Letter pubs.acs.org/NanoLett
Quantum-Confined ZnO Nanoshell Photoanodes for Mesoscopic Solar Cells Aravind Kumar Chandiran,† Mojtaba Abdi-Jalebi,† Aswani Yella,† M. Ibrahim Dar,†,‡ Chenyi Yi,† Srinivasrao A. Shivashankar,‡ Mohammad K. Nazeeruddin,*,† and Michael Graẗ zel*,† †
Laboratory of Photonics and Interfaces, Swiss Federal Institute of Technology (EPFL), Station 6, Lausanne CH 1015, Switzerland Centre for Nano Science and Engineering, Materials Research Centre, Indian Institute of Science, Bangalore 560012, India
‡
S Supporting Information *
ABSTRACT: We present a photoanode for dye-sensitized solar cell (DSC) based on ZnO nanoshell deposited by atomic layer deposition at 150 °C on a mesoporous insulating template. An ultrathin layer of ZnO between 3 and 6 nm, which exhibits quantum confinement effect, is found to be sufficient to transport the photogenerated electrons to the external contacts and exhibits near-unity collection efficiency. A 6 nm ZnO nanoshell on a 2.5 μm mesoporous nanoparticle Al2O3 template yields photovoltaic power conversion efficiency (PCE) of 4.2% in liquid DSC. Perovskite absorber (CH3NH3PbI3) based solid state solar cells made with similar ZnO nanostructures lead to a high PCE of 7%.
KEYWORDS: Dye-sensitized solar cell, zinc oxide, atomic layer deposition, CH3NH3PbI3, quantum confinement, perovskite
D
dye loading which is a must parameter to increase the optical absorption cross section. To increase the dye loading, subbranching or nanoforest concepts were utilized wherein the small nanowire branches were grown from the stem.5,9,12,16−25 Even though the latter concept improved the solar cell performances, the experimental complications on other hand are greater. To circumvent the issue we present a templating concept where in ultrathin ZnO nanoshells are grown by atomic layer deposition (ALD) on any arbitrary insulating template and employed as photoanode for DSC (Figure 1). ALD is employed in this study to deposit ZnO due to versatility
ye-sensitized solar cell (DSC) utilizes mesoporous wide band gap semiconductor nanomaterials as a substrate for dye adsorption and to transport the photogenerated charge carriers to the external contacts.1−3 Several conventional oxides including TiO2, ZnO, and SnO2 have been explored extensively in different nanoarchitectures and crystallographic phases.4−10 While nanoparticles based on anatase TiO2 have displayed the best performance to date, hunt for alternate materials has been pursued for the purpose of improving the transport rate of electrons in the photoanode film.11 The mobility of electrons (μe) in TiO2 is low, and hence for a given electron recombination rate the collection efficiency is low.12 The collection efficiency is given by the ratio of transport rate (ktrans) to the sum of transport and recombination rate (krec) of photogenerated electrons.13 It is quite obvious that a material with higher electron mobility can lead to a better electron transport for high efficiency devices. SnO2 possesses a higher μe but its conduction band is positioned lower than that of TiO2, and hence for a given redox mediator the theoretical attainable maximum open-circuit potential of the device is low.8,14,15 This drawback places a severe restriction in the employment of highly successful dye sensitizers being designed for TiO2. This has motivated the DSC community to exploit ZnO in which both the desired properties, a higher μe and a conduction band position similar to those of anatase TiO2, coexist. In the context of nanoarchitecture, ZnO in several morphologies has been successfully implemented as photoanodes for DSC. Most of the prior efforts were focused on single crystal ZnO nanowires to enhance the transport rate of electrons, which could lead to better collection efficiency of the photogenerated charge carriers at short-circuit. However, this approach presents a sharp restriction on the available surface area for the effective © 2014 American Chemical Society
Figure 1. Block diagram of the photoanode architecture where the ZnO nanoshell is deposited onto a mesoporous insulating Al2O3 template; the dotted arrows show the electron pathways. Received: October 25, 2013 Revised: January 31, 2014 Published: February 13, 2014 1190
dx.doi.org/10.1021/nl4039955 | Nano Lett. 2014, 14, 1190−1195
Nano Letters
Letter
Figure 2. TEM micrographs displays a conformal layer. (A) Bright-field TEM image, (B) HRTEM, and (C) SAED pattern.
surface acid−base properties.28,29 To verify the layer thickness, 50 cycles of ZnO was deposited on the mesoporous alumina substrate and analyzed by transmission electron microscopy (TEM). Bright-field TEM (Figure 2A) evidence conformal layer of ZnO on the Al2O3 nanoparticle and high-resolution TEM (Figure 2B) reveals crystal lattice fringes with d-spacing of 2.47 Å, corresponding to the (101) plane of ZnO, indicating the crystalline nature of the as-deposited films. From HRTEM, the thickness of the shell is found to be ∼10 nm and is consistent with ellipsometric measurement. Selected area electron diffraction (SAED, Figure 2C) shows a ring pattern which could be indexed to the hexagonal phase of ZnO. The absorption spectra of the ZnO nanoshell on the mesoporous Al2O3 is analyzed using UV−visible spectroscopy and is presented in Figure 3 and Supporting Information Figure
of the technique to form conformal layers and provide precise close control over thickness down to the angstrom level.26,27 Photovoltaic characteristics and electron transfer dynamics were systematically studied by depositing different thicknesses of ZnO on mesoporous insulating Al2O3 nanoparticles. Surprisingly, we discovered that a shell of merely 3−6 nm deposited on an insulating scaffold shows quantum confinement and transports the electrons efficiently, leading to a PCE of 4% at AM 1.5G solar illumination. To the best knowledge of the authors, this is the first time a proof has been provided for the concept that quantum-confined ZnO nanoshell deposited at a relatively low temperature (150 °C) exhibits fast electron transport in dye-sensitized solar cells and leads to almost 100% collection efficiency. In the case of titanium oxide, the minimum thickness needed for effective electron transport was shown previously to be 6 nm for a high-temperature processed ALD nanoshell or 15 nm when TiO2 is deposited at 200 °C.28,29 This three-dimensional core−shell architecture highlights the necessity to go down to the nanoregime for effective utilization of the materials and faster device processing that, together, drive the market to low cost solar cells. These nanoarchitectures were further implemented as a photoanode in our high efficient solid-state perovskite absorber (CH3NH3PbI3) based solar cells that lead to a power conversion efficiency of 7%.30,31 A screen printed Al2O3 mesoporous insulating template (2.5 μm) that has an average particle size, porosity, and pore diameter of 23 nm, 76%, and 47 nm, respectively, was used as a scaffold for the deposition of the ZnO “nanoshell”. Material with such a high porosity and pore diameter was selected because it ensures uniform diffusion of ALD metal/oxidizing precursors along the pores down to 2.5 μm and also because it allows studying various thickness of ZnO without narrowing down the pore diameter that can lead to mass transport limitation of Co2+/Co3+ in DSC.28,32,33 The deposition of 10 nm ZnO reduces the pore diameter from 47 to 27 nm but is still sufficient to allow the efficient shuttling of redox species in the electrolyte, across the pores from/to the counter electrode.34 Following the deposition of ZnO, the films were cleaned by oxygen plasma and subjected to different materials and photovoltaic characterization. In this work, the conventional firing process was not followed prior to the sensitization, as that could lead to film shrinkage and eventually distort the conformality of ZnO nanoshell. Complete experimental details are given in the Supporting Information. The growth rate of the film was estimated by employing spectroscopic ellipsometry by depositing different cycles of ZnO on a Si wafer. The growth per cycle was found to be 0.22 nm, that is, roughly 5 cycles yield a 1 nm thick ZnO layer. For the same number of cycles, the thickness of the deposition can vary on Al2O3 scaffold due to the possible difference in the
Figure 3. Absorption spectra of different thicknesses of ZnO nanoshell on alumina substrate.
