Renewable Energy 72 (2014) 134e139
Contents lists available at ScienceDirect
Renewable Energy journal homepage: www.elsevier.com/locate/renene
“Solar tree”: Exploring new form factors of organic solar cells Weiran Cao, Zhifeng Li, Yixing Yang, Ying Zheng, Weijie Yu, Rimza Afzal, Jiangeng Xue* Department of Materials Science and Engineering, University of Florida, Gainesville, FL 32611, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 3 December 2013 Accepted 25 June 2014 Available online
Organic solar cells have great potential as a clean and renewable solar energy conversion system, due to their low cost materials, ease of production, and lack of harmful emissions. The rapid improvement in organic solar cell performance in recent years has triggered signiﬁcant interests in developing organic solar cells for commercial applications. Harnessing the unique set of characteristics of organic solar cells, here we demonstrate a new form factor for organic solar cells, a “solar tree” or an electricity-generating artiﬁcial tree with organic solar cells as leaves. We ﬁrst fabricated polymer:fullerene based organic solar cells on ﬂexible plastic substrates that show similar performance to devices on rigid glass substrates using the inverted device structure. Large-area ﬂexible devices were fabricated and cut into palm leaf shapes with an active device area of 6.5 cm2 using a steel rule die. 12 leaf-shaped organic solar cells were then assembled to form a prototype “solar palm tree”. Two different wiring conﬁgurations among the devices provided different power delivery modes: a low-voltage, high-current “fan mode” and a high voltage, low-current “LED mode”. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Organic photovoltaic cells Flexible solar cells Solar tree Inverted structure
1. Introduction The search for clean and renewable energy sources has become one of the greatest challenges for our society, due to the rapid depletion of fossil fuels and increasing demand on energy supply. One of the most promising alternative energy sources is solar energy, which is clean, renewable, safe and abundant [1,2]. Using solar cells to directly convert sunlight to electricity is one of the major methods to utilize such an abundant energy source. Compared to inorganic semiconductors that are used in commercial solar modules, organic solar cells have several advantages such as the low material cost, easy tunability of material properties and compatibility with large-area and roll-to-roll manufacturing technologies [1e7]. The power conversion efﬁciency, hP, now reaches over 10% with the recent advance in new active material and device architecture designs [8e14]. In addition to conventional applications of solar cells as powergenerating sources, the light weight, high ﬂexibility, and rich colors of organic semiconductors enable new form factors and unique applications of organic solar cells. For example, organic solar cells can be integrated into fabrics, such as jackets, handbags, and tents . They can also be used in landscaping applications to substitute * Corresponding author. Tel.: þ1 352 846 3775; fax: þ1 352 846 3355. E-mail address: [email protected]
ﬂ.edu (J. Xue). http://dx.doi.org/10.1016/j.renene.2014.06.045 0960-1481/© 2014 Elsevier Ltd. All rights reserved.
for natural grass, ﬂowers, or trees. Such “solar plants” not only can preserve the aesthetic appearance of a location, but provide economic and environmental beneﬁts as they avoid irrigation, application of pesticides/fertilizers, trimming/mowing, and other maintenance needs for natural plants and grass. This makes them very appealing for regions where obtaining sufﬁcient water supply is at a premium, and at the same time minimizing potential adverse impact on the environment. Here we demonstrate the viability of assembling a “solar palm tree” based on organic solar cells on ﬂexible polyethylene terephthalate (PET) substrates. Using a poly(3-hexylthiophene) (P3HT) and [6,6]-phenyl-C61-butyric acid methyl ester (PCBM) ﬁlm as the active layers, such ﬂexible devices show a maximum power conversion efﬁciency hP ¼ 2.7 ± 0.1%, compared to hP ¼ 3.0 ± 0.1% for similar devices on rigid glass substrates. The ﬂexible devices show high ﬂexibility and hP was only decreased by ~12% after 1000 bending cycles with a bending radius of 1.2 cm. We also scaled up the device area to more than 6 cm2 on ﬂexible substrates, and demonstrated “solar palm leaves” by cutting the ﬂexible organic solar cells into palm leave shapes using a steel rule die. 12 pieces of leaf-shape devices were then assembled together to construct a prototype “solar palm tree”, in which two different wiring conﬁgurations among the devices provided two power delivery modes: a low-voltage, high-current “fan mode” and a high-voltage, lowcurrent “LED mode”.
