Photoexcitation of neodymium doped TiO2 for improved performance in dye-sensitized solar cells

    Photoexcitation of Neodymium Doped TiO 2 for Improved Performance in Dye - Sensitized Solar Cells Laveena P. D’Souza, R. Shwetharani, Vipin Amoli, C.A.N. Fernando, Anil Kumar Sinha, R. Geetha Balakrishna PII: DOI: Reference:

S0264-1275(16)30585-8 doi: 10.1016/j.matdes.2016.05.007 JMADE 1745

To appear in: Received date: Revised date: Accepted date:

12 December 2015 6 April 2016 4 May 2016

Please cite this article as: Laveena P. D’Souza, R. Shwetharani, Vipin Amoli, C.A.N. Fernando, Anil Kumar Sinha, R. Geetha Balakrishna, Photoexcitation of Neodymium Doped TiO2 for Improved Performance in Dye - Sensitized Solar Cells, (2016), doi: 10.1016/j.matdes.2016.05.007

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Photoexcitation of Neodymium Doped TiO2 for Improved Performance in Dye - Sensitized Solar Cells Laveena P D’Souzaa, Shwetharani Ra, Vipin Amolib, CAN Fernandoc, Anil Kumar Sinhab and R

Center for Nano and Material Sciences, Jain University, Jain Global Campus, Bangalore 562112, India

Indian Institute of Petroleum, Dehradun- 248 005, India Wayamba University of Sri Lanka

NU

c

SC

b

RI

a

PT

Geetha Balakrishnaa*

Abstract:

MA

A series of Nd-doped TiO2 with Nd content ranging from 1 to 5 atom % were synthesized by solid state technique and explored as a photoanode material in dye-sensitized solar cells (DSSCs). Incorporation of Nd3+ into TiO2 lattice resulted in more negative flat band potential for

D

Nd-doped TiO2 than pure TiO2, as determined by impedance spectroscopy (Mott−Schottky

TE

plots), which is highly desirable to achieve higher Voc in DSSCs. In addition, Nd doping in TiO2 provided the opportunities to engender higher density of oxygen vacancies and bang gap

AC CE P

narrowing in Nd-TiO2 photoanode, which collectively enhanced the optical properties of photoanode DSSC in terms of higher dye loading and light absorption respectively. A best efficiency of 6.17% was achieved with 4 % Nd-doped titania, resulting in a remarkable increase in efficiency (102.3%) of the device over the undoped titania cells. Induction of intermediate bands, realignment of energy levels leading to a better electron injection, high sensitizer loading, efficient charge separation due to dopant and its impurity levels contribute favorably towards the superior performance of Nd doped TiO2 DSSC. Spectroscopy (EIS) measurements showed Nd doped TiO2 photoanode possessed higher charge recombination resistance and longer electron lifetime compared to DSSCs with undoped TiO2. Key words: Nd doped TiO2, negative flat band potential, higher dye loading, charge recombination, Dye-sensitized solar cell

1

ACCEPTED MANUSCRIPT

1. Introduction Dye sensitized solar cells (DSSCs) that use dye molecules to absorb photons and convert them to

PT

electric charges, have drawn extensive attention due to their ease of device making, low cost and their comparatively better energy-conversion efficiency [1-3]. A conventional DSSC made of a

RI

dye-activated mesoporous semiconductor oxide film on conducting glass, an iodine electrolyte

SC

solution, and a platinised counter electrode [4]. Contradictory to typical silicon solar cells, wherein the semiconductor takes charge of both the photon absorption and charge transport, in

NU

DSSCs, light is absorbed by the sensitizer (dye) and allows electron injection from the sensitizer to the conduction band of the semiconductor. There are several ways to enhance the performance

MA

of DSSCs namely by (a) increase of light harvesting, which could be achieved by good surface area, and absorption of broader range of solar light; [5] (b) increase of the electron injection speed by improving the electron injection over potential; [6, 7] (c) moving the redox couple

D

Fermi level to enhance the dye regeneration rate; [8, 9] (d) enhancing the lifetime of electrons by

TE

retarding the probability of charge recombination; [10] and (e) improving the charge transfer rate in TiO2 [11, 12]. Appropriate improvement of the photoanode, more so the modification of the

AC CE P

nanoparticles (NPs) used in working electrode can exhibit a relatively strong capability to meet all the above criteria via a tailored microstructure of semiconductor NPs which excels in light harvesting, electron injection, electron transport and charge separation [13]. TiO2, in special, is a broadly accepted material as a good electron transport and separation medium for DSSCs in which the voltage occurs at the interface between the dye-sensitized metal oxide semiconductor and an electrolyte [14]. TiO2 has also been the most promising material among the other semiconductors used in DSSCs due to its exceptional properties such as good oxidative ability, strong chemical stability, being economical and possessing environment friendly features [15]. The mesoporous TiO2 working electrodes that have percolated links of the NPs yield very high photocurrent due to their enhanced surface area apt for dye adsorption and favorable energy level alignment to allow electrons from the dye molecules to nanostructured films [16]. Doping of TiO2 with transition metal ions can further enhance optical absorption, extend band edge energies, and increase electron density of states and favorably align Fermi levels [17-20]. Recently reports indicate that lanthanides are effective dopants for altering the crystal structure of TiO2 due to their unique electronic configuration and spectral characteristics to 2

ACCEPTED MANUSCRIPT achieve the above desired properties of DSSCs [21, 22]. Zhang et al [21] described that doping of lanthanides can cause prominent improvement in the photoresponse of TiO2, and this improvement was susceptible to the ionic, electronic configuration and atomic radius.

PT

Lanthanide dopants can reduce the band gap of TiO2 and the fundamental absorption edge redshifts to visible region, still keeping its strong redox potentials. The Ti 3d states of conduction

RI

band (CB) are split into eg and t2g states in a ligand field with Oh symmetry, and hence the CB is

SC

separated into upper and lower parts [23]. On doping of lanthanide into TiO2 (although a nanoparticle), their 4f/5d states are more or less delocalized, thus adding significantly to form

NU

impurity energy levels (IELs), thus broadening the CB or VB [24]. This results in band gap narrowing. An enhanced utilization of the broader range of solar spectrum could be achieved

MA

with intermediate bands (IBs) formed by IELs [25]. Properly located IBs behave as stepping stones to allow low-energy photons to excite electrons from VB to CB. IBs also serve to trap and detrap electrons thus prolonging the lifetime of charge carriers. In addition lanthanides

D

proportionately plunder oxygen into TiO2 lattice to cause surface oxygen vacancies for dye

TE

absorption and thus electron harvesting [26]. Doping of Sm3+, Nd3+, Pr3+, Eu3+ and La3+ in particular could enhance the photo

