Near Field Enhanced Photocurrent Generation in P-type Dye-Sensitized Solar Cells

Over the past few decades, the field of p-type dye-sensitized solar cell (p-DSSC) devices has undergone tremendous advances, in which Cu-based delafossite nanocrystal is of prime interest. This paper presents an augment of about 87% improvement in photocurrent observed in a particular configuration of organic dye P1 sensitized CuCrO2 delafossite nanocrystal electrode coupled with organic redox shuttle, 1-methy-1H- tetrazole-5-thiolate and its disulfide dimer when Au nanoparticles (NPs, with diameter of about 20 nm) is added into the photocathode, achieving a power convert efficiency of 0.31% (measured under standard AM 1.5 G test conditions). Detailed investigation shows that the local electrical-magnetic field effect, induced by Au NPs among the mesoporous CuCrO2 film, can improve the charge injection efficiency at dye/semiconductor interface, which is responsible for the bulk of the gain in photocurrent.

D ye-sensitized solar cells (DSSC), one of the third generation solar cells, has attracted much attention because of high efficiency, low cost and easy fabrication 1 . It's the first to use a three-dimensional nanocrystalline junction for solar electricity production featured with separating light absorption from charge carrier transport. The adsorption of light in the DSSC occurs by dye molecules and the charge separation by electron injection from the dye to the nanocrystal (TiO 2 for instance) at the semiconductor electrolyte interface. Optimizing kinetic efficiency in each step of this energy hierarchy is a principal goal of electrochemical energy research. Liquid-based DSSCs using porphyrins sensitizers currently attain 12.3% solar energy conversion efficiency and solid-state devices using perovskite light harvester with 15%, rendering it a credible alternative to other thin film photovoltaic cells 2,3 . The concept of the multi-junction or tandem solar cell was proposed with the goal to provide the possibility to absorb more photons, thereby leading to increase power conversion efficiency (PCE) 4,5 . Experimentally, this concept has been shown to be viable using two different sensitizers 5 , which can be matched to absorb the high energy photons on one electrode and low energy photons on the other, thereby utilizing the full solar spectrum 6 . Theoretically, the overall PCE of tandem DSSCs can reach above 40% 7 , superior to their single components. But presently, the efficiency of tandem DSSC devices is limited by the lower efficiency of p-type devices (the maximum reported value being about 1.3%). For this reason, there is a strong need for an improvement of the photovoltaic performance of p-type DSSCs. The operating principle of a p-type DSSC is shown by the schematic diagram in Figure 1. There are two key issues to address in order to increase the relatively low efficiencies found in p-type DSSC. One basic problem is a poor light harvesting found in p-type devices, resulting from the use of a thin nanocrystalline film (i.e., ,2 mm). The other impediment to producing efficient p-type DSSC devices is the low carrier mobility in the p-type semiconductor and mismatch energy level with sensitizers found at the dye-sensitized heterojunction (see Figure 1). One of the most recent achievements of p-type DSSCs research was the utilization of efficient sensitizing dyes 5 . Sun et al reported a series of organic donor-p-acceptor structured dyes for p-type DSSC having the triphyenylamine moieties as the electron donor, malononitrile moieties as the electron acceptor, and a thiophenes or other units as the conjugated chain, showing promising PCEs in the range of 0.1 , 0.2% 8,9 . Further improvements have been observed including an implantation of p-type semiconductor with high carrier mobility and suitable band energy level [10][11][12] , or new redox shuttles with high transparency and the steric bulk 13,14 . ABO 2 oxides (A 5 Cu, Ag, B 5 transition metal element) bearing the delafossite structures have attracted a number of attractions as a triangular lattice antiferromagnet prototype with interesting properties such as transparent conductivity 15 and multiferroicity 16,17 . Among recently studied compounds, the CuCrO 2 delafossite is of prime interest due to its higher transparence, higher carrier mobility and more positive flat band in comparison with NiO 11,14 .
