Available online at www.sciencedirect.com

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

Nano Energy (2015) 18, 109–117

http://dx.doi.org/12211-2855/& 2015 E

nCorresponding auE-mail address: j

RAPID COMMUNICATION

Economical low-light photovoltaics by usingthe Pt-free dye-sensitized solar cell withgraphene dot/PEDOT:PSS counter electrodes

Chuan-Pei Leea,b, Chin-An Lina, Tzu-Chiao Weia, Meng-Lin Tsaia,Ying Menga, Chun-Ting Lic, Kuo-Chuan Hoc, Chih-I Wub,Shu-Ping Laud, Jr-Hau Hea,n

aComputer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division, King Abdullah,University of Science and Technology (KAUST), Thuwal, 23955-6900, Saudi ArabiabInstitute of Photonics and Optoelectronics and Department of Electrical Engineering, National TaiwanUniversity, Taipei 10617, Taiwan, ROCcDepartment of Chemical Engineering, National Taiwan University, Taipei 10617, Taiwan, ROCdDepartment of Applied Physics, The Hong Kong Polytechnic University, Hong Kong Special AdministrativeRegion

Received 12 July 2015; received in revised form 5 September 2015; accepted 10 October 2015Available online 17 October 2015

KEYWORDSDye-sensitized solarcell;Conducting polymer;Graphene quantumdots;Platinum-free counterelectrode;Low light

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

thor.rhau.he@kaust.ed

AbstractGraphene dots (GDs) are used for enhancing the performance of the poly(3,4-ethylenediox-ythiophene):polystyrenesulfonate (PEDOT:PSS)-based counter electrodes in Pt-free dye-sensi-tized solar cells (DSSCs). As compared to PEDOT:PSS CEs, GD–PEDOT:PSS films possess a roughsurface morphology, high conductivity and electrocatalytic activity, and low charge-transferresistance toward I�/I3

� redox reaction, pushing cell efficiency to 7.36%, which is 43% higherthan that of the cell with PEDOT:PSS CEs (5.14%). Without much impact on efficiency, the DSSCswith GD–PEDOT:PSS CEs work well under low-light conditions (light intensity o13.5 mW cm�2

and angle of incidence 4601), such as indoor and low-level outdoor lighting and of the sunwhile the other traditional cells would fail to work. The concurrent advantage in low cost in Pt-free materials, simple fabrication processes, comparable efficiency with Pt CEs, and highperformance under low-light conditions makes the DSSC with GD–PEDOT:PSS CEs suitable toharvest light for a diverse range of indoor and low-level outdoor lighting locations.& 2015 Elsevier Ltd. All rights reserved.

015.10.008hts reserved.

u.sa (J.-H. He).

C.-P. Lee et al.110

Introduction cathodic peak current density) due to the nonconductive

For the practical use of solar cells, the light source usuallydoes not match the ideal solar power density (100 mWcm�2), and the angle of incidence (AOI) of the sun changesduring the day and depends on the installation location.Moreover, a large number of diffused (scattered) light arecreated after sunlight passes through the atmosphere andreaches the earth; on cloudy days, over 90% of the light isdiffused. It is well known that most common Si solar cellsstrongly depend on light incident angle [1–4]. Althoughincorporating the light tracing systems, it still exhibits poorperformance under low-light condition, such as cloudy day,early morning and nightfall of a day [1]. Accordingly, it isimportant to develop the low-light solar cells which canachieve excellent conversion efficiency even in broadlydiffused light and poor weather conditions. Moreover, thelow-light solar cell enables new indoor applications such assmart home and intelligent building, with an average powerconsumption in the range of microwatts, including sensors,chargers, and standby power. For example, indoor light-emitting diode (LED) has been tremendous potential assupplemental lighting systems for crop production andlow-light photovoltaics could be used to harvest the residuallight energy, which is not absorbed by the crop.

In contrast with conventional Si solar cells, the performanceof dye-sensitized solar cells (DSSCs) are less affected by thelight intensity and less sensitive to the AOI [1,5], which makesDSSCs to be the most efficient photo-to-electron conversiondevices under indoor and low-level outdoor lighting. A typicalDSSC consists of a dye-sensitized mesoporous TiO2 photoanode,an electrolyte containing tri-iodide/iodide (I�/I3

�) redox coupleand a counter electrode (CE). The CE, a crucial part in a DSSC,should possess a good conductivity for the transportation ofelectrons flowing through an external circuit and a superiorcatalytic ability for the reduction of I3

� for reducing the energyloss caused by the charge transfer overpotential between theelectrolyte and the CEs [6–8]. In DSSCs, Pt is the most usedcatalyst on the CE because of its high conductivity, stability andhigh electrocatalytic activity toward I3

�/I� redox reaction.However, the high cost and energy-consumptive fabricationprocess of the Pt CE limits its applicability. By replacing Pt withother cheaper electrocatalytic materials, the cost of fabricationof DSSCs on industrial scale can be greatly reduced, that wouldheighten the competitiveness of DSSCs in solar cell markets.Recently, conducting polymers (such as poly(3,4-ethylene diox-ythiophene) (PEDOT) [9], poly(3,3-diethyl-3,4-dihydro-2H-thieno-[3, 4-b][1, 4]-dioxepine) (PProDOT–Et2) [10], polypyrrole(PPy) [11], and polyaniline (PANI) [12]) and carbonaceousmaterials (such as carbon black (CB) [13], graphite [14], carbonnanotube (CNT) [15], and graphene [16]) have become the mostpromising candidates as catalysts of CE for DSSCs because theyare metal element-free, abundant and low cost, and possessgood electrocatalytic activity.

