Photovoltaics
Semiconducting Carbon Nanotube Aerogel Bulk Heterojunction Solar Cells Yumin Ye, Dominick J. Bindl, Robert M. Jacobberger, Meng-Yin Wu, Susmit Singha Roy, and Michael S. Arnold*
Using a novel two-step fabrication scheme, we create highly semiconducting-enriched single-walled carbon nanotube (sSWNT) bulk heterojunctions (BHJs) by first creating highly porous interconnected sSWNT aerogels (sSWNT-AEROs), followed by back-filling the pores with [6,6]-phenyl-C71-butyric acid methyl ester (PC71BM). We demonstrate sSWNT-AERO structures with density as low as 2.5 mg cm−3, porosity as high as 99.8%, and diameter of sSWNT fibers ≤10 nm. Upon spin coating with PC71BM, the resulting sSWNT-AERO-PC71BM nanocomposites exhibit highly quenched sSWNT photoluminescence, which is attributed to the large interfacial area between the sSWNT and PC71BM phases, and an appropriate sSWNT fiber diameter that matches the inter-sSWNT exciton migration length. Employing the sSWNT-AERO-PC71BM BHJ structure, we report optimized solar cells with a power conversion efficiency of 1.7%, which is exceptional among polymer-like solar cells in which sSWNTs are designed to replace either the polymer or fullerene component. A fairly balanced photocurrent is achieved with 36% peak external quantum efficiency (EQE) in the visible and 19% peak EQE in the near-infrared where sSWNTs serve as electron donors and photoabsorbers. Our results prove the effectiveness of this new method in controlling the sSWNT morphology in BHJ structures, suggesting a promising route towards highly efficient sSWNT photoabsorbing solar cells.
1. Introduction Single-walled carbon nanotubes (SWNTs) are excellent candidates as photoabsorbers for next generation solar cells owing to their strong optical absorptivity, fast exciton and
Y. Ye, Dr. D. J. Bindl, R. M. Jacobberger, S. S. Roy, Prof. M. S. Arnold Department of Materials Science and Engineering University of Wisconsin-Madison Madison 53706, USA E-mail: [email protected] M.-Y. Wu Department of Electrical and Computer Engineering University of Wisconsin-Madison Madison 53706, USA DOI: 10.1002/smll.201400696 small 2014, DOI: 10.1002/smll.201400696
charge transport, solution-processability, and chemical and thermal stability.[1–4] Recently, a variety of SWNT sorting techniques have been developed, which allow for the isolation of highly pure semiconducting SWNTs (sSWNTs) and the elimination of metallic SWNTs (mSWNTs) from as-produced, heterogeneous mixtures of these materials.[5–7] The elimination of mSWNTs that otherwise quench photogenerated electron-hole pairs (which are bound as excitons) has allowed sSWNT-absorber based solar cells with significantly enhanced photovoltaic efficiency.[8–11] For example, Bindl et al.[12–14] and Shea et al.[15] have demonstrated efficient photon harvesting from type-controlled sSWNTs films using a planar bilayer structure consisting of a thin film of sSWNTs overcoated by a thin film of C60, where C60 serves as the electron acceptor to drive the dissociation of photogenerated excitons into free charge carriers. Absorbed photon to collected electron efficiency (APCE) > 85% at the sSWNT bandgap has been reported with AM
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
1
full papers
Y. Ye et al.
1.5G power conversion efficiency of 1% using thin films of sSWNTs that are ∼5 nm in thickness.[15] However, maintaining the high APCE in thicker films of sSWNTs, which are needed to increase the photon absorption efficiency of the SWNTs and thus the overall power conversion efficiency, has been difficult.[14] The limitation with these planar heterojunction architectures is that the SWNTs in the active layer form a “lying-down” morphology in which there is fast exciton transport in the plane of the film but slow exciton transport perpendicular to it. The exciton diffusion length perpendicular to the film, towards the sSWNT/C60 interface is limited by slow intertube energy transfer to distances that are 3.7%. One possibility is that the increased concentration of sSWNTs in the BHJs disrupts the charge collection pathways of PC71BM, resulting in a lower charge collection efficiency. The sSWNT fiber diameter and interconnectivity may also change as a function of f (Figure S4c); the relation of which will be studied in more detail in the future. We note that similar trends of JSC, FF, and ηP as a function of carbon nanotube concentration have been frequently observed in nanotube-containing solar cells where nanotubes serve either as an additive or a replacement of the polymer or fullerene component in polymer-fullerene solar cells for different purposes of accepting or collecting charges.[21,35,36] Possible causes to this phenomenon have been suggested, including the bundling of nanotubes and the disruption of the optimum phase separation of the polymer and/or fullerene component as the nanotube concentration increases. Our findings agree well with previous reports as manifested by the changes in sSWNT morphology, series resistance and RBCCF as a function of f. We further optimize the active layer thickness of the sSWNT-AERO-PC71BM BHJ layer, t, while maintaining f around 4 ± 0.5%. As shown in Figure 4c, the VOC and FF exhibit no discernable variation with t, whereas the JSC, and hence ηP, first increase and then decrease with increasing t. This trend is typical of polymer solar cells,[37,38] in which efficiency is limited by poor light absorption at small t and by poor charge collection or optical interference effects at high t. Figure 5 shows the performance of a sSWNT-AERO BHJ device with an optimized f = 3.7% and t = 100 nm. ηP = 1.7%, VOC = 0.56 V, JSC = 7.2 mA cm−2 and FF = 0.41 are measured under simulated AM 1.5G illumination. In comparison, sSWNT BHJ devices using sSWNTs as hole donors and electron acceptors to P3HT show JSC of 1.99 mA cm−2.[21] Likewise, sSWNT BHJ devices using sSWNTs as electron donors and hole acceptors to PC71BM show JSC of 3.1 mA cm−2.[22] The significantly enhanced JSC of the sSWNT-AERO BHJ devices contributes the most to the improved device performance. We further measure the EQE of the optimized device as a function of wavelength from 320 to 1300 nm. A peak EQE of 36% is observed in the visible range which
6 www.small-journal.com
Figure 5. a) Current density-voltage curves from optimized device in the dark (black) and under simulated AM 1.5G 100 mW cm−2 (red). Inset: dark current density-voltage characteristics on a log scale. b) External quantum efficiency (EQE) of the device (black) compared with the AM 1.5 G photon flux (red).
