Optimized CdS quantum dot-sensitized solar cell performance through atomic layer deposition of ultrathin TiO₂ coating

Abstract

Here we demonstrate a CdS quantum dot (QD) sensitized solar cell with significantly enhanced stability and depressed recombination in I⁻/I₃⁻ electrolyte. The CdS QDs were deposited in a mesoporous TiO₂ film using chemical bath deposition. Following the coating of an ultrathin TiO₂ protection layer using atomic layer deposition (ALD), the performance and stability of CdS QD sensitized solar cells were pronouncedly enhanced. Current–voltage measurements and electrochemical impedance spectroscopy analysis show that the ALD-TiO₂ coating slows down the charge recombination and protects the QDs from the photocorrosion of the I⁻/I₃⁻ electrolyte. A systematic study shows that a ∼2 nm ALD-TiO₂ layer can best promote solar cell efficiency by balancing protection with charge transfer. We have obtained a photoconversion efficiency of 1.41% and stable operation (no degradation in photocurrent) for at least 20 min under full sun illumination on CdS QD solar cells. The photocurrent gradually decreased down to ∼70% of its original value after 1 h of operation.

---

Introduction

Quantum dot sensitized solar cells (QDSSCs) have been studied extensively because they are potential low-cost alternatives to existing crystalline silicon cells. Semiconductor QDs such as CdS,¹ CdSe,² PbS,³ PbSe,⁴ InAs,⁵ and InP⁶ have been used to sensitize metal oxides, like nanocrystalline (nc) TiO₂ or SnO₂, by absorbing the most intense regions of the solar spectrum.⁷ In a recent breakthrough, a photoconversion efficiency (PCE) of 6% was achieved using PbS QDs.⁸ Unlike their dye counterparts, most QDs suffer from photodegradation when used with I⁻/I₃⁻ in a photoelectrochemical cell.⁹ In the QDSSC, the optimum redox couple was found to be sulfide/polysulfide (S²⁻/S_n²⁻);^10,11 however, sulfides react very strongly with platinum, the catalyst of the counter electrode (CE), and can poison the surface, thereby reducing the catalytic activity.¹² Alternative CE materials for polysulfide solution, such as CoS, Cu₂S, NiS, carbon, and PbS, have been investigated.¹³ To overcome the attack of iodine, Zaban and co-workers have designed a strategy to employ a TiO₂ barrier layer over CdS QDs using electrophoretic deposition.¹⁴ With the amorphous TiO₂ passivation shell, they obtained a CdS QDSSC that was stable for 120 s in I⁻/I₃⁻ electrolyte. However, control over the TiO₂ layer thickness, which is much needed for further optimization of the process, was quite limited by the means of electrophoretic deposition used in the study.

Atomic layer deposition (ALD) has been used to introduce extremely thin and conformal coating due to its unique excellent step coverage, atomic-scale thickness control, exceptional composition control for nanostructures, and excellent uniformity over large areas. Researchers have grown metal oxide materials like TiO₂ and Al₂O₃ by ALD to improve the performance of dye-sensitized solar cells;^15,16 for example, the Cao group introduced an ALD-TiO₂ film that suppressed surface charge recombination.¹⁵ An ALD-Al₂O₃ film employed by the Hupp group was found to passivate surface states.¹⁶ For a QDSSC, the passivation layer can be neither too thick, which would impede the charge transfer, nor too thin, which may allow the electrolyte to penetrate through the layer. ALD is especially suitable for depositing a TiO₂ thin film as the passivation layer for a stable QDSSC, due to its ultrafine thickness control.

In this study, different thicknesses of ultrathin TiO₂ layers were deposited on the surface of CdS QDSSCs using ALD. In addition, once the intended thickness was reached, the ultrathin TiO₂ coating was subsequently annealed to form a crystalline phase, which has better stability and electronic properties than amorphous TiO₂. Photoanodes with different TiO₂ coating thicknesses were systematically examined using electrochemical impedance spectroscopy (EIS) and photocurrent–voltage measurements. An intermediate coating thickness was found to be optimal, by balancing the protection effect and the charge transfer efficiency. A power efficiency of 1.4% and stability of 20 min were achieved as a result of ultrafine thickness control of TiO₂ coating by ALD.

