DOI: 10.1002/cphc.201402762
Articles
Universal Efficiency Improvement in Organic Solar Cells Based on a Poly(3-hexylthiophene) Donor and an IndeneC60 Bisadduct Acceptor with Additional Donor Nanowires Sung-yoon Joe, Jong Hyuk Yim, Shin Young Ryu, Na Young Ha, Yeong Hwan Ahn, Ji-Yong Park, and Soonil Lee*[a] addition of P3HT NWs did not result in enhanced internal quantum efficiencies for either type of device. However, the difference in light-harvesting efficiency was important in accounting for Jsc variations. Interestingly, there was no correlation between Jsc and PCE variations, whereas the open-circuit voltage (Voc) and fill factor (FF) showed correlations with the PCE. The variation in FF is discussed in terms of Voc and equivalent-circuit parameters based on a nonideal diode model.
With poly(3-hexylthiophene) (P3HT) nanowire (NW) inclusion in active layers (ALs), organic solar cells (OSCs) based on P3HT donor and indene-C60 bisadduct (ICBA) acceptor showed power conversion efficiency (PCE) improvements for both bulk heterojunction (BHJ)- and bilayer (BL)-structure AL devices. The PCE increase was approximately 14 % for both types of P3HT:ICBA OSCs. However, improvements in short-circuit current density (Jsc) were about 4.4 and 6.4 % for BHJ- and BL-type AL devices, respectively. A systematic study showed that the
1. Introduction Generally, active layers (ALs) of organic solar cells (OSCs), in which light absorption and electron–hole generation occur, consist of a donor–acceptor pair. The most common examples are poly(3-hexylthiophene) (P3HT) donor and fullerene derivative [6,6]-phenyl-C61-butyric acid methyl ester (PCBM) acceptor combinations. To make efficient OSCs, it is essential to maximize the donor–acceptor junction area while maintaining continuous hole- and electron-transport paths in donor and acceptor phases, respectively. An AL structure that has been most frequently used to satisfy such requirements is the bulk heterojunction (BHJ).[1–3] In the case of P3HT:PCBM OSCs, many thermal and solvent annealing processes have been developed to optimize junction formation and morphology of a donor– acceptor mixture. Recently, utilization of P3HT nanowires (NWs) has emerged as a novel method for the modification of donor–acceptor morphology in ALs. Crystalline P3HT NWs can be formed by slowly cooling P3HT solutions that are prepared by dissolving P3HT in low-solubility solvents, such as xylene, toluene, cyclohexane, and methylene chloride, at high temperatures.[4–6] As the temperature of the solutions is lowered, P3HT self-assembles into a one-dimensional (1D) crystalline structure due to p–p stacking. The unique optical and electrical properties of 1D P3HT NWs have been proven effective in increasing the ef-
ficiencies of P3HT:PCBM OSCs. For example, J.-H. Kim et al. reported a power conversion efficiency (PCE) of 4.07 % from P3HT:PCBM OSCs with P3HT NWs formed from a P3HT:PCBM solution in chlorobenzene/cyclohexane mixture.[7] On the other hand, J. S. Kim et al. fabricated OSCs with a double AL structure of P3HT/P3HT:PCBM and reported a PCE of 3.94 %. They formed P3HT NWs by dissolving a thin P3HT layer in dichloromethane.[8] Similarly, Berson et al. fabricated P3HT:PCBM OSCs with an AL that consisted of both nanofiber- and disorganizedP3HT phases. After optimization of the ratio between nanofiber and disorganized phases, they obtained a PCE of 3.6 %.[9] Alternatively, OSCs can be fabricated by using ALs that have a bilayer (BL) structure because acceptor molecules such as PCBM can diffuse into a P3HT layer with appropriate processing. Chen et al. fabricated P3HT:PCBM OSCs by annealing a BL of amorphous P3HT and PCBM at 150 8C, and reported a PCE of 2.24 %.[10] In the case of Gevaerts et al., they first formed a neat BL by sequentially spin-coating P3HT and PCBM dissolved in orthogonal solvents, and then used thermal annealing to induce BHJ-like morphology. The PCE of their device was 2.38 %.[11] Another variation in OSC fabrication is using different donor–acceptor pairs. For example, low-bandgap donors have been studied to extend solar photon absorption to an infrared spectral range. Replacing PCBM with another acceptor material is interesting as well, because there is potential to improve PCE without replacing familiar widely available P3HT with more expensive new donor materials. Indene-C60 bisadduct (ICBA) is one such candidate. The advantage of ICBA is its high LUMO (lowest unoccupied molecular orbital) position compared to PCBM.[12, 13] Indeed, P3HT:ICBA OSCs with higher open-circuit voltage and, consequently, higher PCE were fabricated.[14, 15]
[a] S.-y. Joe, Dr. J. H. Yim, S. Y. Ryu, Prof. N. Y. Ha, Prof. Y. H. Ahn, Prof. J.-Y. Park, Prof. S. Lee Department of Physics and Division of Energy Systems Research Ajou University, Suwon 443-749 (South Korea) E-mail: [email protected] Supporting Information for this article is available on the WWW under http://dx.doi.org/10.1002/cphc.201402762. An invited contribution to a Special Issue on Organic Electronics
ChemPhysChem 0000, 00, 0 – 0
These are not the final page numbers! ÞÞ
1
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles In this work we studied the effects of P3HT-NW inclusion in enhancing the performance of OSCs. By using ICBA as acceptor and incorporating P3HT NWs into both BHJ- and BL-structure ALs, we were able to gain some insights into the changes in device operation induced by P3HT-NW inclusion that resulted in PCE variations. Both the application of P3HT NWs in BL OSCs and the combination of P3HT NWs with ICBA were not reported previously.
2. Results and Discussion Figure 1 shows a general schematic of the device architecture based on a bottom indium tin oxide (ITO) anode and the corresponding energy level diagram for P3HT:ICBA solar cells (SCs). While keeping other parts of the devices fixed, we systematically varied the structure and morphology of organic
Figure 2. J–V characteristics of the four types of OSCs at 1 sun under AM 1.5G illumination. Symbols represent measured data, and solid lines are the best-fit curves. For both BHJ- and BL-type ALs based on P3HT donor and ICBA acceptor, the inclusion of P3HT NWs resulted in efficiency improvements.
The BHJ device had a short-circuit current (Jsc) of 8.22 mA cm2, an open-circuit voltage (Voc) of 0.761 V, and fill factor (FF) of 52.3 % under AM 1.5G solar illumination at 1 sun, which resulted in a PCE of 3.27 %. Such a modest PCE was due to the use of LiF/Al as a cathode. We note that higher PCE values (4.84– 7.50 %) of P3HT:ICBA OSCs in the literature were measured from devices having a Ca/Al cathode and/or an additional cathode or anode buffer layer.[16–22] We were able to fabricate P3HT:ICBA BHJ OSCs with PCEs higher than 5 % by simply substituting Ca/Ag for LiF/Al as well (Figure S3 and Table S1). The inclusion of P3HT NWs in a BHJ AL produced a larger PCE of 3.73 %. Such an improvement in PCE resulted from the combination of respective increases in Jsc (8.58 mA cm2), Voc (0.791 V), and FF (55.0 %). On the other hand, the BL device had a PCE of 3.19 % (corresponding to a Jsc of 8.40 mA cm2, a Voc of 0.740 V, and a FF of 51.3 %), which was slightly smaller than that of the BHJ device. It is typical for BL devices to have lower PCEs than their BHJ counterparts. Interestingly, the PCE difference between BHJ and BL devices based on a P3HT–ICBA combination was noticeably smaller than that on the more common P3HT–PCBM combination. Similar to the BHJ case, the addition of P3HT NWs in a BL AL resulted in an improved PCE of 3.63 %, corresponding to a Jsc of 8.94 mA cm2, a Voc of 0.777 V, and a FF of 52.3 %. It is common to assess the per-
Figure 1. a) Schematic device architecture of P3HT:ICBA SCs with a bottom ITO anode and a top LiF/Al cathode. b) Energy level diagram of the materials used in device fabrication. PEDOT = poly(3,4-ethylenedioxythiophene), PSS = poly(styrene sulfonate), w/o = without, w/ = with.