AC CE P

electrochemical properties, improve the photoresponse and the photocurrent conversion efficiency for the range 350 - 400 nm [27, 28]. In a DSSC the power conversion efficiency (PCE) of the device is determined by the open-circuit photovoltage (Voc), short-circuit photocurrent (Jsc), and the fill factor (FF). Nd3+ being a dopant induces IBs which could steeply reduce the band gap to the visible range [29-31] and allow the semiconductor to photoexcite electrons for electron transport, thus favorably influencing the density of electrons and hence Jsc. The mismatch of ionic radius of Nd3+ (0.983 Å) and Ti4+ (0.605Å) tends to induce lattice distortion/defects [13]. The obtained larger diameters of doped particles can contribute to scattering of incident light [1] which can confine the incident light better and reduce the loss of photon energy [32, 33] and hence enhance Jsc. The open circuit potential (Voc) is given by the potential difference between the quasi-Fermi level of oxides and Nernstian potential of the electrolyte [34]. So, if the number of the electrons being injected is enhanced and/or loss of the injected electrons by recombination is less, Voc would improve, increasing the oxide’s quasiFermi level. Nd3+ doping allows feasible CB shift and engineers a more favorable equilibrium

3

ACCEPTED MANUSCRIPT Fermi level position for an enhanced Voc. The substitutional Nd3+ dopant thus was aimed to contribute to both an enhanced Jsc and Voc and hence an overall efficiency of the solar cell. The above findings inspired us to synthesize Nd doped TiO2 NPs for application of these

PT

NPs as photoanode material for DSSCs. There have been reports on the preparation of Nd - TiO2 by hydrothermal and solvothermal method and being utilized as the photoanode of

RI

nanostructured solar cells [35, 36]. Use of chemical methods for dopants tend to cause deep

SC

dopant energy levels which may not be available for charge transfer and may also end in recombination adversely affecting the efficiency. Hence physical process of pulverization has

NU

been attempted to cause surface level doping. Significant changes in the conversion efficiency of DSSCs using different and nominal trace concentrations of Nd dopant (in TiO2) were observed,

MA

analyzed and related to morphological, electronic and optical changes that were studied by spectroscopy and microscopy. The kinetics of charge transport and recombination rate was

D

studied by Electron Impedance Spectroscopy.

2.1. Materials and Methods

TE

2. Experimental

AC CE P

All the chemical reagents were used without any further purification. Nd2O3. 5 H2O (99.9%, Loba chemicals) was used as precursor for the neodymium doping, TiCl4 (99.5% loba Chemie) was used as a titanium source for the preparation of TiO2. Fluorine doped tin oxide (FTO) glass (Sigma Aldrich). The others include sulphuric acid, sodium hydroxide, ammonium hydroxide, acetic

acid,

ethanol

were

from

Merck

and

Triton

X-100,

Ruthenium

535

dye

(Ru(bpy)2(NCS)2H4, Iodolyte TG-50, Platisol were from Solaronix.

2.2. Preparation of TiO2 The preparation of TiO2 NPs is as per our earlier protocol [37]. To brief the preparation, 100 mL aliquot of TiCl4 was added drop wise to 1 L of double distilled water with vigorous stirring and a temperature of <10 °C, to avoid the agglomeration of NPs during hydrolysis. To complete the hydrolysis 1 mL H2SO4 is added followed by the addition of ammonia for precipitation till the pH reaches 7-8. The obtained precipitate was allowed to settle, washed with distilled water, filtered using Whatman 41 filter paper, dried and ground. The TiO2 powder was annealed at 600 °C for 6 hours to obtain an anatase phase of TiO2. 4

ACCEPTED MANUSCRIPT

2.3. Preparation of Nd doped TiO2

PT

Neodymium (Nd) substituted TiO2 was synthesized by solid state technique using Nd2O3. 5 H2O as precursor. 0.98: 0.02 weight ratio of TiO2 and neodymium oxide was used to obtain 1% Nd-

RI

doped TiO2. The mixture were pulverized in an agate vial using a ball mill (SPEX 8000M) for a

SC

time period of 90 minutes and annealed at 600 °C for 6 hours. Similar procedure was adopted by varying the weight ratio of the precursor and TiO2 to obtain 2%, 3%, 4% and 5% Nd doped TiO2.

NU

To cause an effective doping, pulverization and annealing was repeated. Plastic deformation of TiO2 crystal lattice takes place when it undergoes the process of milling, causing stresses and

MA

strains in the lattice. This produces lattice distortion and forms many defects inside TiO2 particles. The high surface energy and lattice distortion energy creates activation energy for atomic and ionic diffusion at room temperature. During the process of pulverization the collision

D

that occurs between the grains of powder and the balls of the mixer mill causes a rise in interface

TE

temperature to induce incorporation of dopants [30].

AC CE P

2.4. Preparation of Paste and Fabrication of DSSC FTO substrate was cleaned in surfactant, deionized water, acetone and ethanol. Paste of TiO2 was prepared with 0.25 g of TiO2, 0.5 ml of acetic acid and 1:1 (10 ml) mixture of deionized water and ethanol ground for 20 mins. Triton - X (0.5 ml) was added, the resultant mixture was ground to get homogenous paste. Doctor blade technique was adapted to cast the TiO2 paste on FTO glass substrate [4]. TiO2 coated films were sintered at 450 oC for 30 minutes. Films were cooled at room temperature and scratched into an active area of 0.16 cm2. To attain sensitization of dye, films were dipped in ethanolic solution of 0.3 mM of cis-bis (2,2’-bipyridyl-4,4’-dicarboxylato) ruthenium(II)-bis-tetrabutylammonium (called N719 dye) for 24 hours. The sensitized electrodes were washed with ethanol to remove the unanchored dye. Similar method was adopted for the preparation of different percentages Nd-TiO2 NPs. Counter electrode was obtained by placing a thin layer of platisol which was squeegee printed using a polyester mesh sintered at 450 °C for 30 minutes. A drop of Iodolyte TG - 50 was casted on the surface of sensitized photoanodes, to penetrate into the porous structure via capillary action. The Pt coated FTO electrode was then clipped onto the top of TiO2 working electrode to form the complete solar cell. The adsorbed 5

ACCEPTED MANUSCRIPT amount of dye was measured by dipping dye sensitized photoanode in 0.1 M NaOH in 1:1 ethanol water solution until complete desorption of dye and the desorbed dye was measured

PT

using UV-Vis spectrometer.

2.5. Characterization Techniques:

RI

Phase identification of the nanomaterials were studied using Shimad u

D

using u

SC

radiation (scan rate o 2 min-1). Crystallite size and lattice parameters were determined by dhkl = kλ/β cos θ and 1/d2=h2/a2+k2/b2+l2/c2 [38]. The crystal planes (004), (101) and (211) were

NU

selected to determine the lattice parameters of TiO2 and Nd-TiO2. VG Scientific FSCA MK II spectrometer were used to record the X-ray photoelectron spectra (XPS) of NPs, at fixed radiation (hv = 1253.6 eV).