A water-dispersible conducting polymer, poly(3,4-ethylenedioxythiophene):poly(4-styrene sulfonate) (PEDOT:PSS), hasattracted much attention as the catalytic CEs of DSSCs mainlydue to its exceptional advantage of aqueous solution proces-sibility [17]. However, pristine PEDOT:PSS films are plagued bya low conductivity (generally below 1 S cm�1) and poorelectrocatalytic activity for the reduction of I3

� (unobvious

counter anion, PSS�, disturbing the conduction path of PEDOTinside the film as well as the poor catalytic surface area of itsflat film [17,18]. Inert solvents [19] or carbon materials [20–24]have been employed to improve conductivity and catalyticsurface areas of PEDOT:PSS films. For example, by introducingCB into the PEDOT:PSS-based CEs for their DSSCs, η can achieveto 7.01% [22]. Multiwall CNT-PEDOT:PSS composite CE for DSSCsexhibits η of 6.50% [23]. A catalytic film composited ofgraphene and PEDOT:PSS for the use of CE in a DSSC hadreached 4.50% efficiency [24]. However, a perfect graphenesheet usually possesses limited active sites for electrocatalyticreaction in spite of its extraordinarily high electrical conduc-tivity [25].

Several strategies are employed to increase the electro-catalytic active sites on graphene sheets, such as chemicalfunctionalization (e.g., oxygen-containing functional groupsreduced graphene oxide) [26], heteroatom doping (e.g.,nitrogen-doped graphene) [25,27,28], and nano-sized gra-phene pieces (e.g., graphene dots (GDs)). Among thesegraphene nanostructures, GDs have attracted great atten-tion and been widely applied in bio-imaging [29–31], LEDs[32,33] and photovoltaics [34–37] due to their uniqueproperties of quantum confinement and edge effects[34,38]. Moreover, nanometer-size and rich oxygen-containing group of GDs facilitate them to be well dispersedin common solvents [39], which is expected to enablevarious solution-processable applications.

In this study, the water-soluble GDs via microwave-assisted hydrothermal technique is incorporated into thePEDOT:PSS-based CEs for Pt-free DSSCs achieve an excellentη of 7.36% under 1 sun AM 1.5 G, which to the best of ourknowledge is the highest ever reported for a DSSC withPEDOT:PSS-based CEs. DSSCs with GD–PEDOT:PSS CEs canmaintain the performance under various low-light condi-tions (light intensity o13.5 mW cm�2 and angle of inci-dence 460 1), such as white-light LEDs, noon of a cloudyday, and early morning and nightfall of a sunny day, ascompared with conventional Si solar cells. Meanwhile, theDSSCs with GD–PEDOT:PSS CEs and with Pt CEs exhibitsimilar η under low-light conditions. Less performancedegradation of DSSCs under low-light, compared with theSi solar cells, is due to reduced Jsc degradation andimproved FF. The concurrent advantage in low-cost materi-als, simple fabrication processes and high performanceunder low light makes our Pt-free DSSCs playing a crucialrole in energy-harvesting systems for smart-home andintelligent buildings.

Experimental

Materials

Lithium iodide (LiI, synthetical grade) and iodine (I2, synthe-tical grade) were obtained from Merck. Acetone (99+%), Tert-butyl alcohol (tBA, 96%), guanidine thiocyanate (GuSCN, 99+%), and 4-tert-butylpyridine (TBP, 96%) were bought fromAcros. Ti (IV) tetraisopropoxide (TTIP, 498%), lithium perchlo-rate (LiClO4, Z98.0%), ethanol (EtOH, absolute), isopropylalcohol (IPA, 99.5%), and 2-methoxyethanol (99.95%), were

111Economical low-light photovoltaics by using the Pt-free dye-sensitized solar cell with graphene dot/PEDOT:

received from Sigma-Aldrich. Acetonitrile (ACN, 99.99%) wasprocured from J. T. Baker. 1,2-dimethyl-3-propylimidazoliumiodide (DMPII) and cis-diisothiocyanato-bis (2,2’-bipyridyl-4,4’-dicarboxylato) ruthenium (II) bis(tetra-butylammonium) (N719dye) were purchased from Solaronix (S.A., Aubonne, Switzer-land). PEDOT:PSS aqueous solution (PH 500, 1.0–1.4 wt% dis-persion in water) was obtained from Heraeus.

Synthesis of water soluble GDs solution

The GD solution was prepared by using glucose anddeionized-water as the source and solvent, respectively[33]. First, 2.5 mL glucose solutions were transferred to aglass bottle with 4 mL volume and a tightened cover. Theglass bottle was then heated with a microwave oven at apower of 595 W for 9 min. Subsequently, the glass bottlewas cooled to room temperature, and the water soluble GDssolution was thus prepared. The GD sample was dialyzed bymolecular weight cutoff (MWCO) of 1000 against DI waterfor characterization. The yield of the GDs was determinedto be around �80%. As we do not know how many glucosemolecules were converted to GDs, thus the concentration ofGDs could not be determined.

Preparation of CEs

The fluorine-doped SnO2 (FTO, TEC–7, 10 Ω sq�1, NSGAmerica Inc., New Jersey, USA) conducting glasses werefirst cleaned with a neutral cleaner and then washed withDI-water, acetone, and IPA sequentially. The standard Pt CEwas prepared by sputtering for 5 nm of Pt on the cleanedFTO substrate. The PEDOT:PSS film was prepared with amixing solution composited of 50 V% PEDOT:PSS aqueoussolution and 50 V% ethanol. For the GDs–PEDOT:PSS mixingsolution, the GD solution content in the above mixingsolution (PEDOT:PSS, GDs, and ethanol solutions) was opti-mized by applying different volume ratios from 10 V% to40 V% as shown in Figure S1 (see Supporting information).The optimized ratio was found to be 30 V% GD solutioncontent. Therefore, another mixing solution composited of50 V% PEDOT:PSS aqueous solution, 30 V% GDs solution, and20 V% ethanol was used for preparing the GDs–PEDOT:PSScomposite film. These films were fabricated on the cleanedFTO substrates via drop-coating method and then heated to60 1C for 30 min.

