure S2, Supporting Information). To fabricate the rear cell, we adapted a modified lamination process of active layers,[17] which is reported to eliminate most of the
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
299
www.advenergymat.de
www.MaterialsViews.com
Communication
distinguished in Figure 2b. In high magnification (Figure 2b, insets), SWCNTs can be seen protruding out from the hole transporting and the interfacial layers. Electrons extracted from the front cell are recombined with the holes driven from the rear cell at the interface comprised dense ZnO nanoparticles supported on GO:SWCNTs. As shown in the J–V characteristics (Figure 2c), the VOC of the tandem cell (red curve) is significantly higher than that of a single subcell (black curve). To compare the performances of single active layer (150 nm) cells with tandem cells, 20 best devices from each group were chosen to calculate the average values. The VOC, shortFigure 1. Schematic drawings illustrating the procedure of fabricating a regular tandem cell: circuit current density (JSC), and fill factor i) GO:SWCNT aqueous dispersion is spin coated onto an ITO glass pretreated with oxygen (FF) of typical single-layer cells prepared plasma. ii) The P3HT:PCBM active layer in the front subcell is transferred onto the ITO/ under the same conditions were measured to GO:SWCNT layer from a PDMS stamp. iii) The same type of P3HT:PCBM/PDMS stamp is −2 used to make the rear subcell by spin casting a layer of GO:SWCNTs and iv) a layer of ZnO be 0.56 ± 0.02 V, 9.39 ± 0.21 mA cm , and nanoparticles. v) The completed rear subcell is then directly stamped onto the prefabricated 0.59 ± 0.04, respectively, giving an average front subcell to construct the tandem structure. Finally, Ca/Al electrodes are thermally evapo- PCE of 3.41% ± 0.10%. On the other hand, rated on top to complete the device. the average VOC of tandem cells was found to be 0.94 ± 0.06 V, 84% of the sum of the processing challenges.[17,18] In addition, the stamping process subcells. Occasionally a VOC of as high as 1.05 V was obtained allows us to fine tune the morphology of the subcells independfor the tandem structure, but the data were not included in the ently before the active layers are brought into electrical contact statistics because of reproducibility issue. Although the JSC of in the final stage. P3HT:PCBM active layer (150 nm) was spin the multijunction cells (7.40 ± 0.42 mA cm−2) was a bit lower coated from chloroform solution (0.85 wt%) onto a polydimethbecause of limited absorption of the rear cell, the overall PCE ylsiloxane (PDMS) stamp at 1500 rpm. Next, the transferwas still increased to 4.10% ± 0.17%. The significant increase ring of the active layer was completed by directly pressing the in VOC suggests that the subcells were well separated but have PDMS stamp against the ITO/GO:SWCNTs substrate on a preheated hotplate at around 85 °C. In parallel, another P3HT:PCBM/ PDMS stamp was used to fabricate the rear cell with slightly smaller thickness (120 nm) to balance photocurrents of the two subcells. GO:SWCNTs and ZnO nanoparticle layers were sequentially spin coated on top of the front cell supported on a PDMS to serve as interconnect layers. ZnO nanoparticles were chosen as the n-type interlayer because of its excellent electron transporting property, matching energy level with PCBM and most importantly, the orthogonal solubility, which preserves the integrity of underlying polymer composites.[4,19] Finally, the two subcells were stamped together and annealed at 150 °C for 20 min before peeling off the PDMS stamp and the thermal deposition of Ca/Al top electrodes. The final annealing process is well known to facilitate preferential phase reorganization to enhance the efficiency of charge separation.[16] Figure 2 schematically illustrates the final Figure 2. a) A schematic drawing and b) a false-colored cross-sectional SEM image showing device architecture and the corresponding the layers in the tandem device. The two subcells (light purple) and the interconnect (light blue) can be clearly identified. The SEM images in the insets show individual SWCNTs protruding false colored, cross-sectional scanning elecfrom the interconnect (top) and anode-modification layers (bottom), respectively. Scale bars tron microscopy (SEM) image. Well-defined are 200 nm (main) and 50 nm (insets). c) J–V characteristics of a representative single subcell interlayer (light blue) as well as two indi- (black) and a tandem cell (red). The nearly doubled VOC of the tandem cell suggests successful vidual subcells (light purple) can be clearly connection of the two subcells. 300
wileyonlinelibrary.com
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Energy Mater. 2012, 2, 299–303
www.advenergymat.de
www.MaterialsViews.com
Adv. Energy Mater. 2012, 2, 299–303
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
Communication
after coating a thin layer of GO:SWCNTs on top of the rear cell to facilitate hole extraction. Finally, the inverted cell was annealed at 150 °C for 20 min before measurements. Figure 4 shows a representative crosssectional SEM image of the resulting inverted tandem cells. Insets are higher magnification SEM images providing a close-up view of the GO:SWCNTs (top) and the ZnO layer (bottom). Figure 4c provides J–V characteristics of the single subcell (black curve) and the inverted tandem cell (red curve), respectively. Under illumination, the inverted tandem Figure 3. Schematic drawings illustrating the fabrication procedure of inverted tandem cells: cell yielded a VOC of 0.83 ± 0.05 V compared with 0.53 ± 0.02 V of the subcell, which on i) ZnO nanoparticles are spin coated onto an ITO glass pretreated with oxygen plasma. ii) The P3HT:PCBM active layer in the front subcell is transferred onto the ITO/ZnO layer from a average is 80% of the sum of VOCs of the two PDMS stamp. iii) The same type of P3HT:PCBM/PDMS stamp is used to make the rear cell by subcells. Occasionally, a VOC of as high as spin casting a layer of ZnO nanoparticles and iv) a layer of GO:SWCNTs. v) The rear subcell is 0.92 V was obtained for the inverted tandem then directly stamped onto the prefabricated front cell to construct the tandem device. vi) An structure, but the data were not included additional GO:SWCNTs layer is spin coated atop the rear subcell, serving as a hole transporting in the statistics because of reproducibility layer. Finally, the device is completed with thermally evaporated Al electrodes. issue. This leads to a PCE of 3.50% ± 0.30%, which is significantly higher than that of the subcell (2.90% ± 0.21%) prepared under the same conditions, good electrical connection through the GO:SWCNTs-based even though same polymer is used. The device performance of interconnect layers. Without adding SWCNTs, tandem devices both regular and inverted tandem solar cells is summarized in made with GO and ZnO as interlayer have much lower perTable 1. formance (Figure S3, Supporting Information). The quality of the GO:SWCNTs thin films such as coverage The solution processability of GO:SWCNTs also makes it an and roughness is crucial to the successful device fabrication. ideal interfacial layer for inverted tandem solar cells. Inverted In our prior work, we showed that adding a small amount of polymer solar cells are attractive because they should have SWCNTs can relax the thickness requirement of the GO layer, better long-term stability by eliminating low work function metal or metal salts such as Ca, which could be gradually oxidized and be detrimental to the polymer active layers over time.