Recently, noble metal nanoparticles (NPs) such as Au and Ag with surface plasmon resonance (SPR) property constitute a large ongoing research field 18 , that have brought these nanostructures to the forefront of nanotechnology research directed towards applications in optoelectronics, including DSSCs 19 . The oscillation of conduction electrons in resonance with incident light promotes the electromagnetic field localized in the vicinity of plasmon resonance site, which has been verified to enhance optical absorption. Therefore the augmented photo-to-electron conversion efficiency could be attributed to the consequence of enhancement of optical absorption when Au NPs were added into DSSCs 20,21 . However, Zhang et al argued that the enhanced photocurrent was obtained from unexpected mechanisms of reducing exciton binding energy with help of Au@SiO 2 NPs, rather than the enhancement of optical absorption 22 .
In this communication, we revisited CuCrO 2 11,15 and an organic sensitizer P1 (see Figure 1) 8,9 in combination with a disulfide/thiolate redox shuttle (see Figure 1), presenting an impressive solar-to-electricity conversion efficiency of 0.31% under AM 1.5 G (air mass 1.5 global) (100 mW cm 22 ) when Au NPs were added into the photocathode. The magnetic field effect (MFE) offers a possible solution to show that the near field induced by Au NPs has an effect on photocurrent. In addition, the analysis supported that the near field promotes more charge injection at dye/semiconductor interface. This result is in agreement with earlier reported phenomena that the external electrical and magnetic fields have an effect on electron transfer reaction in various electrochemical systems, including electroluminescent diodes, fuel cells, organic solar cells, and n-type DSSCs [23][24][25][26] .

Results
CuCrO 2 film characterization. Figure 2a presents the TEM images of ultrasmall CuCrO 2 delafossite nanocrystals prepared by a hydrothermal method. The inset shows the high resolution image of CuCrO 2 nanocrystals with a typical size of about 15 3 5 nm 2 11,15 . The Au@SiO 2 NPs were prepared by a standard chemical reduction synthesis method 27 . A representative TEM image of Au@SiO 2 NPs is shown in Figure 2b. The Au NPs of about 20 nm diameter are coated by an approximate 3 nm SiO 2 shell, showing an absorption peak at ,550 nm in water due to collective oscillation of electrons in nanoparticles (see Figure S1). Figure 2c presents the XRD phase analysis of CuCrO 2 , which confirms the delafossite crystalline structure with Cu 1 in the lattice. The nanocrystal is predominantly c-axis oriented, with a very minor additional Bragg peak at the (101) orientation. Most traces of the Cu 2 O impurity phase were eliminated. Figure 2d presents the UV-visible spectrum of P1-sensitized CuCrO 2 or CuCrO 2 mixed with Au@SiO 2 NPs (coded CuCrO 2 @Au) films. A slightly increased absorbance spectrum was observed in the wavelength range from 550 nm to 750 nm. The absorption peak due to the P1 dye was blue-shifted (from 495 nm to 482 nm) and the half width of absorption peak was broadened when Au NPs were mixed with CuCrO 2 . It was observed that the transparency of a 1.8 mm thick CuCrO 2 or CuCrO 2 @Au film samples was 60% in the whole visible range. The inset of Figure 2d shows the estimated optical bandgap of CuCrO 2 film, being about 3.11 eV.
Photovoltaic performance. Figure 3a presents the photocurrent density-voltage (J-V) and dark current characteristics of DSSCs using CuCrO 2 (black curve, device A) or CuCrO 2 @Au (red curve, device B) as the photocathode (1.8 mm film thickness). The photovoltaic parameters, open circuit voltage (V oc ), fill factor (FF), short circuit current density (J sc ) and PCE of p-type DSSC devices under standard AM 1.5 G illumination at 100 mW cm 22 are tabulated in Table 1. Device A exhibited a V oc of 0.309 V, a J sc of 1.43 mA cm 22 , and a FF of 0.38, giving an overall PCE of 0.17%. It is worthy to note the photocurrent line and the dark current line are crossing each other. The current-voltage characteristics for an inplane file containing P1-sensitized CuCrO 2 or bare CuCrO 2 measured in dark and illumination are presented in Figure S2. A largely increased current for P1-sensitized CuCrO 2 under illumination clearly demonstrated the photo-induced charge transfer process. An increasing hole density gives rise to CuCrO 2 film conductivity   and charge carrier mobility, which can be contributed to the cross of photocurrent and dark current. The larger current-voltage characteristic slope value of CuCrO 2 film sensitized by P1 under illumination comparing to those in the dark implied that the enhanced conductivity only occurs when the photo-generated charge is sustained in CuCrO 2 film 28 .
