Influence of Anti-reflecting Nature of MgF₂ Embedded Electrospun TiO₂ Nanofibers Based Photoanode to Improve the Photoconversion Efficiency of DSSC

Pratheep Panneerselvam,¹ Vignesh Murugadoss,^1,2 Vijayakumar Elayappan,^1,3  Na Lu,^4* Zhanhu Guo^2* and Subramania Angaiah^1­­­*

¹Electro-Materials Research Laboratory, Centre for Nanoscience and Technology,Pondicherry University, Puducherry-605 014, India.

²Integrated Composites Laboratory, Department of Chemical and Biomolecular Engineering, University of Tennessee, Knoxville, Tennessee 37996, USA

³Department of Chemical Engineering, Yeungnam University, Gyeongsan, Gyeongbuk-38541, Republic of Korea

⁴School of Materials Engineering, Purdue University, Indiana 47907-2045, USA

*E-mail: luna@purdue.edu (N.L); zguo10@utk.edu (Z.G); a.subramania@gmail.com (S.A)

 

Abstract

TiO₂ nanofibers (NFs) embedded with different weight percentages of magnesium fluoride (MgF₂) are fabricated by electrospinning technique to use as an anti-reflecting photoanode to enhance the power conversion efficiency (PCE) of DSSC. The thermal behaviour of TiO₂/PVP and TiO₂/PVP/MgF₂ precursors are studied by TG/DTA analysis. The MgF₂ embedded electrospun TiO₂ NFs has mixed phases of anatase and rutile and their weight fractions are investigated by X-ray diffraction analysis. The MgF₂ embedded electrospun TiO₂ NFs exhibited high dye loading compared to bare TiO₂ NFs. The surface morphology of MgF₂ embedded electrospun TiO₂ NFs is investigated by FE-SEM and TEM analysis. Among the different wt.% of MgF₂ embedded electrospun TiO₂ NFs, 10 wt% MgF₂ embedded TiO₂ NFs based DSSC enhanced the light harvesting effectively and thereby increase the power conversion efficiency (PCE) to 5.56%, which is an increase of 17% compared to bare electrospun TiO₂ NFs based DSSC (4.76%).

---

Table of Content

A new approach to integrate the advantage of anti-reflective MgF₂ into the one dimensional randomly aligned TiO₂ nanofibers (NFs) based photoanode for high performance Dye sensitized solar cell.

 

Keywords

Dye-sensitized solar cell     TiO₂ nanofibers      Magnesium fluoride      Anti-reflecting photoanode

---

1. Introduction

Dye-sensitized solar cell (DSSC) is a low-cost solar cell and an alternative to conventional solar cell demonstrated by Michael Gratzel in the year 1991. This invention has attracted much because it does not require ultra-high pure materials, unlike silicon solar cell.1 Many researchers are taking the effort to improve the overall PCE by modifying the photoanode, counter electrode, dye, and electrolyte.2-5 Among them, the light harvesting nature of solar cell mainly depends on the photoanode.6,7 The photoanode fabricated from TiO2 nanoparticles, the electrons transport are limited by the resistance of electrons in traps and the structural disorder at the contact among the nanoparticles leads to a reduction in the collection of the injected electrons.8 On the contrary, one-dimensional (1D) nanostructures consist of an almost defect-free structure which can provide a direct and quick pathway for the electrons with a limited interfacial recombination9 and also provide excellent mobility of charge carriers that enhance the charge collection and transport.10-15 Use of 1D TiO2 nanomaterial instead of TiO nanoparticles decreases the ohmic loss during the transport of injected electrons.16,17 The electrospinning technique is a simple, versatile and cost-effective to produce 1D nanostructured materials.18 There are many strategies to enhance the cell efficiency where one of the effective methods to enhance the light harvesting capability is via improved light transport by introducing anti-reflective material into the photoanode so that the dye has an opportunity to absorb more photons and consequently generate more number of free carriers for DSSC. In the recent scenario, ZnO and SiO are widely used as anti-reflecting materials 2 for DSSC.17,19 To the best of our knowledge, so far no one has reported that MgF as an anti-reflecting material for DSSC 2 applications. The MgF has a low refractive index value of 1.35 with 2 high positive isoelectric point (IEP) which may increase the light harvesting capability and dye adsorption nature of the system.

