Efficient Dye-Sensitized Solar Cells Composed of Nanostructural ZnO Doped with Ti

Photoanode materials with optimized particle sizes, excellent surface area and dye loading capability are preferred in good-performance dye sensitized solar cells. Herein, we report on an efficient dye-sensitized mesoporous photoanode of Ti doped zinc oxide (Ti-ZnO) through a facile hydrothermal method. The crystallinity, morphology, surface area, optical and electrochemical properties of the Ti-ZnO were investigated using X-ray photoelectron spectroscopy, transmission electron microscopy and X-ray diffraction. It was observed that Ti-ZnO nanoparticles with a high surface area of 131.85 m2 g−1 and a controlled band gap, exhibited considerably increased light harvesting efficiency, dye loading capability, and achieved comparable solar cell performance at a typical nanocrystalline ZnO photoanode.


Introduction
Environmental challenges and increasing global energy consumption demand clean and renewable energy to support the activities of modern society [1].To convert solar energy directly into electricity using solar cells is one of the most feasible technologies.As an alternative to the conventional silicon-based solar cell, dye-sensitized solar cells (DSSCs) have attracted tremendous interest due to their simple fabrication process, low manufacturing cost, and environmentally-friendly advantages [2][3][4][5][6].Thus far, an extensive study has been carried out on the TiO 2 nanoparticle as a photoelectrode since 1991 [7,8], although the TiO 2 nanoparticle showed more than 14% power conversion efficiency when used as a photoanode due to its higher surface area.However, its poor long-term stability and low electron mobility are concerns [9,10].Excluding TiO 2 , studies on other alternative semiconductor metal oxides, such as SnO 2 , ZnO, Nb 2 O 5 and Zn 2 SnO 4 , have been conducted [11][12][13][14][15][16][17][18][19].Among these, ZnO has received increasing interest as a promising alternative photoanode material for DSSCs due to similar electron affinities and band gap energy (E g ), i.e., 3.2 eV and 3.3 eV compared with TiO 2 , and has excellent electron diffusivity [20,21].At the same time, ZnO is probably richer in diverse morphologies compared to any known materials [22][23][24].More recently, several studies have focused on the diverse morphology and multidimensionality of TiO 2 and ZnO with regards to robust photocatalytic activity of DSSCs with increasing electron transport and higher surface-to-volume ratios [25][26][27].Additionally, ZnO has amazing merits like stability against photo corrosion, low cost, large excitation binding energy (60 eV), easier synthesis, and surface modification [28].However, the photovoltaic performance of ZnO-based dye-sensitized solar cells is still poor, which could be caused by the narrow absorption spectral range, the quick recombination of photo-generated electron-hole pair occurring in or on the surface photoanode and the instability of film in acidic dye solution in the case of ZnO materials [29].In principle, ZnO-based electrodes could achieve a similar or even better photovoltaic performance if the aforementioned drawbacks could be overcome.Up to now, a number of possible strategies have been adopted to tackle these issues of ZnO electrodes.The morphological surface modification with nanoparticles of noble metals (Au, Pt, Pd, Ag, etc.) has been widely employed to decorate ZnO by doping to suppress carrier recombination [30,31].The light sensing properties can be improved by passivation of the defect in the bulk material and at the oxide interface with dangling bond fixing [32].Considering the economic and limited sources of noble metals, in this prospective study, Ti-doped ZnO and undoped ZnO nanoparticles were fabricated using a hydrothermal route with the aim to achieve an efficient charge separation, interfacial charge transfer to adsorbed dye to increase lifetime of charge carrier, and enhancing the photovoltaic performance of ZnO-based DSSCs.