arises from PC71BM absorption, while a peak EQE of 19% is observed in the NIR range which arises from sSWNT absorption. The shape of the EQE spectra in the NIR region closely matches that of the sSWNT absorption spectrum (Figure 1), including S1 transition peaks of (6,5), (7,5) and (7,6) at 1000 nm, 1050 nm, and 1150 nm, respectively, and the S1+X peak at 900 nm. The shoulder peak at 655 nm is attributed to the S2 transition of the (7,5) sSWNTs. We note that most of the previously reported sSWNT BHJ devices have suffered from low photocurrent arising from the sSWNT components with the reported peak EQE in the NIR region at only 0.9%,[21] 2.5%[22] and 4.5%.[11] Bindl et al.[23] and Isborn et al.[24] reported peak EQE of 18% in the NIR, but the overall photocurrent was limited by the low EQE in the visible range. In comparison, we demonstrate sSWNT-AERO BHJ solar cells with a more balanced photocurrent generation across the solar spectrum. The dark current densityvoltage characteristics of the device are plotted in the inset of Figure 5a. The device shows a rectification ratio of ∼103 at ±1 V, diode ideality factor of ∼2, and a series resistance of Rs = 0.83 Ω cm2. Finally, to confirm the importance of the sSWNTs in the BHJ performance, we fabricate a reference PC71BM device
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2014, DOI: 10.1002/smll.201400696
Semiconducting Carbon Nanotube Aerogel Bulk Heterojunction Solar Cells
with an equivalent architecture but without the sSWNTs. The EQE spectrum of the reference device (Figure S5) shows significantly less photocurrent generation in the visible range with a peak EQE at only ∼1.5% and JSC = 0.064 mA cm−2. The dramatically increased photocurrent of sSWNT-AERO BHJ devices indicates a synergistic effect arising from the incorporation of sSWNT-AEROs: the sSWNTs absorb photons in the NIR range which generates a photocurrent, while the sSWNTs furthermore serve as hole acceptors and collectors for charges photogenerated in the PC71BM phase.
3. Conclusions In summary, we demonstrate the fabrication of nearly monochiral semiconducting carbon nanotube aerogels with density as low as ∼2.5 mg cm−3 and porosity as high as ∼99.8%. The highly porous structure imparts sSWNT-AERO films with optoelectronic properties distinct from densely packed sSWNT films. Using a two-step fabrication scheme, we create sSWNT BHJ structures by back-filling sSWNT-AEROs with PC71BM. The high interfacial area between sSWNT and PC71BM results in highly quenched sSWNT photoluminescence, suggesting an appropriate sSWNT fiber diameter that matches the inter-sSWNT exciton migration length. Employing the sSWNT-AERO-PC71BM structures, we report optimized solar cells with VOC = 0.56 V, JSC = 7.2 mA cm−2, FF = 0.41 and η = 1.7%. The power conversion efficiency is exceptional among polymer-like solar cells in which sSWNTs replace either the polymer or fullerene component and in which the sSWNTs act as NIR photoabsorbers. EQE measurements reveal the photocurrent is generated more balanced, with 36% peak EQE in the visible and 19% peak EQE in the NIR. We also note that by further controlling the diameter and bandgap distribution of the sSWNT photoabsorbers, it should be possible to more evenly cover an even wider spectral range. Compared to polymer-fullerene BHJ solar cells, sSWNT BHJ solar cells are much less studied. While the sSWNTs resemble polymers in certain aspects, the rigidity, long length, and built-in crystallinity bring distinct challenges in controlling BHJ morphology. Therefore the well-established blending and post-processing fabrication schemes for polymer BHJ solar cells cannot be easily adapted to sSWNT BHJ solar cells. We demonstrate that using our two-step method, we are able to control the density, fiber size, and interconnectivity of sSWNTs in the active layer. With further optimization of the active layer morphology, e.g. the sSWNT orientation, the internal structure of sSWNT fibers, and the crystallinity of the donor/acceptor phase, we expect this promising method will provide new opportunities for rapidly and significantly improving the efficiency of sSWNT BHJ solar cells.
