characteristics of sno2 nanofiber tio2 nanoparticle composite for dye sensitized solar cells

Sumathy, Qufu Wei, and Qiquan Qiao Citation: AIP Advances 5, 067134 2015; doi: 10.1063/1.4922626 View online: http://dx.doi.org/10.1063/1.4922626 View Table of Contents: http://scitation

solar cells

Jiawei Gong, Hui Qiao, Sudhan Sigdel, Hytham Elbohy, Nirmal Adhikari, Zhengping Zhou, K Sumathy, Qufu Wei, and Qiquan Qiao

Citation: AIP Advances 5, 067134 (2015); doi: 10.1063/1.4922626

View online: http://dx.doi.org/10.1063/1.4922626

View Table of Contents: http://scitation.aip.org/content/aip/journal/adva/5/6?ver=pdfcov

Published by the AIP Publishing

Articles you may be interested in

Enhanced conversion efficiency of dye-sensitized solar cells using a CNT-incorporated TiO2 slurry-based photoanode

AIP Advances 5, 027118 (2015); 10.1063/1.4908179

Transient photocurrent and photovoltage studies on charge transport in dye sensitized solar cells made from the composites of TiO 2 nanofibers and nanoparticles

Appl Phys Lett 98, 082114 (2011); 10.1063/1.3560057

The effect of annealing on the photoconductivity of carbon nanofiber/ TiO 2 core-shell nanowires for use in dye-sensitized solar cells

Appl Phys Lett 97, 043102 (2010); 10.1063/1.3464965

Electron transport in electrospun TiO 2 nanofiber dye-sensitized solar cells

Appl Phys Lett 95, 012101 (2009); 10.1063/1.3167298

Fabrication of dye-sensitized solar cells with Ti O 2 photoelectrode prepared by sol-gel technique with low annealing temperature

J Vac Sci Technol A 27, 1042 (2009); 10.1116/1.3116588

AIP ADVANCES 5, 067134 (2015)

composite for dye-sensitized solar cells

Jiawei Gong,1,2Hui Qiao,3, aSudhan Sigdel,1Hytham Elbohy,1

Nirmal Adhikari,1Zhengping Zhou,1K Sumathy,2Qufu Wei,3, b

and Qiquan Qiao1, c

1Center for Advanced Photovoltaics, Department of Electrical Engineering,

South Dakota State University, Brookings, SD 57007, USA

2Department of Mechanical Engineering, North Dakota State University,

Fargo, ND 58102, USA

3Key Laboratory of Eco-Textiles, Jiangnan University, Ministry of Education,

Wuxi 214122, China

(Received 1 March 2015; accepted 26 May 2015; published online 18 June 2015)

SnO2nanofibers and their composites based photoanodes were fabricated and

inves-tigated in the application of dye-sensitized solar cells The photoanode made of

SnO2/TiO2 composites yielded an over 2-fold improvement in overall conversion

efficiency The microstructure of SnO2nanofibers was characterized by X-ray

diffrac-tion (XRD) and transmission electron microscopy (TEM) A compact morphology of

composites was observed using scanning electron microscopy (SEM) A long charge

diffusion length (62.42 µm) in the composites was derived from time constant in

tran-sient photovoltage and photocurrent analysis These experimental results demonstrate

that one-dimensional nanostructured SnO2/TiO2 composites have a great potential

for application in solar cells C 2015 Author(s) All article content, except where

otherwise noted, is licensed under a Creative Commons Attribution 3.0 Unported

License.[http://dx.doi.org/10.1063/1.4922626]

I INTRODUCTION

Dye-sensitized solar cells (DSSCs) have attracted tremendous research interest during the past several decades and their efficiency has recently been raised up to ∼ 15%.1 3In the operation of DSSCs, wide band gap semiconductor photoanodes (typically TiO2) play a crucial role in dye sensi-tizer uptake and electron transport Although TiO2nanoparticle based photoanodes provide large surface area for sufficient dye loading, numerous boundary defects at contacts between nanoparti-cles lead to the scattering of free electrons and reduced charge mobility