S1A. One can notice that the reduction in the shell thickness of the ZnO leads to an absorption shift toward blue. The sample with 10 or 15 nm ZnO exhibits an absorption onset at around 405 nm and with decreasing thickness, the onset shifts down to ∼375 nm for a 2 nm ZnO. A clear difference in the onset could be observed even with 1 nm variation in the thickness and it could be attributed to the quantum confinement effect (QC) or scattering. As a consequence of QC, the conduction band tends to move toward vacuum when the thickness of the shell is decreased. Hence, for a given redox potential and recombination rate of electrons, the open-circuit potential of the solar cell is expected to be higher for the thinner layers. It has to be noted that the absolute absorption onset wavelength is higher than the reported values and it could be due to the electronic distortion induced by the underlying Al2O3 susbstrate.35,36 To further confirm the quantum confinement effect, we made 1191
dx.doi.org/10.1021/nl4039955 | Nano Lett. 2014, 14, 1190−1195
Nano Letters
Letter
Figure 4. (A) J−V curves measured at AM1.5G solar illumination for DSC containing photoanodes of different thicknesses of ZnO nanoshell. (B) Transport rate, (C) recombination rate, and (D) collection efficiencies of the corresponding devices are plotted as a function of voltage analyzed using transient decay techniques.
similar ALD ZnO depositions on planar alumina films instead of mesoporous layers. Supporting Information Figure S1B shows the absorption spectra of different thicknesses of ZnO on 20 nm planar Al2O3 substrates. Similar to mesoporous alumina substrates, a trend of shift in the wavelength onset toward blue, is observed when the thickness of the layers is decreased. These nanoshell photoanodes with different thicknesses of ZnO were sensitized with our standard D-π-A dye (Y123, Supporting Information Figure S2A) and devices were made using cobalt(bipyridine)32+/3+ (Supporting Information Figure S2B) redox electrolyte.11,32,37 A mesoporous alumina photoanode without zinc oxide is not expected to show any photovoltaic performance, as the excited dye molecules cannot inject electrons into alumina since latter does not possess any conduction band electronic states below the LUMO level of the dye. The solar cell with 2 nm ZnO exhibits short-circuit current density (JSC), open-circuit potential (VOC) and fill factor (ff), respectively, of 2.5 mAcm−2, 890 mV, and 0.67, leading to a power conversion efficiency (PCE) of 1.5%. When the shell thickness is increased to 3 nm, JSC increased significantly to 6 mAcm−2 and with VOC of 924 mV the cell results in a PCE of ∼4%. Further increase in the thickness enhanced the JSC to a maximum of 8 mA cm−2 with a decline in VOC and ff. Between ZnO shell thickness of 3 and 6 nm, the overall PCE remained at around 4%. The current−voltage curves are presented in Figure 4A and the corresponding photovoltaic data are given in Table 1. The enhancement in current density between 3 and 6 nm ZnO is supported by the improvement in the incident photonto-current collection efficiency (IPCE) spectra shown in Supporting Information Figure S3. To rationalize the observed photovoltaic parameters, dyeloading on the films, transport and recombination rates of the
Table 1. Current−Voltage Characteristics of the DyeSensitized Solar Cells with Different Thicknesses of ZnO on a 2.5 μm Al2O3 Template Measured under AM1.5G Solar Illumination (100 mW/cm2) thickness of ZnO ALD shell (nm)
JSC (mA cm‑2)
VOC (mV)
ff
PCE (%)
2 3 5 6 10
2.5 6.1 7.7 8.1 8.0
890.6 923.9 913.7 903.9 885.0
0.67 0.70 0.54 0.57 0.50
1.5 3.9 3.8 4.2 3.5
photogenerated electrons were investigated. Supporting Information Figure S4 shows the change in the absorption maximum at 463 nm of Y123 dye desorbed from the film in a basic solution of DMF. With increasing thickness, the dye uptake scales down linearly. This reduction in the dye loading could be partly attributed to the reduction in the available surface area as described in our previous reports.28,29 The transport rate of the photogenerated electrons in ZnO was studied using transient photocurrent decay and is displayed in Figure 4B as a function of the voltage. Two distinct type of transport response could be observed: (1) ktrans increases with increasing voltage, in short, when the electron quasi Fermi level moves closer to the conduction band, the transport rate increases steeply. This observation is similar to the transport properties observed in the conventional TiO2 nanoparticles used in the dye-sensitized solar cell and (2) the transport rate increases when shell thickness is increased from 3 to 5 nm and further deposition does not change the parameter significantly.38,39 This result, combined with the dye uptake, explains the evolution of the JSC. At 3 nm, despite higher dye loading the 1192
dx.doi.org/10.1021/nl4039955 | Nano Lett. 2014, 14, 1190−1195
Nano Letters
Letter
current density is probably limited due to the ktrans. However, beyond 5 nm, the current density is governed primarily by the difference in the dye uptake. The evolution of the VOC was studied using transient photovoltage decay technique and the measured recombination rates are presented as a function of their corresponding voltages in Figure 4C.39 A general trend of increasing electron recombination from the ZnO or underlying transparent conducting oxide to the Co3+is observed while moving toward the conduction band of ZnO or at higher voltages. At any given voltage, the recombination rate of electrons is higher for the 3 nm ZnO. The increase in the back flow of electrons could come mainly from the TCO, as underlayer passivation of just 3 nm may not be sufficient to block electron leakage.29 On increasing the thickness to 5 nm, the recombination rate is effectively suppressed. However with further deposition, the krec increases. The reverse trend in the recombination rate observed beyond 6 nm could be due to increased film conductivity or lower dye loading. For a given surface area, the decrease in the dye loading can expose the surface of the photoanode film to the oxidized species in the electrolyte component. This probably has led to an increase in the unwanted back reaction for thicker shell on Al2O3. Even though 3 nm ZnO shows higher back reaction, the open-circuit potential of the corresponding device is higher than in the other devices. The influence of quantum confinement plays an important role for thin layers where the conduction position edge/bottom of the discrete levels above the optical band gap is higher than for the thicker layers as evidenced by the absorption spectra shown in Figure 3. Thus, the upshift in the conduction band for the 3 nm ZnO dominates the recombination resulting in higher VOC of the solar cell. The shift in the conduction band in the dye-sensitized solar cells is studied using charge extraction measurements and the plot is presented in Figure 5. It can be seen that the device possessing 3 nm ZnO overlayer exhibits a negatively shifted (w.r.t. NHE) trap states distribution compared to the 10 nm ZnO overlayer, further confirming the quantum confinement effect in these ZnO films.