ALs that consisted of P3HT donor and ICBA acceptor. Either a BHJ- or a BL-structure AL with or without inclusion of preformed P3HT NWs was used to fabricate four different types of P3HT:ICBA SCs that were accordingly named BHJ, BHJ-NW, BL, and BL-NW devices. The diameter of P3HT NWs that we made and used to prepare ALs was 25 5 nm (see the Supporting Information, Figure S1). It was interesting to find that the respective root-mean-square (RMS) roughness values estimated from AFM images were 5.6, 6.1, 6.1, and 5.8 nm for BHJ, BHJ-NW, BL, and BL-NW ALs, respectively, and that there was no noticeable surface morphology difference among these ALs (Figure S2). Figure 2 shows the current density versus voltage (J–V) characteristics of the four types of P3HT:ICBA SCs. These J–V curves were measured from the best devices of each type, and corresponding photovoltaic parameters are summarized in Table 1.
Table 1. Solar cell parameters corresponding to four types of ALs based on the combination of P3HT donor and ICBA acceptor with or without the inclusion of P3HT NWs. Voc, Jsc, FF, PCE, Rs, Rsh, and n represent open-circuit voltage, short-circuit current, fill factor, power conversion efficiency, series resistance, shunt resistance, and ideality factor, respectively. These parameter values that were extracted from the J–V curves in Figure 2 represent the best performance of each device type. The average photovoltaic parameter values estimated from six devices of each type show a similar trend (Table S2).
&
Sample ID
P3HT NWs
Voc [V]
Jsc [mA cm2]
FF [%]
n
Rs [W cm2]
Rsh [W cm2]
PCE [%]
BHJ BHJ-NW BL BL-NW
no yes no yes
0.761 0.791 0.740 0.777
8.22 8.58 8.40 8.94
52.3 55.0 51.3 52.3
3.78 0.01 2.17 0.03 2.52 0.03 3.38 0.02
6.37 0.02 14.9 0.02 15.8 0.07 4.48 0.02
451 3 486 7 443 6 347 2
3.27 3.73 3.19 3.63
ChemPhysChem 0000, 00, 0 – 0
www.chemphyschem.org
2
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
Articles formance of OSCs by using an equivalent-circuit model based on a nonideal diode with parasitic resistances [Eq. (1)]:[23–25]
crease in Jsc, as confirmed by the estimated values from the following integral equation [Eq. (2)]:
h qðVJR Þ i V JR s JðV Þ ¼ J0 e nk T 1 þ JL Rsh
Jsc ¼
s
B
ð1Þ
To elucidate the PCE improvements of the BHJ-NW and BLNW devices compared to those without P3HT-NW inclusion, external quantum efficiency (EQE) spectra were measured. Interestingly, the EQE spectra of the two BHJ-type devices were distinctly different from those of the BL-type devices, as shown in Figure 3 a. Qualitatively, the shapes of EQE spectra of the two BHJ-type devices were similar regardless of the P3HT-NW inclusion. However, there was a quantitative difference between these two EQE curves. Specifically, the EQEs of the BHJNW device were slightly larger than those of the BHJ device in most of the visible-light spectral range. Such a small but wide spectral range difference was sufficient to produce a 4.4 % in-
Jsc ¼
These are not the final page numbers! ÞÞ
ð2Þ
Z
l F ðlÞLHE ðlÞIQE ðlÞdl 1240 sol
ð3Þ
Naturally, IQE, the ratio between the number of charge carriers and that of photons absorbed by an AL, depends on all electrical/electronic processes. On the contrary, the number of solar photons actually absorbed by an AL, which corresponds to the rest of the integrand, is determined by combined optical effects. In this work, we first approximated LHE(l) as a product of transmittance Tw(l) corresponding to a window structure of glass/ITO/PEDOT:PSS and absorbance AAL(l) of an AL, and then used it to deduce IQE. Both Tw(l) and AAL(l) spectra were measured from extra samples prepared exclusively for optical measurements. In particular, transmittance TAL(l) and reflectance RAL(l) were measured and used to estimate the absorbance of ALs as AAL(l) = 1TAL(l)RAL(l). Figure 4 shows LHE spectra of the four types of ALs. Interestingly, all the LHE spectra had noticeable spectral variations in the wavelength range of 350 to 650 nm. Specifically, LHEs were
Figure 3. a) Comparison of EQE spectra of the four types of P3HT:ICBA OSCs. EQE spectra of devices having identical AL structures, either BHJ or BL, were qualitatively similar regardless of P3HT-NW inclusion. However, there was a noticeable difference in EQE spectra between two sets of devices with different AL structures. The integration of EQE spectra and solar irradiance from 300 to 800 nm produced Jsc values that matched measured ones closely. b) DJsc(l) represents accumulated contributions to Jsc by solar photons up to wavelength l, which was estimated by using Equation (2).
www.chemphyschem.org
l F ðlÞEQE ðlÞdl 1240 sol
in which l is the wavelength in nanometers and Fsol is the solar irradiance in W m2 nm1. The values of Jsc estimated based on the measured EQE spectra were 8.19 and 8.56 mA cm2 for the BHJ and BHJ-NW devices, respectively. The good agreement between measured and estimated values of Jsc was a strong confirmation of the accurate EQE measurements. The resemblance between EQE spectra of the BL-type devices was not as strong as that of BHJ-type devices. Both in the ultraviolet and visible spectral ranges, there were noticeable discrepancies between the BL and BL-NW devices. We note that the larger Jsc of the BL-NW device by 6.4 % was well accounted for by Equation (2). Specifically, estimated Jsc values were 8.37 and 8.91 mA cm2 for the BL and BL-NW devices, respectively. A slight EQE reduction of the BL-NW device near 400 nm compared to BHJ-type devices had no significant effect on Jsc. On the contrary, the EQE reduction of the BL device in the short-wavelength range was significant enough to make the integrated DJsc value smaller than that of the BHJNW device, as shown in Figure 3 b. The difference between integrated DJsc values diminished in the mid-spectral range because of the larger EQE of the BL device, but with another crossing of the EQE spectra at 565 nm, the Jsc of the BL-NW device eventually became smaller than that of the BHJ-NW device. It was convenient to represent EQEs as the product of lightharvesting efficiency (LHE) and internal quantum efficiency (IQE) to separate the changes induced by P3HT-NW inclusion into contributions of optical and electrical/electronic origin. In terms of LHE and IQE, the integral equation for Jsc in Equation (2) can be rewritten as [Eq. (3)]:
in which JL is photocurrent and J0 reverse saturation current, Rs and Rsh respectively are series and shunt resistances, and n is an ideality factor. The values of equivalent-circuit parameters, deduced by fitting the J–V curves in Figure 2 to Equation (1), are listed in Table 1 together with the values of Jsc, Voc, FF, and PCE. We note that FF variations with respect to photovoltaic and equivalent-circuit parameters can be estimated quantitatively by using the analytical expressions derived by Green[26] (Table S3).
