pubs.acs.org/NanoLett
Photovoltaic Devices with a Low Band Gap Polymer and CdSe Nanostructures Exceeding 3% Efficiency Smita Dayal,* Nikos Kopidakis, Dana C. Olson, David S. Ginley, and Garry Rumbles National Renewable Energy Laboratory, 1617 Cole Boulevard, Golden, Colorado 80401 ABSTRACT We report the fabrication and measurement of solar cells approaching a power conversion efficiency of 3.2% using a low band gap conjugated polymer poly[2,6-(4,4-bis-(2-ethylhexyl)-4H-cyclopenta[2,1-b;3,4-b′]dithiophene)-alt-4,7-(2,1,3-benzothiadiazole)] and CdSe nanoparticles. These devices exhibit an external quantum efficiency (EQE) of >30% in a broad range of 350-800 nm with a maximum EQE of 55% in a range of 630-720 nm. We also present certified device efficiencies of 3.13% under AM 1.5 illumination. KEYWORDS Nanoparticles, inorganic-organic photovoltaic devices, low band gap polymer, power conversion efficiency
B
ulk heterojunction photovoltaic cells based on composites of semiconducting nanoparticles and conjugated polymers are attractive due to their solution processability as well as the ability to tune each component in order to achieve composite films optimized for solar energy conversion.1-3 Using semiconductor nanoparticles as electron acceptors combined with the polymer as the electron donor gives these devices the additional advantage that light is absorbed by both components,1 unlike the prototypical polymer/fullerene bulk heterojunction where the PC61BM fullerene contributes very little to the spectral response although PC70BM shows an enhancement in optical absorption.4-6 A number of research groups have been working on the realization of these composite solar cell devices and have studied in detail the effect on the device performance of various processing conditions, shape of the nanoparticles, and choice of the conjugated polymers. It has been demonstrated that branched particles (tetrapods) result in better device performance than both nanorods and quantum dots when blended with polymers.7 Conjugated polymers that have been utilized in these devices are poly(3-hexylthiophene) (P3HT),8-12 poly(phenylene vinylene) derivatives such as: OC1C10 PPV,7,13 MEH-PPV,14 and alternating polyfluorene copolymer APFO-3.15 These studies have led to significant improvements in efficiency, with the current state-of-the-art at 2.6%.12 One of the main factors limiting the efficiency of these solar cells is the large (∼1.9 eV) band gap of the polymer, which prevents the efficient harnessing of the red and near-IR parts of the solar spectrum. Among new classes of polymers being developed, copolymers utilizing an electron-rich cyclopentadithiophene unit in the poly-
mer chain are of particular interest because their band gap of ∼1.4 eV allows for more efficient light harvesting at long wavelengths.16,17 The power conversion efficiency (PCE) of devices using this class of polymer blended with fullerene derivatives has reached a certified efficiency of close to 6%,18,19 which is among the highest reported for organic photovoltaics. We report here on bulk heterojunction solar cells based on composites of CdSe tetrapods and the low band gap polymer poly[2,6-(4,4-bis-(2-ethylhexyl)-4H-cyclopenta[2,1b;3,4-b′]dithiophene)-alt-4,7-(2,1,3-benzothiadiazole)] (PCPDTBT), reaching a power conversion efficiency of ∼3.2% under simulated air-mass 1.5 global irradiation of 1000 W/m2. These devices exhibit a reasonable high external quantum efficiency (EQE) of >30% in a range of 350-800 nm with a maximum EQE of 55% in a range of 630-720 nm which is the primary absorption band of the polymer. For the first time, we also report the National Renewable Energy Laboratory (NREL)-certified efficiency of 3.13% of these nanoparticle-polymer composite solar cells, measured by the Characterization and Measurement group within the National Center for Photovoltaics at NREL.20 CdSe tetrapods were synthesized by following published procedure.7 In brief, 0.35 g of cadmium oxide, 2.6 g of trioctylphosphosphonic acid, and 1.07 g of octylphosphonic acid were slowly heated in a three-neck flask to 310 °C under N2 flow. Heat was removed after the cadmium solution turned colorless, and the solution was stored in the glovebox. The selenium precursor was prepared by dissolving 0.42 g of Se in 1.27 g of tributylphosphine. The solution was loaded in a 10 mL syringe with 0.3 g of toluene added and cooled in a refrigerator for 15 min before injection. After 48 h cadmium precursor was heated back to 300 °C and selenium precursor was injected rapidly at 300 °C. Tetrapods were grown at 250 °C for 50 min. Synthesized tetra-
* Corresponding author,
[email protected]. Received for review: 10/12/2009 Published on Web: 12/11/2009 © 2010 American Chemical Society
239
DOI: 10.1021/nl903406s | Nano Lett. 2010, 10, 239-242
FIGURE 1. (a) Transmission electron microscopy image of CdSe tetrapods. Scale bar, 50 nm. (b) Absorption spectra of films of pure CdSe tetrapods (dot-dashed, blue), PCPDTBT (dotted, red), and their blend (solid, black).
