Received: 05 Oct 2018
Revised: 27 Dec 2018
Accepted: 01 Jan 2019
Published online: 02 Jan 2019
M. A. Waghmare1,2, N. I. Beedri2, A. U. Ubale1*and H. M. Pathan2
1Nanostructured Thin Film Materials Laboratory, Department of Physics,
Govt. Vidarbha Institute of Science and Humanities, Amravati - 444604, India
2Advanced Physics Laboratory, Department of Physics, Savitribai Phule Pune
University, Pune - 411007, India
E-mail: ashokuu@yahoo.com
Titanium oxide (TiO2) electrode has been the most commonly used photo-electrode for the dye-sensitized solar cells (DSSCs). Several research groups have already reported that only TiO2 layer is not yet ideal for electron transfer in the absence of space charge layer and also demonstrated the procedures for coating nanocrystalline semiconducting oxide films with a thin overcoat of a different semiconducting oxide with a higher conduction band energy level (EC). Zirconium oxide (ZrO2) is a suitable material for such overcoat because of its higher EC. The binary TiO2–ZrO2 oxide photo-electrodes were prepared by doctor blading technique. The electrodes were annealed at 450 °C and then sensitized by Rose Bengal (RB) dye. The DSSC fabricated by binary TiO2-ZrO2 photo-electrode showed improved solar energy conversion efficiency than that of fabricated only by pure component of TiO2.
1. Introduction
DSSC based on nanocrystalline TiO as a porous photo-electrode was firstly reported by O'regan and Gratzel.1 As far as all solar energy devices are concerned, DSSCs have attained significant importance due to its low production cost as compared to other solar energy conversion devices. The power conversion efficiencies of TiO based DSSCs 2 considerably get influenced by its geometrical structures such as rutile, anatase and brookite.2 In DSSCs, anatase (3.2 eV) TiO nanopowder is commonly used to prepare photo-electrode films due to its higher Fermi-energy level and greater band gap energy as compared to rutile (3.0 eV).3 Photovoltaic properties of DSSCs can be improved by modifying the properties of photo-electrode films. The charge transport and electrochemical properties of TiO can be enhanced by introducing 2 dopant such as W, Ta, Ni, Zr and Nb.4-7 As far as these dopants are concerned, zirconium (Zr) has attained considerable interest for DSSC as it has the same valence shell structure (n-1) d2 ns2 and same valence state as titanium (Ti).6,8-10 There are many techniques reported to enhance the open circuit voltage (V ) of device. The V of the device can be OC OC increased either by implementing bilayer technique11-14 or by introducing changes in the structure of metal oxide semiconductor.15,16 The bilayer technique is used to decrease recombination reactions. In this technique, a thin layer of another metal oxide having more negative conduction band edge (E ) as compared to photo-electrode material is deposited on C photo-electrode film. Another possible alternative to enhance V is to OC increase E of the photo-electrode material by modifying its structure as C V depends upon the electron energy difference between the OC conduction band edge of photo-electrode and the redox potential level of electrolyte.17 Kitiyanan et al.18 have employed sol-gel method to prepare TiO -ZrO (95 % TiO + 5 % ZrO ) mixed oxide system to 2 2 2 2 increase efficiency of the cell. The cell parameters such as V enhances OC to ~4 % , J to ~11 % and to ~17 % as compared to device SC η prepared using TiO photo-electrode film. Durr et al.19 have revealed 2 that for mixed Ti-Zr oxide system in DSSC, the V of the device OC enhances with increase in Zr percentage. For higher Zr content, the device conversion efficiency decreases due to reduced electron injection. Menzies et al.20 have investigated the photovoltaic performance