Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/nanoenergy

Nano Energy (2015) 11, 728–735

http://dx.doi.org/12211-2855/& 2014 E

nCorresponding au637553 Singapore, S

E-mail address: l1Equal contributio

RAPID COMMUNICATION

TiO2 nanotube arrays based flexibleperovskite solar cells with transparent carbonnanotube electrode

Xiaoyan Wanga,1, Zhen Lia,1, Wenjing Xuc,Sneha A. Kulkarnia, Sudip K. Batabyala, Sam Zhangd,Anyuan Caoc, Lydia Helena Wonga,b,n

aEnergy Research Institute @ NTU (ERI@N), Nanyang Technological University, Techno Plaza,50 Nanyang Drive, 637553 Singapore, SingaporebSchool of Materials Science and Engineering, Nanyang Technological University (NTU), Block N4.1,Nanyang Avenue, 639798 Singapore, SingaporecDepartment of Materials Science and Engineering, College of Engineering, Peking University,Beijing 100871, People’s Republic of ChinadSchool of Mechanical and Aerospace Engineering, Nanyang Technological University, 50 Nanyang Avenue,Singapore 639798, Singapore

Received 4 September 2014; received in revised form 5 November 2014; accepted 25 November 2014Available online 9 December 2014

KEYWORDSPerovskite solar cell;Flexible;Titanium foil;TiO2 nanotubes;Carbon nanotubes;Anodization

0.1016/j.nanoen.2lsevier Ltd. All rig

thor at: Energy Reingapore.ydiawong@ntu.edn to the work.

AbstractA solid-state, flexible solar cell based on titanium (Ti) foil/TiO2 nanotubes (TNTs) with organic–inorganic halide perovskite absorber and transparent carbon nanotube electrode is demon-strated. TNT arrays together with an inherent blocking layer were simultaneously formed on Tifoil during one-step anodization. TNTarrays serve as deposition scaffold and electron conductorfor perovskite absorber. Transparent conductive carbon nanotube network is laminated on topof perovskite and serves as hole collector as well as transparent electrode for light illumination.Under AM 1.5, 100 mW cm�2 illumination, power conversion efficiency of 8.31% has beenachieved, which is among the highest for TiO2 nanotube based flexible solar cells. Interestingly,up to 100 mechanical bending cycles show little deterioration to the device performance,demonstrating good flexibility of the Ti foil based perovskite solar cells. The Ti foil based

014.11.042hts reserved.

search Institute @ NTU (ERI@N), Nanyang Technological University, Techno Plaza, 50 Nanyang Drive,

u.sg (L.H. Wong).

729TiO2 nanotube arrays based flexible perovskite solar cells with transparent carbon nanotube electrode

solid-state, flexible perovskite solar cells have great potential for applications in buildingphotovoltaics and wearable electronic devices.& 2014 Elsevier Ltd. All rights reserved.

Introduction

As a new member of the next generation photovoltaicmaterials, organometal halide perovskite (e.g., CH3NH3PbI3,CH3NH3PbI3�xClx) was first demonstrated for efficient solarcells in 2009 [1], and soon become the most importantcandidate to replace silicon, with low material cost and highefficiency. Owing to the high light absorption coefficient(1.5� 104 cm�1 at 550 nm) [2] and long electron–hole diffu-sion length (�100 nm for CH3NH3PbI3 [3,4] and �1 μm forCH3NH3PbI3�xClx [4]), perovskite solar cells have achieved astunning success in a very short period in terms of efficiency,i.e., from 4% in 2009 [1] to 10% in 2012 [5,6], then 15% in 2013[7] and 19.3% at present [8].