4. Experimental Section Nearly monochiral sSWNT solutions are prepared as previously reported.[5,12,15] Briefly, 1 mg ml−1 SG65i CoMoCAT single-walled small 2014, DOI: 10.1002/smll.201400696
carbon nanotubes (Southwest Nanotechnologies, Lot # SG65i-L38) are dispersed in 3 mg ml−1 PFO (American Dye Source)/toluene solution by ultrasonication for 1 hr at 40% amplitude with a horntype sonic dismembrator (Fisher Scientific, 400 W). The resulting suspension is centrifuged at 50,000 g for 15 min and the top 80% of the supernatant is collected, filtered by a 5 µm PTFE filter, and dried by evaporating the solvent under vacuum, resulting in PFO-wrapped sSWNT powder. To remove the excess PFO, the PFOsSWNT powder is then dispersed in THF at 90 °C, and the solution is centrifuged at 4 °C, 50,000 g for ∼8 hours. The pellets are collected, dissolved in THF and centrifuged for 3 more times. The final sSWNT solution is obtained by dispersing the sSWNT pellet in chlorobenzene using micro-tip sonication at 10% amplitude. The sSWNT-AERO films are made by co-coasting a mixed solution of PMMA (Sigma Aldrich, > 99%, molecular weight = 950,000) and sSWNTs in chlorobenzene at concentrations of 10 mg ml−1 and 10 µg ml−1, respectively. The mixtures are deposited onto ITO-coated glass substrates by drop casting. The PMMA is then removed by repeated rinsing in acetone and isopropanol (IPA). The resulting nanotube films are quickly transferred to the chamber of a critical point dryer (Automegasamdri 915B) filled with IPA while the films are kept in a wet state. The IPA bath solution is then purged for 8 min in exchange for liquid CO2 under 800 psi. The chamber pressure is raised to 1300 psi and the temperature is raised to above 38 °C to reach the critical state. The nanotube films are then dried without introducing surface tension-induced collapse of the structure, resulting in highly porous interconnected aerogel films. The sSWNT-AERO-PC71BM composite film is prepared by spin coating PC71BM/chlorobenzene solutions (7–20 mg ml−1 depending on the desired active layer thickness) on sSWNT-AERO films in a glovebox where the H2O and O2 concentration are kept under 1 ppm. The rotation speed is set at 1000 rpm for 15 s, and 2000 rpm for 3 s. The resulting composite films are dried for 15 min in the glove box. For device fabrication, ITO-coated glass substrates are cleaned in acetone, trichloroethylene, and isopropanol for 20 min at 100 °C, and UV-Ozone exposed for 15 min. A 10 nm of TiO2 film is deposited on ITO substrate by spin coating a mixed precursor solution of titanium isopropoxide (Ti[OCH(CH3)2]4) (Sigma Aldrich, >97%), 2-methoxyethanol (CH3OCH2CH2OH) (Acros Organics, >99%), and ethanolamine (H2NCH2CH2OH) (Acros Organics, >99%) at 4000 rpm for 40 s, followed by sintering at 500 °C in air, as described elsewhere.[39] After depositing sSWNTAERO-PC71BM composite as the active layer on the TiO2 film, devices are completed by thermally evaporating 5 nm of MoO3 and 100 nm of Ag sequentially. The active area of each device is 0.785 mm2. The devices are measured under simulated AM 1.5G illumination using a custom 2-lamp simulator that provides a good match to the AM 1.5G spectrum in both visible and NIR spectra. The detailed description of the setup is provided in the supplementary information. The measured JSC and the JSC calculated by integrating the product of EQE and the AM 1.5G photon flux agree within 8%, indicating that the simulated spectrum is a good match (see Figure S3 for detailed comparison of the spectra). The current density-voltage characteristics are measured using a Keithley source meter (model 2636A). The EQE of the devices are measured following the methods as previously reported.[8,15] The detailed procedures are provided in the supplementary information. The SEM
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com
7
full papers
Y. Ye et al.
images are taken using a LEO 1530 scanning electron microscope. The XPS measurements are performed using a Thermo Scientific K-Alpha XPS with an Al Kα X-ray source and a 400 µm spot-size. The TEM images are taken using a LEO 912 EFTEM transmission electron microscope. The photoluminescence spectra are measured using a Horiba NanoLog iHR320 fluorescence spectrometer.
Acknowledgements Funding was provided by AFOSR FA9550–12–1–0063 (Y. Y., M. S. A.). Additional partial support was provided for D. J. B. by National Science Foundation DMR-0905861 and U.S. Army Research Office W911NF-12–1–0025; R. M. J. by the DOE Office of Science Early Career Research Program (DE- SC0006414) through the Office of Basic Energy Sciences (for TEM measurements) and from the Department of Defense through the National Defense Science & Engineering Graduate Fellowship (NDSEG) Program; and, M.-Y. W. by National Science Foundation UW-CEMRI DMR-1121288.
[1] T. Durkop, S. A. Getty, E. Cobas, M. S. Fuhrer, Nano Lett. 2004, 4, 35. [2] P. Avouris, M. Freitag, V. Perebeinos, Nat. Photonics 2008, 2, 341. [3] S. M. Bachilo, M. S. Strano, C. Kittrell, R. H. Hauge, R. E. Smalley, R. B. Weisman, Science 2002, 298, 2361. [4] R. H. Baughman, A. A. Zakhidov, W. A. de Heer, Science 2002, 297, 787. [5] A. Nish, J.-Y. Hwang, J. Doig, R. J. Nicholas, Nat. Nanotechnol. 2007, 2, 640. [6] H. L. Zhang, B. Wu, W. P. Hu, Y. Q. Liu, Chem. Soc. Rev. 2011, 40, 1324. [7] J. Liu, M. C. Hersam, MRS Bulletin 2010, 35, 315. [8] D. J. Bindl, M.-Y. Wu, F. C. Prehn, M. S. Arnold, Nano Lett. 2011, 11, 455. [9] R. M. Jain, R. Howden, K. Tvrdy, S. Shimizu, A. J. Hilmer, T. P. McNicholas, K. K. Gleason, M. S. Strano, Adv. Mater. 2012, 24, 4436. [10] M. P. Ramuz, M. Vosgueritchian, P. Wei, C. G. Wang, Y. L. Gao, Y. P. Wu, Y. S. Chen, Z. N. Bao, ACS Nano 2012, 6, 10384. [11] Y. Xie, M. G. Gong, T. A. Shastry, J. Lohrman, M. C. Hersam, S. Q. Ren, Adv. Mater. 2013, 25, 3433. [12] D. J. Bindl, M. S. Arnold, J. Phys. Chem. C 2013, 117, 2390. [13] D. J. Bindl, N. S. Safron, M. S. Arnold, ACS Nano 2010, 4, 5657. [14] D. J. Bindl, M. J. Shea, M. S. Arnold, Chem. Phys.2013, 413, 29.
