A Enm 201401185

8/9/2019 A Enm 201401185

1/12

© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim (1 of 12)1401185wileyonlinelibrary.com

Integrated Design of Organic Hole Transport Materialsfor Efcient Solid-State Dye-Sensitized Solar CellsBo Xu, Haining Tian, Lili Lin, Deping Qian, Hong Chen, Jinbao Zhang, Nick Vlachopoulos,Gerrit Boschloo, Yi Luo, Fengling Zhang, Anders Hagfeldt, and Licheng Sun*

DOI: 10.1002/aenm.201401185

suffer from potential leakage problemsassociated with the volatile nature of theliquid electrolyte, limiting this technologyfor large-scale applications. [3] In an effortto address these issues, a p-type organicsemiconductor termed 2,2 ′,7,7′-tetrakis-(N,N-di-p-methoxyphenyl-amine)9,9′ spiro-biuorene (Spiro-OMeTAD) was developedto replace the liquid electrolyte as a redox

couple in the fabrication of “solid-state”dye-sensitized solar cells (ssDSCs) byBach et al. in 1998. [4] To date, the highestefciency of 7.2% for ssDSCs have beenachieved using Spiro-OMeTAD as thehole transport material (HTM), which wasreported by Nazeeruddin and co-workers. [5] Recently, Snaith and co-workers employed

an organometallic halide perovskite with Spiro-OMeTAD asHTM obtained a record efciency of 15.4%. [6] However, pre-vious studies have demonstrated that the low hole mobility andconductivity of Spiro-OMeTAD signicantly limits the deviceperformance near the maximum power point in solar cell. [7,8] In addition, the lengthy synthetic route of Spiro-OMeTAD maymake it impractical for large-scale application in ssDSCs. [9] More importantly, the substantial overpotential required for dye

A series of triphenylamine-based small molecule organic hole transport mate-rials (HTMs) with low crystallinity and high hole mobility are systematicallyinvestigated in solid-state dye-sensitized solar cells (ssDSCs). By using theorganic dye LEG4 as a photosensitizer, devices with X3 and X35 as the HTMsexhibit desirable power conversion efciencies (PCEs) of 5.8% and 5.5%,respectively. These values are slightly higher than the PCE of 5.4% obtainedby using the state-of-the-art HTM Spiro-OMeTAD. Meanwhile, transient

photovoltage decay measurement is used to gain insight into the complexinuences of the HTMs on the performance of devices. The results demon-strate that smaller HTMs induce faster electron recombination in the devicesand suggest that the size of a HTM plays a crucial role in device performance,which is reported for the rst time.

1. IntroductionDye-sensitized solar cells (DSCs) have emerged as a potentiallow-cost alternative energy solution compared to the typicalsilicon-based p-n junction solar cells since the rst report byO′Regan and Grätzel in 1991. [1] Power conversion efciencies(PCEs) exceeded 12% has been obtained with DSCs utilizingcobalt based complex redox mediators. [2] However, such cells

B. Xu, Dr. H. Tian, Prof. L. SunOrganic ChemistryCenter of Molecular DevicesDepartment of ChemistrySchool of Chemical Science and EngineeringKTH Royal Institute of Technology10044 Stockholm, SwedenE-mail: [email protected]. H. Tian, J. Zhang, Dr. N. Vlachopoulos,Dr. G. Boschloo, Prof. A. HagfeldtDepartment of Chemistry-Ångström LaboratoryPhysical ChemistryUppsala UniversityBox 523, 751 20, Uppsala, SwedenDr. L. LinCollege of Physics and ElectronicsShandong Normal University250014 Jinan, ChinaDr. L. Lin, Prof. Y. LuoDepartment of Theoretical Chemistry & BiologySchool of BiotechnologyKTH Royal Institute of Technology106 91 Stockholm, Sweden

D. Qian, Prof. F. ZhangBiomolecular and Organic ElectronicsDepartment of PhysicsChemistry and BiologyLinköping University581 83 Linköping, SwedenH. ChenBerzelii Center EXSELENT on Porous Materialsand Department of Materials and Environmental ChemistryStockholm University106 91 Stockholm, SwedenProf. L. SunState Key Laboratory of Fine ChemicalsDUT-KTH Joint Research Center on Molecular DevicesDalian University of Technology (DUT)116024 Dalian, China