[20,21] To construct the inverted tandem solar cells, transition metal oxides such as vanadium oxide (V2O5) and molybdenum oxide (MoO3) are typically required and need to be vacuum deposited to pair with n-type metal oxides, such as ZnO or titanium oxide.[21] Vacuum deposition step is laborious and interrupts the flow of solution-based device fabrication. By using water-processable GO:SWCNTs as the p-type component at the interface, here we demonstrate that inverted tandem cells can be made by all-solution processing route. Figure 3 shows the fabrication steps of the inverted cells. A thin film of ZnO (≈30 nm) nanoparticles was first spin coated onto an ITO substrate. Analogous to the fabrication of the regular tandem cell (Figure 1), two subcells consisting of bulk heterojunction P3HT:PCBM active layers were made separately. ZnO nanoparticles and GO:SWCNTs were sequentially spin coated on the rear cell Figure 4. a) Device configuration and b) corresponding cross-sectional SEM images of the supported by a PDMS stamp. The rear cell as fabricated inverted tandem solar cells. Cross-sectional SEM image is colorized for clarity. Well-defined subcells (light purple) and the interconnecting layers (light blue) can be clearly was then directly transferred atop the front identified. Insets are high magnification cross-sectional SEM images, providing a close-up view cell on a preheated hotplate at around 85 °C. of the GO:SWCNTs (top) and ZnO nanoparticles (bottom) layers. Scale bars are 200 nm (main) The inverted tandem solar cell is completed and 50 nm (insets). c) J–V characteristics of single subcell (black) and an inverted multijunction by thermal evaporation of the Al electrode, solar cell (red) measured under AM 1.5 G illumination. 301
www.advenergymat.de
www.MaterialsViews.com
Communication
Table 1. Summary of J–V characteristics in both regular and inverted geometry. VOC [V]
JSC [mA cm−2]
FF [%]
PCE [%]
Regular subcell
0.56 ± 0.02
9.39 ± 0.21
59 ± 4.0
3.41 ± 0.10
Regular tandem cell
0.94 ± 0.06
7.40 ± 0.42
58 ± 3.0
4.10 ± 0.17
Inverted subcell
0.53 ± 0.02
9.28 ± 0.50
58 ± 2.1
2.90 ± 0.21
Inverted tandem cell
0.83 ± 0.05
7.27 ± 0.03
55 ± 4.1
3.50 ± 0.30
making it easier to create near-full coverage thin films.[15] Here, we would also like to highlight that to ensure a smooth GO or GO:SWCNTs thin film, the GO product needs to be thoroughly purified. The preparation of GO usually involves the use of large quantities of strong acids such as concentrated H2SO4 in the reaction and HCl during rinsing, and metal ions such as K+ or Na+. Because of the ease of GO gelation during washing, such byproducts are hard to remove because of the slow rate of common purification methods such as dialysis, filtration, or centrifugation. Excess metal ions could cause aggregation of GO in solution,[22] or turn GO into a flammable material in solid states.[23] In this work, we found that if there is excess acid (pH ≤ 4), the resulting GO or GO:SWCNTs thin films made by spin coating have large numbers of wrinkles, especially at the boundaries between neighboring sheets (Figure S4a and c, Supporting Information). Such wrinkles or buckled structures are likely a result of the lateral compressive stresses experienced by the neighboring sheets as they are pushed together during high-speed spin coating and solvent evaporation. Formation of wrinkles under acidic condition has also been observed in Langmuir–Blodgett monolayers of GO sheets.[24] Height profiles scanned along the white line indicated in Figure S4a (Supporting Information) suggest that these wrinkles are typically thicker than 5 nm, often in the range of 5 to 7 nm, although the thickness of GO itself is only around 1 nm. When incorporated with SWCNTs, many rugged terrains with even larger thickness can be observed. Figure S4c (Supporting Information) shows one with height profile peaking at 11 nm. These spiky GO:SWCNTs spots are problematic for device fabrication as they can greatly disturb the deposition of active layers and deteriorate the quality of the contact between polymer and the interconnect layer. In near neutral conditions (pH ≥ 5) where excess acid is removed or neutralized, the carboxylic acid groups on GO sheets are charged and at their more hydrophilic state. Therefore, when they interact during spin coating, a lubricating layer of water could make them slide to overlap with each other.[24] This yields much smoother thin films without the spikes as can be seen at the line scan profiles of the thin films made under thoroughly washed condition (Figure S4b and d, Supporting Information). Acidic GO or GO:SWCNTs dispersion could also cause etching of the ZnO layer upon spin coating.[25] Therefore, it is crucial to have purified GO stock solution with excess metal ions and acids removed. This can be done by a two-step HCl–acetone washing procedure.[23,26] In summary, we demonstrate that water-processable GO:SWCNTs thin films can be used as an effective interfacial layer to construct tandem polymer solar cells in both regular and inverted geometry by all-solution processing routes. Although the same polymer is used in both subcells in this 302
wileyonlinelibrary.com
proof-of-concept study, the significantly increased VOCs and improved PCEs of both types of serially connected tandem cells show that GO:SWCNTs thin films serve as good mechanical separator and electrical interconnect of the two subcells. Because of their high transparency in the NIR region, GO:SWCNTs-based interconnect should work well with complementary tandem cells using a low band gap polymer. By using p- or n-doped SWCNTs[27] with GO, the charge recombination in the interconnect layer could be further improved. The stamping method for transferring of the active layer can greatly facilitate the stacking of materials with complementary absorption profiles while minimizing the problem of intermixing during deposition, which could lead to diverse types of tandem structures.