W. Cao et al. / Renewable Energy 72 (2014) 134e139
2. Experiment 2.1. Device fabrication Comparative devices were fabricated on two types of substrates with different transparent conducting electrodes (TCE): glass substrates coated with a transparent indium tin oxide (ITO) electrode, and ﬂexible PET substrates coated with a transparent In2O3/Au/Ag (IAA, Delta Technologies) electrode. The sheet resistance of ITO is 19 ± 1 U/,, whereas the IAA electrodes have two different sheet resistances, 20 ± 2 U/, and 60 ± 5 U/, (denoted as IAA-L and IAA-H, respectively). All the substrates were successively sonicated in a solution of Liquinox, deionized water, acetone, and isopropanol, followed by 15 min treatment under UV-ozone prior to device fabrication. P3HT (purchased from Rieke Metal Inc.) and PCBM (purchased from Nano-C Inc.) were dissolved in chlorobenzene (1:0.8 weight ratio with a total concentration of 27 mg/mL) and stirred for 24 h before use. ZnO nanoparticles (NPs) were synthesized as reported previously and dispersed in ethanol [9,16,17]. Fig. 1 schematically shows two types of device structures studied here. For the “normal” device structure (Fig. 1(a)), a 40 nm thick poly(3,4ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) layer was ﬁrst spin-coated from an aqueous solution and annealed in air at 150 C for 15 min. The active layer was then spin-coated onto the PEDOT:PSS layer from the P3HT:PCBM solution at 1000 rpm for 1 min in a nitrogen-ﬁlled glove-box. The devices were completed by thermally evaporating a 100 nm thick Al layer, followed by a postfabrication thermal annealing at 150 C for 30 min in a N2 atmosphere. For the “inverted” device structure (Fig. 1(b)), a 40 nm thick ZnO nanoparticle layer was ﬁrst spin-coated onto the TCE, followed by spin-coating the P3HT:PCBM layer. The samples were then annealed at 150 C for 30 min in a N2 atmosphere. The “inverted” devices were completed by thermally evaporating a 5 nm thick layer of MoO3 and a 100 nm thick Al layer on the organic layer. All the annealing temperature was reduced to 110 C for devices on the PET substrates to avoid thermal damage to the PET ﬁlms. For comparison, devices on glass substrate were also annealed at 110 C. In the normal structures, the TCE serves as the anode for hole collection, whereas Al serves as the cathode. In the inverted devices, the low work function of the ZnO NP layer [18e21] and the high work function of the MoO3 layer [22,23] lead to electron and hole extraction at the TCE and Al electrodes, respectively, thus inverting the polarity of the cells. The active device area for the small-area organic solar cells is 4 mm2, and device performance was averaged over 16 devices. For leaf-shaped devices, the active device area is 6.5 mm2.
using an Agilent 4155C semiconductor parameter analyzer. An Oriel solar simulator equipped with a Xe-arc lamp was used to provide simulated AM 1.5G solar illumination. The light intensity was measured using a calibrated single-crystal silicon reference cell with a KG1 ﬁlter . To measure the external quantum efﬁciency (EQE) of these devices, a monochromatic light with varying wavelength was generated from a tungsten lamp through an Oriel monochromator, and chopped at 400 Hz by a mechanical chopper prior to incident on solar cells. The photocurrent was measured using a Stanford Research System 830DPS lock-in ampliﬁer and a Keithley 428 current ampliﬁer. The optical transmittance of ﬁlms was calculated from the incident, transmitted and reﬂected light intensities measured using the same setup as the EQE measurement. 3. Result and discussion 3.1. Devices on rigid glass substrates The JeV characteristics for the normal and inverted organic solar cells on glass substrates under 1 sun AM 1.5G illumination are shown in Fig. 2(a). All the device parameters are summarized in Table 1. The inverted device has a slightly higher open-circuit
2.2. Device characterization The currentevoltage (JeV) characteristics of the organic solar cells in the dark and under white light illumination were measured
Fig. 1. Schematic illustration of organic solar cells with (a) normal device structure and (b) inverted device structure.
