Preparation of nano-crystal N–Zn/TiO 2 anode films and the effects of co-sensitization on the performance of dye-sensitized solar cells photoelectric

A nano-crystal N–Zn/TiO 2 anode film was prepared using a combined technology. X-ray diffraction, transmission electron mi-croscopy, X-ray photoelectron spectroscopy and cyclic voltammetry characterizations showed that the two elements N and Zn were doped into nano-crystal TiO 2 successfully. This resulted in a strong redshift in the UV-Visible spectrum. UV-Visible measurements showed that the light absorption of N719 and P3OT were complementary and covered the entire visible region. This led to a high utilization of visible light. Solar cells based on the N – Zn/TiO 2 anode film were co-sensitized using P3OT and N719. The cells have a short-circuit current density of 7.91 mA/cm 2 , an open-circuit photovoltage of 0.659 V, and a photoelectric conversion efficiency of 2.64%. Also, t he relationship among the N – Zn/TiO 2 -film anode’s electric structure, the dye’s LUMO, electrochemical impedance, and photoelectric conversion efficiency are discussed in the paper. dye-sensitized solar cells, nano-crystal N–Zn/TiO 2 anode film co-sensitization, electrochemical impedance spectroscopy, energy level match,

The rapid depletion of fossil fuels and potentially disastrous environmental problems forcing the development of renewable energy resources for our growing energy demands. Fortunately, solar energy has a great potential to fulfill the demand for sustainable energy for future generations. Dye-sensitized solar cells (DSSCs) have attracted increasing attention for their high efficiency, low-cost, flexibility, and fine-tuned materials, since O'Regan and Grätzel first introduced DSSCs in 1991 [1][2][3].
The most extensively studied DSSCs are based on TiO 2 [4], which is a wide bandgap semiconductor with rutile and anatase phases, and only absorb UV light under 387 nm. Previously, there have been some reports of TiO 2 doped with metal elements or non-metallic elements such as Li, N, Cl, Zn and W [5][6][7][8][9]. However, a nano-crystal N-Zn/TiO 2 anode has not been reported in a DSSC. *Corresponding author (email: zhangjc1@mail.buct.edu.cn) Studies on dye sensitizers have been mainly focused on organic metal complexes or pure organic dyes [10][11][12][13]. However, there has been no report of dye-sensitized solar cells based on N719 and P3OT (poly(3-octylthiophene)). In this paper, a nano-crystal N-Zn/TiO 2 anode film was prepared by doping N and Zn into TiO 2 . Solar cells based on nano-crystal N-Zn/TiO 2 were co-sensitized using P3OT and N719 to improve the performance of the solar cells. The relationships among the microscopic structure of the battery, electrochemical impedance, and photoelectric conversion efficiency were also studied.
This solution was stirred vigorously at room temperature. We adjusted the pH of the solution was adjusted to 5-6 by dripping ammonia. The gel was filtered and washed until there were no remaining traces of Cl -. Then, the gel was transferred into a PTFE autoclave (100 mL) and kept at 240°C for 12 h. Then, the gel was washed using ethanol until almost all the water was removed. The obtained alcogel was transferred into a crucible and dried at room temperature. Finally calcination was performed at 500°C for 1 h. Through this process the nano-TiO 2 was obtained.

Preparation of nano-N/TiO 2
The gel (prepared in section 1.1) was filtered and washed with aqueous ammonia (3 mol/L) until all the chloride ions were removed. The N-doped TiO 2 was obtained following the steps in section 1.1.

Preparation of nano-N-Zn/TiO 2
PEG800 (0.5 mL) was added to 0.3 mol/L TiCl 4 (55 mL), and was vigorously stirred at room temperature. Aqueous Zn(NO 3 ) 2 (0.2 mol/L) was added to the system at the molar ratio, Ti : Zn = 100 : 0.2. The N-Zn-doped TiO 2 was obtained following the remaining steps in section 1.1.