Tin dioxide (SnO2) is a promising alternative to TiO2 due to its large band gap (3.6 eV), high electron mobility (100 − 200 cm2V− 1S− 1), and low conduction band effective mass (0.1mo).4A large band gap can promote device stability by suppressing the generation of oxidative holes under ultraviolet light illumination.5The oxidative holes can decompose organic compounds such as Ru dyes adsorbed on oxide surface In addition, the holes can oxidize I− to I3−; and any holes that oxidize the electrolyte irreversibly rather than regenerating I3−will lead to loss of I3−and decreasing

of DSSC performance A high electron mobility and low conduction band effective mass indicate fast electron transport, which contributes to long diffusion length and efficient charge collection

A considerable amount of effort has been devoted to develop nanostructured SnO2 with diverse morphologies A variety of nanostructured SnO2 such as nanosheets,6 nanofibers,7 nanoflowers,8

hollow mircrospheres,9and urchin-like structures10have been synthesized and widely investigated

in the appplication of DSSCs However, the application of SnO2-based DSSCs has been limited due

a huiqiao@jiangnan.edu.cn

b qfwei@jiangnan.edu.cn

c qiquan.qiao@sdstate.edu ; Tel.: 605-688-6965; Fax: 605-688-4401

2158-3226/2015/5(6)/067134/10 5, 067134-1 © Author(s) 2015

to poor photovoltaic performance caused by fast interfacial electron recombination and insufficient attachment of dye molecules Fast recombination dynamics for SnO2electrode results from a lower trap density and a 300 mV positive shift of the SnO2conduction band.11The insufficient dye attach-ment is a result of less surface area and lower isoelectric point (pH 4-5), which limits the attachattach-ment

of dye molecules with acidic carboxyl groups (e.g N719 dye, (Bu4N)2[Ru(Hdcbpy)2(NCS)2])

To minimize the interfacial charge recombination and increase dye upload, hybrid TiO2-SnO2 structures such as TiO2-coated SnO2hollow microspheres,12TiO2-coated SnO2nanosheets,13 and TiO2@SnO2 core/shell nanoparticles14 have been previously attempted by others Our previous work demonstrated that incorporation of TiO2 nanoparticles into SnO2 nanofibers could signif-icantly improve the device efficiency The excellent performance was expected due to allevi-ated band shift effect and increased dye loading In this study, detailed characterization was per-formed to analyze the microstructure, transient response and recombination dynamics of SnO2

naonofiber/TiO2nanoparticle composites In addition, the nanorod/nanowire geometry can enhance the photogeneration process by inducing strong light scattering /trapping when the diameter of nanofibers is larger 200 nm.15 Thus, SnO2nanofibers with a diameter 200 - 300 nm was used to induce light scattering/harvesting, and therefore enhance the photogeneration process by scatter-ing/trapping light in the photoanode

II EXPERIMENTAL DETAILS

A Synthesis and characterization of SnO 2 nanofibers

SnO2 nanofibers were prepared by electrospinning and calcination from polyvinyl pyrroli-done/stannic chloride pentahydrate (PVP/SnCl4·5H2O) precursors Specifically, the electrospinning solution was prepared by adding SnCl4·5H2O into 10 wt% PVP in an ethanol/DMF mixed solvent (weight ratio 1:1) The weight ratio of SnCl4·5H2O to polymer intermediate (PVP) was fixed at 1:1 The solution was stirred by a magnetic bar at room temperature Subsequently, electrospinning was carried out with this solution The PVP/SnCl4·5H2O precursor was ejected from a stainless steel needle under a high voltage of 17 kV to form fibrous nonwoven mats on the collector The flow rate was kept at 1.0 ml/h, and the needle-to-collector distance was fixed at 21 cm The elec-trospun nanofiber mats were calcinated at 500◦C for 4 h with a heating rate of 0.5◦C/min The structural analysis of SnO2nanofibers was performed on a D8 Advance X-ray diffractometer (XRD, Bruker AXS, Germany) over the 2θ range from 10◦to 80◦ The morphology of SnO2nanofibers was observed by scanning electron microscope (SEM; Quanta-200, Netherlands) and transmission electron microscope (TEM) with selected-area electron diffraction (SAED) (TEM; JEM-2100HR, JEOL) For TEM measurements, precursor nanofibers were directly deposited on the copper grids during electronspinning, and SnO2 nanofibers were dispersed in ethanol by ultrasound and then transferred onto copper grids.16