Experimental Section Materials: GO was synthesized by a modified Hummers’ method[28] and extensively purified by a two-step HCl-acetone washing method as reported elsewhere.[23,26] The purified GO, with typical lateral size of a few hundred nanometers to a few micrometers, was dried in ambient conditions and then redispersed in deionized water to create dispersions with various concentrations. If necessary, ammonium hydroxide can be added into GO stock solutions to neutralize excess HCl used in washing. SWCNTs were purchased from Carbon Solutions, Inc. (P3-SWNT, functionalized with 1.0–3.0 atomic% carboxylic acid) and dispersed in water (1 mg mL−1) by ultrasonication (Misonix, S-4000) for 2 h, followed by centrifugation at 110 000 rpm for 1 h (Eppendorf, 5417C) to decant the supernatant for later use. PEDOT:PSS (Clevios 4083, Heracus), chloroform, chlorobenzene (anhydrous, Sigma–Aldrich), P3HT (4002EE, Rieke Metals, Inc.), and PCBM (>99.5%, Sigma–Aldrich) were used as received. Vertical Current–Voltage Measurement: GO and GO:SWCNTs thin films were spin coated on ITO for these experiments and were subjected to thermal annealing at 150 °C for 20 min prior to measurements. To prevent shorting, ITO substrates were precoated with a layer of ZnO nanoparticles (30 nm). Two terminal measurements were performed using a probe station connected to a Keithley 2400 source meter. Lateral Current–Voltage Measurement: PEDOT:PSS and GO:SWCNTs thin films were spin coated on Si/SiO2 (300 nm) pretreated with oxygen plasma. Because the polymer photovoltaic devices are thermal annealed at 150 °C for 20 min before completion, the GO:SWCNTs and PEDOT:PSS thin films were also annealed at 150 °C for 20 min in ambient conditions prior to thermal deposition of Al electrodes through a shadow mask. Current-voltage measurements were carried out using a probe station connected to a Keithley 2400 source meter. Device Fabrication: (a) Regular tandem: the GO:SWCNTs dispersions were created by directly mixing the two constituents in a combination of deionized water and ethanol (7:3, v/v) by ultra-sonication. The resulting dispersions were then centrifuged to further remove the unwanted metallic catalysts and carbonaceous materials from the SWCNTs. The black stock solution was stable for months without any visible precipitation and can be readily deposited onto various substrates for characterization. GO:SWCNTs dispersions were spin coated at 2000 rpm for 40 s on ITO anodes (20 Ω sq−1) pretreated with oxygen plasma, followed by thermal annealing at 120 °C for 20 min. The modified substrates were then transferred into a nitrogen-filled glove box. A 0.85 wt% P3HT:PCBM (1:1, w/w) in chloroform was spin coated onto a PDMS stamp at 1500 rpm for 10 s. Next, P3HT:PCBM active layer was directly pressed against ITO/GO:SWCNTs atop the hot plate preheated at 85 °C to complete the front cell. For rear cells, the same type of P3HT:PCBM/PDMS stamp was prepared, followed by sequential spin casting of GO:SWCNTs and ZnO (30 mg mL−1 in acetone) at 2000 rpm. The completed rear cell was brought into electrical contact with the prefabricated front cell on a preheated hot plate at around 85 °C. Postannealing at 150 °C for 20 min was found to further improve the morphology of active layers at
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Energy Mater. 2012, 2, 299–303
www.advenergymat.de
www.MaterialsViews.com
Acknowledgements The work was supported by a booster award from the Initiative for Sustainability and Energy at Northwestern (ISEN) and National Science Foundation (CMMI 1130407). We thank the Sony Corporation and the Northrop Grumman Corporation for additional support. JH is an Alfred P. Sloan Research Fellow, and JK gratefully acknowledges support from the Ryan Fellowship and the Northwestern University International Institute for Nanotechnology. We thank Prof. S. I. Stupp for the use of his device fabrication and measurement system, NUANCE for the use of their microscopy facilities, Dr. A. R. Carretero for helping us on preparation of ZnO nanoparticles, and Dr. S. Sista, Prof. C.-W. Chu, A. Koltonow, A. T. Tan, and L. J. Cote for helpful discussions. Received: October 6, 2011 Revised: November 3, 2011 Published online: January 19, 2012
Adv. Energy Mater. 2012, 2, 299–303
[1] H. Y. Chen, J. H. Hou, S. Q. Zhang, Y. Y. Liang, G. W. Yang, Y. Yang, L. P. Yu, Y. Wu, G. Li, Nat. Photon. 2009, 3, 649. [2] T. Ameri, G. Dennler, C. Lungenschmied, C. J. Brabec, Energy Environ. Sci. 2009, 2, 347. [3] K. Lee, J. Y. Kim, N. E. Coates, D. Moses, T. Q. Nguyen, M. Dante, A. J. Heeger, Science 2007, 317, 222. [4] J. Gilot, M. M. Wienk, R. A. J. Janssen, Adv. Mater. 2010, 22, E67. [5] J. Yang, R. Zhu, Z. Hong, Y. J. He, A. Kumar, Y. Li, Adv. Mater. 2011, 23, 3465. [6] S. Sista, Z. R. Hong, L. M. Chen, Y. Yang, Energy Environ. Sci. 2011, 4, 1606. [7] V. C. Tung, J. Kim, L. J. Cote, J. X. Huang, J. Am. Chem. Soc. 2011, 133, 9262. [8] S. S. Li, K. H. Tu, C. C. Lin, M. Chhowalla, C. W. Chen, ACS Nano 2010, 4, 3169. [9] Y. Gao, H. L. Yip, S. K. Hau, K. M. O’Malley, N. C. Cho, H. Z. Chen, A. K. Y. Jen, Appl. Phys. Lett. 2010, 97, 203306. [10] L. J. Cote, F. Kim, J. Huang, J. Am. Chem. Soc. 2009, 131, 1043. [11] L. J. Cote, J. Kim, V. C. Tung, J. Luo, F. Kim, J. Huang, Pure Appl. Chem. 2011, 83, 95. [12] J. Kim, L. J. Cote, F. Kim, W. Yuan, K. R. Shull, J. X. Huang, J. Am. Chem. Soc. 2010, 132, 8180. [13] F. Kim, L. J. Cote, J. X. Huang, Adv. Mater. 2010, 22, 1954. [14] V. C. Tung, J. H. Huang, I. Tevis, F. Kim, J. Kim, C. W. Chu, S. I. Stupp, J. X. Huang, J. Am. Chem. Soc. 2011, 133, 4940. [15] J. Kim, V. C. Tung, J. Huang, Adv. Energy Mater. 