The rapid and significant progress in perovskite solar celltriggers vigorous research interest, as well as commercializationefforts. Flexibility is one of the development directions, whichcan be beneficial for both production (with roll-to-roll appr-oaches) and applications (easy installation on buildings andintegration on wearable devices). Flexible perovskite solar cellshave been reported by a few groups since end of 2013 [9–14].Polyethylene terephthalate (PET) coated by conductive indiumtin oxide (ITO) was adopted as the flexible substrates [9–13].PCBM and PEDOT:PSS were used as electron and hole transportlayers, yielding efficiencies of 4.5% �9.2% [9–12]. In addition,low-temperature processed ZnO layer has also been used aselectron transport layer in flexible perovskite solar cells on PETsubstrate [13]. The efficiency limitation of the flexible per-ovskite solar cells on PET substrates may lie in the high seriesresistance of ITO on PET substrates. In contrast, conductivemetal foils possess better conductivity and mechanical robust-ness compared to PETsubstrates. Furthermore, metal substratecan tolerate high temperature treatment for sintering TiO2.Among the electron transport materials used in perovskite solarcell, TiO2 is still holding the efficiency record [8]. Meso-porousTiO2 scaffold is also essential for solving the hysteresis problemof perovskite solar cells [15].

Titanium (Ti) foils with TiO2 nanoparticles, nanotubes, ornanowires have been applied in flexible dye-sensitized solarcells (DSSCs) as photoanodes [16–23]. In particular, TiO2 nano-tube (TNT) arrays can be grown on Ti foil by a facile elec-trochemical anodization, which is a scaleable productiontechnique. However, flexible DSSCs with TNTarrays have shownrelatively poor performance with efficiencies lower than 4%[17,19] and all these devices required liquid electrolyte, whichwill cause sealing difficulty in large-scale production. Toreplace dye absorber in DSSCs, efficient perovskite absorberswill greatly increase the light absorption. More importantly,the monolithic all-solid-state device structure for perovskiteabsorber will render the flexible solar cells better performancestability during deformations. Recently Gao and co-authorsreported perovskite sensitized liquid solar cells with TNTs onrigid FTO glass [24]. Fiber shaped, flexible perovskite solar cellshave also been demonstrated with TiO2 nanoparticle coated on

stainless steel fibers, yielding efficiency of 3.3% [14]. However,solid-state flexible perovskite solar cells based on TiO2 nano-tubes have not been reported so far.

In Ti foil based perovskite devices, the opaque Ti foilhinders the light absorption from the photoanode, so thedevice will not work using a conventional metallic counterelectrode and therefore a transparent counter electrode isrequired. Transparent graphene [25] and carbon nanotubes(CNT) [20–23,26,27] have been successfully employed as thecounter electrode in DSSC devices. Recently, transparent CNTnetworks had been proved to be good hole conductor forperovskite solar cells [14,28]. These works inspired ourproposed flexible perovskite solar cell architecture with Tifoil as a working electrode, TNT as mesoporous layer forperovskite loading and carbon nanotubes as hole conductorand transparent electrode for light illumination. To date, it isthe first attempt of Ti metal foil substrate based flexibleperovskite solar cell and a decent power conversion efficiencyof 8.31% has been achieved.

Experimental

Fabrication of TiO2 nanotube arrays

Two kinds of Ti foils with different thicknesses (125 μm, 99.7%purity, Sigma-Aldrich; 25 μm, 99.98% purity, Sigma-Aldrich)were employed. Prior to anodization, Ti foils were degreasedultrasonically in acetone, ethanol and deionized (DI) water for20 min each and dried by air stream. Highly-ordered TiO2

nanotube arrays were prepared by electrochemically anodiza-tion at 20 V for 10 min at room temperature (�20 1C). Theanodizations were carried out with a two-electrode configura-tion with Ti foil as the working electrode and platinum gauzeas the counter electrode. The electrolyte solution was ethy-lene glycol (extra pure, Merck) containing 0.3 wt% ammoniumfluoride (98+%, Reagent, Sigma-Aldrich) and 2 vol% DI water.After anodization, the as-anodized TNT samples were rinsed inDI water to remove the electrolyte and then dried in air. Forapplication in solar cells, the as-grown TNTs were subjected tothermal annealing at 450 1C for 3 h to convert amorphoustitania into anatase phase. For better cell performance, theTNTs were also treated in 40 mM TiCl4 aqueous solution at70 1C for 10 h and then rinsed with ethanol and DI water.