Magnetic field effect. Magnetic field effects (MFE) measurements have been carried out with a home-made experimental setup by monitoring the device photocurrent (MFE PC ) 29 . The current response of an in-plane film containing CuCrO 2 sensitized with P1 in a magnetic field was first studied. The result in Figure 2e indicates that the magnetic field has negligible effect on the P1 sensitized CuCrO 2 semiconductor film conductivity at room temperature when there is no redox shuttles. Figure 4a presents the photocurrent response (top) of a complete p-DSSC device using CuCrO 2 /P1 photocathode (red curve, curve 1) and device with bare CuCrO 2 photocathode photocathode (black curve, curve 2) in combination with T 2 /T 2 electrolyte at a bias of 20.1 V in a magnetic field (bottom) under illumination as a function of time. For comparison purpose, the MFE PC measurement in the dark for the P1 based device was performed and the results were shown in Figure 4a (curve 3). Clearly, for a complete device under ,2% sun illumination condition the MFE PC simultaneously increased with magnetic field intensity. A negligible change in the current signal was observed in curve 2 (without sensitizer) or curve 3 (in dark). Because of the instrument limitation, the magnetic field intensity could only reach its maximum value at ,700 mT, giving a MFE PC improvement of ,3% (Figure 4b).
Since the photocurrent depends on the light harvesting capability, the absorbance of P1 sensitized CuCrO 2 film in an increasing magnetic field was recorded as presented in Figure S3. The result showed that MFE had no effect on the absorption spectrum. Earlier studies have demonstrated that the magnetic field can increase photoconduction in terms of a hyperfine scale magnetic-field-dependent mixing of singlet and triplet electron-hole pair states, in which the concentration of singlet excitons can be modulated. The contribution of excitonic injection of holes to the drive current has been reported in Alq3-based electroluminescent diodes 25 . Therefore, based on the observation in Figure 2e and Figure 4 as well as Figure S3, we suggest that the enhanced MFE PC can be due to an improvement in the charge injection efficiency at P1/CuCrO 2 interface.
Near field effect induced by Au NPs. As discussed above, the MFE PC was observed in the p-type DSSC based on P1 sensitized CuCrO 2 films in combination with T 2 /T 2 electrolyte. It is reasonable to argue that photocurrent can be greatly enhanced if there is a magnetic field generated by a localized SPR. In order to verify this idea, silicacovered Au NPs (with the diameter of 20 nm) were synthesized and added into CuCrO 2 film for SPR application in p-type DSSC. The Au NPs were mixed with CuCrO 2 (Au:CuCrO 2 0.1% w/w) during the screen-printing paste preparation and the resulted films were sintered at 550uC in Ar atmosphere to promote a well-junction between Au NPs and CuCrO 2 nanocrystals. Under the same condition for preparation of DSSC devices, changing the photocathode materials from CuCrO 2 to CuCrO 2 @Au (the film thickness, the dye-loading and the rodox species concentration remain unchanged) in device B improved photocurrent by 87% to 2.68 mA cm 22 , giving an overall PCE of 0.31% (Table 1). Figure 3b presents the incident photon to current conversion efficiency (IPCE) as a function of the light excitation wavelength. The IPCEs below 390 nm excitation are deteriorated due to a sheet of anti-reflecting UV cut-off film attached to DSSCs. The features of the spectral response of the photocurrent closely match the adsorption spectrum of the P1 dye onto the CuCrO 2 film (referring to Figure 2e).   The IPCE of device A reaches its highest values of about 15% at about 470 nm, then drops rapidly above 530 nm. For device B, the IPCE increases largely from 440 nm to 530 nm, reaching over 25%. From the overlapped integral of this curve and the standard global AM 1.5 G solar emission spectrum, a short-circuit photocurrent density of 2.70 mA cm 22 is obtained, which is in excellent agreement with the measured photocurrent (J sc 5 2.68 mA cm 22 ), providing that the spectral mismatch between solar simulator and the standard AM 1.5 G solar emission is negligibly small. The ratio of IPCE device B / IPCE device A was evaluated to be about 2 in the range from 400 nm to 580 nm (Figure 3b, right ordinate), evidently ascribing the augmented photocurrent. Considering the optical adsorption of the dyesensitized films and IPCE spectra of the corresponding devices, the increased photocurrent in device B should not be caused by those spectrogram response after 600 nm though the ratio of IPCE device B / IPCE device A increases largely in this wavelength range.