8 www.small-journal.com
[15] M. J. Shea, M. S. Arnold, Appl. Phys. Lett. 2013, 102. [16] R. D. Mehlenbacher, M. Y. Wu, M. Grechko, J. E. Laaser, M. S. Arnold, M. T. Zanni, Nano Lett. 2013, 13, 1495. [17] M. Y. Wu, R. M. Jacobberger, M. S. Arnold, J. Appl. Phys. 2013, 113. [18] G. Li, V. Shrotriya, J. S. Huang, Y. Yao, T. Moriarty, K. Emery, Y. Yang, Nat. Mater. 2005, 4, 864. [19] S. H. Park, A. Roy, S. Beaupre, S. Cho, N. Coates, J. S. Moon, D. Moses, M. Leclerc, K. Lee, A. J. Heeger, Nat. Photonics 2009, 3, 297. [20] X. N. Yang, J. Loos, S. C. Veenstra, W. J. H. Verhees, M. M. Wienk, J. M. Kroon, M. A. J. Michels, R. A. J. Janssen, Nano Lett. 2005, 5, 579. [21] S. Q. Ren, M. Bernardi, R. R. Lunt, V. Bulovic, J. C. Grossman, S. Gradecak, Nano Lett. 2011, 11, 5316. [22] M. Bernardi, J. Lohrman, P. V. Kumar, A. Kirkeminde, N. Ferralis, J. C. Grossman, S. Q. Ren, ACS Nano 2012, 6, 8896. [23] D. J. Bindl, A. S. Brewer, M. S. Arnold, Nano Res. 2011, 4, 1174. [24] C. M. Isborn, C. Tang, A. Martini, E. R. Johnson, A. Otero-de-la-Roza, V. C. Tung, J. Phys. Chem. Lett. 2013, 4, 2914. [25] M. S. Arnold, J. L. Blackburn, J. J. Crochet, S. K. Doorn, J. G. Duque, A. Mohite, H. Telg, Phys. Chem. Chem. Phys. 2013, 15, 14896. [26] E. Verploegen, C. E. Miller, K. Schmidt, Z. N. Bao, M. F. Toney, Chem. Mater. 2012, 24, 3923. [27] G. Li, Y. Yao, H. Yang, V. Shrotriya, G. Yang, Y. Yang, Adv. Funct. Mater. 2007, 17, 1636. [28] T. Bordjiba, M. Mohamedi, L. H. Dao, Adv. Mater. 2008, 20, 815. [29] J. H. Zou, J. H. Liu, A. S. Karakoti, A. Kumar, D. Joung, Q. A. Li, S. I. Khondaker, S. Seal, L. Zhai, ACS Nano 2010, 4, 7293. [30] A. E. Aliev, J. Y. Oh, M. E. Kozlov, A. A. Kuznetsov, S. L. Fang, A. F. Fonseca, R. Ovalle, M. D. Lima, M. H. Haque, Y. N. Gartstein, M. Zhang, A. A. Zakhidov, R. H. Baughman, Science 2009, 323, 1575. [31] H. Y. Sun, Z. Xu, C. Gao, Adv. Mater. 2013, 25, 2554. [32] K. H. Kim, M. Vural, M. F. Islam, Adv. Mater. 2011, 23, 2865. [33] F. Schoppler, C. Mann, T. C. Hain, F. M. Neubauer, G. Privitera, F. Bonaccorso, D. P. Chu, A. C. Ferrari, T. Hertel, J. Phys. Chem. C 2011, 115, 14682. [34] T. Ando, J. Phys. Soc. Jpn. 2010, 79. [35] J. M. Lee, J. S. Park, S. H. Lee, H. Kim, S. Yoo, S. O. Kim, Adv. Mater. 2011, 23, 629. [36] L. Y. Lu, T. Xu, W. Chen, J. M. Lee, Z. Q. Luo, I. H. Jung, H. I. Park, S. O. Kim, L. P. Yu, Nano Lett. 2013, 13, 2365. [37] A. J. Moule, J. B. Bonekamp, K. Meerholz, J. Appl. Phys. 2006, 100. [38] Y. M. Nam, J. Huh, W. H. Jo, Sol. Energy Mater. Sol. Cells 2010, 94, 1118. [39] J. Y. Kim, S. H. Kim, H. H. Lee, K. Lee, W. L. Ma, X. Gong, A. J. Heeger, Adv. Mater. 2006, 18, 572.