Synthesis of Materials
Titanium-doped zinc oxide nanoparticles were synthesized using a facile hydrothermal route.In a typical procedure, 1.25 mmol zinc acetate dihydrate was added to methanol and stirred for 20 min, then 2 mmol of potassium hydroxide (KOH) was dropped to the solution, and a calculated amount (Ti = 1, 2, 5 and 7.5 wt%) of titanium isopropoxide (TTIP) was introduced under continuous stirring.The reaction mixture was further stirred and refluxed at 60-80 • C. The transparent color of the zinc acetate solution changed to transparent white, which indicated the starting of Ti-ZnO synthesis.After several minutes of stirring, the reaction mixture was transferred into a Teflon lined autoclave and kept in oven at 180 • C for 12-15 h.After cooling to room temperature, the yellowish solution was separated by centrifugation (8000 rpm) for 10 min and washed several times with absolute ethanol.Eventually, the powder was dried in vacuum at 60 • C overnight and calcined at 500 • C for 2 h in Ar.As a reference sample, the undoped zinc oxide nanoparticles were also synthesized via the same procedure except for the addition of TTIP.

Paste Preparation
Different pastes of ZnO and titanium-doped zinc oxide (Ti-ZnO) were prepared according to previously reported literature [33,34].Typically, ethyl cellulose, α-terpineol (10, 5% by weight) and zinc oxide or Ti-ZnO nanoparticles were suitably dispersed in ethanol using an Ultra-Turrax T25 homogenizer followed by a rotary evaporator to concentrate up to 20 wt%.

Films Fabrication
Photoanodes were prepared through a screen printing method with as prepared zinc oxide and Ti-ZnO paste applied on an FTO substrate (sheet resistance 8 Ω sq −1 , Pilkington, and followed by sintering at 450 • C for 30 min.

Dye Loading
The dye-loading capability of the photoanodes was estimated photometrically.The ZnO and Ti doped ZnO with various percentages of photoanodes loaded with N719 dyes were immersed in 10 mM NaOH at 30 • C overnight.The most intense band intensity was the region of 550 to 515 nm of the adsorption spectrum and was used to calibrate the dye loading.

Device Fabrication
The Ti-ZnO and ZnO electrodes fabricated as described above were immersed in 0.5 mM solution of N719 dye in isopropyl alcohol and acetonitrile (1:1 v/v) for 2 h at 30 • C and then rinsed with acetonitrile and dried with a nitrogen gas flow.The electrolyte was composed of 0.6 M 1,2-dimethyl-3-n-propylimidazolium iodide (DMPII), 0.05 M I 2 , 0.1 M LiI, in 3-methoxypropinonitrile with 0.5 M 4-tertbutylpyridine (TBP).A counter electrode was fabricated via dropping of 5 mM H 2 PtCl 6 on a FTO substrate and sintered at 400 • C for 30 min.Afterwards, a polyethylene spacer was sandwiched between a dye-sensitized photoanode and a Pt counter electrode to prevent the cell from short circuiting.For the last step of DSSC fabrication, the redox electrolyte was injected into the gap of the spacer.ZnO cells were fabricated for comparison of Ti-ZnO based device photovoltaic performance to a known system.

Characterization of the Materials and Measurement of the Photovoltaic Performance
The phase purity and structural properties of the materials were analyzed by powder X-ray diffraction (XRD) on a Rigaku Ultima IV X-ray diffractometer using Cu Kα radiation (λ = 1.5418Å).A scanning electron microscope (SEM, Hitachi S4800) was applied to determine the size and morphology, and a transmission electron microscopy (TEM) was performed on a Tecnai G2 F20 (FEI, Hillsboro, OR, USA) transmission electron microscope to further characterize the morphology and intrinsic structure of the materials.The N2 adsorption-desorption isotherms of Ti-ZnO were performed on an ASAP 2020 surface area analyzer (Micromeritics, Norcross, GA, USA).The film thickness was measured on a profilometer (Accretech Surfcom 130A).A Shimadzu UV-2450 UV-vis spectrophotometer and a Renishaw inVia Raman microscope were used for spectroscopic measurement.The Keithley 2400 SourceMeter under simulated AM 1.5 G illumination (Peccell, with a 1000 W Xe lamp and AM 1.5 filter) was used to investigate current-voltage (I-V) characteristic.Incident photon-to-current conversion efficiency (IPEC) spectra were recorded on a PEC S-20 (Peccell, Technology Co., Ltd., Yokohama, Japan) multimeter.Electrochemical impedance spectroscopy (EIS) was conducted on the electrochemical workstation IM6 (Zahner, and a bias potential of −0.7 V in dark condition with an 80 MHz-3 kHz range.