In the present investigation, a new anti-reflective nanostructured photoanode is developed by embedding anti-reflecting material (MgF ) in the randomly aligned TiO nanofibers (NFs) by 2 2 electrospinning process and its influence on the adsorption of dye, anti-reflecting effect on its PCE are studied in details.

2. Experimental procedure

Pure and MgF embedded electrospun TiO NFs were fabricated by 2 2 electrospinning method with the following procedure; the precursor solutions composed of 2.7 g of titanium (IV) isopropoxide, 2.5 mL of acetic acid, 2.5 mL of ethanol and 15 mL of 5 wt.% polyvinylpyrrolidone (PVP) along with two different wt.% of MgF (5 and 10 wt.%) added 2 with continuous stirring for 12 h to get viscous solutions.20 They were then loaded separately into a syringe connected with a 27 gauge stainless steel needle. The applied potential and the collector distance from the tip of the needle were fixed as 24 kV and 12 cm, respectively. The flow rate of the solution was 0.5 mL h-1 to get the MgF embedded electrospun TiO 2 2 mats. They were removed from the collector and kept in a vacuum oven at 80 ºC for 6 h. They were then calcined at 500 ºC for 4 h to get both pure and MgF embedded electrospun TiO NFs.

2.1 Materials

MgF embedded electrospun TiO NFs were fabricated from 2 2 titanium (IV) isopropoxide (Sigma-Aldrich), polyvinylpyrrolidone (PVP, Aldrich), acetic acid (Merck) and magnesium fluoride (Himedia). The photoanode paste was prepared from ethyl cellulose (EC, Ottokemi), terpineol (Himedia) and dibutyl phthalate (Merck). A fluorine-doped tin oxide glass plate (FTO, TEC-7 and 2.2 mm) and N719 dye (Sigma-Aldrich) were used.

2.2 Fabrication of pure and MgF₂ embedded TiO₂ NFs

Pure and MgF embedded electrospun TiO NFs were fabricated by 2 2 electrospinning method with the following procedure; the precursor solutions composed of 2.7 g of titanium (IV) isopropoxide, 2.5 mL of acetic acid, 2.5 mL of ethanol and 15 mL of 5 wt.% polyvinylpyrrolidone (PVP) along with two different wt.% of MgF (5 and 10 wt.%) added 2 with continuous stirring for 12 h to get viscous solutions.20 They were then loaded separately into a syringe connected with a 27 gauge stainless steel needle. The applied potential and the collector distance from the tip of the needle were fixed as 24 kV and 12 cm, respectively. The flow rate of the solution was 0.5 mL h-1 to get the MgF embedded electrospun TiO 2 2 mats. They were removed from the collector and kept in a vacuum oven at 80 ºC for 6 h. They were then calcined at 500 ºC for 4 h to get both pure and MgF embedded electrospun TiO NFs.

2.3 Characterization of pure and MgF₂ embedded TiO₂ NFs

The physical characterization such as TG/DTA and X-Ray diffraction analysis (XRD) were carried out as per the procedure given in our previous studies.20-22 The band gap, dye adsorption and anti-reflecting properties of the prepared TiO and MgF embedded TiO NFs were 2 2 2 studied by UV-Vis spectrometer.20,21 The morphology of the prepared TiO and MgF embedded TiO NFs were studied by FE-SEM and TEM 2 2 2 analysis. The Raman spectra and Photoluminescence characterization were also done as per the procedure given in our previous studies.20,21 Electrochemical AC-impedance analysis was made in the frequency range of 1m Hz to 100 kHz with an AC amplitude of 10 mV. For all physical characterization, we used uniform coating thickness of 12 μm for all the studied systems

2.4 Photovoltaic performance of DSSCs

The photovoltaic performance of DSSCs fabricated by using both pure and MgF embedded TiO NFs as photoanodes were examined 2 2 by AM 1.5 solar simulator (Newport, Oriel Instruments, USA 150 W; Model: 67005) and a light intensity of 100 mWcm-2 with a computer controlled digital source meter (Keithley, Model: 2420) as described elsewhere.20,21 We fabricated three DSSCs for each system to study their photovoltaic performances.