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Received: March 14, 2014 Published online:
small 2014, DOI: 10.1002/smll.201400696
Semiconducting Carbon Nanotube Aerogel Bulk Heterojunction Solar Cells Yumin Ye, Dominick J. Bindl, Robert M. Jacobberger, Meng-Yin Wu, Susmit Singha Roy, and Michael S. Arnold*
Using a novel two-step fabrication scheme, we create highly semiconducting-enriched single-walled carbon nanotube (sSWNT) bulk heterojunctions (BHJs) by first creating highly porous interconnected sSWNT aerogels (sSWNT-AEROs), followed by back-filling the pores with [6,6]-phenyl-C71-butyric acid methyl ester (PC71BM). We demonstrate sSWNT-AERO structures with density as low as 2.5 mg cm−3, porosity as high as 99.8%, and diameter of sSWNT fibers ≤10 nm. Upon spin coating with PC71BM, the resulting sSWNT-AERO-PC71BM nanocomposites exhibit highly quenched sSWNT photoluminescence, which is attributed to the large interfacial area between the sSWNT and PC71BM phases, and an appropriate sSWNT fiber diameter that matches the inter-sSWNT exciton migration length. Employing the sSWNT-AERO-PC71BM BHJ structure, we report optimized solar cells with a power conversion efficiency of 1.7%, which is exceptional among polymer-like solar cells in which sSWNTs are designed to replace either the polymer or fullerene component. A fairly balanced photocurrent is achieved with 36% peak external quantum efficiency (EQE) in the visible and 19% peak EQE in the near-infrared where sSWNTs serve as electron donors and photoabsorbers. Our results prove the effectiveness of this new method in controlling the sSWNT morphology in BHJ structures, suggesting a promising route towards highly efficient sSWNT photoabsorbing solar cells.
1. Introduction Single-walled carbon nanotubes (SWNTs) are excellent candidates as photoabsorbers for next generation solar cells owing to their strong optical absorptivity, fast exciton and
Y. Ye, Dr. D. J. Bindl, R. M. Jacobberger, S. S. Roy, Prof. M. S. Arnold Department of Materials Science and Engineering University of Wisconsin-Madison Madison 53706, USA E-mail: [email protected] M.-Y. Wu Department of Electrical and Computer Engineering University of Wisconsin-Madison Madison 53706, USA DOI: 10.1002/smll.201400696 small 2014, DOI: 10.1002/smll.201400696
charge transport, solution-processability, and chemical and thermal stability.[1–4] Recently, a variety of SWNT sorting techniques have been developed, which allow for the isolation of highly pure semiconducting SWNTs (sSWNTs) and the elimination of metallic SWNTs (mSWNTs) from as-produced, heterogeneous mixtures of these materials.[5–7] The elimination of mSWNTs that otherwise quench photogenerated electron-hole pairs (which are bound as excitons) has allowed sSWNT-absorber based solar cells with significantly enhanced photovoltaic efficiency.[8–11] For example, Bindl et al.[12–14] and Shea et al.[15] have demonstrated efficient photon harvesting from type-controlled sSWNTs films using a planar bilayer structure consisting of a thin film of sSWNTs overcoated by a thin film of C60, where C60 serves as the electron acceptor to drive the dissociation of photogenerated excitons into free charge carriers. Absorbed photon to collected electron efficiency (APCE) > 85% at the sSWNT bandgap has been reported with AM
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
1
full papers
Y. Ye et al.
1.5G power conversion efficiency of 1% using thin films of sSWNTs that are ∼5 nm in thickness.[15] However, maintaining the high APCE in thicker films of sSWNTs, which are needed to increase the photon absorption efficiency of the SWNTs and thus the overall power conversion efficiency, has been difficult.[14] The limitation with these planar heterojunction architectures is that the SWNTs in the active layer form a “lying-down” morphology in which there is fast exciton transport in the plane of the film but slow exciton transport perpendicular to it. The exciton diffusion length perpendicular to the film, towards the sSWNT/C60 interface is limited by slow intertube energy transfer to distances that are 3.7%. One possibility is that the increased concentration of sSWNTs in the BHJs disrupts the charge collection pathways of PC71BM, resulting in a lower charge collection efficiency. The sSWNT fiber diameter and interconnectivity may also change as a function of f (Figure S4c); the relation of which will be studied in more detail in the future. We note that similar trends of JSC, FF, and ηP as a function of carbon nanotube concentration have been frequently observed in nanotube-containing solar cells where nanotubes serve either as an additive or a replacement of the polymer or fullerene component in polymer-fullerene solar cells for different purposes of accepting or collecting charges.[21,35,36] Possible causes to this phenomenon have been suggested, including the bundling of nanotubes and the disruption of the optimum phase separation of the polymer and/or fullerene component as the nanotube concentration increases. Our findings agree well with previous reports as manifested by the changes in sSWNT morphology, series resistance and RBCCF as a function of f. We further optimize the active layer thickness of the sSWNT-AERO-PC71BM BHJ layer, t, while maintaining f around 4 ± 0.5%. As shown in Figure 4c, the VOC and FF exhibit no discernable variation with t, whereas the JSC, and hence ηP, first increase and then decrease with increasing t. This trend is typical of polymer solar cells,[37,38] in which efficiency is limited by poor light absorption at small t and by poor charge collection or optical interference effects at high t. Figure 5 shows the performance of a sSWNT-AERO BHJ device with an optimized f = 3.7% and t = 100 nm. ηP = 1.7%, VOC = 0.56 V, JSC = 7.2 mA cm−2 and FF = 0.41 are measured under simulated AM 1.5G illumination. In comparison, sSWNT BHJ devices using sSWNTs as hole donors and electron acceptors to P3HT show JSC of 1.99 mA cm−2.[21] Likewise, sSWNT BHJ devices using sSWNTs as electron donors and hole acceptors to PC71BM show JSC of 3.1 mA cm−2.[22] The significantly enhanced JSC of the sSWNT-AERO BHJ devices contributes the most to the improved device performance. We further measure the EQE of the optimized device as a function of wavelength from 320 to 1300 nm. A peak EQE of 36% is observed in the visible range which
6 www.small-journal.com
Figure 5. a) Current density-voltage curves from optimized device in the dark (black) and under simulated AM 1.5G 100 mW cm−2 (red). Inset: dark current density-voltage characteristics on a log scale. b) External quantum efficiency (EQE) of the device (black) compared with the AM 1.5 G photon flux (red).
