Control of charge dynamics through a charge-separation interface for all-solid perovskite-sensitized solar cells.

CHEMPHYSCHEM ARTICLES DOI: 10.1002/cphc.201301153

Control of Charge Dynamics through a Charge-Separation Interface for All-Solid Perovskite-Sensitized Solar Cells Yuhei Ogomi,*[a, e] Kenji Kukihara,[a] Shen Qing,*[b, e] Taro Toyoda,[b, e] Kenji Yoshino,[c, e] Shyam Pandey,[a] Hisayo Momose,[d] and Shuzi Hayase*[a, e] The relationship between the structure of the charge-separation interface and the photovoltaic performance of all-solid dye-sensitized solar cells is reported. This cell is composed of porous a TiO2/perovskite (CH3NH3PbIxCl3x)/p-type organic conductor. The porous titania layer was passivated with Al2O3 or Y2O3 to remove surface traps of the porous titania layer. Both passivations were effective in increasing the efficiency of the solar cell. Especially, the effect of Y2O3 passivation was remark-

able. After passivation, the efficiency increased from 6.59 to 7.5 %. The increase in the efficiency was discussed in terms of the electron lifetime in TiO2, the thermally stimulated current, the measurement of the microwave refractive carrier lifetime, and transition absorption spectroscopy. It was proven that surface passivation resulted in retardation of charge recombination between the electrons in the porous titania layers and the holes in the p-type organic conductors.

1. Introduction Solar cells prepared by printing technology are expected to be candidates for next-generation solar cells. All-solid-state dyesensitized solar cells (DSCs) are printable solar cells. Various allsolid DSCs (S-DSCs) consisting of p-type inorganic, organic, and polymer materials have been reported.[1] Recently, high-efficiency S-DSCs consisting of perovskite-sensitized solar cells (PEROV-S-DSC) were reported. Park and his co-workers reported PEROV-S-DSCs with 9.7 % efficiency that consisted of TiO2/ perovskite (CH3NH3PbI3)/(2,2’,7,7’-tetrakis[N,N-di(4-methoxyphe-

[a] Dr. Y. Ogomi, K. Kukihara, Prof. Dr. S. Pandey, Prof. Dr. S. Hayase Graduate School of Life Science and Systems Engineering Kyushu Institute of Technology 2-4 Hibikino, Wakamatsu-ku Kitakyushu, Fukuoka 808-0196 (Japan) E-mail: [email protected] [email protected] [b] Prof. Dr. S. Qing, Prof. Dr. T. Toyoda Graduate School of Informatics and Engineering University of Electro-Communications 1-5-1 Chofugaoka, Chofu Tokyo 182-8585 (Japan) E-mail: [email protected] [c] Prof. Dr. K. Yoshino Department of Electrical and Electronic Engineering University of Miyazaki 1-1, Gakuen Kibanadai Nishi Miyazaki, 889-2192 (Japan) [d] Dr. H. Momose Center for Semiconductor Research and Development Toshiba Corporation 580-1 Horikawa-cho, Saiwai-ku, Kawasaki-shi Kanagawa 212-8520 (Japan) [e] Dr. Y. Ogomi, Prof. Dr. S. Qing, Prof. Dr. T. Toyoda, Prof. Dr. K. Yoshino, Prof. Dr. S. Hayase CREST (Japan) Science and Technology Agency (JST) 4-1-8 Honcho Kawaguchi Saitama 332-0012 (Japan)