Fabrication of DSSCs

The conducting surface of the FTO was treated with asolution of TTIP in 2-methoxyethanol (weight ratio of 1:3)for obtaining a good contact between the conducting glassand the TiO2 film. For the mesoporous double-layer TiO2

film, 12-μm-thick transparent layer (diameter of 20–50 nm,Eversolar P-300) and 5-μm-thick light-scattering layer (PST-400, JGC Catalysts and Chemicals, Japan) were depositedon the treated FTO sequentially by a conventional screen-printing method. A portion of 0.4� 0.4 cm2 was selectedfrom the double-layer TiO2 film as the active area byremoving the side portions by scrapping. The TiO2 film wasgradually heated to 500 1C in an oxygen atmosphere andsubsequently sintered at 500 1C for 30 min. After sintering

at 500 1C and cooling to 80 1C, the TiO2 electrode wasimmersed in a 3� 10�4 M N719 dye solution with a mixedsolvent of ACN/tBA (volume ratio of 1/1) at room tempera-ture for 24 h. Such a prepared TiO2 electrode was coupledwith a CE; these two electrodes were separated by a 25 μm-thick Surlyns film (SX1170-25, Solaronix S.A., Aubonne,Switzerland) and sealed by heating. A mixture of 1.2 MDMPII, 0.35 M I2, 0.1 M GuSCN, and 0.5 M TBP in ACN/MPN(volume ratio of 8/2) was used as the electrolyte. Theelectrolyte was injected into the gap between the twoelectrodes by capillarity.

Measurements and instruments

Transmission electron microscopy (TEM, JEM-2100F, JEOL) wasperformed at operating voltage of 200 kV. The height of theGDs was characterized by atomic force microscope (AFM,NanoScope IV, Digital Instruments) under tapping mode. UV-absorption spectrum was obtained for the GD solution by usingan UV–visible spectrophotometer (UV–VIS–NIR, V670, Jasco Co.,Japan). For characterizations of photoluminescence (PL) spec-tra, the excitation and emission spectra of the GD solutionwere recorded using an FLS920P Edinburgh Analytical Instru-ment apparatus with Xe lamp as an excitation source. Thesurface morphologies of various catalytic films were observedby field emission scanning electron microscopy (FE-SEM, NovaNano-SEM 230, FEI, Oregon, USA).

The surface of the DSSC was covered by a mask leaving alight-illuminating area of 0.16 cm2 and then illuminated bya class – A quality solar simulator (XES-301S, AM 1.5G, SAN-EI ELECTRIC CO., LTD.). Incident light intensity (100 mWcm�2) was calibrated with a standard Si cell (PECSI01,Peccell Technologies, Inc., Yokohama, Japan). Photoelec-trochemical characteristics of the DSSC were recorded witha potentiostat/galvanostat (PGSTAT 30, Autolab, Eco-Che-mie, the Netherlands). Electrochemical impedance spectro-scopy (EIS) analysis was performed using the above-mentioned potentiostat/galvanostat, equipped with anFRA2 module. Herein, a symmetric sandwich-type cell wasused to investigate the electrocatalytic ability of the CE byEIS with a frequency range from 10 mHz to 100 kHz; thesymmetric cell consisting of an electrolyte and two identicalelectrodes with an area of 1 cm2, separated by a Surlynsfilm of 60 μm in thickness. Impedance spectra data wereanalyzed by fitting them to an equivalent circuit model[40,41], using Z-view software. Cyclic voltammetry (CV) wasperformed to investigate electrocatalytic abilities of theCEs. The CV was carried out with a three-electrode electro-chemical system, by using an electrode with PEDOT or GDs–PEDOT:PSS or sputtered Pt catalytic film as the workingelectrode, a Pt foil as the CE, and an Ag/Ag+ electrode asthe reference electrode in an ACN solution, containing10 mM I�, 1 mM I2, and 0.1 M LiClO4.

Results and discussions

GDs are edge-bound nano-sized graphene pieces and exhibitunique electronic and optical properties due to the quantumconfinement and edge effects [34,38]. The synthesized GDswere investigated by TEM, UV–visible spectroscopy, and PL

C.-P. Lee et al.112

spectroscopy. Figure 1a shows the TEM image of the mono-dispersed GDs, which exhibit uniform diameters of �3.50 nm.As shown in the inset of Figure 1a, the HRTEM image indicates

Figure 2 Cross-sectional SEM images of (a) PEDOT:PSS and (b)(c) PEDOT:PSS and (d) GD–PEDOT:PSS films.

200 300 400 5000

1

2

3

Abs

orba

nce

Wavelength (nm)

GDs

Figure 1 (a) TEM image of GDs and its corresponding HRTEM imageprofile. (c) Absorbance spectrum of the GDs. (d) PL spectra of GDs. Tby ambient light (left) and UV light (right).

high crystallinity of GDs with a lattice parameter of 0.246 nmcorresponding to the (1120) lattice fringes of graphene.Figure. 1b shows the AFM image of the monodispersed GDs,

GD–PEDOT:PSS films. Corresponding top-view SEM images of

400 600 800

Inte

nsity

(a.u

.)