B Fabrication of dye-sensitized solar cells

To prepare SnO2/TiO2composite based photoanodes, SnO2nanofibers and TiO2nanoparticles (P25, Degussa) were mixed at an optimized weight ratio (1:1) with ethyl cellulose, α-terpineol and ethanol to form a paste through successive sonication and stirring Specifically, 0.125 g SnO2 and 0.125 g TiO2 were added to 4.9 ml ethanol and left for dispersion The solution was sonicated and mechanically stirred alternatively at 1-hour intervals for a total of 4 hours Subse-quently, 0.125 g ethyl cellulose as binder and 0.89 ml α-terpineol as solvent were added to the mixture for consecutive sonication and stirring until all grains disappeared and the solution became homogeneous The solution was heated in a vacuum oven at 80 ◦C to remove excess ethanol until it turned into a form of homogeneous slurry ready for doctor-blading A thin layer of the slurry was doctor-bladed onto fluorine doped tin oxide (FTO) glass substrate with an active area of 0.16 cm2 An optimal thickness (8-9 µm) was confirmed by a Dektak profilometer.17 This sample was annealed at 500◦C for 30 min to form a mesoporous film On the top of the film, a scattering layer (Solaronix Ti-Nanoxide R/SP) was coated that enhanced light absorption of the underneath

067134-3 Gong et al. AIP Advances 5, 067134 (2015)

mesoporous active layer Then the sample was dipped in 40 mM TiCl4 solution at 80◦C for 30 min, followed by sintering at 500◦C to form a TiO2blocking layer This layer can passivate the 3D interpenetrated nanofiber/nanoparticle network, and therefore can improve the electron transport This TiO2blocking layer effectively prevents electron recombination This thin blocking layer was coated on all the samples using the same methods in order to exclude its effects on the perfor-mance enhancement for different samples including TiO2nanoparticle sample The thickness of this blocking layer was typically tens of nm, which is quite smooth

The resulting photoanodes were immersed in dye solution containing 0.5 mM Ruthenizer 535-bisTBA dye (Dyesol N-719) in acetonitrile/t-butanol (volume ratio: 1:1) for 24 hours In the final step, any excess dye molecules on the photoanode were rinsed in acetonitrile for several hours Counter electrode was prepared by sputtering 40 nm Pt onto FTO glass substrates In the end, the photoanode and counter electrode were sandwiched and sealed with 60 µm thick plastic and injected with I−/I3−electrolyte through reversed channels All devices were fabricated in the same procedure.18

C Dye desorption from photoanodes

10 mM NaOH solution in ethanol and DI water (volume raio 1:1) was used to desorb dyes from SnO2nanofiber, and SnO2/TiO2composite Each of the dye attached photoanodes were dipped

FIG 1 X-Ray powder diffraction pattern of SnO2nanofibers and (b) transmission electron microscopy image of a single SnO2nanofiber.