Experimental

Equipment

A field-emission scanning electron microscope (SEM) (Hitachi S 4800) equipped with a Bruker Quantax energy-dispersive X-ray spectrometry (EDS) system was used to observe the morphology of TiO₂–CdS structures and element distribution; the SEM renders a resolution of 1.4 nm at 1 kV acceleration voltage. High-resolution transmission electron microscopy (HRTEM) characterization of TiO₂–CdS was carried out using a Hitachi H 9000 NAR TEM, which has a point resolution of 0.18 nm at 300 kV in the phase contrast HRTEM imaging mode. The EIS measurements were conducted using an electrochemical workstation (CH Instruments 600D).

Preparation of QDSSCs

Mesoporous TiO₂ films were prepared by doctor-blade casting of a titania paste (1.0 g Degussa P25 TiO₂ powder, 50 μl titanium(IV) butoxide, 2 ml isopropanol) onto fluorine-doped tin oxide (FTO, Solaronix) coated glass substrates with 12 Ω □⁻¹ sheet resistance. A TiO₂ blocking/protection layer was grown on all FTO glass for 1,000 ALD cycles before applying mesoporous TiO₂. The detailed ALD process is described below. After drying in air, the TiO₂–FTO glass was sintered at 500 °C for 30 min in air.

The CdS QDs were assembled on TiO₂ films using chemical bath deposition (CBD).¹⁷ After cooling down, the TiO₂–FTO glass was transferred immediately to a Teflon-lined stainless steel autoclave (100 ml) containing Cd(CH₃COO)₂·2H₂O (0.110 g, 98%, Sigma Aldrich) and dimethyl sulfoxide (DMSO, 40 ml). The autoclave was tightly sealed and reacted at 180 °C for 12 h. The obtained anodes were then rinsed extensively with DI water and isopropanol and dried in air. The resulting anodes were then annealed in Ar flow (1.0 liter min⁻¹, lpm) at 500 °C for 30 min.

The ultrathin TiO₂ protection layers were coated on the TiO₂–CdS photoanodes at 200 °C using the Cambridge NanoTech Savannah S100 ALD System. Titanium isopropoxide (Ti(OⁱPr)₄) and H₂O were used as precursors, with high-purity nitrogen as a carrier and purging gas. The reactants were alternatively pulsed and purged in the following sequence: pulsing H₂O for 0.3 s, purging for 6 s, pulsing Ti(OⁱPr)₄ for 0.015 s, purging for 4 s. The thicknesses of the TiO₂ layers were measured with a Horiba spectroscopic ellipsometer to be 1.5 ± 0.08 nm, 2.2 ± 0.09 nm, and 3.2 ± 0.12 nm for coatings of 50 cycles, 100 cycles, and 150 cycles, respectively. The measurements of thickness were carried out on silicon wafers coated with a TiO₂ thin film under the same conditions used for the mesoporous photoanodes. The same procedure was used for the deposition of the 1000-cycle blocking layer; the TiO₂-coated films were annealed at 500 °C in Ar (1.0 lpm) for 30 min to increase crystallinity.

The Pt/FTO counter electrodes (CEs) were prepared by drop casting a 0.005 M H₂PtCl₆ aqueous solution on FTO glass. After drying in air at room temperature, the CEs were elevated to 500 °C in air and maintained for 30 min.

A hot-melt sealing foil (25 microns, Meltonix 1170-25, Solaronix) was sandwiched between the photoanode and CE. Light pressure was applied on the anode–spacer–cathode assembly on a hot plate at 150 °C to seal the cell. An I⁻/I₃⁻ electrolyte (Iodolyte AN-50, Solaronix) was introduced into the cells by capillary force. No further sealing process was adopted.

