In situ electrochemical doping enhances the efficiency of polymer photovoltaic
devices†
Ming-Shin Su,
aHai-Ching Su,
bChih-Yin Kuo,
aYi-Ren Zhou
aand Kung-Hwa Wei*
a Received 20th October 2010, Accepted 9th February 2011DOI: 10.1039/c0jm03550e
In this study, we found that the formation of a p–i–n junction through in situ electrochemical doping is a promising way to enhance the performance of polymer photovoltaic devices. We applied a pre-bias to metal triflate [LiOTf, KOTf, Ca(OTf)2, Zn(OTf)2]/poly(ethylene oxide) (PEO)–incorporated poly[5-(20-ethylhexyloxy)-2-methoxy-1,4-phenylenevinylene] (MEH-PPV)/{6}-1-(3-(methoxycarbonyl) propyl)-{5}-1-phenyl-[5,6]-C61(PCBM) photovoltaic devices to form p–i–n junctions in their active layers. Auger depth profile analyses and alternating-current capacitance analyses of these doped devices revealed that the positive and negative ions were distributed unequally to form an asymmetrical p–i–n structure in a thin layer of ca. 100 nm of the polymer, and the intrinsic layer became thinner when formed under a higher pre-bias voltage. Atomic force microscopy and transmission electron
microscopy revealed that the addition of metal triflate/PEO to MEH-PPV/PCBM improved the morphology of the composite films. Among the various doped devices, the MEH-PPV/PCBM device incorporating a LiOTf/PEO mixture exhibited the highest power conversion efficiency, a 40% increase relative to that of the reference device (MEH-PPV/PCBM).
Introduction
Organic photovoltaic devices are attracting much attention because of their potential for use as cheap, large-area, flexible devices. Such devices featuring a bulk heterojunction (BHJ) structure exhibit much higher power conversion efficiencies (PCEs) than corresponding non-heterojunction devices because of their larger interfacial areas for exciton dissociation.1–6Much effort has been made to optimize the internal structures of polymer solar cells to increase their lifetimes and PCEs; for example, modifying the work functions of the electrodes7,8 by
incorporating metal oxides such as TiOx, MoO3, WO3and V2O5 between the active layers and electrodes9,10and taking advantage of nano-scale effects, have been carried out.11,12Phase compati-bility and overall uniformity of the film roughness are critically important properties for the optimal performance of polymer solar cells.13,14 Nevertheless, the built-in potential across the
disordered active layer, with a thickness of ca. 100 nm, which is essential for efficient light absorption, leads to a relatively lower built-in electric field and thus, lower carrier collection
efficiency.15–19 These obstacles, which are also encountered in
small-molecule solar cells, can be partially overcome by fabri-cating organic solar cells with a p–i–n architecture.20–23Recently, a small-organic-molecule p–i–n structure was developed by doping small-molecule organic layers to increase the conductivity and, therefore, ensure efficient charge transport and the forma-tion of ohmic contacts at organic–metal interfaces.24As a result,
the built-in potential is predominant and strong across the resistive and thin intrinsic regions, rendering a higher built-in electric field and, consequently, a higher carrier collection efficiency.
Polymer photovoltaic devices can be conveniently fabricated using cost-effective solution processing techniques, such as spin-coating or inkjet printing, making them attractive for large-area applications. Nevertheless, multilayered polymer p–i–n struc-tures are not readily fabricated through solution processing because of interfacial mixing problems. One promising approach toward polymer p–i–n photovoltaic devices is the use of in situ electrochemical doping.25The active layer of such photovoltaic devices contains salts with mobile ions, which can drift towards electrodes under an applied bias. For light-emitting electro-chemical cells (LECs), the spatial separation of ions that induce doping (oxidation and reduction) of the active materials near the electrodes (i.e., p-type doping near the anode and n-type doping near the cathode) has been demonstrated previously.26,27 The p–i–n structure is, therefore, formed in situ under bias in a single-layered thin film, which is readily fabricated through solution processing. Two types of polymer-based chemically fixed p–i–n photovoltaic devices have been prepared using such
aDepartment of Materials Science and Engineering, National Chiao Tung
University, 1001 Ta Hsueh Road, Hsinchu, 30050, Taiwan, ROC. E-mail: [email protected]; Fax: 5724727; Tel: +886-3-5731771
bInstitute of Lighting and Energy Photonics, National Chiao Tung
University, 301 Gaofa 3rd. Road, Tainan, 71150, Taiwan, ROC † Electronic supplementary information (ESI) available: current density–voltage and EQE curves of p–i–n MEH-PPV/PCBM photovoltaic devices, enlarged Auger depth profile consequence. See DOI: 10.1039/c0jm03550e
Materials Chemistry
Cite this: J. Mater. Chem., 2011, 21, 6217
www.rsc.org/materials
PAPER
in situ electrochemical doping.28,29 Both exhibit improved photovoltaic device performance because of the presence of the p–i–n architecture. We believed that the PCEs of polymer p–i–n photovoltaic devices could be further improved for practical applications by optimizing the donor/acceptor systems with appropriate ions. The effect of varying the nature of the salt on the morphologies and electrical properties at different pre-bias voltages of donor/acceptor blend films has not been explored previously.