of TiO and ZrO -coated TiO photo-electrodes calcined at 2 2 2 450 oC in a 2.45 GHz microwave furnace. The ZrO shell-coating acts 2 as an energy barrier between the photo-injected electrons in the conduction band of TiO and the oxidized electrolyte species/dye 2 molecules. The presence of ZrO coating reduces the recombination rate 2 of the injected photo-electrons into the conduction band of TiO to 2 either the oxidized electrolyte species or the oxidized dye molecules. The microwave calcined TiO photo-electrode was found to be 2.4 % 2 efficient whereas ZrO -coated TiO showed efficiency up to 3.1 %. The 2 2 presence of ZrO shell showed enhancement in all the cell parameters 2 such as J V and FF. In DSSC, optical absorption can be enhanced by SC, OC introducing light scattering layer forming double layer structure. Due to this light harvesting layer, the incident light remains confined within the photo-electrode which increases photocurrent density and efficiency of the device. Moradzaman et al.21 have studied absorption effects, light harvesting and light scattering in DSSC using Zr-doped TiO photo- 2 electrode. The double layer photo-electrode consisting of mixture 0.025 wt % carbon nanotubes (CNTs) with TiO nanoparticles and 0.025 mol 2 % Zr doped TiO nanoparticles as an over layer and 0.025 mol % Zr 2 doped TiO nanoparticles as an under layer shows highest efficiency of 2 8.19 %. Sayyed et al.22 have studied Rose Bengal (RB) dye-sensitized TiO –CeO bilayer photo-electrodes for DSSC. They demonstrated that 2 2 the introduction of CeO layer reduces the recombination of photo- 2 injected electrons with either the oxidized dye molecules or oxidized redox species which resulted into 38.10 % increase in J and 66.67 % SC increase in V as compared to the pure TiO photo-electrode. Mohamed OC 2 et al.23 have revealed the influence of nanofibrous morphology and Zrdoping on the optical and electrical properties TiO . The DSSC 2 fabricated using 1 % Zr-doped TiO nanofibers showed highest 2 photovoltaic efficiency of 4.51 % as compared to other doped samples (0.5, 1.5 and 2 % Zr-doped TiO nanofibers). Beedri et al.24 have 2 enlightened the use of Nb O as an energy barrier layer in N3 dye- 2 5 sensitized ZnO/Nb O photo-electrode based DSSC. They have noticed 2 5 10 % enhancement in the power conversion efficiency in ZnO/Nb O 2 5 photo-electrode due to increase in the lifetime of photo-injected electron as compared to the pure ZnO photo-electrode. They have also demonstrated the use of bilayer structure in N3 dye-sensitized TiO /Nb O photo-electrode based DSSC to improve the overall power 2 2 5 conversion efficiency.25
Anchoring group of dye sensitizer plays a vital role in deciding efficiency of the device. In the present work, Rose Bengal is used as a dye sensitizer. RB is a xanthene class of organic dye sensitizer which is widely used due to its noticeable merits such as low cost, moderate oxidation/reduction ability and high coefficient of absorption.26-27
The purpose of the present work is to investigate photovoltaic properties of DSSC with a porous TiO -ZrO layer. ZrO layer acts as an 2 2 2 energy barrier layer between the oxidized redox couple/dye and photoinjected electrons to minimize recombination pathways.
2. Experimental section
2.1 Preparation of TiO photo-electrodes
TiO paste was prepared as per the literature method mentioned by 2 Waghmare et al.28. The TiO paste was prepared by mixing 4 g TiO2 2 nanopowder (SRL, APS. ~ 50 nm), 2 g ethyl cellulose (SRL), 1.8 ml terpineol (SRL), 5-6 drops of acetylacetone (SRL) and 20 ml ethanol (SRL). The TiO paste was uniformly spread over the area of 0.6 cm x 2 0.4 cm (0.24 cm2) on the FTO glass by doctor blade technique. The TiO photo-electrode films were annealed at 450 oC for 1 h.