MA

analy er transmission mode with pass energy o 2 eV, using Mg

The absorption and diffused reflectance spectra were recorded by a Shimadzu 1700 PC UVVisible spectrophotometer. Kubelka - Munk plots of (1-R∞)2/2R∞ versus wavelength were

D

plotted, where R∞ is the ratio of relative reflectance to reflectance of non absorbing medium

TE

(100%) and band gap of the material was calculated using Eg = hc/λ. The absorption and emission spectra were recorded using a Shimadzu 1700 PC UV-Visible and Shimadzu RF 5301

AC CE P

PC spectro fluorometer. Field Emission Scanning Electron Microscope (Quanta 200 FEI Carl Zeiss (Germany), used to analyze the surface morphology. The high resolution transmission election microscopy (HRTEM) images were obtained from JEM-2010 electron microscope (JEOL, Tokyo, Japan) at an accelerating voltage of 200 kV. Random selections of 18 particles were chosen to determine the size distribution of the nanostructures. The surface area of the samples were determined by Smart Sorb 93 BET surface analyzer with sorb 93 reduction software Electrochemical measurements were performed using a CHI660D potentiostat (CH instruments, Austin, USA). The impedance spectra of devices were measured from 1 Hz to 105 Hz in dark at the applied bias of open circuit potential 0.7 V with an alternating current amplitude of 10 mV. The Mott - Schottky analysis was carried out using electrochemical impedance spectroscopy (Potentiostat, BioLogic VSP). The reference and counter electrodes were Ag/AgCl in 3 M KCl and Pt wire. The photocurrent transient measurements on the films were conducted at 20+2 °C with a BAS100B electrochemical workstation in a three-electrode cell. The working electrode in the 0.1 M Na2SO4 electrolyte solution was illuminated through the quartz window by a white light generated from a 300 Wcm2 UV lamp. Photovoltaic performance 6

ACCEPTED MANUSCRIPT of the DSSCs were measured by integrated I–V Test station (PVIV-211V) with Kiethley 2420 source meter and 94023A Oriel Sol3A, class AAA Solar Simulator (power output: 100 mW/cm2,

PT

lamp power: 450 W), equipped with an AM 1.5 filter. 3. Result and Discussions

RI

Fig. 1 illustrates the X - ray diffraction pattern of TiO2 and 1 - 5% percentages of Nd-TiO2 NPs. The peaks indicate the complete retention of anatase phase after Nd doping [38, 39]. A slight

SC

shift to lower theta values of (101) peak (as in inset of Fig. 1.) was in accordance with the Bragg’s equation, 2d sin  = nλ wherein the larger ionic radius o Nd3+ substitutes Ti4+, and this

NU

effect appeared more prominent at higher theta [11]. The calculated crystallite size as per Scherer’s equation or (1 1) peak intensity is observed to enhance rom 1 .9 nm (TiO2) to 25.8

MA

nm (4% Nd-TiO2) due to dopants. Doping results in change of lattice constants as observed in Table 1. The cell volume increases from 135.9 Å3 (TiO2) to 136.0 Å3 (4% Nd-TiO2). On Nd doping, bond lengths of Ti - O are extended slightly, with Nd - O bond lengths being appreciably

D

longer than Ti - O, it results in dilation of cell volume of doped TiO2 with an arise of local strain

TE

field at dopant sites [24]. This can cause deviation of atomic interactions and promote lattice

AC CE P

distortion [19]. However 5% doped sample shows a decrease in crystallite size and cell volume which can be attributed to high distortion.

Table. 1 Lattice parameters, average crystallite size and d - spacing of Nd-TiO2 NPs Sample

Average

crystallite

Cell

d-spacing 3

volume (Å )

(Å)

a (Å)

c (Å)

size (nm)

TiO2

17.9

135.9

3.51

3.7823

9.50

1% Nd-TiO2

25.8

136.3

3.51

3.7850

9.51

2% Nd-TiO2

25.8

136.5

3.51

3.7850

9.51

3% Nd-TiO2

25.8

136.0

3.51

3.7823

9.51

4% Nd-TiO2

25.8

136.0

3.51

3.7823

9.51

7

ACCEPTED MANUSCRIPT 17.2

135.8

3.51

3.7796

9.51

D

MA

NU

SC

RI

PT

5% Nd-TiO2

TE

Fig. 1: X-ray diffraction pattern of (a) 1% Nd-TiO2, (b) 2% Nd-TiO2, (c) 3% Nd-TiO2 (d) 4% Nd-

AC CE P

TiO2 (e) 5% Nd-TiO2 and (f) TiO2

Fig. 2 represents the XPS spectrum of undoped and Nd doped TiO2. It can be observed that the Nd 3d spectrum (Fig. 2c) is split into two peaks at 982 eV and 1003.8 eV, ascribed to Nd 3d5/2 and Nd 3d3/2 states [40]. Anatase TiO2 display peaks at 529.7 eV and 531.7 eV (Fig. 2b) attributed to Ti-O bond in TiO2 and hydroxyl groups on the surface respectively [41]. Oxygen (O 1s) binding energy is obtained at 528.3 eV compared to the undoped TiO2 reveals the Nd-O boding [42]. This lower binding energy value can be explained using electronegativities [43] of the two metal atoms, Ti (1.54) and Nd (1.14). The peak obtained at 530.2 eV could be attributed to O2- of TiO2 lattice in Nd-TiO2. The peaks obtained at 458.5 eV and 464.3 eV can be assigned to core level of Ti4+ 2p3/2 and Ti4+ 2p1/2 and the peaks at 456.9 eV and 462.7 eV corresponds to Ti3+ 2p3/2 and Ti3+ 2p1/2 (Fig 2a) respectively. The presence of Ti3+ in Nd doped TiO2 indicates oxygen vacancies engendered by Nd doping [26]. The chemical bond strength for Ti-O is 662 kJ/mol, much less than Nd-O of 703 kJ/mol [26]. Induction of Nd3+ hence into surface of TiO2 creates oxygen vacancies. Nd atom has stronger affinity and bond strength for oxygen atoms than Ti atoms. Also in DSSCs, each dye molecule allows the transfer of one electron to Ti4+ 3d0 8

ACCEPTED MANUSCRIPT band of TiO2 and form Ti3+ 3d1 level. Since the energy gap is very small between Ti4+ 3d0 and Ti3+ 3d1 energy level the transfer of electron to neighboring Ti4+ from Ti 3+ is very easy and this tends to form a space charge [11]. Presence of oxygen vacancies and Ti3+/Nd3+ can extend light

PT

absorption range, absorption intensity and better electron transfer [44-46]. A wide scan spectra of TiO2 and Nd -TiO2 (Fig. 2d) shows the presence of Ti, O and Nd peaks. In TiO2 wide scan

RI

spectrum the peaks obtained after 900 eV is attributed to the Auger peaks of titanium and oxygen

SC

[47]. A peak at 286 eV can be assigned to the C 1s is a reference peak of XPS instrument used. A small peak around 122 eV corresponds to Nd 4d, [48] indicating the dopant Nd in the doped

AC CE P

TE

D

MA

NU

sample.