In this study, we employed electrochemical doping by applying different pre-bias voltages to poly[5-(20 -ethylhexy-loxy)-2-methoxy-1,4-phenylenevinylene] (MEH-PPV)/{6}-1-(3-(methoxycarbonyl)propyl)-{5}-1-phenyl-[5,6]-C61 (PCBM) photovoltaic devices incorporating metal triflate salts [LiOTf, KOTf, Ca(OTf)2, Zn(OTf)2] into poly(ethylene oxide) (PEO) to form ionic complex mixtures. We used atomic force microscopy (AFM) and transmission electron microscopy (TEM) to examine the surface roughness and the morphology of the active layers, respectively. After the p–i–n device architecture had formed under direct current (DC) biasing, the photovoltaic properties of the devices incorporating different metal triflate/PEO mixtures were tested under AM 1.5G conditions. We used capacitance– voltage analysis to deduce the change in the width of the intrinsic layer as the formation of the p–i–n junction experienced the various DC pre-biases. With a relatively unchanged built-in potential that was predominant across a thinner intrinsic layer, the built-in electric field increased upon decreasing the thickness of the intrinsic region, rendering high carrier collection efficien-cies and, thus, enhanced PCEs. We also carried out Auger depth profile analysis on the active layer to elucidate the distribution of positive and negative ions in the active layer that had experienced the pre-biases, to confirm the formation of a p–i–n junction structure.
Experimental
MEH-PPV (Mn: 150 000–250 000), LiOTf (purity: 99.995%), KOTf (purity: 98%), Ca(OTf)2(purity: 99.9%), Zn(OTf)2(purity: 98%), PEO (Mv: 600 000), and chlorobenzene (anhydrous, 99.8%, <0.005% water) were purchased from Sigma–Aldrich. PCBM (purity: 99%) was purchased from Solenne BV. All materials were used as received without further purification.
Indium tin oxide (ITO)–coated glass substrates were pre-cleaned and treated with UV-ozone prior to use. A layer of poly (3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT: PSS; Clevios PVP AI4083, H. C. Starck) was spin-coated (5000 rpm) onto the ITO glass, followed by baking at 150C for 15 min in ambient atmosphere. The PEDOT:PSS–coated ITO glass was then loaded in a N2-filled glove box for subsequent solution processing. The active layer was spin-coated (1500 rpm) onto the PEDOT:PSS–coated ITO glass using an appropriate solution. The thickness of the active layer (ca. 100 nm) was measured using a profilometer. Top contacts of Al (100 nm) were deposited through thermal evaporation at a base pressure below 1 106torr. A MEH-PPV/PCBM (1 : 4, w/w; 20 mg) solution in chlorobenzene (1 mL) was stirred overnight at ca. 60 C in a N2-filled glove box prior to use. Metal triflate (2 mg for LiOTf, 2.4 mg for KOTf, 4.3 mg for Ca(OTf)2and 4.7 mg for Zn(OTf)2)/ PEO (8 mg) solutions in chlorobenzene (1 mL) were stirred at
room temperature (ca. 25C) for several days prior to use. The mole ratio of metal triflate (1.282 105mol)/CH2CH2O unit
(1.82 101mol) solution is 7 104. MEH-PPV/PCBM/metal
triflate/PEO mixtures were prepared by mixing a metal triflate/ PEO solution (0.075 mL) with the MEH-PPV/PCBM solution (1 mL) and stirring overnight at ca. 60C in a N2-filled glove box prior to use.
For photovoltaic analysis, devices were tested under simulated AM 1.5G irradiation (100 mW cm2) using a Xe lamp–based
Newport 66902 150 W solar simulator. A Xe lamp equipped with an AM 1.5G filter was used as the white light source; the optical power at the surface of the sample was 100 mW cm2. The J–V characteristics were measured using a Keithley 2400 source meter. The spectrum of the solar simulator was calibrated using a PV-measurement (PVM-154) mono-Si solar cell (NREL-cali-brated); a Si photodiode (Hamamatsu S1133) was used to check the illumination uniformity of the exposed area.30,31 Reported efficiencies are the averages obtained from four devices prepared on each substrate. External quantum efficiencies (EQEs) were measured using the Optosolar SPF50 spectrum response.32The
morphologies of the films of the active layers, which were fabricated through the same processes used for the devices, were recorded using an atomic force microscope (Digital Instruments Nanoscope IIIa) operated at a scan rate of 1.0 Hz in the tapping mode. The AC capacitance analysis was measured by Keithley 4200. The Auger depth profile analysis was studied by Microlab Auger electron microprobe (VG Scientific, UK). At the Auger depth profile etching process, we used Ar gas as the ion gun to etch the device surface and the etching area was 4 mm2.