Finally the collection efficiency of the photogenerated electrons that cumulatively takes account of the transport and recombination rate was calculated and is given in Figure 4D. For a 3 nm ZnO, due to the low electron transport rate and high recombination rate the collection efficiency is relatively less than in the other devices. One has to note that for a 5 nm ZnO nanoshell the collection efficiency is almost 100% which is a significant factor for achieving record-high efficiency in devices.13,40 Despite higher dye uptake and almost similar collection efficiency, 3 nm ZnO showed low current density. It could be assigned to the low driving force available for the excited state electron injection from the dye to the conduction band of ZnO, due to the observation of quantum confinement effect.41 The electron recombination/transfer properties could not be studied for the device with 2 nm ZnO nanoshell as the transient current involved was below the resolution limit of our Keithley instrument. The low performance of this device, however, is attributed to the sluggish transport of photogenerated electrons within the ultrathin ZnO and to the external contact. Recently, solid-state mesoscopic solar cells based on CH3NH3PbI3 perovskite absorber have shown a significant potential for the production of low cost third generation solar cells. Power conversion efficiencies of these devices reached 15% with standard titanium dioxide photoanodes using organic hole conductor and gold counter electrodes.30,31 In this study, we employed our modified ZnO nanostructures (on a 300 nm Al2O3 scaffold) and for a shell thickness of 5 nm, we reach a high PCE of 7% with an impressive short-circuit current density of 18.5 mAcm−2. The device without any ZnO overlayer exhibited very low PCE of 1.7%. This is probably due to high recombination of photogenerated electrons to perovskite absorber or hole conductors, at the exposed FTO surface. The J−V curves and data are presented in Figure 6, Supporting
Figure 6. J−V curves measured under solar illumination for perovskite absorber mesoscopic solar cells containing photoanodes of different thicknesses of ZnO nanoshell. The values displayed in brackets within the legends give the solar illumination with respect to 1 sun.
Information Figure S5, and Table 2. Even though the ultrathin ZnO overlayers transport current densities of 18.5 mAcm−2, the effect of quantum confinement is not clear. The shift in the conduction band for different ZnO devices is evidenced in Figure 5 for dye-sensitized solar cells. But only a slight or marginal variation in the open-circuit potential is observed for the perovskite absorber based devices and the possible reason
Figure 5. Distribution of density of states measured by charge extraction technique for the DSC photoanodes containing 3 and 10 nm ALD ZnO overlayer. 1193
dx.doi.org/10.1021/nl4039955 | Nano Lett. 2014, 14, 1190−1195
Nano Letters
Letter
Foundation as a part of the 2009 Balzan prize awarded to Michael Grätzel.
Table 2. Current-Voltage Characteristics of the Perovskite Solar Cells with Different Thicknesses of ZnO on a 0.3 μm Al2O3 Template Measured under AM1.5G Solar Illumination (100 mW cm−2) thickness of ZnO ALD shell (nm)
JSC (mA cm‑2)
VOC (mV)
ff
PCE (%)
3 4 5 6
8.4 13.3 18.5 16.4
796.5 800.8 758.5 793.8
0.38 0.47 0.50 0.37
2.6 5.0 7.0 4.8
■
(1) O’Regan, B.; Grätzel, M. Nature 1991, 353, 737−740. (2) Hagfeldt, A.; Boschloo, G.; Sun, L.; Kloo, L.; Pettersson, H. Chem. Rev. 2010, 110, 6595−6663. (3) Gratzel, M. Nature 2001, 414, 338−344. (4) Ito, S.; Murakami, T. N.; Comte, P.; Liska, P.; Grätzel, C.; Nazeeruddin, M. K.; Grätzel, M. Thin Solid Films 2008, 516, 4613− 4619. (5) Fan, J.; Hao, Y.; Cabot, A.; Johansson, E. M. J.; Boschloo, G.; Hagfeldt, A. ACS Appl. Mater. Interfaces 2013, 5, 1902−1906. (6) Law, M.; Greene, L. E.; Johnson, J. C.; Saykally, R.; Yang, P. Nat. Mater. 2005, 4, 455−459. (7) Baxter, J. B.; Aydil, E. S. Appl. Phys. Lett. 2005, 86, 53114. (8) Gubbala, S.; Chakrapani, V.; Kumar, V.; Sunkara, M. K. Adv. Funct. Mater. 2008, 18, 2411−2418. (9) Martinson, A. B. F.; Elam, J. W.; Hupp, J. T.; Pellin, M. J. Nano Lett. 2007, 7, 2183−2187. (10) Mor, G. K.; Shankar, K.; Paulose, M.; Varghese, O. K.; Grimes, C. A. Nano Lett. 2005, 6, 215−218. (11) 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. (12) Zhang, Q.; Dandeneau, C. S.; Zhou, X.; Cao, G. Adv. Mater. 2009, 21, 4087−4108. (13) Grätzel, M. Inorg. Chem. 2005, 44, 6841−6851. (14) Park, S.-S.; Mackenzie, J. D. Thin Solid Films 1995, 258, 268− 273. (15) Bruneaux, J.; Cachet, H.; Froment, M.; Messad, A. Thin Solid Films 1991, 197, 129−142. (16) Westermark, K.; Rensmo, H.; Siegbahn, H.; Keis, K.; Hagfeldt, A.; Ojamäe, L.; Persson, P. J. Phys. Chem. B 2002, 106, 10102−10107. (17) Schölin, R.; Quintana, M.; Johansson, E. M. J.; Hahlin, M.; Marinado, T.; Hagfeldt, A.; Rensmo, H. J. Phys. Chem. C 2011, 115, 19274−19279. (18) Shi, Y.; Zhu, C.; Wang, L.; Zhao, C.; Li, W.; Fung, K. K.; Ma, T.; Hagfeldt, A.; Wang, N. Chem. Mater. 