ChemPhysChem 0000, 00, 0 – 0
Z
3
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
Articles were not severely limited by electrical and/or electronic processes. Additionally, a reverse correlation between IQE and Jsc indicated that a slight change in LHE was more influential to result in the Jsc increase. Presumably, the inclusion of P3HT NWs induced changes in a planar-diffused structure/morphology of ALs and, consequently, their absorption profile. Formation of planar-diffused ALs seemed to be an extra advantage of P3HT– ICBA devices. A similar reverse correlation between IQE and Jsc was observed from the BHJ-type devices. In spite of a lower IQE, the BHJ-NW device had a larger Jsc because it had a higher LHE at most wavelengths relative to the other one. On the contrary, a significantly smaller IQE ( 0.76), together with a slightly smaller LHE at long wavelengths, made the Jsc of the BHJ-NW device smaller than that of the BL-NW device. One interesting observation was that there was no strong correlation between the variation in Jsc and that of PCE, and/or those of other SC and equivalent-circuit parameters. As shown in Table 1, the BL-NW and BHJ devices showed, respectively, the largest and smallest Jsc, whereas the BHJ-NW and BL devices showed the largest and smallest PCE, respectively. On the other hand, there was a clearer correlation between the PCE variation and the Voc and FF variations. The FF is known to increase monotonically with respect to Voc in an ideal diode model, but decreases as a function of Rs and Rsh in a nonideal diode model. Therefore, the largest FF of the BHJ-NW device that had the largest Voc and Rsh was not surprising. However, the FF of the BL-NW device was smaller than that of the BHJNW device because of the smaller Voc and Rsh. Accordingly, the BL-NW device’s PCE was smaller than that of the BHJ-NW device in spite of it having the largest Jsc. In the case of the BL device, the smallest Voc and the largest Rs made its FF and, consequently, PCE the lowest. We attribute the smallest Voc of the BL device to unfavorable charge recombination effects. Arguably, charge recombination was not significant in the BL-NW device because holes migrated fast along P3HT NWs from the junction of donor and acceptor layers. On the contrary, hole transport was less efficient without P3HT NWs, and more recombination occurred near a diffused junction area in the BL device.
Figure 4. Comparison of LHE spectra of the four types of P3HT:ICBA OSCs. LHE(l) was approximated as a product of transmittance Tw(l) through a window structure of glass/ITO/PEDOT:PSS and AL absorbance AAL(l) that is deduced from transmittance and reflectance of an AL: LHE(l) = T(l) AAL(l) = T(l) [1TAL(l)RAL(l)].
larger in the order of BHJ-NW, BL, BL-NW, and BHJ devices in short-wavelength ranges. However, the order changed to BLNW, BHJ, BHJ-NW, and BL devices in long-wavelength ranges. Such variations suggested that the IQEs of these devices were neither identical nor constant. IQE spectra of the four types of ALs shown in Figure 5 were deduced from the EQE and LHE spectra as IQE(l) = EQE(l)/LHE(l).[27, 28] IQE spectra were generally flat in the wavelength range of 450–550 nm. However, there were some variations in shorteror longer-wavelength ranges. Mid-wavelength IQE values of the BHJ-NW, BHJ, BL-NW, and BL devices were approximately 0.76, 60.79, 0.91, and 0.94, respectively. Surprisingly, both the BHJ-NW and BL-NW devices had slightly lower IQE values than their respective counterparts. Furthermore, the IQE values of the BL-type devices were higher than those of the BHJ-type devices. Both of these observations were counterintuitive. In particular, IQE values larger than 0.9 indicated that planar-diffused ALs were formed, and Jsc values of the BL-type devices
3. Conclusions We have found that adding P3HT NWs to ALs was a viable option to improve the PCEs of SCs based on the P3HT–ICBA donor–acceptor pair. The improvements in PCE were about 14 % for both the BHJ- and BL-type devices. It was particularly interesting to note that the 3.63 % PCE of the BL-NW device was higher than that of the BHJ device. We argued that such a high PCE was possible because of the P3HT–ICBA pair’s unique advantage of forming a planar-diffuse interface. Apparently, Jsc improved as a result of P3HT-NW inclusion for both the BHJ- and BL-type devices. Systematic analysis to manifest the origin of the Jsc increase showed that there were no IQE improvements. On the contrary, the difference in LHE appeared to be consistent with Jsc variations. Such discrepancy in optical effects was indicative of different distributions of absorbing
Figure 5. Comparison of IQE spectra of the four types of P3HT:ICBA OSCs. IQE(l) was deduced from EQE(l) and LHE(l): IQE(l) = EQE(l)/LHE(l). In the mid-wavelength range of 450–550 nm, IQE spectra were generally flat with values of approximately 0.76, 0.79, 0.91, and 0.94 for the BHJ-NW, BHJ, BLNW, and BL devices, respectively.
&
ChemPhysChem 0000, 00, 0 – 0
www.chemphyschem.org
4
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
Articles was spin-coated at 3000 rpm for 10 s and heat-treated at 150 8C for 30 min. The DA-layer thicknesses of the BL and BL-NW devices were 160 and 165 nm, respectively.
P3HT within inhomogeneous ALs, which warrants further studies. One interesting observation was that there was no correlation between Jsc and PCE variations. However, PCE showed correlation with Voc and FF. Analysis based on an equivalent-circuit model that consisted of nonideal diode and parasitic resistances allowed us to understand PCE variations qualitatively. The BL device had the lowest PCE because of the largest Rs and the smallest Voc that, presumably, was due to adverse recombination effects with no provision to effectively transport holes from the diffuse junction region.
The cell size, defined by the overlap of cathode and anode, was 0.24 cm2. All devices were measured under AM 1.5G illumination at 1 sun by using a solar simulator (PEC-L01, Peccell) that was calibrated to 100 mW cm2 with a reference Si cell (PEC-S101, Peccell). The EQEs of the fabricated OSCs were measured by using a photoelectrochemistry incident photon-to-current efficiency (IPCE) system (PE-IPCE, HS Technologies). Transmittance, reflectance, and absorbance spectra were measured by using a spectrophotometer (Cary 5000, Varian).
Experimental Section Acknowledgements
We fabricated OSCs with a structure of ITO/PEDOT:PSS/donor–acceptor (DA)/LiF/Al on ITO-coated glasses (12 W sq.1, GEOMATEC). PEDOT:PSS (Clevios P, Heraeus) were spin-coated onto a thoroughly cleaned ITO anode and baked for 10 min at 140 8C to form 30 nm thick layers. Organic DA layers were formed by spin-coating mixtures of P3HT donor and ICBA acceptor. ICBA (SOL5064A) and regioregular P3HT (4000-E) were purchased from Solaris Chem and Reike Metals, respectively. LiF and Al layers were evaporated to form an upper cathode. In this work, we fabricated four types of OSCs by using either a BHJ or a BL DA-layer structure, and additionally, by changing DA-layer morphologies with the inclusion of preformed P3HT NWs. A name convention of BHJ, BHJ-NW, BL, and BL-NW was used to identify the four types of OSCs according to their DA-layer structures and morphologies.