2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

nyl)amino]-9,9’-spirobifluorene) (spiro).[2] The efficiency was increased to about 15 % by Grtzel and his co-workers.[4] Snaith and his co-workers reported TiO2-free perovskite solar cells (PEROV-SOLAR CELLs) with 10.9 % efficiency, in which Al2O3 nanoparticles were employed instead of porous TiO2 layers.[5] Recently, it was reported that these cells have long electron diffusion lengths that reach 1 mm.[6, 7] By using unique properties, PEROV-SOLAR CELLs with planar structures and 15.4 % efficiency were reported.[8] In addition, PEROV-SOLAR CELLs with other organic p-type semiconductors in addition to spiro were reported.[9] Figure 1 shows a representative structure of a perovskite-sensitized solar-cell structure. There are two electroncollection mechanisms, namely, by a TiO2 porous layer and by a perovskite layer that covers Al2O3. In this paper, we take the former model in which electrons are collected by porous TiO2 layers and discuss the relationship between the structure of the charge-generation interface and the performance of the solar cells. It was reported that TiO2 surface states of DSCs consisting of liquid electrolytes (I/I3) are one type of charge-recombination centers, and the charge recombination is retarded by surface passivation of the TiO2 nanoparticles.[10–16] We already reported that the number of surface traps was decreased by passivating the surface of the TiO2 nanoparticles and that by doing so the solar cell efficiency was increased.[14, 15] This prompted us to examine the surface passivation effects on PEROV-S-DSCs. In this paper, surface passivation effects by using Y2O3 and Al2O3 thin layers are discussed.

Experimental Section Cells were fabricated by the following process and materials. Fdoped SnO2-layered glass (FTO glass, Nippon Sheet Glass Co., Ltd) was patterned by using Zn and 6 n HCl aqueous solution. Titanium ChemPhysChem 0000, 00, 1 – 9

&1&

These are not the final page numbers! ÞÞ

CHEMPHYSCHEM ARTICLES

www.chemphyschem.org were released from the trap sites. These electrons were detected as the TSCs. At higher temperatures, the TSC is due to the electrons released from deeper traps, and at lower temperatures, the TSC is associated with electrons released from shallow traps. Then, trap depths (Et) and trap densities (N) were obtained from the temperature and current density at the given temperature by Equations (2) and (3), respectively: 4 T Et ¼ kB Tm ln m b

ð2Þ

in which kB is the Boltzmann constant, Tm is the temperature [K], and b is the programming rate [K s1], N¼ Figure 1. Structure of perovskite-sensitized solar cells (PEROV-S-DSC).

kB T dVoc 1 dt e

ð1Þ

in which kB is the Boltzmann constant, T is the temperature, e is the elementary charge, and Voc is the open-circuit voltage. Thermally stimulated currents (TSCs) were measured by using a TSFETT apparatus (Rigaku) in the same way as that described in our previous report.[14] At a low temperature (180 8C), traps were filled with electrons by exposing ultraviolet light to the substrate. As the temperature of the sample was increased, these electrons 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ð3Þ

dI

diisopropoxide bis(acetylacetonate) solution in ethanol was sprayed at 300 8C onto this patterned FTO glass to prepare compact TiO2 layers. A porous TiO2 layer was fabricated by spin-coating TiO2 paste (PST-18NR, JGC Catalysts and Chemicals, Ltd.) in ethanol (TiO2 paste/ethanol = 1:2.5 weight ratio), followed by heating the substrate at 550 8C for 30 min. Surface passivation with Al2O3 or Y2O3 was performed by dipping the substrate in its precursor solution. The substrate was heated at 500 8C for 30 min. CH3NH3I and PbCl2 were mixed in a 3:1 molar ratio to prepare a 40 % solution of perovskite in N,N-dimethylformamide, and the mixture was spincoated on the surface-passivated substrate. The substrate was heated at 100 8C for 45 min, followed by spin-coating a mixture of 55 mm tert-butylpyridine, 9 mm lithium bis(trifluoromethylsulfonyl)imide salt, and 68 mm spiro. Finally, Ag and Au electrodes were fabricated by vacuum deposition. The photovoltaic performance was evaluated by using an AM1.5G 100 mW cm2 irradiance solar simulator (CEP-2000, Bunkoukeiki, Inc.) with 0.4 0.4 cm mask. The cell size was 0.5 0.5 cm. Solar cells without passivation layers, those with Al2O3 passivation layers, and those with Y2O3 passivation layers are denoted as PEROV-S-DSC, PEROV-S-DSC-Al, and PEROV-S-DSC-Y, respectively. The electron lifetime (tn) was measured by the open-circuit voltage decay (OCVD) method by using Equation (1).[14]