Fig. 2. Current densityevoltage (JeV) characteristics of (a) normal and inverted P3HT:PCBM organic solar cells on glass substrates with different annealing temperature (110 C and 150 C) and (b) normal and inverted devices on PET substrates with different sheet resistance (annealing at 110 C) in the dark and under simulated 1 sun AM 1.5 solar illumination.
W. Cao et al. / Renewable Energy 72 (2014) 134e139
voltage (Voc), 0.63 ± 0.01 V vs. 0.61 ± 0.01 V, and a slightly lower short-circuit current density (Jsc), 8.3 ± 0.3 mA/cm2 vs. 8.6 ± 0.3 mA/cm2, than the normal structure device. The inverted device has a much higher ﬁll factor (FF), FF ¼ 0.69 ± 0.01, compared to that of the normal structure device (FF ¼ 0.58 ± 0.01), leading to a higher power conversion efﬁciency (hP) (3.6 ± 0.2% for inverted device and 3.1 ± 0.3% for normal device). As the devices on the ﬂexible PET substrates were only annealed at 110 C due to the limited thermal tolerance of the PET substrates, instead of 150 C for those on glass substrates, we also reduce the annealing temperature (Ta, including the PEDOT:PSS and P3HT:PCBM active layer annealing) from 150 C to 110 C for the devices on glass. We observed a reduction in both Voc and Jsc by <7% relative change; however, the reduction in FF is signiﬁcantly larger, decreasing from 0.58 ± 0.01 to 0.50 ± 0.02 for the normal structure devices and from 0.69 ± 0.01 to 0.60 ± 0.01 for the inverted devices, when Ta was reduced from 150 C to 110 C. This results in a decrease in hP, from 3.1 ± 0.3% to 2.4 ± 0.2% for the normal devices and from 3.6 ± 0.2% to 3.0 ± 0.1% for the inverted devices. It has been reported extensively that this different device performance is related to dependence of the phase separation of the P3HT:PCBM active layer on the annealing temperature [25e28]. 3.2. Devices on ﬂexible PET substrates As shown in Fig. 2(a) and Table 1, the normal and inverted devices on glass substrates show mostly similar PV performance. However, for devices on ﬂexible PET substrates, the normal and inverted device structures lead to drastically different device performance. As shown in Fig. 2(b), the normal structure devices on the PET substrates shows rather poor photovoltaic characteristics, with Voc ¼ 0.10 ± 0.04 V and Jsc ¼ 1.3 ± 0.4 mA/cm2, compared to Voc ¼ 0.61 ± 0.01 V and Jsc ¼ 8.0 ± 0.2 mA/cm2 for similar devices on glass substrates (also see Table 1). We speculate that such poor device performance is attributed to the unfavorable interface between the IAA electrodes and the PEDOT:PSS layer, likely due to the low work function of the IAA electrodes. In past years, the inverted device architecture has been investigated to remove the PEDOT:PSS layer that limits the device stability, and also to avoid the thermal vacuum deposition of the top metal electrode which is not suitable for roll-to-roll processing [29e33] Here, we show that the inverted device structure also helps to obtain reasonable photovoltaic characteristics from devices on the PET substrates. As already summarized in Table 1, the ﬂexible devices (either on IAA-L or IAA-H) have a slightly lower Voc (0.59 ± 0.01 V vs. 0.61 ± 0.01 V), compared to that of the rigid inverted devices annealed at the same temperature (110 C). The FF of the ﬂexible devices is also lower than that of the rigid devices, although it was found that reducing the sheet resistance of the IAA electrode helps to improve the FF (0.61 ± 0.01 for IAA-L vs. 0.56 ± 0.02 for IAA-H), due to reduction in the device series resistance. On the other hand, while the ﬂexible device with the IAA-H electrode has similar Jsc to the rigid device (Jsc ¼ 8.3 ± 0.3 mA/cm2),