Results and Discussion
The crystallographic phases of ZnO and Ti-ZnO were analyzed from X-ray diffraction patterns, as shown in Figure 1A.All of the diffraction peaks match the standard phase (JCPDS # 01-079-0207), indicating that the synthesized materials are well crystallized and phase pure hexagonal ZnO.However, slight shifting of the peaks to a higher angle was observed for titanium-containing materials, Ti(7.5%)-ZnO compared with the ZnO, which implied the inward relaxation of the lattice due to the substitution of the bigger Zn 2+ atom (ionic radius = 0.74 Å) with smaller ion (Ti 4+ ionic radius = 0.68 Å) which had larger size as ions in the crystalline structure, and caused a change in the lattice parameter value without changing the crystal structure of the ZnO lattice.The average value of crystallite size and strain estimated by the Debye-Scherrer equation, D = Kλ/βCosθ [35], and the XRD patterns of all samples are shown in Figure S2 and their physiochemical characteristics, including surface area, strain, crystallite and pore size, are summarized in Table S1.
Two small negligible peaks were found at 25 and 27 degrees with higher dopant concentrations, which may mean that trace amount of Ti was converted into TiO 2 which was not detected by any other characterization technique.The chemical state of the atoms on the surface of ZnO and the elemental composition were examined via XPS measurements.Figure 1B represents the XPS survey spectra of ZnO and Ti-ZnO.The well defined peaks centered at 1044, 1021, 531, 475, 459, 285, 140, 98 and 10 eV are ascribed to the binding energy of Zn2p 1/2 , Zn2p 3/2 , O1s, ZnLMM1, Ti2p, C1s, Zn3s, Zn3d and Zn3p respectively.The high-resolution spectra of Zn2p and Ti2P with different dopant consideration are shown in Figure 1C,D.The morphological changes and particle size of the synthesized ZnO with controllable titanium doping concentration were observed by SEM measurements.The SEM images shown in Figure 2A,B prove that the prepared ZnO and Ti-ZnO samples were composed of nanoparticles with particle sizes of ca.9-13 nm, while the incorporation of the doping particle did not induce any significant change in their morphology but decreased the particle size.The typical crystalline behavior, morphology and Ti ion distribution in the ZnO crystal lattice are further confirmed from TEM images depicted in Figure 2C,D.The corresponding TEM images reveal that the obtained materials were well crystallized, and the selected area electron diffraction (SAED) pattern inset in Figure 2C indicates ((101), ( 102) and ( 110)) a well-crystalized plane of ZnO.In addition, the lattice distance 0.24 nm shown in Figure 2D matches with the (101) plane ZnO crystal (slightly decrease of pure ZnO: 0.25 nm), indicating that Ti ions were evenly distributed in ZnO crystal lattice.
Catalysts 2019, 9, x FOR PEER REVIEW 4 of 11 samples were composed of nanoparticles with particle sizes of ca.9-13 nm, while the incorporation of the doping particle did not induce any significant change in their morphology but decreased the particle size.The typical crystalline behavior, morphology and Ti ion distribution in the ZnO crystal lattice are further confirmed from TEM images depicted in Figure 2C, D. The corresponding TEM images reveal that the obtained materials were well crystallized, and the selected area electron diffraction (SAED) pattern inset in Figure 2C indicates ((101), ( 102) and ( 110)) a well-crystalized plane of ZnO.In addition, the lattice distance 0.24 nm shown in Figure 2D matches with the (101) plane ZnO crystal (slightly decrease of pure ZnO: 0.25 nm), indicating that Ti ions were evenly distributed in ZnO crystal lattice.samples were composed of nanoparticles with particle sizes of ca.9-13 nm, while the incorporation of the doping particle did not induce any significant change in their morphology but decreased the particle size.The typical crystalline behavior, morphology and Ti ion distribution in the ZnO crystal lattice are further confirmed from TEM images depicted in Figure 2C, D. The corresponding TEM images reveal that the obtained