tn ¼

dI 1 dt enP

in which dt is the current per unit time, e is the elementary charge, n is the volume (thickness gap), and P is the porosity. Transient absorption (TA) measurements for characterization of the charge separation and recombination dynamics were performed by the following method. Two TA apparatuses were setup to characterize charge separation (electron injection) and recombination dynamics in the samples. For measuring charge separation dynamics, a titanium/sapphire laser (CPA-2010, Clark-MXR Inc.) with a wavelength of 775 nm, a repetition rate of 1 kHz, and a pulse width of 150 fs was employed. The light was separated into two parts. One part (775 nm light) was used as the probe pulse, and the other part was used to pump an optical parametric amplifier (OPA) (a TOAPS from Quantronix) to generate light pulses with a wavelength tunable from 290 nm to 3 mm. The latter was used as a pump light to excite the sample. In this study, 470 nm light was used for pumping. In the TA setup for measuring charge recombination dynamics, an OPO (Surelite II-10FP) output excited by a Nd:YAG nanosecond pulse laser (Panther, Continuum, Electro-Optics, Inc.) was used. Light (470 nm) with a 5 ns pulse width and 0.5 Hz repetition was employed as the excitation. A fiber-coupled CW semiconductor laser with a wavelength of 658 nm (measurement of charge recombination between electrons in TiO2 and holes in perovskite) or 1500 nm (measurement of charge recombination between electrons in TiO2 and holes in spiro) was used as the prove light. The carrier lifetime was measured by microwave reflection photoconductive decay by using a KOBELCO model LTA-1510EP.

2. Results and Discussion The surface trap density and trap depth of the porous TiO2 layers before and after Y2O3 surface passivation were evaluated by thermally stimulated current, and the results are shown in Figure 2. The conduction band level of titania was 4.0 eV ChemPhysChem 0000, 00, 1 – 9

&2&



www.chemphyschem.org concentrations of Cl were observed. The perovskite layer was prepared by spin-coating the mixture solution of (A) in Equation (4) (3 MeNH3 + I + PbCl2). It is expected that the perovskite material (MeNH3 + PbI3) and the byproduct (MeNH3 + Cl) may form in the solar cell area, as shown in Equation (4). The nonuniform distribution of Cl and Pb suggests that domains of MeNH3 + PbI3 and MeNH3 + Cl may be present separately. 3 MeNH3 þ I þ PbCl2 ! MeNH3 þ PbI3 þ 2 MeNH3 þ Cl ðAÞ

Figure 2. Trap density and depth of the porous titania layer before and after Y2O3 passivation; spiro-OMeTAD = spiro in Figure 1.

ðBÞ

ð4Þ

Figure 5 shows the photovoltaic performance of the solar cells (PEROV-S-DSC, PEROV-S-DSC-Al, and PEROV-S-DSC-Y). The photovoltaic performance of PEROV-S-DSC was as follows: short circuit current (Jsc): 11.81 mA cm2, open-circuit voltage (Voc): 0.79 V, fill factor (FF): 0.71, and efficiency: 6.59 %. After surface passivation with Y2O3 (PEROV-S-DSC-Y), Jsc increased from 11.81 to 16.55 mA cm2 without changing Voc (0.79 V), and 7.53 % efficiency was observed. Voc increased from 0.79 to 0.83 V with a small increase in Jsc from 11.81 to 12.56 mA cm2, and the efficiency increased from 6.59 to 6.96 % by passivating porous TiO2 layers with Al2O3 (PEROV-S-DSC-Al). An increase in the value of Voc is usually observed for cells with a longer electron lifetime. However, in our case of PEROVS-DSC-Y (perovskite-sensitized solar cells passivated with Y2O3), and increase in Voc was not observed. Probably, Y2O3 passivation may shift the TiO2 conduction band to a lower level. Therefore, the increase in Voc brought about by the longer electron lifetime may be compensated by a shift in the titania conduction band to a lower level. It was reported that high efficiency was observed for porous Al2O3-based perovskite solar cells in which the perovskite layer collected the electrons.[5] In our experiment, only a small increase in the efficiency was observed after Al2O3 passivation. Under our experimental conditions, the perovskite layer may