Adv. Energy Mater . 2015, 5, 1401185

www.MaterialsViews.comwww.advenergymat.de

http://doi.wiley.com/10.1002/aenm.201401185

8/9/2019 A Enm 201401185

2/12

© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1401185 (2 of 12) wileyonlinelibrary.com

regeneration by Spiro-OMeTAD limits the maximum obtainablevoltage of the system; therefore, the open-circuit voltage has to

be tuned by adjusting the highest occupied molecular orbital(HOMO) of the HTM [10] (Figure 1 ). In an effort to solve theseproblems, many new HTMs have been proposed to improvehole mobility and optimize HOMO levels, yet the measureddevice efciencies still remain lower than Spiro-OMeTAD. [11–15] Most recently, some studies have used thiophene-based mon-omers to ll the pores followed by electropolymerization. [16] However, this method requires special processing proceduresthat may be difcult to actualize in large-scale production pro-cesses. Therefore, it would be of great importance to developnew alternatives and systematically investigate the effect of theproperties of different HTMs on the performance of ssDSCs.

In this paper, two new triphenylamine-based small mole-cule organic HTMs X2 and X35 with low cost, high solubilityand low crystallinity were synthesized and characterized. Thethermal properties, electrochemical properties and photophys-ical properties of these two new HTMs together with our earlierreported X1 [8] and X3 [9] were systematically investigated and cal-culated (Figure 2 ).

2. Results and Discussion2.1. Design and Synthesis

The materials design rationale was to prepare HTMs with dif-ferent triphenylamine units while systematically investigating

the effect of the HOMO level, charge carriers mobility and mol-ecule size on performance of the devices. The synthetic routesof X1 , X2 and X35 are shown in Scheme 1 . The synthetic proce-dures of X1 and X2 were straightforward. Inexpensive startingmaterials were utilized to synthesize the desired materials X1 and X2 by one step Palladium-catalyzed Buchwald Hartwigreaction, rendering overall yield up to 95% and 90% after nalcolumn chromatography purication, respectively. Meanwhile,we designed a star burst HTM X35 with the same molecularweight with X3 [9] but different molecular structure, which wassynthesized through two steps reactions (Buchwald Hartwigreaction and Suzuki coupling-cross reaction) with the nal yield

of 55%. All of the HTMs employ methoxy groups which arereported to have a high tendency to stabilize the radical cationsand also increase the hole mobility and solubility. [15,20,21] Thematerials were characterized by 1 H/ 13 C NMR and high reso-lution mass spectrometry (HR-MS) (Supporting InformationFigure S1–S8). Additionally, Spiro-OMeTAD is employed as areference HTM in this study.

2.2. Thermal Properties

Thermal properties of the HTMs were estimated by differ-ential scanning calorimetry (DSC) (Supporting Informa-tion Figure S9), and the corresponding data are collected inTable S1. The glass transition temperatures ( T g ) of HTMs havebeen determined to be 71.2 °C for X1 , 110 °C for X2 , 137.2 °Cfor X3 and 132.5 °C for X35 . From the DSC curves, we haveobserved that the molecular weight of HTMs brought aboutdrastic change of the T g . The HTMs with similar molecularweight show similar T g values, such as Spiro-OMeTAD, [13] X3

and X35 . The T g value of linear linkage molecule X3 is onlyslightly higher than the branch linkage molecular X35 andSpiro-OMeTAD. We furthermore observed that the cold crystal-lization temperature ( T cc ) of X1 is 156.6 °C, while no signal ofT cc temperature of other HTMs, illustrating that the crystallinetemperature will be lower while increasing the oligomerizationnumber of triphenylamine for HTMs, and then exhibit bettersolubility and amorphous nature. Moreover, the T g of X1 ismuch lower than Spiro-OMeTAD, which has a potential for amelt inltration process in thicker ssDSC devices.

