www.MaterialsViews.com
FULL P
APER
Highly Effi cient Polymer Tandem Cells and Semitransparent
Cells for Solar Energy
Chih-Yu Chang , Lijian Zuo , Hin-Lap Yip , Chang-Zhi Li , Yongxi Li , Chain-Shu Hsu ,*
Yen-Ju Cheng , Hongzheng Chen ,* and Alex K-Y. Jen *
integration, which permits daylight to pass through while generating elec-tricity. [ 3–6 ] Currently, the most effi cient OPVs are based on the bulk-heterojunc-tion (BHJ) architecture, in which an elec-tron-donor (e.g., a conjugated polymer) is blended with an electron-acceptor (e.g., [6,6]-phenyl-C 61 -butyric acid methyl ester (PC 61 BM) or [6,6]-phenyl-C 71 -butyric acid methyl ester (PC 71 BM)) to form the active layer. [ 1,2 ] One of the factors limiting the power conversion effi ciency (PCE) of BHJ OPVs is insuffi cient absorption of solar radiation due to the active layer usually suffers from: 1) narrow absorption band-width [ 1,2 ] and 2) limited thickness because the low charge carrier mobility of organic semiconductors. [ 1,2,7 ] In addition, the ther-malization losses of hot charge carriers generated by high energy photons reduce the maximum achievable cell voltage. [ 8–10 ] A possible approach to address these issues is to stack indi-vidual cells with complementary absorption characteristics into a tandem cell structure. [ 7–10 ] This confi guration usually consists of a front cell with a large bandgap ( E g ) polymer ( E g > 1.7 eV), an interconnection layer (ICL), and a rear cell with a low bandgap (LBG) polymer ( E g < 1.5 eV). [ 10 ] However, the progress in polymer tandem cells is still limited due to lacking of high-performance LBG polymers. [ 8,9 ] Additionally, although most of the high-performance single-junction devices reported to date are based on medium bandgap polymers ( E g ≈ 1.6 eV), [ 10 ] their utilization in tandem cells remains quite challenging because their absorption have signifi cant overlap with other commonly used polymers. [ 10 ] Considering that PC
61 BM and PC 71 BM possess different absorption characteristics in the vis-ible region, [ 11 ] the combination of a medium bandgap polymer with PC 61 BM and PC 71 BM as the active layer would be an promising approach to realize the complementary absorption characteristics.
In this work, we demonstrate highly effi cient tandem polymer solar cells and ST solar cells utilizing the same donor polymer blended with PC 61 BM or PC 71 BM as active layers in two cells. The tandem device structure consists of two sub-cells that are electrically connected in series via an ICL of modi-fi ed poly(3,4-ethylene-dioxythiophene):poly(styrenesulfonate) (hereafter referred to as m-PEDOT:PSS)/high conductivity PEDOT:PSS (hereafter referred to as PH1000)/ZnO fi lms, as illustrated in Figure 1 a. Two polymers are used in this study:
DOI: 10.1002/aenm.201301645
Dr. C.-Y. Chang, L. Zuo, Dr. H.-L. Yip, Dr. C.-Z. Li, Y. X. Li, Prof. A. K.-Y. Jen
Department of Materials Science and Engineering University of Washington
Seattle, WA, 98195 , USA E-mail: [email protected]
Dr. C.-Y. Chang, Prof. C.-S. Hsu, Prof. Y.-J. Cheng Department of Applied Chemistry
National Chiao Tung University
1001 Ta Hseuh Road , Hsin-Chu , 30010 , Taiwan E-mail: [email protected]
L. Zuo, Prof. H. Z. Chen
State Key Laboratory of Silicon Materials MOE Key Laboratory of Macromolecule Synthesis and Functionalization
Zhejiang-California International Nanosystems Institute Zhejiang University
Hangzhou , 310027 , P. R. China E-mail: [email protected]
Highly effi cient tandem and semitransparent (ST) polymer solar cells utilizing the same donor polymer blended with [6,6]-phenyl-C 61 -butyric
acid methyl ester (PC 61 BM) and [6,6]-phenyl-C 71 -butyric acid methyl
ester (PC 71 BM) as active layers are demonstrated. A high power
conver-sion effi ciency (PCE) of 8.5% and a record high open-circuit voltage of 1.71 V are achieved for a tandem cell based on a medium bandgap polymer poly(indacenodithiophene- co -phananthrene-quinoxaline) (PIDT-phanQ). In addition, this approach can also be applied to a low bandgap polymer poly[2,6-(4,4-bis(2-ethylhexyl)-4H-cyclopenta[2,1-b;3,4-b′]dithiophene)-alt-4,7-(5-fl
uoro-2,1,3-benzothia-diazole)] (PCPDTFBT), and PCEs up to 7.9% are achieved. Due to the very thin total active layer thickness, a highly effi cient ST tandem cell based on PIDT-phanQ exhibits a high PCE of 7.4%, which is the highest value reported to date for a ST solar cell. The ST device also possesses a desirable average visible transmittance (≈40%) and an excellent color rendering index (≈100), permitting its use in power-generating window applications.