from the vacuum level. Shallow traps were focused because they seriously affect photovoltaic performance. The trap density at 4.2 eV from the vacuum level (0.2 eV under the TiO2 conduction band) was about 1016–17 cm3, and the density gradually decreased to 1011–12 cm3 at 4.7 eV as the energy level became lower. After the surface was passivated with Y2O3, the number of traps decreased to 1014 cm3 at 4.2 eV and to 1011–12 cm3 at 4.7 eV. Y2O3 passivation decreased the trap density by about five digits at the shallow energy level area. Figure 3 shows the structure of the cell and the cross-section bright-field scanning transmission electron microscopy (BFSTEM) image of PEROV-S-DSC-Y. On an FTO layered glass, a compact layer (thickness of 20– 30 nm), a porous TiO2/perovskite/spiro layer (thickness of 500–600 nm), and a spiro layer (thickness of 150 nm) were observed. Figure 4 shows the elemental distribution profile of PEROV-S-DSC-Y. Au, Ag, spiro, porous titania, and compact TiO2 layers were clearly observed. Pb, I, and Cl atoms of the perovskite were dispersed in the porous titania layer from the top to the bottom; however, the distribution was not uniform, and this suggests that the perovskite dye is not uniformly distributed under these experimental conditions. For example, there were some areas in which the high Figure 3. Cell structure and cross-section BF-STEM image of PEROV-S-DSC. 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemPhysChem 0000, 00, 1 – 9

&3&



www.chemphyschem.org These properties of the thin Al2O3 layer prepared by surface passivation may be different from those for nanocrystalline Al2O3 employed for high-efficiency solar cells.[5] Spin-coating conditions, including holding time before spinning; first spinning rate; holding time for the first spinning; second spinning rate; holding time; space of the spinning area; and controlling air flow of the spin-coater head area during spinning, which may affect drying speed of the perovskite layers, may be other factors. Figure 6 shows incident photon to current efficiency (IPCE) curves for PEROV-S-DSC before and after passivation. PEROV-SDSC-Y passivated with Y2O3 had higher IPCE values in the range from 400 to 700 nm than PEROV-S-DSC without passivation.

Figure 4. Elemental distribution profile of PEROV-S-DSC.

have a discontinuous structure and may not be able to collect electrons. Even after Al2O3 passivation, electrons would be collected by the porous titania layers for the cells. The meaning of “continuous” is that the perovskite layer is connected continuously to the electrode (transparent conductive oxide). The surface energy of the porous surface and the pore-size distribution (pore filling), in addition to other factors, may affect the fabrication of the continuous structure of the perovskite layer.

Figure 6. Incident photon to current efficiency (IPCE) curves for PEROV-SDSC before and after passivation. Experimental conditions and abbreviations: see Figure 5.

Figure 5. Photovoltaic performance of all-solid perovskite-sensitized solar cells before and after surface passivation. I–V curves of PEROV-S-DSC with TiO2 under light illumination (TiO2) and in the dark [dark current, TiO2(dark)], I–V curves of PEROV-S-DSC-Al with TiO2 passivated with Al2O3 under light illumination (TiO2/Al2Ox) and in the dark [dark current,TiO2/Al2Ox(dark)], and I–V curves of PEROV-S-DSC-Y with TiO2 passivated with Y2O3 under light illumination (TiO2/Y2Ox) and in the dark [dark current, TiO2/Y2Ox(dark)]. AM1.5G 1 sun 100 mW cm2, cell area 4 mm 4 mm. Measured with a photomask.