Jsc of the ﬂexible device with the IAA-L electrode is signiﬁcantly lower (Jsc ¼ 6.3 ± 0.1 mA/cm2). This is consistent with the EQE spectra of these inverted devices as shown in Fig. 3. Integrating these EQE spectra with the standard 1 sun AM 1.5G solar spectrum gives a Jsc of 8.2 ± 0.1 mA/cm2, 6.2 ± 0.2 mA/cm2, and 8.1 ± 0.2 mA/ cm2 for the glass/ITO, PET/IAA-L and PET/IAA-H devices, respectively, in good agreement with the experimental data. The lower EQE in the IAA-H devices could at least partially be attributed to the lower optical transmittance. Also shown in Fig. 3, the PET/IAA-L/ ZnO ﬁlm has a lower transmittance (~70%), compared to that of the Glass/ITO/ZnO (~90%) and the PET/IAA-H/ZnO (80e85%), which leads to a lower Jsc and EQE for the corresponding devices. While we do not have the speciﬁc material information about the IAA electrodes, we speculate that the increased metal content leads to the lower sheet resistance and the lower transmittance in the IAA-L electrodes, as compared to the IAA-H electrodes. To better evaluate the performance of these ﬂexible inverted organic solar cells, we have measured the device performance after multiple bending cycles with a bending radius of r ¼ 1.2 cm. As shown in Fig. 4(a), only a ~12% drop in hP after 1000 bending cycles was observed for the ﬂexible device with the IAA-L electrode. However, for the device with IAA-H electrode, the reduction in Jsc and FF after bending is more severe, as shown in Fig. 4(b), resulting in a more than 75% decrease in hP after 1000 bending cycles. The results of the bending test suggest that the device with the IAA-L electrode is more robust upon bending or has a better electrical path stability, than the device with IAA-H electrode. While we do not have the material details of the two IAA electrodes, it is believed that the metal layers in IAA-H is thinner than in IAA-L, hence the former has a higher sheet resistance and higher transmittance, but less robust under repeated bending stressing. Hence, we choose the PET ﬁlms pre-coated with the lower resistance IAA electrodes for the large-area ﬂexible device fabrication discussed in the next section. 3.3. “Solar leaves” and “solar tree” Palm leaf shaped organic solar cells were fabricated step by step as illustrated in Fig. 5. The PET substrates were ﬁrst trimmed to 200 200 squares and the IAA-L ﬁlms were patterned by conventional photolithography followed by etching in aqua regia solution (Fig. 5(a)). Similar to the fabrication of the small area devices, a ZnO nanoparticle layer and a P3HT:PCBM active layer were spin-coated onto the substrates (Fig. 5(b)). The organic solar cells were completed by thermally evaporating MoO3 and Al as the anode using a shadow mask to form the desired electrode pattern (Fig. 5(c)). Next, the devices were cut into the palm leaf shape using a steel rule die with an alignment guide, leaving a 2 mm edge around the Al electrode pattern. Copper tapes were attached to the two electrodes to form the stems (Fig. 5(d)). The cut-out leaf device was then inserted in a lamination pouch and laminated using a commercial GBC Heatseal H425 laminator (Fig. 5(e)). The solar palm leaves were completed by cutting out the laminated devices
Table 1 Photovoltaic performance parameters for P3HT:PCBM organic solar cells on glass or plastic substrates with different annealing temperatures under 1 sun AM 1.5 solar illumination.
Annealing temp ( C)
Normal Normal Inverted Inverted Inverted Inverted
150 110 150 110 110 110
0.61 0.61 0.63 0.61 0.59 0.59
± ± ± ± ± ±
Jsc (mA/cm2) 0.01 0.01 0.01 0.01 0.01 0.01
8.6 8.0 8.3 8.2 6.3 8.3
± ± ± ± ± ±
0.3 0.2 0.3 0.1 0.1 0.3
FF 0.58 0.50 0.69 0.60 0.61 0.56
± ± ± ± ± ±
0.01 0.02 0.01 0.01 0.01 0.02
3.1 2.4 3.6 3.0 2.3 2.7
± ± ± ± ± ±
0.3 0.2 0.2 0.1 0.1 0.2
W. Cao et al. / Renewable Energy 72 (2014) 134e139
Fig. 3. External quantum efﬁciency (EQE) of normal devices on glass/ITO substrate and inverted device on PET/IAA-L and PET/IAA-H substrates as a function of wavelength, l; Transmittance of ITO electrode on glass and IAA-L and IAA-H electrodes on PET substrates as a function of wavelength, l.
using a second steel rule die (Fig. 5(f)). The encapsulated leaf device has an active area of 6.5 cm2. The currentevoltage (IeV) characteristics of the leaf device under 1 sun AM 1.5G illumination is shown in Fig. 6. Only the center area of the leaf device was illuminated due to equipment limitation, resulting in an illumination area of 5.1 cm2. The leaf device has a Voc ¼ 0.58 ± 0.01 V, just slightly lower than the small area device. It has a short-circuit current (Isc) of 9.6 ± 0.8 mA. The FF of the leaf device is only 0.29 ± 0.03, much lower than that of the small area
Fig. 4. Normalized photovoltaic performance parameters, Voc, Jsc, FF and hP, as a function of the bending cycles N (bending radius r ¼ 1.2 cm) for device on (a) IAA-L and (b) IAA-H substrates.