Fig. 2: X-ray photoelectron spectra of TiO2 and 4% Nd-TiO2 Fig. 3 represents an orderly, well distributed sphericular particles of the Nd doped TiO 2 with particles of size averaging around 70 ± 20 nm. Solid state technique of pulverization employed to cause doping with subsequent sintering, facilitates the formation of an oxide network which can act as a pathway for charge transport [49]. The act of charge transport is greatly influenced by 9

ACCEPTED MANUSCRIPT sintering [49]. The formation of large spheroids has been observed to reduce the surface area from 35 m2/g to 15 m2/g and it is well in agreement with particle size measured from TEM [19]. Spheroids allows scattering of incident light which results in an increase of photon travelling

PT

distance [49] thus confining the incident light within them to enhance electron conduction and reduce photon loss [32, 33]. The dye loading capacity seems higher in doped samples as

RI

tabulated in Table 3, and is mainly due to the enormous oxygen vacancies created by lanthanide

TE

D

MA

NU

SC

doping [26].

AC CE P

Fig. 3: FESEM of 4% Nd-TiO2

Fig. 4a and b depict TEM images of TiO2 and 4% Nd -TiO2, suggesting an increase in mean particle size from 27 - 32 nm and 40 - 50 nm, as measured in size distribution histograms (Fig. 4c and d). Their insets depict discrete selected - area electron diffraction (SAED) patterns depicting a sequence of diffraction rings indicating polycrystallinity of both the samples [18]. Fig. 4e displays the corresponding clear lattice fringes for the doped and undoped sample indicating good crystallinity [13]. An obvious defect layer (indicated by an arrow in Fig. 4f) that could be due to the mismatch of ionic radius of Nd and Ti causes a slight distortion in the lattice structure [19]. The results obtained from this study corroborates with XRD, BET and FESEM results.

10

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

TE

Fig. 4: TEM images of (a) TiO2, (b) 4% Nd -TiO2 (inset: SAED patterns) (c) and (d) size

AC CE P

distribution histogram of TiO2 and 4%Nd -TiO2, (e) and (f) HRTEM images of 4% Nd-TiO2 The optical property of any semiconductor is greatly influenced by its electronic structure and so the extent of absorbing light is analyzed by investigating studying the relation between its electronic structure and optical property. Nd being a lanthanide, on doping, allows its 4f states to be more or less delocalized, favorably adding to formation of IELs [24], thus broadening the CB and narrowing the band gap [24]. Unlike transition metal doped TiO2, the 4f states of lanthanide do not actually hybridize with O 2p states. The IELs are located below the conduction band and have little distance between conduction band minimum and IELs and if neglected, the band gap narrows down to 2.5 eV, as evident from the Kubelka - Munk plot. The inset in Fig. 5 shows a bathochromic shift of 95 nm from the fundamental band edge (Eg) of TiO2 of 390 nm to 485 nm. A proportional shift is observed with increase in dopant concentration as shown in supporting information. This modified band edge energy enables the doped analogue to be active in the visible region indicating increased performance due to the broad range of solar light absorption. The decrease in band gap could also be corroborated with the fact of increase in crystallite size and particle size as shown in inset (a) of Scheme 1, and is as per the statement of V. I. Klimov 11

ACCEPTED MANUSCRIPT “si e dependent control to energy gap is simply proportional to 1/r2” (where r is the particle radius) [50]. The arise of new absorption peaks in the range of 360 to 390 nm can be attributed to inter band transitions while all curves (of negligible intensity) above 400 nm gradually decrease

PT

due to intraband transitions among the IELs [24]. IELs /IBs can retard the rate of electron hole

AC CE P

TE

D

MA

NU

SC

RI

recombination and enhance the photoresponse of TiO2.

Fig. 5: UV-Vis spectra of (a) TiO2 and (b) 4% Nd-TiO2 inset: Kubelka - Munk plots to measure their band gaps

Arise of IELs (IBs) and their function as electron traps to suppress recombination is substantiated by PL studies. The cause of main photoluminescence (PL) emission in any semiconductor is prominently due to radiative recombination of charge carriers. When the semiconductor is irradiated with light of wavelength  its band electron holes pairs. The photogenerated electrons return back from CB to VB with an emission of energy as photoluminescence radiation [51]. Other photoluminescence processes that can occur, includes, the transfer of excited electrons from the conduction band to different IBs, via non-radiative phenomenon, and subsequently from the IBs to the VB by radiative transition causing photoluminescence signals [51]. As observed in the Fig. 6, doping causes a significant decrease in the PL intensity, suggesting the occurrence of the latter process leading to a retarded recombination of electrons and holes. Decrease in recombination rate evidently supports the enhanced availability of electrons for charge transfer. The small emissions below 450 nm are attributed to the transitions occurring due to 12

ACCEPTED MANUSCRIPT recombination of electrons - holes in surface defects/surface oxygen vacancies [52], the extent of decrease in PL with dopant concentration indicates the proportional enhancement in defects/oxygen vacancies as shown in Fig. 6, while shoulder peak at 462 nm can be attributed to

TE

D

MA

NU

SC

RI

PT

an emission of IB transition of Nd - TiO2.

AC CE P

Fig. 6: Photoluminescence spectra of TiO2 and 1-5% percentages of Nd-TiO2 samples Although band bending is negligible for NPs, it can still exist for NPs of bigger diameters ( 10 nm) but without the clear evidence for space charge region in such mesoscopic films [53]. Assuming that band bending is present in our films, analysis of impedance for doped and undoped TiO2 films were done to evaluate flat band potentials which could approximately denote the conduction band position in the titania particles. The flat band potentials measured were found to be -0.60 V, -0.62 V and -0.63 V vs Ag/AgCl, for TiO2, 4% Nd-TiO2 and 5% Nd-TiO2 respectively. As observed in Fig. 7 the CB evidently shifts negatively with an increase in doping amount. The establishment of high Fermi level, to balance the electron injection in TiO2 and avoid recombination to I3- in the electrolyte is an important factor to be influenced by flat band potentials [53]. Hence contribution to higher Voc is majorly from the upward shift of CB which promotes easy electron injection from the excited dye molecules. Simultaneous decrease in band gap and shifting of band edge upon doping is not surprising as shown in Scheme 1. The VB of TiO2 on doping allows the corresponding electrons of the VB of Nd to contribute significantly, thus uplifting the VB as well. 13

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

D

Fig. 7: Mott-Schottky plots of a) TiO2, b) 4% Nd-TiO2 and c) 5% Nd-TiO2

Scheme 1: Schematic representation of electronic transitions in Nd-TiO2. Inset (a) size effect on semiconductor band gap The electron transfer occurring at electrodes, the electron recombination and the redox specie diffusion can be discretely identified by the shape of impedance spectra. Since the present device output stems mostly from enhanced Voc, which is a direct result of the retarded charge 14