Figure 7 shows the relationship between dipping time of the porous TiO2 sheet (surface passivation time) and solar cell performance. At first, the photovoltaic performance increased as the passivation time was increased, and a maximum efficiency was obtained at a passivation time of 10–20 min. After that, the efficiency decreased as the passivation time was increased further. This behavior is often observed for passivation of porous titania layers in liquidtype dye-sensitized solar cells consisting of I/I3 redox species. A thick surface passivation layer would retard charge injection from the perovskite layer to the porous TiO2 layer (route 2 in ChemPhysChem 0000, 00, 1 – 9

&4&



www.chemphyschem.org

Figure 7. Relationship between passivation time and photovoltaic performance. Ref = PEROV-S-DSC with TiO2 under light illumination. With Al2Ox = PEROV-SDSC-Al with Al2O3 passivation. With Y2Ox = PEROV-S-DSC-Y with Y2O3 passivation. AM1.5G 1 sun 100 mW cm2, cell area 4 4 mm. Measured with a photomask.

Figure 8. Working principle for solar cells.

Figure 8). It was difficult to measure the thickness of Y2O3 covering the porous titania. Instead of the porous titania, we prepared a Y2O3 thin layer on a flat compact titanium layer under the same passivation conditions. The thickness of the Y2O3 layer was measured by focus variation with a white light interferometer (Nikon: BW-D501). The thickness of Y2O3 increased as the passivation time was increased. The thickness of Y2O3 was 7 nm after passivation for 10 min and 15 nm after passivation for 20 min. The thickness of the Al2O3 layer was almost the same as that of the Y2O3 layer. These experimental results are explained well by a mechanism in which the electrons are collected by the TiO2 layers (route 3 in Figure 8). Figures 9 and 10 shows Voc decay curves (relationship between Voc and time) and the relationship between electron lifetime and Voc for PEROV-S-DSC and PEROV-S-DSC-Y. If compared at the same Voc, the electron lifetime of PEROV-S-DSC-Y was 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Figure 9. Relationship between Voc and time (Voc decay curve) of PEROV-SDSC before and after Y2O3 passivation measured by the open-circuit voltage decay method: ~ PEROV-DSC and & PEROV-DSC-Y.

five to ten times longer than that of PEROV-S-DSC, which strongly demonstrates that charge recombination between electrons in TiO2 and the holes in the hole-transporting layer (route 5, Figure 8) are retarded by surface passivation of the Y2O3 layer. In addition, the carrier lifetime was measured by microwave reflection photoconductive decay. The carrier lifetime of PEROV-S-DSC was 59 ns and that of PEROV-S-DSC-Y was 134 ns. This also supports the explanation that the Y2O3 layer retards charge recombination. The carrier dynamics were measured by transient absorption spectroscopy. We excited CH3NH3PbClxI3x both on Y2O3 and on TiO2 with a pump wavelength of 470 nm and detected the lifeChemPhysChem 0000, 00, 1 – 9

&5&



www.chemphyschem.org and t2 of the two charge recombination processes were 34 ps [A1/(A1+A2) = 9 %], which is much longer than 10 ns [A2/(A1+A2) = 91 %]. The former can be considered to correspond to the nonradiative recombination of electrons and holes through trap states at CH3NH3PbClxI3x/Y2O3 interfaces. The latter can be considered to be a radiative recombination of electrons and holes with a lifetime longer than 100 ns, as previously reported.[6, 7] Figure 12 shows a normalized TA response of CH3NH3PbClxI3x/Y2O3 with spiro as a hole-transport layer. By fitting Figure 10 to Equation (5), only monoexponential decay with a time constant of 16 ns appeared. Relative to the long

Figure 10. Electron lifetime of PEROV-S-DSC before and after Y2O3 passivation measured by the open-circuit voltage decay method: ~ PEROV-S-DSC and & PEROV-S-DSC-Y.

time of the photoexcited charge carriers and the charge separation dynamics, that is, the hole-transfer time and the electron-injection time, by measuring the TA responses with a probe wavelength (775 nm) that was just at the optical edge of CH3NH3PbClxI3x. Figure 11 shows the normalized TA responses of CH3NH3PbClxI3x/Y2O3. A bleaching signal with a very slow decay can be observed clearly; this corresponds to the slow recombination process of the photogenerated excitons (electron