1 . Introduction
Organic photovoltaics (OPVs) are of great interest as an alter-native renewable energy source to typical silicon-based photo-voltaic cells due to their potential for cost-effective large-area manufacturing, light-weight, mechanical fl exibility, and semi-transparent (ST) characteristics. [ 1–6 ] An emerging market seg-ment for OPVs is ST solar cells that can be used for window
www.MaterialsViews.com
FULL P
APER
a medium bandgap polymer poly(indacenodithiophene- co -phananthrene-quinoxaline) (PIDT-phanQ; chemical structure shown in Figure 1 b) and a LBG polymer poly[2,6-(4,4-bis(2- ethylhexyl)-4H-cyclopenta[2,1-b;3,4-b′]dithiophene)-alt-4,7-(5-fl uoro-2,1,3-benzothia-diazole)] (PCPDTFBT; chemical structure shown in Figure 1 c). The tandem cell based on PIDT-phanQ exhibits a very high PCE up to 8.5%, which is the fi rst demon-stration of a tandem cell based on a medium bandgap polymer with such a high PCE. This approach can also be applied to a LBG polymer PCPDTFBT-based tandem cell to achieve a high PCE up to 7.9%. More importantly, the thin active layers used in these tandem cells make them ideal for ST solar cells. A record high PCE of 7.4% with an average visible transmittance (AVT) of ≈40% and a general color rendering index (CRI) of ≈100 could be achieved for PIDT-phanQ-based ST cells.
2 . Results and Discussion
2.1 . Single-Junction Solar Cells
To achieve highly-effi cient tandem solar cells, the optimization of active layer thickness of the sub-cells is critical for balanced light absorption and matched photocurrent. [ 7–10 ] Therefore, single-junction devices based on PIDT-phanQ blended with PC 61BM or PC 71BM (device structure: indium tin oxide (ITO)-coated glass/ZnO/fullerene self-assembled monolayer (C 60 -SAM)/PIDT-phanQ:PC 61BM or PC 71 BM/m-PEDOT:PSS/ Ag) were preliminarily studied. A C 60 -SAM layer was used to passivate ZnO surface traps and enhance electronic coupling between ZnO and the active layer. [ 12,13 ] The current–voltage ( J – V ) characteristics of the devices were measured under AM1.5 illumination (Supporting Information Figure S1), and the cor-responding device parameters were summarized in Supporting Information Table S1. It is clearly shown that PCE decreased when the thickness of the active layer increased (Table S1 and Figure S1) independent from the type of fullerene derivatives used. This is consistent with previous observations that thicker active layer tends to increase bimolecular charge recombination and cause a space charge effect. [ 2,14,15 ]
Based on the optimal active layer thickness, Device A with PC 61 BM acceptor exhibited a short-circuit current density ( J sc ) of 8.56 mA cm −2 , an open-circuit voltage ( V
oc ) of 0.85 V, a fi ll factor (FF) of 68.56%, and a PCE of 5.0%, which was inferior to that of Device B with PC 71 BM acceptor (PCE = 6.5%, with