arises from PC71BM absorption, while a peak EQE of 19% is observed in the NIR range which arises from sSWNT absorption. The shape of the EQE spectra in the NIR region closely matches that of the sSWNT absorption spectrum (Figure 1), including S1 transition peaks of (6,5), (7,5) and (7,6) at 1000 nm, 1050 nm, and 1150 nm, respectively, and the S1+X peak at 900 nm. The shoulder peak at 655 nm is attributed to the S2 transition of the (7,5) sSWNTs. We note that most of the previously reported sSWNT BHJ devices have suffered from low photocurrent arising from the sSWNT components with the reported peak EQE in the NIR region at only 0.9%,[21] 2.5%[22] and 4.5%.[11] Bindl et al.[23] and Isborn et al.[24] reported peak EQE of 18% in the NIR, but the overall photocurrent was limited by the low EQE in the visible range. In comparison, we demonstrate sSWNT-AERO BHJ solar cells with a more balanced photocurrent generation across the solar spectrum. The dark current densityvoltage characteristics of the device are plotted in the inset of Figure 5a. The device shows a rectification ratio of ∼103 at ±1 V, diode ideality factor of ∼2, and a series resistance of Rs = 0.83 Ω cm2. Finally, to confirm the importance of the sSWNTs in the BHJ performance, we fabricate a reference PC71BM device
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2014, DOI: 10.1002/smll.201400696
Semiconducting Carbon Nanotube Aerogel Bulk Heterojunction Solar Cells
with an equivalent architecture but without the sSWNTs. The EQE spectrum of the reference device (Figure S5) shows significantly less photocurrent generation in the visible range with a peak EQE at only ∼1.5% and JSC = 0.064 mA cm−2. The dramatically increased photocurrent of sSWNT-AERO BHJ devices indicates a synergistic effect arising from the incorporation of sSWNT-AEROs: the sSWNTs absorb photons in the NIR range which generates a photocurrent, while the sSWNTs furthermore serve as hole acceptors and collectors for charges photogenerated in the PC71BM phase.
3. Conclusions In summary, we demonstrate the fabrication of nearly monochiral semiconducting carbon nanotube aerogels with density as low as ∼2.5 mg cm−3 and porosity as high as ∼99.8%. The highly porous structure imparts sSWNT-AERO films with optoelectronic properties distinct from densely packed sSWNT films. Using a two-step fabrication scheme, we create sSWNT BHJ structures by back-filling sSWNT-AEROs with PC71BM. The high interfacial area between sSWNT and PC71BM results in highly quenched sSWNT photoluminescence, suggesting an appropriate sSWNT fiber diameter that matches the inter-sSWNT exciton migration length. Employing the sSWNT-AERO-PC71BM structures, we report optimized solar cells with VOC = 0.56 V, JSC = 7.2 mA cm−2, FF = 0.41 and η = 1.7%. The power conversion efficiency is exceptional among polymer-like solar cells in which sSWNTs replace either the polymer or fullerene component and in which the sSWNTs act as NIR photoabsorbers. EQE measurements reveal the photocurrent is generated more balanced, with 36% peak EQE in the visible and 19% peak EQE in the NIR. We also note that by further controlling the diameter and bandgap distribution of the sSWNT photoabsorbers, it should be possible to more evenly cover an even wider spectral range. Compared to polymer-fullerene BHJ solar cells, sSWNT BHJ solar cells are much less studied. While the sSWNTs resemble polymers in certain aspects, the rigidity, long length, and built-in crystallinity bring distinct challenges in controlling BHJ morphology. Therefore the well-established blending and post-processing fabrication schemes for polymer BHJ solar cells cannot be easily adapted to sSWNT BHJ solar cells. We demonstrate that using our two-step method, we are able to control the density, fiber size, and interconnectivity of sSWNTs in the active layer. With further optimization of the active layer morphology, e.g. the sSWNT orientation, the internal structure of sSWNT fibers, and the crystallinity of the donor/acceptor phase, we expect this promising method will provide new opportunities for rapidly and significantly improving the efficiency of sSWNT BHJ solar cells.