Fig. 5. Leaf-shape device fabrication procedures: (a) pattern of the transparent conducting electrode (TCE); (b) spin-coat ZnO and P3HT:PCBM active layer; (c) thermally evaporate MoO3 and Al electrode; (d) cut the device into leaf shape (2 mm around the original leaf patter) through a steel rule die with an alignment guide; (e) laminate the device in a lamination pouch; (f) cut the device into the leaf shape for the second time.
devices. This is due to the increased contribution of device series resistance (Rs) in large area solar cell . The series resistance is 32 ± 3 U for the leaf device and 70 ± 5 U for the small area device, which is mostly dominated by the ﬁnite sheet resistance of the TCE (note most of the series resistance of the small area device came from the 2 mm wide and approximately 8 mm long TCE strip from the device active region to the edge of the sample where contacts were made). However, the voltage loss due to the series resistance is proportional to device area A, V ¼ J$A$Rs, where J is the current density, hence the impact of Rs on device performance is more severe for larger area devices. Practically, narrow metal lines on the TCE could be deposited to reduce Rs of the OPV cell to improve FF. Nevertheless, it is important to note that the encapsulated leaf device shows good stability. As shown in Fig. 6, only a ~18% decrease in Isc was observed after storing the device in laboratory ambient for 90 days. The incorporation of the ZnO nanoparticle layer has been previously shown to drastically improve the environmental stability of organic-based solar cells, especially for devices in the inverted geometry [16,29,35]. More sophisticated encapsulation method has been used to demonstrate organic solar cells with operational lifetime of more than 7 years . 12 leaf-shaped devices were further assembled on a trunk to mimic a palm tree, as shown by the photograph in Fig. 7(a). The 12 leaf-shaped devices were wired into four units, each with three
Fig. 6. Current densityevoltage (JeV) characteristics of original leaf-shape device and device kept in air for up to 90 days under simulated 1 sun AM 1.5 solar illumination.
W. Cao et al. / Renewable Energy 72 (2014) 134e139
transparent electrode show good ﬂexibility, with only 12% drop in efﬁciency after 1000 bending cycles with a bending radius of 1.2 cm. Large-area, leaf-shaped devices with the same device structure were also fabricated by lamination, which showed good environmental stability although the photovoltaic performance was limited by the effect of the series resistance. Such leaf-shaped devices were assembled together to build a solar palm tree prototype, which can supply enough power to drive a mini-fan or LEDs under outdoor sunlight. Although the color of those leaf devices based on P3HT:PCBM is not green, it is obvious that other photoactive organic materials can be used to yield the desired colors, especially using specially designed green polymers to mimic natural greeneries .
Fig. 7. (a) A photograph of a solar palm tree prototype with 12 leaf-shape devices. Circuit conﬁguration for prototype demonstration: (b) 4 units (each with 3 cells) of leaf-shaped devices put into parallel to power a mini-fan (“Fan-mode”); (c) 4 units (each with 3 cells) of leaf-shaped devices put into series to power a LED display (“LEDmode”).
Fig. 8. Current densityevoltage (JeV) characteristics of the solar palm tree prototype under the outdoor sunlight (~0.8 sun) for (a) “Fan-mode” and (b) “LED-mode”.
