Preparation and Characterization of Nickel Oxide Photocathodes

Oct 2, 2013 - Ankita Kolay , Ramesh K. Kokal , Ankarao Kalluri , Isaac Macwan , Prabir K. ... J. Roland , Arvydas Ruseckas , Ifor D.W. Samuel , Fabric...
0 downloads 0 Views 394KB Size
Article pubs.acs.org/JPCC

Preparation and Characterization of Nickel Oxide Photocathodes Sensitized with Colloidal Cadmium Selenide Quantum Dots Irene Barceló, Elena Guillén,† Teresa Lana-Villarreal, and Roberto Gómez* Institut Universitari d’Electroquímica i Departament de Química Física, Universitat d’Alacant, Apartat 99, E-03080 Alacant, Spain S Supporting Information *

ABSTRACT: Quantum dot sensitized solar cells (QDSCs) are receiving a lot of attention as promising third generation solar cells, being virtually all of them based on sensitized photoanodes. Finding efficient QD-sensitized photocathodes would pave the way toward the implementation of tandem QDSCs. In this context, NiO photocathodes have been sensitized with colloidal CdSe quantum dots directly attached to the semiconductor oxide surface. The emission spectra indicate effective hole injection from the excited state of the quantum dots to the valence band of the NiO. A maximum incident current to photon conversion efficiency of 17% at 420 nm has been achieved. For the sake of comparison, other ways to prepare and anchor the QDs have been tested. Sensitization routes based on presynthesized colloidal quantum dots show better results than in situ growth techniques such as successive ionic layer adsorption and reaction. Electrochemical impedance measurements have identified transport resistance in NiO as one of the limiting factors in the performance of the system under study. Interestingly, surface treatments based on the deposition of very thin films of either SiO2 or Al2O3 can diminish recombination at the NiO/CdSe/electrolyte interface. This work also identifies a number of possible routes for the improvement of this kind of electrodes, unveiling their potential use in tandem quantum dot solar cells.



INTRODUCTION Since the discovery of dye sensitized solar cells (DSCs) in 1991,1 many studies have appeared dealing with the efficiency enhancement of these photovoltaic devices. The efforts have been focused among other things on the improvement of the photoactive anode, which is based on an n-type semiconductor oxide. At the end of the 90s, the possibility of using p-type semiconductor photocathodes in DSCs was first reported.2 Ever since, a number of relevant works have been published proving the feasibility of this new configuration.3−5 p-DSCs operate in an inverse mode to their n-type counterparts: upon light absorption and excitation of the dye, charge separation occurs by electron transfer from the valence band (VB) of the semiconductor to the excited dye (henceforth, dye hole injection into the semiconductor). Subsequently, the dye is regenerated to its ground state by electron transfer to the oxidized species in the electrolyte. Although several n-type oxides such as TiO2,6 ZnO,7 and SnO28 have been employed as electron conductors in DSCs, only NiO has been used as a hole conductor in p-DSCs. This is related to the paucity of p-type metal oxides and to the use of NiO in a variety of applications, particularly as an electrochromic material for smart windows.9 Thus, this p-type oxide has been extensively studied and a great amount of information about the preparation of thin film (nanostructured) electrodes can be found in the literature.10−16 In the context of solar cells, the main interest of the p-type materials that can be used as photocathodes is that they can be © 2013 American Chemical Society

combined with photoactive anodes in a tandem configuration. In theory, the maximum photoconversion efficiency attainable increases from 30% for a photovoltaic device using a single junction to 42% when two semiconductors are employed.17 Furthermore, this configuration allows harvesting light from a broader window of the solar spectrum since two sensitizers, one at each electrode, can be used. In principle, these advantages can be obtained by simply replacing the platinized counter electrode of the classical DSC configuration by a dye-sensitized p-type semiconductor electrode. This is very attractive from a commercial point of view since minor extra manufacturing and material costs would be involved. However, the overall efficiencies of tandem devices are still substantially lower than those of typical DSCs. In a series connection of a photocathode and a photoanode, the photovoltage will be given by the sum of the photopotentials of each electrode, whereas the photocurrent will be that of the least efficient electrode. Until now, the photocurrents achieved by p-type sensitized NiO electrodes are lower than those of sensitized photoanodes. This is one of the reasons for the poor performance of tandem solar cells. The low photocurrent values obtained for photocathodes stem from (1) poor sensitizer loading, (2) recombination of holes injected into the NiO VB with reduced dye molecules, and (3) the fact that p-type Received: July 15, 2013 Revised: October 2, 2013 Published: October 2, 2013 22509

dx.doi.org/10.1021/jp406989v | J. Phys. Chem. C 2013, 117, 22509−22517

The Journal of Physical Chemistry C

Article

approach and encourages future research on photoactive cathodes based on colloidal QDs as sensitizers.