2013, 25, 1000−1012. (19) Keis, K.; Lindgren, J.; Lindquist, S.-E.; Hagfeldt, A. Langmuir 2000, 16, 4688−4694. (20) Quintana, M.; Edvinsson, T.; Hagfeldt, A.; Boschloo, G. J. Phys. Chem. C 2006, 111, 1035−1041. (21) Rensmo, H.; Keis, K.; Lindström, H.; Södergren, S.; Solbrand, A.; Hagfeldt, A.; Lindquist, S.-E.; Wang, L. N.; Muhammed, M. J. Phys. Chem. B 1997, 101, 2598−2601. (22) Willis, R. L.; Olson, C.; O’Regan, B.; Lutz, T.; Nelson, J.; Durrant, J. R. J. Phys. Chem. B 2002, 106, 7605−7613. (23) Bauer, C.; Boschloo, G.; Mukhtar, E.; Hagfeldt, A. J. Phys. Chem. B 2001, 105, 5585−5588. (24) Jiang, C. Y.; Sun, X. W.; Lo, G. Q.; Kwong, D. L.; Wang, J. X. Appl. Phys. Lett. 2007, 90, 263501. (25) Zhang, Q.; Chou, T. P.; Russo, B.; Jenekhe, S. A.; Cao, G. Angew. Chem. 2008, 120, 2436−2440. (26) George, S. M. Chem. Rev. 2009, 110, 111−131. (27) Leskelä, M.; Ritala, M. Thin Solid Films 2002, 409, 138−146. (28) Chandiran, A. K.; Yella, A.; Stefik, M.; Heiniger, L.-P.; Comte, P.; Nazeeruddin, M. K.; Grätzel, M. ACS Appl. Mater. Interfaces 2013, 5, 3487−3493. (29) Chandiran, A. K.; Comte, P.; Humphry-Baker, R.; Kessler, F.; Yi, C.; Nazeeruddin, M. K.; Grätzel, M. Adv. Funct. Mater. 2013, 23, 2775−2781. (30) Lee, M. M.; Teuscher, J.; Miyasaka, T.; Murakami, T. N.; Snaith, H. J. Science 2012, 338, 643−647. (31) Burschka, J.; Pellet, N.; Moon, S.-J.; Humphry-Baker, R.; Gao, P.; Nazeeruddin, M. K.; Gratzel, M. Nature 2013, 499, 316−319. (32) Feldt, S. M.; Gibson, E. A.; Gabrielsson, E.; Sun, L.; Boschloo, G.; Hagfeldt, A. J. Am. Chem. Soc. 2010, 132, 16714−16724.
could be the variation in the recombination rates on different ZnO films. Further studies on the passivation of these ZnO nanomaterials by insulating tunneling layers, to obtain higher VOC, are in progress.42,43 Thus, the present work highlights the efficacy of ultrathin ZnO nanoshells on insulating templates for a prospective high efficiency dye-sensitized solar cell. With a nanoshell thickness between 3 and 6 nm on a 2.5 μm Al2O3 mesoporous scaffold, the maximum power conversion efficiency of 4% was achieved with our standard organic sensitizer and Co2+/Co3+ redox mediator. Further work has to involve the implementation of this nanoshell on thicker alumina layers to efficiently harvest the solar spectrum efficiently, so as to increase the current density of the solar cell. As of now, the deposition of ZnO on thicker alumina templates is limited by the inhomogenous diffusion of ALD precursors across the thickness of the film.28 Solid-state solar cells made with similar quantum-confined ZnO nanostructure led to a power conversion efficiency of 7%. As the ZnO for liquid DSC is deposited at 150 °C, this work can be extended to any arbitrary flexible plastic substrate that can withstand 150 °C, leading to low-temperature fabrication of DSC for portable device applications.
■
ASSOCIATED CONTENT
S Supporting Information *
Complete experimental details, absorption spectra of ZnO on mesoporous and planar alumina films, structure of dye sensitizer and cobalt complex, IPCE spectra of 3 and 6 nm ZnO, dye uptake, and current−voltage curves of a device without ZnO nanoshell. This material is available free of charge via the Internet at http://pubs.acs.org.
■
REFERENCES
AUTHOR INFORMATION
Corresponding Authors
*E-mail: (M.K.N.) mdkhaja.nazeeruddin@epfl.ch. *E-mail: (M.G.) michael.graetzel@epfl.ch. Author Contributions
The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. A.K.C. and M.A.-J. contributed equally. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors acknowledge the financial contribution from EU FP7 project “ORION” (Grant NMP-229036). M.I.D gratefully acknowledges financial support from the Swiss confederation under Swiss Government Scholarship programme. The authors are grateful for the financial support from the Balzan 1194
dx.doi.org/10.1021/nl4039955 | Nano Lett. 2014, 14, 1190−1195
Nano Letters
Letter
(33) Nelson, J. J.; Amick, T. J.; Elliott, C. M. J. Phys. Chem. C 2008, 112, 18255−18263. (34) Tsao, H. N.; Comte, P.; Yi, C.; Grätzel, M. ChemPhysChem 2012, 13, 2976−2981. (35) Bahnemann, D. W.; Kormann, C.; Hoffmann, M. R. J. Phys. Chem. 1987, 91, 3789−3798. (36) Lin, K.-F.; Cheng, H.-M.; Hsu, H.-C.; Lin, L.-J.; Hsieh, W.-F. Chem. Phys. Lett. 2005, 409, 208−211. (37) Tsao, H. N.; Yi, C.; Moehl, T.; Yum, J.-H.; Zakeeruddin, S. M.; Nazeeruddin, M. K.; Grätzel, M. ChemSusChem 2011, 4, 591−594. (38) O’Regan, B. C.; Durrant, J. R. Acc. Chem. Res. 2009, 42, 1799− 1808. (39) Peter, L. M. Phys. Chem. Chem. Phys. 2007, 9, 2630−2642. (40) Chandiran, A. K.; Sauvage, F.; Casas-Cabanas, M.; Comte, P.; Zakeeruddin, S. M.; Graetzel, M. J. Phys. Chem. C 2010, 114, 15849− 15856. (41) Koops, S. E.; O’Regan, B. C.; Barnes, P. R. F.; Durrant, J. R. J. Am. Chem. Soc. 2009, 131, 4808−4818. (42) Chandiran, A. K.; Tetreault, N.; Humphry-Baker, R.; Kessler, F.; Baranoff, E.; Yi, C.; Nazeeruddin, M. K.; Grätzel, M. Nano Lett. 2012, 12, 3941−3947. (43) Chandiran, A. K.; Nazeeruddin, M. K.; Grätzel, M. Adv. Funct. Mater. 2013, n/a−n/a.