FIGURE 2. (a) Molecular structure of PCPDTBT. (b) Photovoltaic device structure of the CdSe-PCPDTBT devices. FIGURE 3. (a) J-V characteristics measured under AM 1.5 global solar spectrum with average value of Voc ) 678 mV, Jsc ) 10.1 mA/ cm2, fill factor ) 0.51, and η ) 3.19%. Inset shows the dark current. (b) External quantum efficiency (EQE) action spectrum of the photovoltaic device containing CdSe and PCPDTBT. The absorption spectrum of the blends is also shown for comparison.
pods were washed four times with a mixture of toluene and ethanol to remove the excess capping ligand, and the remaining phosphonic acid ligands were exchanged with pyridine by heating the particles in pyridine overnight at 107 °C. Pyridine-treated particles were recovered by precipitation with hexane and dissolved in a 9:1 mixture of chloroform and pyridine. The resulting solution was sonicated for 1 h and filtered through a 1 µm PTFE filter. The final concentration of CdSe tetrapods was 30 mg/mL as determined by drying and weighing a known volume of the sample. The transmission electron micrograph of CdSe tetrapods is shown in Figure 1a. The majority of nanoparticles have multiple arms, and the average diameter of the arm is ∼5 nm with the arm length in a range of 30-50 nm. Figure 1b shows the absorption spectra of films of pure CdSe, pure PCPDTBT and their blend (9:1 CdSe:PCPDTBT weight ratio cast from a chloroform/pyridine/trichlorobenzene solution). The film of pyridine-treated CdSe nanoparticles shows an exciton absorption peak at ∼625 nm. The PCPDTBT film exhibits a very broad absorption spectrum and shows two distinct absorption bands at 410 and 717 nm. The absorption onset of PCPDTBT is red-shifted in thin film compared to the absorption peak in chloroform (not shown here); a shift that is often seen for conjugated polymers.16 The absorption spectrum of the nanoparticle-PCPDTBT composite is simply a superposition of the respective absorption spectra of pure PCPDTBT and nanoparticle features as shown in Figure 1b. In a composite film having ∼90% by © 2010 American Chemical Society
weight nanoparticles, only 34% absorption is due to absorption by the nanoparticles, as compared to 66% by the PCPDTBT. The nanoparticle contribution to the absorption spectrum of the blend can be clearly seen as a shoulder at 628 nm that corresponds to the first exciton peak of the nanoparticles and also as increased absorption at wavelengths below 570 nm. The structure of the PCPDTBT low band gap polymer is shown in Figure 2a and consists of an electron-rich cyclopentadithiophene group and an alternating electron-poor benzothiadiazole group. The optical band gap of PCPDTBT is ∼1.4 eV, which is very close to the optimum optical band gap of ∼1.5 eV,16 that is required for efficient solar harvesting by a single-junction device. Figure 2b also shows the architecture of a fabricated device. For device fabrication a solution containing 0.9 mL of 30 mg/mL CdSe tetrapods in chloroform and pyridine and 0.2 mL of 20 mg/mL PCPDTBT in trichlorobenzene was mixed thoroughly overnight. Solar cells were fabricated in sandwich geometry, as shown in Figure 2b. Prepatterned ITO coated glass substrates (purchased from Thin Film Devices) were cleaned with water, sonicated in acetone and isopropyl alcohol, and dried with a stream of N2. Substrates were O2 plasma cleaned immediately prior to the deposition of a 30 240
DOI: 10.1021/nl903406s | Nano Lett. 2010, 10, 239-242
duced the JSC of the device with a 10% accuracy. These J-V data were acquired with a home-built solar simulator inside aN2-filledglovebox.23 AcalibratedfilteredSidiode(Hamamatsu, S1787-04) was used as the reference cell for J-V measurements. EQE measurements were performed in the same glovebox with a fiber-coupled monochromator, a Stanford Research SR830 lock-in amplifier, and a NIST-calibrated silicon diode (UDI, uv-100) for visible wavelengths. The current density versus voltage (J-V) characteristic of the cell measured under Air Mass 1.5 Global solar spectrum is shown in Figure 3a. After accounting for spectral mismatch, the optimized solar cell has an efficiency of 3.19% with JSC ) 10.1 mA/cm2, VOC ) 678 