Synthesis of carbon nanotubes

CNT network films were synthesized using the floatingcatalyst chemical vapor deposition (CVD) method using atube furnace [29]. Ferrocene (0.36 M) as catalyst and sulfur(0.036 M) as growth promotion agent were dissolved inxylene to form a uniform precursor solution. The tempera-ture for CVD was set to 1150 1C. Then 2500 sccm Ar and600 sccm H2 were introduced into the quartz tube as carrier

X. Wang et al.730

gas. When the temperature and gas flow stabilized, theprecursor solution was injected into the quartz tube at apreheating zone of the furnace with a temperature of180 1C. The precursor vaporized and was transported intothe center zone of the furnace by the gas flow for CNTgrowth. CNTs grew from the floating Fe catalyst in the carriergas flow. As shown in the Supporting materials (Fig. S1), theattained carbon nanotubes are mainly single and double-walled carbon nanotubes with diameters between 1 and2 nm. After growth, the CNTs formed an aerosol carried by

Fig. 1 Schematic of solid-state perovskite solar cells based onTi foil/TiO2 nanotubes and carbon nanotubes.

Fig. 2 (a) XRD patterns of the phase structures of TiO2 nanotubes/Ti, Pand perovskite peaks are marked by black box, circle and star respemorphology of perovskite/TiO2 nanotubes/Ti electrode; (c) tilted scansurface; (d) magnified top morphology of CNT covered perovskite.

the gas flow to the end of the quartz tube with a temperatureof 100–150 1C, where it was collected on a nickel foil.Individual nanotubes assembled into bundles with diameterof tens of nanometers and interweave to form a free standingCNT film.

Deposition of perovskite absorber

Perovskite absorber was deposited on TNT arrays by asequential method. 1 M lead iodide (PbI2) were dissolved inN,N-dimethylformamide overnight under stirring condition at70 1C. The PbI2 solution was spin coated on TNTs at 6000 rpmfor 5 s, followed by drying on a hot plate at 70 1C for 30 min. Inorder to convert PbI2 into CH3NH3PbI3, the PbI2 loaded TNTsamples were immersed in 8 mg mL�1 CH3NH3I solution in2-propanol for 30 min. Subsequently, the samples were rinsedwith 2-propanol and then dried at 70 1C for 30 min again.

Assembly of perovskite solar cell device

CNT films on nickel foil was lifted off by a taped substrate andtransferred on the top of perovskite sensitized TNT/Ti foil[24]. Several drops of toluene were used to wet CNT film forimproving the contact between CNTs and perovskite surface.After toluene vaporization, a hole transport materials, namely

bI2/TiO2 nanotubes/Ti and perovskite/TiO2 nanotubes/Ti. TiO2, PbI2ctively and the rest peaks from Ti substrate; (b) cross-sectionalning electron image of CNT film covering partially on perovskite

731TiO2 nanotube arrays based flexible perovskite solar cells with transparent carbon nanotube electrode

spiro-OMeTAD (2,20,7,70-tetrakis-(N,N-di-p-methoxyphenyla-mined) 9,90-spirobifluorene) in chlorobenzene (120 mg mL�1)were spin coated on CNTs covered substrate with a speed of4000 rpm for better hole collection. Prior to cell testing,both Ti foil and CNT electrodes were soldered for betterelectrical contact.

Film characterization and cell testing

The phase structure of the TNT arrays, the deposited PbI2and perovskite films were investigated by X-ray diffraction(XRD, Bruker-AXS D8 Advance). The cross-sectional mor-phology of perovskite sensitized TNTs and top views of CNTcovered perovskite layers were examined by field-emissionscanning electron microscope (FESEM, JSM-7600). The pho-tocurrent density–voltage (J–V) performance of the celldevices was characterized using solar simulator (San-EIElectric, XEC-301S) under AM 1.5 with illumination fromthe CNT side. The illumination area was determined by theblack mask with an area of 0.16 cm2 (small area testing) and0.36 cm2 (larger area for bending testing). Incident photonto current conversion efficiency (IPCE) was determinedusing PVE300 (Bentham), with a dual xenon/quartz halogenlight source, measured in DC mode and no bias light used.