Preparation of P3OT
The conducting P3OT was synthesized using Fe(III) as catalyst, following the method in references [14,15]. In an N 2 environment, chloroform (20 mL) and anhydrous FeCl 3 (0.03 mol) were added to a 100 mL flask, and the mixture was stirred for 15 min. 3-Octylthiophene (0.01 mol) in chloroform (20 mL) was added dropwise to the solution. The reaction temperature of system was kept between 0 -1°C for 7 h. The product was slowly added to methanol (300 mL), separated from the solvent, and filtered. The remaining black solid was added to chloroform (50 mL) in a conical flask and was stirred for 60 min while being gently heated to 40°C. A chloroform solution remained. After removing the chloroform by distillation, the small amount of viscous liquid residue was slowly added to methanol (200 mL) and a product precipitated. The precipitate was dried in vacuum at room temperature, and then it was subjected to Soxhlet extraction with methanol for 20 h. Finally the polymer was precipitated with methanol (200 mL) and vacuum dried at room temperature. Through this process the target polymer was obtained.

Preparation of dye-sensitized solar cells
(1) Preparation of TiO 2 anode, N/TiO 2 anode and N-Zn/ TiO 2 anode. Equal volumes of 4 mol/L TiCl 4 , deionized water and ethanol were mixed to obtain a TiCl 4 solution. In an agate grinding bowl, 0.25 g of TiO 2 powder, 1 mL of deionized water, 0.05 mL of acetylacetone and 0.15 mL of PEG800 were ground for 20 min to obtain a TiO 2 paste. Four sides of a fluorine doped tin oxide glass were covered with adhesive tape to obtain a 0.5 cm 2 (0.5 cm×1 cm) area. The glass was then coated with a layer of sol-TiCl 4 . We then added some TiO 2 paste to the surface and smoothed the surface using a scraping method. After being dried in air, the tape was removed. The TiO 2 anode was obtained after heating the dried paste at 450°C for half an hour. The N/TiO 2 anode and N-Zn/TiO 2 anode were prepared using the same method.
(2) Preparation of Pt electrode. H 2 PtCl 6 was placed evenly on another fluorine doped tin oxide substrate. It was then dried at 25°C and calcinated at 380°C for 30 min to prepare Pt electrode.
(3) Sensitization and assembly the DSSCs. We prepared three dyes: 5 mmol/L P3OT in a chloroform solution, 5 mmol/L N719 in an ethanol solution, and mixtures of 5 mmol/L P3OT in a chloroform solution and 5 mmol/L N719 in an ethanol solution with different molar ratios. The N-Zn/TiO 2 anode thin films prepared using the steps in section 1.5 (1) were soaked in different kinds of dyes at 80°C for 12 h. Dye attached to the surface layer was washed with anhydrous ethanol. The sample was dried at 80°C for 15 min. Through this method the sensitized N-Zn/TiO 2 anode thin film was prepared. After the dye-sensitized solar cells were obtained, the N-Zn/TiO 2 anode was packaged with a Pt electrode using an epoxy resin film. Then, we injected the electrolyte (0.5 mol/L LiI, 0.05 mol/L I 2 , 0.3 mol/L DMPII, 0.5 mol/L 4-TBP) into the system and sealed it.
Transmission electron microscope (TEM) measurements were performed on a HITACHI-800 (Tokyo, Japan) to test the power sample.

Cyclic Voltammetry
Cyclic voltammetry (CV) measurements were performed using a Potentiostat/Galvanostat (EG&G PAR Model 283, Princeton Applied Research, Oak Ridge, TN, USA). The counter electrode was made of platinum, and the reference electrode was non-aqueous Ag/Ag + . The CV curves were recorded in acetonitrile with 0.1 mol/L LiClO 4 at a scanning rate of 50 mV/s. Prior to use, all the solutions were insufflated with N 2 for 30 min.

Testing the DSSCs
Electrochemical Impedance Spectroscopy (EIS) measurements were performed using the Potentiostat/Galvanostat (EG&G PAR Model 283). The measured frequency ranged from 50 mHz to 100 kHz. R ct1 , R ct2 and R diff were corresponded to the charge transfer processes at the Pt counter electrode, the TiO 2 /dye/electrolyte interface and the Nernst diffusion process of I -/I 3 in the electrolyte, respectively. Current-voltage (I-V) measurements were taken in air at room temperature (25°C) using a Keithley 236 high-currentsource power meter (Cleveland, OH, USA) under whitelight illumination from a 500-W xenon lamp. The light intensity was about 100 mW/cm 2 on the sample surfaces, and was measured using a photodetector at 25°C. The tested battery parameters were the short-circuit current density (I sc ), the open-circuit photovoltage (V oc ), the maximum power (P max ), the incident light power P in (100 mW/cm 2 ), and the photoelectric conversion efficiency (η) [16]: These peaks correspond to the anatase phase. There are no N or ZnO diffraction peaks in the XRD spectra, which indicates that doping elements were dispersed in TiO 2 to a high degree. We calculated that the average TiO 2 , N/TiO 2 and N-Zn/TiO 2 particles are 22.22, 13.69 and 11.83 nm respectively using the Scherrer formula (D hkl = kλ /β cosθ ). This indicates the doping of N and Zn can inhibit the growth and increase the surface area of nano-TiO 2 powders. This is favorable for the adsorption of dye molecules in N/

