Effect of TiO2 blocking layer on TiO2 nanorod arrays based dye sensitized solar cells

Highly ordered rutile titanium dioxide nanorod (TNR) arrays (1.2 to 6.2 μm thickness) were grown on TiO2 blocking layer chemically deposited on fluorine doped tin oxide (FTO) substrate and were used as photo-electrodes to fabricate dye sensitized solar cells (DSSC's). Homogeneous layer of TiO2 on FTO was achieved by using aqueous peroxo- titanium complex (PTC) solutions via chemical bath deposition. Structural and morphological properties of the prepared samples were investigated using X-ray diffraction (XRD), scanning electron microscopy (SEM) measurements. TNR arrays (6.2 μm) with TiO2 blocking layer showed higher energy conversion efficiency (1.46%) than that without TiO2 blocking layer. The reason can be ascertained to the suppression of electron-hole recombination at the semiconductor/electrolyte interface by the effect of TiO2 blocking layer.


Introduction
In DSSC's, the TiO 2 photo-electrode essentially collects and transports the photo-generated electrons [1,2]. However, the electron transport in nanoparticle based device is suffered by trap limited diffusion process, a slow recombination step that results in low efficiency. Also the large grain boundaries of TiO 2 nanoparticles lead to fast recombination of the photo-generated electrons with I 3 − in the electrolyte prior to their collection at the electrode. Hence, to provide uninterrupted transport of photo-generated electrons in TiO 2 , one dimensional (1D) nanostructure such as nanorods [3], nanowires [4], and nanotubes [5] etc. have been designed and demonstrated for their potential application in DSSC's. It has been proved that the electron transport in 1D nanostructure is expected to be several orders of magnitude faster than that of nanoparticle network [6,7]. Several researchers fabricated thick 1D nanostructure arrays and achieved higher energy conversion efficiency [8,9]. Another approach to improve the charge transport properties and collection efficiency is by introducing TiO 2 seed layer on the FTO layer prior to the growth of 1D nanostructure [10][11][12][13][14][15]. Recently, Qian et al. [16] and Yang et al. [17] derived TiO 2 blocking layers from peroxo-titanium solution and studied their effect on dye sensitized solar cells. Increase in photocurrent by the introduction of PTS overlayer was mainly attributed to the retardation of the back-transport reaction in the FTO/TiO 2 interfaces. In a systematic approach, hereby we made an endeavour to introduce TiO 2 blocking layer through per-oxo titanium method on TNR arrays based DSSCs.

Fabrication of TNR arrays.
TiO 2 nanorod arrays were grown on FTO PTC substrate using Teflonlined stainless steel autoclave (45 mL volume, Parr Instrument Co.) following the hydrothermal method [3]. Briefly, 0.30 mL of TiCl 4 was added to a mixture solution containing 5 mL of deionized water and 5 mL of hydrochloric acid (37 wt%) in a teflon container and was stirred for 5 minutes for homogenization. FTO PTC glass plates were placed at an angle against the wall of the Teflon-liner with the conducting side facing downwards and the teflon liner was placed in a stainless steel autoclave to initiate the hydrothermal reaction. The reaction was carried out at 150 °C in an electric oven for 4, 8, 12, 20, 30 and 40 hours. For comparison TNR arrays were grown on bare FTO glass plates for 20 hours at 150 °C. After the set reaction period, the samples were taken out and rinsed extensively with deionized water and left to dry in ambient air. Finally, all the samples were annealed at 450 °C for 30 minutes.

Analytical measurements
X-ray diffraction patterns were recorded using a Bruker D4 Endeavor advance diffractometer with Cu K radiation at a scanning rate of 2º/min. Morphology and microstructure analysis of the samples were performed using a JEOL JSM-6380LV scanning electron microscope (SEM). The samples under study were first sputtered with gold for 3 min. to obtain a layer of 150 A° thick using an Edwards S 150 Sputter Coater.

Fabrication of dye sensitized solar cells
TNR array electrodes were dipped in dye solution (3×10 −4 M) in ethanol for 24 h. After the dye adsorption was complete, the electrode were withdrawn from the solution and dried under a stream of argon. The platinum counter electrode and the dye-coated TiO 2 electrode was then put together with a thin transparent film of Surlyn polymer frame (25 m). The sandwiched electrodes were tightly held and then heated to 130 °C, so that the Surlyn frame would seal the two electrodes intact. A thin layer of electrolyte, consisting 0.6 M propyl methylimidazolium iodide (PMII), 0.05 M I 2 , 0.1 M LiI, and 0.5 M tert-butylpyridine in 1:1 acetonitrile/valeronitrile, was introduced into the inter-electrode space from the counter electrode side through predrilled holes. The drilled holes were sealed with a microscope cover slide to avoid leakage of the electrolyte solution.