Results and discussion

The cell configuration is shown in Fig. 1. From bottom to top insequence are Ti foil, TNTarrays loaded with perovskite absorber

Fig. 3 (a) Characteristic photocurrent–voltage curves of TNT and Cand with/without TiCl4 treatment; (b) IPCE of 25-μm-thick Ti basedsummarizes performance parameters with standard deviations caidentical fabrication conditions, together with the attained best va

and CNT networks composite with spiro-OMeTAD. Dense TNTarrays grown on Ti foil by electrochemical anodization serve bothas a scaffold for perovskite deposition and as an electroncollector. CNT network acts as hole collector and transparentelectrode. For better hole collection, the hole transport materialspiro-OMeTAD is infiltrated in carbon nanotubes network [28].Light comes from CNTside, as indicated by arrow in Fig. 1. SinceTi foil and CNT network are flexible materials, the integratedsolar cell device is expected to show good flexibility.

Highly ordered TiO2 nanotubes arrays are formed on Ti foil byelectrochemical anodization [30,31]. The as-anodized TNTs areamorphous in nature. To facilitate electron transport, amor-phous nanotubes are converted into anatase phase by thermalannealing at 450 1C for 3 h, as shown in the XRD patterns inFig. 2(a). CH3NH3PbI3 perovskite are formed on TNT arrays by asequential deposition method [7]. In the first step, PbI2 wasdeposited on nanotubes by spin coating, as revealed by the twopeaks at 12.72 1 and 39.52 1 in the XRD patterns Fig. 2(a)).Thereafter, soaking of PbI2 loaded nanotube substrates inCH3NH3I solution and subsequent drying process at 70 1C leadto formation of CH3NH3PbI3. The characteristics XRD peaks ofCH3NH3PbI3 are indicated by black stars in Fig. 2(a). It is in wellagreement with previous report [32].

Cross-sectional morphology of perovskite loaded TiO2 nano-tubes is presented in Fig. 2(b). The tube arrays formed on Ti foilare �300 nm in length and �60 nm in diameter. A dense layerof perovskite nanocrystals with a size of 100–400 nm completelycovers the nanotubes. The flexible CNT network is transferredon top of the perovskite layer as the counter electrode, asshown in the tilted SEM image of Fig. 2(c). The CNT transfer

NT based perovskite solar cells with different Ti foil thicknessperovskite solar cells with/without TiCl4 treatment. The tablelculated from different batches of devices fabricated underlue shown in brackets.

X. Wang et al.732

procedure is described in previous report [28]. The CNT film ishighly transparent with transmittance between 60% and 80% allover the CH3NH3PbI3 absorption wavelength range from 300 to800 nm (see Fig. S2 in Supporting information). The CNTnetwork is closely adhered to the perovskite by van der Waalforce. From the magnified top morphology of CNT/perovskite inFig. 2(d), it shows that the bundled CNT networks are sparsewith pores for light transmittance. In order to enhance holecollection in perovskite solar cells, spiro-OMeTAD are infiltratedinto CNT networks by spin coating [28].

Noticeably, there is a very thin TiO2 compact layerformed between the TNT arrays and Ti foil during anodiza-tion (Fig. 2(b)) [33–35]. The simultaneous anodic formationof the TiO2 blocking layer and nanotube scaffold has greatadvantages. The one-step anodization exempts the complexfabrication process of sequential depositing blocking layerand meso-porous TiO2 layers. It is also highly controllablewith the ability of forming uniform coating over large area,which is desirable for large scale production.

The thickness of Ti foil affects the device flexibility.Herein, two kinds of Ti foils with different thicknesses

Fig. 4 (a) Photograph of Ti foil/TNT and CNT based flexible perovsdifferent bending cycles; (c) plot of photovoltaic parameters as a f

(125 μm and 25 μm) were used in perovskite solar cellfabrication. For both thick and thin Ti foil based perovskitedevices, TiCl4 treatment was employed to improve thephotovoltaic performance. TiCl4 treatment has been widelyused in dye-sensitized solar cells [36–38] and perovskitesolar cells [5,7,28]. It fills the voids and cracks in TiO2

blocking layer and therefore decreases recombination insolar cells. Fig. 3 shows the device performances of theTNT/CNT perovskite solar cells with different Ti foil thick-nesses and with/without TiCl4 treatment. The combinedJ–V curves are presented in Fig. 3(a) and the correspondingphotovoltaic parameters are summarized in the table. Solarcells fabricated on 25 μm Ti foil exhibited higher photo-voltage (0.83 vs. 0.70 V for non-TiCl4-treated tubes; 0.99 vs.0.78 V for TiCl4-treated tubes) and improved fill factor (0.63vs. 0.62 for non-TiCl4-treated tubes; 0.68 vs. 0.62 for TiCl4-treated tubes), in comparison to 125 μm Ti foil based solarcells. The Voc and fill factor improvement may be ascribedto the smaller surface roughness of the thinner Ti foil, asshown in Fig. S3 of the Supporting information. It ispresumable that the smoother Ti surface improves the

kite solar cells; (b) combined J–V curves of flexible device withunction of bending cycles.