4. Experimental Section Nearly monochiral sSWNT solutions are prepared as previously reported.[5,12,15] Briefly, 1 mg ml−1 SG65i CoMoCAT single-walled small 2014, DOI: 10.1002/smll.201400696
carbon nanotubes (Southwest Nanotechnologies, Lot # SG65i-L38) are dispersed in 3 mg ml−1 PFO (American Dye Source)/toluene solution by ultrasonication for 1 hr at 40% amplitude with a horntype sonic dismembrator (Fisher Scientific, 400 W). The resulting suspension is centrifuged at 50,000 g for 15 min and the top 80% of the supernatant is collected, filtered by a 5 µm PTFE filter, and dried by evaporating the solvent under vacuum, resulting in PFO-wrapped sSWNT powder. To remove the excess PFO, the PFOsSWNT powder is then dispersed in THF at 90 °C, and the solution is centrifuged at 4 °C, 50,000 g for ∼8 hours. The pellets are collected, dissolved in THF and centrifuged for 3 more times. The final sSWNT solution is obtained by dispersing the sSWNT pellet in chlorobenzene using micro-tip sonication at 10% amplitude. The sSWNT-AERO films are made by co-coasting a mixed solution of PMMA (Sigma Aldrich, > 99%, molecular weight = 950,000) and sSWNTs in chlorobenzene at concentrations of 10 mg ml−1 and 10 µg ml−1, respectively. The mixtures are deposited onto ITO-coated glass substrates by drop casting. The PMMA is then removed by repeated rinsing in acetone and isopropanol (IPA). The resulting nanotube films are quickly transferred to the chamber of a critical point dryer (Automegasamdri 915B) filled with IPA while the films are kept in a wet state. The IPA bath solution is then purged for 8 min in exchange for liquid CO2 under 800 psi. The chamber pressure is raised to 1300 psi and the temperature is raised to above 38 °C to reach the critical state. The nanotube films are then dried without introducing surface tension-induced collapse of the structure, resulting in highly porous interconnected aerogel films. The sSWNT-AERO-PC71BM composite film is prepared by spin coating PC71BM/chlorobenzene solutions (7–20 mg ml−1 depending on the desired active layer thickness) on sSWNT-AERO films in a glovebox where the H2O and O2 concentration are kept under 1 ppm. The rotation speed is set at 1000 rpm for 15 s, and 2000 rpm for 3 s. The resulting composite films are dried for 15 min in the glove box. For device fabrication, ITO-coated glass substrates are cleaned in acetone, trichloroethylene, and isopropanol for 20 min at 100 °C, and UV-Ozone exposed for 15 min. A 10 nm of TiO2 film is deposited on ITO substrate by spin coating a mixed precursor solution of titanium isopropoxide (Ti[OCH(CH3)2]4) (Sigma Aldrich, >97%), 2-methoxyethanol (CH3OCH2CH2OH) (Acros Organics, >99%), and ethanolamine (H2NCH2CH2OH) (Acros Organics, >99%) at 4000 rpm for 40 s, followed by sintering at 500 °C in air, as described elsewhere.[39] After depositing sSWNTAERO-PC71BM composite as the active layer on the TiO2 film, devices are completed by thermally evaporating 5 nm of MoO3 and 100 nm of Ag sequentially. The active area of each device is 0.785 mm2. The devices are measured under simulated AM 1.5G illumination using a custom 2-lamp simulator that provides a good match to the AM 1.5G spectrum in both visible and NIR spectra. The detailed description of the setup is provided in the supplementary information. The measured JSC and the JSC calculated by integrating the product of EQE and the AM 1.5G photon flux agree within 8%, indicating that the simulated spectrum is a good match (see Figure S3 for detailed comparison of the spectra). The current density-voltage characteristics are measured using a Keithley source meter (model 2636A). The EQE of the devices are measured following the methods as previously reported.[8,15] The detailed procedures are provided in the supplementary information. The SEM
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com
7
full papers
Y. Ye et al.
images are taken using a LEO 1530 scanning electron microscope. The XPS measurements are performed using a Thermo Scientific K-Alpha XPS with an Al Kα X-ray source and a 400 µm spot-size. The TEM images are taken using a LEO 912 EFTEM transmission electron microscope. The photoluminescence spectra are measured using a Horiba NanoLog iHR320 fluorescence spectrometer.
Acknowledgements Funding was provided by AFOSR FA9550–12–1–0063 (Y. Y., M. S. A.). Additional partial support was provided for D. J. B. by National Science Foundation DMR-0905861 and U.S. Army Research Office W911NF-12–1–0025; R. M. J. by the DOE Office of Science Early Career Research Program (DE- SC0006414) through the Office of Basic Energy Sciences (for TEM measurements) and from the Department of Defense through the National Defense Science & Engineering Graduate Fellowship (NDSEG) Program; and, M.-Y. W. by National Science Foundation UW-CEMRI DMR-1121288.
[1] T. Durkop, S. A. Getty, E. Cobas, M. S. Fuhrer, Nano Lett. 2004, 4, 35. [2] P. Avouris, M. Freitag, V. Perebeinos, Nat. Photonics 2008, 2, 341. [3] S. M. Bachilo, M. S. Strano, C. Kittrell, R. H. Hauge, R. E. Smalley, R. B. Weisman, Science 2002, 298, 2361. [4] R. H. Baughman, A. A. Zakhidov, W. A. de Heer, Science 2002, 297, 787. [5] A. Nish, J.-Y. Hwang, J. Doig, R. J. Nicholas, Nat. Nanotechnol. 2007, 2, 640. [6] H. L. Zhang, B. Wu, W. P. Hu, Y. Q. Liu, Chem. Soc. Rev. 2011, 40, 1324. [7] J. Liu, M. C. Hersam, MRS Bulletin 2010, 35, 315. [8] D. J. Bindl, M.-Y. Wu, F. C. Prehn, M. S. Arnold, Nano Lett. 2011, 11, 455. [9] R. M. Jain, R. Howden, K. Tvrdy, S. Shimizu, A. J. Hilmer, T. P. McNicholas, K. K. Gleason, M. S. Strano, Adv. Mater. 2012, 24, 4436. [10] M. P. Ramuz, M. Vosgueritchian, P. Wei, C. G. Wang, Y. L. Gao, Y. P. Wu, Y. S. Chen, Z. N. Bao, ACS Nano 2012, 6, 10384. [11] Y. Xie, M. G. Gong, T. A. Shastry, J. Lohrman, M. C. Hersam, S. Q. Ren, Adv. Mater. 2013, 25, 3433. [12] D. J. Bindl, M. S. Arnold, J. Phys. Chem. C 2013, 117, 2390. [13] D. J. Bindl, N. S. Safron, M. S. Arnold, ACS Nano 2010, 4, 5657. [14] D. J. Bindl, M. J. Shea, M. S. Arnold, Chem. Phys.2013, 413, 29.