semiconductor oxides have lower carrier mobility than n-type ones (the hole diffusion coefficient in NiO is 2 orders of magnitude lower than the electron diffusion coefficient in TiO2).18,19 Furthermore, in many cases, the light harvesters employed in this kind of photovoltaic device have been designed for n-type semiconductors. This fact has two negative consequences. First, the inappropriate location of the dye lowest unoccupied molecular orbital (LUMO), close to the NiO VB edge, which facilitates the aforementioned recombination route. Second, light absorbers in both photoactive electrodes have similar spectral responses and therefore compete for the same photons. The substitution of dyes by alternative light absorbers such as quantum dots (QDs) appears as a promising strategy to improve the response of p-type photocathodes. Among the advantages of using these sensitizers, there is the possibility of both tuning the visible response and properly locating the VB edge (enhancing consequently the hole injection efficiency) by controlling the QD size and taking advantage of hot carriers to get larger photocurrents (through multiple exciton generation).20,21 Moreover, they have molar extinction coefficients higher than those of dyes, which is important bearing in mind the reported difficulty to prepare thick NiO films in comparison with TiO2.5 However, it has been demonstrated that the performance of QD-sensitized TiO2 photoanodes depends on how light harvesters are anchored to the oxide surface. Sensitization methods employing presynthesized colloidal QDs (direct adsorption and adsorption via linker) present some advantages over popular in situ growth techniques, such as successive ionic layer adsorption and reaction (SILAR) and chemical bath deposition (CBD). Concretely, when colloidal solutions are employed there is a higher degree of control of the size and quantum confinement properties of the light harvesters. Furthermore, recombination processes are diminished since (i) colloidal QDs are generally passivated, and (ii) it is easier to obtain a single monolayer, preventing the formation of QD clusters. However, the electron injection rate from in situ grown QDs is higher as the distance between electron donor and acceptor is minimized. Along the same lines, directly adsorbed colloidal QDs show faster carrier injection than those adsorbed through a linker.22,23 However, the use of QDs instead of dyes imposes some changes in the NiO photocathode configuration. For example, a redox couple different from the mostly used iodide/triiodide is necessary as QDs are unstable in the presence of this redox couple.24,25 In addition, this redox pair has a potential close to that of the NiO VB edge, which limits to a great extent the theoretical open circuit voltage attainable by the cell. In this respect, cobalt-based redox couples have been proposed as a suitable alternative for p-type solar cells based on NiO.26 In this context, Gibson et al.27 found that the open circuit voltage (Voc) for dye-sensitized NiO solar cells with a cobalt electrolyte was almost three times higher than that for the iodine-based one. As a drawback, mass transport limitations are usually associated to cobalt-based redox couples.28,29 This work reports on the fabrication of photocathodes based on NiO nanostructured films sensitized with presynthesized CdSe QDs. For the first time, a comprehensive study on the feasibility of this kind of photocathodes as working electrodes in p-QDSCs has been carried out. Incident photon to current efficiencies as high as 17% (at 420 nm) have been achieved. To the best of our knowledge, this is the highest IPCE reported for this kind of electrode. This work proves the potentialities of this



EXPERIMENTAL SECTION Nanoporous NiO films were fabricated by doctor blading an aqueous suspension of NiO commercial nanoparticles (15−20 nm mean size) on FTO substrates and sintering at 400 °C. The metal oxide films were sensitized using presynthesized CdSe QDs capped with trioctylphosphine (TOP) or growing the QDs in situ by SILAR. The TOP-capped QDs were prepared using a solvothermal route.30 These light harvesters were adsorbed on the NiO surface directly or by means of a bifunctional linker. For direct adsorption, the oxide films were immersed in a CH2Cl2 dispersion of the QDs. In the case of linker-assisted adsorption, the molecular wires, L-cysteine (Cys), 3-mercaptopropionic acid (MPA), or 11-mercaptoundecanoic acid (MUA), were initially adsorbed on the NiO surface by immersion in an organic solution of the corresponding linker for one day. Afterward, the electrodes were thoroughly washed and soaked in a colloidal dispersion of CdSe QDs in toluene. The SILAR sensitization of the NiO electrodes was carried out by following the route described by Lee et al.31 In some cases, blocking layers of Al2O3 or SiO2 were grown on the NiO films prior to their modification with CdSe QDs.32,33 Absorption and diffuse reflectance spectra were obtained with a Shimadzu UV-2401 PC spectrophotometer equipped with an integrating sphere. Emission spectra were measured by means of a Fluoro-Max-4 spectrofluorometer. The amount of deposited CdSe QDs was determined by measuring the Cd content by inductively coupled plasma atomic emission spectrometry (ICP-AES) in a Perkin-Elmer Optima 7300 DV spectrophotometer. Transmission electron microscopy (TEM) analysis was carried out using a JEOL JEM-2010 instrument. Photoelectrochemical measurements were performed at room temperature in a three-electrode cell equipped with a fused silica window using a computer-controlled Autolab PGSTAT30 potentiostat. An aqueous Ag/AgCl/KCl(sat) electrode and a Pt wire were employed as a reference and a counter electrode, respectively. Two different N2-purged working electrolytes were used: an aqueous solution of 1 mM methyl viologen (MV2+) and 0.5 M NaOH, or a propylene carbonate solution of 0.005 M tris(4,4′-ditertbutyl-2,2′bipyridine)cobalt(III) perchlorate (Co(dtb-bpd)3(ClO4)3) complex27 and 0.1 M LiClO4. A 300 W Xe arc lamp with both a water filter and a UV filter (cutoff < 370 nm) was employed for electrode illumination from the substrate side. IPCE measurements were obtained by placing a monochromator (Oriel model 74100) between the light source and the cell. The light intensity was measured with an optical power meter (Oriel model 70310) equipped with a photodetector (Thermo Oriel 71608). Electrochemical impedance spectroscopy measurements were carried out in two different configurations: a threeelectrode and a sandwich-type cell. For the latter, the working electrode was clamped together with a platinum counter electrode using two binder clips. The electrolyte, 0.005 M Co(dtb-bpd)3(ClO4)2 and 0.005 M Co(dtb-bpd)3(ClO4)3 + 0.1 M LiClO4 in propylene carbonate, was drop casted on the working electrode before assembling the cell. The cells were illuminated from the NiO side at 1 sun (AM 1.5G) using a solar simulator (Sun 2000−11018 from Abet Technologies, Inc.) at open circuit and also in the dark applying the bias observed under illumination at open circuit. Measurements were carried 22510