Photovoltaic Measurements
Solar simulator device (Model: CT80AAA, PET Photo Emission Tech., Inc., USA) incorporated with a 300 W xenon arc light source was used for measuring the photovoltaic properties. A 1 sun, AM 1.5G, light irradiance of 100 mW/cm 2 was achieved over the surface of the test cell. During each measurement, temperature of the platen which holds the cell was controlled at 25°C using a control system attached to the solar simulator. The power of simulated light was calibrated using a reference silicon photodiode (20mm x 20mm) provided with the instrument. The test cell was placed at the same position as the calibration cell. Photocurrent-voltage plots for all cells were recorded using a Keithley 2400 digital source meter, from +950 to −300 mV, in steps of 10 mV. Figure 1 a and b shows the scanning electron microscopic images of bare FTO and FTO PTC . A homogeneous layer of TiO 2 was observed for the TiO 2 blocking layer deposited over bare FTO (Fig.  1b). Small grains with dense structure and well surface coverage are observed for chemically deposited TiO 2 blocking layer. Accordingly, this blocking layer consisting of anatase TiO 2 nanocrystal formed upon high temperature treatment of per-oxo titanium complex [18]. Apparently, TNR arrays grown on FTO PTC (Fig. 2a) and bare FTO (Fig. 2c) reveal uniform surface coverage and the shape of the nanorod is tetragonal with square top facets as expected for rutile TiO 2 structure. From cross sectional SEM images, the thickness of TNR arrays grown on FTO PTC were found to be 1.2 m, 2.5 m, 4.2 m and 6.2 m for the hydrothermal duration set to 4, 8, 12 and 20 h respectively. For comparison, TNR arrays grown on bare FTO results in similar length, 6.1 m (Fig. 2d). On prolonging the hydrothermal reaction to 30 and 40 h, no increase in nanorod length was observed in FTO PTC and bare FTO.   Figure 3 shows the X-ray diffraction patterns of bare FTO and FTO PTC and TNR arrays grown on FTO PTC and bare FTO. Diffraction peaks of bare FTO (Fig. 3a) correspond to tetragonal rutile phase of SnO 2 . The diffraction peaks of TiO 2 crystallites were not observed for TiO 2 blocking layer deposited FTO (Fig. 3b) due to the strong diffraction peaks of FTO. However, after the hydrothermal growth of TNR arrays on FTO PTC and bare FTO, peaks corresponding to tetragonal TiO 2 with rutile phase were observed ( Fig. 5c-g). The intensity of the peak (002) increases as the hydrothermal duration increases. Significantly, the enhancement of diffraction peak (002) is a clear indication of directed growth of TNR arrays perpendicular to the FTO substrate. Absence of diffraction peaks at (110), (111), and (211) is an evidence for uni-directional growth of TNR arrays with respect to the substrate surface.  Table 1. It was observed that the efficiency increases as the thickness of the TNR arrays increases. DSSC assembled with 6.2 m thick TNR arrays grown on FTO PTC exhibits a photocurrent of 3.56 mA/cm 2 , open circuit voltage of 0.79 V and overall energy conversion efficiency of 1.46%. Whereas, TNR arrays (6.1 m) grown on bare FTO showed marginally low efficiency (1.19%). It is noteworthy that the DSSCs assembled from TNR arrays with TiO 2 blocking layer show higher V OC (0.79 V) than that without TiO 2 blocking layer (0.74 V). This indicates that TiO 2 blocking layer plays a vital part on the performance of rutile TiO 2 nanorod based DSSC's. This is because, TiO 2 blocking layer covering the vacant spaces unoccupied by TNR arrays on FTO are effective in blocking the direct contact of the redox electrolyte and FTO and diminishes the dark current that arises from reduction of I 3 at the FTO/electrolyte interface. Also the increase in J SC proves the lower recombination rate of the whole photo anode in the presence of TiO 2 blocking layer. We note that such a per-oxo titanium route presented here offers a significant advantage for the DSSCs.

Conclusions
TiO 2 nanorod arrays were grown on TiO 2 blocking layer chemically deposited on FTO substrate from per-oxo titanium complex solution. The peroxo-titanium route provides oxygen rich environment to enhance the Ti-O-Ti linkage which avoids the atomic incorporation of chlorine. DSSC assembled using TNR arrays (6 m) in the presence of TiO 2 blocking layer produces 1.46% efficiency which is higher than that assembled in the absence of TiO 2 blocking layer (1.19 %). This enhancement was attributed to the effect of TiO 2 blocking layer on suppressing the dark current that arises from the direct contact of electrolyte with the FTO. These findings provide a new route to preparing highly pure TiO 2 nanocrystals on FTO and further extend their application into solid state DSSCs.