Photovoltaic characterization

The photocurrent–voltage (J–V) characteristics were obtained under simulated solar illumination (solar simulator, Newport, 94021A) at one sun (AM 1.5G, 100 mW cm⁻²) using a Keithley 2420 source meter equipped with a calibrated Si-reference cell (Oriel, P/N 91150V). The incident-photon-to-electron conversion efficiency (IPCE) of solar cells was analyzed using a system consisting of a 300 W Xenon lamp simulated light source (Newport 66902), a monochromator (Newport cornerstone 74125), and a radiometer (Newport Merlin 70104). The photon flux of light incident on the samples was calibrated using a silicon photodiode (Newport 70356). Measurements were typically made at 10 nm wavelength intervals between 350 and 1100 nm.

Electrochemical characterization

The EIS measurements were conducted under AM 1.5G simulated solar illumination at 100 mW cm⁻². The same solar cells were measured for EIS immediately after J–V characterization. EIS spectra were recorded over a frequency range of 100 kHz–10 mHz. The applied bias voltage and ac amplitude were set at open circuit voltage (V_OC) of the cells and at 5 mV, respectively.

Results and discussion

Structure of photoanodes

The photoanodes of QDSSCs were fabricated by chemical bath deposition of CdS QDs on mesoporous TiO₂ nanocrystal films. Conformal ultrathin TiO₂ films were coated on the CdS–TiO₂ nanocomposite using ALD. Typically, the photoanodes were deposited for 50 cycles, 100 cycles, and 150 cycles, denoted as 50c, 100c, and 150c, respectively. The thicknesses of the TiO₂ coating layers were measured with an ellipsometer as 1.5 ± 0.08 nm, 2.2 ± 0.09 nm, and 3.2 ± 0.12 nm for the 50c, 100c, and 150c cases, respectively. Comparisons were made among three differently coated solar cells and a non-coated solar cell, and the schematic diagrams of coated and non-coated QDSSCs are shown in Fig. 1a and b. The thickness of the CdS–TiO₂ photoanode film is 15–20 μm, confirmed by a cross-sectional SEM image (Fig. 1c). EDS reveals a homogeneous CdS distribution across the mesoporous TiO₂ film (Fig. 1d). The ultrathin ALD-TiO₂ coating can be clearly observed in an HRTEM image, as shown in Fig. 2, in which a thin layer of TiO₂ (∼2 nm) can be discerned covering the CdS–TiO₂ nanostructure. The observed thickness of the 100c ALD-TiO₂ agrees well with the ellipsometer measurement of 2.2 ± 0.09 nm.

Fig. 1 Schematics of the CdS QD-sensitized nanocrystalline TiO₂ electrode (a) without and (b) with an ALD-TiO₂ protection layer. (c) Cross-sectional SEM image of the CdS QD-sensitized mesoporous TiO₂ film with 100c ALD-TiO₂ coating supported on FTO glass. (d) EDS analysis of Ti and Cd along the white arrow in (c).

Fig. 2 HRTEM image of 100c ALD-TiO₂-coated CdS–TiO₂ nanocomposite. TiO₂ and CdS particles are outlined by dashed lines.

Electrochemical properties

EIS analysis revealed noticeable differences in charge transfer resistance at the CdS–ALD-TiO₂–electrolyte interfaces with different TiO₂ coating thicknesses. Fig. 3a shows EIS Nyquist plots of coated QDSSCs; they all exhibit Gerischer impedances, indicating low charge collection efficiencies.¹⁸ Apart from the components corresponding to the electrolyte and the counter electrode, the EIS Nyquist plot ranging from 100 kHz to 1 Hz is of particular interest because it is directly related to the charge transfer at the CdS–ALD-TiO₂–electrolyte interface.^15,19 The diameter of the Gerischer arc represents the active layer resistance of the photoanode, R_AL = (R_recR_tr)^1/2, where R_tr is the total transport resistance in the mesoporous TiO₂ film and at the CdS–ALD-TiO₂–electrolyte interface, and R_rec is the recombination resistance across the surface of the nanoporous solid.¹⁸ The Gerischer impedance happens when the electron diffusion length is much smaller than the thickness of the active layer, that is, R_tr ≫ R_rec. It is obvious that R_AL for the 100c case is appreciably lower than that of the 150c. A better charge transfer can be achieved at the CdS–100c-ALD-TiO₂–electrolyte interface since the ALD-TiO₂ layer is simultaneously serving as a protection layer and a thinner charge transfer barrier. For the similar interface of the 150c case, although the ALD-TiO₂ layer is thicker and better for protection, it hinders the hole tunneling and displays slower recombination and poor charge transfer; however, the ALD-TiO₂ coating thickness cannot be further reduced such that the effective protection of the CdS against the electrolyte is not sacrificed. A rapid bleaching was observed in the 50c cell for both the QDs and the electrolyte, that is the redox I⁻/I₃⁻ couple concentration decreased and the QDs were dissolved, thus resulting in a slow charge transfer (50c, Nyquist plot).