dx.doi.org/10.1021/jp406989v | J. Phys. Chem. C 2013, 117, 22509−22517

The Journal of Physical Chemistry C

Article

luminescence also increases (inset in Figure 1B). This is probably related to a hindered hole injection from the QDs not in direct contact with the oxide.23 It should be mentioned that the degree of QD coverage of the NiO film sensitized via the MUA linker is within the range of those obtained by direct adsorption using sensitization times between 1 and 6 h. The QD coverage values have been determined by chemical analysis (see Supporting Information). In all cases, it is extremely low, ranging between 2 and 4% (see Table S1, Supporting Information). The difficulty to attain high sensitizer loadings when QD colloids are employed has also been observed with ntype oxides.22 In this respect, the remarkable QD coverage (34%) obtained recently by Zhang and co-workers36 for TiO2 electrodes modified with MPA-capped CdSe nanoparticles is noteworthy. It has already been proven for n-type semiconductors such as TiO237 and ZnO38 that the performance of QD-sensitized photoanodes strongly depends on the mode of attachment of the QDs to the oxide. Accordingly, different methods to sensitize NiO have been compared: direct adsorption, MPAand Cys-assisted adsorption, SILAR, and cosensitization (combining direct adsorption and SILAR). TEM images for the different electrodes can be found in Figure 2. Concretely, Figure 2b corresponds to directly adsorbed CdSe QDs, while Figure 2c,d shows TEM images for QDs attached to NiO through MPA and Cys linkers, respectively. Figure 2e,f corresponds to NiO samples modified by SILAR and cosensitization, respectively, and finally, Figure 2a shows a TEM image for an unsensitized NiO sample. In most cases, just a small fraction of the NiO surface is covered by CdSe QDs. Relatively high coverages could be attained only for electrodes prepared through SILAR and cosensitization. In fact, after 5 SILAR cycles, a homogeneous CdSe layer seems to be formed on some NiO nanoparticles, while in the case of the cosensitized electrode a number of large aggregates can be observed. Obviously, the QD size distribution is also wider for both SILAR and cosensitization methods. In Figure 3, the photocurrent transients obtained under the same conditions for electrodes sensitized by the different methods in contact with a Co3+ complex-based electrolyte are compared. It should be taken into account that initially the reduced form of the redox couple is absent from the working solution (only a negligible amount of it being produced during illumination on account of the small magnitude of the observed photocurrents), which renders negligible the recombination with the electrolyte. The photocurrent varies from sample to sample with the amount of adsorbed QDs, which in turn depends on the anchoring mode. In Table 1 the values of maximum photocurrent (jph) and QD coverage degree (θQD) attained in the different cases, together with those for the corrected photocurrent ((jph − jph0)/θQD), being jph0 the blank photocurrent delivered by an unsensitized NiO electrode, are presented. The ratio (jph − jph0)/θQD allows us to compare the relative performance of the different electrodes regardless of the sensitizer loading. From the values shown in the table, it is deduced that higher sensitizer loadings do not necessarily lead to higher photocurrents. This is particularly remarkable bearing in mind the very low coverage degree achieved in all cases. Paying attention to the (jph − jph0)/θQD ratio, it is observed that the electrode sensitized by direct adsorption is that giving the best result, followed far behind by that modified using Cys as a linker. In the case of electrodes sensitized by SILAR, the corrected photocurrent is reduced to half the value resulting

out in the frequency range of 1 mHz to 100 kHz with an ac amplitude of 10 mV. For the three-electrode cell, the experiments were performed in the dark under the same experimental conditions, but at 0 V (vs Ag/AgCl/KCl(sat)). Additional experimental details can be found in the Supporting Information.



RESULTS AND DISCUSSION Sensitization of NiO Films. Figure 1A shows the emission and absorption spectra of the colloidal dispersion of CdSe QDs

Figure 1. (A) UV−vis absorption and emission spectra for a colloidal dispersion of CdSe QDs in toluene. (B) Emission spectra for NiO films (∼7 μm-thick) supported on conducting glass and sensitized with CdSe QDs by either direct adsorption for 1 to 12 h or via MUA linker-mediated adsorption for 1 h. The excitation wavelength was 450 nm. The inset shows a detail of the spectra corresponding to electrodes sensitized by direct adsorption.

employed to sensitize the NiO films. A well-defined absorption peak at 588 nm is observed, corresponding to a QD diameter of around 4.1 nm.34 The corresponding photoluminescence spectrum, as expected, shows a peak at higher wavelengths (around 606 nm). In Figure 1B, the emission spectra of NiO nanostructured films sensitized with CdSe QDs by either direct adsorption or linker-mediated adsorption using MUA molecules are presented, along with that of a bare NiO film. When the sensitizer is deposited by means of the MUA linker, an intense emission signal is observed. This suggests that the long length of this bifunctional molecule favors radiative recombination of the photogenerated electron−hole pairs over hole injection into the oxide. In contrast, there is a severe photoluminescence quenching for NiO films sensitized by QDs directly deposited on the oxide surface. The fluorescence intensity is reduced by a factor of ∼7, indicating that holes are effectively transferred to the NiO nanoparticles. In this regard, Wu and Yeow35 have shown similar results using CdSe-ZnS core−shell QDs and NiO nanoparticles. These authors also concluded that hole injection takes place upon light excitation, which leads to a diminution of radiative recombination. Furthermore, as the amount of directly adsorbed QDs grows (by increasing the adsorption time), the sample photo22511

dx.doi.org/10.1021/jp406989v | J. Phys. Chem. C 2013, 117, 22509−22517

The Journal of Physical Chemistry C

Article

Figure 2. TEM images of the NiO nanoparticles prior (a) and after being sensitized with CdSe QDs by different methods: direct adsorption for 24 h (b), adsorption through MPA (c) and Cys (d) linkers for 24 h, SILAR (5 cycles) (e), and cosensitization (direct adsorption for 24 h and 2 SILAR cycles) (f). For guiding the eye, a few CdSe QDs are marked with a red dotted circular line.