Figure 12. Normalized TA response of Y2O3/CH3NH3Pb(ClI)3/spiro. The gray solid line represents the fitting result with Equation (5).

lifetime of the TA decay of CH3NH3PbClxI3x/Y2O3, this faster decay process can be considered to originate from photoexcited hole injection from CH3NH3PbClxI3x to spiro. Figure 13 shows normalized TA responses of CH3NH3PbClxI3x on TiO2 substrates with and without Y2O3 surface passivation. The TA responses can be fitted very well to a monoexponential decay equation with time constants of 1.8 and 3.3 ns for the sample without and with Y2O3 surface passivation, respectively.

Figure 11. Normalized TA responses of CH3NH3Pb(ClI)3/Y2O3 and theoretical fitting results of the TA response with a biexponential function [Eq. (5)].

and hole pairs) in CH3NH3PbClxI3x, as there is no electron injection from perovskite to Y2O3, as mentioned above. There were two processes in the TA decay, and the TA signal can be fitted very well to a biexponential function as follows [Eq. (5)]: Y ¼ A1 et=t1 þ A2 et=t2

ð5Þ

in which t1 and t2 are the time constants and A1 and A2 are the contributions from the two components. The time constants t1 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Figure 13. Dependence of normalized TA responses of CH3NH3Pb(ClI)3/TiO2 without and with Y2O3 surface passivation. The theoretical fitting results of the TA responses with a monoexponential decay function with a decay time of 1.8 (without Y2O3) and 3.3 ns (with Y2O3) are also shown. The pump light wavelength was 470 nm and the probe light wavelength was 775 nm.


&6&



www.chemphyschem.org

These decay times are much shorter (about one or two orders of magnitude shorter) than the charge carrier lifetimes (@ ns) of the most photoexcited charge carriers (92 %) in CH3NH3Pb(ClI)3/Y2O3. So the electron injection time tET can be calculated to be about 1.8 and 3.3 ns by using Equation (6): 1 1 1 ¼ tET tPero=TiO2 tPero=Y2 O3

ð6Þ

ly. These results indicate that recombination between electrons in TiO2 and the holes in CH3NH3Pb(ClI)3 can be suppressed greatly by Y2O3 surface passivation. Figure 15 shows the TA responses of CH3NH3Pb(ClI)3/spiro without and with Y2O3 surface passivation measured with a pump light wavelength of 470 nm and a probe light wavelength of 1500 nm, which monitors the holes in spiro. For the

in which tPero=TiO2 and tPero=Y2 O3 are the TA signal decay times of perovskite on TiO2 and on Y2O3 substrates, respectively. Then, we detected the recombination dynamics in CH3NH3PbClxI3x/TiO2 with and without spiro, including that between the electrons in TiO2 and the holes in CH3NH3PbClxI3x, and that between the electrons in TiO2 and the holes in spiro by using nanosecond-TA. Figure 14 shows the TA responses of CH3NH3Pb(ClI)3/TiO2 without and with Y2O3 surface passivation, which were measured with a pump light wavelength of 470 nm and a probe light wavelength of 658 nm. As shown in Figure 12, the TA decays can be fitted very well with a monoexponential function with time constants of 0.14 and 1 ms for the samples without and with Y2O3 surface passivation, respective-

Figure 15. TA responses of TiO2/CH3NH3Pb(ClI)3/spiro without (a) and with (b) Y2O3 surface passivation. The responses were measured with a pump light wavelength of 470 nm and a probe light wavelength of 1500 nm. The solid lines represent the fitting results with a monoexponential function with a decay time of 60 (a) and 180 ms (b), respectively.