2011, 1, 1052. [16] G. Li, V. Shrotriya, J. S. Huang, Y. Yao, T. Moriarty, K. Emery, Y. Yang, Nat. Mater. 2005, 4, 864. [17] J. H. Huang, Z. Y. Ho, T. H. Kuo, D. Kekuda, K. C. Ho, C. W. Chu, J. Mater. Chem. 2009, 19, 4077. [18] H. Wang, E. D. Gomez, J. Kim, Z. L. Guan, C. Jaye, D. A. Fischer, A. Kahn, Y. L. Loo, Chem. Mater. 2011, 23, 2020. [19] R. A. J. Janssen, W. J. E. Beek, M. M. Wienk, M. Kemerink, X. N. Yang, J. Phys. Chem. B 2005, 109, 9505. [20] H. H. Liao, L. M. Chen, Z. Xu, G. Li, Y. Yang, Appl. Phys. Lett. 2008, 92, 173303. [21] C. H. Chou, W. L. Kwan, Z. R. Hong, L. M. Chen, Y. Yang, Adv. Mater. 2011, 23, 1282. [22] S. O. Kim, J. E. Kim, T. H. Han, S. H. Lee, J. Y. Kim, C. W. Ahn, J. M. Yun, Angew. Chem. Int. Ed. 2011, 50, 3043. [23] F. Kim, J. Y. Luo, R. Cruz-Silva, L. J. Cote, K. Sohn, J. X. Huang, Adv. Funct. Mater. 2010, 20, 2867. [24] L. J. Cote, J. Kim, Z. Zhang, C. Sun, J. X. Huang, Soft Matter 2010, 6, 6096. [25] J. Gilot, M. M. Wienk, R. A. J. Janssen, Appl. Phys. Lett. 2007, 90, 143512. [26] A. V. Nikolaev, A. S. Nazarov, V. V. Lisitsa, Russ. J. Inorg. Chem. 1974, 19, 3396. [27] J. M. Lee, J. S. Park, S. H. Lee, H. Kim, S. Yoo, S. O. Kim, Adv. Mater. 2011, 23, 629. [28] W. S. Hummers, R. E. Offeman, J. Am. Chem. Soc. 1958, 80, 1339.
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
Communication
nanoscale as well as electrical contact of interconnecting layers. Finally, Ca (30 nm)/Al (100 nm) were thermally evaporated at
www.MaterialsViews.com
Communication
Graphene Oxide:Single-Walled Carbon Nanotube-Based Interfacial Layer for All-Solution-Processed Multijunction Solar Cells in Both Regular and Inverted Geometries Vincent C. Tung, Jaemyung Kim, and Jiaxing Huang* Over the past decade, advances in material synthesis and interfacial engineering have produced more efficient polymer solar cells with improved absorption of the solar irradiation and charge carrier transport that have led to continuously increased power conversion efficiencies (PCE), recently surpassing 8%.[1] In parallel to the material design and engineering in the active layer, novel organic photovoltaic architectures have emerged. A representative example is tandem photovoltaics in which subcells are stacked along the optical path to increase optical absorption.[2–5] When the subcells are connected in series, the ideal open-circuit voltage (VOC) is equivalent to the sum of the VOCs of the individual subcells. PCE enhancement is most significant when high-efficiency polymers with complementary absorption profiles are stacked. Central to the construction of tandem devices is the interconnect layer that spatially separates but electrically connects the two subcells. Ideal candidates for interconnect layers require (a) high optical transparency over the entire solar spectrum; (b) sufficient vertical conductivity; (c) complete coverage to avoid intermixing of the two subcells; (d) low surface roughness to avoid interrupting the deposition of neighboring subcells; and (e) appropriate energetics to facilitate efficient charge recombination. For solution-processable interconnect materials, there are additional challenges such as orthogonal solubility, chemical and mechanical stability, and so on. Thus far, the conducting polymer poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) is predominantly used as a component in solution-processable interconnect layers.[3–7] However, low transmittance of PEDOT:PSS in the red to near infrared (NIR) region results in a loss of available solar flux, which could adversely affect the overall photocurrent. In addition, PEDOT:PSS could also react with some common interfacial materials such as metal oxides, thus degrading the quality of interfaces over time.[6] Recently, graphene oxide (GO) was found to be capable of replacing PEDOT:PSS as a hole transporting layer in polymer solar cells.[8,9] Since GO is an insulating soft carbon sheet[10,11] with thickness of around 1 nm and lateral dimension extends readily into tens of micrometers, an effective GO modifying layer needs to have as small thickness
Dr. V. C. Tung, J. Kim, Prof. J. Huang Department of Materials Science and Engineering Northwestern University 2220 Campus Drive, Evanston, IL 60208, USA E-mail: [email protected]
DOI: 10.1002/aenm.201100595
Adv. Energy Mater. 2012, 2, 299–303
as possible (e.g., 1–2 monolayers) while achieving full coverage. In earlier works, we revealed that GO is amphiphilic and can act as surfactant to disperse insoluble materials such as carbon nanotubes in water.[12–14] Later we discovered that adding a small amount of single-walled carbon nanotubes (SWCNTs) into GO does not affect its solution processability but improves the vertical resistance of the resulting thin film (see Figure S1, Supporting Information). This relaxes the spincoating conditions and allows the use of thicker (i.e., easier to make) 3–4 nm GO thin films. Thus, GO:SWCNTs can replace PEDOT:PSS as the anode-modifying layer to construct highperformance poly(3-hexylthiophene):[6,6]-phenyl C61 butyric acid methyl ester (P3HT:PCBM) bulk heterojunction solar cells with efficiency up to 4.1%.[15] Recently, we reported the first use of graphene-based sheets in the interconnect layer for polymer tandem solar cell.