1195
dx.doi.org/10.1021/nl4039955 | Nano Lett. 2014, 14, 1190−1195
Quantum-Confined ZnO Nanoshell Photoanodes for Mesoscopic Solar Cells Aravind Kumar Chandiran,† Mojtaba Abdi-Jalebi,† Aswani Yella,† M. Ibrahim Dar,†,‡ Chenyi Yi,† Srinivasrao A. Shivashankar,‡ Mohammad K. Nazeeruddin,*,† and Michael Graẗ zel*,† †
Laboratory of Photonics and Interfaces, Swiss Federal Institute of Technology (EPFL), Station 6, Lausanne CH 1015, Switzerland Centre for Nano Science and Engineering, Materials Research Centre, Indian Institute of Science, Bangalore 560012, India
‡
S Supporting Information *
ABSTRACT: We present a photoanode for dye-sensitized solar cell (DSC) based on ZnO nanoshell deposited by atomic layer deposition at 150 °C on a mesoporous insulating template. An ultrathin layer of ZnO between 3 and 6 nm, which exhibits quantum confinement effect, is found to be sufficient to transport the photogenerated electrons to the external contacts and exhibits near-unity collection efficiency. A 6 nm ZnO nanoshell on a 2.5 μm mesoporous nanoparticle Al2O3 template yields photovoltaic power conversion efficiency (PCE) of 4.2% in liquid DSC. Perovskite absorber (CH3NH3PbI3) based solid state solar cells made with similar ZnO nanostructures lead to a high PCE of 7%.
KEYWORDS: Dye-sensitized solar cell, zinc oxide, atomic layer deposition, CH3NH3PbI3, quantum confinement, perovskite
D
dye loading which is a must parameter to increase the optical absorption cross section. To increase the dye loading, subbranching or nanoforest concepts were utilized wherein the small nanowire branches were grown from the stem.5,9,12,16−25 Even though the latter concept improved the solar cell performances, the experimental complications on other hand are greater. To circumvent the issue we present a templating concept where in ultrathin ZnO nanoshells are grown by atomic layer deposition (ALD) on any arbitrary insulating template and employed as photoanode for DSC (Figure 1). ALD is employed in this study to deposit ZnO due to versatility
ye-sensitized solar cell (DSC) utilizes mesoporous wide band gap semiconductor nanomaterials as a substrate for dye adsorption and to transport the photogenerated charge carriers to the external contacts.1−3 Several conventional oxides including TiO2, ZnO, and SnO2 have been explored extensively in different nanoarchitectures and crystallographic phases.4−10 While nanoparticles based on anatase TiO2 have displayed the best performance to date, hunt for alternate materials has been pursued for the purpose of improving the transport rate of electrons in the photoanode film.11 The mobility of electrons (μe) in TiO2 is low, and hence for a given electron recombination rate the collection efficiency is low.12 The collection efficiency is given by the ratio of transport rate (ktrans) to the sum of transport and recombination rate (krec) of photogenerated electrons.13 It is quite obvious that a material with higher electron mobility can lead to a better electron transport for high efficiency devices. SnO2 possesses a higher μe but its conduction band is positioned lower than that of TiO2, and hence for a given redox mediator the theoretical attainable maximum open-circuit potential of the device is low.8,14,15 This drawback places a severe restriction in the employment of highly successful dye sensitizers being designed for TiO2. This has motivated the DSC community to exploit ZnO in which both the desired properties, a higher μe and a conduction band position similar to those of anatase TiO2, coexist. In the context of nanoarchitecture, ZnO in several morphologies has been successfully implemented as photoanodes for DSC. Most of the prior efforts were focused on single crystal ZnO nanowires to enhance the transport rate of electrons, which could lead to better collection efficiency of the photogenerated charge carriers at short-circuit. However, this approach presents a sharp restriction on the available surface area for the effective © 2014 American Chemical Society
Figure 1. Block diagram of the photoanode architecture where the ZnO nanoshell is deposited onto a mesoporous insulating Al2O3 template; the dotted arrows show the electron pathways. Received: October 25, 2013 Revised: January 31, 2014 Published: February 13, 2014 1190
dx.doi.org/10.1021/nl4039955 | Nano Lett. 2014, 14, 1190−1195
Nano Letters
Letter
Figure 2. TEM micrographs displays a conformal layer. (A) Bright-field TEM image, (B) HRTEM, and (C) SAED pattern.
surface acid−base properties.28,29 To verify the layer thickness, 50 cycles of ZnO was deposited on the mesoporous alumina substrate and analyzed by transmission electron microscopy (TEM). Bright-field TEM (Figure 2A) evidence conformal layer of ZnO on the Al2O3 nanoparticle and high-resolution TEM (Figure 2B) reveals crystal lattice fringes with d-spacing of 2.47 Å, corresponding to the (101) plane of ZnO, indicating the crystalline nature of the as-deposited films. From HRTEM, the thickness of the shell is found to be ∼10 nm and is consistent with ellipsometric measurement. Selected area electron diffraction (SAED, Figure 2C) shows a ring pattern which could be indexed to the hexagonal phase of ZnO. The absorption spectra of the ZnO nanoshell on the mesoporous Al2O3 is analyzed using UV−visible spectroscopy and is presented in Figure 3 and Supporting Information Figure
of the technique to form conformal layers and provide precise close control over thickness down to the angstrom level.26,27 Photovoltaic characteristics and electron transfer dynamics were systematically studied by depositing different thicknesses of ZnO on mesoporous insulating Al2O3 nanoparticles. Surprisingly, we discovered that a shell of merely 3−6 nm deposited on an insulating scaffold shows quantum confinement and transports the electrons efficiently, leading to a PCE of 4% at AM 1.5G solar illumination. To the best knowledge of the authors, this is the first time a proof has been provided for the concept that quantum-confined ZnO nanoshell deposited at a relatively low temperature (150 °C) exhibits fast electron transport in dye-sensitized solar cells and leads to almost 100% collection efficiency. In the case of titanium oxide, the minimum thickness needed for effective electron transport was shown previously to be 6 nm for a high-temperature processed ALD nanoshell or 15 nm when TiO2 is deposited at 200 °C.28,29 This three-dimensional core−shell architecture highlights the necessity to go down to the nanoregime for effective utilization of the materials and faster device processing that, together, drive the market to low cost solar cells. These nanoarchitectures were further implemented as a photoanode in our high efficient solid-state perovskite absorber (CH3NH3PbI3) based solar cells that lead to a power conversion efficiency of 7%.30,31 A screen printed Al2O3 mesoporous insulating template (2.5 μm) that has an average particle size, porosity, and pore diameter of 23 nm, 76%, and 47 nm, respectively, was used as a scaffold for the deposition of the ZnO “nanoshell”. Material with such a high porosity and pore diameter was selected because it ensures uniform diffusion of ALD metal/oxidizing precursors along the pores down to 2.5 μm and also because it allows studying various thickness of ZnO without narrowing down the pore diameter that can lead to mass transport limitation of Co2+/Co3+ in DSC.28,32,33 The deposition of 10 nm ZnO reduces the pore diameter from 47 to 27 nm but is still sufficient to allow the efficient shuttling of redox species in the electrolyte, across the pores from/to the counter electrode.34 Following the deposition of ZnO, the films were cleaned by oxygen plasma and subjected to different materials and photovoltaic characterization. In this work, the conventional firing process was not followed prior to the sensitization, as that could lead to film shrinkage and eventually distort the conformality of ZnO nanoshell. Complete experimental details are given in the Supporting Information. The growth rate of the film was estimated by employing spectroscopic ellipsometry by depositing different cycles of ZnO on a Si wafer. The growth per cycle was found to be 0.22 nm, that is, roughly 5 cycles yield a 1 nm thick ZnO layer. For the same number of cycles, the thickness of the deposition can vary on Al2O3 scaffold due to the possible difference in the
Figure 3. Absorption spectra of different thicknesses of ZnO nanoshell on alumina substrate.