a J sc of 11.45 mA cm −2 , an V oc of 0.87 V, and a FF of 65.04%), as shown in Table 1 and Figure 2 . The good agreement between the measured J sc and that extrapolated from the external quantum effi ciency (EQE) spectra confi rms the accuracy of the reported PCE values (Table 1 ). The difference in PCE between Device A and Device B was mainly due to different J sc values (Table 1 and Figure 2 ), which can be rationalized by comparing the spectral response in the EQE spectra ( Figure 3 a). Device A revealed two dominant peaks at 435 nm and 645 nm (with the max-imum EQE of ≈55% at 435 nm), while Device B revealed a broad spectral response in the range between 350 and 700 nm (with the maximum EQE of ≈65% at 540 nm). This distinction was also manifested in the absorption spectra of the blend fi lms (Figure 3 b), which can be ascribed to the fact that PC 71 BM possesses better light absorption in the visible region compared to that of PC 61 BM. [ 11 ]
It is worthy noting that both Device A and Device B pos-sessed high FF values (>65%), as shown in Table 1 . The reason
Figure 1. a) Schematic representation of the tandem device architecture used in this study.
Chemical structures of: b) PIDT-phanQ and c) PCPDTFBT.
Table 1. Summary of the photovoltaic properties of solar cells.
Device Condition a) V oc [V] J sc [mA cm −2 ] FF [%] PCE [%] A PIDT-phanQ, front cell 0.85 8.56 (8.45) d) 68.56 5.0
B PIDT-phanQ, rear cell 0.87 11.45 (11.24) d) 65.04 6.5
C PIDT-phanQ, tandem cell 1.71 6.91 65.70 7.8 D PCPDTFBT, front cell 0.75 10.61 62.09 4.9 E PCPDTFBT, rear cell 0.76 13.90 58.08 6.1 F PCPDTFBT, tandem cell 1.51 8.39 57.04 7.2 G PIDT-phanQ, tandem cell b) 1.68 5.93 68.58 8.5
H PCPDTFBT, tandem cell c) 1.47 5.75 58.64 7.9
I PIDT-phanQ, ST cell 1.70 5.81 67.40 6.7 J PIDT-phanQ, ST cell b) 1.63 5.23 68.94 7.4
a) Polymer and device architecture; b) Light intensity = 80 mw cm −2 ; c) Light intensity
= 63 mW cm −2 ; d) Calculated from the EQE spectra shown in Figure 3 a.
Figure 2. J – V characteristics of solar cells based on PIDT-phanQ (see
www.MaterialsViews.com
FULL P
APER
high, with the maximum values approaching 100%, confi rming the effi cient exciton dissociation, charge transport, and charge collection within the devices.
2.2 . Tandem Soar Cells
The device characteristics of a double-junction tandem solar cell comprising a front cell of PIDT-phanQ:PC 61 BM (70 nm) and a rear cell of PIDT-phanQ:PC 71 BM (85 nm) was then studied. An ICL consisting of m-PEDOT:PSS/PH1000/ZnO is used to connect the sub-cells effectively because it possesses desired high optical transparency (≈85% in 700–900 nm), reasonable conductivity (≈15 S cm −1 ), smooth surface (root-mean-square roughness = 0.88 nm), and excellent robustness against solvent erosion. [ 16 ] Under the AM1.5G illumination (with an intensity of 100 mW cm −2 ), Device C showed a high PCE of 7.8%, with a J sc of 6.91 mA cm −2 , a FF of 65.70%, and an exceptional V oc of 1.71 V (Table 1 and Figure 2 ). This represents the highest V oc value reported to date for a double-junction organic tandem cell. This high V oc value is also close to the sum of the V oc of two sub-cells (Table 1 and Figure 2 ), confi rming that they are electronically coupled in series through the ICL. It is worth noting that the measured J sc of the tandem cell (6.91 mA cm −2 ; Device C in Table 1 ) agrees well with the predicted value from optical simulation (6.97 mA cm −2 ), indicating there is no sig-nifi cant electrical loss in the tandem cell.