mV, and FF ) 0.51. To date, this is the highest efficiency observed for any polymer-nanoparticle solar cell. Figure 3b shows the measured EQE spectrum of a typical CdSe tetrapod-PCPDTBT device. These devices show very broad EQE spectra spanning a wavelength range of 350-850 nm. A maximum EQE of ∼55% was observed in a range of 630-720 nm. The small peak observed at 480 nm clearly shows the evidence of the CdSe tetrapods where PCPDTBT has very low absorption. The comparison of a blend absorption spectrum with an EQE spectrum shows the evidence of both CdSe and PCPDTBT response in a wavelength range of 350-850 nm. One of the devices was loaded in an air-free sample holder and submitted for certification by the Characterization and Measurement group of the National Center for Photovoltaics. For certification measurement an aperture was used to subtract any inaccuracy due to shadow effects and the NREL Measurement and Characterization group independently measured the area of aperture.24 The NREL certified J-V characteristics of the device are shown in Figure 4 and demonstrate an efficiency of 3.13%; a value very close to the efficiency of 3.19%, measured in our lab. These devices were stored in glovebox in N2 atmosphere, and after a period of more then 3 months since the time of fabrication, the devices still exhibit efficiencies over 3%. The J-V characteristics of cells were measured with different irradiation light intensity in a range of 1-100 mW/
FIGURE 4. NREL-certified J-V characteristic of the CdSe tetrapodPCPDTBT solar cell.
nm thick hole-conducting layer of poly(3,4-ethylene dioxythiophene)/poly(styrenesulfonate) (PEDOT/PSS) (HC Stark CLEVIOS 4083). The PEDOT/PSS layer was annealed at 120 °C in air for 1 h. Substrates were subsequently transferred to a N2-filled glovebox, and the active layer was spin-cast. The active layer was dried slowly in a covered Petri dish and subsequently annealed at 100 °C for 5 min. A 0.6 nm thick lithium fluoride layer followed by a 100 nm aluminum electrode layer was deposited by thermal evaporation in a vacuum of 10-7 Torr or better. The thin layer of LiF was found to enhance the fill factor of the devices, as reported previously.21 The completed device area was 0.11 cm2. The average thickness of the active layer was 100-120 nm. The spectral mismatches for the samples were calculated according to Shrotriya et al.,22 overestimating short-circuit current JSC and conversion efficiency by 20% for our PCPDTBT-CdSe system; these were subsequently corrected. The integrated EQE and AM 1.5 spectrum repro-
FIGURE 5. (a) J-V characteristic of photovoltaic cell for different incident light intensities. (b) JSC, VOC, FF, and efficiency as a function of incident light intensity for CdSe-PCPDTBT solar cells. Incident light intensity was varied from 1 to 100 mW/cm2. © 2010 American Chemical Society
241
DOI: 10.1021/nl903406s | Nano Lett. 2010, 10, 239-242
cm2, and these are shown in Figure 5a. Various cell parameters (JSC, VOC, FF, and efficiency) are plotted as a function of incident light intensity in Figure 5b. JSC shows a linear dependence on incident light intensity indicating no charge buildup. The open circuit voltage, VOC, increases logarithmically with incident light intensity and reaches a maximum of 680 mV at 100 mW/cm2. The fill factor, FF, approaches a maximum value of 58% at 1 mW/cm2 and then decreases with further increase in irradiation intensity. The overall cell efficiency increases with increase in incident light intensity and reaches a maximum of 3.56% between 24 and 56 mW/ cm2. In conclusion, we have constructed and studied a bulk heterojunction photovoltaic device that contains a low band gap polymer, PCPDTBT, and CdSe tetrapods and exhibits a certified solar to power conversion efficiency of ∼3.2%. This work emphasizes the importance of low band gap polymers with a broad absorption spectrum for better solar spectrum harvesting and also demonstrates the contribution of nanoparticles toward achieving high power conversion efficiencies. The charge transfer dynamics of these blends studied with time-resolved microwave conductivity will be reported in an upcoming publication.