FIG 2 Scanning electron micrographs of (a) electrospun nanofiber SnO 2 network calcinated at 500 ◦ C, (b) Short SnO 2

nanofibers randomly packed on FTO glass substrate, and (c) SnO 2 nanofiber/TiO 2 nanoparticle composite (weight ratio: 1:1).

in the 10 mM NaOH solution for 24 h at room temperature to desorb dye molecules The dye molecules were peeled off from the photoanode into solution by neutralizaiton reaction between acidic carboxylic group and basic solution.19The volume of each solution was kept as 2 mL for dye desorption Spectra of the dye molecules desorbed from different photoanodes were measured

III RESULTS AND DISCUSSION

The XRD pattern shown in Fig.1(a) indicates a high purity of SnO2 nanofibers annealed in air at 500◦C Peaks with 2θ values of 26.48, 33.87, 37.91, 51.72, 54.85, and 57.97 were observed,

067134-5 Gong et al. AIP Advances 5, 067134 (2015)

which correspond to SnO2crystal planes of (110), (101), (200), (211), (220), and (002), respec-tively No obvious impurity peaks (e.g., unreacted Sn metal and other tin oxides) were observed, indicating the high purity of the rutile SnO2nanofibers These signature diffraction peaks indicate a tetragonal rutile structure of SnO2with lattice constants of a, b = 4.74 Å and c = 3.18 Å that agree well with documented values for the SnO2crystals (JCPDS card, No 41-1445) Scherrer’s equation was adopted to estimate the size of SnO2crystals in the form of powder It is stated that the average crystallite size D = 0.89λ/β cos θ, where λ is the wavelength for the Cu Kα(= 1.54056 Å), β is the line broadening at half the maximum intensity (FWHM) expressed in radian, and θ is Bragg’s angle The average crystallite size was calculated to be c.a 6 nm for SnO2nanofibers based on the (211) peak.20 , 21 The TEM image shown in Fig.1(b)reveals a structure of single SnO2nanofiber which retains an intact and uniform fibrous morphology

Figure2(a)shows scanning electron micrographs (SEM) of the as-spun PVP/SnCl4·5H2O pre-cursor nanofibers It is seen that SnO2 nanofibers exhibit fibrous morphology, good rigidity and are separated from each other Due to weak adhesion between the original SnO2nanofiber sheets and FTO substrate, the nanofiber sheets were dispersed into short fibers by sonication The short fibers were made into a paste from using the procudure discribed in Section II The paste was doctor-bladed onto FTO glass and anealed to form photoanode, which has the topology shown in Fig.2(b) Due to the fact that continuous SnO2naonfibers were crushed into randomly packed short fragments, significant amount of voids were formed in the film which reduced the surface area of photoanode To minimize the number of voids and increase the surface area, closely-packed TiO2

nanoparticles were introduced into SnO2nanofibers to form a compact film morphology as shown in Fig.2(c)

Three samples of each type of devices were tested under AM 1.5 illumination at a light intensity of 100 mW/cm2 Figure3 shows the comparison of J-V characteristics of DSSCs based

on SnO2 nanofiber, and SnO2/TiO2composite The photovoltaic parameters are listed in TableI The SnO2 nanofiber based device shows poor performance with open-circuit voltage (Voc) of 0.7 V, short-circuit current density (Jsc) of 5.9 mAcm− 2, and an overall conversion efficiency (η)

FIG 3 Comparison of current density versus voltage (J-V) curves of SnO 2 nanofiber, and SnO 2 /TiO 2 composite based DSSCs.

TABLE I PHOTOVOLTAIC PARAMETERS OF DSSCs BASED ON SnO2NANOFIBER, AND SnO2/TiO 2 COMPOSITE PHOTOANODE.