3. Results and discussion

3.1 Thermal behaviour and structural properties

Fig. 1a shows the thermal behaviour of electrospun pure PVP/TiO 2 and MgF /PVP/TiO precursors. According to the TG curve, ~25% of 2 2 weight loss has identified for PVP/TiO and MgF /PVP/TiO samples 2 2 2 in the temperature range of 35 °C to 200 °C. The major weight loss of ~44% and ~35% have identified in the region of 350 °C to 550 °C for PVP/TiO and MgF /PVP/TiO samples, respectively. These 2 2 2 results indicate that while embedding MgF in TiO NFs, the weight 2 2 loss reduced and there is no decomposition of MgF in that region. 2 The evaporation of solvent molecules and other organic residues are indicated by a long endothermic peak at ~90 °C for MgF /PVP/TiO 2 2 sample and ~100 °C for PVP/TiO sample. A small exothermic peak 2 in the range of ~340 °C to 400 °C is observed for both the samples is due to the decomposition of PVP on the main chain.23 A weak endothermic peak at ~450 °C indicates the phase change of TiO 2 from amorphous to anatase is found in the DTA curve. At the same time, there is no weight loss after 470 °C in the TG curve.

Fig. 1b shows the XRD pattern of pure TiO NFs and MgF 2 2 embedded TiO NFs calcined at 500 °C. For pure TiO NFs, the XRD 2 2 pattern well index to the anatase phase and non-appearance of diffraction peaks at 27° and 31° (Fig. S1) specify that the TiO 2 sample is free from the rutile and brookite structure. The MgF 2 embedded TiO NFs exhibit mixed phases of anatase and rutile. 2 From the XRD result, the weight fraction of these two phases calculated from the following equation:24

where W , I and I represent rutile weight percentage, the integrated R A, R intensity of anatase (101) and rutile (110) peaks, respectively. The calculated rutile content by the addition of 5 and 10 wt.% MgF in 2 TiO NFs are approximately 26% and 17%, respectively. The rutile 2 content is reduced by addition of 10 wt.% MgF in TiO NFs because 2 2 MgF is a rutile compound, and hence while adding, it induces the 2 rutile nature of TiO . Beside that, the interaction of Mg2+ with Ti4+ 2 also induces the rutile nature. But, while adding 10 wt.% MgF , Mg2+ 2 may form MgO during calcination process and hence the rutile nature is reduced to17% when compared to the addition of 5 wt.% MgF in TiO NFs.25-27

Fig. 1 (a) Thermal analysis of pure TiO and MgF embedded TiO precursor samples; (b) XRD patterns of pure 2 2 2 and MgF embedded TiO NFs calcined at 500 °C.

Fig. 2 shows the Raman spectra of pure and MgF embedded 2 TiO NFs. There is no secondary peak related to rutile, MgF Mg and 2 2 its oxide. The normal modes of vibration for pure TiO tetragonal 2 anatase phase are identified at 142 (E ), 196 (E ), 396 (B ), 515 (A g g 1g 1g + B ) and 638 cm-1 (E ). For 5 wt.% NFs, the 1g g MgF embedded TiO 2 2 normal modes of vibration are observed at 145 (E ), 196 (E ), 393 g g (B ), 522(A + B ) and 640 cm-1 (E ). Raman peaks are observed at 1g 1g 1g g 144 (E ), 196 (E ), 396 (B ), 515 (A + B ) and 638 cm-1 (E ) for 10 g g 1g 1g 1g g wt.% MgF embedded TiO NFs. These results are matched with 2 2 previously reported values.27 The E peak corresponds to O−Ti−O g symmetric stretching vibration of TiO , B peak corresponds to 2 1g symmetric bending vibration of O−Ti−O, and the A peak 1g corresponds to O−Ti−O anti-symmetric bending vibration of TiO 28 2. Changes in the Raman shift is mainly at low-frequency E peak and g high-frequency E peak. The intensity of Raman lines at low- g frequency E peak value drastically decrease for 5 wt.% g MgF 2 embedded TiO MgF 2 2 NFs compared to the pure and 10 wt.% embedded TiO2 NFs and the intensity of Raman line increases for 10 wt.% MgF embedded TiO NFs when compared to pure and 5 wt.% 2 2 MgF embedded TiO 2 2 NFs. The reason for changes in the E peak is g due to the disorder induced and phonon confinement effect by the embedded MgF , which affects the lattice vibrational characteristics 2 and also induces the change in the Raman lines of TiO NFs.29

Fig. 2 Raman spectra of (a) pure TiO NFs; (b) 5 wt.% MgF embedded TiO Nfs; (c) 10 wt.% MgF embedded TiO NFs.