ACCEPTED MANUSCRIPT recombination, the electrochemical impedance spectra (EIS) recorded at intermediate frequency range is analyzed and presented here. Fig. 8 illustrates the semicircle that reveals the spectra recorded at intermediate frequency region and it indicates the resistance offered to recombination

PT

of electrons in the electrode material and the redox species in the electrolyte of DSSCs [53, 54] and this is supported with an equivalent circuit as shown in Fig. 8 (right image). The obtained

RI

chemical capacitance (Cµ), recombination resistance (Rr), ohmic serial resistance (Rs), and the

SC

electron li e time (Ʈn= Rr x Cµ) [55] determined from EIS is as shown in Table 2. The larger radius of the semicircle in 4% Nd-TiO2 implies an increase in its charge recombination resistance

NU

and therefore an enhanced electron life time. Recombination of electrons dramatically slows down after doping as evident in PL studies and thus reduces the current loss enhancing the

MA

phenonmenen of electron transport. Recombination is an opposite flow that internally annihilates the moving carriers. Lower value of Cµ indicates lesser photogenerated electrons being stagnated by empty trap states in doped sample [13] and this is favorable for electron transport. The

D

electron life time obtained for the 4% Nd-TiO2 is longer than that of other two cells. A relatively

TE

high value of resistance to recombination and a lower value of chemical capacitance put together yields an excellent electron life time that contributes to the survival of recombination of charge

AC CE P

carriers thus keeping Voc at high levels.

Table 2 Resistance (R), capacitance (Cµ) and electron life time for the DSSCs Sample TiO2 4% Nd -TiO2 5% Nd - TiO2

Rs(ohm)

Rr(ohm)

Cµ (µF)

Ʈ (RrxCµ) (ms)

43.93

295.1

2.94

869.6

49.33

627

2.08

1304.1

57.85

358.3

2.43

872.4

15

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Fig. 8: EIS spectra of TiO2, 4% Nd-TiO2 and 5% Nd-TiO2, right image: equivalent circuit

MA

simulated to fit EIS of the DSSC. Evidently the energy conversion efficiency went up with an increase in dopant concentration as

D

seen from the I-V graph of Fig. 9, and this was attributed to an increase in J sc, FF and Voc. Only

TE

the results of good fits are shown in the Fig. 9. An energy conversion efficiency of 6.19 % was observed with more than 100% enhancement over the undoped cell. It is however presumed that

AC CE P

the overall efficiency of both the doped and undoped cells drastically enhances when the fabrication of solar cell includes a layer of TiO2 compact layer and encapsulation which is done for better charge transfer. The rise in Jsc was very nominal upto an extent of 10.6% for 4 % NdTiO2 and this is because of the steep decline in the surface area on doping. The adsorbed amount of dye was proportional to the dopant concentration, being maximum, for 4% Nd doped TiO2. The characteristic feature of lanthanides to plunder oxygen into titania by doping, thus causing an enhanced dye absorption and light harvesting surpasses the negative effect of the decrease in surface area. As mentioned earlier, the decrease in the band gap of the Nd doped TiO2 to visible range can cause the photoexcitation of the semiconductor, leading to an increased electron hole pairs. The electrons can be pumped directly from IB to CB or from VB to IB and then finally to CB, thus increasing photoelectron density in CB and hence J sc. The occurrence of photoexcitation of semiconductor together with enhanced photoinjection contributes sufficiently to enhance the electron density. Thus the above two factors, together with the scattering ability to reduce loss of photons improves Jsc. 5% doped Nd-TiO2 shows an unexpectedly high Jsc of 15.8 mA/cm2. The enhanced up shift of the conduction band in it or in other words the favorable 16

ACCEPTED MANUSCRIPT alignment of energy levels, resulting in better rate of electron injection and electron mobility together with the phenomenon of photoexcitation of the semiconductor leads to an improved electron density (which increases with dopant concentration), and hence high J sc. Thus the

PT

electrons generated from photoexcitation of semiconductor play a very prominent role together with the electrons generated from photosensitization of dye to increase the Jsc of doped cells. One

RI

DSSC fabricated without the dye sensitization was tested for its I-V characteristics to support the

SC

concept of photoexcitation occurring in these doped cells. The output was negligible implying the need of the dye to create the necessary driving force to push the photoexcited electrons of

NU

doped semiconductor towards the FTO. Thus it can be presumed that photoexcitation can contribute to Jsc only in the presence of photosensitization.

MA

A marginal difference is observed in the fill factor among the doped samples; except for 5% Nd doped TiO2 for which FF is very low. It is well known that fill factor is influenced by series resistance and built in voltage [34]. Since the built in voltage given by the difference

D

between the electrode work function and the electrolyte redox potential is increased owing to the

TE

upward shift of CB, the FF also increases. However the drastic lowering of FF in 5% Nd-TiO2 can be accounted for. The enhanced lattice defects induced by higher concentration of dopant

AC CE P

reduce the quality of the NPs [53] adversely affecting the charge transfer resistance with respect to Cµ. Hence both the packing defects [19] and electronic defects are necessary to be optimized to obtain a good FF.

The generation of Voc appears high for all the doped samples. The theoretical maximum achievable value of Voc for TiO2 based DSSCs is 0.95 V [56]. Dopant induction causes visible light absorption which is unlikely in undoped cells. The photoexcitation of both the semiconductor and the dye, on absorption of visible light leads to a competing effect of electron capture (donated by electrolyte) by the holes, created both in the dye and the semiconductor. The strong redox potential of the doped TiO2 induced by lanthanide induction can easily cause oxidation of its holes. This strongly survives the photoinjected electrons and photoexcited electrons from recombination. The most prominent and known factor to enhance Voc remains to be the suppressed recombination induced by the dopants, causing IELs or IBs, that can act as electron traps to trap electrons and detrap the same to the surface causing an efficient interfacial charge transfer [57, 58]. The concept of forbidden energy gap lapses in a nano metal oxide semiconductor, unlike 17

ACCEPTED MANUSCRIPT their bulk analogue and the electron transfer goes through a number of trap-detrap events occurring amongst the IBs [59-61]. Beyond the optimum concentration of 4%, probability of recombination of trapped charge carrier via quantum tunneling tend to be more due to lack of [30]

. On the other hand Ti3+ induces oxygen vacancy which can

PT

driving force to separate them

also act as electron traps to promote the charge separation [62-64]. But Ti3+ content increases

RI

with the increase in Nd3+ dosage. Beyond the optimal amount of 4%, both the lattice and

SC

electronic defects tend to hinder charge transfer leading to a decrease in Voc. Increase of Voc is also on account of its better negative flat-band potential and increased

NU

electron lifetime. Also, the local internal field (aroused because of the dipole moment of distorted and doped octahedra of titania) can influence charge separation [65]. The charge on the atoms due to the new Ti-O-Nd bond results in center of gravity of negative electric charges

MA

deviating from the position of Ti4+ ions in TiO6 octahedra, causing a dipole moment [65]. Such increase in internal field allows easy separation of charge carriers. As recombination occurs on

D

the oxide surface, the decline in surface area can also suppress the recombination [49]. All the

TE

above factors contribute significantly to overcome the drawback of recombination at the TiO2electrolyte or TiO2 - dye interface, thus facilitating Voc to reach almost nearer to theoretical

AC CE P

maximum of 0.95 eV.