733TiO2 nanotube arrays based flexible perovskite solar cells with transparent carbon nanotube electrode

flatness of perovskite layer and reduces the unfavorablecurrent shunting, thus in return increases the Voc and fillfactor.

As evident in Fig. 3(a), TiCl4 treatment after anodizationis beneficial to enhance all photovoltaic parameters, includ-ing photocurrent, Voc, fill factor and thus power conversionefficiency. Notably, Voc is greatly improved after TiCl4treatment (0.78 vs. 0.70 V for 125-μm-thick Ti; 0.99 vs.0.83 V for 25-μm-thick Ti). It can be ascribed to thereduction of recombination sites by the newly formednanoparticles from TiCl4 treatment. The slight improvementof photocurrent can be also shown from the IPCE ofperovskite solar cells with/without TiCl4 treatment, asdisplayed in Fig. 3(b). From 400 nm to 700 nm, TiCl4 treatedTNTs exhibit higher IPCE, which indicates a higher chargeseparation efficiency thus results in photocurrent increase.Noticeably, the IPCE is relatively low at wavelengthbetween 300 and 400 nm due to the strong light absorptionof spiro-OMeTAD in this wavelength region (Fig. S2 in theSupporting information). Further improvement of photocur-rent can be expected by adapting HTM materials with bettertransparency. For the 25-μm-thick Ti foils, the best per-ovskite solar cells with TiCl4 treatment yield efficiency of8.31%. The obtained efficiency is among the highestreported for flexible perovskite solar cells [10–13].

Besides photovoltaic performance, tolerance to mechan-ical bending is another important factor of consideration forflexible perovskite solar cells. A photograph of flexible TNT/CNT perovskite solar cell is shown in Fig. 4(a). The 25-μm-thick Ti foil based devices were used to investigate solar cellflexibility. The solar cell with length of 2.5 cm was bendedto a bending radius of 0.75 cm by mechanical force up to100 cycles. The dependence of the device performances onbending cycles is presented in Fig. 4(b and c). J–V curves inFig. 4(b) show that the photocurrent remains identical andthe photovoltage is slightly decreased through the bendingtests. As the J–V curve shape moves inwards with bending,as indicated by the arrow, the most affected device para-meter after bending is the fill factor. During repeatingbending, micro-sized cracks and delamination could begenerated at the interfaces between different layers ofthe solar cells, which would deteriorate the interface andincrease the series resistance of solar cells. The decrease offill factor could be a result of the increased series resistancein solar cells after bending. The photovoltaic parameterswith bending cycles are plotted in Fig. 4(c). After 100bending cycles, the photocurrents are just slightlydecreased from 9.56 to 9.37 mA cm�2 and photovoltage isreduced from 0.98 V to 0.95 V. The fill factor is mostlyaffected, from 0.64 to 0.57, with 11% decrease. Theefficiency is thus decreased from 6.01% to 5.06%. Theresults show that the mechanical bending does not signifi-cantly affect the cell performance. Ti foil/TNT/CNT basedperovskite solar cells maintain good performance after 100bending cycles, demonstrating their high flexibility.

Conclusions

In summary, flexible, solid-state perovskite solar cells basedon Ti foil/TNTs and CNTs have been demonstrated. To ourbest knowledge, it is the first demonstration of flexible

perovskite solar cells on Ti metal foil substrate. The Ti foil/TNTs act as scaffold for perovskite loading and electrontransport layer, while the transparent CNT top electrodeacts as hole collecting layer and light transmission. With25 μm Ti foil and TiCl4 treatment to TiO2 nanotube arrays,power conversion efficiency up to 8.31% has been achieved.The solar cells on Ti foil maintain good performance after100 mechanical bending cycles, indicating their excellentflexibility. Considering the high efficiency, good flexibilityand simple fabrication technique, Ti foil/TNTs based flexibleperovskite solar cells holds a promising future for roof-topphotovoltaics and power sources for wearable devices.