from Cys-mediated adsorption. The superior performance of the modes of attachment using presynthesized colloidal QDs over SILAR was already observed in our previous works using n-type oxides.37 In situ growth techniques such as SILAR lead to the formation of QD multilayers and clusters with a large number of defects, which may retard hole injection and enhance recombination between holes in the oxide and electrons trapped at the QDs. The cosensitized electrode shows a corrected photocurrent similar to that of the SILAR method, which can be related to the fact that the previous adsorption of colloidal QDs on the oxide surface triggers a subsequent massive deposition of the sensitizer when the SILAR method is applied (Figure 2f). Comparing linker-mediated and direct adsorption, the higher corrected photocurrents obtained in the latter case can be explained in terms of a faster hole injection due to the smaller distance between hole donor and acceptor.23,37 Interestingly, the response of the electrode sensitized through the MPA linker is significantly lower than the rest. The cyclic voltammogram in the dark obtained for this photocathode proves that there is an obvious clogging of the oxide pores after QD deposition (see Figure S1, Supporting Information), which would explain the poor results achieved in this case. Finally, it is important to point out that, for the attainment of a significant photocurrent, it is necessary to work with relatively thick NiO films (see Supporting Information). This is related to the fact that a substantial QD loading was only achieved when thick NiO films were used, which is in agreement with the very low QD coverages observed in all cases. Incident Photon to Current Efficiency (IPCE). The IPCE spectrum for a NiO electrode modified with CdSe QDs by direct adsorption can be found in Figure 4. Significant IPCE values are measured in a broad region of the spectrum, from 420 to 625 nm. As expected, the IPCE spectrum resembles the absorption spectrum of colloidal CdSe QDs in toluene (Figure 1A). However, there is a broadening of the IPCE maximum peak with respect to the excitonic peak of CdSe in solution as a consequence of the different environments of the QDs within

Figure 3. Photocurrent transients measured at −0.1 V vs Ag/AgCl under 260 mW·cm−2 of white light (Xe arc lamp with cutoff filter, λ > 370 nm) for NiO nanoporous films (∼8.5 μm-thick) before (dashed line) and after sensitization with CdSe QDs through different methods (solid lines). A N2-purged 0.005 M Co(dtb-bpd)3(ClO4)3 + 0.1 M LiClO4 propylene carbonate solution was used as an electrolyte.

Table 1. Performance of the Electrodes Employed in the Experiments of Figure 3 mode of attachment

jph (μA·cm−2)

θQD (%)

(jph − jph0)/θQD (μA·cm−2)

direct adsorption SILAR cosensitized Cys linker MPA linker

107 72 104 67 39

0.80 1.34 2.25 0.62 1.18

114 42 39 83 20

22512

dx.doi.org/10.1021/jp406989v | J. Phys. Chem. C 2013, 117, 22509−22517

The Journal of Physical Chemistry C

Article

light absorbed by the electrolyte in the three-electrode cell, the theoretical maximum photocurrent attainable by the characterized electrode can be estimated to be around 2.4 mA·cm−2. Unfortunately, this value is much larger than that obtained experimentally (Figure 3), which can be probably ascribed to mass transport limitations of the cobalt complex through the mesoporous metal oxide network. This is one of the limitations of working with a cobalt-based electrolyte.28,29 The use of a solvent such as propylene carbonate, much more viscous than water, could also play a role. Recombination in p-Type Quantum Dot Solar Cells. To get further insights into the ionic and electronic processes limiting the final response of the system under study, electrochemical impedance spectra have been measured. Two different configurations have been studied: a three-electrode cell and a sandwich cell setup (referred to as solar cell in the following). The results obtained with the three-electrode setup are presented and discussed in the Supporting Information. These results confirmed the aforementioned mass transport limitations of the cobalt-based electrolyte. Likewise, they also point out the high charge transport resistance showed by the NiO as a limiting factor in the attainable efficiency for photocathodes using this p-type semiconductor. In the case of the sandwich cell setup, three different types of photocathodes have been tested: (1) an electrode based on a NiO film, (2) a NiO film sensitized with CdSe QDs by direct adsorption, and finally (3) a photocathode where a SiO2 layer was deposited on the NiO surface before sensitization. The use of blocking layers to reduce losses by recombination has already been explored in pDSCs based on NiO. Specifically, Al2O3 coatings have been successfully employed in this kind of devices.4,41 We have also implemented this strategy for QD-sensitized NiO photocathodes (see Supporting Information). The charge transfer processes and recombination routes taking place in the three studied cases are depicted in Figure 5. Figure 5a corresponds to an unsensitized NiO electrode, while Figure 5b corresponds to a QD-sensitized NiO electrode. In the case of Figure 5c, the sensitized electrode includes a blocking layer. Upon quantum dot excitation, the reduced form of the redox couple in the

Figure 4. IPCE spectra for a NiO nanostructured electrode prior and after sensitization with CdSe QDs by direct adsorption for 24 h. Measurements were carried out in a three-electrode cell setup at an applied potential of −0.1 V vs Ag/AgCl and using a N2-purged solution of 0.005 M Co(dtb-bpd)3(ClO4)3 + 0.1 M LiClO4 in propylene carbonate as an electrolyte.

the NiO thin film.39 Additionally, subtle structural changes associated to a partial removal of the TOP molecules capping the QDs after their direct adsorption on the oxide surface cannot be discarded. No response was found for bare NiO at wavelengths higher than 420 nm. Therefore, it can be concluded that the IPCEs measured are unequivocally a consequence of NiO sensitization with CdSe QDs. The IPCE reaches a maximum of 17% at 420 nm. To the best of our knowledge, this is the highest IPCE reported so far for QDsensitized NiO electrodes. Kang et al.40 obtained an IPCE of 12% at 370 nm for NiO films sensitized with CdS QDs by SILAR using an aqueous electrolyte based on polysulfide. However, in that work, the IPCE yields were significantly lower (less than 0.1%) when a cobalt-based electrolyte was employed instead. In our case, the NiO electrodes were very unstable in the presence of aqueous electrolytes. Dalavi et al.10 also observed instability of NiO films in contact with aqueous electrolytes. Moreover, the efficiency of the sensitized electrode increased below 420 nm, as a consequence of NiO excitation (results not shown). By considering the IPCE curve between 420 and 700 nm, the spectral irradiance of the Xe lamp employed for the photocurrent transient measurements of Figure 3, and the