We have also examined if similar strategy can be used for LBG polymer-based tandem cells. PCPDTFBT was previously proven to be an effi cient polymer for OPV, possessing good sol-ubility in common organic solvents, broad absorption band that extends into near IR with an onset of ≈920 nm ( E g = 1.35 eV), high charge carrier mobility (≈10 −2 cm 2 V −1 S −1 ), and good air stability. [ 16,17 ] The optimal PCE for PCPDTFBT:PC
61 BM (Device D) and PCPDTFBT:PC 71 BM (Device E) single-junction cells was 4.9% and 6.1%, respectively (Table 1 and Figure 5 ). A double-junction tandem solar cell consisting of a PCPDTFBT:PC 61 BM (80 nm) front cell and a PCPDTFBT:PC 71BM (85 nm) rear cell was fabricated (Device F). Under the AM1.5G illumina-tion (100 mW cm −2 ), Device F exhibited a high PCE of 7.2% for the high FF is most probably the thin active layer thickness
(≈80 nm), which can decrease the charge carrier recombination losses within the active layer. [ 1,2 ] To gain further insight into the effectiveness of the overall photoconversion process within the devices, the internal quantum effi ciencies (IQEs) of the devices were evaluated by measuring their total absorption spectra (based on the refl ection geometry) and EQE spectra. [ 6 ] As shown in Figure 4 , the IQEs of both devices were remarkably
Figure 3. a) EQE spectra of the as-fabricated devices (see Table 1 for
descriptions of the device types) and b) absorption spectra of PIDT-phanQ:PC 61 BM and PIDT-phanQ:PC 71 BM fi lms.
Figure 4. IQE spectra of the as-fabricated devices (see Table 1 for descriptions of the device types).
Figure 5. J – V characteristics of the solar cells based on PCPDTFBT (see
www.MaterialsViews.com
FULL P
APER
improved PCE (7.4%) was obtained under a low light intensity of 80 mW cm −2 (Device J; Table 1 and Figure 7 ), representing the highest value reported for a ST OPV. Moreover, Device I also possessed a high AVT (≈40%) in the range between 380 and 700 nm ( Figure 8 ). The photographic image of Device I is shown in the inset of Figure 8 ; a University of Washington campus building can be clearly visualized through the device.
In addition to optical transmittance, suitable solar cells for window application also require good color rendering proper-ties and transparency perception by the human eye for realistic scene illumination. [ 4,5 ] Therefore, the Commission Interna-tionale de l’Eclairage (CIE) 1931 color coordinates, correlative color temperature (CCT), and general CRI of Device I were evaluated. Device I had a CCT of 5622 K and CIE color coor-dinates of (0.3297, 0.3387), which is close to AM1.5G light illu-mination ( Figure 9 ), indicating good achromatic when looking through the device under AM1.5G illumination. In addition, Device I possessed an excellent CRI of 97.2, clearly demon-strating the exceptional color rendering properties of the device. The combined high PCE, good AVT, excellent color perception, and rendering properties of the devices enable them to be used in power-generating window applications.
with a J sc of 8.39 mA cm −2 , a FF of 57.04%, and an V oc of 1.51 V (Table 1 and Figure 5 ). These results show the feasibility of using the same donor polymer to blend with PC 61 BM and PC 71 BM to serve as active layer in two sub-cells for constructing highly effi cient tandem cells.