Acknowledgement

Funding from National Research Foundation (NRF), Singa-pore, is acknowledged through CRP Award no. NRF-CRP4-2008-03 and the Singapore-Berkeley Research Initiative forSustainable Energy (SinBeRISE) CREATE program. X. Wangwishes to thank the support from the World Future Founda-tion (WFF) as a recipient for the 2014 WFF PhD Prize.

Appendix A. Supporting information

Supplementary data associated with this article can befound in the online version at http://dx.doi.org/10.1016/j.nanoen.2014.11.042.

References

[1] A. Kojima, K. Teshima, Y. Shirai, T. Miyasaka, J. Am. Chem.Soc. 131 (2009) 6050–6051.

[2] J.-H. Im, C.-R. Lee, J.-W. Lee, S.-W. Park, N.-G. Park,Nanoscale 3 (2011) 4088–4093.

[3] G. Xing, N. Mathews, S. Sun, S.S. Lim, Y.M. Lam, M. Grätzel,S. Mhaisalkar, T.C. Sum, Science 342 (2013) 344–347.

[4] S.D. Stranks, G.E. Eperon, G. Grancini, C. Menelaou, M.J.P. Alcocer, T. Leijtens, L.M. Herz, A. Petrozza, H.J. Snaith,Science 342 (2013) 341–344.

[5] H.-S. Kim, C.-R. Lee, J.-H. Im, K.-B. Lee, T. Moehl, A. Marchioro,S.-J. Moon, R. Humphry-Baker, J.-H. Yum, J.E. Moser,M. Graetzel, N.-G. Park, Sci. Rep. 2 (2012) 591.

[6] M.M. Lee, J. Teuscher, T. Miyasaka, T.N. Murakami, H.J. Snaith,Science 338 (2012) 643–647.

[7] J. Burschka, N. Pellet, S.-J. Moon, R. Humphry-Baker, P. Gao,M.K. Nazeeruddin, M. Graetzel, Nature 499 (2013) 316–319.

[8] H. Zhou, Q. Chen, G. Li, S. Luo, T.-b. Song, H.-S. Duan, Z. Hong,J. You, Y. Liu, Y. Yang, Science 345 (2014) 542–546.

[9] J. You, Z. Hong, Y. Yang, Q. Chen, M. Cai, T.-B. Song, C.-C. Chen,S. Lu, Y. Liu, H. Zhou, ACS Nano 8 (2014) 1674–1680.

[10] C. Roldan-Carmona, O. Malinkiewicz, A. Soriano, G. MinguezEspallargas, A. Garcia, P. Reinecke, T. Kroyer, M.I. Dar, M.K. Nazeeruddin, H.J. Bolink, Energy Environ. Sci 7 (2014)994–997.

[11] Y.-F. Chiang, J.-Y. Jeng, M.-H. Lee, S.-R. Peng, P. Chen, T.-F. Guo,T.-C. Wen, Y.-J. Hsu, C.-M. Hsu, Phys. Chem. Chem. Phys 16(2014) 6033–6040.

[12] P. Docampo, J.M. Ball, M. Darwich, G.E. Eperon, H.J. Snaith,Nat. Commun 4 (2013) 2761.

[13] D. Liu, T.L. Kelly, Nat. Photon 8 (2014) 133–138.

X. Wang et al.734

[14] L. Qiu, J. Deng, X. Lu, Z. Yang, H. Peng, Angew. Chem. Int. Ed.126 (2014) 10425–10428.

[15] N.J. Jeon, J.H. Noh, Y.C. Kim, W.S. Yang, S. Ryu, S.I. Seok,Nat. Mater. (2014). http://dx.doi.org/10.1038/NMAT4014.

[16] S. Ito, N.-L.C. Ha, G. Rothenberger, P. Liska, P. Comte, S.M. Zakeeruddin, P. Pechy, M.K. Nazeeruddin, M. Gratzel,Chem. Commun. (2006) 4004–4006.