Light harvesting property. Figure 3c presents the light harvesting efficiency (LHE) for the dye-sensitized nanocrystalline films examined by their UV-visible absorption spectra. As shown in Figure 3c (right coordinate), LHE is slightly increased for wavelength range from 550 nm to 800 nm due to the addition of Au@ SiO 2 NPs into the CuCrO 2 film. As discussed above, the J sc of device B had 87% enhancement while the V oc values (Table 1) were nearly the same for both devices. We reasoned that the expected gain in J sc could be caused by an in-situ electromagnetic field produced by Au NPs, due to the resonant collective oscillation of free electrons in noble metal nanostructures driven by light.
Photovoltage transient decay measurement. The interfacial charge recombination kinetics and diffusion kinetics of injected holes were investigated by photovoltage transient decay experiments and charge extraction measurements. Figure 5a exhibits the V oc as a function of the extracted charge density, n t (where n t 5 DQ/(A r d/(1 2 p)), DQ equals to the extracted charge, A r being the real interface area of the photocathode, d being the film thickness of the photocathode, and p being the porosity of the cathode, respectively). The charge density was chosen as the abscissa in order to compare the V oc values obtained from the two photocathodes at the same hole concentration in the dye-sensitized mesoporous films. All date points from both devices fall on the same straight line, suggesting that the addition of Au NPs to the CuCrO 2 nanocrystals has negligible effect on the CuCrO 2 valence band edge. Figure 5b shows the apparent charge lifetime (t) as a function of injected hole density, allowing a comparison of the recombination rates for the different photocathodes at equal hole concentration in the films. The hole lifetime decreases with increasing the extracted charge density due to the higher hole density in the CuCrO 2 layer and a larger driving force for the charge recombination. Figure 5c shows the charge collection efficiency (g cc , g CC~ffi ffiffiffiffiffiffiffiffi ffi and d are the hole lifetime, hole diffusion coefficient and the film thickness, respectively) at the back TCO (FTO) as a function of opencircuit potential (i.e., under different light intensity irradiation). Both devices give similar t and g cc values, clarifying that addition of Au NPs into CuCrO 2 film has no effect on the interfacial charge recombination and the charge collection efficiency.