Figure 5. Schematic representation of the charge transfer processes and recombination routes for three different photocathodes: an electrode based on a NiO film (a), a NiO film sensitized with CdSe QDs (b), and a photocathode where a blocking layer (BL) has been deposited on the NiO surface before sensitization (c). The arrow width qualitatively indicates the relative rate of the corresponding process. 22513

dx.doi.org/10.1021/jp406989v | J. Phys. Chem. C 2013, 117, 22509−22517

The Journal of Physical Chemistry C

Article

electrodes (Figure 6c), the size of the midfrequency semicircle can be directly related to the resistance to hole transfer at the NiO/CdSe/electrolyte interface.45 As can be observed in Figure 6, this size strongly varies under illumination. This behavior agrees well with the promotion of recombination. It is usually observed in n-type cells, but it has been also reported for p-type cells.46 In the dark, only recombination with the electrolyte is possible (R2 in Figure 5). However, under illumination two new recombination routes are envisageable: recombination of electrons photogenerated in CdSe with holes in NiO and hole transfer from the QD to the electrolyte before hole injection into NiO (R1 and R3 in Figure 5, respectively). In addition, under prolonged illumination a small increase in the concentration of reduced species is expected in the pores of the semiconductor structure due to the quantum dot regeneration process (T2 in Figure 5), which would enhance the R2 and R3 processes.46,47 In Figure 7, the spectra obtained under illumination for the three types of cells are compared. An analogous comparison for

electrolyte can transfer an electron to the QD VB (R3 in Figure 5) before hole injection into NiO (T1 in Figure 5). Once the hole has been injected into the semiconductor oxide VB, prior to extraction to the external circuit, it can recombine with either reduced species in the electrolyte or photogenerated electrons in the conduction band (CB) of the QDs (R2 and R1 in Figure 5, respectively). The latter is considered to be the main recombination pathway in p-DSCs.5 Note that in these sketches the possibility of photoexciting the NiO substrate has been considered because the samples were illuminated with 1 sun (AM 1.5G), which includes wavelengths smaller than 420 nm. In Figure 6, the impedance spectra for the different solar cell configurations are shown. The measurements were performed

Figure 7. Impedance spectra under illumination for solar cells based on NiO, NiO/CdSe, and NiO/SiO2/CdSe photocathodes. Blow-ups of the high frequency region of each spectrum can be found as insets. Figure 6. Nyquist plots obtained from EIS measurements in a twoelectrode configuration under illumination (open symbols) and in the dark under the same bias (solid symbols) for cells based on NiO (a), CdSe-sensitized NiO (b), and NiO/SiO2/CdSe (c) photocathodes. A propylene carbonate solution of 0.005 M Co(dtb-bpd)3(ClO4)2 and 0.005 M Co(dtb-bpd)3(ClO4)3 complexes + 0.1 M LiClO4 was used as an electrolyte.

the spectra measured in the dark can be found in Figure S9, Supporting Information. The equivalent circuit employed to fit the spectra is also shown (inset in Figure 7). This circuit is similar to that used to fit the impedance spectra for the threeelectrode configuration (see Figure S8, Supporting Information), but including an additional RC element to consider the charge transfer resistance (Rpt) and the double layer capacitance (Cpt) associated to the platinum counter electrode. The impedance due to diffusion in the electrolyte has been excluded from the fits. For n-type solar cells, an analogous circuit is usually employed.44 Blow-ups of the different Nyquist diagrams in the high frequency range are also presented as insets in Figure 7 to show the accuracy of the fitting. In Table 2 the obtained main fitting parameters are gathered, together with

under 1 sun illumination at open circuit and in the dark at the same bias. Two different regions are observed in the Nyquist plots, a small semicircle at high frequencies (left side of the spectra), which accounts for the counter electrode contribution, and a depressed semicircle at intermediate frequencies (right side of the spectra), which can be assigned to hole transport in the mesoscopic NiO film and to the back-reaction at the NiO/ CdSe/electrolyte interface. For a nonsensitized NiO film, the contributions of the counter electrode and the mesoporous film to the impedance spectrum partially overlap, and therefore, they cannot be straightforwardly singled out (Figure 6a). The contribution to the spectra in the low frequency region accounting for ionic transport in the electrolyte was neglected in contrast to the case of the three-electrode configuration (see Supporting Information). The spectra resemble a Gerischer impedance. This type of response has already been observed for solar cells based on cobalt electrolytes.42,43 It appears when transport resistance in the oxide is considerably larger than charge-transfer resistance (Figure S7, Supporting Information).44 For the NiO/CdSe (Figure 6b) and NiO/SiO2/CdSe

Table 2. Open Circuit Voltage (Voc) Measured under 1 Sun Illumination (AM 1.5) for Solar Cells Based on NiO, NiO/ CdSe, and NiO/SiO2/CdSe Electrodes, Together with Charge Transfer Resistance (rct) and Transport Resistance (rt) in the Mesoporous Oxide Layer (see Figure S8 in Supporting Information) for the Different Cells As Obtained by Fitting the Impedance Spectra in Figure 7

Voc (mV) rct (Ω) rt (Ω) 22514

NiO

NiO/CdSe

NiO/SiO2/CdSe

0 50

14 182 471

45 498 358

dx.doi.org/10.1021/jp406989v | J. Phys. Chem. C 2013, 117, 22509−22517

The Journal of Physical Chemistry C

Article

QDs and the electrolyte through the oxide network and provide a direct pathway for a rapid collection of photogenerated holes by the conducting substrate, (iii) using electrolytes with less mass transport limitations and/or (iv) employing light harvesters more suitable for p-type semiconductors. Moreover, surface treatments based on thin layers focused on recombination blockage and/or improving hole diffusion through the oxide matrix are equally interesting. Finally, the employment of a different p-type semiconductor with a higher hole diffusion coefficient could boost the efficiency of these systems. In a more general vein, we believe that this work reveals the potentialities of QD-sensitized photocathodes and opens up new routes for further research.