Wavelength (nm)

Excitation wavelengths 300 nm 400 nm 500 nm

(inset). (b) AFM image of the GDs and its corresponding heighthe inset in (d) shows photographs of the GDs solutions irradiated

0 200 400 600 8000

4

8

12

16

Phot

ocur

rent

den

sity

(mA

cm

-2)

Voltage (mV)

DSSC with PEDOT:PSS CE GD-PEDOT:PSS CE

-0.6 -0.4 -0.2 0.0 0.2 0.4-2

0

2

4

Cur

rent

den

sity

(mA

cm

-2)

Potential (V vs. Ag/Ag+)

CE with PEDOT:PSS GD-PEDOT:PSS

10 15 20 25 300

4

8

12

CE with PEDOT:PSS GD-PEDOT:PSS

- Z"

(ohm

)

Z' (ohm)

Equivalent circuit model

CPE

RsRct Zw

Figure 3 (a) The J–V curves of the DSSCs with PEDOT:PSS andGD–PEDOT:PSS CEs, measured at 100 mWcm�2 (AM 1.5 G).(b) The CV curves of the CEs with PEDOT:PSS and GD–PEDOT:PSS films. (c) Nyquist plots of the symmetric sandwich-typecells with PEDOT:PSS and GD–PEDOT:PSS CEs obtained at zerobias potential. The inset in (c) is the equivalent circuit.

Table 1 Photovoltaic parameters of the DSSCs withPEDOT:PSS and with GD–PEDOT:PSS counter electrodes.

DSSC with η (%) VOC (mV) JSC (mA cm�2) FF

PEDOT:PSS CE 5.14 668 12.82 0.60GD–PEDOT:PSS

CE7.36 718 14.70 0.70

113Economical low-light photovoltaics by using the Pt-free dye-sensitized solar cell with graphene dot/PEDOT:

three GDs are selected randomly and labeled as A, B, and C.The average height of these GDs is 2.90 nm as shown in theinset of Figure 1b. As shown in Figure. 1c, two absorption peaksat 228 and 282 nm are observed for the diluted GD solution,which consists with the result in the previous literature [33].Figure. 1d shows the PL spectra of the GDs solution. A broademission peak at 450 nm is observed when the sample isexcited by 300 nm. When the excitation wavelength changedfrom 300 to 500 nm, the PL peak shifts from 450 to 537 nm,and shows the decrease of PL intensity. The excitationwavelength-dependent emission wavelength and intensityobserved here is a common phenomenon in carbon-basedquantum dots [31,33]. Insets of Figure 1d are the photographsof GD solution taken under visible (left) and UV (right) lightillumination, showing the excitation wavelength-dependentfluorescence of GDs.

Figure 2 shows SEM images of PEDOT:PSS and GD–PEDOT:PSScomposite films on the FTO substrate. As shown in Figures 2aand b, the cross-sectional SEM image reveals that both PEDOT:PSS and GD–PEDOT:PSS composite films possess the similar filmthickness of �6.3 μm. It is observed from Figure. 2b that thePEDOT:PSS film has flat surfaces with smooth morphology, whichimplies the unfavorable electrochemical surface area for theredox reaction. On the other hand, a rough surface morphologyconsisting of irregular island-like structures can be seen in theGD–PEDOT:PSS film, as shown in Figure. 2c. The catalytic filmwith the high superficial roughness is expected to enhance itselectrocatalytic activity for I�/I3

� redox couple, as well asreduce the charge-transfer resistance at the interface of CE/electrolyte [7,37]. Moreover, we measured the sheet resistanceof the PEDOT:PSS film with and without GDs and found that theGD–PEDOT:PSS film (1.5� 10�4Ω cm) possesses lower sheetresistance than the PEDOT:PSS film (9.6� 10�4 Ω cm). There-fore, the attractive performance of the GD–PEDOT:PSS film as aCE in a DSSC can be excepted due to the large surfaceroughness and the improved conductivity.

Photocurrent density–voltage (J–V) characteristics of theDSSCs with PEDOT:PSS CEs and GD–PEDOT:PSS CEs are shownin Figure 3a, and the corresponding photovoltaic parametersare listed in Table 1. The cell with GD–PEDOT:PSS CEs shows thebest performance with a cell efficiency (η) of 7.36%, an open-circuit voltage (VOC) of 718 mV, a fill factor (FF) of 0.70, and ashort-circuit current density (JSC) of 14.70 mA cm�2; this η ismuch higher than that of the cell with PEDOT:PSS CE (5.14%).The η of the DSSC with GD–PEDOT:PSS CE (7.36%) is close to thatof the cell with sputtered Pt CE (8.46%), as shown in Figure. S2in Supporting information. The surface morphology of thesputtered Pt CE is also shown in Figure S2 in Supportinginformation. It can be observed that Pt nanoparticles areuniformly distributed on the FTO substrate and the effectivecatalytic area of the Pt CE is mainly determined by theroughness of FTO layers. The high efficiency of the DSSC withGD–PEDOT:PSS CEs is due to its high JSC and FF, with referencesto that of the cell with PEDOT:PSS CEs. High JSC of the DSSCwith GD–PEDOT:PSS CE may be attributed to more activeelectrocatalytic sites on rough catalytic GD–PEDOT:PSS filmsfor reducing triiodide (I3

�) ions, compared to that in the flatPEDOT:PSS films. A faster movement of the redox couple isexpected in the electrolyte of the DSSC with GD–PEDOT:PSS CEsdue to the fast reduction of I3

� ions at the CE of the cell, whichin turn can lead to fast electron transfer kinetics in the cell andto a high FF for the cell (0.70), compared to that of the cell

with PEDOT:PSS CE (0.60). These mechanisms will be discussedand verified later through CV and EIS.