S1A. One can notice that the reduction in the shell thickness of the ZnO leads to an absorption shift toward blue. The sample with 10 or 15 nm ZnO exhibits an absorption onset at around 405 nm and with decreasing thickness, the onset shifts down to ∼375 nm for a 2 nm ZnO. A clear difference in the onset could be observed even with 1 nm variation in the thickness and it could be attributed to the quantum confinement effect (QC) or scattering. As a consequence of QC, the conduction band tends to move toward vacuum when the thickness of the shell is decreased. Hence, for a given redox potential and recombination rate of electrons, the open-circuit potential of the solar cell is expected to be higher for the thinner layers. It has to be noted that the absolute absorption onset wavelength is higher than the reported values and it could be due to the electronic distortion induced by the underlying Al2O3 susbstrate.35,36 To further confirm the quantum confinement effect, we made 1191
dx.doi.org/10.1021/nl4039955 | Nano Lett. 2014, 14, 1190−1195
Nano Letters
Letter
Figure 4. (A) J−V curves measured at AM1.5G solar illumination for DSC containing photoanodes of different thicknesses of ZnO nanoshell. (B) Transport rate, (C) recombination rate, and (D) collection efficiencies of the corresponding devices are plotted as a function of voltage analyzed using transient decay techniques.
similar ALD ZnO depositions on planar alumina films instead of mesoporous layers. Supporting Information Figure S1B shows the absorption spectra of different thicknesses of ZnO on 20 nm planar Al2O3 substrates. Similar to mesoporous alumina substrates, a trend of shift in the wavelength onset toward blue, is observed when the thickness of the layers is decreased. These nanoshell photoanodes with different thicknesses of ZnO were sensitized with our standard D-π-A dye (Y123, Supporting Information Figure S2A) and devices were made using cobalt(bipyridine)32+/3+ (Supporting Information Figure S2B) redox electrolyte.11,32,37 A mesoporous alumina photoanode without zinc oxide is not expected to show any photovoltaic performance, as the excited dye molecules cannot inject electrons into alumina since latter does not possess any conduction band electronic states below the LUMO level of the dye. The solar cell with 2 nm ZnO exhibits short-circuit current density (JSC), open-circuit potential (VOC) and fill factor (ff), respectively, of 2.5 mAcm−2, 890 mV, and 0.67, leading to a power conversion efficiency (PCE) of 1.5%. When the shell thickness is increased to 3 nm, JSC increased significantly to 6 mAcm−2 and with VOC of 924 mV the cell results in a PCE of ∼4%. Further increase in the thickness enhanced the JSC to a maximum of 8 mA cm−2 with a decline in VOC and ff. Between ZnO shell thickness of 3 and 6 nm, the overall PCE remained at around 4%. The current−voltage curves are presented in Figure 4A and the corresponding photovoltaic data are given in Table 1. The enhancement in current density between 3 and 6 nm ZnO is supported by the improvement in the incident photonto-current collection efficiency (IPCE) spectra shown in Supporting Information Figure S3. To rationalize the observed photovoltaic parameters, dyeloading on the films, transport and recombination rates of the
Table 1. Current−Voltage Characteristics of the DyeSensitized Solar Cells with Different Thicknesses of ZnO on a 2.5 μm Al2O3 Template Measured under AM1.5G Solar Illumination (100 mW/cm2) thickness of ZnO ALD shell (nm)
JSC (mA cm‑2)
VOC (mV)
ff
PCE (%)
2 3 5 6 10
2.5 6.1 7.7 8.1 8.0
890.6 923.9 913.7 903.9 885.0
0.67 0.70 0.54 0.57 0.50
1.5 3.9 3.8 4.2 3.5
photogenerated electrons were investigated. Supporting Information Figure S4 shows the change in the absorption maximum at 463 nm of Y123 dye desorbed from the film in a basic solution of DMF. With increasing thickness, the dye uptake scales down linearly. This reduction in the dye loading could be partly attributed to the reduction in the available surface area as described in our previous reports.28,29 The transport rate of the photogenerated electrons in ZnO was studied using transient photocurrent decay and is displayed in Figure 4B as a function of the voltage. Two distinct type of transport response could be observed: (1) ktrans increases with increasing voltage, in short, when the electron quasi Fermi level moves closer to the conduction band, the transport rate increases steeply. This observation is similar to the transport properties observed in the conventional TiO2 nanoparticles used in the dye-sensitized solar cell and (2) the transport rate increases when shell thickness is increased from 3 to 5 nm and further deposition does not change the parameter significantly.38,39 This result, combined with the dye uptake, explains the evolution of the JSC. At 3 nm, despite higher dye loading the 1192
dx.doi.org/10.1021/nl4039955 | Nano Lett. 2014, 14, 1190−1195
Nano Letters
Letter
current density is probably limited due to the ktrans. However, beyond 5 nm, the current density is governed primarily by the difference in the dye uptake. The evolution of the VOC was studied using transient photovoltage decay technique and the measured recombination rates are presented as a function of their corresponding voltages in Figure 4C.39 A general trend of increasing electron recombination from the ZnO or underlying transparent conducting oxide to the Co3+is observed while moving toward the conduction band of ZnO or at higher voltages. At any given voltage, the recombination rate of electrons is higher for the 3 nm ZnO. The increase in the back flow of electrons could come mainly from the TCO, as underlayer passivation of just 3 nm may not be sufficient to block electron leakage.29 On increasing the thickness to 5 nm, the recombination rate is effectively suppressed. However with further deposition, the krec increases. The reverse trend in the recombination rate observed beyond 6 nm could be due to increased film conductivity or lower dye loading. For a given surface area, the decrease in the dye loading can expose the surface of the photoanode film to the oxidized species in the electrolyte component. This probably has led to an increase in the unwanted back reaction for thicker shell on Al2O3. Even though 3 nm ZnO shows higher back reaction, the open-circuit potential of the corresponding device is higher than in the other devices. The influence of quantum confinement plays an important role for thin layers where the conduction position edge/bottom of the discrete levels above the optical band gap is higher than for the thicker layers as evidenced by the absorption spectra shown in Figure 3. Thus, the upshift in the conduction band for the 3 nm ZnO dominates the recombination resulting in higher VOC of the solar cell. The shift in the conduction band in the dye-sensitized solar cells is studied using charge extraction measurements and the plot is presented in Figure 5. It can be seen that the device possessing 3 nm ZnO overlayer exhibits a negatively shifted (w.r.t. NHE) trap states distribution compared to the 10 nm ZnO overlayer, further confirming the quantum confinement effect in these ZnO films.
