Letter pubs.acs.org/NanoLett
Nanocrystalline Rutile Electron Extraction Layer Enables LowTemperature Solution Processed Perovskite Photovoltaics with 13.7% Efficiency Aswani Yella,* Leo-Philipp Heiniger, Peng Gao, Mohammad Khaja Nazeeruddin,* and Michael Graẗ zel* Laboratory of Photonics and Interfaces, Swiss Federal Institute of Technology (EPFL), Station 6, CH 1015, Lausanne, Switzerland S Supporting Information *
ABSTRACT: We demonstrate low-temperature (70 °C) solution processing of TiO2/CH3NH3PbI3 based solar cells, resulting in impressive power conversion efficiency (PCE) of 13.7%. Along with the high efficiency, a strikingly high open circuit potential (VOC) of 1110 mV was realized using this lowtemperature chemical bath deposition approach. To the best of our knowledge, this is so far the highest VOC value for solutionprocessed TiO2/CH3NH3PbI3 solar cells. We deposited a nanocrystalline TiO2 (rutile) hole-blocking layer on a fluorinedoped tin oxide (FTO) conducting glass substrate via hydrolysis of TiCl4 at 70 °C, forming the electron selective contact with the photoactive CH3NH3PbI3 film. We find that the nanocrystalline rutile TiO2 achieves a much better performance than a planar TiO2 (anatase) film prepared by hightemperature spin coating of TiCl4, which produces a much lower PCE of 3.7%. We attribute this to the formation of an intimate junction of large interfacial area between the nanocrystalline rutile TiO2 and the CH3NH3PbI3 layer, which is much more effective in extracting photogenerated electrons than the planar anatase film. Since the complete fabrication of the solar cell is carried out below 100 °C, this method can be easily extended to plastic substrates. KEYWORDS: Chemical bath deposition, titanium dioxide, solar cells, perovskite absorber, CH3NH3PbI3, low-temperature fabrication
D
efficiencies of over 15% were obtained using, for example, a sequential deposition approach.16 Even though mesoporous oxides such as TiO2, ZrO2, ZnO, and Al2O3 were found to be very effective scaffolds to support the perovskite absorber material, it was soon realized, because of the ambipolar semiconducting nature of the perovskite, that planar device architectures17−24 can be employed instead. Here the perovskite layer is sandwiched between the selective electron and hole collecting contacts.25 The planar device architecture simplifies their fabrication on flexible substrates.22 Several different deposition techniques have been reported to achieve high efficiencies using such a planar device architecture, that is, vapor deposition18 and vapor-assisted solution processed deposition.25 However devices based entirely on low-temperature solution processing gave lower efficiencies compared to those made by vapor deposition. Herein we employ a chemical bath deposition method to deposit a judiciously designed nanocrystalline TiO2 (rutile) layer27−30 onto the fluorine doped tin oxide (FTO) substrates. The perovskite is formed on the nanocrystalline rutile film using a sequential deposition method.17 This results in devices
ye-sensitized solar cells have been recognized as one of the promising alternatives to the silicon solar cells because of their low cost, transparency, spectral tunability, and the possibility to make them on flexible substrates.1,2 Even though high efficiencies were obtained using liquid electrolytes,3 the solid state analogues showed inferior performance to their liquid counterparts.4 This was attributed to limited light harvesting as the mesoporous titanium dioxide film thickness is kept below 2 μm in order to ensure better pore filling by the solid hole transporting material (HTM), reduce carrier recombination, and enhance the charge collection.5 To increase the efficiency of the solid state devices, the optical absorption cross-section of the light harvesters needs to be increased. Organic−inorganic hybrid perovskite materials have gained much attention in the past years because of their high efficiency, low cost, and the ease to make these materials solution processable.6−26 The methyl ammonium lead iodide (CH3NH3PbI3) perovskite has a direct band gap of 1.5 eV,10 a large absorption coefficient (1.5 × 104 cm−1 at 550 nm), and very high charge carrier mobility.11 These properties render the organolead halide materials very attractive as photovoltaic light harvesters. Organo-lead iodide perovskites such as CH3NH3PbI3 have been applied in solid state mesoscopic solar cells where they were found to act not only as a sensitizer,12 but also as an electron,13 and hole conductor.14,15 In combination with a mesoporous TiO2 scaffold impressive © 2014 American Chemical Society
Received: January 31, 2014 Revised: March 9, 2014 Published: March 14, 2014 2591
dx.doi.org/10.1021/nl500399m | Nano Lett. 2014, 14, 2591−2596
Nano Letters
Letter
Figure 1. Cross-section SEM images of the TiO2 particles formed by chemical bath deposition using different concentrations of TiCl4 at (a) 100 mM, (b) 200 mM, (c) 250 mM, (d) 300 mM, (e) 400 mM, and (f) 500 mM. Panels a−c have the same scale bar of 250 nm, while the scale bar of panels d−f is 500 nm.