TEM measurements of TiO 2 , N/TiO 2 and N-Zn/TiO 2
The TEM images of TiO 2 , N/TiO 2 and N-Zn/TiO 2 particles are shown in Figure 2. These images show that the sizes of the particle are 22 nm (a), 14 nm (b), and 12 nm (c). The results are basically the same as those predicted using the Scherer formula. The doped particles were found to have very good dispersivity. Alcohol exchange during the powder preparation eliminated the liquid surface tension and reduces agglomeration of particles during the calcining process. Therefore, nano-particles with high surface area, small size, and good dispersivity were obtained. Figure 3(a)-(e) is high-resolution XPS spectra of sample N-Zn/TiO 2 film. From Figure 3(a), it can be seen that the N and Zn two elements were doped successfully. It can be seen from Figure 3(b) that Ti in N-Zn / TiO 2 is primarily tetravalent [17]. Figure 3(b) shows a close-up of the scan of the Ti2p peaks located at 464.62 eV (Ti 2p 1/2 ) and 458.86 eV (Ti 2p 3/2 ) in the N-Zn / TiO 2 film. Compared with TiO 2 , a high binding energy-shift took place. This is because the radius of Zn 2+ (74 pm) is larger than that of Ti 4+ (68 pm), and Zn is difficult to dope into the internal lattice of TiO 2 . This leads to Ti atoms being replaced by Zn atoms to form Ti-O-Zn on the interface of the anatase crystal TiO 2 . In Figure 3 Figure 4 shows the UV-Vis spectra of TiO 2 , N/TiO 2 and N-Zn/TiO 2 anode thin film electrodes. In Figure 4, the UV-Vis absorption spectra show that the maximum absorption wavelengths of TiO 2 , N/TiO 2 and N-Zn/TiO 2 anode thin films are 249.5, 254.7 and 260.3 nm, respectively. The corresponding absorption wavelengths are 388, 438 and 456 nm. Compared with pure TiO 2 , the significant red-shift took place in N-Zn/TiO 2 films, which has strong visible light absorption [20]. Figure 5 shows the UV-Vis spectra of P3OT, N719 and P3OT/N719. The three dyes have strong absorption between 200 -700 nm. The main absorption peaks of N719 located at 212, 317, 400 and 551 nm, which nearly cover the entire UV-Vis region. The maximum absorption wavelengths of P3OT are 220 and 518 nm. From Figure 5, it can be seen that N719 has a valley at 490 nm. The absorption peak of P3OT at 500 nm supplements the absorption of N719. The combined absorption of N719 and P3OT can achieve a full absorption of visible light spectrum. Therefore, we used N719/P3OT dye to broaden the light absorption of anode in the visible range to improve the photoelectric conversion efficiency of the system.

Cyclic voltammetry
CV curves of TiO 2 , N/TiO 2 and N-Zn/TiO 2 nanocrystalline thin film electrodes and different dyes are shown in Figure  6(a) and (b). The frontier orbital and energy gap were listed in Table 1. According to E g = 1239.8/λ (E g is the band gap, λ is the maximum wavelength of absorption), the maximum absorption wavelengths of TiO 2 , N/TiO 2 and N-Zn/TiO 2 films were 387, 441 and 459 nm, respectively. The results are basically the same with the UV-Vis test.
It can be seen from Figure 7 that the energy gaps of N719 and N719/P3OT are 1.57 and 1.54 eV, respectively. These are smaller than that of P3OT and N719. The electrons in N719/P3OT are easier to excite with light. In addition, the lowest unoccupied molecular orbitals (LUMOs) of N719 and N719/P3OT are 3.11 and 3.12 eV, respectively. It is easier for an electron in the LUMO of N719 or N719/P3OT to transfer to the LUMO of the N/TiO 2 or N-Zn/TiO 2 film electrodes (Figure 7). However, the LUMO of the N-Zn/TiO 2 film electrode is lower than that of the N/TiO 2 film electrode. Therefore, N719/P3OT matches the LUMO of N-Zn/TiO 2 thin film electrode more closely (Figure 7).