devices connected in series. External switches can switch the four three-cell units between the parallel or series connection to provide high current (but low voltage) or high voltage (but high current), respectively. The IeV characteristics of the solar palm tree prototype measured in outdoor sunny conditions (z0.8 sun intensity, as measured by the silicon reference cell) are shown in Fig. 8. In the parallel conﬁguration (called the “fan-mode”), Voc ¼ 1.70 ± 0.05 V, Isc ¼ 32.5 ± 0.9 mA, and maximum output power Pmax ¼ 13.8 ± 0.9 mW were observed, which can power a mini-fan (see Fig. 7(a)). For the series conﬁguration, called the “LED mode”, we obtained Voc ¼ 5.1 ± 0.3 V, Isc ¼ 9.3 ± 0.8 mA, and Pmax ¼ 12.8 ± 0.8 mW. This is suitable for powering light-emitting device (LED) operation. The slightly lower Pmax in the LED mode is attributed to the series connection among all devices, which has higher series resistance within the whole circuit. 4. Conclusion In summary, we have successfully fabricated organic solar cells on ﬂexible substrates by using an inverted device architecture. With 100 nm thick P3HT:PCBM as the active material, a maximum hP of 2.7 ± 0.3% is observed for the device on ﬂexible PET substrate, compared to that of 3.0 ± 0.1% for the device on rigid glass substrate. The devices on PET substrates with a low resistance IAA
The authors gratefully acknowledge the ﬁnancial support from Sestar Technologies, the Florida Energy System Consortium, and the University of Florida Ofﬁce of Research.
References  Forrest SR. The path to ubiquitous and low-cost organic electronic appliances on plastic. Nature 2004;428:911e8.  Ginley D, Green MA, Collins R. Solar energy conversion toward 1 terawatt. MRS Bull 2008;33:355e64.  Espinosa N, Hosel M, Angmo D, Krebs FC. Solar cells with one-day energy payback for the factories of the future. Energy Environ Sci 2012;5:5117e32.  Liang Y, Yu L. Development of semiconducting polymers for solar energy harvesting. Polym Rev 2010;50:454e73.  Powell C, Bender T, Lawryshyn Y. A model to determine ﬁnancial indicators for organic solar cells. Sol Energy 2009;83:1977e84.  Xue J. Perspectives on organic photovoltaics. Polym Rev 2010;50:411e9.  Zheng Y, Xue J. Organic photovoltaic cells based on molecular donoreacceptor heterojunctions. Polym Rev 2010;50:420e53.  Andersson BV, Wuerfel U, Inganas O. Full day modelling of v-shaped organic solar cell. Sol Energy 2011;85:1257e63.  Cao W, Myers JD, Zheng Y, Hammond WT, Wrzesniewski E, Xue J. Enhancing light harvesting in organic solar cells with pyramidal rear reﬂectors. Appl Phys Lett 2011;99:023306.  Dou L, You J, Yang J, Chen C-C, He Y, Murase S, et al. Tandem polymer solar cells featuring a spectrally matched low-bandgap polymer. Nat Photonics 2012;6:180e5.  Green MA, Emery K, Hishikawa Y, Warta W, Dunlop ED. Solar cell efﬁciency tables (version 42). Prog Photovolt 2013;21:827e37.  Li G, Zhu R, Yang Y. Polymer solar cells. Nat Photonics 2012;6:153e61.  You J, Dou L, Yoshimura K, Kato T, Ohya K, Moriarty T, et al. A polymer tandem solar cell with 10.6% power conversion efﬁciency. Nat Commun 2013;4:1446.  Cao W, Xue J. Recent progress in organic photovoltaics: device architecture and optical design. Energy Environ Sci 2014;7:2123.  Shrotriya V. Polymer power. Nat Photonics 2009;3:447e9.  Qian L, Yang J, Zhou R, Tang A, Zheng Y, Tseng T-K, et al. Hybrid polymer-cdse solar cells with a zno nanoparticle buffer layer for improved efﬁciency and lifetime. J Mater Chem 2011;21:3814e7.  Qian L, Zheng Y, Xue J, Holloway PH. Stable and efﬁcient quantum-dot lightemitting diodes based on solution-processed multilayer structures. Nat Photonics 2011;5:543e8.  Cao W, Zheng Y, Li Z, Wrzesniewski E, Hammond WT, Xue J. Flexible organic solar cells using an oxide/metal/oxide trilayer as transparent electrode. Org Electron 2012;13:2221e8.  White MS, Olson DC, Shaheen SE, Kopidakis N, Ginley DS. Inverted bulkheterojunction organic photovoltaic device using a solution-derived zno underlayer. Appl Phys Lett 2006;89:143517.  Wei W, Zhang C, Chen D, Wang Z, Zhu C, Zhang J, et al. Efﬁcient “lightsoaking”-free inverted organic solar cells with aqueous solution processed low-temperature zno electron extraction layers. ACS Appl Mater Inter 2013;5: 13318e24.  Jouane Y, Colis S, Schmerber G, Dinia A, Leveque P, Heiser T, et al. Inﬂuence of ﬂexible substrates on inverted organic solar cells using sputtered zno as cathode interfacial layer. Org Electron 2013;14:1861e8.  Kyaw AKK, Sun XW, Jiang CY, Lo GQ, Zhao DW, Kwong DL. An inverted organic solar cell employing a solegel derived zno electron selective layer and thermal evaporated moo3 hole selective layer. Appl Phys Lett 2008;93:221107.  Tao C, Ruan S, Zhang X, Xie G, Shen L, Kong X, et al. Performance improvement of inverted polymer solar cells with different top electrodes by introducing a moo(3) buffer layer. Appl Phys Lett 2008;93:193307.