2.2 Preparation of TiO - ZrO photo-electrodes
ZrO paste preparation was adopted from Waghmare et al.28. The ZrO2 2 paste was prepared by mixing 4 g ZrO nanopowder (SRL, APS. ~ 45 2 nm), 2 g ethyl cellulose (SRL), 1.8 ml terpineol (SRL), 5-6 drops of acetylacetone (SRL) and 20 ml ethanol (SRL). The TiO –ZrO photo- 2 2 electrode films were prepared by spreading the ZrO paste uniformly 2 over the area of 0.6 cm x 0.4 cm (0.24 cm2) on the preformed TiO2 photo-electrode film by doctor blade technique. The TiO –ZrO photo- 2 2 electrode films were annealed at 450 oC for 1 h. The photo-electrodes were prepared with different over layers of ZrO on TiO films. All the 2 2 samples are named as TiZr , where x is the number of over layers of x ZrO on TiO films, e.g. TiZr is one over layer of ZrO deposited on 2 2 1 2 TiO film (see Table 1).
2.3 DSSC fabrication
The TiO -ZrO photo-electrodes were kept in 0.5 mM Rose Bengal dye 2 2 solution prepared in ethanol at room temperature for 24 h. Thin Pt coated FTO glass was used as a counter electrode. The counter electrode was then kept in contact with the TiO -ZrO photo-electrode 2 2 films. A spacer was inserted between the TiO -ZrO photo-electrode and 2 2 the counter electrode to prevent the direct contact between them. The redox electrolyte was consisting of a 0.5 M lithium iodide (LiI) solution, a 0. 05 M iodine (I ) solution and a 0.5 M 4-tert-butylpyridine 2 solution in acetonitrile. It was introduced between the gap of the photoelectrode film and the counter electrode.
2.4 Characterization
X–ray diffraction (XRD) of the photo-electrode films was performed for phase identification and crystallite size determination by X-ray diffractometer (XRD, Rigaku ‘‘D/B max-2400’’, λ = 1.54 Å). The surface morphological and elemental analyses of the photo-electrodes were carried out by scanning electron microscopy (SEM, JEOL-JSM 6360-A) and energy-dispersive X-ray spectroscopy (EDX), respectively. The UV- vis absorption spectra of Rose Bengal in ethanol and adsorbed on TiO -ZrO photo-electrode films were measured by UV–vis 2 2 spectrometer (JascoV-670) in the range 200–800 nm. Electrochemical impedance spectroscopic (EIS) analyses were carried out to study the electron transfer at the photo-electrode / electrolyte interface. The applied ac signal voltage is 0.01 V in the frequency range of 1 Hz–1 MHz. EIS performed using Potentiostat/ Galnostat (Ivium Soft: Vertex) under dark. The photovoltaic performance of the fabricated cells was studied under an illumination of Xenon lamp of ~ 72 mW/cm2 intensity using kiethley source meter (2420).
3. Results and discussion
3.1 XRD investigations of the photo-electrodes
Fig. 1 illustrates the diffraction patterns of ZrO , TiO and ZrO coated 2 2 2 TiO photo-electrode films on FTO substrate. The diffraction peaks 2 observed at 26.44o, 33.64o, 37.71o, 51.48o, 61.48o and 65.40o are corresponding to crystal planes of (110), (101), (200), (211), (310) and (301), respectively can be ascribed to FTO [JCPDS card No. 46-1088]. The diffraction peaks of ZrO can be attributed to monoclinic ZrO 2 2 [JCPDS card No. 37-1484] at 2θ values of 24.14o, 24.55o, 28.27o, 31.56o, 34.25o, 35.41o, 40.84o, 49.35o, 50.23o, 54.14o and 55.47o corresponding to the crystal planes of (110), (011), (-111), (111), (200), (002), (-112), (220), (022), (003) and (310) respectively (see Fig. 1a and Table 2). The crystallite size of the ZrO nanopowder was calculated according to the 2 Scherrer formula29 along the most intense peak (-111). The crystallite


Fig 1. XRD patterns of (a) ZrO2 , (b) TiO2 and (c) ZrO2 coated TiO2 photo-electrodes on FTO substrate