the open circuit voltage measured under 1 sun illumination. Because of the interference between the impedances of counter and NiO electrodes mentioned above, it was not possible to obtain a reliable value for transport resistance in the case of nonsensitized NiO films. When NiO is sensitized with QDs, recombination resistance increases. In spite of the low degree of QD coverage achieved, it seems that QDs passivate many of the traps/defects of the NiO nanoparticle surface that act as recombination centers (i.e., sites of preferential hole transfer to the electrolyte). At the same time, the oxide surface area exposed to the electrolyte is reduced.40 Both effects can also explain the significant increase in the resistance to recombination when a SiO2 layer is deposited between the NiO particles and the sensitizer. On the contrary, transport resistance decreases upon coating with silica. For cells based on sensitized NiO films without the SiO2 overlayer, resistance to transport is much higher than recombination resistance. This has been also observed for the three-electrode cell setup and can be attributed to the low hole diffusion coefficient in the NiO.18,40 The opposite result is obtained upon silica coating, where resistance to recombination is higher than resistance to transport. In agreement with these results, Zhu et al.48 found that charge transport in NiO is surface-mediated. If it is assumed that transport is slowed down by surface states, the lower resistance to diffusion can be due to the reduction of the surface trap density by the conformal coverage of the NiO surface with SiO2.40 The higher Voc obtained for the NiO sensitized photocathode upon SiO2 coating could be related to the favorable ratio between recombination resistance and transport resistance. In this regard, it should be mentioned that covering the NiO surface with Al2O3 (instead of SiO2) also led to an increase in Voc in aqueous media (Figure S5, Supporting Information), pointing to the fact that the observed improvement is not exclusive of the SiO2 coating.



ASSOCIATED CONTENT

S Supporting Information *

Experimental details, determination of quantum dot coverage, cyclic voltammograms for NiO electrodes sensitized with CdSe QDs via mercaptopropionic acid linker, effect of NiO film thickness, effect of Al2O3 and SiO2 blocking layers, and additional information on EIS measurements. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*(R.G.) E-mail: [email protected]. Phone: (+34) 96 590 3748. Present Address

Abengoa Research, C/Energiá Solar, Palmas Altas, E-41014 Sevilla, Spain. †

Notes

The authors declare no competing financial interest.



CONCLUSIONS In summary, an efficient sensitization of relatively thick nanoparticulate NiO films by CdSe QDs has been proven by means of photoluminescence and photoelectrochemical measurements. The performance of the sensitized photocathodes depends on the mode of attachment of the QDs to the oxide surface, being direct adsorption the method giving the best results. In all cases, very low QD coverages (0.6−4.1%) were obtained. Interestingly, an incident photon to current efficiency as high as 17% at 420 nm was obtained, which is, to the best of our knowledge, the highest IPCE obtained for QD-sensitized ptype electrodes. Higher photovoltages were achieved when blocking layers of Al2O3 and SiO2 were deposited on the oxide surface before sensitization. EIS measurements were undertaken to further analyze the limiting processes in the systems under study, pointing out diffusion difficulties in the cobaltbased electrolyte employed together with a high resistance to hole transport in NiO. In addition, these measurements also show that the presence of a SiO2 blocking layer increases the charge transfer resistance and decreases the hole transport resistance in agreement with the observed increase in Voc. In spite of the encouraging results, the photocurrents and photopotentials achieved by these electrodes should be increased to enable their future integration in tandem cells. A better response could be obtained by (i) increasing the sensitizer loading, (ii) using well-defined open one-dimensional NiO nanostructures, which favor the penetration of both the



ACKNOWLEDGMENTS



REFERENCES

This work was financially supported by the Spanish Ministry of Science and Innovation through projects HOPE CSD200700007 (Consolider-Ingenio2010) and MAT2012-37676 (FONDOS FEDER).

(1) O’Regan, B.; Grätzel, M. A Low-Cost, High-Efficiency Solar Cell Based on Dye-Sensitized Colloidal TiO2 Films. Nature 1991, 353, 737−740. (2) He, J.; Lindström, H.; Hagfeldt, A.; Lindquist, S.-E. DyeSensitized Nanostructured p-Type Nickel Oxide Film as a Photocathode for a Solar Cell. J. Phys. Chem. B 1999, 103, 8940−8943. (3) Nattestad, A.; Mozer, A. J.; Fischer, M. K. R.; Cheng, Y. B.; Mishra, A.; Bäuerle, P.; Bach, U. Highly Efficient Photocathodes for Dye-Sensitized Tandem Solar Cells. Nat. Mater. 2010, 9, 31−35. (4) Uehara, S.; Sumikura, S.; Suzuki, E.; Mori, S. Retardation of Electron Injection at NiO/Dye/Electrolyte Interface by Aluminium Alkoxide Treatment. Energy Environ. Sci. 2010, 3, 641−644. (5) Odobel, F.; Le Pleux, L.; Pellegrin, Y.; Blart, E. New Photovoltaic Devices Based on the Sensitization of p-Type Semiconductors: Challenges and Opportunities. Acc. Chem. Res. 2010, 43, 1063−1071. (6) 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. Porphyrin-Sensitized Solar Cells with Cobalt (II/III)Based Redox Electrolyte Exceed 12% Efficiency. Science 2011, 334, 629−634. (7) Anta, J. A.; Guillén, E.; Tena-Zaera, R. ZnO-Based Dye-Sensitized Solar Cells. J. Phys. Chem. C 2012, 116, 11413−11425. 22515