W. Cao et al. / Renewable Energy 72 (2014) 134e139  Myers JD, Cao W, Cassidy V, Eom S-H, Zhou R, Yang L, et al. A universal optical approach to enhancing efﬁciency of organic-based photovoltaic devices. Energy Environ Sci 2012;5:6900e4.  Chen L-M, Hong Z, Li G, Yang Y. Recent progress in polymer solar cells: manipulation of polymer: fullerene morphology and the formation of efﬁcient inverted polymer solar cells. Adv Mater 2009;21:1434e49.  Hoppe H, Sariciftci NS. Morphology of polymer/fullerene bulk heterojunction solar cells. J Mater Chem 2006;16:45e61.  Ma W, Yang C, Gong X, Lee K, Heeger AJ. Thermally stable, efﬁcient polymer solar cells with nanoscale control of the interpenetrating network morphology. Adv Funct Mater 2005;15:1617e22.  Pivrikas A, Neugebauer H, Sariciftci NS. Inﬂuence of processing additives to nano-morphology and efﬁciency of bulk-heterojunction solar cells: a comparative review. Sol Energy 2011;85:1226e37.  Hau SK, Yip H-L, Jen AK-Y. A review on the development of the inverted polymer solar cell architecture. Polym Rev 2010;50:474e510.  Sun Y, Seo JH, Takacs CJ, Seifter J, Heeger AJ. Inverted polymer solar cells integrated with a low-temperature-annealed sol-gel-derived zno ﬁlm as an electron transport layer. Adv Mater 2011;23:1679e83.  Savva A, Petraki F, Elefteriou P, Sygellou L, Voigt M, Giannouli M, et al. The effect of organic and metal oxide interfacial layers on the
performance of inverted organic photovoltaics. Adv Energy Mater 2013;3: 391e8. Docampo P, Ball JM, Darwich M, Eperon GE, Snaith HJ. Efﬁcient organometal trihalide perovskite planar-heterojunction solar cells on ﬂexible polymer substrates. Nat Commun 2013;4:2761. Intemann JJ, Yao K, Li Y-X, Yip H-L, Xu Y-X, Liang P-W, et al. Highly efﬁcient inverted organic solar cells through material and interfacial engineering of indacenodithieno[3,2-b]thiophene-based polymers and devices. Adv Funct Mater 2014;24:1465e73. Xue J, Uchida S, Rand BP, Forrest SR. 4.2% efﬁcient organic photovoltaic cells with low series resistances. Appl Phys Lett 2004;84:3013e5. Hau SK, Yip H-L, Baek NS, Zou J, O'Malley K, Jen AK-Y. Air-stable inverted ﬂexible polymer solar cells using zinc oxide nanoparticles as an electron selective layer. Appl Phys Lett 2008;92:253301. Peters CH, Sachs-Quintana IT, Kastrop JP, Beaupre S, Leclerc M, McGehee MD. High efﬁciency polymer solar cells with long operating lifetimes. Adv Energy Mater 2011;1:491e4. Subbiah J, Beaujuge PM, Choudhury KR, Ellinger S, Reynolds JR, So F. Efﬁcient green solar cells via a chemically polymerizable donor-acceptor heterocyclic pentamer. ACS Appl Mater Interfaces 2009;1:1154e8.