Herein, the CV analysis was performed to understand thereaction kinetics and electrocatalytic activities of the CEswith PEDOT:PSS and GD–PEDOT:PSS for I�/I3

� redox reactionby using a three-electrode electrochemical system. Figure 3b

C.-P. Lee et al.114

displays the CV curves of the electrodes with PEDOT:PSS andGD–PEDOT:PSS. Table 2 lists the corresponding cathodiccurrent density. The electrode with GD–PEDOT:PSS shows ahigher cathodic current density (�1.58 mA cm�2), corre-sponding to the reduction reaction of I3

–, than that of theelectrode with PEDOT:PSS (�0.39 mA cm�2). A higher catho-dic current density of the CE with GD–PEDOT:PSS can beattributed to its higher electrocatalytic surface area andconductivity for this CE for I3

� reduction via the introductionof GDs, than those of PEDOT:PSS. The peak separation (ΔEP)values of the CV curves are recorded in Table 2. The ΔEPvaries inversely with the charge transfer rate (kS) [8]. Thepeak separation for the redox reaction of I�/I3

� ions at theGD–PEDOT:PSS CE (347 mV) is smaller than that for thereduction at the PEDOT:PSS CE (436 mV), indicating that thecharge transfer kinetics at the GD–PEDOT:PSS CE are faster,echoing the higher performance of corresponding DSSCs.

0 20 40 60 80 1003

6

9

12

Si solar cell DSSC with Pt CE DSSC with GD-PEDOT:PSS CE

η(%

)

Light intensity (mW cm-2

)

-90 -60 -30 0 30 60 900

3

6

9

12

15

η (%

)

AOI (degree)

Si solar cell DSSC with Pt CE DSSC with GD-PEDOT:PSS CE

Figure 4 (a) Illumination intensity dependence of cell efficiencyshown in (b). (c) Incident angle dependence of cell efficiency formeasurement is also shown in the figure as an inset; the correspon

Table 2 Electrochemical parameters for differentcounter electrodes.

CE with Cathodic current den-sitya (mA cm�2)

ΔEPa

(mV)RCt

b

(ohm)

PEDOT:PSS �0.39 436 10.38GD–PEDOT:

PSS�1.58 347 7.92

aThe values are obtained from CV;bThe values are obtained from EIS.

In order to further investigate the charge transfer resis-tances (RCt) for the reduction of I3

� ions at the interfaces ofPEDOT:PSS and GD–PEDOT:PSS with the electrolyte, we per-formed EIS analysis with symmetric sandwich-type cells havingtwo identical electrodes. Figure 3c shows Nyquist plots of thesymmetric sandwich-type cells with PEDOT:PSS and GD–PEDOT:PSS electrodes, and the equivalent circuit is shown in the insetof Figure 3c. An EIS of a symmetric sandwich-type cell can bedivided into two parts. The first semicircle in the highfrequency range (10–105 Hz) represents the RCt at the elec-trode/electrolyte interface. The second semicircle in the lowfrequency range (0.1–10 Hz) represents the Warburg diffusionresistance (ZW) within the electrolyte. The RCt values of thesymmetric sandwich-type cells with PEDOT:PSS and GD–PEDOT:PSS electrodes are listed in Table 2. The symmetric sandwich-type cells with PEDOT:PSS and GD–PEDOT:PSS electrodespossess RCt values of 10.38 and 7.92 Ω, respectively. A lowerRCt values is observed for the symmetric sandwich-type cellwith GD–PEDOT:PSS electrodes due to its higher electrocataly-tic ability for the reduction of I3

� ions. This result shows thatGD–PEDOT:PSS catalytic films can effectively catalyze thereduction of I3

� ions to I� ions due to low charge transferresistances at the electrode/electrolyte interface, leading tothe high performance of the DSSC with GD–PEDOT:PSS CEs.

In literatures, almost all of photovoltaics (PVs) characteriza-tions were discussed with normal incident light. However, theconversion of light into electric energy occurring at low-light (inthe fog or on overcast days, for example) should be consideredfor practical PVapplications. Additionally, in real life, the AOI ofsun light is always changed by time and depends on theinstallation location. Therefore, illumination intensity depen-dence of cell efficiency, incident angle dependence of cell

0 20 40 60 80 1000.2

0.4

0.6

0.8

1.0

Si solar cell DSSC with Pt CE DSSC with GD-PEDOT:PSS CEN

orm

aliz

ed e

ffic

ienc

y

Light intensity (mW cm-2

)

-90 -60 -30 0 30 60 900.0

0.5

1.0

Nor

mal

ized

eff

icie

ncy

AOI (degree)

Si solar cell DSSC with Pt CE DSSC with GD-PEDOT:PSS CE

for various solar cells; the corresponding normalized data arevarious solar cells; a schematic of incident angle dependent

ding normalized data are shown in (d).

Scheme 1 A schematic of the charge transport in a DSSC.

0

2

4

6

8

IVIIII

Si solar cell DSSC with Pt CE DSSC with GD-PEDOT:PSS CE

η (%

)

II

(I) (II)

(III) (IV)

Figure 5 The real-world case with weak light and large AOIconditions: (a) the photographs of the location for the mea-surements and (b) corresponding efficiencies. (I) white-lightLED (0.1 W; 10 cm distance), (II) the noon of a cloudy day (9 Dec2013), (III) the early morning and (IV) the nightfall of a sunnyday (11 May 2014) at the outdoor location with latitude of25.019526 and longitude of 121.534621.

115Economical low-light photovoltaics by using the Pt-free dye-sensitized solar cell with graphene dot/PEDOT:

efficiency, and cell performance under practical ambientconditions for the fabricated DSSCs are investigated andcompared with those with conventional Si solar cells.Figure 4a and c shows the illumination intensity dependenceand AOI dependence of cell efficiency measured under simu-lated solar light (AM 1.5 G, 100 mW cm�2), and the correspond-ing normalized data are shown in Figure 4b and d, respectively.Moreover, all of their photovoltaic parameters are shown inFigures S3 and S4 in Supporting information. As shown inFigure 4a, our DSSCs show better performance than the Si solar