2.3. Photophysical and Electrochemical Properties

The optical absorption and uorescence (Supporting InformationFigure S10) spectra in DCM were measured at a concentrationof 10−5 M. All photophysical data of HTMs are listed in Table 1 . X1 exhibits strong absorption peak at 353 nm and the otherHTMs have similar absorption peaks around 370 nm. In addi-tion, photoluminescence spectra were also recorded (SupportingInformation Figure S10). The measurements show that all ofthe HTMs have a homologous peak around 430 nm. From theintersection of emission and absorption spectra we can obtainthe E0–0 transition and estimate the optical band gap (Table 1 ).

The oxidation potential of HTM and energy level alignmentwith the HOMO level of the dye are crucial parameters for con-structing high-performance ssDSCs. [15] Herein, the oxidation

potentials of the HTMs were determined by solution-basedcyclic voltammetry (CV) (Supporting Information Figure S11) indichloromethane. All of the HTMs have more positive oxidationpotential and lower molecular weight than Spiro-OMeTAD (seeTable 1). Comparing these to the dye LEG4, [22] the HTMs havingoverpotential between 160 mV to 110 mV are not sufcientenough to drive the hole-transfer process in the device accordingto the previous report that an overpotential of approximately180 mV should be sufcient to achieve a 90% efcient holetransfer. [10,23] However, the protonation/deprotonation of sur-face-adsorbed hydroxyl groups has been shown to shift the den-sity of states of the nanocrystalline TiO 2 band and to modulate

Adv. Energy Mater.2015, 5, 1401185


+

EF

E v s

N H E ( V )

+

2

e - e -

+

V m a x

EOXO

e -

e -

D*

D/D +

Figure 1. Schematic diagram and operating principle of a ssDSC.

8/9/2019 A Enm 201401185

3/12


the reduction/oxidation potentials of dyes and HTMs. [12] Inorder to get information on the actual energy levels of the HTMsand dye, we further measured the oxidation potentials of HTMsand dye on TiO 2 electrode (Supporting Information Figure S12)according to a published procedure. [24] In consideration of theoxidation potential value (1.12 V vs NHE) (Table 1 ) of the LEG4dye on the TiO 2 electrode, all hole conductors having overpoten-tials between 370 mV and 190 mV should be able to regeneratethe oxidized dye effectively. Meanwhile, we noted that the oxi-dation potentials of the HTMs are more positive than that of

Spiro-OMeTAD both in solution and on TiO 2 electrode, so thatit can be estimated that the devices with the X-series of HTMsshould exhibit higher open-circuit voltage in the devices uponthe same conditions, as we will discuss afterwards.

2.4. Computational Study

To elucidate the electronic properties of the HTMs, DFT cal-culations were performed at the B3LYP 6–31G (d) level with

Adv. Energy Mater . 2015, 5, 1401185


Figure 2. Molecular structures of LEG4, Spiro-OMeTAD,X1 , X2 , X3 and X35; the molecular weights are given in g mol−1

.

8/9/2019 A Enm 201401185

4/12


Gaussian 09 program. The electronic distribution of the highestoccupied molecular orbital (HOMO) and the lowest unoccupiedmolecular orbital (LUMO) of the HTMs are shown in Sup-

porting Information Figure S13. The HOMOs of these HTMsare almost delocalized in the whole systems, which predictgood transport properties for holes. The LUMOs are typicallylocalized at the biphenyl group, and relative poor electron trans-port mobility can be implied. The results of the calculationshave been summarized in Supporting Information Table S2,which make it clear that the energy of HOMO levels of allHTMs show little difference.

The charge transport rate is one of the most important fac-tors to indicate the transport property of materials, and the gen-eral formula to calculate the hole transfer rate is the Marcusrate equation.

exp4

2 2∓π

λ λ

λ ( )= − ∆ +W V

K T G

K T ji

ji

B

ji

B

(1)

Here, V ji is the transfer integral between site i and site j , andλ is the reorganization energy. ∆ G ji is the site energy differencebetween the site i and site j , and it is approximately zero for thepure molecular crystals. One can see that the transfer integraland the reorganization energy are the most important factorsthat determine the charge transfer rate. The transfer integralindicates the strength of the interaction between two neighborsites (molecules here), and it relies on the relative positions ofthe hopping sites (the network of the lm) and overlap of theorbitals of two molecules involved. Generally, the larger thetransfer integral is, the higher the charge transfer rate becomes.