Figure 2. (a) Cross-sectional SEM image of the CH3NH3PbI3 film deposited on the nanocrystalline TiO2 film formed from a 200 mM TiCl4 chemical bath in water at 70 °C. The thickness of the perovskite layer is about 300 nm. (b) J−V curves obtained at full solar intensity using different concentrations of TiCl4 to produce the electron selective contact layer.
plays a crucial role in determining the electronic property of the junction formed with the perovskite light harvester. Figure 1 shows cross sectional scanning electron micrographs (SEMs) of TiO2 layers formed by chemical bath deposition using concentrations of TiCl4 in the range from 100 to 500 mM. Using 100 mM TiCl4 concentration resulted in the formation of crystalline nanoparticles that are approximately 5− 10 nm in diameter, as determined by scanning electron microscopy. Importantly, the XRD measurement shown in the Supporting Information section (Figure S1) reveals that the nanoparticles are composed of the rutile phase of TiO2. This is unexpected, as the stable low-temperature TiO2 phase is anatase. The likely reasons for the formation of rutile are (i) the use of TiCl4 as a precursor which is prone to produce rutile nanoparticles on hydrolysis31 and (ii) induction of epitaxial growth of rutile nanoparticles by the FTO support which has
that can be made at very low temperatures with a simple manufacturing process, which is also applicable to plastic substrates. The combination of the low temperature deposition of both the TiO2 electron collection layer and the perovskite light absorber is of great importance to the industrial scale up of this new photovoltaic technology. We use a nanocrystalline rutile TiO2 film as a selective contact, extracting electrons but blocking holes, in the fabrication of the CH3NH3PbI3 perovskite based PV devices. We employed a chemical bath deposition process, similar to the one previously used for dye-sensitized solar cells.27−30 Titanium tetrachloride (TiCl4) was hydrolyzed in water at 70 °C for 1 h, resulting in the formation of crystalline TiO2 (rutile) nanoparticles. We use the concentration of the TiCl4 in the chemical bath to control the size and agglomeration of the TiO2 nanoparticles formed during the hydrolysis process which 2592
dx.doi.org/10.1021/nl500399m | Nano Lett. 2014, 14, 2591−2596
Nano Letters
Letter
the fill factor and open-circuit voltage (VOC) are very low because of increased recombination at the FTO surface. The chemical bath deposition produces a nanoparticulate compact TiO2 (rutile) layer whose thickness can be varied systematically by increasing the concentration of TiCl4 in the chemical bath ranging from approximately 10 nm to few micrometers. From the data shown in Table 1, it is clear that the electron collecting and hole blocking layer is absolutely necessary to obtain a high open circuit potential and good fill factors. Employing a very thin (∼5−10 nm) TiO2 layer, prepared by treating the FTO substrate with 100 mM TiCl4, substantially improves both the VOC and the fill factor, resulting in PCEs above 10%. While the short-circuit current density was similar to the cell with no TiO2 layer, the fill factor almost doubled, and the open-circuit potential increased dramatically by nearly 600 mV. Upon doubling the concentration to 200 mM, the rutile nanoparticle layer became ∼25 nm, and the open circuit potential further increased to 1110 mV. To the best of our knowledge, this is the highest open circuit potential obtained with the TiO2/CH3NH3PbI3 solar cells based on solution processing. The short-circuit current density also rose by approximately 1 mA/cm2, yielding an overall power conversion efficiency (PCE) of 13% using a solution processable, all low temperature fabrication of TiO2/CH3NH3PbI3/Spiro-MeOTAD solar cells. A further increase in the concentration of TiCl4 treatment did not result in any improvement in device performance. On the contrary, all three PV device parameters, that is, the shortcircuit current density, the open circuit potential, and