materials were well crystallized, and the selected area electron diffraction (SAED) pattern inset in Figure 2C indicates ((101), ( 102) and ( 110)) a well-crystalized plane of ZnO.In addition, the lattice distance 0.24 nm shown in Figure 2D matches with the (101) plane ZnO crystal (slightly decrease of pure ZnO: 0.25 nm), indicating that Ti ions were evenly distributed in ZnO crystal lattice.The surface area and pore size distribution of ZnO and Ti-ZnO nanoparticles were performed with Brunauer-Emmett-Teller (BET) by using N2 adsorption-desorption isotherm.The XRD and BET results were in good alignment.From the results in Figure 3 and Table S1, it confirms that the surface area of (Ti-ZnO) under the same synthetic conditions dramatically increased from 45.38 to 131.85 m 2 g −1 and the average pore size distribution of Ti-ZnO (3.4 nm) was slightly greater compared to ZnO (2.8 nm).The increment in BET surface area and pore size may be due to the inhibition of the ZnO crystallites growth with the incorporation of Ti dopant.The smaller size of Ti-ZnO would possess of higher BET surface area and larger pore size according to the small size effect of nanomaterials.The higher BET surface area and larger pore size would contribute to the higher dye loading, leading to enhanced light harvesting efficiency.
Catalysts 2019, 9, x FOR PEER REVIEW 5 of 11 The surface area and pore size distribution of ZnO and Ti-ZnO nanoparticles were performed with Brunauer-Emmett-Teller (BET) by using N2 adsorption-desorption isotherm.The XRD and BET results were in good alignment.From the results in Figure 3 and Table S1, it confirms that the surface area of (Ti-ZnO) under the same synthetic conditions dramatically increased from 45.38 to 131.85 m 2 g −1 and the average pore size distribution of Ti-ZnO (3.4 nm) was slightly greater compared to ZnO (2.8 nm).The increment in BET surface area and pore size may be due to the inhibition of the ZnO crystallites growth with the incorporation of Ti dopant.The smaller size of Ti-ZnO would possess of higher BET surface area and larger pore size according to the small size effect of nanomaterials.The higher BET surface area and larger pore size would contribute to the higher dye loading, leading to enhanced light harvesting efficiency.The doped materials were further analyzed for elemental percentage of titanium in Ti-ZnO samples via ICP-OES, the data are summarized in Table 1.The ICP-OES analytical and EDS data are in best alignment with calculated values as observed from the ICP-OES analytical results and EDS spectra depicted in Figure S2.UV-visible characterization is a significant technique to assess the quantum confinement effects and band gap.In order to analyze the optical changes (effect of doping on the band gap), the UV absorption spectra were measured at room temperature in the range of 200 to 800 nm as displayed in Figure 4, the broad absorption band at 399 nm in ZnO was shifted to a higher wavelength of 426 nm with incorporation of Ti, which was assigned to the ligand-metal-charge transfer in the octahedral environment.After the successive incorporation of Ti, the resultant products showed more visible light absorbance (visible light utilizing efficiency) compared to pure ZnO, which might be attributable to the fact that the light scattering effect of ZnO was enhanced by Ti doping.Meanwhile, the decrease in bandgap, was well corresponded to the visible spectrum.The bandgap energy (Eg) was calculated from UV-visible spectrum using Equation (1).Eg = 1240/λ (1) The doped materials were further analyzed for elemental percentage of titanium in Ti-ZnO samples via ICP-OES, the data are summarized in Table 1.The ICP-OES analytical and EDS data are in best alignment with calculated values as observed from the ICP-OES analytical results and EDS spectra depicted in Figure S2.UV-visible characterization is a significant technique to assess the quantum confinement effects and band gap.In order to analyze the optical changes (effect of doping on the band gap), the UV absorption spectra were measured at room temperature in the range of 200 