3.2 Optical and morphology properties

Fig. 3a shows the absorption of pure TiO NFs and MgF embedded 2 2 TiO NFs. The maximum absorption in the range of 380 nm to 200 2 nm is due to the addition of MgF in TiO . The band gaps of the 2 2 prepared sample have changed from 3.20 eV to 3.19 eV by the addition of MgF in TiO NFs (Fig. S2). The bandgap of 5 wt. % of 2 2 MgF embedded TiO NFs has 3.13 eV, where the rutile phase is 2 2 26%. But the band gap of pure TiO rutile phase is 3.0 eV and its 2 pure anatase phase is 3.2 eV.30 The decrease in band gap is due to the high rutile phase concentration. The 10 wt.% of MgF embedded 2 TiO NFs has a bandgap of 3.19 eV where the rutile phase is 17%, 2 due to the decrease of rutile phase of the system, which increased the band gap to 3.19 eV. The relationship between the flat-band potential, V (NHE) and the band gap, E is given by Eq. (2).31 fb g Accordingly, the flat-band potential of 10 wt.% MgF embedded TiO 2 2 NFs (-0.25 eV) sample lies at the position slightly lesser than pure TiO (-0.26 eV) and for 5 wt.% MgF embedded TiO NFs (-0.19 eV) 2 2 2 sample, the V is also lying at more positive side than other two fb samples. Due to more shift in the V affects the electron transfer to fb the FTO glass for DSSC fabricated by using 5 wt.% MgF embedded 2 TiO NFs as photoanode.

Fig. 3b shows the transmittance spectra of pure TiO NFs and MgF 2 2 embedded TiO NFs. The pure TiO NFs sample has exhibited less 2 2 transmittance of light in the region of 400-550 nm, whereas MgF 2 embedded TiO NFs exhibited an increase in transmittance of light. 2 This is due to the low refractive index of MgF (1.35), which produce 2 high anti-reflecting nature. The increment in the transmittance is greatly dependent on their refractive index.32-34 This result indicates that 10 wt.% MgF embedded TiO NFs works as a good anti- 2 2 reflective material and the incident light loss has much lesser than other two photoanodes fabricated by using pure TiO NFs and 5 2 wt.% of MgF embedded TiO NFs. The absorption peak of N719 2 2 dye on the photoanode fabricated by using pure TiO NFs and MgF 2 2 embedded TiO NFs are observed at 500-550 nm, where the higher 2 dye absorption has observed in 5 wt.% and 10 wt.% of MgF2 embedded TiO NFs (Fig. 3c). This can be explained by using the 2 isoelectric point (IEP) of MgF and TiO which have IEPs of 9.5 and 2 2 6.2, respectively.8,35 While immersing the photoanodes in the dye solution, the oxide surface gets a positive charge and dye molecules

Fig. 3 (a) UV-Vis absorption spectra; (b) Transmittance spectra of pure TiO NFs and MgF embedded TiO NFs; (c) Absorption spectra of 2 2 2 N719 dye adsorbed on pure TiO NFs and MgF embedded TiO NFs photoanodes.

get a negative charge. MgF more positive than TiO and hence the 2 2 MgF embedded TiO NFs has higher dye loading than the pure TiO 2 2 2 NFs.8 The transmittance spectrum of the low refractive index of MgF has higher transmittance and this is one of the key points to 2 explain the higher dye loading on the system, which has a lower refractive index compared to the pure TiO NFs. The refractive index 2 is also related to porosity and hence the introduction of porosity also yields the low refractive index on the system.32,33,36,37 These results revealed the enhanced dye loading and photovoltaic performance of MgF embedded TiO NFs.