Figure 14. TA responses of CH3NH3Pb(ClI)3/TiO2 without (a) and with (b) Y2O3 surface passivation. The responses were measured with a pump light wavelength of 470 nm and a probe light wavelength of 658 nm. The solid line represents the fitting result with a monoexponential function with a decay time of 0.14 (a) and 1 ms (b), respectively.


sample without spiro, no TA signal was observable with a 1500 nm probe wavelength. However, for the sample with spiro, an absorption signal was very clearly observed, which was fitted very well with a monoexponential decay function with a time constant of 60 (without Y2O3) and 180 ms (with Y2O3). On the basis of the above results, the photoexcited chargecarrier relaxation and transfer dynamics of CH3NH3Pb(ClI)3 on TiO2 without and with Y2O3 surface passivation are illustrated in Figure 16. Perovskite had two excitation lifetime components, namely, 40 ps (9 %) and one longer than 1 ns (91 %). Electron injection (route 2) from perovskite to titania occurred ChemPhysChem 0000, 00, 1 – 9

&7&



www.chemphyschem.org Acknowledgement I would like to thank Prof. K. Nazeeruddin for discussions on perovskite reactions. Keywords: dye-sensitized solar cells · electron lifetimes · energy conversion · sensitizers · surface passivation [1] A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo, H. Pettersson, Chem. Rev. 2010, 110, 6595 – 6663. [2] H.-S. Kim, C.-R. Lee, J.-H. Im, K.-B. Lee, T. Moehl, A. Marchioro, S.-J. Moon, R. Humphry-Baker, J.-H. Yum, J. E. Moser, M. Grtzel, N.-G. Park, Sci. Rep. 2012, 2, 591. J.-H. Noh, S.-H. Im, J.-H. Heo, T.-N. Mandal, S. I. Seok, Nano Lett. 2013, 13, 1764 – 1769. J. Burschka, N. Pellet, S.-J. Moon, R. Humphry-Baker, P. Gao, M. K. Nazeeruddin, M. Grtzel, Nature 2013, 499, 316 – 319. M. M. Lee, J. Teuscher, T. Miyasaka, T. N. Murakami, H. J. Snaith, Science 2012, 338, 643 – 647. G. Xing, N. Mathews, S. Sun, S. S. Lim, Y. M. Lam, M. Grtzel, S. Mhaisalkar, T. C. Sum, Science 2013, 342, 344 – 347. S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou, M. J. P. Alcocer, T. Leijtens, L. M. Herz, A. Petrozza, H. J. Snaith, Science 2013, 342, 341 – 344. M. Liu, M. B. Johnston, H. J. Snaith, Nature 2013, 501, 395 – 398. N. J. Jeon, J. Lee, J. H. Noh, M. K. Nazeeruddin, M. Grtzel, S. I. Seok, J. Am. Chem. Soc. 2013, 135, 19087 – 19090. A. Kay, M. Grtzel, Chem. Mater. 2002, 14, 2930 – 2935. Y. Diamant, S. Chappel, S. G. Chen, O. Melamed, A. Zaban, Coord. Chem. Rev. 2004, 248, 1271 – 1276. E. Palomares, J. N. Clifford, S. A. Haque, T. Lutz, J. R. Durrant, J. Am. Chem. Soc. 2003, 125, 475 – 482. F. Fabregat-Santiago, J. Bisquert, L. Cevey, P. Chen, M. Wang, S. M. Zakeeruddin, M. Grtzel, J. Am. Chem. Soc. 2009, 131, 558 – 562. Y. Noma, T. Kado, D. Ogata, Y. Hara, S. Hayase, Jpn. J. Appl. Phys., Part 1 2008, 47, 505 – 508. F. Fabregat-Santiago, J. Garca-CaÇadas, E. Palomares, J. N. Clifford, S. A. Haque, J. R. Durrant, G. Garcia-Belmonte, J. Bisquert, J. Appl. Phys. 2004, 96, 6903 – 6907. Y. Ogomi, S. Sakaguchi, T. Kado, M. Kono, Y. Yamaguchi, S. Hayase, J. Electrochem. Soc. 2006, 153, A2294 – A2297.