Finally the collection efficiency of the photogenerated electrons that cumulatively takes account of the transport and recombination rate was calculated and is given in Figure 4D. For a 3 nm ZnO, due to the low electron transport rate and high recombination rate the collection efficiency is relatively less than in the other devices. One has to note that for a 5 nm ZnO nanoshell the collection efficiency is almost 100% which is a significant factor for achieving record-high efficiency in devices.13,40 Despite higher dye uptake and almost similar collection efficiency, 3 nm ZnO showed low current density. It could be assigned to the low driving force available for the excited state electron injection from the dye to the conduction band of ZnO, due to the observation of quantum confinement effect.41 The electron recombination/transfer properties could not be studied for the device with 2 nm ZnO nanoshell as the transient current involved was below the resolution limit of our Keithley instrument. The low performance of this device, however, is attributed to the sluggish transport of photogenerated electrons within the ultrathin ZnO and to the external contact. Recently, solid-state mesoscopic solar cells based on CH3NH3PbI3 perovskite absorber have shown a significant potential for the production of low cost third generation solar cells. Power conversion efficiencies of these devices reached 15% with standard titanium dioxide photoanodes using organic hole conductor and gold counter electrodes.30,31 In this study, we employed our modified ZnO nanostructures (on a 300 nm Al2O3 scaffold) and for a shell thickness of 5 nm, we reach a high PCE of 7% with an impressive short-circuit current density of 18.5 mAcm−2. The device without any ZnO overlayer exhibited very low PCE of 1.7%. This is probably due to high recombination of photogenerated electrons to perovskite absorber or hole conductors, at the exposed FTO surface. The J−V curves and data are presented in Figure 6, Supporting
Figure 6. J−V curves measured under solar illumination for perovskite absorber mesoscopic solar cells containing photoanodes of different thicknesses of ZnO nanoshell. The values displayed in brackets within the legends give the solar illumination with respect to 1 sun.
Information Figure S5, and Table 2. Even though the ultrathin ZnO overlayers transport current densities of 18.5 mAcm−2, the effect of quantum confinement is not clear. The shift in the conduction band for different ZnO devices is evidenced in Figure 5 for dye-sensitized solar cells. But only a slight or marginal variation in the open-circuit potential is observed for the perovskite absorber based devices and the possible reason
Figure 5. Distribution of density of states measured by charge extraction technique for the DSC photoanodes containing 3 and 10 nm ALD ZnO overlayer. 1193
dx.doi.org/10.1021/nl4039955 | Nano Lett. 2014, 14, 1190−1195
Nano Letters
Letter
Foundation as a part of the 2009 Balzan prize awarded to Michael Grätzel.
Table 2. Current-Voltage Characteristics of the Perovskite Solar Cells with Different Thicknesses of ZnO on a 0.3 μm Al2O3 Template Measured under AM1.5G Solar Illumination (100 mW cm−2) thickness of ZnO ALD shell (nm)
JSC (mA cm‑2)
VOC (mV)
ff
PCE (%)
3 4 5 6
8.4 13.3 18.5 16.4
796.5 800.8 758.5 793.8
0.38 0.47 0.50 0.37
2.6 5.0 7.0 4.8
■
(1) O’Regan, B.; Grätzel, M. Nature 1991, 353, 737−740. (2) Hagfeldt, A.; Boschloo, G.; Sun, L.; Kloo, L.; Pettersson, H. Chem. Rev. 2010, 110, 6595−6663. (3) Gratzel, M. Nature 2001, 414, 338−344. (4) Ito, S.; Murakami, T. N.; Comte, P.; Liska, P.; Grätzel, C.; Nazeeruddin, M. K.; Grätzel, M. Thin Solid Films 2008, 516, 4613− 4619. (5) Fan, J.; Hao, Y.; Cabot, A.; Johansson, E. M. J.; Boschloo, G.; Hagfeldt, A. ACS Appl. Mater. Interfaces 2013, 5, 1902−1906. (6) Law, M.; Greene, L. E.; Johnson, J. C.; Saykally, R.; Yang, P. Nat. Mater. 2005, 4, 455−459. (7) Baxter, J. B.; Aydil, E. S. Appl. Phys. Lett. 2005, 86, 53114. (8) Gubbala, S.; Chakrapani, V.; Kumar, V.; Sunkara, M. K. Adv. Funct. Mater. 2008, 18, 2411−2418. (9) Martinson, A. B. F.; Elam, J. W.; Hupp, J. T.; Pellin, M. J. Nano Lett. 2007, 7, 2183−2187. (10) Mor, G. K.; Shankar, K.; Paulose, M.; Varghese, O. K.; Grimes, C. A. Nano Lett. 2005, 6, 215−218. (11) 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. (12) Zhang, Q.; Dandeneau, C. S.; Zhou, X.; Cao, G. Adv. Mater. 2009, 21, 4087−4108. (13) Grätzel, M. Inorg. Chem. 2005, 44, 6841−6851. (14) Park, S.-S.; Mackenzie, J. D. Thin Solid Films 1995, 258, 268− 273. (15) Bruneaux, J.; Cachet, H.; Froment, M.; Messad, A. Thin Solid Films 1991, 197, 129−142. (16) Westermark, K.; Rensmo, H.; Siegbahn, H.; Keis, K.; Hagfeldt, A.; Ojamäe, L.; Persson, P. J. Phys. Chem. B 2002, 106, 10102−10107. (17) Schölin, R.; Quintana, M.; Johansson, E. M. J.; Hahlin, M.; Marinado, T.; Hagfeldt, A.; Rensmo, H. J. Phys. Chem. C 2011, 115, 19274−19279. (18) Shi, Y.; Zhu, C.; Wang, L.; Zhao, C.; Li, W.; Fung, K. K.; Ma, T.; Hagfeldt, A.; Wang, N. Chem. Mater. 2013, 25, 1000−1012. (19) Keis, K.; Lindgren, J.; Lindquist, S.