to 800 nm as displayed in Figure 4, the broad absorption band at 399 nm in ZnO was shifted to a higher wavelength of 426 nm with incorporation of Ti, which was assigned to the ligand-metal-charge transfer in the octahedral environment.After the successive incorporation of Ti, the resultant products showed more visible light absorbance (visible light utilizing efficiency) compared to pure ZnO, which might be attributable to the fact that the light scattering effect of ZnO was enhanced by Ti doping.Meanwhile, the decrease in bandgap, was well corresponded to the visible spectrum.The bandgap energy (E g ) was calculated from UV-visible spectrum using Equation (1).
where λ is the wavelength measured from UV-visible absorption spectra Figure 4.The Ti-ZnO possesses a narrow forbidden Eg of 2.91 ev, is well corresponding to higher visible light harvesting efficiency, which is conductive to produce more photo-generated electrons and enhancement of short current density, as confirmed from IPCE measurement.The valence (VB) and conduction bands (CB) were estimated from oxidation potential (E ox ) by cyclic voltammetry (Figure S4) through equations in literature [36].The corresponding results were manifested in Table S3.Based on optical and electrochemical properties comparison, the band positions of dye N719 and Ti-ZnO are in good alignment which is very important for charge separation.In the DSSC, the dye-sensitized dye adsorbs the photon under solar illumination at its interface, then separation of charge take place at the interface.Due to the well-aligned band positions for the charge separation, the photo-generated electrons can be quickly transferred from excited dye to the conduction band of Ti-ZnO, and the external circuit resulting in the photocurrent.The LUMO level of the N719 dye is higher than that of Ti-ZnO and ZnO, so for the electron transport has a single directional pathway.where λ is the wavelength measured from UV-visible absorption spectra Figure 4.The Ti-ZnO possesses a narrow forbidden Eg of 2.91 ev, is well corresponding to higher visible light harvesting efficiency, which is conductive to produce more photo-generated electrons and enhancement of short current density, as confirmed from IPCE measurement.The valence (VB) and conduction bands (CB) were estimated from oxidation potential (Eox) by cyclic voltammetry (Figure S4) through equations in literature [36].The corresponding results were manifested in Table S3.Based on optical and electrochemical properties comparison, the band positions of dye N719 and Ti-ZnO are in good alignment which is very important for charge separation.In the DSSC, the dye-sensitized dye adsorbs the photon under solar illumination at its interface, then separation of charge take place at the interface.Due to the well-aligned band positions for the charge separation, the photo-generated electrons can be quickly transferred from excited dye to the conduction band of Ti-ZnO, and the external circuit resulting in the photocurrent.The LUMO level of the N719 dye is higher than that of Ti-ZnO and ZnO, so for the electron transport has a single directional pathway.Raman spectroscopy is a sensitive technique for the analysis of subtle phase information, structural defects and crystallinity.ZnO has a hexagonal wurtzite structure with C6vP63mc space group, the group theory predicts that the Raman spectra of ZnO has two A1, two E2, two E1 and two B1 modes.To further study the crystallinity and effect of Ti in the framework structure of ZnO, both the Raman spectra of pure ZnO and Ti-ZnO were performed as shown in Figure 5. From the comparison of Raman spectra, it is evident that the well resolved peaks (TO, E2 and LO phonon modes) small frequency shifting, broadening and intensity decrease in the Ti-ZnO sample might be ascribed to the disorder induced defects or oxygen vacancies and crystal lattice distortion [37].Raman spectroscopy is a sensitive technique for the analysis of subtle phase information, structural defects and crystallinity.ZnO has a hexagonal wurtzite structure with C6vP63mc space group, the group theory