Figure 16. Carrier dynamics of perovskite-sensitized solar cell before and after Y2O3 passivation.

on the timescale of 1.8 ns. The injection (route 2) became slightly slow (3.3 ns) after Y2O3 passivation. The charge recombination between electrons in TiO2 and the spiro layer (route 5) occurred on the timescale of 0.14 ms before passivation. The rate became slower (180 ms) after Y2O3 passivation. Charge recombination (route 4) between electrons in the porous titania layer and oxidized perovskite occurred on a timescale that was shorter than 1 ms. After Y2O3 surface passivation, the charge recombination rate became slower (1 ms). These results strongly demonstrated that the Y2O3 thin layer retards the charge recombination of routes 5 and 4. These results support well the explanation that charge recombination between electrons in the TiO2 layer and the holes in the spiro layer (route 5) as well as that between electrons in the TiO2 layer and oxidized perovskite (route 4) was retarded by surface passivation of the porous titania layer.

[3] [4] [5] [6] [7]

[8] [9] [10] [11] [12]

3. Conclusions

[13]

It was proven that the performance of perovskite-sensitized solar cells was improved by Y2O3 passivation, which decreased the shallow trap density from 1017 to 1014 cm3. A longer charge lifetime after passivation was observed by open-circuit voltage decay and microwave reflection photoconductive decay and through studying the carrier dynamics by using transient spectroscopy. Control of charge-generation interfaces is crucial for increasing the photovoltaic performance of allsolid perovskite-sensitized solar cells.

[14]


[15]

[16]

Received: December 4, 2013 Revised: January 27, 2014 Published online on && &&, 2014


&8&


ARTICLES Y. Ogomi,* K. Kukihara, S. Qing,* T. Toyoda, K. Yoshino, S. Pandey, H. Momose, S. Hayase* && – && Control of Charge Dynamics through a Charge-Separation Interface for AllSolid Perovskite-Sensitized Solar Cells Dye cell, dye! Carrier dynamics of perovskite sensitized solar cells with and

without Y2O3 layer are reported.



&9&


Charge Dynamics in Solution-Processed Nanocrystalline CuInS2 Solar Cells.

Dynamic interface charge governing the current-voltage hysteresis in perovskite solar cells.

Charge generation mechanism in organic solar cells.

HTM interface.

Probing Charge Transfer and Hot Carrier Dynamics in Organic Solar Cells with Terahertz Spectroscopy.

Interface Engineering through Atomic Layer Deposition towards Highly Improved Performance of Dye-Sensitized Solar Cells.

β Subunit Interface of a Photochemical Ribonucleotide Reductase.

Models of charge pair generation in organic solar cells.

Photovoltaics. Interface engineering of highly efficient perovskite solar cells.

Hot photocarrier dynamics in organic solar cells.

Ultrafast charge transfer in operating bulk heterojunction solar cells.

Influence of counter-anions during electrochemical deposition of ZnO on the charge transport dynamics in dye-sensitized solar cells.

Heterojunction PbS nanocrystal solar cells with oxide charge-transport layers.

A quantitative model for charge carrier transport, trapping and recombination in nanocrystal-based solar cells.

Charge transport limitations in perovskite solar cells: The effect of charge extraction layers.

Morphology control of the perovskite films for efficient solar cells.

Interplay of nanoscale domain purity and size on charge transport and recombination dynamics in polymer solar cells.

Molecular dynamics study of salt-solution interface: solubility and surface charge of salt in water.

Ultrafast charge separation dynamics in opaque, operational dye-sensitized solar cells revealed by femtosecond diffuse reflectance spectroscopy.

Interface Engineering of Organic Schottky Barrier Solar Cells and Its Application in Enhancing Performances of Planar Heterojunction Solar Cells.

Antiferroelectric Nature of CH3NH3PbI3-xClx Perovskite and Its Implication for Charge Separation in Perovskite Solar Cells.

Design and application of carbon nanomaterials for photoactive and charge transport layers in organic solar cells.

A nanoscale study of charge extraction in organic solar cells: the impact of interfacial molecular configurations.

Charge-induced equilibrium dynamics and structure at the Ag(001)-electrolyte interface.