the fill factor, declined. The drop in photocurrent density can be explained by the poor interconnectivity of the TiO2 nanoparticles, with increasing TiCl4 concentrations since no sintering was applied during the whole fabrication process. Poor particle interconnectivity slows down the electron transport, resulting in lower photocurrents due to a reduction in the electron collection efficiency.32 The lowering of the open circuit potential upon increasing the TiCl4 concentrations can be attributed to the poor coverage of the TiO2 surface by the perovskite, hence increasing the contact between TiO2 and spiro-OMeTAD and therefore favoring charge carrier recombination over extraction. This is evident from the SEM image shown in Figure S3. The lowering of the fill factor with increasing TiCl4 concentration could also be the result of the poor connectivity between the particles and hence a higher series resistance. The IPCE spectrum shows the characteristic shape of devices based on CH3NH3PbI3 with an external quantum efficiency of approximately 70% across almost the entire spectrum (360− 750 nm). The onset of photocurrent at 800 nm is consistent with the reported band gap of CH3NH3PbI3. Integrating the product of the AM1.5G photon flux with the IPCE spectrum yields a predicted Jsc of 17.1 mA/cm2, which is in good agreement with the measured value of 17.4 mA/cm2 showing that the spectral mismatch is below 2%. Figure S6 presents the IPCE of the best performing device using 200 mM TiCl4 during the chemical bath deposition. To further improve light harvesting by the perovskite, the loading with CH3NH3PbI3 was increased by augmenting the concentration of the PbI2 in DMF from 460 mg/mL to 700 mg/mL. At such high concentrations the conversion of the PbI2 to CH3NH3PbI3 was found to be incomplete even after prolonged dipping for ∼5 min in the methyl amonium iodide in
also a rutile crystal structure. The rutile nanoparticles formed via chemical bath deposition adhere very strongly to the FTO, and hence no further calcination or sintering step is required for the formation of the electron-selective TiO2 contact. Increasing the concentration of the TiCl4 to 200 mM resulted in the formation of a compact layer consisting of slightly larger particles compared to the case of 100 mM. The particles now completely cover the FTO substrate resulting in a thin, nanoparticular film of approximately 25 nm thickness. At 250 mM, a thicker compact layer is grown together with star-shaped TiO2 particles on top. Increasing the concentration above 250 mM led to the formation of microporous TiO2 beads. The size of these beads increased further as the concentration is raised to 500 mM. The thickness of the compact nanoparticle layer beneath the beads also increases with the concentration of TiCl4 during the chemical bath deposition. At 500 mM TiCl4, a compact layer as thick as 500 nm is formed. Figure S2 shows the top view SEM images of the TiO2 samples prepared using different concentrations of the TiCl4 with the chemical bath deposition process. Further experimental details on the photoanode film preparation, device fabrication, and other characterization are given in the methods and materials (Supporting Information). A two-step deposition process was used to grow the CH3NH3PbI3 perovskite layer as reported previously.10 The procedure consists of spin coating a layer of PbI2 onto the nanoparticulate TiO2 surface, followed by immersion of the PbI2 coated substrate into a solution of CH3NH3I in isopropanol. The conversion of PbI2 to CH3NH3PbI3 perovskite was found to vary with CH3NH3I concentration and dipping time. In this study, the concentration of CH3NH3I isopropanol was kept constant at 8 mg/mL, and the dipping time was 20 s. Figure 2a shows the cross-sectional SEM image of the CH3NH3PbI3 film deposited on the TiO2 with a dipping time of 20 s, indicating a thickness of approximately 300 nm. The monolithic cell fabrication is completed by spin coating the spiro-MeOTAD hole conductor and evaporating the gold back contact. The current−voltage (J−V) characteristics of the devices with and without TiO2 layer are investigated under AM1.5G solar irradiance (Figure 2b, Table 1). To probe the effect of Table 1. Detailed Photovoltaic Parameters of the Devices Made with Different Concentrations of TiCl4 Used in the Chemical Bath Deposition TiO2 no TiO2 100 mM 200 mM 250 mM 300 mM 400 mM