cell when the light intensity is lower than 13.5 mW cm�2, andFigure 4b reveals that fabricated DSSCs show less decreasedegree of cell efficiency than the Si solar cell as the lightintensity is decreased. In Figure 4c, the cell efficiency of solarcells is recorded under different incident angles as shown in theinset of Figure 4c. We found that our DSSCs show betterperformance when the angle of incident (AOI) is larger than601, as compared with conventional Si solar cells, and thecorresponding normalized data (Figure. 4d) indicates fabricatedDSSCs are less sensitive to AOI. These results well consist withthe previous reports [1–4], which reported that the cellefficiency of conventional Si solar cells strongly depend onAOI, and they exhibit poor performance under low-lightirradiation. On the contrary, the efficiency of DSSCs are lessaffected by the light intensity and less sensitive to the AOI[1,5], which make DSSCs to be the most efficient photo-to-electron conversion devices under low-light irradiation. Thesuperior performances of DSSCs under low-light irradiation orunder large AOIs are due to the significant increases in the FFvalues whereas the Si solar cell shows decreased FF values(Figures S3 and S4 in supporting information). Actually, underlow-light irradiation, surface and bulk recombination in Sisemiconductor becomes serious due to the unfilled defects.Thus, the FF of Si solar cell is decreased because of the highenergy losses, resulting in poor cell performance. The tradi-tional Si solar cell would suffer a cutoff at some lowerillumination limit when charge carrier mobility is low andrecombination becomes a major issue. However, the semicon-ductor in a DSSC is merely to conduct the injected majoritycharge carriers and there are no minority carriers involving inthe photoconversion process of DSSCs [1]. Therefore, energylosses, encountered in Si solar cell, are not observed in DSSCsunder low-light or at large AOI. The cutoff in a DSSC is very lowso DSSCs are even been proposed for indoor use, collectingenergy for small devices from the lights in the house.

On the other hand, the DSSCs with GD–PEDOT:PSS CE andwith Pt CE exhibit similar η when the light intensity is lowerthan 13.5 mW cm�2 or the AOI is larger than 601, which can beunderstood from the charge transport kinetics in a DSSC.Scheme 1 shows the schematics of the charge transportkinetics in a DSSC under light illumination. In the case oflow-light irradiation, less oxidized dyes (n h+) are generated,and then those oxidized dyes would be quickly regenerated bythe I� ions in the electrolyte, thus less I3

� ions should beregenerated at the surface of the counter electrode. There-fore, under low-light irradiation, the GD–PEDOT:PSS CE iscapable of providing sufficient charges (m≒n) for the reductionof I3

– ions and showing a comparable performance to the Pt CEsdespite it usually has larger overpotential than Pt CE. Theabove photovoltaic performances of the solar cells were allperformed under the simulated solar light. Herein, we alsoinvestigated the cell performances under various real cases ofindoor light (white light LEDs) (Figure 5a (I)), the noon of acloudy day (Figure 5a (II)), the early morning (Figure 5a (III))and the nightfall (Figure 5a (IV)) of a sunny day. As shown inFigure 5b, the DSSCs show better performance than Si solarcells under mostly real conditions despite the later exhibit thehighest cell efficiency (Figure S5a in Supporting information) inthe noon of a sunny day (Figure S5b in Supporting information).The DSSCs with GD–PEDOT:PSS CEs and Pt CEs show the similarperformance under these field test environments. Briefly, thisfield tests demonstrate that, for the use of photo-to-electron

C.-P. Lee et al.116

conversion device at large AOIs or under low-light irradiation,the cost-effective GD–PEDOT:PSS CEs, as compared to expen-sive Pt CEs, pave the way for commercializing DSSCs and thusheighten the competitiveness of DSSCs in solar cell markets.

4. Conclusions

The GD–PEDOT:PSS film with a rough surface, high electronconductivity, excellent electrocatalytic activity and lowcharge-transfer resistance toward I�/I3

� redox achieves a ηup to 7.36%, which is much higher than that of cell withpristine PEDOT:PSS CE (5.14%) and is close to that of cell withPt CE (8.46%). The DSSCs with The GD–PEDOT:PSS CEs showsuperior performance under low-light conditions (light intensityo13.5 mW cm�2 and AOI 4601), as compared with conven-tional Si solar cell. This study indicates that GD–PEDOT:PSS asan Pt-free CE for DSSCs paves the way for efficient light-harvesting devices for intelligent building and smart home.

Acknowledgment

This work was financially supported by KAUST baseline fund,the Research Grants Council of Hong Kong (Project no. PolyU153012/14P), the National Natural Science Foundation ofChina (Grant no. 11374250), and National Taiwan University(102R4000).

Appendix A. Supplementary material

Supplementary data associated with this article can be foundin the online version at http://dx.doi.org/10.1016/j.nanoen.2015.10.008.

References

[1] H.J. Kim, Y.C. Kim, J.T. Hong, M.J. Kim, H.W. Seo, J.W. Park,J.Y. Choi, J. Electr. Eng. Technol. 2 (2007) 513–517.

[2] W.-R. Wei, M.-L. Tsai, S.-T. Ho, S.-H. Tai, C.-R. Ho, S.-H. Tsai, C.-W. Liu, R.-J. Chung, J.-H. He, Nano Lett. 13 (2013) 3658–3663.

[3] H.-P. Wang, T.-Y. Lin, C.-W. Hsu, M.-L. Tsai, C.-H. Huang, W.-R. Wei, M.-Y. Huang, Y.-J. Chien, P.-C. Yang, C.-W. Liu, L.-J. Chou, J.-H. He, ACS Nano 7 (2013) 9325–9335.

[4] H.-P. Wang, T.-Y. Lin, M.-L. Tsai, W.-C. Tu, M.-Y. Huang, C.-W. Liu, Y.-L. Chueh, J.-H. He, ACS Nano 8 (2014) 2959–2969.

[5] H. Tian, X. Yu, J. Zhang, W. Duan, F. Tian, T. Yu, Int. J.Electrochem. Sci. 7 (2012) 4686–4691.