(3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18)
Acknowledgment. Authors would like to thank Dr. Andrew Norman for his help with TEM and P. Ciszek and K. Emery for certified solar cell measurements at NREL. We also thank Dr. Russ Gaudiana of Konarka Technologies, Inc., for providing PCPDTBT. The Department of Energy EERE Solar Technology Program through the National Center for Photovoltaics Seed Fund Program is acknowledged for funding.
(19) (20)
(21) (22)
REFERENCES AND NOTES (1) (2)
(23)
Greenham, N. C.; Peng, X.; Alivisatos, A. P. Phys. Rev. B 1996, 54 (24), 17628–17637. Saunders, B. R.; Turner, M. L. Adv. Colloid Interface Sci. 2008, 138 (1), 1–23.
© 2010 American Chemical Society
(24)
242
Huynh, W. U.; Peng, X. G.; Alivisatos, A. P. Adv. Mater. 1999, 11 (11), 923. Wang, X. J.; Perzon, E.; Oswald, F.; Langa, F.; Admassie, S.; Andersson, M. R.; Inganas, O. Adv. Funct. Mater. 2005, 15 (10), 1665–1670. Kazaoui, S.; Minami, N. Synth. Met. 1997, 86 (1-3), 2345–2346. Diener, M. D.; Alford, J. M. Nature 1998, 393 (6686), 668–671. Sun, B. Q.; Marx, E.; Greenham, N. C. Nano Lett. 2003, 3 (7), 961– 963. Han, L.; Qin, D.; Jiang, X.; Liu, Y.; Wang, L.; Chen, J.; Cao, Y. Nanotechnology 2006, 17 (18), 4736–4742. Gur, I.; Fromer, N. A.; Chen, C.; Kanaras, A. G.; Alivisatos, A. P. Nano Lett. 2007, 7 (2), 409–414. Liu, J. S.; Tanaka, T.; Sivula, K.; Alivisatos, A. P.; Frechet, J. M. J. J. Am. Chem. Soc. 2004, 126 (21), 6550–6551. Huynh, W. U.; Dittmer, J. J.; Alivisatos, A. P. Science 2002, 295 (5564), 2425–2427. Sun, B.; Greenham, N. C. Phys. Chem. Chem. Phys. 2006, 8 (30), 3557–3560. Sun, B. Q.; Snaith, H. J.; Dhoot, A. S.; Westenhoff, S.; Greenham, N. C. J. Appl. Phys. 2005, 97 (1), 014914. Zhou, Y.; Li, Y.; Zhong, H.; Hou, J.; Ding, Y.; Yang, C.; Li, Y. Nanotechnology 2006, 17 (16), 4041–4047. Wang, P.; Abrusci, A.; Wong, H. M. P.; Svensson, M.; Andersson, M. R.; Greenham, N. C. Nano Lett. 2006, 6 (8), 1789–1793. Soci, C.; Hwang, I.; Moses, D.; Zhu, Z.; Waller, D.; Gaudiana, R.; Brabec, C. J.; Heeger, A. Adv. Funct. Mater. 2007, 17 (4), 632– 636. Morana, M.; Wegscheider, M.; Bonanni, A.; Kopidakis, N.; Shaheen, S.; Scharber, M.; Zhu, Z.; Waller, D.; Gaudiana, R.; Brabec, C. Adv. Funct. Mater. 2008, 18 (12), 1757–1766. Park, S.; Roy, A.; Beaupre´, S.; Cho, S.; Coates, N.; Moon, J.; Moses, D.; Leclerc, M.; Lee, K.; Heeger, A. Nat. Photonics 2009, 3 (5), 297–302. Kim, J.; Lee, K.; Coates, N.; Moses, D.; Nguyen, T.; Dante, M.; Heeger, A. Science 2007, 317 (5835), 222–225. Although this is a certified result it only qualifies as a “Notable Exception” as the area of device is