We also studied the light-intensity dependence of the pho-tocurrent, which can provide the information about which type of recombination (e.g. monomolecular or bimolecular) is dominant and whether space charge effects play a role. [ 18,19 ] The short-circuit current densities of PIDT-phanQ- and PCP-DTFBT-based tandem cells as function of light intensity are shown in Figure 6 . For both devices, we observed a linear relationship (the slopes are 1.02 and 1.04 for PIDT-phanQ- and PCPDTFBT-based device, respectively) between the light intensity and short-circuit current density, suggesting that the bimolecular recombination and space charge effects can be neglected. [ 18,19 ] In addition, the light-intensity dependence PCE was also evaluated. The J – V characteristics of the devices were shown in Figure S2 and Figure S3 (Supporting Information). Very encouragingly, the device performance of both devices can be further increased under lower light intensity. For the PIDT-phanQ-based tandem cells, a high PCE of up to 8.5% can be reached under a light intensity of 80 mW cm −2 (Device G; Table 1 and Figure 2 ), while a PCE up to 7.9% can be reached under a light intensity of 63 mW cm −2 for the PCPDTFBT-based tandem cells (Device H; Table 1 and Figure 5 ). These results suggest that these devices can be used for indoor applications.
2.3 . Semitransparent Solar Cells
The thin active layers (total thickness of 155 nm) in the tandem cell enable them to be used for making ST cell. To verify this, the device characteristics of the PIDT-phanQ-based ST cell (Device I) were evaluated. The device fabrication procedure was similar as that used for Device H, except that the thickness of the top Ag electrode was reduced to ≈10 nm. Device I exhib-ited a high PCE of 6.7% (light intensity = 100 mW cm −2 ), with a J sc of 5.75 mA cm −2 , an V oc of 1.47 V, and a FF of 58.64% (Table 1 and Figure 7 ). Similar to previous tandem cells, an
Figure 6. Short-circuit current density plotted against incident light
inten-sity for PIDT-phanQ- and PCPDTFBT-based tandem cells.
Figure 7. J – V characteristics of ST solar cells based on PIDT-phanQ (see
Table 1 for descriptions of the device types).
Figure 8. Optical transmittance of Device I (the description of the type
www.MaterialsViews.com
FULL P
APER
spin-coated onto pre-cleaned ITO-coated glass, followed by thermal annealing in air at 130 °C for 5 min (≈20 nm). Subsequently, a C 60 -SAM
layer was deposited on ZnO using a spin-coating process as previously reported. [ 12 ] The substrates were washed with tetrahydrofuran (THF)
twice to remove unbound C 60-SAM molecules. The active layer was
then spin-cast from a solution containing a mixture of either PIDT-phanQ:acceptor (1:3 w/w) or PCPDTFBT:acceptor (1:2.5 w/w), followed by annealing at 110 °C for 5 min. The modifi ed PEDOT:PSS layer (≈70 nm) was then spin-coated from PEDOT:PSS solution (Clevious P VP A1 4083) diluted with equal volume of isopropyl alcohol and 0.2 wt% of Zonyl FSO fl uorosurfactant. Finally, an Ag layer (150 nm) was then deposited under high vacuum (<10 −6 torr) through a shadow mask,
which defi ned an active area of 0.036 cm 2 .
Tandem Cell Fabrication : The front cells were made according to the procedure for the single-junction cell fabrication. A PH1000 layer (≈40 nm) was spin-coated onto the m-PEDOT:PSS layer from its solution (Clevious PH1000) and annealed at 120 °C for 5 min. The surface of the PH1000 layer was then washed with methanol to increase the conductivity and reduce the surface roughness as reported elsewhere. [ 21 ]
Next, the rear cells were made according to the procedure for the single-junction cell fabrication. Finally, MoO 3 (7 nm)/Ag (150 nm) layers were
deposited under high vacuum (<10 −6 torr) through a shadow mask,
which defi ned an active area of 0.036 cm 2 .