Fig. 3 Electrochemical impedance spectra of the three differently coated cells (-Δ- 50c, -□- 100c, and -○- 150c) under illumination of one sun (AM 1.5G, 100 mW cm⁻²) at open circuit voltage. (a) Nyquist plots, frequencies of peak points are indicated; (b) Bode phase plots.

The charge recombination in the solar cells can be evaluated by EIS. As proposed by Kern et al.²⁰ and Bisquert,²¹ the “recombination” reflected by the EIS represents reaction of the electrons either on the conduction band or in the trap states of TiO₂ with oxidized redox species I₃⁻ in a dye-sensitized solar cell. For QDSSCs, the QDs introduce additional surface states, which make the recombination more complicated. However, the EIS analysis is still effective since the recombination is generally defined as the back reaction of electrons. The effective rate constant for recombination k_eff can be estimated with

where ω_max is the frequency at the peak point of the EIS Nyquist plots (as indicated in Fig. 3a).²² Effective lifetime of electrons is thus defined as

As shown in Fig. 3a, the peak frequencies ω_max decrease from 50c to 150c. The recombination rate is lower for the anode with the thicker coating, which is calculated using eqn (1). This result clearly shows that the ALD coating passivates the surface states or trap states. A TiO₂ shell structure is often employed to suppress recombination, which is usually caused by surface traps (defect states) and back electron transfer from mesoporous TiO₂ to the electrolyte.²³ Also, the passivation effect is presented directly by the EIS Bode phase plots. The characteristic peak shifts to a lower frequency with increasing ALD coating thickness, suggesting an increase in electron lifetime (Fig. 3b). It should be noted that frequencies at characteristic peaks in the Bode phase plots (Fig. 3b) are slightly different from peak frequencies in the Nyquist plots (Fig. 3a). The lifetimes of electrons calculated with eqn (2), as well as the rate constants of recombination, are listed in Table 1. We observed the same trend on additional batches of solar cells, which is shown in the ESI, Fig. S1.†

Table 1 Solar cell parameters of the CdS quantum dot sensitized mesoporous TiO₂ electrode with and without an ALD-TiO₂ protection layer

		J SC (mA cm^−2)	V OC (mV)	FF	η (%)	k eff (s^−1)	τ eff (ms)
Non-coated	1st	1.47	572	0.71	0.6	—	—
	2nd	1.36	449	0.13	0.08
	3rd	1.32	468	0.13	0.08
50c	1st	1.69	563	0.5	0.47	84.9	11.8
	2nd	1.40	543	0.53	0.41
	3rd	1.68	554	0.48	0.45
100c	1st	3.16	668	0.56	1.19	39.3	25.4
	2nd	3.02	667	0.57	1.15
	3rd	2.97	667	0.57	1.14
	Al-back	3.55	673	0.59	1.41
150c	1st	1.39	668	0.62	0.58	22.1	45.2
	2nd	1.52	660	0.60	0.60
	3rd	1.54	660	0.59	0.60