Fig. 4a&b shows FE-SEM images of electrospun pure TiO and 2 MgF embedded TiO mat before calcination. The electrospun TiO 2 2 2 NFs appear quite smooth due to its amorphous nature. Each individual TiO NFs are quite uniform in cross-section an 2 with average diameter of 200 nm. Fig. 4c&d show the FE-SEM image of pure TiO NFs and MgF embedded TiO NFs after calcination 2 2 2 . Its diameter is appeared to decrease from 200 to 90 nm. Energy dispersive spectra (EDS) of pure TiO NFs and 10 wt.% of MgF 2 2 (Fig. S3) embedded TiO NFs confirmed that the MgF present in the 2 2 TiO NFs.

Fig. 5a&b shows the typical TEM images of 10 wt.% MgF 2 embedded TiO NFs and its average diameter is found to be ~100 nm 2 and whereas the individual TiO NF is made up of arrays of 2 nanocrystals. The HRTEM image (Fig. 5c) shows the plane orientation of different phases present in 10 wt.% MgF embedded 2 TiO NFs. Fig. 5c indicates that the interplanar spacing of 0.35 nm, 2 0.25 nm and 0.32 nm corresponding to TiO anatase, TiO rutile, and 2 2 MgF phases, respectively. Fig. 5d shows the selected area electron 2 diffraction (SAED) pattern which confirms that the MgF embedded 2 TiO NFs is composed of polycrystalline phase. The Debye-Scherrer 2 concentric rings correspond to the phases denoted as A, R, and M which correspond to anatase, rutile and magnesium fluoride, respectively and their planes are indexed in SAED pattern (Fig. 5d).

Fig. 6 shows PL emission spectra of pure TiO NFs, 5 and 10 2 wt.% MgF embedded TiO NFs excited at 290 nm. All the three 2 2 samples showed a similar PL pattern with a strong emission band at 378 nm. The photocurrent generation process in DSSC generally includes charge separation, charge transport, and charge recombination. It is well reported that the external embedded carriers will control the electron-hole separation at the host semiconductor electrode. Among the three samples, 10 wt.% MgF embedded TiO 2 2 NFs have exhibited much lower emission intensity than pure TiO 2 NFs and 5 wt.% MgF embedded TiO . It indicates that the 2 2 NFs recombination of photoinduced charge carrier was greatly inhibited in 10 wt.% MgF embedded TiO NFs based heterojunction 2 2 system,38,39 which implies high PCE of the system.

3.3 Electrochemical and photovoltaic performance of DSSCs

The electrochemical impedance studies explain the internal resistance of DSSC. The impedance spectra of DSSCs fabricated using pure TiO NFs and MgF embedded TiO NFs as photoanodes 2 2 2 are shown in Fig. 7a. The Nyquist plot exhibits two semicircles, the small semicircle at higher frequency corresponds to the charge transfer resistance at the interface of the redox electrolyte/Pt counter electrode (R ) and the larger semicircle at lower frequency ct1 corresponds to charge transfer resistance at TiO NFs/dye/electrolyte 2 interface (R ). The Warburg diffusion process of I-/I - redox couple ct2 3 in an electrolyte (Z ) is virtually overlapped by R . The larger

Figure 4. FE-SEM image of (a) pure TiO mat; (b) 10 wt.% MgF 2 2 embedded TiO mat before calcination; (c) pure TiO NFs and (d) 2 2 10 wt.% MgF embedded TiO NFs after calcination at 500°C.

Fig. 5 (a & b) TEM images of 10 wt.% of MgF embedded TiO 2 2 NFs; (c & d) HRTEM image and SAED pattern of 10 wt.% of MgF embedded TiO NFs.

Fig. 6 Photoluminescence emission spectra of (a) pure TiO NFs; (b) 2 5 wt.% MgF embedded TiO NFs; (c) 10 wt.% MgF embedded TiO 2 2 2 2 NFs.

Fig. 7 (a) Impedance spectra; (b) Mott – Schottky electrochemical analysis of pure TiO NFs and MgF embedded TiO NFs.

Table 1 Electrochemical impedance parameter values of DSSCs fabricated using pure TiO₂ NFs and MgF₂ embedded TiO₂ NFs as photoanodes.