Trans-mission electron microscopy (TEM) images were recorded using an FEI Tecnai G2 instrument operated at 120 keV. The TEM samples were first prepared using the same processing steps for preparing the devices; a sandwiched structure consisting of an active layer between two electrodes. Then, this structure was subject to a pre-bias of 3 V for 180 s prior to being immersed into a solution that consisted of 3 vol% detergent (Merck MA-03) in DI water for the electrodes exfoliation. Subsequently, we trans-ferred the active layer that has been floating on the solution to a copper foil and submerged it in DI water for 10 s prior to vacuuming for 6 h before entering the TEM.
Results and discussion
Fig. 1a presents the molecular structures of MEH-PPV and PCBM, which we used as donor and acceptor materials, respectively, for the active layers. The metal triflate salts [LiOTf, KOTf, Ca(OTf)2, Zn(OTf)2] were added in the active layers to provide mobile ions. The presence of an ion-conducting polymer, PEO, served as a solid-state solvent in blend films for undergoing salts dissociation, and also attained the ionic mobility. Fig. 1b provides a schematic representation of the operating principle of using in situ electrochemical doping to form p–i–n junctions in polymer photovoltaic devices. After a bias is applied, the mobile ions in the active layer drift toward the corresponding electrodes. Typically, the resultant n-type doped region would be smaller than that of p-type doped region18,19as indicated in a conducting
atomic force microscopy (AFM) study. Fig. 1b also displays the spatially separated ions inducing doping (oxidation or reduction) of the active materials near the electrodes—p-type doping near
Published on 23 March 2011. Downloaded by National Chiao Tung University on 25/04/2014 02:48:39.
the anode and n-type doping near the cathode. The doped regions induce ohmic contacts between the active layer and the electrodes and, consequently, reduce the voltage drop at the polymer–metal interface—a phenomenon that is beneficial for retaining the open-circuit voltage. As compared to the p- and n-type doped layers, the relatively neutral thin region—the intrinsic layer—exhibits higher resistance and, thus, the built-in potential is predominantly across this layer. Under fixed built-in potential, the built-in electric field in a thinner intrinsic layer can dissociate photogenerated excitons more effectively.25
To explore the phase compatibility, as well as overall unifor-mity, of MEH-PPV/PCBM blend films containing metal triflate/ PEO mixtures, we used AFM to examine the surface morphol-ogies of these films. Fig. 2 presents AFM images of MEH-PPV/ PCBM blend films formed in the absence and presence of metal triflate/PEO mixtures. The MEH-PPV/PCBM films formed in the presence of metal triflate/PEO mixtures exhibited a much smaller root-mean-square (RMS) roughnesses (ca. 0.5–0.6 nm) than those of the MEH-PPV/PCBM films formed in the absence of metal triflate/PEO mixtures, indicating that addition of metal triflate/PEO mixtures into the MEH-PPV/PCBM blend films may have suppressed the phase separation of MEH-PPV and PCBM. Similar phenomena have been observed in MEH-PPV/ C60blend films containing LiOTf/PEO mixtures.33Therefore, we suspected that the morphological manipulation caused by the additives might help to optimize the device performance. The transmission electron microscopy studies on these blend films provides the internal structure and morphology of them.
Fig. 3 displays the TEM images of the MEH-PPV/PCBM and MEH-PPV/PCBM films incorporating LiOTf, KOTf, Ca(OTf)2, or Zn(OTf)2/PEO mixtures with or without applied pre-bias of 3 V for 180 s. In Fig. 3a, the bright and the dark regions can be identified as the polymer rich and the PCBM rich areas, respec-tively, because of the large electron density difference. In Fig. 3 b1–e1, the TEM images of the MEH-PPV/PCBM film incorpo-rating LiOTf, KOTf, Ca(OTf)2, or Zn(OTf)2/PEO mixtures shows much smaller bright and dark domains than that of the MEH-PPV/PCBM film, indicating that the incorporation of the metal triflate/PEO makes the dispersion of MEH-PPV and PCBM more uniform. This compatibilizing effect is possibly due to the interaction of charged ions between MEH-PPV and PCBM. Fig. 3 b2–e2 present the TEM images of MEH-PPV/ PCBM incorporating LiOTf, KOTf, Ca(OTf)2, and Zn(OTf)2/ PEO mixtures with pre-bias of 3 V for 180 s; we can see that the
Fig. 1 a) Device structures and molecular structures of the materials used in this study. b) Operating principle of using in situ electrochemical doping to form p–i–n junctions in polymer photovoltaic devices. The doping area width was a relative viewpoint and not the real width of the ion distribution.