-E.; Hagfeldt, A. Langmuir 2000, 16, 4688−4694. (20) Quintana, M.; Edvinsson, T.; Hagfeldt, A.; Boschloo, G. J. Phys. Chem. C 2006, 111, 1035−1041. (21) Rensmo, H.; Keis, K.; Lindström, H.; Södergren, S.; Solbrand, A.; Hagfeldt, A.; Lindquist, S.-E.; Wang, L. N.; Muhammed, M. J. Phys. Chem. B 1997, 101, 2598−2601. (22) Willis, R. L.; Olson, C.; O’Regan, B.; Lutz, T.; Nelson, J.; Durrant, J. R. J. Phys. Chem. B 2002, 106, 7605−7613. (23) Bauer, C.; Boschloo, G.; Mukhtar, E.; Hagfeldt, A. J. Phys. Chem. B 2001, 105, 5585−5588. (24) Jiang, C. Y.; Sun, X. W.; Lo, G. Q.; Kwong, D. L.; Wang, J. X. Appl. Phys. Lett. 2007, 90, 263501. (25) Zhang, Q.; Chou, T. P.; Russo, B.; Jenekhe, S. A.; Cao, G. Angew. Chem. 2008, 120, 2436−2440. (26) George, S. M. Chem. Rev. 2009, 110, 111−131. (27) Leskelä, M.; Ritala, M. Thin Solid Films 2002, 409, 138−146. (28) Chandiran, A. K.; Yella, A.; Stefik, M.; Heiniger, L.-P.; Comte, P.; Nazeeruddin, M. K.; Grätzel, M. ACS Appl. Mater. Interfaces 2013, 5, 3487−3493. (29) Chandiran, A. K.; Comte, P.; Humphry-Baker, R.; Kessler, F.; Yi, C.; Nazeeruddin, M. K.; Grätzel, M. Adv. Funct. Mater. 2013, 23, 2775−2781. (30) Lee, M. M.; Teuscher, J.; Miyasaka, T.; Murakami, T. N.; Snaith, H. J. Science 2012, 338, 643−647. (31) Burschka, J.; Pellet, N.; Moon, S.-J.; Humphry-Baker, R.; Gao, P.; Nazeeruddin, M. K.; Gratzel, M. Nature 2013, 499, 316−319. (32) Feldt, S. M.; Gibson, E. A.; Gabrielsson, E.; Sun, L.; Boschloo, G.; Hagfeldt, A. J. Am. Chem. Soc. 2010, 132, 16714−16724.
could be the variation in the recombination rates on different ZnO films. Further studies on the passivation of these ZnO nanomaterials by insulating tunneling layers, to obtain higher VOC, are in progress.42,43 Thus, the present work highlights the efficacy of ultrathin ZnO nanoshells on insulating templates for a prospective high efficiency dye-sensitized solar cell. With a nanoshell thickness between 3 and 6 nm on a 2.5 μm Al2O3 mesoporous scaffold, the maximum power conversion efficiency of 4% was achieved with our standard organic sensitizer and Co2+/Co3+ redox mediator. Further work has to involve the implementation of this nanoshell on thicker alumina layers to efficiently harvest the solar spectrum efficiently, so as to increase the current density of the solar cell. As of now, the deposition of ZnO on thicker alumina templates is limited by the inhomogenous diffusion of ALD precursors across the thickness of the film.28 Solid-state solar cells made with similar quantum-confined ZnO nanostructure led to a power conversion efficiency of 7%. As the ZnO for liquid DSC is deposited at 150 °C, this work can be extended to any arbitrary flexible plastic substrate that can withstand 150 °C, leading to low-temperature fabrication of DSC for portable device applications.
■
ASSOCIATED CONTENT
S Supporting Information *
Complete experimental details, absorption spectra of ZnO on mesoporous and planar alumina films, structure of dye sensitizer and cobalt complex, IPCE spectra of 3 and 6 nm ZnO, dye uptake, and current−voltage curves of a device without ZnO nanoshell. This material is available free of charge via the Internet at http://pubs.acs.org.
■
REFERENCES
AUTHOR INFORMATION
Corresponding Authors
*E-mail: (M.K.N.) mdkhaja.nazeeruddin@epfl.ch. *E-mail: (M.G.) michael.graetzel@epfl.ch. Author Contributions
The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. A.K.C. and M.A.-J. contributed equally. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors acknowledge the financial contribution from EU FP7 project “ORION” (Grant NMP-229036). M.I.D gratefully acknowledges financial support from the Swiss confederation under Swiss Government Scholarship programme. The authors are grateful for the financial support from the Balzan 1194
dx.doi.org/10.1021/nl4039955 | Nano Lett. 2014, 14, 1190−1195
Nano Letters
Letter
(33) Nelson, J. J.; Amick, T. J.; Elliott, C. M. J. Phys. Chem. C 2008, 112, 18255−18263. (34) Tsao, H. N.; Comte, P.; Yi, C.; Grätzel, M. ChemPhysChem 2012, 13, 2976−2981. (35) Bahnemann, D. W.; Kormann, C.; Hoffmann, M. R. J. Phys. Chem. 1987, 91, 3789−3798. (36) Lin, K.-F.; Cheng, H.-M.; Hsu, H.-C.; Lin, L.-J.; Hsieh, W.-F. Chem. Phys. Lett. 2005, 409, 208−211. (37) Tsao, H. N.; Yi, C.; Moehl, T.; Yum, J.-H.; Zakeeruddin, S. M.; Nazeeruddin, M. K.; Grätzel, M. ChemSusChem 2011, 4, 591−594. (38) O’Regan, B. C.; Durrant, J. R. Acc. Chem. Res. 2009, 42, 1799− 1808. (39) Peter, L. M. Phys. Chem. Chem. Phys. 2007, 9, 2630−2642. (40) Chandiran, A. K.; Sauvage, F.; Casas-Cabanas, M.; Comte, P.; Zakeeruddin, S. M.; Graetzel, M. J. Phys. Chem. C 2010, 114, 15849− 15856. (41) Koops, S. E.; O’Regan, B. C.; Barnes, P. R. F.; Durrant, J. R. J. Am. Chem. Soc. 2009, 131, 4808−4818. (42) Chandiran, A. K.; Tetreault, N.; Humphry-Baker, R.; Kessler, F.; Baranoff, E.; Yi, C.; Nazeeruddin, M. K.; Grätzel, M. Nano Lett. 2012, 12, 3941−3947. (43) Chandiran, A. K.; Nazeeruddin, M. K.; Grätzel, M. Adv. Funct. Mater. 2013, n/a−n/a.
1195
dx.doi.org/10.1021/nl4039955 | Nano Lett. 2014, 14, 1190−1195