---

Photovoltaic performances of QDSSCs

A comparison of the performance of CdS QDSSCs with various ALD-TiO₂ coating thicknesses is shown in Fig. 4. Fig. 4a and b show the J–V characteristics of the 50c, 100c, and 150c cells, under illumination and in the dark, respectively, using I⁻/I₃⁻ as a redox electrolyte. Each cell was tested three times at 1 min intervals, and then measured in the dark. Under illumination (Fig. 4a), the performance of the 100c cell is significantly better than that of the other cells. Insignificant degradation was observed during the J–V test, indicating a relatively good photo-stability. Table 1 summarizes the parameters of three measurements of all the cells, and a power conversion efficiency (η) of 1.19% was achieved by the 100-cycle coating. Based on the three measurements, the open circuit voltages V_OC,100c ∼ V_OC,150c > V_OC,50c > V_OC,non were observed. Carrier recombination is considered as one of the factors for the reduction in open circuit voltage.^11,15,18 Although a thicker coating can favor the V_OC due to less recombination, the 150-cycle coating hinders charge separation and leads to hole accumulation. Based on our observation, the 100-cycle coating promotes the V_OC most by balancing the surface passivation and charge separation.

Fig. 4 Characterization of the CdS-QD sensitized TiO₂ solar cells with and without the ALD-TiO₂ protection layer. J–V measurements of three differently coated cells (50c, 100c, and 150c) (a) under illumination of one sun (AM 1.5G, 100 mW cm⁻²) and (b) under dark conditions. For the curves measured under illumination, each cell was measured three times: solid line (first time), dashed line (second time), and dash-dotted line (third time). (c) J–V measurement of a non-coated cell under illumination of one sun (AM 1.5G, 100 mW cm⁻²) three times. (d) IPCE of the three differently coated cells.

The fill factor FF_150c > FF_100c > FF_50c > FF_non is a direct indication of the slower recombination for thicker coatings. Although a larger series resistance (resistance of the TiO₂ coating) usually leads to a smaller FF, it is clear that the FFs here are mainly affected by recombination control.¹⁸Fig. 4b shows that electron transfer (or dark current) from the electrode with a thicker coating into the electrolyte at forward bias is significantly reduced compared to the electrode with a thinner coating. This result is directly correlated with the reduced recombination of electrons from the mesoporous electrode into the redox electrolyte caused by the ALD-TiO₂ shell, which is consistent with the EIS analysis. In contrast, a non-coated cell degraded quickly during the three consecutive measurements as the QDs dissolved quickly in the I⁻/I₃⁻ electrolyte (Fig. 4c). Considering all factors, an intermediate coating thickness is optimal to achieve slow recombination, stability against corrosion, and fast charge transfer in the nc-TiO₂–CdS–ALD-TiO₂ structure; as a result, the highest short circuit current density J_SC and η were obtained for the 100c cell. This conclusion is based on observations of multiple batches of solar cells. The J–V characteristics of additional solar cells are shown in the ESI, Fig. S2.† The incident photon-to-current efficiency (IPCE) results show the same trend as the J–V characteristics (Fig. 4d)—the better the cell was protected, the smaller the loss of charge during the photoconversion.

We also noticed that the TiO₂–QD film did not absorb 100% incident light. Increasing the TiO₂–QDs film thickness will not lead to improved light absorption because a thicker film will lower the charge collection efficiency, as suggested by the EIS result. A reflecting aluminum foil was placed at the back of the 100c cell (100c Al-back) for a further test, and a power efficiency of η = 1.41% was achieved. Other solar cell parameters are listed in Table 1. We note that the Cd²⁺ ion concentration (Cd : Ti = 1 : 4) in our anodes produced using CBD is lower than that reported by Zaban et al. (Cd : Ti ∼ 1 : 2).¹⁴ Lower CdS loading will lead to a lower photoelectron concentration, which is a direct reason for a large R_tr. It also contributes to a low IPCE as many incident photons cannot be captured by CdS. Additionally, only a fraction of the visible light was utilized by CdS QDs because of their relatively large band gap (2.4 eV). This issue can be overcome by replacing CdS QDs with other low band gap QDs, such as CdSe, PbS, PbSe, or a combination of them.