Fig. 9: Current -Voltage curve of DSSCs made from different percentages of Nd - TiO2 18

ACCEPTED MANUSCRIPT Table 3: Photovoltaic parameters and quantity of the adsorbed dye for various DSSCs Jsc (mA/cm2)

Voc (V)

Fill

(Conversion

Adsorbed

(Best)

(Best)

Factor

efficiency, η%)

dye [molcm-2]

(%)

(Best)

PT

Sample

3.06

5.87x10-7

3.50

8.91x10-7

70.13

4.04

1.63x10-6

61.32

4.49

1.98x10-6

76.35

6.19

2.81x10-6

47.8

5.99

2.01x10-6

8.70

0.69

50.47

1% Nd-TiO2

7.30

0.81

59.19

2% Nd-TiO2

6.85

0.84

3% Nd-TiO2

8.59

0.85

4% Nd-TiO2

9.63

0.85

5% Nd-TiO2

15.8

0.78

MA

NU

SC

TiO2

D

4. Conclusion

RI

(Best)

The efficient, nanocrystalline Nd doped titania of different percentage were successfully

TE

prepared by solid sate technique (pulverization) and its properties were exploited as photo anode

AC CE P

material in dye sensitized solar cell. The existence of lanthanide induces oxygen vacancies and better dye absorption together with formation of IELs/IBs leading to a band gap narrowing, visible light absorption and photoexcitation of titania in DSSC. Favorable alignment of energy levels, resulting in better rate of electron injection and electron mobility, together with the phenomenon of photoexcitation of the semiconductor leads to an improved electron density (which increases with dopant concentration), and improved Jsc. Higher negative flat-band potential and lower recombination of electrons and holes enhances electron life time and thus improves Voc in doped samples. This study gains signi icance in utili ation o ‘the highly efficient lanthanide doped TiO2” as photo anode material in DSS s.

Acknowledgement The authors wish to acknowledge Ministry of New and Renewable Energy and Department of Science and Technology (Nanomission), India for funding. Authors wish to thank C.A.N Fernando, Nano-Technology Research Lab, Wayamba University of Sri Lanka for Mott– Schottky measurements

19

ACCEPTED MANUSCRIPT References

[1]

S. Ito, T. N. Murakami, P. Comte, P. Liska, C. Gra¨tzel, M. K. Nazeeruddin, M. Gratzel,

PT

Fabrication of thin film dye sensitized solar cells with solar to electric power conversion efficiency over 10%, Thin Solid Films, 516 (2008) 4613-4619.

M.Grätzel, Dye-sensitized solar cells. J. Photochem. Photobiol.C: Photochem. Rev, 4

RI

[2]

[3]

SC

(2003) 145-153.

T.Y. Lee, Alegaonkar P.S, J.B.Yoo, Fabrication of dye sensitized solar cell using TiO 2

[4]

NU

coated carbon nanotubes, Thin Solid Films, 515 (2007) 5131-5135. V. Amoli, R. Tiwari, A. Datta, A. Bhaumik, A.K. Sinha, Structurally stabilized

MA

organosilane-templated thermostable mesoporous titania, Chemphyschem, 15 (2014) 187-94 [5]

F. Huang, Q. Li, G. J. Thorogood, Y. B. Cheng, R. A. Caruso, Zn-doped TiO2 electrodes

D

in dye-sensitized solar cells for enhanced photocurrent. J. Materi. Chem, 22 (2012)

[6]

S. E.

oops, B.

TE

17128.

. O’ egan, P.

. F. Barnes, J. D. Durran, Parameters In luencing the

AC CE P

Efficiency of Electron injectionin Dye Sensitized solar cell, J. Am. Chem. Soc, 131 (2009) 4808-4818 [7]

Q. Wang, S. Ito, M. Gratzel, F.F Santiago, I. Mora-Sera, J. Bisquert, T.Bessho, H. Imai, Characteristics of high efficiency dye sensitized solar cells.J. Phys. Chem. B, 110 (2006) 25210-25221

[8]

S.M. Feldt, G. Wang, G. Boschloo, A. Hagfeldt, Effects of Driving Forces for Recombination and Regeneration on the Photovoltaic Performance of Dye-Sensitized Solar Cells using Cobalt Polypyridine Redox Couples, J. Phys. Chem. C, 115 (2011) 21500-21507.

[9]

H. Wang, M. Liu, C. Yan, J. Bell, Reduced electron recombination of dye-sensitized solar cells based on TiO2 spheres consisting of ultrathin nanosheets with [001] facet exposed, Beilstein J Nanotechnol, 3 (2012) 378-87.

[10]

F. Fabregat-Santiago, J. Bisquert, G. Garcia-Belmonte, G. Boschloo, A. Hagfeldt, Influence of electrolyte in transport and recombination in dye-sensitized solar cells studied by impedance spectroscopy, Sol. Energ. Mat. Sol. Cells, 87 (2005) 117-131. 20

ACCEPTED MANUSCRIPT [11]

X. Lü, X. Mou, J. Wu, D. Zhang, L. Zhang, F. Huang, F. Xu, S. Huang, ImprovedPerformance Dye-Sensitized Solar Cells Using Nb-Doped TiO2 Electrodes: Efficient Electron Injection and Transfer, Adv. Func. Mater, 20 (2010) 509-515. J. Liu, H. Yang, W. Tan, X. Zhou, Y. Lin, Photovoltaic performance improvement of

PT

[12]

dye-sensitized solar cells based on tantalum-doped TiO2 thin films. Electrochim. Acta, 56 Y. Duan, N. Fu, Q. Liu, Y. Fang, X. Zhoy, J. Zhang, Y. Lin, Sn-Doped TiO2 Photoanode

SC

[13]

RI

(2010) 396-400.

for Dye-Sensitized Solar Cells, J. Phys. Chem. C, 116 (2012) 8888-8893. R. Jose, V. Thavasi, S. Ramakrishna, Metal Oxides for Dye-Sensitized Solar Cells, J. Am. Ceram. Soc, 92 (2009) 289-301

C. Xiaobo, S.S. Mao, Titanium Dioxide Nanomaterials - Synthesis, Properties,

MA

[15]

NU

[14]

Modifications, and Applications, Chem. Rev, 107 (2007) 2891-2959. [16]

S.S. Mali, C. A. Betty, P. N. Bhonsle, P. S. Patil, C. K. Hong, From nanocorals to

D

nanorods to nanoflowers nanoarchitecture for efficient dye-sensitized solar cells at

[17]

TE

relatively low film thickness: All Hydrothermal Process, Sci Rep, 4 (2014) 5451. P.S. Archana, E. N. Kumar, C. Vijila, S. Ramakrishna, M. M. Yusoff, R. Jose, Random

AC CE P

nanowires of nickel doped TiO2 with high surface area and electron mobility for high efficiency dye-sensitized solar cells, Dalton Trans, 42 (2013) 1024-32. [18]

P.S. Archana, A. Gupta, M. M. Yusoff, R. Jose, Tungsten doped titanium dioxide nanowires for high efficiency dye-sensitized solar cells, Phys Chem Chem Phys, 16 (2014) 7448-54.