Photoanode	Rs	Rct1 ( Ωcm-2)	Rct2 ( cm-2)
Pure TiO2 NFs	15.06	6.89	23.84
5 wt.% MgF2 embedded TiO2 NFs	14.90	7.28	22.35
10 wt.% MgF2 embedded TiO2 NFs	13.81	6.42	18.09

semicircle decreased with increasing amount of MgF . DSSC 2 fabricated by using 10 wt.% MgF embedded TiO NFs as the 2 2 photoanode has exhibited less R (Table 1) than other two systems.40

The MgF embedded system influence the TiO band structure 2 2 and energy levels which is evaluated with Mott – Schottky (MS) analysis. The MS plots are obtained for pure TiO NFs and MgF 2 2 embedded TiO NFs photoanodes are shown in Fig. 7b. The plot 1/C2 2 vs Applied Potential (V) exhibits a typical positive slope characteristic of n-type semiconducting behaviour. The flat band potential possibilities are gained by MS plot as acquired by a linear extrapolation to C=0, i.e., the capture at the X-axis.41 The flat band potential of three samples are -0.63, -0.75, -0.55 V (vs Ag/AgCl). It is noteworthy that the embedded system changes the Fermi level and flat band potential of TiO NFs this is a further witness from UV-Vis 2 analysis.

Fig. 8 represents the photocurrent density-voltage (J-V) curves for DSSCs fabricated by using pure TiO NFs and MgF embedded 2 2 TiO NFs as photoanodes at a light intensity of 100 mW cm-2 under 2 the standard global AM 1.5 irradiation and their corresponding photovoltaic parameters are summarized in Table 2. It shows that the DSSC fabricated by using 10 wt.% MgF embedded TiO NFs as the 2 2 photoanode has the maximum efficiency of 5.56 % with a short- circuit photocurrent density (J ) of 11.2 mA/cm2. MgF is an anti-reflective material which helps to reduce the incident photon loss by reducing the reflection of light and increasing light absorption of dye molecules in the system. This leads to more electron generation which results in high photocurrent density.

Fig. 8 J-V curves of DSSCs fabricated with different wt.% of MgF₂ embedded TiO₂NFs as photoanodes.

Table 2 Photovoltaic parameter values of DSSCs fabricated with different weight pecent of MgF₂ embedded TiO₂NFs as photoanodes.

MgF2 wt.%	J sc (mA cm-2 )	Voc (V)	FF	η (%)
0	10.3	0.66	0.70	4.76
5	10.6	0.67	0.71	4.95
10	11.2	0.69	0.72	5.56

The photocurrent density (J ) is directly proportional to the sc physical parameter of the cell which is given by the Eq. (3):42

where I is the incident photon flux, Φ is the fraction of incident 0 LH light absorbed by the dye is termed as light harvesting efficiency and is determined by the amount of dye present in the system, Φ is the CS quantum efficiency of charge separation in the event. Φ is the COL fraction of the separated charges measured as photocurrent under short circuit condition is termed as collection efficiency. The DSSC fabricated by using 10 wt.% MgF embedded TiO NFs as the 2 2 photoanode exhibited higher photocurrent density compared to pure TiO NFs and 5 wt.% MgF embedded TiO NFs based DSSC 2 2 2 systems due to the higher light absorption of dye (Φ ) and the LH mixed phase of anatase and an optimum amount of rutile structure was good for electron transport. According to the PL study, the highest charge separation efficiency (Φ ) of photoinduced electron- CS hole pairs occurred in the heterojunction system. According to the equation (3), the higher charge separation efficiency (Φ ) and higher CS light absorption of dye (Φ ) imply the higher photocurrent density LH and higher power conversion efficiency than the pure TiO NFs and 5 2 wt.% MgF embedded TiO NFs based DSSCs. But the DSSC 2 2 fabricated by using 5 wt.% MgF embedded TiO NFs as the 2 2 photoanode has higher light absorption of dye which leads to producing more electrons but due to the reduction in band gap the conduction band level has much reduced so the electrons diffused through the TiO NFs is much reduced which reduced the charge 2 separation efficiency (Φ ) that reduced the PCE.