[7] In that study, two serially connected subcells were glued together using a sticky interconnect based on GO and PEDOT:PSS composites, in which the polymer became more rigid and conductive by interacting with GO. Here, we show that GO:SWCNTs thin films can also replace PEDOT:PSS as a foundation of the interconnect layer for constructing serially connected P3HT:PCBM tandem devices. Both regular and inverted tandem cells are fabricated, producing significantly increased VOCs reaching 84% and 80% of the sum of the subcell VOCs, respectively, suggesting successful serial connection of subcells. Although polymers with complementary absorptions are needed to achieve higher PCE of tandem solar cells, to prevent the complexity introduced by optimization of each subcell and to facilitate the evaluation of the interconnect layer, here tandem architectures with both subcells made of the prototypical P3HT:PCBM materials are constructed[7] because they have been developed to a rather mature stage with reliable output characteristics.[16] The key index of successful tandem structure is the value of VOC, which ideally should be the combination of the values of the individual subcells. Figure 1 illustrates the step-by-step procedure of constructing a regular tandem cell using GO:SWCNTs as the hole transporting layer for individual subcells as well as in the interconnect. GO:SWCNT stock solution (1:0.2 weight ratio with GO concentration of 0.15 wt%) was deposited on oxygen plasma treated indium tin oxide (ITO) substrates and then annealed at 120 °C to remove the residual solvent. Such prepared GO:SWCNTs thin film is only a few nanometers thick, but the conductance is already on the same order of magnitude as that of PEDOT:PSS thin film, which is usually tens of nanometers thick (Figure S2, Supporting Information). To fabricate the rear cell, we adapted a modified lamination process of active layers,[17] which is reported to eliminate most of the
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
299
www.advenergymat.de
www.MaterialsViews.com
Communication
distinguished in Figure 2b. In high magnification (Figure 2b, insets), SWCNTs can be seen protruding out from the hole transporting and the interfacial layers. Electrons extracted from the front cell are recombined with the holes driven from the rear cell at the interface comprised dense ZnO nanoparticles supported on GO:SWCNTs. As shown in the J–V characteristics (Figure 2c), the VOC of the tandem cell (red curve) is significantly higher than that of a single subcell (black curve). To compare the performances of single active layer (150 nm) cells with tandem cells, 20 best devices from each group were chosen to calculate the average values. The VOC, shortFigure 1. Schematic drawings illustrating the procedure of fabricating a regular tandem cell: circuit current density (JSC), and fill factor i) GO:SWCNT aqueous dispersion is spin coated onto an ITO glass pretreated with oxygen (FF) of typical single-layer cells prepared plasma. ii) The P3HT:PCBM active layer in the front subcell is transferred onto the ITO/ under the same conditions were measured to GO:SWCNT layer from a PDMS stamp. iii) The same type of P3HT:PCBM/PDMS stamp is −2 used to make the rear subcell by spin casting a layer of GO:SWCNTs and iv) a layer of ZnO be 0.56 ± 0.02 V, 9.39 ± 0.21 mA cm , and nanoparticles. v) The completed rear subcell is then directly stamped onto the prefabricated 0.59 ± 0.04, respectively, giving an average front subcell to construct the tandem structure. Finally, Ca/Al electrodes are thermally evapo- PCE of 3.41% ± 0.10%. On the other hand, rated on top to complete the device. the average VOC of tandem cells was found to be 0.94 ± 0.06 V, 84% of the sum of the processing challenges.[17,18] In addition, the stamping process subcells. Occasionally a VOC of as high as 1.05 V was obtained allows us to fine tune the morphology of the subcells independfor the tandem structure, but the data were not included in the ently before the active layers are brought into electrical contact statistics because of reproducibility issue. Although the JSC of in the final stage. P3HT:PCBM active layer (150 nm) was spin the multijunction cells (7.40 ± 0.42 mA cm−2) was a bit lower coated from chloroform solution (0.85 wt%) onto a polydimethbecause of limited absorption of the rear cell, the overall PCE ylsiloxane (PDMS) stamp at 1500 rpm. Next, the transferwas still increased to 4.10% ± 0.17%. The significant increase ring of the active layer was completed by directly pressing the in VOC suggests that the subcells were well separated but have PDMS stamp against the ITO/GO:SWCNTs substrate on a preheated hotplate at around 85 °C. In parallel, another P3HT:PCBM/ PDMS stamp was used to fabricate the rear cell with slightly smaller thickness (120 nm) to balance photocurrents of the two subcells. GO:SWCNTs and ZnO nanoparticle layers were sequentially spin coated on top of the front cell supported on a PDMS to serve as interconnect layers. ZnO nanoparticles were chosen as the n-type interlayer because of its excellent electron transporting property, matching energy level with PCBM and most importantly, the orthogonal solubility, which preserves the integrity of underlying polymer composites.[4,19] Finally, the two subcells were stamped together and annealed at 150 °C for 20 min before peeling off the PDMS stamp and the thermal deposition of Ca/Al top electrodes. The final annealing process is well known to facilitate preferential phase reorganization to enhance the efficiency of charge separation.[16] Figure 2 schematically illustrates the final Figure 2. a) A schematic drawing and b) a false-colored cross-sectional SEM image showing device architecture and the corresponding the layers in the tandem device. The two subcells (light purple) and the interconnect (light blue) can be clearly identified. The SEM images in the insets show individual SWCNTs protruding false colored, cross-sectional scanning elecfrom the interconnect (top) and anode-modification layers (bottom), respectively. Scale bars tron microscopy (SEM) image. Well-defined are 200 nm (main) and 50 nm (insets). c) J–V characteristics of a representative single subcell interlayer (light blue) as well as two indi- (black) and a tandem cell (red). The nearly doubled VOC of the tandem cell suggests successful vidual subcells (light purple) can be clearly connection of the two subcells. 300