This work was supported by the National Research Foundation of Korea (NRF) (Grant No. NRF-535 2014R1A2A2A01005632), which is funded by the Korea government (MEST), and through the Priority Research Centers Program (NRF-2009-0094049). Keywords: donor–acceptor systems · energy conversion · nanowires · organic solar cells · quantum efficiency [1] G. Li, V. Shrotriya, J. Huang, Y. Yao, T. Moriarty, K. Emery, Y. Yang, Nat. Mater. 2005, 4, 864. [2] W. Ma, C. Yang, X. Gong, K. Lee, A. J. Heeger, Adv. Funct. Mater. 2005, 15, 1617. [3] M. C. Quiles, T. Ferenczi, T. Agostinelli, P. G. Etchegoin, Y. Kim, T. D. Anthopoulos, P. N. Stavrinou, D. C. Bradley, J. Nelson, Nat. Mater. 2008, 7, 158. [4] J. Liu, M. Arif, J. Zou, S. I. Khondaker, L. Zhai, Macromolecules 2009, 42, 9390. [5] S. Samitsu, T. Shimomura, S. Heike, T. Hashizume, K. Ito, Macromolecules 2008, 41, 8000. [6] D. H. Kim, J. T. Han, Y. D. Park, Y. Jang, J. H. Cho, M. Hwang, K. Cho, Adv. Mater. 2006, 18, 719. [7] J.-H. Kim, J. H. Park, J. H. Lee, J. S. Kim, M. Sim, C. Shim, K. Cho, J. Mater. Chem. 2010, 20, 7398. [8] J. S. Kim, J. H. Lee, J. H. Park, C. Shim, M. Sim, K. Cho, Adv. Funct. Mater. 2011, 21, 480. [9] S. Berson, R. De Bettignies, S. Bailly, S. Guillerez, Adv. Funct. Mater. 2007, 17, 1377. [10] D. Chen, F. Liu, C. Wang, A. Nakahara, T. P. Russell, Nano Lett. 2011, 11, 2071. [11] V. S. Gevaerts, L. J. Koster, M. M. Wienk, R. A. Janssen, ACS Appl. Mater. Interfaces 2011, 3, 3252. [12] Y. He, H.-Y. Chen, J. Hou, Y. Li, J. Am. Chem. Soc. 2010, 132, 1377. [13] D. W. Laird, R. Stegamat, H. Richter, V. Vejins, L. Scott, T. A. Lada, Patent WO 2008/018931A2. [14] G. Zhao, Y. He, Y. Li, Adv. Mater. 2010, 22, 4355. [15] F. Cheng, G. Fang, X. Fan, H. Huang, Q. Zheng, P. Qin, H. Lei, Y. Li, Sol. Energy Mater. Sol. Cells 2013, 110, 63. [16] Y.-J. Cheng, C.-H. Hsieh, Y. J. He, C.-S. Hsu, Y. F. Li, J. Am. Chem. Soc. 2010, 132, 17381. [17] C.-Y. Chang, C.-E. Wu, S.-Y. Chen, C. H. Cui, Y.-J. Cheng, C.-S. Hsu, Y.-L. Wang, Y. F. Li, Angew. Chem. Int. Ed. 2011, 50, 9386; Angew. Chem. 2011, 123, 9558. [18] S.-H. Liao, Y.-L. Li, T.-H. Jen, Y.-S. Cheng, S.-A. Chen, J. Am. Chem. Soc. 2012, 134, 14271. [19] Z. A. Tan, L. J. Li, C. H. Cui, Y. Q. Ding, Q. Xu, S. S. Li, D. P. Qian, Y. F. Li, J. Phys. Chem. C 2012, 116, 18626. [20] Z. A. Tan, W. Q. Zhang, C. H. Cui, Y. Q. Ding, D. P. Qian, Q. Xu, L. J. Li, S. S. Li, Y. F. Li, Phys. Chem. Chem. Phys. 2012, 14, 14589. [21] Z. A. Tan, D. P. Qian, W. Q. Zhang, L. J. Li, Y. Q. Ding, Q. Xu, F. Z. Wang, Y. F. Li, J. Mater. Chem. A 2013, 1, 657.
The DA layers of BHJ devices were spin-coated from a 1:1 mixture of P3HT and ICBA in dichlorobenzene (DCB), the total concentration of which was 34 mg mL1. In the case of BHJ-NW devices, DA layers were spin-coated from a P3HT NW-added mixture solution that was prepared in two steps. Initially, a P3HT solution was made by thoroughly mixing a P3HT solution (24 mg mL1) in DCB with the same volume of a solution (4 mg mL1) of preformed P3HT NWs in dichloromethane (CH2Cl2) under continuous stirring at room temperature. The total concentration of P3HT in the DCB/ CH2Cl2 mixture was 14 mg mL1, and 1/7 of the P3HT was NWs. Next, ICBA (17 mg) was added to this NW-containing P3HT solution to prepare a mixture at the total concentration of 31 mg mL1. In both device types, DA layers were formed by spin-coating the aforementioned P3HT/ICBA mixture solutions at 800 rpm for 30 s, which were followed with a 40 min waiting period for solvents to evaporate. Finally, a heat treatment was carried out at 140 8C for 10 min on a hot plate. The DA-layer thicknesses of the BHJ and BHJ-NW devices were 190 and 220 nm, respectively. Contrary to the BHJ and BHJ-NW cases, the BL and BL-NW DA layers were formed by sequential spin-coating of donor and acceptor solutions. A donor solution for the BL device was prepared by dissolving P3HT (20 mg) in DCB (1 mL), and that for the BL-NW device by 1:1 mixing of a P3HT solution (24 mg mL1) with a solution (4 mg mL1) of preformed P3HT NWs in CH2Cl2. The NW-containing donor solution was stirred over 24 h at room temperature before spin-coating. The single acceptor solution prepared by dissolving ICBA (10 mg) in CH2Cl2 at 60 8C under continuous stirring over 24 h was used for both the BL and BL-NW devices. The coating of the donor part, spin-coating of a donor solution at 800 rpm for 30 s and 10 min of heat treatment at 140 8C on a hot plate, was straightforward. However, the subsequent coating of the acceptor part required great care not to damage the donor part. In particular, the selection of orthogonal solvent for the donor solution, such as CH2Cl2 for ICBA, was of utmost importance. The ICBA solution ChemPhysChem 0000, 00, 0 – 0
www.chemphyschem.org
These are not the final page numbers! ÞÞ
5
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
Articles [27] S. H. Park, A. Roy, S. Beaupr, S. Cho, N. Coates, J. S. Moon, D. Moses, M. Leclerc, K. Lee, A. J. Heeger, Nat. Photonics 2009, 3, 297. [28] J. Burschka, N. Pellet, S.-J. Moon, R. Humphry-Baker, P. Gao, M. K. Nazeeruddin, M. Grtzel, Nature 2013, 499, 316.
[22] Z. A. Tan, W. Q. Zhang, D. P. Qian, C. H. Cui, Q. Xu, L. J. Li, S. S. Li, Y. F. Li, Phys. Chem. Chem. Phys. 2012, 14, 14217. [23] B. Kippelen, J.-L. Brdas, Energy Environ. Sci. 2009, 2, 251. [24] J. D. Servaites, S. Yeganeh, T. J. Marks, M. A. Ratner, Adv. Funct. Mater. 2010, 20, 97. [25] J. H. Yim, S.-Y. Joe, C. Pang, K. M. Lee, H. Jeong, J.-Y. Park, Y. H. Ahn, J. C. de Mello, S. Lee, ACS Nano 2014, 8, 2857. [26] M. A. Green, Solar Cells 1982, 7, 337.
&
ChemPhysChem 0000, 00, 0 – 0
www.chemphyschem.org
Received: October 27, 2014 Revised: January 21, 2015 Published online on && &&, 2015
6
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
ARTICLES Energy efficiency: Preformed poly(3hexylthiophene) (P3HT) nanowires (NWs) added to active layers of organic solar cells based on the P3HT–indeneC60 bisadduct (ICBA) donor–acceptor pair result in improved power conversion efficiencies for both bulk heterojunction- and bilayer-type devices. The differences in the light-harvesting efficiencies account for variations in the short-circuit current density.