TiCl4 TiCl4 TiCl4 TiCl4 TiCl4
intensity (mW/cm2)
VOC (mV)
Jsc (mA/cm2)
FF
η (%)
99.7 99.1 97.4 98.5 98.5 100.5
330 916 1110 996 773 749
16.35 16.48 17.41 15.24 13.19 11.61
0.327 0.685 0.656 0.586 0.605 0.492
1.77 10.5 13.03 9.10 6.26 4.21
TiO2 layer thickness on device performance, devices were prepared by using the different concentrations of the TiCl4 used in the chemical bath deposition (Table 1). The reference cell with no hole blocking layer (perovskite on bare FTO) shows a short-circuit photocurrent density (JSC), open-circuit voltage (VOC), and fill factor (FF) of 16.35 mA/cm2, 330 mV, and 0.327, respectively, leading to a PCE of 1.77%. Thus in the absence of the electron extracting nanocrystalline rutile layer 2593
dx.doi.org/10.1021/nl500399m | Nano Lett. 2014, 14, 2591−2596
Nano Letters
Letter
spin-coating a 2 M solution of TiCl4, followed by thermal treatment at 450 °C. This produces a smooth and compact 50− 80 nm thick TiO2 (anatase) film as shown by the crosssectional SEM picture in Figure 5. The purpose of this analysis was to compare a smooth TiO2 layer with a layer of rutile nanoparticles which is rough on a nanoscale but nonporous. In contrast to the nanocrystalline rutile layer, the mesoporous TiO2 anatase scaffold that is generally used has a porosity over 60%. The CH3NH3PbI3 perovskite was deposited as a planar layer on this flat film and was contacted by the spiro-MeOTAD hole transporting layer and gold as a back contact. From the J− V curve shown in Figure 5 one derives the following device parameters: Jsc = 7.58 mA/cm2. VOC = 872 mV, ff = 0.56, and PCE = 3.7%. Thus perovskite cells employing the planar hightemperature TiO2 hole blocking layer show much inferior performance to those using nanocrystalline rutile particles. A key finding of our investigation is that the nanocrystalline TiO2 (rutile), forming an intimate junction of large interfacial area with the CH3NH3PbI3 film, is much more effective in extracting photogenerated electrons from the perovskite than a conventional planar TiO2 film. This corroborates very recent findings with ZnO electron collection layers by Liu et al.22 In conclusion, we introduce a low-temperature route for the fabrication of TiO2/CH3NH3PbI3 solar cells using TiCl4 chemical bath deposition technique. The chemical bath deposition resulted in the formation of a layer of rutile nanoparticles forming a nanocrystalline junction with the perovskite layer, which was found to exhibit outstanding properties as an electron collector. A key finding of our investigation is that the nanocrystalline TiO2 (rutile) forming an intimate junction of large interfacial area with the CH3NH3PbI3 film is much more effective in extracting photogenerated electrons from the perovskite than a conventional planar TiO2 (anatase) film. The highest solar cell efficiencies were realized using a 200 mM TiCl4 chemical bath deposition with a resulting TiO2 layer thickness of approximately 25−30 nm. The measured VOC of 1110 mV is to our knowledge the highest reported open circuit potential with TiO2 based solution processable CH3NH3PbI3 perovskite devices. Increasing the loading of the PbI2 resulted in higher photocurrents and higher power conversion efficiencies of 13.7%. A much lower PCE of 3.7% was obtained with the
isopropanol. To enhance the conversion reaction, prewetting of the PbI2 coated films was carried out with isopropanol as previously described.14 Figure 3 shows the SEM images of the
Figure 3. Top-view SEM images of the perovskite crystals grown on a FTO glass treated in 200 mM TiCl4 (a) without and (b) with prewetting of the PbI2 film by isopropanol.