4. Conclusion

Both, pure TiO NFs and MgF embedded TiO NFs were 2 2 2 successfully fabricated by electrospinning process. The thermal studies showed complete crystallization temperature of the prepared precursor sample and it was found to be 500 °C. The evolution of rutile phase was analyzed by XRD and the weight fraction of the anatase and rutile phases were calculated from the anatase (101) and rutile (110) peaks. The shift in flat band potential, optical band gap and the dye adsorption nature of the photoanodes were analyzed by UV-Vis spectroscopy. The surface morphology was revealed by FE-SEM and TEM images of the prepared TiO2 NFs and MgF embedded TiO NFs. The photovoltaic performance 2 2 of DSSC fabricated by using 10 wt.% MgF embedded TiO NFs as 2 2 photoanode exhibited the maximum light to electricity efficiency (η) of 5.56% with 17% improvement than that fabricated with pure TiO NFs as photoanode.

Conflicts of interest

The authors declare that they have no conflict of interest.

Acknowledgements

The authors gratefully acknowledge the CIF of Pondicherry University for extending the instrumentation facilities. Dr. AS gratefully acknowledges the CSIR, New Delhi for providing the financial support (No.01(2810)/14/EMR-II). Mr. MV is grateful to the Indo-US Science and Technology Forum (IUSSTF), Department of Science and Technology (DST), New Delhi for providing a fellowship under the BASE program (IUSSTF BASE Internships 2018/ 12/ Vignesh M, dt. 09/04/2018).