Characterization : The current–voltage characteristics of un-encapsulated solar cell devices were measured under ambient conditions using a Keithley 2400 source-measurement unit. An Oriel xenon lamp (450 W) with an AM1.5 G fi lter was used as the solar simulator. Contributions to the J sc from regions outside the active
area were eliminated using illumination masks with an aperture size of 0.0314 cm 2 . A Hamamatsu silicon solar cell with a KG5 color fi lter,
which is traced to the National Renewable Energy Laboratory (NREL), was used as the reference cell. To calibrate the light intensity of the solar simulator, the power of the xenon lamp was adjusted to make the J SC
of the reference cell under simulated sun light as high as it was under the calibration condition. The spectral mismatches resulting from the test cells, the reference cell, the solar simulator, and the AM1.5 were calibrated with mismatch factors ( M ). The mismatch factor is defi ned as according to Shrotriya et al. [ 22 ]
M = 82 81 ERef(8)SR(8)d8 82 81 ES(8)ST(8)d8 82 81 ERef(8)ST(8)d8 82 81 ES(8)SR(8)d8
where E Ref ( λ ) is the reference spectral irradiance (AM1.5), E S ( λ ) is the
source spectral irradiance, S R ( λ ) is the spectral responsivity of the
reference cell, and S T ( λ ) is the spectral responsivity of the test cell,
each as a function of wavelength ( λ ). The spectral responsivities of the test cells and the reference cell were calculated from the corresponding external quantum effi ciencies (EQE) by the relationship
S(8) = qhc8EQE (8)
where the constant term q / hc equals 8.0655 × 10 5 for wavelengths in
units of meters and S ( λ ) in units of A W −1 . The Hamamatsu solar cell
was also used as the detector for determining the spectral irradiance of the solar simulator. To minimize the spectral transformation, the irradiance spectrum was calibrated with the spectral responsively of the Hamamatsu cell and the grating effi ciency curve of the monochromator (Oriel Cornerstone 130). UV-Vis absorption spectra were recorded with Perkin-Elmer Lambda-9 spectrophotometer at room temperature. The optical simulations were performed based on the transfer matrix formalism. [ 4,23 ]
Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.
3 . Conclusions
In summary, highly effi cient tandem and ST polymer solar cells were demonstrated by utilizing the same donor polymer blended with PC 61BM and PC 71BM as the active layers in two sub-cells. A high PCE of 8.5% and a record high V oc of 1.71 V were achieved for a tandem cell using the same medium bandgap polymer PIDT-phanQ. Similarly, a high PCE (7.9%) was obtained in a tandem cell using an LBG polymer PCP-DTFBT. Due to its very thin total active layer thickness, a highly effi cient ST tandem cell could also be fabricated using PIDT-phanQ as a donor; it exhibited a high PCE of 7.4%, which is the highest value reported to date for a ST OPV. This ST device also possessed a high AVT (≈40%) and an excellent CRI (97.2), which enables its use for power-generating window applications.
4 . Experimental Section
Materials : PIDT-phanQ and PCPDTFBT were synthesized in-house, and the detailed synthesis was reported previously. [ 17,20 ] The number
average molecular weights of PIDT-phanQ and PCPDTFBT were ≈82.4 kDa (polydispersity index = 2.42) and ≈25.3 kDa (polydispersity index = 1.5), respectively, as determined by gel permeation chromatography. PC 61 BM
and PC 71 BM were purchased from American Dye Source. Unless otherwise
stated, all chemicals were purchased from Aldrich and used as received. Single-Junction Cell Fabrication : A ZnO precursor solution consisting of 20 mg mL −1 zinc acetylacetonate hydrate in anhydrous ethanol was
Figure 9. a) The color coordinates of Device I under AM1.5G
illumina-tion on the CIE chromaticity diagram and b) an enlarged image from (a). The color coordinate representation of AM1.5G illumination is also presented.
www.MaterialsViews.com
FULL P
APER
[9] J. You , L. Dou , K. Yoshimura , T. Kato , K. Ohya , T. Moriarty , K. Emery , C.-C. Chen , J. Gao , G. Li , Y. Yang , Nat. Commun. 2013 , 4 , 1446 . [10] J. You , L. Dou , Z. Hong , G. Li , Y. Yang , Prog. Polym. Sci. 2013 , 38 ,
1909 .
[11] Y. Yao , C. Shi , G. Li , V. Shrotriya , Q. Pei , Y. Yang , Appl. Phys. Lett. 2006 , 89 , 153507 .
[12] S. K. Hau , H.-L. Yip , O. Acton , N. S. Baek , H. Ma , A. K. Y. Jen , J.