Discussion
In this communication, it is the first time that an enhanced photocurrent due to the magnetic field effect was observed for DSSC devices endowed with p-type delafossite CuCrO 2 and P1 dye in combination with organic thiolate-based electrolyte. The positive magnetic field effect on photocurrent of p-type DSSC based on CuCrO 2 photocathode can be related to the tunning electron-hole pair dissociation in exciting singlet state dye 18,25,[30][31][32] . Previously investigations suggested that an increase in photocurrent of DSSC devices associated with Au or other metal NPs was mainly attributed to the SPR effect [33][34][35][36][37][38] . In a majority of cases the evidence includes an increase in optical cross sectional areas of absorbers, thus a corresponding enhancement in PCE. Kamat et al also argued that the SPR effects gave rise to the photocurrent of DSSC while the charging effects of metal NPs leaded to an increase in photovoltage 31 . However, in this study we found that the addition of Au NPs into the CuCrO 2 film brought negligible influence on the photon adsorption ability of the film and the interfacial recombination kinetic dynamics as well as the electronic energy levels of the semiconductor. Those parameters are related to the light-harvesting efficiency and the charge collection efficiency. It is well known that the IPCE spectra are determined by the sensitizer light-harvesting efficiency, the quantum yield for hole injection from the excited state of the dyes to the valence band of the semiconductor, and the collection efficiency of the hole at the back contact. Therefore, the IPCE gain observed in this study could be expected from an increase of the charge injection from the excited P1 dyes into CuCrO 2 valence band. In order to investigate the nature of the charge injection efficiency enhancements at visible wavelengths, the internal quantum efficiency, or absorbed photon conversion efficiency (APCE) of the tested devices were determined. These efficiencies were calculated through the following equations: where A l is the absorbance of a component in the layered structure, corrected for the substrates and the electrolyte, LHE l is the light harvesting efficiency. The APCE proves a precise representation of the intrinsic light conversion efficiency. The LHE and APCE values determined by equation 1a and 1b are plotted in Figure 3c. The device B shows a broader and higher APCE response above 480 nm than that of device A. At wavelength of 500 nm (close to the adsorption peak of P1 sensitizer, see Figure 2d), the APCEs values are about 16% and 28% for devices A and B (Figure 3c) respectively. This observation confirms the proposed benefits of the heterojunction architecture as discussed above: stronger APCE response at visible wavelengths suggesting an enhancement of charge separation. This result also indicates that the Au NPs-doped CuCrO 2 can improve the absorbed photon-to-current conversion efficiency through an increased dissociation of excitons. Several works have demonstrated that the electromagnetic field can be remarkably changed by the light polarization 39 . Theoretical and experimental results have disclosed the relation between the SPR and magnetic filed, showing that the excitation of surface plasmons can greatly enhance electromagnetic fields [40][41][42] . The effect of the light wavelength on the intensity of local electromagnetic field around Au@SiO 2 surface was evaluated (see Figure 6). The simulation was carried out based on a two-dimensional model at the finite element method supported by COMSOL Multiphysics software (CnTech). Simulation results showed that a strong electromagnetic field (about 50 V m 21 ) could be generated near onto the surface of Au NPs for the incident irradiation in the range from 450 to 650 nm, but decreased quickly with increasing distance from the surface of Au@SiO 2 NPs.
The dye molecules can collect the light energy and transfer their singlet excitation (functioning as an antenna) and, at the same time, dissipate the singlet and triplet energies (functioning as a selfquencher). For a series of changes among the singlet-excited and redox states having a singlet character in P1-sensitized CuCrO 2 device, upon absorption of photon, electron is transferred to a higher singlet level. Then hole injection takes place to generate a chargeseparated state having a singlet character on the P1/CuCrO 2 boundary. Hole is transferred further into CuCrO 2 to form a stable chargeseparated state. A reverse electron transfer followed by charge recombination takes place to relax into the ground state. If considering the generation of the triplet-excited and radical-cation states both having a triplet character, singlet excitons can be partially converted into triplet excitons through intersystem crossing 26 . Both singlet and triplet excited states have different contributions to the charge dissociation owing to their different binding energies. A positive magnetic field effect was observed in small molecules, which was attributed to a change in the conversion rate between singlet and triplet prior to recombination 43 . In this study, the tripheneylanmine moiety of the P1 sensitizer presents an aromatic molecular structure. The internal splitting energy was assumed to be about 1-10 meV 44 . This small internal splitting energy suggested that a magnetic field of 100 mT or a local electromagnetic field of 50 V m 21 could influence the intersystem crossing process and consequently change the singlet and triple ratios.