I-V and EIS
(1) Effects of N-Zn/TiO 2 nanocrystalline thin film electrodes on the performance of dye-sensitized solar cell. Current density-voltage curves for the DSSCs based on the thin film electrodes are shown in Figure 8. The electrodes were sensitized using N719, and are under a constant illumination of 100 mW/cm 2 . The estimated performance parameters are summarized in Table 2. Photoelectric chemical characteristics, such as short circuit current density (J sc ) and photoelectric conversion efficiency (η), of the DSSCs based on the N-Zn/TiO 2 film electrode are greater than both of the DSSCs based on the N/TiO 2 and TiO 2 film electrodes. The J sc and η of the DSSC based on the N-Zn/TiO 2 film electrode are 6.22 mA/cm 2 and 2.28%, respectively.
From the EIS spectra (Figure 9), it can be seen that the R ct2 of the DSSC based on N-Zn/TiO 2 film electrode is the lowest. We concluded that N719 matches with N-Zn/TiO 2 film electrode the best (Figure 7). The η of the DSSCs based on N-Zn/TiO 2 film electrode is the highest. Also, the particle size of the N-Zn/TiO 2 film is the smallest, and it has with a high surface area. This increases the adsorption for both the dye and anode, which improves the photoelectric conversion efficiency.
(2) Effects of co-sensitization on the performance of the dye-sensitized solar cells. Current density-voltage curves for DSSCs sensitized using different dyes are plotted in Figure 10. These measurements were made at a constant illumination of 100 mW/cm 2 . The estimated performance parameters are summarized in Table 3. It can be seen that the performance of the DSSCs co-sensitized using P3OT and N719 is better than that of the DSSCs sensitized using N719 and P3OT. When n(P3OT): n(N719) = 1 : 2.5, the dye-sensitized solar cells have a short-circuit current density of 7.91 mA/cm 2 , an open-circuit photovoltage of 0.659 V, and a photoelectric conversion efficiency of 2.64%. However, photoelectric chemical characteristics, such as J sc , V oc and η, of the DSSCs decrease with increasing N719. This is because the DSSCs become increasingly similar to those sensitized using N719 alone. These findings match those of 3.5, which stated that the P3OT/N719 dye matches the N-Zn/TiO 2 film electrode best. From the EIS spectra (Figure 11), it can be seen that the R ct2 of the DSSCs co-sensitized using P3OT and N719 is the lowest when n(P3OT) : n(N719) = 1 : 2.5. In this case, the short-circuit current density and the photoelectric conversion efficiency are at their highest values.

Conclusion
In this paper, TiO 2 , N/TiO 2 and N-Zn/TiO 2 anode films were prepared using a combination of sol and powder technologies of. DSSCs based on the N-Zn/TiO 2 anode film were co-sensitized using N719 and P3OT.
(1) Anatase N-Zn/TiO 2 was prepared using the hydrothermal method. The XPS measurements showed that the N and Zn elements were doped successfully. From the XRD and TEM characterization, we found that the particles sizes were distributed between 10-13 nm. The doping of N and Zn inhibited the growth of nano-TiO 2 particles.
(2) The TiO 2 , N/TiO 2 and N-Zn/TiO 2 anode films were prepared using a combination of techniques. The UV-vis spectral measurements showed the light absorption of N719 and P3OT could supplement one another and cover the entire visible region. CV measurements showed that the P3OT/N719 matched the N-Zn/TiO 2 film electrode the best.   a) The Φ ox and Φ red potentials (vs. 0.01 mol/L Ag + /Ag in CH 3 CN) were determined using cyclic voltammetry. They were calculated using E g =Φ ox -Φ red . E HOMO and E LUMO from E HOMO =-(Φ ox +4.44) eV and E LUMO = -(Φ red +4.44) eV [21].