[17] D. Kuang, J. Brillet, P. Chen, M. Takata, S. Uchida, H. Miura,K. Sumioka, S.M. Zakeeruddin, M. Grätzel, ACS Nano 2 (2008)1113–1116.

[18] J.-Y. Liao, B.-X. Lei, H.-Y. Chen, D.-B. Kuang, C.-Y. Su, EnergyEnviron. Sci 5 (2012) 5750–5757.

[19] Z. Liu, V. Subramania, M. Misra, J. Phys. Chem. C 113 (2009)14028–14033.

[20] T. Chen, L. Qiu, Z. Yang, Z. Cai, J. Ren, H. Li, H. Lin, X. Sun,H. Peng, Angew. Chem. Int. Ed. 51 (2012) 11977–11980.

[21] H. Sun, Z. Yang, X. Chen, L. Qiu, X. You, P. Chen, H. Peng,Angew. Chem. Int. Ed. 52 (2013) 8276–8280.

[22] S. Pan, Z. Yang, H. Li, L. Qiu, H. Sun, H. Peng, J. Am. Chem.Soc. 135 (2013) 10622–10625.

[23] X. Fang, Z. Yang, L. Qiu, H. Sun, S. Pan, J. Deng, Y. Luo,H. Peng, Adv. Mater. 26 (2014) 1694–1698.

[24] X. Gao, J. Li, J. Baker, Y. Hou, D. Guan, J. Chen, C. Yuan,Chem. Commun. 50 (2014) 6368–6371.

[25] X. Wang, L. Zhi, K. Müllen, Nano Lett. 8 (2007) 323–327.[26] S. Zhang, C. Ji, Z. Bian, R. Liu, X. Xia, D. Yun, L. Zhang,

C. Huang, A. Cao, Nano Lett. 11 (2011) 3383–3387.[27] D. Liu, M. Zhao, Y. Li, Z. Bian, L. Zhang, Y. Shang, X. Xia, S. Zhang,

D. Yun, Z. Liu, A. Cao, C. Huang, ACS Nano 6 (2012) 11027–11034.[28] Z. Li, S.A. Kulkarni, P.P. Boix, E. Shi, A. Cao, K. Fu,

S.K. Batabyal, J. Zhang, Q. Xiong, L.H. Wong, N. Mathews,S.G. Mhaisalkar, ACS Nano 8 (2014) 6797–6804.

[29] Z. Li, Y. Jia, J. Wei, K. Wang, Q. Shu, X. Gui, H. Zhu, A. Cao,D. Wu, J. Mater. Chem. 20 (2010) 7236–7240.

[30] D. Gong, C.A. Grimes, O.K. Varghese, W. Hu, R.S. Singh,Z. Chen, E.C. Dickey, J. Mater. Res. 16 (2001) 3331–3334.

[31] X. Wang, L. Sun, S. Zhang, D. Zhao, Electrochim. Acta (2013)200–208.

[32] T. Baikie, Y. Fang, J.M. Kadro, M. Schreyer, F. Wei,S.G. Mhaisalkar, M. Graetzel, T.J. White, J. Mater. Chem. A1 (2013) 5628–5641.

[33] S.P. Albu, A. Ghicov, S. Aldabergenova, P. Drechsel,D. LeClere, G.E. Thompson, J.M. Macak, P. Schmuki, Adv.Mater. 20 (2008) 4135–4139.

[34] A. Ghicov, P. Schmuki, Chem. Commun. (2009) 2791–2808.[35] Q.A.S. Nguyen, Y.V. Bhargava, V.R. Radmilovic, T.M. Devine,

Electrochim. Acta 54 (2009) 4340–4344.[36] P. Roy, D. Kim, I. Paramasivam, P. Schmuki, Electrochem.

Commun. 11 (2009) 1001–1004.[37] D. Kim, P. Roy, K. Lee, P. Schmuki, Electrochem. Commun. 12

(2010) 574–578.[38] X. Wang, L. Sun, S. Zhang, X. Wang, ACS Appl. Mater.

Interfaces (2014) 1361–1365.