SnO2/TiO 2 composite 0.79±0.04 10.1±0.07 57±0.3 4.54±0.1 62.42±1.25

FIG 4 (a) UV-Vis absorbance spectra from the solutions of dyes that were desorbed from SnO2nanofiber, and SnO2/TiO 2

composite based photoanodes (b) Transmittance spectra before dye soaking.

of 1.68% A relatively low open circuit voltage (0.7 V) was expected because of a more positive conduction-band edge of SnO2with respect to nanocrystalline TiO2.15 , 22 , 23The low current density can be mainly ascribed to insufficient dye attachment This was confirmed by UV-Vis absorbance spectra, shown in Fig.4(a), that dye solution derived from SnO2nanofiber photoanode has a lower absorbance, which implies the least amount of dye molecules attached onto photoanodes in DSSCs

A comparatively low fill factor (FF) of 0.41 was observed, which was caused by high charge resistance in porous SnO2film due to randomly packed short fiberous morphology For each device,

067134-7 Gong et al. AIP Advances 5, 067134 (2015)

FIG 5 (a) Nyquist plots of SnO 2 nanofiber, and SnO 2 /TiO 2 composite based DSSCs in dark at bias of V oc from 0.1 Hz to

100 Hz with an amplitude of 10 mV, and (b) Equivalent circuit model for full cell.

TABLE II FITTED VALUE Of R S , R CT , AND R CR FOR SnO 2 NANOFIBER AND TiO 2 NANOPARTICLE BASED DSSCs.

charge diffusion length (L) was calculated based on transient photovoltage and photocurrent anlysis, shown in Fig 6 A long diffusion length (62.42 µm) of SnO2/TiO2composites indicates a small magnitude of recombination dynamics

In contrast, SnO2/TiO2 composite showed an over 2-fold improvement compared to SnO2

nanofiber with a Vocof 0.79 V, a Jsc of 10.1 mAcm−2, a FF of 0.57, and an overall efficiency η

of 4.54% The increase in Voc from 0.7 V to 0.79 V can be attributed to the alleviated shift of conduction band of SnO2nanofiber by incorporating TiO2nanoparticles The improvement in short circuit current density can be ascribed to the increase in dye attachment as shown in Fig.4(a) The enhanced fill factor is a result of low electron recombination rate due to the formation of TiO2 blocking layer, which is confirmed by transient analysis and electrochemical impedance spectros-copy (EIS).24

The UV-Vis absorbance spectrum of each dye solution is shown in Figure 4(a) It can be seen that SnO2/TiO2composites absorbed ∼30% more of dye molecules than the SnO2nanofibers, whereas it generated over 2-fold conversion efficiency Such significant improvement is a result

of increased surface area and more compact morphology introduced by TiO2nanoparticles It has

FIG 6 Normalized transient (a) photovoltage and (b) photocurrent decay of SnO2nanofiber, SnO2/TiO 2 composite based DSSCs.

been found that when the diameters of nanofibers increased to 200 nm or beyond, the light scat-tering becomes substantially stronger.15 Since the diameters of the electrospun SnO2 nanofibers were ca 200–300 nm, these nanofibers were expected to induce strong light scattering and thereby significantly enhance the light harvesting This prediction was confirmed by comparing transimit-tance spectra of SnO2 nanofiber, and SnO2/TiO2composite based photoanodes in Fig.4(b) The SnO2 nanofiber based cells showed a broad light absorption property from 400 nm to 800 nm with a minimum tranmittance of 14% at green light (530 nm) This light harvesting capability can be attributed to multiple light scattering in large SnO2agglomerates scattered in the electrode SnO2/TiO2 composite based photoanode has a higher transmittance compared to SnO2nanofiber photoanode

To further understand the interfacial charge transfer, electrochemical impedance spectroscopy (EIS) analysis was carried out Figure5(a)shows EIS spectra of SnO2 nanofiber, and SnO2/TiO2 composite based DSSCs Two semicircles were clearly observed: one on the left side (high fre-quency region) that represents the charge transfer at the electrolyte/counter electrode interface; and the other on the right (low frequency region) that represents the back charge transfer from the photoanode to electrolyte.25 , 26 Equivalent circuit used to analyze the EIS spectra is presented in Fig.5(b) Each of the two interfaces was modeled by a parallel combination of a resistance and a capacitor RSrepresents the total series resistance of a device, RCTis the charge transfer resistance

at the electrolyte/counter electrode interface, and RCR is the charge recombination resistance at