The photo-stability of the ALD-TiO₂ protected electrode was tested under periodic on–off one-sun illumination in the presence of the I⁻/I₃⁻ electrolyte. The short circuit current of a 100c cell was measured as a function of time with shuttered light at 12 s intervals. Fig. 5a shows the reproducible photo-current of the coated photoanode. A long-term stability test of a 100c cell was carried out under the illumination of AM 1.5G, 100 mW cm⁻². The J_SC was recorded for over 1 h (Fig. 5b); the current was very stable during the first 20 min, and then gradually decreased down to ∼70% of its original value in the next 40 min. After the measurements, the photoanode (originally yellow) was found slightly bleached (light yellow). Nevertheless, such a stability is much better than what has been reported by Zaban et al.¹⁴

Fig. 5 Photo-stability test of a 100c solar cell. (a) Short circuit current density as a function of time using periodic illumination intervals (12 s interval length). (b) Short circuit current versus time of a 100c solar cell under illumination of one sun (AM 1.5G, 100 mW cm⁻²) over 1 h.

An Al₂O₃ coating was introduced to improve the performance of dye-sensitized solar cells,¹⁶ and as a comparison we also made a similar Al₂O₃ coating to protect CdS QDs using ALD. The J_SC was as low as 0.33, 0.31, and 0.24 mA cm⁻² when the thickness of the Al₂O₃ layer was 1.3, 1.9, and 2.6 nm, respectively. Obviously, this non-conductive coating limited the photocurrent to a very low level. A thinner Al₂O₃ coating could reduce the series resistance and increased the J_SC to 2.8 and 1.28 mA cm⁻² for coating thicknesses of 2.6 and 3.9 Å, respectively. However, QDs with an extremely thin coating (e.g. 1.3 Å) were even less stable. For example, the photoanode with the 1.3 Å ALD-Al₂O₃ coating degraded quickly during the test and the resulting J_SC was as low as 0.29 mA cm⁻². Detailed J–V curves are shown in the ESI (Fig. S3 and S4).†

Conclusions

ALD can introduce an ultrathin protection layer of TiO₂ to passivate and enhance the stability of CdS QD sensitized solar cells. The ALD-TiO₂ layer slows down charge recombination and increases V_OC and FF; however, ALD-TiO₂ coating that is too thick may hinder the hole tunneling, resulting in a low charge separation efficiency. And a protection layer that is too thin cannot effectively provide the protection effect. A systematic study shows that a ∼2 nm ALD-TiO₂ layer, corresponding to 100 ALD cycles, can best promote solar cell efficiency by balancing the protection and charge transfer. An ALD-Al₂O₃ coating was employed to compare with the TiO₂ coating. But the ALD-Al₂O₃ must be thinner to obtain a sufficiently high photocurrent because of its insulating nature. As a result, the protection effect of ALD-Al₂O₃ is worse than that of ALD-TiO₂. For future work, a deposition process with both chemical stability and a no hole-transfer barrier is desired for photoanode coating in QDSSCs. The ALD coating method is applicable not only to the current solar cell, but also to other QDSSC systems, such as solar cells sensitized with CdSe, PbS, and PbSe QDs. Therefore, this study can potentially guide the design of QDSSCs with enhanced performance.

Acknowledgements

This work was financially supported by the NSF (CMMI-1001039) and DOE (DE-EE0003208). SEM/EDS and TEM were performed in the UWM Electron Microscope Laboratory and UWM HRTEM Laboratory, respectively. We thank M. Gajdardziska-Josifovska for providing TEM access, D. Robertson for technical support with TEM, H. A. Owen for technical support with SEM.