Fig. 2 AFM images of blend films: a) MEH-PPV/PCBM (RMS¼ 2.003 nm), b) MEH-PPV/PCBM with LiOTf/PEO (RMS¼ 0.510 nm), c) MEH-PPV/PCBM with KOTf/PEO (RMS ¼ 0.510 nm), d) MEH-PPV/PCBM with Ca(OTf)2/PEO (RMS ¼ 0.645 nm), and e)
MEH-PPV/PCBM with Zn(OTf)2/PEO (RMS¼ 0.568 nm). Area of each
image: 5mm 5 mm.
morphology changed after experiencing the pre-bias, and the bright and dark domains become larger for any of the four types of p–i–n junction devices, implying that the applied pre-bias would not only alter the ion distribution but also change the active layer’s morphology.
To clarify the mechanism for the formation of the p–i–n junctions through in situ electrochemical doping, we performed alternating current (AC) capacitance analyses34,35of the
photo-voltaic devices. Fig. 4 displays the capacitance of the MEH-PPV/ PCBM device formed in the absence of a metal triflate/PEO mixture, measured under zero bias, as the reference. When incorporating LiOTf/PEO mixtures and applying different pre-biases for 180 s, the devices exhibit enhanced capacitance over the entire frequency range from 0.1 to 500 kHz because the ionic doping induced an increase in the dielectric constant of the active layer. Furthermore, the capacitance of the device increased upon increasing the pre-bias voltage up to 5 V. We observed similar trends for the MEH-PPV/PCBM devices incorporating the KOTf, Ca(OTf)2, and Zn(OTf)2/PEO mixtures. The capacitance of the MEH-PPV/PCBM devices incorporating the LiOTf/PEO mixtures that had experienced pre-biasing at 3 V for 180 s were enhanced by 10–22% relative to that of the device that had not been subjected to pre-biasing, in the frequency range from 0.1 to 500 kHz. These results are reminiscent of those obtained from AC impedance analyses of polymer LECs composed of MEH-PPV and LiOTf/PEO mixtures that had experienced a 3 V bias.34The increase in the capacitance can be attributed to the formation of the p–i–n junction, where the highly conductive positive ion (n-type)-doped region and negative ion (p-type)-doped region function as electrodes for the insulating intrinsic (i) region in between.34 This phenomenon can be quantified; the
capacitance of a parallel plate capacitor with area A, dielectric constant3, and thickness d can be described using equation (1):
C¼ 3A
d (1)
In situ electrochemical doping reduces the thickness (d) of the relatively insulating intrinsic layer and, therefore, enhances the capacitance. A higher pre-bias voltage would result in more accumulated ions in the active layer, thereby broadening the doped regions and leading to a thinner intrinsic layer, rendering a higher capacitance. These results imply an enhanced built-in electric field in the p–i–n MEH-PPV/PCBM photovoltaic devices because the built-in potential is positioned across a thinner intrinsic layer.
In order to directly confirm the formation of the p–i–n junc-tion structure, we performed Auger depth profile analysis on the device consisting of KOTf/PEO incorporated MEH-PPV/PCBM active layer, for detecting the ion distribution. The device structure used in Auger depth profile analysis consists of Al as the top layer, MEH-PPV/PCBM incorporating KOTf/PEO as the active layer and ITO as the bottom layer. The reason that the device studied was without a PEDOT:PSS layer is that we do not want the sulfur energy signal peak contributed to by the PEDOT:PSS layer to interfere with that contributed to by the
Fig. 3 TEM images of the MEH-PPV/PCBM and MEH-PPV/PCBM films incorporating LiOTf, KOTf, Ca(OTf)2, and Zn(OTf)2/PEO
mixtures with or without applied pre-bias of 3 V for 180 s. (a) represents the PPV/PCBM device, and (b1–e1) represent the devices of MEH-PPV/PCBM incorporating LiOTf, KOTf, Ca(OTf)2, and Zn(OTf)2/PEO
mixtures. (b2–e2) represent the devices of MEH-PPV/PCBM incorporating LiOTf, KOTf, Ca(OTf)2, and Zn(OTf)2/PEO mixtures
with a prebias of 3 V for 180 s.
Published on 23 March 2011. Downloaded by National Chiao Tung University on 25/04/2014 02:48:39.
CF3SO3ion. In the Auger depth profile analysis, we identified the potassium (K) and sulfur (S) element energy peak intensities as indicators of the concentration of the positively charged ions (K+) and negatively charged ions (CF
3SO3), respectively, while collecting the energy peak signal of the Al element and In element intensity to represent the cathode (Al) and anode (ITO) concentrations. Fig. 5a showed the Auger depth profile of a device of MEH-PPV/PCBM incorporating the KOTf/PEO mixture: the concentration profile displayed is almost an equal distribution of the positively charged ions, K+, and the negatively
charged ions, CF3SO3, in the active layer, and therefore the active layer maintains electroneutrality for the doped device without applying a pre-bias voltage. In contrast, Fig. 5b presents an asymmetrical concentration profile for the KOTf/PEO doped MEH-PPV/PCBM device after applying a pre-bias: the positively charged ions shift to cathode (Al) and the negatively charged ions moved to anode (ITO). At the point where the concentration of the K and S elements are equivalent, it is defined as an intrinsic region. Moreover, we also observed that the p-type region width is larger than the n-type region and this phenomenon is consis-tent with previous reports.18,19The p-type region and the n-type
region thicknesses can be deduced to be ca. 55 nm and ca. 40 nm, respectively, while the intrinsic region thickness is ca. 5 nm. In the Auger depth profile analysis, we used Ar gas to etch the sample and estimated the etched layer thickness during an etching period (10 s). The etched layer thickness for every etching period was about 5 nm. Fig. S3, ESI,†shows that less than 3% difference between the positive and negative ion distribution was found for one etching period. Therefore, we estimated that the i region thickness is about 5 nm. Hence, we directly confirm the forma-tion of an asymmetrical p–i–n juncforma-tion structure in the KOTf/ PEO doped MEH-PPV/PCBM device after applied bias.