[19]

P.S. Archana, R. Jose,T. M. Jin, C. Vijila, M. M. Yusoff, S. Ramakrishna, Structural and Electrical Properties of Nb-Doped Anatase TiO2 Nanowires by Electrospinning, J. Am. Ceram. Soc, 93 (2010) 4096-4102.

[20]

Munir, S., et al., Effect of carrier concentration on the optical band gap of TiO2 nanoparticles. Materials & Design, 2016. 92: p. 64-72.

[21]

Y. Zhang, H. Zhang,Y. Xu, Y. Wang, Significant effect of lanthanide doping on the texture and properties of nanocrystalline mesoporous TiO2, J. Solid. State. Chem, 177 (2004) 3490 - 3498

21

ACCEPTED MANUSCRIPT [22]

W. Zhou, Y. Zheng, G. Wu, Novel luminescent RE/TiO2 (RE = Eu, Gd) catalysts prepared by in-situation sol–gel approach construction of multi-functional precursors and their photo or photocatalytic oxidation properties, Appl. Surf. Sci, 253 (2006) 1387-1392. R. Asahi, Y. Taga, W. Mannstadt, A. Freeman, Electronic and optical properties of

PT

[23]

anatase TiO2, Phys Revi. B, 61 (2000) 7459-7465.

Z. Zhao, Q. Liu, Effects of lanthanide doping on electronic structures and optical

RI

[24]

SC

properties of anatase TiO2 from density functional theory calculations, J. Phys. D: Appl.Phys, 41 (2008) 085417

Z. Tong, T. Peng, W. Sun, W. Liu, S. Guo, X.Z. Zhau, Introducing an Intermediate Band

NU

[25]

into Dye-Sensitized Solar Cells by W6+ Doping into TiO2 Nanocrystalline Photoanodes, J. [26]

MA

Phys.Chem. C, 118 (2014) 16892-16895.

J. Zhang, Z. Zhao, X. Wang, T. Yu, J. Guan,Z. Yu, Z. Li, Z. Zou, Increasing the oxygen vacancy density on the TiO2 surface by La-doping for Dye sensitized solar cell, J. Phys. Y. Wang, H. Cheng, Y. Hao, J. Ma, B. Xu, W. Li, S. Cai, The Photoelectrochemistry of

TE

[27]

D

chem. C, 114 (2010) 18396-18400.

Nd3+- sdoped TiO2 nanocrystelline electrodes, J. Mater. Sci. Lett, 18 (1999) 127-129. Y. Wang, H. Cheng, Y. Hao, J. Ma, W. Li, S. Cai, Photoelectrochemical propeties of

AC CE P

[28]

metal-ion -doepd TiO2 nanocrystelline electrodes, Thin Solid Films, 349 (1999) 120-125. [29]

W. Li, A. I. Frenkel, J. C. Woicik, C. Ni, S, I, Shah, Dopant location identification in Nd3+- doped TiO2 nanoparticles, Phys. Rev. B, 72 (2005) 155315

[30]

K.U. Minchitha, R. Geetha Balakrishna, Structural modification and property tailoring in titania for high efficiency in sunlight, Mater. Chem. Phys, 136 ( 2012) 20-728.

[31]

A. Burns,G. Hayes,W. Li, J. Hirvonen, J. D. Demaree, S. I. Shah, Neodymium ion dopant effects on the phase transformation in sol–gel derived titania nanostructures, Mater. Sci. Eng. B, 111 (2004) 150-155.

[32]

S. Hore, S. Hore, C.Vetter, R. Kerna, H. Smit, A. Hinsch, Influence of scattering layers on efficiency of dye-sensitized solar cells, Sol. Energ. Mater. Sol.Cells, 90 (2006) 11761188

[33]

F. Huang, D. Chen, X. L. Zhang, R. A. Caruso, Y. B. Cheng, Dual-Function Scattering Layer of Submicrometer-Sized Mesoporous TiO2 Beads for High-Efficiency DyeSensitized Solar Cells. Adv. Funct. Mater, 20 (2010) 1301-1305. 22

ACCEPTED MANUSCRIPT [34]

G. Kron, U. Rau, J. H. Werner, Influence of the Built-in Voltage on the Fill Factor of Dye-Sensitized Solar Cell, J.Phys. Chem. B, 107 (2003) 13258-13261.

[35]

S. Shogh, R. Mohammadpour, A. I, Zad, N. Thaghavinia, Improved photovoltaic

PT

performance of nanostructured solar cells by neodymium-doped TiO2 photoelectrode. Mater. Lett, 159 (2015) 273-275.

Q. Yao, J. Liu, Q. Peng, X. Wang, Y. Li, Nd-doped TiO2 nanorods: preparation and

RI

[36]

[37]

SC

application in dye-sensitized solar cells. Chem. Asian. J, 5 (2006) 737-41. S. Swetha, M. K.Singh, K. U. Minchitha, R. Geetha Balakrishna, Elucidation of cell

NU

killing mechanism by comparative analysis of photoreactions on different types of bacteria, Photochem. Photobiol, 88 (2012) 414-22. D’Sou a, L. P, S. Shree, and G.R. Balakrishna, Bifunctional Titania Float for Metal Ion

MA

[38]

Reduction and Organics Degradation, via Sunlight, Ind. Eng. Chem. Res, 52 (2013) 16162-16168.

G. R. Balakrishna, G. Devi, A study of photocatalytic oxidation of indanthrene red LGG

D

[39]

[40]

TE

an anthraquinone vat dye on TiO2, Pol. J. Chem, 79 (2005) 919-939. F. Rivera-López, M. Pérez, X-ray photoelectron spectroscopy: surface and depth

AC CE P

profiling studies of glasses doped with Nd and Yb ions. Surf, Interface Anal, 44 (2012) 927-930. [41]

D. Nasoko, Y. Li, J. Li, X. Li, Y. Yu, Neodymium doped TiO2 with Anatase and Brookite Two Phases: Mechanism for Photocatalytic Activity Enhancement Under Visible

Light

and

the

Role

of

Electron,

Int.