perovskite crystals grown without and with such a prewetting step. Clearly, the latter results in the growth of larger perovskite crystals that enhance light scattering. The combination of increasing the PbI2 concentration and prewetting resulted in significantly higher photocurrents (∼20 mA/cm 2) and accordingly higher efficiencies compared to the ones reported in Table 1 above. Table 2 and Figure 4a show the J−V Table 2. Detailed Photovoltaic Parameters of the Device Made with Higher Concentrations of PbI2 and Prewetting the Sample in Isopropanol TiO2
light intensity (mW/cm2)
VOC (V)
Jsc (mA/cm2)
FF
η (%)
200 mM TiCl4
97
1.05
19.8
0.64
13.7
characteristics of a device made with higher concentration of PbI2 and including a prewetting step. Figure 4b shows the IPCE of the same device with an impressive external quantum efficiency exceeding 80% across the whole visible region. Integrating the product of the AM1.5G photon flux with the IPCE spectrum yields a predicted Jsc of 20.7 mA/cm2 in excellent agreement with the measured photocurrent of 20.4 mA/cm2. We compared the performance of the above champion device with a perovskite cell that was produced in the same manner apart from using a TiO2 layer which was deposited by
Figure 4. J−V curves obtained in the dark and under full sun illumination for the device made with higher concentration of PbI2 and including prewetting in isopropanol. (b) IPCE spectrum of the corresponding cell. 2594
dx.doi.org/10.1021/nl500399m | Nano Lett. 2014, 14, 2591−2596
Nano Letters
Letter
Figure 5. (a) Cross-sectional scanning electron micrograph picture of a smooth compact TiO2 electron collection layer deposited by spin-coating a 2 M TiCl4 solution on the surface of fluorine-doped tin dioxide (FTO) glass, followed by sintering at 450 °C. (b) J−V curves measured in the dark (black curve) and under AM 1.5 solar light (red curve) of a perovskite photovoltaic device using the planar TiO2 layer instead of the nanocrystalline rutile film as an electron-collecting layer. (3) Yella, A.; Lee, H.-W.; Tsao, H. N.; Yi, C.; Chandiran, A. K.; Nazeeruddin, M. K.; Diau, E. W.-G.; Yeh, C.-Y.; Zakeeruddin, S. M.; Grätzel, M. Science 2011, 334, 629−634. (4) Bach, U.; Lupo, D.; Comte, P.; Moser, J. E.; Weissortel, F.; Salbeck, J.; Spreitzer, H.; Gratzel, M. Nature 1998, 395, 583−585. (5) Docampo, P.; Hey, A.; Guldin, S.; Gunning, R.; Steiner, U.; Snaith, H. J. Adv. Funct. Mater. 2012, 22, 5010−5019. (6) Im, J.-H.; Lee, C.-R.; Lee, J.-W.; Park, S.-W.; Park, N.-G. Nanoscale 2011, 3, 4088−4093. (7) Kojima, A.; Teshima, K.; Shirai, Y.; Miyasaka, T. J. Am. Chem. Soc. 2009, 131, 6050−6051. (8) Park, N.-G. J. Phys. Chem. Lett. 2013, 4, 2423−2429. (9) Kim, H.-S.; Lee, J.-W.; Yantara, N.; Boix, P. P.; Kulkarni, S. A.; Mhaisalkar, S.; Grätzel, M.; Park, N.-G. Nano Lett. 2013, 13, 2412− 2417. (10) Kim, H.-S.; Lee, C.-R.; Im, J.-H.; Lee, K.-B.; Moehl, T.; Marchioro, A.; Moon, S.-J.; Humphry-Baker, R.; Yum, J.-H.; Moser, J. E.; Gratzel, M.; Park, N.-G. Sci. Rep. 2012, 2, 591. (11) Stoumpos, C. C.; Malliakas, C. D.; Kanatzidis, M. G. Inorg. Chem. 2013, 52, 9019−9038. (12) Liang, K.; Mitzi, D. B.; Prikas, M. T. Chem. Mater. 1998, 10, 403−411. (13) Snaith, H. J. J. Phys. Chem. Lett. 2013, 4, 3623−3630. (14) Lee, M. M.; Teuscher, J.; Miyasaka, T.; Murakami, T. N.; Snaith, H. J. Science 2012, 338, 643−647. (15) Etgar, L.; Gao, P.; Xue, Z.; Peng, Q.; Chandiran, A. K.; Liu, B.; Nazeeruddin, M. K.; Grätzel, M. J. Am. Chem. Soc. 2012, 134, 17396− 17399. (16) Laban, W. A.; Etgar, L. Energy Environ. Sci. 2013, 6, 3249−3253. (17) Burschka, J.; Pellet, N.; Moon, S.-J.; Humphry-Baker, R.; Gao, P.; Nazeeruddin, M. K.; Gratzel, M. Nature 2013, 499, 316−319. (18) Eperon, G. E.; Burlakov, V. M.; Docampo, P.; Goriely, A.; Snaith, H. J. Adv. Funct. Mater. 2014, 24, 151−157. (19) Jeng, J.-Y.; Chiang, Y.-F.; Lee, M.-H.; Peng, S.-R.; Guo, T.-F.; Chen, P.; Wen, T.-C. Adv. Mater. 2013, 25, 3727−3732. (20) Sun, S.; Salim, T.; Mathews, N.; Duchamp, M.; Boothroyd, C.; Xing, G.; Sum, T. C.; Lam, Y. M. Energy Environ. Sci. 2014, 7, 399− 407. (21) Liu, M.; Johnston, M. B.; Snaith, H. J. Nature 2013, 501, 395− 398. (22) Malinkiewicz, O.; Yella, A.; Lee, Y. H.; Espallargas, G. M.; Graetzel, M.; Nazeeruddin, M. K.; Bolink, H. J. Nat. Photonics 2014, 8, 128−132. (23) Docampo, P.; Ball, J. M.; Darwich, M.; Eperon, G. E.; Snaith, H. J. Nat. Commun. 2013, 4, 2761. (24) Liu, D.; Kelly, T. L. Nat. Photonics 2014, 8, 133−138.