Fig. 6 displays the measured open-circuit voltages (Voc), short-circuit current densities (Jsc), fill factors (FFs), and PCEs of the p–i–n MEH-PPV/PCBM photovoltaic devices incorporating metal triflate/PEO that had been pre-biased at various voltages for 180 s. Fig. 6a reveals that the value of Vocincreased as the applied pre-bias voltages increased, presumably because the decrease in the voltage drops at the polymer–electrode interfaces,
due to ohmic contacts induced by in situ electrochemical doping. Moreover, a common trend in Fig. 6a is that all of the devices pre-biased at 5 V and featuring metal triflate/PEO mixtures exhibited increased values of Vocthat approached ca. 0.8 V, the ideal value of Vocfor MEH-PPV/PCBM systems, as estimated from the energy level offset between the lowest unoccupied molecular orbital (LUMO) of PCBM and the highest occupied molecular orbital (HOMO) of MEH-PPV.36–38 Because the
accumulation of ions would be accelerated under a higher bias,39
the resulting greater accumulation of ions would further reduce the potential drop in voltage, across the active layer and the cathode; the decreases in the voltage at the polymer–electrode interface would enhance the Vocof the device. Fig. 6b reveals that the changes in the Jscof the p–i–n photovoltaic devices occurred in two stages: initial increases upon increasing the pre-bias voltage, reaching a maximum at a pre-bias voltage of 3 V, and then decreases upon increasing the pre-bias voltage beyond that value. This phenomenon results from two counteracting factors: (i) the strength of the built-in electric field for exciton dissociation and (ii) the intrinsic layer thickness for generating majority excitons. In the first stage (pre-bias: 1–3 V), the improvement in
Fig. 4 AC capacitance analysis, under zero DC biasing, at frequencies of 0.1–500 kHz, for p–i–n photovoltaic devices incorporating the LiOTf/ PEO mixture that had been subjected to various pre-bias voltages.
Fig. 5 Auger depth profile of the MEH-PPV/PCBM devices, incorpo-rating KOTf/PEO with and without a pre-bias of 3 V for 180 s, forming a p–i–n junction. (a) shows the Auger depth profile of the device without a pre-bias of 3 V for 180 s, and (b) shows the Auger depth profile of the device with a pre-bias of 3 V for 180 s, forming p–i–n junction.
the exciton dissociation efficiency, due to the enhanced built-in electric field, was more dominant than the decrease in exciton generation, due to the thinner intrinsic layer, relative to that in the control device (pre-bias: 0 V), leading to a higher value of Jsc. When the devices experienced a higher pre-bias (4–5 V), the weighing factors were reversed: much fewer excitons were generated for a much thinner intrinsic layer, despite the larger built-in electric field, leading to a lower value of Jsc. Notably, the values of Jsc for devices pre-biased at 3 V for 180 s increased significantly (up to >1.5 times) relative to those of devices that had not been subjected to pre-biasing. The pre-biased devices doped with metal triflate/PEO mixtures all exhibited much higher current densities than those of pure MEH-PPV/PCBM devices, due to the presence of p–i–n junctions induced by in situ elec-trochemical doping.25,28,29Among the devices, those doped with
LiOTf/PEO featured the most significant enhancements in their current densities, presumably because the Li+ cation, with its
smallest ionic radius and lightest ionic mass, compared with those of K+, Ca2+and Zn2+, resulted in the development of
well-formed p–i–n junctions within the limited pre-biasing time. Consequently, the improved exciton dissociation efficiency induced by the enhanced built-in electric field in the intrinsic layer increased the PCEs of the LiOTf/PEO-doped devices up to 2.07%. Fig. 6c presents the FFs of the p–i–n photovoltaic devices; we observe the same phenomenon as that for the values of Jsc, with an initial increase up to a maximum value at 3 V followed by a decrease upon increasing the pre-bias voltage. We attribute the initial increase in the FF to the decreased series resistance of the active layer after in situ electrochemical doping. Further increasing the pre-bias voltage resulted in a reduced thickness of the intrinsic layer, thereby facilitating doping-induced micro shorts40and, thereby lowering the shunt resistance of the active layer, leading to a lower FF. Fig. 6d presents a plot of the PCEs of the photovoltaic devices with respect to the pre-bias voltage; again, the trend follows those of the plots of Jsc and FF. Among the devices doped with the LiOTf, KOTf, Ca(OTf)2, and Zn(OTf)2/PEO mixtures, the MEH-PPV/PCBM device incor-porating the LiOTf/PEO mixture exhibited the highest PCE (2.07%), 40% greater than that of the control device (1.47%) that was fabricated under state of art conditions different to those of the doped devices. The MEH-PPV/PCBM device was fabricated under the same processing conditions as those of the doped devices and showed a PCE of only 1.26%. Again, we attribute this phenomenon to the Li+ions having the smallest ion radius
and lightest ion mass; therefore, they would exhibit the best arrangement condition under the same electrical field, leading to well formed p–i–n structures. Our results indicate that in situ electrochemical doping is a promising technique for enhancing the PCEs of polymer photovoltaic devices.