ChemPhysChem 0000, 00, 0 – 0
www.chemphyschem.org
These are not the final page numbers! ÞÞ
S.-y. Joe, J. H. Yim, S. Y. Ryu, N. Y. Ha, Y. H. Ahn, J.-Y. Park, S. Lee* && – && Universal Efficiency Improvement in Organic Solar Cells Based on a Poly(3-hexylthiophene) Donor and an Indene-C60 Bisadduct Acceptor with Additional Donor Nanowires
7
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
Articles
Universal Efficiency Improvement in Organic Solar Cells Based on a Poly(3-hexylthiophene) Donor and an IndeneC60 Bisadduct Acceptor with Additional Donor Nanowires Sung-yoon Joe, Jong Hyuk Yim, Shin Young Ryu, Na Young Ha, Yeong Hwan Ahn, Ji-Yong Park, and Soonil Lee*[a] addition of P3HT NWs did not result in enhanced internal quantum efficiencies for either type of device. However, the difference in light-harvesting efficiency was important in accounting for Jsc variations. Interestingly, there was no correlation between Jsc and PCE variations, whereas the open-circuit voltage (Voc) and fill factor (FF) showed correlations with the PCE. The variation in FF is discussed in terms of Voc and equivalent-circuit parameters based on a nonideal diode model.
With poly(3-hexylthiophene) (P3HT) nanowire (NW) inclusion in active layers (ALs), organic solar cells (OSCs) based on P3HT donor and indene-C60 bisadduct (ICBA) acceptor showed power conversion efficiency (PCE) improvements for both bulk heterojunction (BHJ)- and bilayer (BL)-structure AL devices. The PCE increase was approximately 14 % for both types of P3HT:ICBA OSCs. However, improvements in short-circuit current density (Jsc) were about 4.4 and 6.4 % for BHJ- and BL-type AL devices, respectively. A systematic study showed that the
1. Introduction Generally, active layers (ALs) of organic solar cells (OSCs), in which light absorption and electron–hole generation occur, consist of a donor–acceptor pair. The most common examples are poly(3-hexylthiophene) (P3HT) donor and fullerene derivative [6,6]-phenyl-C61-butyric acid methyl ester (PCBM) acceptor combinations. To make efficient OSCs, it is essential to maximize the donor–acceptor junction area while maintaining continuous hole- and electron-transport paths in donor and acceptor phases, respectively. An AL structure that has been most frequently used to satisfy such requirements is the bulk heterojunction (BHJ).[1–3] In the case of P3HT:PCBM OSCs, many thermal and solvent annealing processes have been developed to optimize junction formation and morphology of a donor– acceptor mixture. Recently, utilization of P3HT nanowires (NWs) has emerged as a novel method for the modification of donor–acceptor morphology in ALs. Crystalline P3HT NWs can be formed by slowly cooling P3HT solutions that are prepared by dissolving P3HT in low-solubility solvents, such as xylene, toluene, cyclohexane, and methylene chloride, at high temperatures.[4–6] As the temperature of the solutions is lowered, P3HT self-assembles into a one-dimensional (1D) crystalline structure due to p–p stacking. The unique optical and electrical properties of 1D P3HT NWs have been proven effective in increasing the ef-
ficiencies of P3HT:PCBM OSCs. For example, J.-H. Kim et al. reported a power conversion efficiency (PCE) of 4.07 % from P3HT:PCBM OSCs with P3HT NWs formed from a P3HT:PCBM solution in chlorobenzene/cyclohexane mixture.[7] On the other hand, J. S. Kim et al. fabricated OSCs with a double AL structure of P3HT/P3HT:PCBM and reported a PCE of 3.94 %. They formed P3HT NWs by dissolving a thin P3HT layer in dichloromethane.[8] Similarly, Berson et al. fabricated P3HT:PCBM OSCs with an AL that consisted of both nanofiber- and disorganizedP3HT phases. After optimization of the ratio between nanofiber and disorganized phases, they obtained a PCE of 3.6 %.[9] Alternatively, OSCs can be fabricated by using ALs that have a bilayer (BL) structure because acceptor molecules such as PCBM can diffuse into a P3HT layer with appropriate processing. Chen et al. fabricated P3HT:PCBM OSCs by annealing a BL of amorphous P3HT and PCBM at 150 8C, and reported a PCE of 2.24 %.[10] In the case of Gevaerts et al., they first formed a neat BL by sequentially spin-coating P3HT and PCBM dissolved in orthogonal solvents, and then used thermal annealing to induce BHJ-like morphology. The PCE of their device was 2.38 %.[11] Another variation in OSC fabrication is using different donor–acceptor pairs. For example, low-bandgap donors have been studied to extend solar photon absorption to an infrared spectral range. Replacing PCBM with another acceptor material is interesting as well, because there is potential to improve PCE without replacing familiar widely available P3HT with more expensive new donor materials. Indene-C60 bisadduct (ICBA) is one such candidate. The advantage of ICBA is its high LUMO (lowest unoccupied molecular orbital) position compared to PCBM.[12, 13] Indeed, P3HT:ICBA OSCs with higher open-circuit voltage and, consequently, higher PCE were fabricated.[14, 15]
[a] S.-y. Joe, Dr. J. H. Yim, S. Y. Ryu, Prof. N. Y. Ha, Prof. Y. H. Ahn, Prof. J.-Y. Park, Prof. S. Lee Department of Physics and Division of Energy Systems Research Ajou University, Suwon 443-749 (South Korea) E-mail: [email protected] Supporting Information for this article is available on the WWW under http://dx.doi.org/10.1002/cphc.201402762. An invited contribution to a Special Issue on Organic Electronics
ChemPhysChem 0000, 00, 0 – 0
These are not the final page numbers! ÞÞ
1
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles In this work we studied the effects of P3HT-NW inclusion in enhancing the performance of OSCs. By using ICBA as acceptor and incorporating P3HT NWs into both BHJ- and BL-structure ALs, we were able to gain some insights into the changes in device operation induced by P3HT-NW inclusion that resulted in PCE variations. Both the application of P3HT NWs in BL OSCs and the combination of P3HT NWs with ICBA were not reported previously.
2. Results and Discussion Figure 1 shows a general schematic of the device architecture based on a bottom indium tin oxide (ITO) anode and the corresponding energy level diagram for P3HT:ICBA solar cells (SCs). While keeping other parts of the devices fixed, we systematically varied the structure and morphology of organic
Figure 2. J–V characteristics of the four types of OSCs at 1 sun under AM 1.5G illumination. Symbols represent measured data, and solid lines are the best-fit curves. For both BHJ- and BL-type ALs based on P3HT donor and ICBA acceptor, the inclusion of P3HT NWs resulted in efficiency improvements.
The BHJ device had a short-circuit current (Jsc) of 8.22 mA cm2, an open-circuit voltage (Voc) of 0.761 V, and fill factor (FF) of 52.3 % under AM 1.5G solar illumination at 1 sun, which resulted in a PCE of 3.27 %. Such a modest PCE was due to the use of LiF/Al as a cathode. We note that higher PCE values (4.84– 7.50 %) of P3HT:ICBA OSCs in the literature were measured from devices having a Ca/Al cathode and/or an additional cathode or anode buffer layer.[16–22] We were able to fabricate P3HT:ICBA BHJ OSCs with PCEs higher than 5 % by simply substituting Ca/Ag for LiF/Al as well (Figure S3 and Table S1). The inclusion of P3HT NWs in a BHJ AL produced a larger PCE of 3.73 %. Such an improvement in PCE resulted from the combination of respective increases in Jsc (8.58 mA cm2), Voc (0.791 V), and FF (55.0 %). On the other hand, the BL device had a PCE of 3.19 % (corresponding to a Jsc of 8.40 mA cm2, a Voc of 0.740 V, and a FF of 51.3 %), which was slightly smaller than that of the BHJ device. It is typical for BL devices to have lower PCEs than their BHJ counterparts. Interestingly, the PCE difference between BHJ and BL devices based on a P3HT–ICBA combination was noticeably smaller than that on the more common P3HT–PCBM combination. Similar to the BHJ case, the addition of P3HT NWs in a BL AL resulted in an improved PCE of 3.63 %, corresponding to a Jsc of 8.94 mA cm2, a Voc of 0.777 V, and a FF of 52.3 %. It is common to assess the per-
Figure 1. a) Schematic device architecture of P3HT:ICBA SCs with a bottom ITO anode and a top LiF/Al cathode. b) Energy level diagram of the materials used in device fabrication. PEDOT = poly(3,4-ethylenedioxythiophene), PSS = poly(styrene sulfonate), w/o = without, w/ = with.