predicts that the Raman spectra of ZnO has two A1, two E2, two E1 and two B1 modes.To further study the crystallinity and effect of Ti in the framework structure of ZnO, both the Raman spectra of pure ZnO and Ti-ZnO were performed as shown in Figure 5. From the comparison of Raman spectra, it is evident that the well resolved peaks (TO, E2 and LO phonon modes) small frequency shifting, broadening and intensity decrease in the Ti-ZnO sample might be ascribed to the disorder induced defects or oxygen vacancies and crystal lattice distortion [37].where λ is the wavelength measured from UV-visible absorption spectra Figure 4.The Ti-ZnO possesses a narrow forbidden Eg of 2.91 ev, is well corresponding to higher visible light harvesting efficiency, which is conductive to produce more photo-generated electrons and enhancement of short current density, as confirmed from IPCE measurement.The valence (VB) and conduction bands (CB) were estimated from oxidation potential (Eox) by cyclic voltammetry (Figure S4) through equations in literature [36].The corresponding results were manifested in Table S3.Based on optical and electrochemical properties comparison, the band positions of dye N719 and Ti-ZnO are in good alignment which is very important for charge separation.In the DSSC, the dye-sensitized dye adsorbs the photon under solar illumination at its interface, then separation of charge take place at the interface.Due to the well-aligned band positions for the charge separation, the photo-generated electrons can be quickly transferred from excited dye to the conduction band of Ti-ZnO, and the external circuit resulting in the photocurrent.The LUMO level of the N719 dye is higher than that of Ti-ZnO and ZnO, so for the electron transport has a single directional pathway.Raman spectroscopy is a sensitive technique for the analysis of subtle phase information, structural defects and crystallinity.ZnO has a hexagonal wurtzite structure with C6vP63mc space group, the group theory predicts that the Raman spectra of ZnO has two A1, two E2, two E1 and two B1 modes.To further study the crystallinity and effect of Ti in the framework structure of ZnO, both the Raman spectra of pure ZnO and Ti-ZnO were performed as shown in Figure 5. From the comparison of Raman spectra, it is evident that the well resolved peaks (TO, E2 and LO phonon modes) small frequency shifting, broadening and intensity decrease in the Ti-ZnO sample might be ascribed to the disorder induced defects or oxygen vacancies and crystal lattice distortion [37].To determine the photovoltaic performance of ZnO and Ti-ZnO nanoparticles based DSSCs, three cells were fabricated for comparison.We have checked all the corresponding parameters which affect the power conversion efficiency of the devices like film thickness, dye immersion time, synthesis temperature and morphology of the materials etc. Finally three types of cells that showed high power conversion efficiencies, due to the quality of devices and photoanode films were investigated in this study.Cells were prepared with as synthesized ZnO, Ti(5%)-ZnO, and Ti(7.5%)-ZnO, respectively.The current-voltage curves are plotted in Figure 5 and the corresponding characteristic are summarized in Table 2.It can be observed that the cell composed of pure ZnO nanoparticles possessed a short-circuit current Jsc of 8.01 mA cm −2 and an efficiency of 3.82%.In comparison, higher Jsc and open-circuit voltage Voc along with a PCE of 5.56% were obtained for the cell composed of Ti(7.5%)-ZnO.The increase in Jsc and Voc might be attributed to improved light harvesting efficiency of the photoanode, compared with the highest PCE value of the cell composed of ZnO, the PCE was increased by 44.7%.As shown in Figure 6 both results (Jsc and IPCE) of the DSSCs based on ZnO, Ti(5%)-ZnO and Ti(7.5%)-ZnO are in good alignment with the results displayed in Table 2.