[6] M. Wu, X. Lin, Y. Wang, L. Wang, W. Guo, D. Qi, X. Peng,A. Hagfeldt, M. Grätzel, T. Ma, J. Am. Chem. Soc. 134 (2012)3419–3428.

[7] M.-H. Yeh, L.-Y. Lin, C.-P. Lee, H.-Y. Wei, C.-Y. Chen, C.-G. Wu,R. Vittal, K.-C. Ho, J. Mater. Chem. 21 (2011) 19021–19029.

[8] M. Wu, T. Ma, ChemSusChem 5 (2012) 1343–1357.[9] T.-L. Zhang, H.-Y. Chen, C.-Y. Su, D.-B. Kuang, J. Mater. Chem.

A 1 (2013) 1724–1730.[10] M.H. Yeh, C.P. Lee, L.Y. Lin, P.C. Nien, P.Y. Chen, R. Vittal, K.

C. Ho, Electrochim. Acta 56 (2011) 6157–6164.[11] C. Bu, Q. Tai, Y. Liu, S. Guo, X. Zhao, J. Power Sour. 221 (2013)

78–83.[12] Q. Qin, J. Tao, Y. Yang, X. Dong, Polym. Eng. Sci. 51 (2011)

663–669.

[13] T.N. Murakami, S. Ito, Q. Wang, M.K. Nazeeruddin, T. Bessho,I. Cesar, P. Liska, R. Humphry-Baker, P. Comte, P. Péchy,M. Grätzel, J. Electrochem. Soc. 153 (2006) A2255–A2261.

[14] G. Veerappan, K. Bojan, S.-W. Rhee, ACS Appl. Mater. Interfac.3 (2011) 857–862.

[15] L.-Y. Chang, C.-P. Lee, K.-C. Huang, Y.-C. Wang, M.-H. Yeh, J.-J. Lin, K.-C. Ho, J. Mater. Chem. 22 (2012) 3185–3191.

[16] M.-H. Yeh, C.-L. Sun, J.-S. Su, L.-Y. Lin, C.-P. Lee, C.-Y. Chen,C.-G. Wu, R. Vittal, K.-C. Ho, Carbon 50 (2012) 4192–4202.

[17] C.-H. Chiang, C.-G. Wu, Org. Electron. 14 (2013) 1769–1776.[18] Y. Xu, Y. Wang, J. Liang, Y. Huang, Y. Ma, X. Wan, Y. Chen, Nano

Res. 2 (2009) 343–348.[19] J. Ouyang, C.W. Chu, F.C. Chen, Q. Xu, Y. Yang, Adv. Funct.

Mater. 15 (2005) 203–208.[20] J.-G. Chen, H.-Y. Wei, K.-C. Ho, Sol. Energy Mater. Sol. Cells 91

(2007) 1472–1477.[21] K. Kitamura, S. Shiratori, Nanotechnology 22 (2011) 195703.[22] G. Yue, J. Wu, Y. Xiao, J. Lin, M. Huang, Electrochim. Acta 67

(2012) 113–118.[23] B. Fan, X. Mei, K. Sun, J. Ouyang, Appl. Phys. Lett. 93 (2008)

143103.[24] W. Hong, Y. Xu, G. Lu, C. Li, G. Shi, Electrochem. Commun. 10

(2008) 1555–1558.[25] Y. Xue, J. Liu, H. Chen, R. Wang, D. Li, J. Qu, L. Dai, Angew.

Chem.-Int. Edit. 51 (2012) 12124–12127.[26] J.D. Roy-Mayhew, D.J. Bozym, C. Punckt, I.A. Aksay, ACS Nano

4 (2010) 6203–6211.[27] H. Wang, T. Maiyalagan, X. Wang, ACS Catal. 2 (2012) 781–794.[28] M.J. Ju, J.C. Kim, H.-J. Choi, I.T. Choi, S.G. Kim, K. Lim, J. Ko, J.-

J. Lee, I.-Y. Jeon, J.-B. Baek, H.K. Kim, ACS Nano 7 (2013)5243–5250.

[29] J. Shen, Y. Zhu, X. Yang, C. Li, Chem. Commun. 48 (2012)3686–3699.

[30] H. Sun, L. Wu, N. Gao, J. Ren, X. Qu, ACS Appl. Mater.Interfac. 5 (2013) 1174–1179.

[31] S. Zhu, J. Zhang, C. Qiao, S. Tang, Y. Li, W. Yuan, B. Li, L. Tian,F. Liu, R. Hu, H. Gao, H. Wei, H. Zhang, H. Sun, B. Yang, Chem.Commun. 47 (2011) 6858–6860.

[32] D.I. Son, B.W. Kwon, D.H. Park, W.-S. Seo, Y. Yi, B. Angadi, C.-L. Lee, W.K. Choi, Nat. Nanotechnol. 7 (2012) 465–471.

[33] L. Tang, R. Ji, X. Cao, J. Lin, H. Jiang, X. Li, K.S. Teng, C.M. Luk, S. Zeng, J. Hao, S.P. Lau, ACS Nano, 6, 5102–5110.

[34] V. Gupta, N. Chaudhary, R. Srivastava, G.D. Sharma, R. Bhardwaj,S. Chand, J. Am. Chem. Soc. 133 (2011) 9960–9963.

[35] J.K. Kim, M.J. Park, S.J. Kim, D.H. Wang, S.P. Cho, S. Bae, J.H. Park, B.H. Hong, ACS Nano 7 (2013) 7207–7212.

[36] E. Lee, J. Ryu, J. Jang, Chem. Commun. 49 (2013) 9995–9997.[37] L. Chen, C.X. Guo, Q. Zhang, Y. Lei, J. Xie, S. Ee, G. Guai, Q. Song,

C.M. Li, ACS Appl. Mater. Interfac. 5 (2013) 2047–2052.[38] L.A. Ponomarenko, F. Schedin, M.I. Katsnelson, R. Yang, E.