8 www.small-journal.com
[15] M. J. Shea, M. S. Arnold, Appl. Phys. Lett. 2013, 102. [16] R. D. Mehlenbacher, M. Y. Wu, M. Grechko, J. E. Laaser, M. S. Arnold, M. T. Zanni, Nano Lett. 2013, 13, 1495. [17] M. Y. Wu, R. M. Jacobberger, M. S. Arnold, J. Appl. Phys. 2013, 113. [18] G. Li, V. Shrotriya, J. S. Huang, Y. Yao, T. Moriarty, K. Emery, Y. Yang, Nat. Mater. 2005, 4, 864. [19] S. H. Park, A. Roy, S. Beaupre, S. Cho, N. Coates, J. S. Moon, D. Moses, M. Leclerc, K. Lee, A. J. Heeger, Nat. Photonics 2009, 3, 297. [20] X. N. Yang, J. Loos, S. C. Veenstra, W. J. H. Verhees, M. M. Wienk, J. M. Kroon, M. A. J. Michels, R. A. J. Janssen, Nano Lett. 2005, 5, 579. [21] S. Q. Ren, M. Bernardi, R. R. Lunt, V. Bulovic, J. C. Grossman, S. Gradecak, Nano Lett. 2011, 11, 5316. [22] M. Bernardi, J. Lohrman, P. V. Kumar, A. Kirkeminde, N. Ferralis, J. C. Grossman, S. Q. Ren, ACS Nano 2012, 6, 8896. [23] D. J. Bindl, A. S. Brewer, M. S. Arnold, Nano Res. 2011, 4, 1174. [24] C. M. Isborn, C. Tang, A. Martini, E. R. Johnson, A. Otero-de-la-Roza, V. C. Tung, J. Phys. Chem. Lett. 2013, 4, 2914. [25] M. S. Arnold, J. L. Blackburn, J. J. Crochet, S. K. Doorn, J. G. Duque, A. Mohite, H. Telg, Phys. Chem. Chem. Phys. 2013, 15, 14896. [26] E. Verploegen, C. E. Miller, K. Schmidt, Z. N. Bao, M. F. Toney, Chem. Mater. 2012, 24, 3923. [27] G. Li, Y. Yao, H. Yang, V. Shrotriya, G. Yang, Y. Yang, Adv. Funct. Mater. 2007, 17, 1636. [28] T. Bordjiba, M. Mohamedi, L. H. Dao, Adv. Mater. 2008, 20, 815. [29] J. H. Zou, J. H. Liu, A. S. Karakoti, A. Kumar, D. Joung, Q. A. Li, S. I. Khondaker, S. Seal, L. Zhai, ACS Nano 2010, 4, 7293. [30] A. E. Aliev, J. Y. Oh, M. E. Kozlov, A. A. Kuznetsov, S. L. Fang, A. F. Fonseca, R. Ovalle, M. D. Lima, M. H. Haque, Y. N. Gartstein, M. Zhang, A. A. Zakhidov, R. H. Baughman, Science 2009, 323, 1575. [31] H. Y. Sun, Z. Xu, C. Gao, Adv. Mater. 2013, 25, 2554. [32] K. H. Kim, M. Vural, M. F. Islam, Adv. Mater. 2011, 23, 2865. [33] F. Schoppler, C. Mann, T. C. Hain, F. M. Neubauer, G. Privitera, F. Bonaccorso, D. P. Chu, A. C. Ferrari, T. Hertel, J. Phys. Chem. C 2011, 115, 14682. [34] T. Ando, J. Phys. Soc. Jpn. 2010, 79. [35] J. M. Lee, J. S. Park, S. H. Lee, H. Kim, S. Yoo, S. O. Kim, Adv. Mater. 2011, 23, 629. [36] L. Y. Lu, T. Xu, W. Chen, J. M. Lee, Z. Q. Luo, I. H. Jung, H. I. Park, S. O. Kim, L. P. Yu, Nano Lett. 2013, 13, 2365. [37] A. J. Moule, J. B. Bonekamp, K. Meerholz, J. Appl. Phys. 2006, 100. [38] Y. M. Nam, J. Huh, W. H. Jo, Sol. Energy Mater. Sol. Cells 2010, 94, 1118. [39] J. Y. Kim, S. H. Kim, H. H. Lee, K. Lee, W. L. Ma, X. Gong, A. J. Heeger, Adv. Mater. 2006, 18, 572.
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Received: March 14, 2014 Published online:
small 2014, DOI: 10.1002/smll.201400696