I–V curves and EQE spectra for the MEH-PPV/PCBM photovoltaic devices prepared with and without LiOTf, KOTf, Ca(OTf)2, and Zn(OTf)2/PEO mixtures that had been pre-biased at 3 V for 180 s are shown in the ESI.†
Table 1 summarizes the photovoltaic characteristics of different devices which experienced a pre-bias of 3 V for 180 s. The enhancement in the EQE was consistent with the increased values of Jsc which were induced by the improved exciton dissociation efficiency in the presence of an increased built-in electric field. Although the nature of the cation had a relatively
Fig. 6 Plots of a) Voc, b) Jsc, c) FF, and d) PCE of p–i–n photovoltaic
devices prepared with LiOTf (square), KOTf (circle), Ca(OTf)2(triangle),
and Zn(OTf)2 (inverted triangle)/PEO mixtures pre-biased at various
voltages for 180 s.
Published on 23 March 2011. Downloaded by National Chiao Tung University on 25/04/2014 02:48:39.
slight effect on the values of Vocfor the pre-biased metal triflate/ PEO–incorporated MEH-PPV/PCBM photovoltaic devices, it had a large effect on the values of Jsc.
Our MEH-PPV/PCBM devices doped with metal triflate/PEO mixtures exhibited shorter response times (180 s), relative to those (1 to 6 h) of ionic liquid–doped PPV derivatives,42 for
forming their p–i–n junctions under the applied pre-bias volt-ages. The anions and cations, however, would relax back when the applied bias voltage was removed.41–44 To examine the stability of the p–i–n junction structures within the LiOTF/PEO– doped MEH-PPV/PCBM devices that had experienced pre-biasing at 3 V for 180 s, we performed endurance tests by exposing the devices under simulated AM 1.5G solar radiation at an intensity of 100 mW cm2 in a N2 glove box for 5 h of continuous operation.
Table 2 lists the percentage decreases in the values of Voc, Jsc, FF, and PCE of the devices prepared with and without LiOTf/ PEO mixtures after 5 h continuous illumination. The PCE of the LiOTf/PEO–doped MEH-PPV/PCBM device decreased to ca. 55% of its initial value, while that of the MEH-PPV/PCBM device decreased to ca. 70% to its initial value. With subsequent re-biasing at 3 V for 180 s, the PCE of the LiOTf/PEO–doped device recovered to ca. 68% of its initial value. We observed similar phenomena for the values of Voc, Jsc, and FF for the re-biased LiOTf/PEO–doped devices (Table 2). Therefore, it appears that the larger decreases in Voc, Jsc, FF, and PCE for the LiOTf/PEO–doped devices, relative to those of the correspond-ing pure MEH-PPV/PCBM devices, after 5 h of continuous operation can be attributed to the relaxation of the p–i–n junc-tions, a phenomenon commonly observed in LECs.26,27
Improvements of the relaxation times of p–i–n structure within devices have been demonstrated using ‘‘frozen’’ p–i–n junctions42,45,46and ‘‘covalent bonding.’’47
Conclusions
We have applied a pre-bias to metal triflate/PEO–incorporated MEH-PPV/PCBM photovoltaic devices to form p–i–n junctions in the active layers. AC capacitance analyses of the device indi-cated the presence of p–i–n junctions and revealed that the intrinsic layer was thinner when formed under higher pre-bias voltages. Auger depth profile analysis revealed the distribution of positive and negative ions and confirmed the formation of an asymmetrical p–i–n junction structure in a thin layer of ca. 100 nm. AFM imaging revealed that addition of metal triflate/ PEO to MEH-PPV/PCBM improved the morphology of the composite films. Among the devices doped with the metal triflate/ PEO mixtures, the MEH-PPV/PCBM device incorporating the LiOTf/PEO mixture exhibited the highest PCE, an increase of 40% relative to that of the control device. Thus, the formation of p–i–n junctions through in situ electrochemical doping is a promising approach toward enhancing the performance of polymer photovoltaic devices.