References

1. (a) C.-H. Chang and Y.-L. Lee, Appl. Phys. Lett., 2007, 91, 053503; (b) S. Hotchandani and P. V. Kamat, J. Phys. Chem., 1992, 96, 6834; (c) S. Lin, Appl. Phys. Lett., 2007, 90, 143517; (d) P. Ardalan, T. P. Brennan, H. B. R. Lee, J. R. Bakke, I. K. Ding, M. D. McGehee and S. F. Bent, ACS Nano, 2011, 5, 1495; (e) W. T. Sun, Y. Yu, H. Y. Pan, X. F. Gao, Q. Chen and L. M. Peng, J. Am. Chem. Soc., 2008, 130, 1124; (f) L. Sheeney-Haj-Khia, B. Basnar and I. Willner, Angew. Chem., Int. Ed., 2005, 44, 78; (g) I. Robel, B. A. Bunker and P. V. Kamat, Adv. Mater., 2005, 17, 2458.
2. (a) D. Liu and P. V. Kamat, J. Phys. Chem., 1993, 97, 10769; (b) S. Q. Fan, B. Fang, J. H. Kim, J. J. Kim, J. S. Yu and J. Ko, Appl. Phys. Lett., 2010, 96, 063501; (c) S. R. Sun, L. A. Gao, Y. Q. Liu and J. Sun, Appl. Phys. Lett., 2011, 98, 093112; (d) B. Farrow and P. V. Kamat, J. Am. Chem. Soc., 2009, 131, 11124; (e) A. Kongkanand, K. Tvrdy, K. Takechi, M. Kuno and P. V. Kamat, J. Am. Chem. Soc., 2008, 130, 4007; (f) N. Fuke, L. B. Hoch, A. Y. Koposov, V. W. Manner, D. J. Werder, A. Fukui, N. Koide, H. Katayama and M. Sykora, ACS Nano, 2010, 4, 6377; (g) H. Lee, M. Wang, P. Chen, D. R. Gamelin, S. M. Zakeeruddin, M. Grätzel and M. K. Nazeeruddin, Nano Lett., 2009, 9, 4221; (h) R. Loef, A. J. Houtepen, E. Talgorn, J. Schoonman and A. Goossens, Nano Lett., 2009, 9, 856; (i) P. Brown and P. V. Kamat, J. Am. Chem. Soc., 2008, 130, 8890; (j) K. S. Leschkies, R. Divakar, J. Basu, E. Enache-Pommer, J. E. Boercker, C. B. Carter, U. R. Kortshagen, D. J. Norris and E. S. Aydil, Nano Lett., 2007, 7, 1793; (k) I. Robel, V. Subramanian, M. Kuno and P. V. Kamat, J. Am. Chem. Soc., 2006, 128, 2385.
3. (a) R. Plass, S. Pelet, J. Krueger, M. Gratzel and U. Bach, J. Phys. Chem. B, 2002, 106, 7578; (b) J. B. Gao, J. M. Luther, O. E. Semonin, R. J. Ellingson, A. J. Nozik and M. C. Beard, Nano Lett., 2011, 11, 1002.
4. (a) K. S. Leschkies, T. J. Beatty, M. S. Kang, D. J. Norris and E. S. Aydil, ACS Nano, 2009, 3, 3638; (b) J. J. Choi, Y. F. Lim, M. B. Santiago-Berrios, M. Oh, B. R. Hyun, L. F. Sung, A. C. Bartnik, A. Goedhart, G. G. Malliaras, H. D. Abruna, F. W. Wise and T. Hanrath, Nano Lett., 2009, 9, 3749; (c) M. C. Beard, A. G. Midgett, M. Law, O. E. Semonin, R. J. Ellingson and A. J. Nozik, Nano Lett., 2009, 9, 836; (d) M. Law, M. C. Beard, S. Choi, J. M. Luther, M. C. Hanna and A. J. Nozik, Nano Lett., 2008, 8, 3904.
5. (a) P. R. Yu, K. Zhu, A. G. Norman, S. Ferrere, A. J. Frank and A. J. Nozik, J. Phys. Chem. B, 2006, 110, 25451; (b) W. Wei, X. Y. Bao, C. Soci, Y. Ding, Z. L. Wang and D. Wang, Nano Lett., 2009, 9, 2926.
6. A. Zaban, O. I. Micic, B. A. Gregg and A. J. Nozik, Langmuir, 1998, 14, 3153.