planar TiO2/perovskite junction using a high-temperature compact layer. We note that most of these planar TiO2 devices exhibit in addition a strong hysteresis in the J−V curves. The reason for the drop in the VOC and fill factor upon forward scanning is presently further investigated.
■
ASSOCIATED CONTENT
S Supporting Information *
Complete experimental details, X-ray diffraction pattern for the TiO 2 nanoparticles, top-view SEM images of the TiO2 nanoparticles deposited using different concentrations of TiCl4, top-view SEM image of the perovskite grown on TiO2, cross-section SEM image of the nanoparticulate compact layer formed with 200 mM TiCl4, cross-section SEM image of the device with TiO2/CH3NH3PbI3/Spiro-MeOTAD, and IPCE of the device made with the TiO2 deposited by hydrolyzing 200 mM TiCl4 and approximately 200 nm perovskite layer. This material is available free of charge via the Internet at http:// pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail: aswani.yella@epfl.ch. *E-mail: mdkhaja.nazeeruddin@epfl.ch. *E-mail: michael.graetzel@epfl.ch. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The research leading to these results has received funding from the Balzan foundation as part of the 2009 Balzan Prize awarded to Prof. Michael Graetzel, European Union Seventh Framework Programme under grant agreement number: 309194 “GLOBASOL” and Grant 308997 “NANOMATCELL”. The authors would like to acknowledge Dr. Morgan Stefik for the SEM of the high-temperature processed compact layer and Dr. Kurt Schenk for the XRD measurements.
■
REFERENCES
(1) Hagfeldt, A.; Boschloo, G.; Sun, L.; Kloo, L.; Pettersson, H. Chem. Rev. 2010, 110, 6595−6663. (2) O’Regan, B.; Grätzel, M. Nature 1991, 353, 737−740. 2595
dx.doi.org/10.1021/nl500399m | Nano Lett. 2014, 14, 2591−2596
Nano Letters
Letter
(25) Chen, Q.; Zhou, H.; Hong, Z.; Luo, S.; Duan, H.-S.; Wang, H.H.; Liu, Y.; Li, G.; Yang, Y. J. Am. Chem. Soc. 2014, 136, 622−625. (26) Xing, G.; Mathews, N.; Sun, S.; Lim, S. S.; Lam, Y. M.; Grätzel, M.; Mhaisalkar, S.; Sum, T. C. Science 2013, 342, 344−347. (27) Ito, S.; Liska, P.; Comte, P.; Charvet, R.; Pechy, P.; Bach, U.; Schmidt-Mende, L.; Zakeeruddin, S. M.; Kay, A.; Nazeeruddin, M. K.; Gratzel, M. Chem. Commun. 2005, 4351−4353. (28) Sommeling, P. M.; O’Regan, B. C.; Haswell, R. R.; Smit, H. J. P.; Bakker, N. J.; Smits, J. J. T.; Kroon, J. M.; van Roosmalen, J. A. M. J. Phys. Chem. B 2006, 110, 19191−19197. (29) O’Regan, B. C.; Durrant, J. R.; Sommeling, P. M.; Bakker, N. J. J. Phys. Chem. C 2007, 111, 14001−14010. (30) Lee, S.-W.; Ahn, K.-S.; Zhu, K.; Neale, N. R.; Frank, A. J. J. Phys. Chem. C 2012, 116, 21285−21290. (31) Wang, L.; Yuan, Z.; Egerton, T. A. Mater. Chem. Phys. 2012, 133, 304−310. (32) Carnie, M. J.; Charbonneau, C.; Barnes, P. R. F.; Davies, M. L.; Mabbett, I.; Watson, T. M.; O’Regan, B. C.; Worsley, D. A. J. Mater. Chem. A 2013, 1, 2225−2230.
2596
dx.doi.org/10.1021/nl500399m | Nano Lett. 2014, 14, 2591−2596