wileyonlinelibrary.com
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Energy Mater. 2012, 2, 299–303
www.advenergymat.de
www.MaterialsViews.com
Adv. Energy Mater. 2012, 2, 299–303
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
Communication
after coating a thin layer of GO:SWCNTs on top of the rear cell to facilitate hole extraction. Finally, the inverted cell was annealed at 150 °C for 20 min before measurements. Figure 4 shows a representative crosssectional SEM image of the resulting inverted tandem cells. Insets are higher magnification SEM images providing a close-up view of the GO:SWCNTs (top) and the ZnO layer (bottom). Figure 4c provides J–V characteristics of the single subcell (black curve) and the inverted tandem cell (red curve), respectively. Under illumination, the inverted tandem Figure 3. Schematic drawings illustrating the fabrication procedure of inverted tandem cells: cell yielded a VOC of 0.83 ± 0.05 V compared with 0.53 ± 0.02 V of the subcell, which on i) ZnO nanoparticles are spin coated onto an ITO glass pretreated with oxygen plasma. ii) The P3HT:PCBM active layer in the front subcell is transferred onto the ITO/ZnO layer from a average is 80% of the sum of VOCs of the two PDMS stamp. iii) The same type of P3HT:PCBM/PDMS stamp is used to make the rear cell by subcells. Occasionally, a VOC of as high as spin casting a layer of ZnO nanoparticles and iv) a layer of GO:SWCNTs. v) The rear subcell is 0.92 V was obtained for the inverted tandem then directly stamped onto the prefabricated front cell to construct the tandem device. vi) An structure, but the data were not included additional GO:SWCNTs layer is spin coated atop the rear subcell, serving as a hole transporting in the statistics because of reproducibility layer. Finally, the device is completed with thermally evaporated Al electrodes. issue. This leads to a PCE of 3.50% ± 0.30%, which is significantly higher than that of the subcell (2.90% ± 0.21%) prepared under the same conditions, good electrical connection through the GO:SWCNTs-based even though same polymer is used. The device performance of interconnect layers. Without adding SWCNTs, tandem devices both regular and inverted tandem solar cells is summarized in made with GO and ZnO as interlayer have much lower perTable 1. formance (Figure S3, Supporting Information). The quality of the GO:SWCNTs thin films such as coverage The solution processability of GO:SWCNTs also makes it an and roughness is crucial to the successful device fabrication. ideal interfacial layer for inverted tandem solar cells. Inverted In our prior work, we showed that adding a small amount of polymer solar cells are attractive because they should have SWCNTs can relax the thickness requirement of the GO layer, better long-term stability by eliminating low work function metal or metal salts such as Ca, which could be gradually oxidized and be detrimental to the polymer active layers over time.[20,21] To construct the inverted tandem solar cells, transition metal oxides such as vanadium oxide (V2O5) and molybdenum oxide (MoO3) are typically required and need to be vacuum deposited to pair with n-type metal oxides, such as ZnO or titanium oxide.[21] Vacuum deposition step is laborious and interrupts the flow of solution-based device fabrication. By using water-processable GO:SWCNTs as the p-type component at the interface, here we demonstrate that inverted tandem cells can be made by all-solution processing route. Figure 3 shows the fabrication steps of the inverted cells. A thin film of ZnO (≈30 nm) nanoparticles was first spin coated onto an ITO substrate. Analogous to the fabrication of the regular tandem cell (Figure 1), two subcells consisting of bulk heterojunction P3HT:PCBM active layers were made separately. ZnO nanoparticles and GO:SWCNTs were sequentially spin coated on the rear cell Figure 4. a) Device configuration and b) corresponding cross-sectional SEM images of the supported by a PDMS stamp. The rear cell as fabricated inverted tandem solar cells. Cross-sectional SEM image is colorized for clarity. Well-defined subcells (light purple) and the interconnecting layers (light blue) can be clearly was then directly transferred atop the front identified. Insets are high magnification cross-sectional SEM images, providing a close-up view cell on a preheated hotplate at around 85 °C. of the GO:SWCNTs (top) and ZnO nanoparticles (bottom) layers. Scale bars are 200 nm (main) The inverted tandem solar cell is completed and 50 nm (insets). c) J–V characteristics of single subcell (black) and an inverted multijunction by thermal evaporation of the Al electrode, solar cell (red) measured under AM 1.5 G illumination. 301
www.advenergymat.de
www.MaterialsViews.com
Communication
Table 1. Summary of J–V characteristics in both regular and inverted geometry. VOC [V]
JSC [mA cm−2]
FF [%]
PCE [%]
Regular subcell
0.56 ± 0.02