References

1. A. Jena, S. P. Mohanty, P. Kumar, J. Naduvath, V. Gondane, P. Lekha, J. Das, H. K. Narula, S. Mallick and P. Bhargava, Trans. Indian Ceram. Soc., 2012, 71, 1-16.
2. Subramania A, Vighnesh M, Subasri A and Sarathkumar K, Eng. Sci., 2018, In press, www.doi.org/10.30919/es8d756.
3. V. Murugadoss, N. Wang, S. Tadakamalla, B. Wang, Z. Guo and S. Angaiah, J. Mater. Chem. A, 2017, 5, 14583-14594.
4. V. Elayappan, V. Murugadoss, S. Angaiah, Z. Fei and P. Dyson, J. Appl. Polym. Sci., 2015, 132, 42777.
5. Z. Salam, E. Vijayakumar, A. Subramania, N. Sivasankar and S. Mallick, Sol. Energy Mater. Sol. Cells, 2015, 143, 250-259.
6. T. Liu, J. Hou, B. Wang, F. Bai, H. Chen, L. Gao, Y. Cao, H. He, J. Wang, N. Wang, G. Cao and Z. Guo, J. Mater. Chem. A, 2016, 4,10794-10800.
7. S. K. Kim, M. K. Son, S. Park, M. S. Jeong, H. Seo, K. Prabakar and H. J. Kim, Electron. Mater. Lett., 2014, 10, 229-234.
8. H. Park, D. J. Yang, H. G. Kim, S. J. Cho, S. C. Yang, H. Lee and W. Y. Choi, J. Electroceram., 2009, 23, 146-149.
9. E. H. Kong, Y. H. Yoon, Y. J. Chang and H. M. Jang, Mater. Chem. Phys., 2014, 143, 1440-1445.
10. L. Yang, X. Wang, X. Mai, T. Wang, C. Wang, X. Li, V. Murugadoss, Q. Shao, S. Angaiah and Z. Guo, J. Colloid Interface Sci., 2018, In press, https://doi.org/10.1016/j.jcis.2018.09.045.
11. X. Feng, K. Shankar, O. K. Varghese, M. Paulose, T. J. Latempa and C. A. Grimes, Nano Lett., 2008, 8, 3781-3786.
12. Z. R. Tian, J. A. Voigt, J. Liu, B. McKenzie and H. Xu, J. Am. Chem. Soc., 2003, 125, 12384-12385.
13. K. Zhu, N. R. Neale, A. Miedaner and A. J. Frank, Nano Lett., 2007, 7, 69-74.
14. G. K. Mor, K. Shankar, M. Paulose, O. K. Varghese and C. A. Grimes, Nano Lett., 2006, 6, 215-218.
15. P. Zhu, A. S. Nair, P. Shengjie, Y. Shengyuan and S. Ramakrishna, ACS Appl. Mater. Interfaces, 2012, 4, 581-585.
16. J. Jiu, S. Isoda, F. Wang and M. Adachi, J. Phys. Chem. B, 2006, 110, 2087-2092.
17. H. Li, B. Jiang, R. Schaller, J. Wu and J. Jiao, J. Phys. Chem. C, 2010, 114, 11375-11380.
18. G. Zhu, L. Pan, J. Yang, X. Liu, H. Sun and Z. Sun, J. Mater. Chem., 2012, 22, 24326-24329.
19. Y. Wang, E. Chen, H. Lai, B. Lu, Z. Hu, X. Qin, W. Shi and G. Du, Ceram. Int., 2013, 39, 5407-5413.
20. V. Elayappan, P. Panneerselvam, S. Nemala, K. S. Nallathambi and S. Angaiah, Appl. Phys. A, 2015, 120, 1211-1218.
21. P. Pratheep, E. Vijayakumar and A. Subramania, Appl. Phys. A, 2015, 119, 497-502.
22. A. S. Nair, Z. Peining, V. J. Babu, Y. Shengyuan and S. Ramakrishna, Phys. Chem. Chem. Phys., 2011, 13, 21248-21261.
23. Y. K. Du, P. Yang, Z. G. Mou, N. P. Hua and L. Jiang, J. Appl. Polym. Sci., 2005, 99, 23-26.
24. J. Ding, Y. Li, H. Hu, L. Bai, S. Zhang and N. Yuan, Nanoscale Res. Lett., 2013, 8, 9.
25. D. S. Hwang, N. H. Lee, D. Y. Lee, J. S. Song, S. H. Shin and S. J. Kim, Smart Mater. Struct., 2006, 15, S74.
26. K. K. Kim and A. S. Nowick, J. Phys. C: Solid State Phys., 1977, 10, 509.
27. D. S. Hwang, Y. I. Cho, N. H. Lee, H. G. Lee, D.-H. Cho and S. J. Kim, Key Eng. Mater., 2004, 264-268, 21-24.
28. F. Tian, Y. Zhang, J. Zhang and C. Pan, J. Phys. Chem. C, 2012, 116, 7515-7519.
29. W. F. Zhang, Y. L. He, M. S. Zhang, Z. Yin and Q. Chen, J. Phys. D: Appl. Phys., 2000, 33, 912.
30. B. Choudhury, S. Bayan, A. Choudhury and P. Chakraborty, J. Colloid Interface Sci., 2016, 465, 1-10.
31. S. Iwamoto, Y. Sazanami, M. Inoue, T. Inoue, T. Hoshi, K. Shigaki, M. Kaneko and A. Maenosono, ChemSusChem, 2008, 1, 401-403.
32. F. Chi, L. Yan, H. Lv, H. Yan, X. Yuan and B. Jiang, Nanosci. Nanotechnol. Lett., 2012, 4, 441-444.
33. S. Chhajed, M. F. Schubert, J. K. Kim and E. F. Schubert, Appl. Phys. Lett., 2008, 93, 251108.
34. L. Zhang, Y. Li, J. Sun and J. Shen, Langmuir, 2008, 24, 10851- 10857.
35. I. Sevonkaev and E. Matijević, Langmuir, 2009, 25, 10534- 10539.
36. Y. Du, L. E. Luna, W. S. Tan, M. F. Rubner and R. E. Cohen, ACS Nano, 2010, 4, 4308-4316.
37. M. S. Park, Y. Lee and J. K. Kim, Chem. Mater., 2005, 17, 3944- 3950.
38. J. Choi, P. Sudhagar, P. Lakshmipathiraj, J. W. Lee, A. Devadoss, S. Lee, T. Song, S. Hong, S. Eito, C. Terashima, T. H. Han, J. K. Kang, A. Fujishima, Y. S. Kang and U. Paik, RSC Adv., 2014, 4, 11750-11757.
39. Z. Zhang, C. Shao, X. Li, Y. Sun, M. Zhang, J. Mu, P. Zhang, Z. Guo and Y. Liu, Nanoscale, 2013, 5, 606-618.
40. A. M. Cant, F. Huang, X. L. Zhang, Y. Chen, Y.-B. Cheng and R. Amal, Nanoscale, 2014, 6, 3875-3880.
41. X. Jin, Q. Li, Y. Li, Z. Chen, T.-H. Wei, X. He and W. Sun, Sci. Rep., 2014, 4, 5983.
42. B. C. O'Regan, J. R. Durrant, P. M. Sommeling and N. J. Bakker, J. Phys. Chem. C, 2007, 111, 14001-14010.