References
1 Changduk Yang, Jae Kwan Lee, Alan J. Heeger and Fred Wudl, J. Mater. Chem., 2009, 19, 5416.
2 M. Lenes, F. B. Kooistra, J. C. Hummelen, I. V. Severen, L. Lutsen, D. Vanderzande, T. J. Cleij and P. W. M. Blom, J. Appl. Phys., 2008, 104, 114517.
Table 1 Photovoltaic characteristics of p–i–n MEH-PPV/PCBM photovoltaic devices prepared with and without metal triflate/PEO mixtures pre-biased at 3 V for 180 s
Components of active layer Voc(V) Jsc(mA cm2) FF (%) PCE (%) Enhancement a (%) MEH-PPV/PCBMb(1 : 4, w/w) 0.72 4.67 42.8 1.47 — MEH-PPV/PCBM: PEOb 0.79 4.07 44.1 1.43 — MEH-PPV/PCBM: LiOTf/PEO 0.76 6.37 42.9 2.07 40 MEH-PPV/PCBM: KOTf/PEO 0.72 6.14 41.2 1.82 24 MEH-PPV/PCBM: Ca(OTf)2/PEO 0.70 6.00 42.8 1.80 22
MEH-PPV/PCBM: Zn(OTf)2/PEO 0.70 5.48 42.4 1.62 10 a
Relative to the PCE of the pure MEH-PPV/PCBM (1 : 4, w/w) device.bThe PCE of the MEH-PPV/PCBM and MEH-PPV/PCBM:PEO devices were fabricated under state of the art conditions. The PCE of the MEH-PPV/PCBM device was fabricated at the same conditions as the doped devices was 1.26%.
Table 2 Endurance testadata of p–i–n MEH-PPV/PCBM photovoltaic devices prepared with and without LiOTf/PEO mixtures pre-biased at 3 V for 180 s
Components
of active layer (endurance test time)
Normalized Voc(%) Normalized Jsc(%) Normalized FF (%) Normalized PCE (%) MEH-PPV:PCBMb(0 h) 100 100 100 100 MEH-PPV:PCBM (5 h) 83.14 92.53 89.85 69.12 MEH-PPV/PCBM: LiOTf/PEO (0 h) 100 100 100 100 MEH-PPV/PCBM: LiOTf/PEO (5 h) 77.14 85.08 83.14 54.56 Re-biasedcMEH-PPV/PCBM: LiOTf/PEO (5 h)
85.22 94.3 84.72 68.48
aPhotovoltaic data was collected under simulated AM 1.5G solar radiation (100 mW cm2) in a N
2-filled glove box for 5 h of continuous operation. b
When the MEH-PPV/PCBM devices were also subjected to pre-biasing at 3 V for 180 s, the device performance was unaffected.cThe device was re-biased at 3 V for 180 s after 5 h of continuous operation.
3 T. Kirchartz, J. Mattheis and U. Rau, Phys. Rev. B: Condens. Matter Mater. Phys., 2008, 78, 235320.
4 J. Nelson, J. J. Kwiatkowski, J. Kirkpartrick and J. M. Frost, Acc. Chem. Res., 2009, 42, 1768.
5 C. Groves, J. C. Blakesley and N. C. Greenham, Nano Lett., 2010, 10, 1063.
6 M. C. Scharber, D. M€uhlbacher, M. Koppe, P. Denk, C. Waldauf, A. J. Heeger and C. J. Brabec, Adv. Mater., 2006, 18, 789.
7 H. L. Yip, S. K. Hau, N. S. Baek, H. Ma and A. K. Y. Jen, Adv. Mater., 2008, 20, 2376.
8 V. D. Mihailetchi, P. W. M. Blom, J. C. Hummelen and M. T. Rispens, J. Appl. Phys., 2003, 94, 6849.
9 L. M. Chen, Z. Xu, Z. Hong and Y. Yang, J. Mater. Chem., 2010, 20, 2575.
10 K. StevenHau, Hin-Lap Yip, Orb Acton, Nam Seob Baek, Hong Ma and Alex K.-Y. Jen, J. Mater. Chem., 2008, 18, 5113.
11 H. S. Wang, L. H. Lin, S. Y. Chen, Y. L. Wang and K. H. Wei, Nanotechnology, 2009, 20, 075201.
12 P. Yu, C. H. Chang, M. S. Su, M. H. Hsu and K. H. Wei, Appl. Phys. Lett., 2010, 96, 153307.
13 M. Y. Chiu, U. S. Jeng, C. H. Su, K. S. Liang and K. H. Wei, Adv. Mater., 2008, 20, 2573.
14 M. Y. Chiu, U. S. Jeng, M. S. Su and K. H. Wei, Macromolecules, 2010, 43, 428.
15 A. Liu, S. Zhao, S. B. Rim, J. Wu, M. Konemann, P. Erk and P. Peumans, Adv. Mater., 2008, 20, 1065.
16 B. A. Greeg, Soft Matter, 2009, 5, 2985.
17 J. M. Leger, S. A. Carter and B. Ruhstaller, J. Appl. Phys., 2005, 98, 124907.
18 L. S. C. Pingree, D. B. Rodovsky, D. C. Coffey, G. P. Bartholomew and D. S. Ginger, J. Am. Chem. Soc., 2007, 129, 15903.