dx.doi.org/10.1021/jp406989v | J. Phys. Chem. C 2013, 117, 22509−22517

The Journal of Physical Chemistry C

Article

Mesoporous TiO2 Photoanodes. J. Phys. Chem. C 2008, 112, 18255− 18263. (29) Klahr, B. M.; Hamann, T. W. Performance Enhancement and Limitations of Cobalt Bipyridyl Redox Shuttles in Dye-Sensitized Solar Cells. J. Phys. Chem. C 2009, 113, 14040−14045. (30) Wang, Q.; Pan, D.; Jiang, S.; Ji, X.; An, L.; Jiang, B. A Solvothermal Route to Size- and Shape-Controlled CdSe and CdTe Nanocrystals. J. Cryst. Growth 2006, 286, 83−90. (31) Lee, H.; Wang, M.; Chen, P.; Gamelin, D. R.; Zakeeruddin, S. M.; Grätzel, M.; Nazeeruddin, M. K. Efficient CdSe Quantum DotSensitized Solar Cells Prepared by an Improved Successive Ionic Layer Adsorption and Reaction Process. Nano Lett. 2009, 9, 4221−4227. (32) Palomares, E.; Clifford, J. N.; Haque, S. A.; Lutz, T.; Durrant, J. R. Control of Charge Recombination Dynamics in Dye Sensitized Solar Cells by the Use of Conformally Deposited Metal Oxide Blocking Layers. J. Am. Chem. Soc. 2002, 125, 475−482. (33) Liu, Z.; Miyauchi, M.; Uemura, Y.; Cui, Y.; Hara, K.; Zhao, Z.; Sunahara, K.; Furube, A. Enhancing the Performance of Quantum Dots Sensitized Solar Cell by SiO2 Surface Coating. Appl. Phys. Lett. 2010, 96, 233107. (34) Yu, W. W.; Qu, L.; Guo, W.; Peng, X. Experimental Determination of the Extinction Coefficient of CdTe, CdSe, and CdS Nanocrystals. Chem. Mater. 2003, 15, 2854−2860. (35) Wu, X.; Yeow, E. K. L. Charge-Transfer Processes in Single CdSe/ZnS Quantum Dots with p-Type NiO Nanoparticles. Chem. Commun. 2010, 46, 4390−4392. (36) Zhang, H.; Cheng, K.; Hou, Y. M.; Fang, Z.; Pan, Z. X.; Wu, W. J.; Hua, J. L.; Zhong, X. H. Efficient CdSe Quantum Dot-Sensitized Solar Cells Prepared by a Postsynthesis Assembly Approach. Chem. Commun. 2012, 48, 11235−11237. (37) Barceló, I.; Campiña, J. M.; Lana-Villarreal, T.; Gómez, R. A Solid-State CdSe Quantum Dot Sensitized Solar Cell Based on a Quaterthiophene as a Hole Transporting Material. Phys. Chem. Chem. Phys. 2012, 14, 5801−5807. (38) Barceló, I.; Lana-Villarreal, T.; Gómez, R. Efficient Sensitization of ZnO Nanoporous Films with CdSe QDs Grown by Successive Ionic Layer Adsorption and Reaction (SILAR). J. Photochem. Photobiol., A 2011, 220, 47−53. (39) Leatherdale, C. A.; Bawendi, M. G. Observation of Solvatochromism in CdSe Colloidal Quantum Dots. Phys. Rev. B 2001, 63, 165315. (40) Kang, S. H.; Zhu, K.; Neale, N. R.; Frank, A. J. Hole Transport in Sensitized CdS-NiO Nanoparticle Photocathodes. Chem. Commun. 2011, 47, 10419−10421. (41) Natu, G.; Huang, Z.; Ji, Z.; Wu, Y. The Effect of an Atomically Deposited Layer of Alumina on NiO in p-Type Dye-Sensitized Solar Cells. Langmuir 2011, 28, 950−956. (42) Liu, Y.; Jennings, J. R.; Huang, Y.; Wang, Q.; Zakeeruddin, S. M.; Grätzel, M. Cobalt Redox Mediators for Ruthenium-Based DyeSensitized Solar Cells: A Combined Impedance Spectroscopy and Near-IR Transmittance Study. J. Phys. Chem. C 2011, 115, 18847− 18855. (43) Liberatore, M.; Burtone, L.; Brown, T. M.; Reale, A.; Carlo, A. D.; Decker, F.; Caramori, S.; Bignozzi, C. A. On the Effect of Al2O3 Blocking Layer on the Performance of Dye Solar Cells with Cobalt Based Electrolytes. Appl. Phys. Lett. 2009, 94, 173113. (44) Fabregat-Santiago, F.; Bisquert, J.; Garcia-Belmonte, G.; Boschloo, G.; Hagfeldt, A. Influence of Electrolyte in Transport and Recombination in Dye-Sensitized Solar Cells Studied by Impedance Spectroscopy. Sol. Energy Mater. Sol. Cells 2005, 87, 117−131. (45) Yang, H.; Guai, G. H.; Guo, C.; Song, Q.; Jiang, S. P.; Wang, Y.; Zhang, W.; Li, C. M. NiO/Graphene Composite for Enhanced Charge Separation and Collection in p-Type Dye Sensitized Solar Cell. J. Phys. Chem. C 2011, 115, 12209−12215. (46) Huang, Z.; Natu, G.; Ji, Z.; Hasin, P.; Wu, Y. p-Type DyeSensitized NiO Solar Cells: A Study by Electrochemical Impedance Spectroscopy. J. Phys. Chem. C 2011, 115, 25109−25114. (47) Chan, X.-H.; Jennings, J. R.; Hossain, M. A.; Yu, K. K. Z.; Wang, Q. Characteristics of p-NiO Thin Films Prepared by Spray Pyrolysis