Catalysts 2019, 9, x FOR PEER REVIEW 7 of 11 To determine the photovoltaic performance of ZnO and Ti-ZnO nanoparticles based DSSCs, three cells were fabricated for comparison.We have checked all the corresponding parameters which affect the power conversion efficiency of the devices like film thickness, dye immersion time, synthesis temperature and morphology of the materials etc. Finally three types of cells that showed high power conversion efficiencies, due to the quality of devices and photoanode films were investigated in this study.Cells were prepared with as synthesized ZnO, Ti(5%)-ZnO, and Ti(7.5%)-ZnO, respectively.The current-voltage curves are plotted in Figure 5 and the corresponding characteristic are summarized in Table 2.It can be observed that the cell composed of pure ZnO nanoparticles possessed a short-circuit current Jsc of 8.01 mA cm −2 and an efficiency of 3.82%.In comparison, higher Jsc and open-circuit voltage Voc along with a PCE of 5.56% were obtained for the cell composed of Ti(7.5%)-ZnO.The increase in Jsc and Voc might be attributed to improved light harvesting efficiency of the photoanode, compared with the highest PCE value of the cell composed of ZnO, the PCE was increased by 44.7%.As shown in Figure 6 both results (Jsc and IPCE) of the DSSCs based on ZnO, Ti(5%)-ZnO and Ti(7.5%)-ZnO are in good alignment with the results displayed in Table 2.All cells show the maximum incident-to-current conversion efficiency (IPCE) percentage at 550 nm wavelength corresponding to N719 dye adsorption peak as depicted in Figure 7. Furthermore, the adsorption amount of sensitizer dye directly influenced the quantum efficiency of the DSSCs.The IPCE peaks at 550 nm for Ti(7.5%)-ZnO is 73.76%, which is higher than 64.37% for Ti(5%)-ZnO and 57.23% for ZnO.In comparison the whole spectral region the Ti(7.5%)-ZnOshowed considerable maximum IPCE value, which might be owing to the fact that Ti(7.5%)-ZnO, with a specific doping concentration of Ti, and narrow forbidden gap of Eg, increased the surface area and dye loading which directly influenced visible light absorbance and utilizing efficiency.All cells show the maximum incident-to-current conversion efficiency (IPCE) percentage at 550 nm wavelength corresponding to N719 dye adsorption peak as depicted in Figure 7. Furthermore, the adsorption amount of sensitizer dye directly influenced the quantum efficiency of the DSSCs.The IPCE peaks at 550 nm for Ti(7.5%)-ZnO is 73.76%, which is higher than 64.37% for Ti(5%)-ZnO and 57.23% for ZnO.In comparison the whole spectral region the Ti(7.5%)-ZnOshowed considerable maximum IPCE value, which might be owing to the fact that Ti(7.5%)-ZnO, with a specific doping concentration of Ti, and narrow forbidden gap of Eg, increased the surface area and dye loading which directly influenced visible light absorbance and utilizing efficiency.
To support the photovoltaic performance enhancement and investigate the charge-transfer mechanism, the cells were analyzed via electrochemical impedance spectroscopy (EIS) under dark condition with −0.7 V applied voltage, their Nyquist plots are shown in Figure 8 and corresponding fitted results are summarized in Table 3.Two well-resolved semi-circles were observed in the measured frequency range.The Rs is ascribed to the photoanode film and FTO contact resistance.The small semicircle (R ct1 ) that lies at high frequency range is described to transport resistance at the interface of the counter electrode, while the large semicircle (R ct2 ) at the low frequency range is charge recombination resistance at the interface of photoanode/electrolyte/dye.As observed from the EIS spectra of ZnO and Ti-ZnO cells, the R ct2 value increased with doping concentration indicating that the charge recombination resistance at the interface of ZnO/dye/electrolyte decreased remarkably.All cells show the maximum incident-to-current conversion efficiency (IPCE) percentage at 550 nm wavelength corresponding to N719 dye adsorption peak as depicted in Figure 7. Furthermore, the adsorption amount of sensitizer dye directly influenced the quantum efficiency of the DSSCs.The IPCE peaks at 550 nm for Ti(7.5%)-ZnO is 73.76%, which is higher than 64.37% for Ti(5%)-ZnO and 57.23% for ZnO.In comparison the whole spectral region the Ti(7.5%)-ZnOshowed considerable maximum IPCE