J.

Photoenergy,

http://dx.doi.org/10.1155/2012/716087 [42]

C.W. Kim, Y. H. Kim, U. Pal, Y. S. Kang, Facile synthesis and magnetic phase transformation of Nd–Fe–B nanoclusters by oxygen bridging, J. Mater. Chem. C, 1 (2013) 275-281.

[43]

K. Jayanthi, S.V. Manorama, S. Chawla, Observation of Nd3+visible line emission in ZnO : Nd3+prepared by a controlled reaction in the solid state. J. Phys. D: Appl Phys, 46 (2013) 325101.

[44]

Y. Li D. S. Hwang, N. H. Lee, S. J. Kim, Synthesis and characterization of carbon-doped titania as an artificial solar light sensitive photocatalyst, Chem. Phys. Lett, 404 (2005) 25-29. 23

ACCEPTED MANUSCRIPT [45]

H. H. Lo, N.O. Gopal, S. C. Ke, Origin of photoactivity of oxygen-deficient TiO2 under visible light, Appl. Phys. Lett, 95 (2009) 083126.

[46]

S. H. Lee, E. Yamasue, K. N. Ishihara, H. Okumura, Photocatalysis and surface doping

PT

states of N-doped TiOx films prepared by reactive sputtering with dry air, Appl. Catal. B: Environ, 93 (2010) 217-226.

A.K. Rumaiz, B. Ali, A. Ceylan, M. Boggs, T. Beebe, S. I. Shah, Experimental studies on

RI

[47]

SC

vacancy induced ferromagnetism in undoped TiO2. Solid State Communications, 144 (2007) 334-338

R. Shwetharani, M. S. Jyothi, P. D. Laveena, R. G. Balakrishna, Photoactive titania float

NU

[48]

for disinfection of water; evaluation of cell damage by bioanalytical techniques,

[49]

MA

Photochem Photobiol, 90 (2014) 1099-107.

K. Park, J. Xi, Q. Zhang, G. Cao, Charge Transport Properties of ZnO Nanorod Aggregate Photoelectrodes for DSCs. J. Physi. Chem. C, 115 (2011) 20992-20999. V. I. Klimov Semiconductor and metal nano crystals. Syntheisis, electronic and optical

D

[50]

[51]

TE

properties, 1st edition, CRS press, Tailor and Francis group, 2003. R. Shwetharani, C.A.N. Fernando, G.R. Balakrishna, Excellent hydrogen evolution by a

39130 [52]

AC CE P

multi approach via structure–property tailoring of titania, RSC Adv, 5 (2015) 39122-

J. Yan, G.Wu, N. Guan, L. Li, Z. Li, X. Cao, Understanding the effect of surface/bulk defects on the photocatalytic activity of TiO2: anatase versus rutile, Phys Chem Chem Phys, 15 (2013) 10978-88.

[53]

X. Zhang, F. Liu, Q. L. Huang, G. Zhou, Z. S. Wang, Dye-Sensitized W-Doped TiO2 Solar Cells with a Tunable Conduction Band and Suppressed Charge Recombination, J. Physi. Chem. C, 115 (2011) 12665-12671.

[54]

P. S. Archana, A. Gupta, M. M. Yusoff, R. Jose, Charge transport in zirconium doped anatase nanowires dye-sensitized solar cells: Trade-off between lattice strain and photovoltaic parameters, Appl. Phys. Lett, 105 (2014) 153901.

[55]

H. Zhang, K. Cheng, Y. M. Hou, Z. Fang, Z. X. Pan, W. J. Wu, J. L. Hua, X. H. Zhong, Efficient CdSe quantum dot-sensitized solar cells prepared by a postsynthesis assembly approach, Chem. Commun, 48 (2012) 11235-7.

[56]

H. Zhang, K. Cheng, Y. M. Hou, Z. Fang, Z. X. Pan, W. J. Wu, J. L. Hua, X. H. Zhong, 24

ACCEPTED MANUSCRIPT Tantalum-doped titanium dioxide nanowire arrays for dye-sensitized solar cells with high open circuit voltage, Angew. Chem, 48 (2009) 8095-8098. [57]

J. Orenstein, M. Kastner, Photocurrent Transient Spectroscopy: Measurement of the

[58]

PT

Density of Localized States ina-As2Se3, Phys. Rev. Lett, 46 (1981) 1421-1424. H. J. Snaith, L. Schmidt-Mende, Advances in Liquid - Electrolyte and Solid-State Dye-

J. Bisquert, Physical electrochemistry of nanostructured devices, Phys Chem Chem Phys,

SC

[59]

RI

Sensitized Solar Cells, Adv.Mater, 19 (2007) 3187-3200.

10 (2008). 49-72.

B. O'Regan, M. Grätzel, A low-cost, high-efficiency solar cell based on dye-sensitized

NU

[60]

colloidal TiO2 films, Nature, 353 (1991) 737

J. Bisquert, F. F. Santiago, I. Mora-Sera, G. Gracia-Belmonte, E. M. Barea, E. Palomares,

MA

[61]

A review of recent results on electrochemical determination of the density of electronic states of nanostructured metal-oxide semiconductors and organic hole conductors, Inorg.

J. Peter, H.Gabriele, W. Thomas, K. Wieghardt B. Nuber, Synthesis and Characterization

TE

[62]

D

Chim. Acta, 361 (2008) 684-698.

of Mononuclear Octahedral Titanium (VI) complexes containing Ti =O, Ti (O2) and Ti [63]

AC CE P

(OCH3)x (x=1-3) structural units, Inorg. Chem, 33 (1994) 2462-2471. I. Nakamura, N. Negishi, S. Kutsuna, T. Ihara, S. Sugihare ,K. Takeuchi, Role of oxygen vacancy in the plasma treated TiO2 photocatalyst with visible light activity for NO removal, J. Mol. Catal. A, 161 (2000) 205-212. [64]

M. Anpo, The design and development of highly reactive titanium oxide photocatalysts operating under visible light irradiation, J. Catal, 216 (2003) 505-516.

[65]

J. Sato, H. Kobayashi, Y. Inosue, Photocatalytic Activity for Water Decomposition of Indates with Octahedrally Coordinated d10 configuration. II. Roles of geometric and electronic structures, J.Physi. Chem. B, 107 (2003) 7970-7975.

25

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

Graphical abstract

26

ACCEPTED MANUSCRIPT Highlights 

Physical process of pulverization has been attempted to cause surface level doping. Doping of a lanthanide facilitates the formation of impurity levels, induces

PT



oxygen vacancies and band gap narrowing leading to visible light absorption in

The photoexcitation of the doped semiconductor and the dye leads to a competing

SC



RI

DSSC

effect of electron capture by the holes, avoiding recombination.

TE

D

MA

further enhances Jsc and Voc.

NU

Higher negative flat-band potential and favorable alignment of energy levels

AC CE P



27