W. Hill, K.S. Novoselov, A.K. Geim, Sci. 320 (2008) 356–358.[39] M. Li, W. Ni, B. Kan, X. Wan, L. Zhang, Q. Zhang, G. Long, Y. Zuo,

Y. Chen, Phys. Chem. Chem. Phys. 15 (2013) 18973–18978.[40] L. Han, N. Koide, Y. Chiba, A. Islam, T. Mitate, C.R. Chim., 9

(2006) 645–651.[41] L. Han, N. Koide, Y. Chiba, T. Mitate, Appl. Phys. Lett. 84

(2004) 2433–2435.

Dr. Chuan-Pei Lee received his PhD degreein Chemical Engineering at National TaiwanUniversity in 2012. Currently, he is a post-doctoral research fellow in National TaiwanUniversity. His research interests includesolar energy and electrochemical energymaterials/systems. Besides, he is also famil-iar with electrochemical-analysis andmicroemulsions-synthesis techniques.

117Economical low-light photovoltaics by using the Pt-free dye-sensitized solar cell with graphene dot/PEDOT:

Dr. Chin-An Lin received his B.S degree inPhysics from the National Kaohsiung NormalUniversity, Kaohsiung, Taiwan, in 2006. In2007, he joined Dr. Jr- Hau He's group at theInstitute of Photonics and Optoelectronics,National Taiwan University, Taipei, Taiwan,and received his M.S. degree in 2008 andPh. D degree in 2014. His current researchinterests include antireective coatings insolar cells and dye-sensitized solar cells.

Mr. Tzu-Chiao Wei received his B.S. degreein physics from Fu Jen Catholic University,Taipei, Taiwan, in 2008 and M.S. degree inoptics and photonics from National CentralUniversity, Taoyuan, Taiwan, in 2010. Afterthat, he entered the Graduate Institute ofPhotonics and Optoelectronics at NationalTaiwan University, Taipei, Taiwan, for doc-toral study under supervision of ProfessorJr-Hau He. His doctoral research areas

include photo sensors, photoelectronic devices, Raman spectro-scopy, ferromagnetism, nano-structured composite thin film andtheir applications.

Mr. Meng-Lin Tsai received his B.S. degreefrom the National Tsing Hua University,Hsinchu, Taiwan. He is currently a PhDstudent in Department of Materials Scienceand Technology, National Tsing Hua Univer-sity, Hsinchu, Taiwan. His research interestsinclude the fabrication of 2D materialdevices, novel photodetecting and photo-voltaic devices, and electron beam litho-graphy techniques.

Miss Ying Meng received her B.S. degreefrom the school of EE at Shanghai JiaotongUniversity and M.S. degree in School ofSoftware and Microelectronics, Peking Uni-versity. From 2012 to 2013, she has been avisiting scholar in the institute of Photonicsand Optoelectronics, National Taiwan Uni-versity. Her research interests are nanoma-terials for clean energy such as electrodematerials for dye-sensitized solar cells.

Dr. Chun-Ting Li received her BS and MSdegrees in Department of Chemical andMaterial Engineering at Chang Gung Univer-sity, Taiwan, in 2010 and in 2011, respec-tively. She received her PhD degree inDepartment of Chemical Engineering atNational Taiwan University, Taiwan, in2015. Her research interests principallyfocus on synthesizing and developingelectro-catalyst materials in electrochemi-

cal devices, including dye-sensitized solar cells as well as energyconversion & storage materials/systems with particular attention totransition metal compounds. Besides, her specialty is material andelectrochemical analyses techniques.

Professor Kuo-Chuan Ho received his BS andMS degrees in Department of ChemicalEngineering from National Cheng Kung Uni-versity, Taiwan in 1978 and 1980, respec-tively. He received his PhD in ChemicalEngineering at the University of Rochester,USA in 1986. Currently he is a DistinguishedProfessor jointly appointed by the Depart-ment of Chemical Engineering and Instituteof Polymer Science and Engineering at

National Taiwan University. His research interests mainly surroundapplications of chemically modified electrodes to sensing andelectro-optical devices, including dye-sensitized solar cells andelectrochromic devices.

Professor Chih-I Wu got his B.S. degree fromNational Taiwan University in 1990 and M.S.degree from Northwestern University in1994, both in Physics. Then he went to theDepartment of Electrical Engineering atPrinceton University, where he received hisPh.D. degree in 1999. He joined the Grad-uate Institute of Electro-Optical Engineer-ing and the Department of ElectricalEngineering of National Taiwan University

in 2004. His main research area focuses on optical-electronicdevices and materials and semiconductor physics, which includesorganic light-emitting materials, metal-semiconductor interfaces,and heterojunctions in electronic devices and optical-electronics.

Professor Shu-Ping Lau obtained his Ph.Ddegree in Materials Engineering from Uni-versity of Swansea in 1995. He joinedNanyang Technological University (NTU) asAssistant Professor in 1998. He joined theHong Kong Polytechnic University (PolyU) in2008 as a Professor. He is now the head ofthe Department of Applied Physics at PolyU.He has published 4 invited book chapters,and over 250 international refereed papers.

His current research interest is in two-dimensional materialsincluding graphene, graphene oxide, MoS2 and black phosphorous.

Professor Jr-Hau He received his B.S. andPh.D. degrees from the National Tsing HuaUniversity, Hsinchu, Taiwan, in 1999 and2005, respectively. He is currently anAssociate Professor of Electrical Engineeringprogram, King Abdullah, University ofScience and Technology (KAUST), Kingdomof Saudi Arabia. He is involved in the designof new nanostructured architectures fornanophotonics and the next-generation

nanodevices, including photovoltaics, and resistive memory.