067134-9 Gong et al. AIP Advances 5, 067134 (2015)

the photoanode/electrolyte interface Values of RS, RCTand RCRare extracted by fitting equivalent circuit and presented in Table II A large value of Rs (20.58 Ω) in SnO2 nanofiber compared

to SnO2/TiO2 composite (17.45 Ω) indicates a slower charge transport in SnO2 due to a slow interparticle electron motion in the porous SnO2 film.27 This high total series resistance was in accordance with the low fill factor in SnO2nanofiber based devices The RCTvalues in all devices were consistent as the electrolyte and counter electrode were kept the same A significantly large charge recombination resistance (303.7 Ω) of SnO2nanofibers was observed This implies a slow recombination on the SnO2nanofibers, and a fast charge transport and collection in SnO2nanofiber based photoanode

Transient photovoltage is one of the major characterization techniques to probe recombination dynamics in solar cells Figure 6(a) shows the normalized transient photovoltage decay of SnO2 nanofiber, and SnO2/TiO2 composite Exponential decay of transient photovoltage was fitted in Origin® to derive the value of time constant (τe) The competition between recombination and charge transport determines the diffusion length (L) of electrons, which measures the average dis-tance that the electrons can travel in the photoanode without recombination Diffusion length, L, is calculated using the equation:

L=Dnτe=



w2τe

2.35τtrans

(1)

where w is the thickness of photoanode, τe is the electron recombination lifetime, τtrans is the electron transport lifetime, and Dn is the effective diffusion coefficient.28Electron recombination time of SnO2/TiO2composite is comparable to SnO2nanoiber (21.8 ms); and both showed a long recombination time The long transient photovoltage decay in the composite is attributed to slow recombination dynamics due to TiCl4 post treatment It has been found that an energy barrier of aproximately 300 mV created by TiO2layer prevents back charge transfer from SnO2to the electro-lyte or dye.10It can be seen, in Figure6(b), that incorprating TiO2nanoparticles in SnO2nanofibers reduces the decay time constant τ from 0.2 ms to 0.15 ms, which indicates an improvement in charge transport and collection A relatively slow charge collection in SnO2nanofiber based devices can be attributed to lower bulk electron mobility that is caused by boundaries and defects of SnO2

nanofibers in the film

IV CONCLUSIONS

In summary, incorporation of TiO2 nanoparticles into SnO2nanofibers shows an over 2-fold efficiency improvement over SnO2nanofibers as photoanode in DSSCs Such improved cell perfor-mance can be attributed to a compact morpholgy of SnO2/TiO2composites, a large surface area introduced by the TiO2nanoparicles, and a reduced interfacial charge recombination resulting from TiCl4treatment In addition, a thin film (8-9 µm) photoanode would mitigate charge recombination and reduce series resistance, which is desirable in cell efficiency

ACKNOWLEDGEMENT

This work was partially supported by US-Pakistan joint Science and Technology through Na-tional Academy of Science, NaNa-tional Natural Science Foundation of China (21201083), and Coop-erative Innovation Fund-Prospective Project of Jiangsu Province (29 and BY2014023-23)

1 J Burschka, N Pellet, S.-J Moon, R Humphry-Baker, P Gao, M K Nazeeruddin, and M Graetzel, Nature 499, 316 (2013).

2 J Gong, K Sumathy, and J Liang, Renewable Energy 39, 419 (2012).

3 J Gong, K Sumathy, and J Liang, International Journal of Sustainable Energy , 1 (2014).

4 Q Wali, A Fakharuddin, I Ahmed, M H Ab Rahim, J Ismail, and R Jose, Journal of Materials Chemistry A 2, 17427 (2014).

5 A Kay and M Gratzel, Chemistry of Materials 14, 2930 (2002).

6 F S Cai, Z H Yuan, Y Q Duan, and L J Bie, Thin Solid Films 519, 5645 (2011).