value, which might be owing to the fact that Ti(7.5%)-ZnO, with a specific doping concentration of Ti, and narrow forbidden gap of Eg, increased the surface area and dye loading which directly influenced visible light absorbance and utilizing efficiency.To support the photovoltaic performance enhancement and investigate the charge-transfer mechanism, the cells were analyzed via electrochemical impedance spectroscopy (EIS) under dark condition with −0.7 V applied voltage, their Nyquist plots are shown in Figure 8 and corresponding fitted results are summarized in Table 3.Two well-resolved semi-circles were observed in the measured frequency range.The Rs is ascribed to the photoanode film and FTO contact resistance.The small semicircle (Rct1) that lies at high frequency range is described to transport resistance at the interface of the counter electrode, while the large semicircle (Rct2) at the low frequency range is charge recombination resistance at the interface of photoanode/electrolyte/dye.As observed from the EIS spectra of ZnO and Ti-ZnO cells, the Rct2 value increased with doping concentration indicating that the charge recombination resistance at the interface of ZnO/dye/electrolyte decreased remarkably.Dye instability and carrier recombination in ZnO based DSSCs can occur across both the filmdye interface and film-electrolyte interface.Herein, we have discussed modifying the film/dye/electrolyte interface by way of Ti doping that is conformally doped into the oxide crystal lattice.It is well understood that the Ti doping can improve the ZnO based dye-sensitized solar cell efficiency by inhibiting the surface recombination of electrons at the film/dye/electrolyte interface by creating an energy barrier for physical separation of electrons.The dye and electrolyte stability of film significantly increases after Ti incorporation to the oxide crystal structure as shown in Figure S1B-D, fresh and stretched cells while the Figure S1A shows the cross section of the Ti-ZnO film.A  Dye instability and carrier recombination in ZnO based DSSCs can occur across both the film-dye interface and film-electrolyte interface.Herein, we have discussed modifying the film/dye/electrolyte interface by way of Ti doping that is conformally doped into the oxide crystal lattice.It is well understood that the Ti doping can improve the ZnO based dye-sensitized solar cell efficiency by inhibiting the surface recombination of electrons at the film/dye/electrolyte interface by creating an energy barrier for physical separation of electrons.The dye and electrolyte stability of film significantly increases after Ti incorporation to the oxide crystal structure as shown in Figure S1B-D, fresh and stretched cells while the Figure S1A shows the cross section of the Ti-ZnO film.A photoanode with an increase in material surface area enhanced dye loading capability, which could be ascribed

Conclusions
In summary, Ti doped ZnO nanoparticles with mesoporous structure were synthesized via a simple synthetic route, and then successfully applied in dye-sensitized solar cells as a photoanode.It was found that ZnO nanoparticles decorated with Ti exhibited a superior light harvesting performance and the best efficiency of 5.56% could be obtained.Our experiments prove that the 44.7% photovoltaic performance enhancement can be possibly ascribed to the improved Fermi level and large surface area, which enhance dye loading ability, reduce charge recombination resistance and increase the harvest of photoexcited electrons.In conclusion the present work will provide a new route for the synthesis of efficient and unique working electrode materials.

Table 3 .
EIS parameters of DSSCs based on ZnO and Ti-ZnO, with different dopant.

Table 1 .
Elemental analysis of Ti percentage in Ti-ZnO samples.

Table 1 .
Elemental analysis of Ti percentage in Ti-ZnO samples.

Table 2 .
performance for the cells composed of ZnO and Ti-ZnO photoanodes

Table 2 .
Photovoltaic performance for the cells composed of ZnO and Ti-ZnO photoanodes

Table 3 .
EIS parameters of DSSCs based on ZnO and Ti-ZnO, with different dopant.