ALs that consisted of P3HT donor and ICBA acceptor. Either a BHJ- or a BL-structure AL with or without inclusion of preformed P3HT NWs was used to fabricate four different types of P3HT:ICBA SCs that were accordingly named BHJ, BHJ-NW, BL, and BL-NW devices. The diameter of P3HT NWs that we made and used to prepare ALs was 25 5 nm (see the Supporting Information, Figure S1). It was interesting to find that the respective root-mean-square (RMS) roughness values estimated from AFM images were 5.6, 6.1, 6.1, and 5.8 nm for BHJ, BHJ-NW, BL, and BL-NW ALs, respectively, and that there was no noticeable surface morphology difference among these ALs (Figure S2). Figure 2 shows the current density versus voltage (J–V) characteristics of the four types of P3HT:ICBA SCs. These J–V curves were measured from the best devices of each type, and corresponding photovoltaic parameters are summarized in Table 1.
Table 1. Solar cell parameters corresponding to four types of ALs based on the combination of P3HT donor and ICBA acceptor with or without the inclusion of P3HT NWs. Voc, Jsc, FF, PCE, Rs, Rsh, and n represent open-circuit voltage, short-circuit current, fill factor, power conversion efficiency, series resistance, shunt resistance, and ideality factor, respectively. These parameter values that were extracted from the J–V curves in Figure 2 represent the best performance of each device type. The average photovoltaic parameter values estimated from six devices of each type show a similar trend (Table S2).
&
Sample ID
P3HT NWs
Voc [V]
Jsc [mA cm2]
FF [%]
n
Rs [W cm2]
Rsh [W cm2]
PCE [%]
BHJ BHJ-NW BL BL-NW
no yes no yes
0.761 0.791 0.740 0.777
8.22 8.58 8.40 8.94
52.3 55.0 51.3 52.3
3.78 0.01 2.17 0.03 2.52 0.03 3.38 0.02
6.37 0.02 14.9 0.02 15.8 0.07 4.48 0.02
451 3 486 7 443 6 347 2
3.27 3.73 3.19 3.63
ChemPhysChem 0000, 00, 0 – 0
www.chemphyschem.org
2
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
Articles formance of OSCs by using an equivalent-circuit model based on a nonideal diode with parasitic resistances [Eq. (1)]:[23–25]
crease in Jsc, as confirmed by the estimated values from the following integral equation [Eq. (2)]:
h qðVJR Þ i V JR s JðV Þ ¼ J0 e nk T 1 þ JL Rsh
Jsc ¼
s
B
ð1Þ
To elucidate the PCE improvements of the BHJ-NW and BLNW devices compared to those without P3HT-NW inclusion, external quantum efficiency (EQE) spectra were measured. Interestingly, the EQE spectra of the two BHJ-type devices were distinctly different from those of the BL-type devices, as shown in Figure 3 a. Qualitatively, the shapes of EQE spectra of the two BHJ-type devices were similar regardless of the P3HT-NW inclusion. However, there was a quantitative difference between these two EQE curves. Specifically, the EQEs of the BHJNW device were slightly larger than those of the BHJ device in most of the visible-light spectral range. Such a small but wide spectral range difference was sufficient to produce a 4.4 % in-
Jsc ¼
These are not the final page numbers! ÞÞ
ð2Þ
Z
l F ðlÞLHE ðlÞIQE ðlÞdl 1240 sol
ð3Þ
Naturally, IQE, the ratio between the number of charge carriers and that of photons absorbed by an AL, depends on all electrical/electronic processes. On the contrary, the number of solar photons actually absorbed by an AL, which corresponds to the rest of the integrand, is determined by combined optical effects. In this work, we first approximated LHE(l) as a product of transmittance Tw(l) corresponding to a window structure of glass/ITO/PEDOT:PSS and absorbance AAL(l) of an AL, and then used it to deduce IQE. Both Tw(l) and AAL(l) spectra were measured from extra samples prepared exclusively for optical measurements. In particular, transmittance TAL(l) and reflectance RAL(l) were measured and used to estimate the absorbance of ALs as AAL(l) = 1TAL(l)RAL(l). Figure 4 shows LHE spectra of the four types of ALs. Interestingly, all the LHE spectra had noticeable spectral variations in the wavelength range of 350 to 650 nm. Specifically, LHEs were
Figure 3. a) Comparison of EQE spectra of the four types of P3HT:ICBA OSCs. EQE spectra of devices having identical AL structures, either BHJ or BL, were qualitatively similar regardless of P3HT-NW inclusion. However, there was a noticeable difference in EQE spectra between two sets of devices with different AL structures. The integration of EQE spectra and solar irradiance from 300 to 800 nm produced Jsc values that matched measured ones closely. b) DJsc(l) represents accumulated contributions to Jsc by solar photons up to wavelength l, which was estimated by using Equation (2).
www.chemphyschem.org
l F ðlÞEQE ðlÞdl 1240 sol
in which l is the wavelength in nanometers and Fsol is the solar irradiance in W m2 nm1. The values of Jsc estimated based on the measured EQE spectra were 8.19 and 8.56 mA cm2 for the BHJ and BHJ-NW devices, respectively. The good agreement between measured and estimated values of Jsc was a strong confirmation of the accurate EQE measurements. The resemblance between EQE spectra of the BL-type devices was not as strong as that of BHJ-type devices. Both in the ultraviolet and visible spectral ranges, there were noticeable discrepancies between the BL and BL-NW devices. We note that the larger Jsc of the BL-NW device by 6.4 % was well accounted for by Equation (2). Specifically, estimated Jsc values were 8.37 and 8.91 mA cm2 for the BL and BL-NW devices, respectively. A slight EQE reduction of the BL-NW device near 400 nm compared to BHJ-type devices had no significant effect on Jsc. On the contrary, the EQE reduction of the BL device in the short-wavelength range was significant enough to make the integrated DJsc value smaller than that of the BHJNW device, as shown in Figure 3 b. The difference between integrated DJsc values diminished in the mid-spectral range because of the larger EQE of the BL device, but with another crossing of the EQE spectra at 565 nm, the Jsc of the BL-NW device eventually became smaller than that of the BHJ-NW device. It was convenient to represent EQEs as the product of lightharvesting efficiency (LHE) and internal quantum efficiency (IQE) to separate the changes induced by P3HT-NW inclusion into contributions of optical and electrical/electronic origin. In terms of LHE and IQE, the integral equation for Jsc in Equation (2) can be rewritten as [Eq. (3)]:
in which JL is photocurrent and J0 reverse saturation current, Rs and Rsh respectively are series and shunt resistances, and n is an ideality factor. The values of equivalent-circuit parameters, deduced by fitting the J–V curves in Figure 2 to Equation (1), are listed in Table 1 together with the values of Jsc, Voc, FF, and PCE. We note that FF variations with respect to photovoltaic and equivalent-circuit parameters can be estimated quantitatively by using the analytical expressions derived by Green[26] (Table S3).