9.39 ± 0.21
59 ± 4.0
3.41 ± 0.10
Regular tandem cell
0.94 ± 0.06
7.40 ± 0.42
58 ± 3.0
4.10 ± 0.17
Inverted subcell
0.53 ± 0.02
9.28 ± 0.50
58 ± 2.1
2.90 ± 0.21
Inverted tandem cell
0.83 ± 0.05
7.27 ± 0.03
55 ± 4.1
3.50 ± 0.30
making it easier to create near-full coverage thin films.[15] Here, we would also like to highlight that to ensure a smooth GO or GO:SWCNTs thin film, the GO product needs to be thoroughly purified. The preparation of GO usually involves the use of large quantities of strong acids such as concentrated H2SO4 in the reaction and HCl during rinsing, and metal ions such as K+ or Na+. Because of the ease of GO gelation during washing, such byproducts are hard to remove because of the slow rate of common purification methods such as dialysis, filtration, or centrifugation. Excess metal ions could cause aggregation of GO in solution,[22] or turn GO into a flammable material in solid states.[23] In this work, we found that if there is excess acid (pH ≤ 4), the resulting GO or GO:SWCNTs thin films made by spin coating have large numbers of wrinkles, especially at the boundaries between neighboring sheets (Figure S4a and c, Supporting Information). Such wrinkles or buckled structures are likely a result of the lateral compressive stresses experienced by the neighboring sheets as they are pushed together during high-speed spin coating and solvent evaporation. Formation of wrinkles under acidic condition has also been observed in Langmuir–Blodgett monolayers of GO sheets.[24] Height profiles scanned along the white line indicated in Figure S4a (Supporting Information) suggest that these wrinkles are typically thicker than 5 nm, often in the range of 5 to 7 nm, although the thickness of GO itself is only around 1 nm. When incorporated with SWCNTs, many rugged terrains with even larger thickness can be observed. Figure S4c (Supporting Information) shows one with height profile peaking at 11 nm. These spiky GO:SWCNTs spots are problematic for device fabrication as they can greatly disturb the deposition of active layers and deteriorate the quality of the contact between polymer and the interconnect layer. In near neutral conditions (pH ≥ 5) where excess acid is removed or neutralized, the carboxylic acid groups on GO sheets are charged and at their more hydrophilic state. Therefore, when they interact during spin coating, a lubricating layer of water could make them slide to overlap with each other.[24] This yields much smoother thin films without the spikes as can be seen at the line scan profiles of the thin films made under thoroughly washed condition (Figure S4b and d, Supporting Information). Acidic GO or GO:SWCNTs dispersion could also cause etching of the ZnO layer upon spin coating.[25] Therefore, it is crucial to have purified GO stock solution with excess metal ions and acids removed. This can be done by a two-step HCl–acetone washing procedure.[23,26] In summary, we demonstrate that water-processable GO:SWCNTs thin films can be used as an effective interfacial layer to construct tandem polymer solar cells in both regular and inverted geometry by all-solution processing routes. Although the same polymer is used in both subcells in this 302
wileyonlinelibrary.com
proof-of-concept study, the significantly increased VOCs and improved PCEs of both types of serially connected tandem cells show that GO:SWCNTs thin films serve as good mechanical separator and electrical interconnect of the two subcells. Because of their high transparency in the NIR region, GO:SWCNTs-based interconnect should work well with complementary tandem cells using a low band gap polymer. By using p- or n-doped SWCNTs[27] with GO, the charge recombination in the interconnect layer could be further improved. The stamping method for transferring of the active layer can greatly facilitate the stacking of materials with complementary absorption profiles while minimizing the problem of intermixing during deposition, which could lead to diverse types of tandem structures.
Experimental Section Materials: GO was synthesized by a modified Hummers’ method[28] and extensively purified by a two-step HCl-acetone washing method as reported elsewhere.[23,26] The purified GO, with typical lateral size of a few hundred nanometers to a few micrometers, was dried in ambient conditions and then redispersed in deionized water to create dispersions with various concentrations. If necessary, ammonium hydroxide can be added into GO stock solutions to neutralize excess HCl used in washing. SWCNTs were purchased from Carbon Solutions, Inc. (P3-SWNT, functionalized with 1.0–3.0 atomic% carboxylic acid) and dispersed in water (1 mg mL−1) by ultrasonication (Misonix, S-4000) for 2 h, followed by centrifugation at 110 000 rpm for 1 h (Eppendorf, 5417C) to decant the supernatant for later use. PEDOT:PSS (Clevios 4083, Heracus), chloroform, chlorobenzene (anhydrous, Sigma–Aldrich), P3HT (4002EE, Rieke Metals, Inc.), and PCBM (>99.5%, Sigma–Aldrich) were used as received. Vertical Current–Voltage Measurement: GO and GO:SWCNTs thin films were spin coated on ITO for these experiments and were subjected to thermal annealing at 150 °C for 20 min prior to measurements. To prevent shorting, ITO substrates were precoated with a layer of ZnO nanoparticles (30 nm). Two terminal measurements were performed using a probe station connected to a Keithley 2400 source meter. Lateral Current–Voltage Measurement: PEDOT:PSS and GO:SWCNTs thin films were spin coated on Si/SiO2 (300 nm) pretreated with oxygen plasma. Because the polymer photovoltaic