(8) Chappel, S.; Zaban, A. Nanoporous SnO2 Electrodes for DyeSensitized Solar Cells: Improved Cell Performance by the Synthesis of 18 nm SnO2 Colloids. Sol. Energy Mater. Sol. Cells 2002, 71, 141−152. (9) Lampert, C. M. Electrochromic Materials and Devices for Energy Efficient Windows. Sol. Energy Mater. 1984, 11, 1−27. (10) Dalavi, D. S.; Suryavanshi, M. J.; Patil, D. S.; Mali, S. S.; Moholkar, A. V.; Kalagi, S. S.; Vanalkar, S. A.; Kang, S. R.; Kim, J. H.; Patil, P. S. Nanoporous Nickel Oxide Thin Films and Its Improved Electrochromic Performance: Effect of Thickness. Appl. Surf. Sci. 2011, 257, 2647−2656. (11) Boschloo, G.; Hagfeldt, A. Spectroelectrochemistry of Nanostructured NiO. J. Phys. Chem. B 2001, 105, 3039−3044. (12) Lepleux, L.; Chavillon, B.; Pellegrin, Y.; Blart, E.; Cario, L.; Jobic, S.; Odobel, F. Simple and Reproducible Procedure to Prepare Self-Nanostructured NiO Films for the Fabrication of p-Type DyeSensitized Solar Cells. Inorg. Chem. 2009, 48, 8245−8250. (13) Velevska, J.; Ristova, M. Electrochromic Properties of NiOx Prepared by Low Vacuum Evaporation. Sol. Energy Mater. Sol. Cells 2002, 73, 131−139. (14) Vera, F.; Schrebler, R.; Muñoz, E.; Suarez, C.; Cury, P.; Gómez, H.; Córdova, R.; Marotti, R. E.; Dalchiele, E. A. Preparation and Characterization of Eosin B- and Erythrosin J-Sensitized Nanostructured NiO Thin Film Photocathodes. Thin Solid Films 2005, 490, 182−188. (15) Urbano, A.; Ferreira, F. F.; deCastro, S. C.; Landers, R.; Fantini, M. C. A.; Gorenstein, A. Electrochromism in Lithiated Nickel Oxide Films Deposited by RF Sputtering. Electrochim. Acta 2001, 46, 2269− 2273. (16) Nattestad, A.; Ferguson, M.; Kerr, R.; Cheng, Y.-B.; Bach, U. Dye-Sensitized Nickel(II)Oxide Photocathodes for Tandem Solar Cell Applications. Nanotechnology 2008, 19, 295304. (17) Vos, A. D. Detailed Balance Limit of the Efficiency of Tandem Solar Cells. J. Phys. D: Appl. Phys. 1980, 13, 839. (18) Mori, S.; Fukuda, S.; Sumikura, S.; Takeda, Y.; Tamaki, Y.; Suzuki, E.; Abe, T. Charge-Transfer Processes in Dye-Sensitized NiO Solar Cells. J. Phys. Chem. C 2008, 112, 16134−16139. (19) Li, L.; Gibson, E. A.; Qin, P.; Boschloo, G.; Gorlov, M.; Hagfeldt, A.; Sun, L. Double-Layered NiO Photocathodes for p-Type DSSCs with Record IPCE. Adv. Mater. 2010, 22, 1759−1762. (20) Sambur, J. B.; Novet, T.; Parkinson, B. A. Multiple Exciton Collection in a Sensitized Photovoltaic System. Science 2010, 330, 63− 6. (21) Semonin, O. E.; Luther, J. M.; Choi, S.; Chen, H.-Y.; Gao, J.; Nozik, A. J.; Beard, M. C. Peak External Photocurrent Quantum Efficiency Exceeding 100% via MEG in a Quantum Dot Solar Cell. Science 2011, 334, 1530−1533. (22) Guijarro, N.; Lana-Villarreal, T.; Mora-Seró, I.; Bisquert, J.; Gómez, R. CdSe Quantum Dot-Sensitized TiO2 Electrodes: Effect of Quantum Dot Coverage and Mode of Attachment. J. Phys. Chem. C 2009, 113, 4208−4214. (23) Guijarro, N.; Shen, Q.; Giménez, S.; Mora-Seró, I.; Bisquert, J.; Lana-Villarreal, T.; Toyoda, T.; Gómez, R. Direct Correlation between Ultrafast Injection and Photoanode Performance in Quantum Dot Sensitized Solar Cells. J. Phys. Chem. C 2010, 114, 22352−22360. (24) Hodes, G. Comparison of Dye- and Semiconductor-Sensitized Porous Nanocrystalline Liquid Junction Solar Cells. J. Phys. Chem. C 2008, 112, 17778−17787. (25) Shalom, M.; Dor, S.; Rühle, S.; Grinis, L.; Zaban, A. Core/CdS Quantum Dot/Shell Mesoporous Solar Cells with Improved Stability and Efficiency Using an Amorphous TiO2 Coating. J. Phys. Chem. C 2009, 113, 3895−3898. (26) Hamann, T. W. The End of Iodide? Cobalt Complex Redox Shuttles in DSSCs. Dalton Trans. 2012, 41, 3111−3115. (27) Gibson, E. A.; Smeigh, A. L.; Le Pleux, L.; Hammarström, L.; Odobel, F.; Boschloo, G.; Hagfeldt, A. Cobalt Polypyridyl-Based Electrolytes for p-Type Dye-Sensitized Solar Cells. J. Phys. Chem. C 2011, 115, 9772−9779. (28) Nelson, J. J.; Amick, T. J.; Elliott, C. M. Mass Transport of Polypyridyl Cobalt Complexes in Dye-Sensitized Solar Cells with 22516

dx.doi.org/10.1021/jp406989v | J. Phys. Chem. C 2013, 117, 22509−22517

The Journal of Physical Chemistry C

Article

and Their Application in CdS-Sensitized Photocathodes. J. Electrochem. Soc. 2011, 158, H733−H740. (48) Zhu, H.; Hagfeldt, A.; Boschloo, G. Photoelectrochemistry of Mesoporous NiO Electrodes in Iodide/Triiodide Electrolytes. J. Phys. Chem. C 2007, 111, 17455.

22517

dx.doi.org/10.1021/jp406989v | J. Phys. Chem. C 2013, 117, 22509−22517