ChemPhysChem 0000, 00, 0 – 0
Z
3
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
Articles were not severely limited by electrical and/or electronic processes. Additionally, a reverse correlation between IQE and Jsc indicated that a slight change in LHE was more influential to result in the Jsc increase. Presumably, the inclusion of P3HT NWs induced changes in a planar-diffused structure/morphology of ALs and, consequently, their absorption profile. Formation of planar-diffused ALs seemed to be an extra advantage of P3HT– ICBA devices. A similar reverse correlation between IQE and Jsc was observed from the BHJ-type devices. In spite of a lower IQE, the BHJ-NW device had a larger Jsc because it had a higher LHE at most wavelengths relative to the other one. On the contrary, a significantly smaller IQE ( 0.76), together with a slightly smaller LHE at long wavelengths, made the Jsc of the BHJ-NW device smaller than that of the BL-NW device. One interesting observation was that there was no strong correlation between the variation in Jsc and that of PCE, and/or those of other SC and equivalent-circuit parameters. As shown in Table 1, the BL-NW and BHJ devices showed, respectively, the largest and smallest Jsc, whereas the BHJ-NW and BL devices showed the largest and smallest PCE, respectively. On the other hand, there was a clearer correlation between the PCE variation and the Voc and FF variations. The FF is known to increase monotonically with respect to Voc in an ideal diode model, but decreases as a function of Rs and Rsh in a nonideal diode model. Therefore, the largest FF of the BHJ-NW device that had the largest Voc and Rsh was not surprising. However, the FF of the BL-NW device was smaller than that of the BHJNW device because of the smaller Voc and Rsh. Accordingly, the BL-NW device’s PCE was smaller than that of the BHJ-NW device in spite of it having the largest Jsc. In the case of the BL device, the smallest Voc and the largest Rs made its FF and, consequently, PCE the lowest. We attribute the smallest Voc of the BL device to unfavorable charge recombination effects. Arguably, charge recombination was not significant in the BL-NW device because holes migrated fast along P3HT NWs from the junction of donor and acceptor layers. On the contrary, hole transport was less efficient without P3HT NWs, and more recombination occurred near a diffused junction area in the BL device.
Figure 4. Comparison of LHE spectra of the four types of P3HT:ICBA OSCs. LHE(l) was approximated as a product of transmittance Tw(l) through a window structure of glass/ITO/PEDOT:PSS and AL absorbance AAL(l) that is deduced from transmittance and reflectance of an AL: LHE(l) = T(l) AAL(l) = T(l) [1TAL(l)RAL(l)].
larger in the order of BHJ-NW, BL, BL-NW, and BHJ devices in short-wavelength ranges. However, the order changed to BLNW, BHJ, BHJ-NW, and BL devices in long-wavelength ranges. Such variations suggested that the IQEs of these devices were neither identical nor constant. IQE spectra of the four types of ALs shown in Figure 5 were deduced from the EQE and LHE spectra as IQE(l) = EQE(l)/LHE(l).[27, 28] IQE spectra were generally flat in the wavelength range of 450–550 nm. However, there were some variations in shorteror longer-wavelength ranges. Mid-wavelength IQE values of the BHJ-NW, BHJ, BL-NW, and BL devices were approximately 0.76, 60.79, 0.91, and 0.94, respectively. Surprisingly, both the BHJ-NW and BL-NW devices had slightly lower IQE values than their respective counterparts. Furthermore, the IQE values of the BL-type devices were higher than those of the BHJ-type devices. Both of these observations were counterintuitive. In particular, IQE values larger than 0.9 indicated that planar-diffused ALs were formed, and Jsc values of the BL-type devices
3. Conclusions We have found that adding P3HT NWs to ALs was a viable option to improve the PCEs of SCs based on the P3HT–ICBA donor–acceptor pair. The improvements in PCE were about 14 % for both the BHJ- and BL-type devices. It was particularly interesting to note that the 3.63 % PCE of the BL-NW device was higher than that of the BHJ device. We argued that such a high PCE was possible because of the P3HT–ICBA pair’s unique advantage of forming a planar-diffuse interface. Apparently, Jsc improved as a result of P3HT-NW inclusion for both the BHJ- and BL-type devices. Systematic analysis to manifest the origin of the Jsc increase showed that there were no IQE improvements. On the contrary, the difference in LHE appeared to be consistent with Jsc variations. Such discrepancy in optical effects was indicative of different distributions of absorbing
Figure 5. Comparison of IQE spectra of the four types of P3HT:ICBA OSCs. IQE(l) was deduced from EQE(l) and LHE(l): IQE(l) = EQE(l)/LHE(l). In the mid-wavelength range of 450–550 nm, IQE spectra were generally flat with values of approximately 0.76, 0.79, 0.91, and 0.94 for the BHJ-NW, BHJ, BLNW, and BL devices, respectively.
&
ChemPhysChem 0000, 00, 0 – 0
www.chemphyschem.org
4
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
Articles was spin-coated at 3000 rpm for 10 s and heat-treated at 150 8C for 30 min. The DA-layer thicknesses of the BL and BL-NW devices were 160 and 165 nm, respectively.
P3HT within inhomogeneous ALs, which warrants further studies. One interesting observation was that there was no correlation between Jsc and PCE variations. However, PCE showed correlation with Voc and FF. Analysis based on an equivalent-circuit model that consisted of nonideal diode and parasitic resistances allowed us to understand PCE variations qualitatively. The BL device had the lowest PCE because of the largest Rs and the smallest Voc that, presumably, was due to adverse recombination effects with no provision to effectively transport holes from the diffuse junction region.
The cell size, defined by the overlap of cathode and anode, was 0.24 cm2. All devices were measured under AM 1.5G illumination at 1 sun by using a solar simulator (PEC-L01, Peccell) that was calibrated to 100 mW cm2 with a reference Si cell (PEC-S101, Peccell). The EQEs of the fabricated OSCs were measured by using a photoelectrochemistry incident photon-to-current efficiency (IPCE) system (PE-IPCE, HS Technologies). Transmittance, reflectance, and absorbance spectra were measured by using a spectrophotometer (Cary 5000, Varian).
Experimental Section Acknowledgements
We fabricated OSCs with a structure of ITO/PEDOT:PSS/donor–acceptor (DA)/LiF/Al on ITO-coated glasses (12 W sq.1, GEOMATEC). PEDOT:PSS (Clevios P, Heraeus) were spin-coated onto a thoroughly cleaned ITO anode and baked for 10 min at 140 8C to form 30 nm thick layers. Organic DA layers were formed by spin-coating mixtures of P3HT donor and ICBA acceptor. ICBA (SOL5064A) and regioregular P3HT (4000-E) were purchased from Solaris Chem and Reike Metals, respectively. LiF and Al layers were evaporated to form an upper cathode. In this work, we fabricated four types of OSCs by using either a BHJ or a BL DA-layer structure, and additionally, by changing DA-layer morphologies with the inclusion of preformed P3HT NWs. A name convention of BHJ, BHJ-NW, BL, and BL-NW was used to identify the four types of OSCs according to their DA-layer structures and morphologies.