This study showed that the addition of Au NPs into a p-type oxide semiconductor delafossite CuCrO 2 film used for dye-sensitized solar cells significantly increased the cell performance. The presence of 0.1 wt% Au NPs in a CuCrO 2 based p-type DSSCs increased the PCE to 0.31%. This result should be extendable to other p-type delafossite materials of appreciate conductivity. It leads to the general importance of increasing charge injection in p-type DSSC, and their adverse effects on both J sc and power conversion efficiency.

Methods
Synthesis of spherical Au@SiO 2 nanoparticles. 1 mL sodium citrate (10 mM) and 1 mL HAuCl 4 (10 mM) were dissolved in 37 mL deionized water. Then, 1 mL NaBH 4 (0.1 M) was added into the above solution. This produces Au seeds solution. Meanwhile, 50 mM CTABCl was added into 100 mL HAuCl 4 (0.25 mM) aqueous solution, followed by 0.5 mL ascorbic acid (100 mM), 0.5 mL NaOH (100 mM) solution,and finally the 0.1 mL above-mentioned Au seeds solution. After 10 h, the mixture was dissolved into 5 mL NaOH (1 mM) aqueous solution again. Lastly, 10 mL TEOS was added into the solution and stirred for 10 h.
Devices fabrication and characterization. The CuCrO 2 film was sintered at 420uC for 40 min then 550uC for 40 min in Ar. After cooled down to about 80uC, the film electrodes were dipped into a 300 mM P1 solution in acetonitrileat room temperature for 16 hrs. After washed with acetonitrile and dried by air flow, the sensitized CuCrO 2 electrodes were assembled with counter electrodes. The working electrodes and counter electrodes were separated by a 45 mm thick hot melt ring (Surlyn, Dupont) and sealed by heating. The internal space was filled with liquid electrolytes using a vacuum back filling system. The electrolyte for devices was 0.3 M T 2 and 0.9 M T 2 with the tetramethylammonium cation in the mixture of acetonitrile and propylene carbonate (volume ratio, 753).
A 450 W xenon light source solar simulator (Oriel, model 9119) with AM 1.5 G filter (Oriel, model 91192) was used to generate an irradiance of 100 mW cm 22 at the surface of the test cells. The current-voltage characteristics of the cells under these conditions were obtained by applying external potential bias to the cell and measuring the generated photocurrent with a Keithley model 2400 digital source meter (Keithley, USA). A similar data acquisition system was used to control the IPCE measurement. A white light bias (1% sunlight intensity) was applied onto the sample during the IPCE measurements with ac model (10 Hz).
Photovoltage/photocurrent transient decay and charge extraction measurements. For the photovoltage transient decay measurement, a white light bias was generated by an array of diodes. Red light pulse diodes (0.05 s square pulse width, 100 ns rise and fall time) controlled by a fast solid-state switch were used as the perturbation source. The voltage dynamics were recorded on a PC-interfaced Keithley 2602A source meter with a 100 ms response time. The perturbation light source was set to a suitably low level in order for the voltage decay kinetics to be monoexponential. By varying the white light bias intensity, the recombination rate constant and the hole diffusion rate constant could be estimated over a range of applied biases, which were used to evaluate the charge collection efficiency (g cc ). Before the LEDs switched to the next light intensity, a charge extraction routine was executed to measure the electron density in the film. In the charge extraction techniques, the LED illumination source was turned off in ,1 ms, while simultaneously, the cell was switched from open to short circuit, in which a resistance is series connected to the device. The resistance is adjusted close to the corresponding value at maximum power point of the device. The resulting current, as the cell returns to V 5 0 and J 5 0, was integrated to give a direct measurement of the excess charge in the film at that V oc , which is the minimum level of charge in the semiconducting photocathode.
Magnetic field effect measurements. To define the effect of external magnetic fields, the test cells were placed between the pole pieces of an electromagnet. The surface of sample was parallel to the magnetic field. A simulated sunlight (with of about 1% of full sun intensity) was introduced by optical fiber (Newport, 77632). The generated photocurrent dates were recorded on a PC-interfaced Keithley 2400 source meter.