7. P. V. Kamat, K. Tvrdy, D. R. Baker and J. G. Radich, Chem. Rev., 2010, 110, 6664.
8. J. Tang, K. W. Kemp, S. Hoogland, K. S. Jeong, H. Liu, L. Levina, M. Furukawa, X. Wang, R. Debnath, D. Cha, K. W. Chou, A. Fischer, A. Amassian, J. B. Asbury and E. H. Sargent, Nat. Mater., 2011, 10, 765.
9. C. Liu, P. C. Wu, T. Sun, L. Dai, Y. Ye, R. M. Ma and G. G. Qin, J. Phys. Chem. C, 2009, 113, 14478.
10. P. K. Mahapatra and A. R. Dubey, Sol. Energy Mater. Sol. Cells, 1994, 32, 29.
11. G. Hodes, J. Phys. Chem. C, 2008, 112, 17778.
12. (a) R. Mohtadi, W. K. Lee, S. Cowan, J. W. Van Zee and M. Murthy, Electrochem. Solid-State Lett., 2003, 6, A272; (b) Z. Paal, K. Matusek and M. Muhler, Appl. Catal., A, 1997, 149, 113.
13. (a) G. Hodes, J. Manassen and D. Cahen, J. Appl. Electrochem., 1977, 7, 181; (b) G. Hodes, J. Manassen and D. Cahen, J. Electrochem. Soc., 1980, 127, 544; (c) Z. R. Yu, J. H. Du, S. H. Guo, H. Y. Zhang and Y. Matsumoto, Thin Solid Films, 2002, 415, 173; (d) Z. S. Yang, C. Y. Chen, C. W. Liu and H. T. Chang, Chem. Commun., 2010, 46, 5485; (e) Q. X. Zhang, Y. D. Zhang, S. Q. Huang, X. M. Huang, Y. H. Luo, Q. B. Meng and D. M. Li, Electrochem. Commun., 2010, 12, 327; (f) Z. Tachan, M. Shalom, I. Hod, S. Ruhle, S. Tirosh and A. Zaban, J. Phys. Chem. C, 2011, 115, 6162.
14. M. Shalom, S. Dor, S. Rühle, L. Grinis and A. Zaban, J. Phys. Chem. C, 2009, 113, 3895.
15. K. Park, Q. Zhang, B. B. Garcia, X. Zhou, Y.-H. Jeong and G. Cao, Adv. Mater., 2010, 22, 2329.
16. T. W. Hamann, O. K. Farha and J. T. Hupp, J. Phys. Chem. C, 2008, 112, 19756.
17. (a) A. Cao, Z. Liu, S. Chu, M. Wu, Z. Ye, Z. Cai, Y. Chang, S. Wang, Q. Gong and Y. Liu, Adv. Mater., 2010, 22, 103; (b) R. Elbaum, S. Vega and G. Hodes, Chem. Mater., 2001, 13, 2272.
18. F. Fabregat-Santiago, G. Garcia-Belmonte, I. Mora-Sero and J. Bisquert, Phys. Chem. Chem. Phys., 2011, 13, 9083.
19. (a) Q. Wang, J. E. Moser and M. Gratzel, J. Phys. Chem. B, 2005, 109, 14945; (b) F. Fabregat-Santiago, J. Bisquert, G. Garcia-Belmonte, G. Boschloo and A. Hagfeldt, Sol. Energy Mater. Sol. Cells, 2005, 87, 117.
20. R. Kern, R. Sastrawan, J. Ferber, R. Stangl and J. Luther, Electrochim. Acta, 2002, 47, 4213.
21. J. Bisquert, J. Phys. Chem. B, 2001, 106, 325.
22. M. Adachi, M. Sakamoto, J. Jiu, Y. Ogata and S. Isoda, J. Phys. Chem. B, 2006, 110, 13872.
23. (a) F. Fabregat-Santiago, J. Garcia-Canadas, E. Palomares, J. N. Clifford, S. A. Haque, J. R. Durrant, G. Garcia-Belmonte and J. Bisquert, J. Appl. Phys., 2004, 96, 6903; (b) J. Nelson, Phys. Rev. B: Condens. Matter, 1999, 59, 15374; (c) J.-Y. Hwang, S.-A. Lee, Y. H. Lee and S.-I. Seok, ACS Appl. Mater. Interfaces, 2010, 2, 1343.

---

Footnote

† Electronic Supplementary Information (ESI) available. See DOI: 10.1039/c2ra20979a

---