19 Juhyun Park, Corey V. Hoven, Renqiang Yang, Namsung Cho, Hongbin Wu, Thuc-Quyen Nguyen and Guillermo C. Bazan, J. Mater. Chem., 2009, 19, 211.
20 F. Maddalena, E. J. Meijer, K. Asadi, D. M. de Leeuw and P. W. M. Blom, Appl. Phys. Lett., 2010, 97, 043302.
21 J. Drechsel, B. M€annig, F. Kozlowski, D. Gebeyehu, A. Werner, M. Koch, K. Leo and M. Pfeiffer, Thin Solid Films, 2004, 451, 515. 22 B. M€annig, J. Drechsel, D. Gebeyehu, P. Simon, F. Kozlowski,
A. Werner, F. Li, S. Grundmann, S. Sonntag, M. Koch, K. Leo, M. Pfeiffer, H. Hoppe, D. Meissner, N. S. Sariciftci, I. Reidel, V. Dyakonov and J. Parisi, Appl. Phys. A: Mater. Sci. Process., 2004, 79, 1.
23 C. Falkenberg, C. Uhrich, S. Olthof, B. Maennig, M. K. Riede and K. Leo, J. Appl. Phys., 2008, 104, 034506.
24 W. Y. Gao and A. Kahn, Org. Electron., 2002, 3, 53. 25 J. Gao, G. Yu and A. J. Heeger, Adv. Mater., 1998, 10, 692. 26 Q. Pei, G. Yu, C. Zhang, Y. Yang and A. J. Heeger, Science, 1995,
269, 1086.
27 Q. Pei, Y. Yang, G. Yu, C. Zhang and A. J. Heeger, J. Am. Chem. Soc., 1996, 118, 3922.
28 J. M. Leger, D. G. Patel, D. B. Rodovsky and G. P. Bartholomew, Adv. Funct. Mater., 2008, 18, 1212.
29 Z. Yu, M. Sun and Q. Pei, J. Phys. Chem. B, 2009, 113, 8481. 30 V. Shrotriya, G. Li, Y. Yao, T. Moriarty, K. Emery and Y. Yang,
Adv. Funct. Mater., 2006, 16, 2016.
31 Y. T. Chang, S. L. Hsu, M. H. Su and K. H. Wei, Adv. Mater., 2009, 21, 2093.
32 K. Emery, Sol. Cells, 1986, 18, 301.
33 F. C. Chen, Q. Xu and Y. Yang, Appl. Phys. Lett., 2004, 84, 3181. 34 Y. Li, J. Gao, G. Yu, Y. Cao and A. J. Heeger, Chem. Phys. Lett.,
1998, 287, 83.
35 G. Yu, Y. Cao, C. Zhang, Y. Li, J. Gao and A. J. Heeger, Appl. Phys. Lett., 1998, 73, 111.
36 Fred Wudl, J. Mater. Chem., 2002, 12, 1959.
37 B. C. Thompson and J. M. J. Frechet, Angew. Chem., Int. Ed., 2008, 47, 58.
38 L. J. A. Koster, E. C. P. Smits, V. D. Mihailetchi and P. W. M. Blom, Phys. Rev. B: Condens. Matter Mater. Phys., 2005, 72, 085205. 39 P. Matyba, K. Maturova, M. Kemerink, N. D. Robinson and
L. Edman, Nat. Mater., 2009, 8, 672.
40 J. H. Shin, S. Xiao and L. Edman, Adv. Funct. Mater., 2006, 16, 949.
41 L. Edman, M. Pauchard, D. Moses and A. J. Heeger, J. Appl. Phys., 2004, 95, 4357.
42 Y. Shao, G. C. Bazan and A. J. Heeger, Adv. Mater., 2007, 19, 365. 43 C. Hoven, R. Yang, A. Garcia, A. J. Heeger, T. Q. Nguyen and
G. C. Bazan, J. Am. Chem. Soc., 2007, 129, 10976.
44 F. P. Wenzl, P. Pachler, C. Suess, A. Haase, E. J. W. List, P. Poelt, D. Somitsch, P. Knoll, U. Scherf and G. Leising, Adv. Funct. Mater., 2004, 14, 441.
45 J. Gao, G. Yu and A. J. Heeger, Appl. Phys. Lett., 1997, 71, 1293. 46 G. Yu, Y. Cao, M. Anderson, J. Gao and A. J. Heeger, Adv. Mater.,
1998, 10, 385.
47 C. V. Hoven, H. Wang, M. Elbing, L. Garner, D. Winkelhaus and G. C. Bazan, Nat. Mater., 2010, 9, 249.
Published on 23 March 2011. Downloaded by National Chiao Tung University on 25/04/2014 02:48:39.