Mater. Chem. 2008 , 18 , 5113 .
[13] S. K. Hau , H.-L. Yip , H. Ma , A. K. Y. Jen , Appl. Phys. Lett. 2008 , 93 , 233304 .
[14] M. Lenes , L. J. A. Koster , V. D. Mihailetchi , P. W. M. Blom , Appl.
Phys. Lett. 2006 , 88 , 243502 .
[15] C. G. Shuttle , B. O'Regan , A. M. Ballantyne , J. Nelson , D. D. C. Bradley , J. R. Durrant , Phy. Rev. B 2008 , 78 , 113201 . [16] C.-Y. Chang , L. Zuo , H.-L. Yip , Y. Li , C.-Z. Li , C.-S. Hsu , Y.-J. Cheng ,
H. Chen , A. K. Y. Jen , Adv. Funct. Mater. 23 , 5084 .
[17] Y. Zhang , J. Zou , C.-C. Cheuh , H.-L. Yip , A. K. Y. Jen , Macromolecules 2012 , 45 , 5427 .
[18] L. J. A. Koster , V. D. Mihailetchi , H. Xie , P. W. M. Blom , Appl. Phys.
Lett. 2005 , 87 , 203502 .
[19] S. R. Cowan , J. Wang , J. Yi , Y.-J. Lee , D. C. Olson , J. W. P. Hsu , J.
Appl. Phys. 2013 , 113 , 154504 .
[20] Y. Zhang , J. Zou , H.-L. Yip , K.-S. Chen , D. F. Zeigler , Y. Sun , A. K. Y. Jen , Chem. Mater. 2011 , 23 , 2289 .
[21] D. Alemu , H.-Y. Wei , K.-C. Ho , C.-W. Chu , Energy Environ. Sci. 2012 , 5 , 9662 .
[22] V. Shrotriya , G. Li , Y. Yao , T. Moriarty , K. Emery , Y. Yang , Adv. Funct.
Mater. 2006 , 16 , 2016 .
[23] L. A. A. Pettersson , L. S. Roman , O. Inganas , J. Appl. Phys. 1999 , 86 , 487 .
Acknowledgements
C.-Y.C. and L.Z. contributed equally to this work. This work was supported by the AFOSR (FA9550–09–1–0426), the Offi ce of Naval Research (N00014–11–1–0300), and the Asian Offi ce of Aerospace Research and Development (FA2386–11–1–4072). A. K.-Y.J. thanks the Boeing-Johnson Foundation for fi nancial support.
Received: October 29, 2013 Revised: November 7, 2013 Published online: December 12, 2013
[1] K. M. Coakley , M. D. McGehee , Chem. Mater. 2004 , 16 , 4533 . [2] B. C. Thompson , J. M. J. Fréchet , Angew. Chem. Int. Ed. 2008 , 47 ,
58 .
[3] S. K. Hau , H.-L. Yip , J. Zou , A. K. Y. Jen , Org. Electron. 2009 , 10 , 1401 .
[4] K.-S. Chen , J.-F. Salinas , H.-L. Yip , L. Huo , J. Hou , A. K. Y. Jen , Energy Environ. Sci. 2012 , 5 , 9551 .
[5] C.-C. Chueh , S.-C. Chien , H.-L. Yip , J. F. Salinas , C.-Z. Li , K.-S. Chen , F.-C. Chen , W.-C. Chen , A. K. Y. Jen , Adv. Energy Mater. 2013 , 3 , 417 . [6] Z. Tang , Z. George , Z. Ma , J. Bergqvist , K. Tvingstedt , K. Vandewal , E. Wang , L. M. Andersson , M. R. Andersson , F. Zhang , O. Inganäs , Adv. Energy Mater. 2012 , 2 , 1467 .
[7] J. Y. Kim , K. Lee , N. E. Coates , D. Moses , T.-Q. Nguyen , M. Dante , A. J. Heeger , Science 2007 , 317 , 222 .
[8] T. Ameri , G. Dennler , C. Lungenschmied , C. J. Brabec , Energy