devices are thermal annealed at 150 °C for 20 min before completion, the GO:SWCNTs and PEDOT:PSS thin films were also annealed at 150 °C for 20 min in ambient conditions prior to thermal deposition of Al electrodes through a shadow mask. Current-voltage measurements were carried out using a probe station connected to a Keithley 2400 source meter. Device Fabrication: (a) Regular tandem: the GO:SWCNTs dispersions were created by directly mixing the two constituents in a combination of deionized water and ethanol (7:3, v/v) by ultra-sonication. The resulting dispersions were then centrifuged to further remove the unwanted metallic catalysts and carbonaceous materials from the SWCNTs. The black stock solution was stable for months without any visible precipitation and can be readily deposited onto various substrates for characterization. GO:SWCNTs dispersions were spin coated at 2000 rpm for 40 s on ITO anodes (20 Ω sq−1) pretreated with oxygen plasma, followed by thermal annealing at 120 °C for 20 min. The modified substrates were then transferred into a nitrogen-filled glove box. A 0.85 wt% P3HT:PCBM (1:1, w/w) in chloroform was spin coated onto a PDMS stamp at 1500 rpm for 10 s. Next, P3HT:PCBM active layer was directly pressed against ITO/GO:SWCNTs atop the hot plate preheated at 85 °C to complete the front cell. For rear cells, the same type of P3HT:PCBM/PDMS stamp was prepared, followed by sequential spin casting of GO:SWCNTs and ZnO (30 mg mL−1 in acetone) at 2000 rpm. The completed rear cell was brought into electrical contact with the prefabricated front cell on a preheated hot plate at around 85 °C. Postannealing at 150 °C for 20 min was found to further improve the morphology of active layers at
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Energy Mater. 2012, 2, 299–303
www.advenergymat.de
www.MaterialsViews.com
Acknowledgements The work was supported by a booster award from the Initiative for Sustainability and Energy at Northwestern (ISEN) and National Science Foundation (CMMI 1130407). We thank the Sony Corporation and the Northrop Grumman Corporation for additional support. JH is an Alfred P. Sloan Research Fellow, and JK gratefully acknowledges support from the Ryan Fellowship and the Northwestern University International Institute for Nanotechnology. We thank Prof. S. I. Stupp for the use of his device fabrication and measurement system, NUANCE for the use of their microscopy facilities, Dr. A. R. Carretero for helping us on preparation of ZnO nanoparticles, and Dr. S. Sista, Prof. C.-W. Chu, A. Koltonow, A. T. Tan, and L. J. Cote for helpful discussions. Received: October 6, 2011 Revised: November 3, 2011 Published online: January 19, 2012
Adv. Energy Mater. 2012, 2, 299–303
[1] H. Y. Chen, J. H. Hou, S. Q. Zhang, Y. Y. Liang, G. W. Yang, Y. Yang, L. P. Yu, Y. Wu, G. Li, Nat. Photon. 2009, 3, 649. [2] T. Ameri, G. Dennler, C. Lungenschmied, C. J. Brabec, Energy Environ. Sci. 2009, 2, 347. [3] K. Lee, J. Y. Kim, N. E. Coates, D. Moses, T. Q. Nguyen, M. Dante, A. J. Heeger, Science 2007, 317, 222. [4] J. Gilot, M. M. Wienk, R. A. J. Janssen, Adv. Mater. 2010, 22, E67. [5] J. Yang, R. Zhu, Z. Hong, Y. J. He, A. Kumar, Y. Li, Adv. Mater. 2011, 23, 3465. [6] S. Sista, Z. R. Hong, L. M. Chen, Y. Yang, Energy Environ. Sci. 2011, 4, 1606. [7] V. C. Tung, J. Kim, L. J. Cote, J. X. Huang, J. Am. Chem. Soc. 2011, 133, 9262. [8] S. S. Li, K. H. Tu, C. C. Lin, M. Chhowalla, C. W. Chen, ACS Nano 2010, 4, 3169. [9] Y. Gao, H. L. Yip, S. K. Hau, K. M. O’Malley, N. C. Cho, H. Z. Chen, A. K. Y. Jen, Appl. Phys. Lett. 2010, 97, 203306. [10] L. J. Cote, F. Kim, J. Huang, J. Am. Chem. Soc. 2009, 131, 1043. [11] L. J. Cote, J. Kim, V. C. Tung, J. Luo, F. Kim, J. Huang, Pure Appl. Chem. 2011, 83, 95. [12] J. Kim, L. J. Cote, F. Kim, W. Yuan, K. R. Shull, J. X. Huang, J. Am. Chem. Soc. 2010, 132, 8180. [13] F. Kim, L. J. Cote, J. X. Huang, Adv. Mater. 2010, 22, 1954. [14] V. C. Tung, J. H. Huang, I. Tevis, F. Kim, J. Kim, C. W. Chu, S. I. Stupp, J. X. Huang, J. Am. Chem. Soc. 2011, 133, 4940. [15] J. Kim, V. C. Tung, J. Huang, Adv. Energy Mater. 2011, 1, 1052. [16] G. Li, V. Shrotriya, J. S. Huang, Y. Yao, T. Moriarty, K. Emery, Y. Yang, Nat. Mater. 2005, 4, 864. [17] J. H. Huang, Z. Y. Ho, T. H. Kuo, D. Kekuda, K. C. Ho, C. W. Chu, J. Mater. Chem. 2009, 19, 4077. [18] H. Wang, E. D. Gomez, J. Kim, Z. L. Guan, C. Jaye, D. A. Fischer, A. Kahn, Y. L. Loo, Chem. Mater. 2011, 23, 2020. [19] R. A. J. Janssen, W. J. E. Beek, M. M. Wienk, M. Kemerink, X. N. Yang, J. Phys. Chem. B 2005, 109, 9505. [20] H. H. Liao, L. M. Chen, Z. Xu, G. Li, Y. Yang, Appl. Phys. Lett. 2008, 92, 173303. [21] C. H. Chou, W. L. Kwan, Z. R. Hong, L. M. Chen, Y. Yang, Adv. Mater. 2011, 23, 1282. [22] S. O. Kim, J. E. Kim, T. H. Han, S. H. Lee, J. Y. Kim, C. W. Ahn, J. M. Yun, Angew. Chem. Int. Ed. 2011, 50, 3043. [23] F. Kim, J. Y. Luo, R. Cruz-Silva, L. J. Cote, K. Sohn, J. X. Huang, Adv. Funct. Mater. 2010, 20, 2867. [24] L. J. Cote, J. Kim, Z. Zhang, C. Sun, J. X. Huang, Soft Matter 2010, 6, 6096. [25] J. Gilot, M. M. Wienk, R. A. J. Janssen, Appl. Phys. Lett. 2007, 90, 143512. [26] A. V. Nikolaev, A. S. Nazarov, V. V. Lisitsa, Russ. J. Inorg. Chem. 1974, 19, 3396. [27] J. M. Lee, J. S. Park, S. H. Lee, H. Kim, S. Yoo, S. O. Kim, Adv. Mater. 2011, 23, 629. [28] W. S. Hummers, R. E. Offeman, J. Am. Chem. Soc. 1958, 80, 1339.
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
Communication
nanoscale as well as electrical contact of interconnecting layers. Finally, Ca (30 nm)/Al (100 nm) were thermally evaporated at