Dr. Xiaoyan Wang received her Bachelor andMaster degree in Materials Science and Engi-neering from BeiHang University, Beijing, in2006 and 2009, respectively. In February2014, she received her PhD degree fromNanyang Technological University, Singapore.She won the 2014 “World Future Foundation”outstanding Ph.D thesis prize. She is currentlya research fellow in Energy Research Institute@ NTU (ERI@N). She is now working on CIGS

thin film solar cell and organic–inorganicperovskite solar cells.

Dr. Zhen Li received his B.S. and Ph.D.degree in Material Science and Engineeringfrom Tsinghua University, China in 2008 and2013. Currently, He is research fellow inEnergy Research Institute (ERI@N), NanyangTechnological University. His research inter-est includes perovksite-based solar cells andsynthesis and photovoltaic applications ofcarbon nanomaterials (carbon nanotubes,graphene etc.)

Miss. Wenjing Xu received her B.S. degree inSchool of Physics And Engineering fromZhengzhou University, China in 2013. She iscurrently a Ph. D. student, majoring in Mate-rial Science and Engineering in College ofEngineering in Peking University. Her researchinterest focuses on fabrication and character-ization of carbon-based thin-film solar-cell forefficient energy conversion.

Dr. Sneha Avinash Kulkarni received herPh.D. in Physical and Materials Chemistryfrom National Chemical Laboratory (NCL),University of Pune, India in 2008. Sheworked as a Postdoctoral Fellow in NationalUniversity of Singapore (NUS) from 2008–2011. Presently, she is working as a SeniorResearch Fellow in Energy Research Insti-tute (ERI@N), Nanyang Technological Uni-versity. Her research interests focus in

synthesis and application of nano materialsfor energy harvesting and storage. Her

current research is involved in fabrication of the perovskite basedsolar cell.

Dr. Sudip Kumar Batabyal obtained his PhDdegree in physics from Indian Associationfor the Cultivation of Science (JadavpurUniversity), India in 2007. After completingthe Postdoctoral Research work in NationalUniversity of Singapore and in NanyangTechnological University he joined as asenior scientist in Energy Research Institutein NTU (ERI@N). His research work is insynthesis and application of nanostructured

materials for energy harvesting and storage.Solution processing of inorganic and hybrid

materials for device fabrication is of his special interest. Hisresearch focus is on the development of absorber materials andelectrode materials for photovoltaic’s devices.

Prof. Sam Zhang received Ph.D. degree inCeramics in 1991 from The University ofWisconsin-Madison, USA and is a tenured fullprofessor (since 2006) at School of Mechanicaland Aerospace Engineering, NTU. He serves asEditor-in-Chief for Nanoscience and Nano-technology Letters and Principal Editor forJournal of Materials Research (USA). He hasbeen in processing and characterization ofnanocomposite thin films and coatings for

more than 20 years and has published morethan 280 peer reviewed international papers,

12 books, 20 book chapters and guest-edited more than 10 Journalvolumes. His papers have been cited 4532 times and his h-index is 37.

735TiO2 nanotube arrays based flexible perovskite solar cells with transparent carbon nanotube electrode

Prof. Anyuan Cao received his PhD degreein Mechanical Engineering from TsinghuaUniversity. He has spent 3 years in Rensse-laer Polytechnic Institute as a postdocresearcher, and 3 years in the University ofHawaii at Manoa as an assistant professor.He is currently a professor in the Depart-ment of Materials Science and Engineering,College of Engineering, Peking University.His research areas include controlled synth-

esis of macroscopic structures based oncarbon nanotubes and graphene, self-

assembly, nanocomposites, nanoelectronics, energy and environ-mental applications. He has published over 100 peer-reviewedjournal papers.

Prof. Lydia Helena Wong received B. Appl.Sci. with Honors and Ph.D. in MaterialsScience and Engineering from NTU. After-wards, she worked as a Senior Engineer atthe Technology Development Department ofChartered Semiconductor Manufacturing(Global Foundries) and was a Visiting Scien-tist at Stanford University developingorganic photovoltaic materials at theDepartment of Chemical Engineering. She

is currently an Assistant Professor at theSchool of Materials Science and Engineer-

ing, NTU. Her research group currently focuses on the investigationof non-toxic and abundant metal oxides and chalcopyrite materialsfor solar harvesting applications.