size of the monoclinic ZrO was 40 nm. Fig. 1b clearly demonstrates 2 that the diffraction peaks of TiO photo-electrode film can be attributed 2 to anatase TiO [JCPDS card No. 21-1272] at 2 values of 25.32o, 2 θ 37.73o, 48.07o, 55.07o and 62.68o , corresponding to the crystal planes of (101), (004), (200), (211) and (204), respectively.23 The crystallite size of the TiO sample was calculated according to the Scherrer formula 2 along the most intense peak (101). The crystallite size of the anatase TiO was 15.64 nm. The 2 and (hkl) values of TiO are summarized in 2 2 θ Table 3. The ZrO coated TiO film shows diffraction peaks of anatase 2 2 phase of TiO as well as monoclinic phase of ZrO structures. All the 2 2 films samples (TiO , TiZr , TiZr and TiZr ) exhibit identical diffraction 2 1 2 3 peaks of TiO which confirm that anatase TiO phase was retained in all 2 2 photo-electrode films as it is required because of its photo-active nature.18, 30
3.2 SEM and EDX studies of the photo-electrodes
Fig. 2a-c presents SEM images of TiO , ZrO and ZrO coated TiO 2 2 2 2 photo-electrode films. The images show porous nature of the film samples. Porous films play a vital role in efficient light harvesting because light harvesting depends upon amount of dye adsorbed on the surface of photo-electrode film. Achieving efficient light harvesting, efficiency of the device can be enhanced. The compositional analyses were carried out using EDX technique. Fig. 3a-c displays the EDX patterns of TiO , ZrO and ZrO coated TiO photo-electrode films. The 2 2 2 2 existence of Ti, Zr and O was confirmed from EDX data.
3.3 Optical properties of the photo-electrodes
Fig. 4 shows the plot of ( h )2 vs h of TiO photo-electrode. The 2 α υ υ optical band gap of TiO is estimated from the plot of ( h ) 2 vs h . The 2 α υ υ band gap was found to be 3.22 eV. Many research groups have reported only an indirect band gap of 3.23 eV for anatase TiO . The obtained 2 results are in good agreement with the previously reported results.31-33 The optical band gap of ZrO is calculated from the graph of ( h ) 2 vs 2 α υ hυ (see Fig. 5). The band gap was found to be 5.14 eV. The obtained band gap value of ZrO is consistent with the previously reported band 2 gap values.34-37 Absorption spectra of Rose Bengal dye is shown in Fig. 6a. The absorption spectra generally exhibits two absorption peaks. The peaks are observed in the short wavelength region (258 nm) as well as in the long (554 nm) wavelength region. The absorption peak in the long wavelength region attributed to the intra-molecular electron transfer within the highest occupied molecular orbital energy level (HOMO) to the lowest unoccupied molecular orbital (LUMO) energy level while the peak in the short wavelength region could be assigned to π to π* electron transition.38, 39 It is well known that the absorption of TiO is limited to ultraviolet region (Fig. 6b) whereas the absorption of 2 RB sensitized TiO (Fig. 6c) and RB sensitized ZrO coated TiO (Fig. 2 2 2 6d) is extended to visible region (400 nm–600 nm).

Fig 2. SEM images of (a) TiO2 , (b) ZrO2 and (c) ZrO2 coated TiO2 photo-electrode films.
Fig 3. EDX measurements of (a) TiO2 , (b) ZrO2 and (c) ZrO2 coated TiO2 photo-electrode films.
Fig 4. Plot of ( h )2 vs h of TiO2 photo-electrode film.
Fig 5. Plot of ( h )2 vs h of ZrO photo-electrode film.
Fig 6. UV-vis spectra of (a) RB dye, (b) bare TiO2 photo-electrode, (c) RB sensitized TiO2 and (d) RB sensitized ZrO2 coated TiO2 photo- electrodes.
Fig 7. J/V characteristics of DSSCs fabricated using Rose Bengal sensitized (a) TiO2 , (b) TiZr1 , (c) TiZr2 and (d) TiZr3 films. The data is obtained for the films of areas 0.24 cm2 and under AM 1.5 simulated sunlight, ~72 mW /cm2.