This work was supported by the National Research Foundation of Korea (NRF) (Grant No. NRF-535 2014R1A2A2A01005632), which is funded by the Korea government (MEST), and through the Priority Research Centers Program (NRF-2009-0094049). Keywords: donor–acceptor systems · energy conversion · nanowires · organic solar cells · quantum efficiency [1] G. Li, V. Shrotriya, J. Huang, Y. Yao, T. Moriarty, K. Emery, Y. Yang, Nat. Mater. 2005, 4, 864. [2] W. Ma, C. Yang, X. Gong, K. Lee, A. J. Heeger, Adv. Funct. Mater. 2005, 15, 1617. [3] M. C. Quiles, T. Ferenczi, T. Agostinelli, P. G. Etchegoin, Y. Kim, T. D. Anthopoulos, P. N. Stavrinou, D. C. Bradley, J. Nelson, Nat. Mater. 2008, 7, 158. [4] J. Liu, M. Arif, J. Zou, S. I. Khondaker, L. Zhai, Macromolecules 2009, 42, 9390. [5] S. Samitsu, T. Shimomura, S. Heike, T. Hashizume, K. Ito, Macromolecules 2008, 41, 8000. [6] D. H. Kim, J. T. Han, Y. D. Park, Y. Jang, J. H. Cho, M. Hwang, K. Cho, Adv. Mater. 2006, 18, 719. [7] J.-H. Kim, J. H. Park, J. H. Lee, J. S. Kim, M. Sim, C. Shim, K. Cho, J. Mater. Chem. 2010, 20, 7398. [8] J. S. Kim, J. H. Lee, J. H. Park, C. Shim, M. Sim, K. Cho, Adv. Funct. Mater. 2011, 21, 480. [9] S. Berson, R. De Bettignies, S. Bailly, S. Guillerez, Adv. Funct. Mater. 2007, 17, 1377. [10] D. Chen, F. Liu, C. Wang, A. Nakahara, T. P. Russell, Nano Lett. 2011, 11, 2071. [11] V. S. Gevaerts, L. J. Koster, M. M. Wienk, R. A. Janssen, ACS Appl. Mater. Interfaces 2011, 3, 3252. [12] Y. He, H.-Y. Chen, J. Hou, Y. Li, J. Am. Chem. Soc. 2010, 132, 1377. [13] D. W. Laird, R. Stegamat, H. Richter, V. Vejins, L. Scott, T. A. Lada, Patent WO 2008/018931A2. [14] G. Zhao, Y. He, Y. Li, Adv. Mater. 2010, 22, 4355. [15] F. Cheng, G. Fang, X. Fan, H. Huang, Q. Zheng, P. Qin, H. Lei, Y. Li, Sol. Energy Mater. Sol. Cells 2013, 110, 63. [16] Y.-J. Cheng, C.-H. Hsieh, Y. J. He, C.-S. Hsu, Y. F. Li, J. Am. Chem. Soc. 2010, 132, 17381. [17] C.-Y. Chang, C.-E. Wu, S.-Y. Chen, C. H. Cui, Y.-J. Cheng, C.-S. Hsu, Y.-L. Wang, Y. F. Li, Angew. Chem. Int. Ed. 2011, 50, 9386; Angew. Chem. 2011, 123, 9558. [18] S.-H. Liao, Y.-L. Li, T.-H. Jen, Y.-S. Cheng, S.-A. Chen, J. Am. Chem. Soc. 2012, 134, 14271. [19] Z. A. Tan, L. J. Li, C. H. Cui, Y. Q. Ding, Q. Xu, S. S. Li, D. P. Qian, Y. F. Li, J. Phys. Chem. C 2012, 116, 18626. [20] Z. A. Tan, W. Q. Zhang, C. H. Cui, Y. Q. Ding, D. P. Qian, Q. Xu, L. J. Li, S. S. Li, Y. F. Li, Phys. Chem. Chem. Phys. 2012, 14, 14589. [21] Z. A. Tan, D. P. Qian, W. Q. Zhang, L. J. Li, Y. Q. Ding, Q. Xu, F. Z. Wang, Y. F. Li, J. Mater. Chem. A 2013, 1, 657.
The DA layers of BHJ devices were spin-coated from a 1:1 mixture of P3HT and ICBA in dichlorobenzene (DCB), the total concentration of which was 34 mg mL1. In the case of BHJ-NW devices, DA layers were spin-coated from a P3HT NW-added mixture solution that was prepared in two steps. Initially, a P3HT solution was made by thoroughly mixing a P3HT solution (24 mg mL1) in DCB with the same volume of a solution (4 mg mL1) of preformed P3HT NWs in dichloromethane (CH2Cl2) under continuous stirring at room temperature. The total concentration of P3HT in the DCB/ CH2Cl2 mixture was 14 mg mL1, and 1/7 of the P3HT was NWs. Next, ICBA (17 mg) was added to this NW-containing P3HT solution to prepare a mixture at the total concentration of 31 mg mL1. In both device types, DA layers were formed by spin-coating the aforementioned P3HT/ICBA mixture solutions at 800 rpm for 30 s, which were followed with a 40 min waiting period for solvents to evaporate. Finally, a heat treatment was carried out at 140 8C for 10 min on a hot plate. The DA-layer thicknesses of the BHJ and BHJ-NW devices were 190 and 220 nm, respectively. Contrary to the BHJ and BHJ-NW cases, the BL and BL-NW DA layers were formed by sequential spin-coating of donor and acceptor solutions. A donor solution for the BL device was prepared by dissolving P3HT (20 mg) in DCB (1 mL), and that for the BL-NW device by 1:1 mixing of a P3HT solution (24 mg mL1) with a solution (4 mg mL1) of preformed P3HT NWs in CH2Cl2. The NW-containing donor solution was stirred over 24 h at room temperature before spin-coating. The single acceptor solution prepared by dissolving ICBA (10 mg) in CH2Cl2 at 60 8C under continuous stirring over 24 h was used for both the BL and BL-NW devices. The coating of the donor part, spin-coating of a donor solution at 800 rpm for 30 s and 10 min of heat treatment at 140 8C on a hot plate, was straightforward. However, the subsequent coating of the acceptor part required great care not to damage the donor part. In particular, the selection of orthogonal solvent for the donor solution, such as CH2Cl2 for ICBA, was of utmost importance. The ICBA solution ChemPhysChem 0000, 00, 0 – 0
www.chemphyschem.org
These are not the final page numbers! ÞÞ
5
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
Articles [27] S. H. Park, A. Roy, S. Beaupr, S. Cho, N. Coates, J. S. Moon, D. Moses, M. Leclerc, K. Lee, A. J. Heeger, Nat. Photonics 2009, 3, 297. [28] J. Burschka, N. Pellet, S.-J. Moon, R. Humphry-Baker, P. Gao, M. K. Nazeeruddin, M. Grtzel, Nature 2013, 499, 316.
[22] Z. A. Tan, W. Q. Zhang, D. P. Qian, C. H. Cui, Q. Xu, L. J. Li, S. S. Li, Y. F. Li, Phys. Chem. Chem. Phys. 2012, 14, 14217. [23] B. Kippelen, J.-L. Brdas, Energy Environ. Sci. 2009, 2, 251. [24] J. D. Servaites, S. Yeganeh, T. J. Marks, M. A. Ratner, Adv. Funct. Mater. 2010, 20, 97. [25] J. H. Yim, S.-Y. Joe, C. Pang, K. M. Lee, H. Jeong, J.-Y. Park, Y. H. Ahn, J. C. de Mello, S. Lee, ACS Nano 2014, 8, 2857. [26] M. A. Green, Solar Cells 1982, 7, 337.
&
ChemPhysChem 0000, 00, 0 – 0
www.chemphyschem.org
Received: October 27, 2014 Revised: January 21, 2015 Published online on && &&, 2015
6
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
ARTICLES Energy efficiency: Preformed poly(3hexylthiophene) (P3HT) nanowires (NWs) added to active layers of organic solar cells based on the P3HT–indeneC60 bisadduct (ICBA) donor–acceptor pair result in improved power conversion efficiencies for both bulk heterojunction- and bilayer-type devices. The differences in the light-harvesting efficiencies account for variations in the short-circuit current density.
ChemPhysChem 0000, 00, 0 – 0
www.chemphyschem.org
These are not the final page numbers! ÞÞ
S.-y. Joe, J. H. Yim, S. Y. Ryu, N. Y. Ha, Y. H. Ahn, J.-Y. Park, S. Lee* && – && Universal Efficiency Improvement in Organic Solar Cells Based on a Poly(3-hexylthiophene) Donor and an Indene-C60 Bisadduct Acceptor with Additional Donor Nanowires
7
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