3.4 Photo-electrochemical characterization
Fig. 7 shows photovoltaic properties of Rose Bengal-sensitized TiO , 2 TiZr , TiZr and TiZr photo-electrode films under light intensity of ~72 1 2 3 mW/cm2. The photovoltaic properties of Rose Bengal-sensitized TiO , 2 TiZr , TiZr and TiZr photo-electrode films prepared from commercial 1 2 3 nanopowder by doctor blade technique are listed in Table 4. The J was SC 0.121 mA/cm2 for TiO , 0.106 mA/cm2 for TiZr , 0.075 mA/cm2 for 2 1 TiZr , 0.084 mA/cm2 for TiZr . For TiO , TiZr , TiZr and TiZr , the V 2 3 2 1 2 3 OC was 0.37 V, 0.49 V, 0.35 V and 0.33 V, respectively. A large J was SC observed for TiO photo-electrode but V was less as compared to 2 OC TiZr . The power conversion efficiency for TiZr was 0.038 % while it 1 1 was 0.029% for TiO , 0.020% for TiZr and 0.019 % for TiZr . It was 2 2 3 noticed that the device fabricated with TiZr photo-electrode film 1 performed better as compared to other photo-electrodes. The ZrO layer 2 on TiO photo-electrode film increases the V by 32.43 % as compared 2 OC to pure TiO . It was well reported that the interfacial electron 2 recombination reactions decide V of the device. The V can be OC OC increased if the electron recombination reactions are minimized.40 The implementation of ZrO as an energy barrier layer between the TiO 2 2 photo-electrode and electrolyte reduces the back electron transfer from CB of TiO to I - in the electrolyte or to dye, thus improves V 41. The 2 3 OC overall device efficiency shows a large enhancement 31.03 % as compared to pure TiO photo-electrode film. The open circuit voltage decreases with increase in number of over layers of ZrO which 2 decreases the photovoltaic efficiency (0.020 % for TiZr and 0.019% for 2 TiZr ). It was observed that the introduction of ZrO over layer 3 2 decreases the recombination pathways which enhances the device performance but it was also noticed that with increase in number of over layers, the electron tunneling path length through the ZrO barrier layer 2 into TiO photo-electrode film increases which decreases the overall 2 device performance for TiZr and TiZr .42 If ZrO film thickness is 2 3 2 increased then the photo-excited electrons will not readily get injected into the conduction band of TiO photo-electrode film and they may 2 recombine with the oxidized dye or electrolyte. These recombination reactions decrease the photo-current density.30 Hence, for thicker ZrO 2 over layers, a significant decrease in the photovoltaic parameters was observed.43
In DSSCs, the electrochemical impedance spectroscopy (EIS) is a powerful technique to explore the charge transfer dynamics at the film interfaces.44 Fig. 8 shows the Nyquist plots recorded in dark under forward bias (0.7 V) condition of the fabricated DSSCs using TiO and 2 different over layers of ZrO on TiO photo-electrode films (TiZr , TiZr 2 2 1 2 and TiZr ). As observed in Fig. 8 the Nyquist plots exhibit two 3 semicircles, the first semicircle explains the charge-transfer resistance at counter electrode/electrolyte and the second circle reveals charge recombination resistance at the TiO /electrolyte interface (R ).45 The rate at which the charge transfer takes place depends upon the radius of semicircles. As listed in the Table 5, the R value of the TiO , TiZr , rec 2 1 TiZr and TiZr photo-electrode films were 64.6 , 92.9 , 102.2 and 2 3 128.8 , respectively. The TiO photo-electrode film shows low charge 2 recombination resistance as compared to other photo-electrode films. The implementation of ZrO over layers on TiO photo-electrode films 2 2 increases the R values. The introduction of ZrO over layer plays a rec 2 decisive role in minimizing charge recombination losses as it acts as a barrier layer between the TiO and dye or redox electrolyte. Fig. 9 2 displays Bode phase plot of the fabricated DSSCs using TiO and 2 different over layers of ZrO on TiO photo-electrode films. The electron 2 2 lifetime ( ) in TiO , TiZr , TiZr and TiZr photo-electrode films can be e 2 1 2 3 τ calculated using the equation: = 1/2 f , where f is the maximum e max max τ π peak frequency. As listed in Table 5, the f values of TiO , TiZr , TiZr , max 2 1 2 and TiZr are 251, 174, 912 and 1320 Hz and the electron life time 3 values are estimated to be 0.63, 0.91, 0.17 and 0.12 ms. The f values max exhibit a decrease from TiO to TiZr . The electron life time in TiZr 2 1 1 was prolonged from 0.63 to 0.91 ms. The prolonged electron life time indicates minimization in charge recombination losses which enhances the overall power conversion efficiency of the device. The obtained results are in good agreement with the J/V analysis data.
Fig 8. Nyquist plots observed for the DSSCs fabricated using Rose Bengal sensitized (a) TiO2 (b) TiZr1 (c) TiZr2 and (d) TiZr3 films.
Fig 9. Bode phase plots observed for the DSSCs fabricated using Rose Bengal sensitized (a) TiO2 (b) TiZr1 (c) TiZr2 and (d) TiZr3 films.

4. Conclusions
We have demonstrated a novel approach to improve the efficiency of dye-sensitized solar cell. We have proven that the introduction of ZrO2 layer over TiO photo-electrode film improves the photovoltaic 2 parameters significantly which leads to increase in the overall power conversion efficiency.
Acknowledgement
Authors are thankful to Board of College and University Development (BCUD), Savitribai Phule Pune University, Pune for financial support under the project.
References
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
18.A. Kitiyanan, S. Pavasupree, T. Kato, Y. Suzuki and S. Yoshikawa, As. J. Energy Env., 2005, 6, 165-174.
19. M. Dürr, S. Rosselli, A. Yasuda and G. Nelles, J. Phys. Chem. B, 2006, 110, 21899-21902.
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
CrossRef View Record in Scopus
37. C. R. Chintaparty, Optik, 2016, 127, 4889-4893.
CrossRef View Record in Scopus
38. S. P. Singh, M. S. Roy, K. J. Thomas, S. Balaiah, K. Bhanuprakash and G. D. Sharma, J. Phys. Chem. C, 2012, 116, 5941-5950.
CrossRef View Record in Scopus
39. R. Schlaf, P. G. Schroeder, M. W. Nelson, B. A. Parkinson, C. D. Merritt, L. A. Crisafulli, H. Murata and Z. H. Kafafi, Surf. Sci., 2000, 450, 142-152.
CrossRef View Record in Scopus
40. G. Boschloo and A. Hagfeldt, Acc. Chem. Res., 2009, 42, 1819–1826.
CrossRef View Record in Scopus
41. B. Oregan, S. Scully and A. Mayer, J. Phys. Chem. B, 2005, 109, 4616–4623.
CrossRef View Record in Scopus
42. S. G. Chen, S. Chappel, Y. Diamant and A. Zaban, Chem. Mater., 2001, 13, 4629-4634.
CrossRef View Record in Scopus
43. T. C. Li, M. S. Goes, F. Fabregat-Santiago, J. Bisquert, P. R. Bueno, C. Prasittichai, J. T. Hupp and T. J. Marks, J. Phys. Chem. C, 2009, 113, 18385−18390.
CrossRef View Record in Scopus
44. Q. Wang, J. E. Moser and M. Gratzel, J. Phys. Chem. B, 2005, 109, 14945-14953.
CrossRef View Record in Scopus
45. K. M. Lee, C. W. Hu, H. W. Chen and K. C. Ho, Sol. Energy Mater. Sol. Cells, 2008, 92, 1628-1633.