Scheme 1. The synthetic routes ofX1 , X2 and X35 .

8/9/2019 A Enm 201401185

5/12


The reorganization energy represents the relaxation of the mol-ecule when one electron is withdrawn or donated. The smallreorganization energy implies the fast transfer rate for electronsor holes to some extent. As indicated in Table 2 , the molecule X3 has the smallest calculated reorganization energy and there-fore the relative fast hole transport properties. This is in goodagreement with the measured hole mobilities (see photoelec-trical properties section below). Their good properties might beassociated with the larger conjugation length of X3 and X35 with respect to Spiro-OMeTAD from the structural point of view,which indicate that the design of high hole mobility HTM shouldconsider conjugational aspects within the molecule in the future.

By frozen the dihedral angle of 1–2–3–4 (as illustrated withthe cyan color in the gure), we have optimized the molecule X3 at different angles changing from 140.57 (Trans congu-ration) to 38.79 (Cis conguration) degree. The energy valuesof the molecule with different dihedral angles are shown inFigure 3 . One can deduce that there exists another possible con-guration with the dihedral angle 40.57 degree (named as Vertconguration) additional to the Cis and Trans congurations.The three congurations at the minima are shown in Figure 3 .The energy barrier for the molecule to transform from the Cis

conguration to the Vert conguration is about 0.141 eV, andthe barrier from the Vert conguration to the Trans congura-tion is about 0.156 eV. This implies that the congurations of X3 are chaotic and amorhous in the mesopores TiO 2 , which isbenecial for pore penetration and lm formation.

2.5. Photoelectrical Properties

As previously described, charge transport is another signi-cant parameter to consider in the design of new HTMs forssDSCs. Due to the low mobility of charge carriers in organicsemiconductors, the injected carrier forms a space charge. Thisspace charge creates a eld that opposes the applied bias andthus decreases the voltage drop across junction; as a result,space charge limited currents (SCLCs) have been proposed asthe dominant conduction mechanism in organic semiconduc-tors. [19] Ohmic conduction can be described by

µε ε = J

V d

r 98

0

2

3

(2)

where J is the current density, µ is the hole mobility, ε o is thevacuum permittivity (8.85 × 10−12 F m −1 ), ε r is the dielectric con-stant of the material (normally taken to approach 3 for organicsemiconductors), V is the applied bias, and d is the lm thickness.Fitting the J –V curves (see Supporting Information Figure S14)

for each material to this expression gives the mobility datalisted in Table 3 . The value obtained here for Spiro-OMeTADis similar with the data reported in the literature. [13] The holemobility values of X1 , X2 and X3 are higher than that of Spiro-OMeTAD, while X35 is comparable with Spiro-OMeTAD in thisstudy. We have noted that the hole mobility of X3 is almost oneorder of magnitude higher than that of Spiro-OMeTAD, whichmight contribute to obtain superior performance in the device.

The conductivities of the lms were determined by usingtwo-probe electrical conductivity measurements, which wereperformed by following a published procedure. [25,26] The cur-rent–voltage characteristics curves are showed in SupportingInformation Figure S15, and the measured values for the pureand Li-TFSI-doping HTMs are depicted in Table 3 . The conduc-tivity of pure Spiro-OMeTAD closely matches that reported inthe literature, [25] and all of the HTMs have a similar conduc-tivity below 10−7 S m−1 without Li-TFSI doping. However, theconductivity can be signicantly increased by two orders ofmagnitude upon Li-TFSI-doping. We found that the conduc-tivity of X3 is slightly higher than that of Spiro-OMeTAD, bothunder doping and non-doping circumstances, which might dueto the better conjugated system for the former, leading to moreefcient π –π stacking of HTMs lms. In this case, the conduc-tivity and mobility values of X3 are higher than that of Spiro-OMeTAD while the related values of other HTMs in this seriesare at the same level as for Spiro-OMeTAD, so that at least thehole transfer in our novel HTMs should not limit the device

performance, and possibly could contribute to better photovol-taic parameters than Spiro-OMeTAD-based devices, as we willdiscuss afterwards.

2.6. Photovoltaic Properties

The photovoltaic properties of these HTMs were investigatedby fabricating ssDSCs using the D- π -A dye LEG4.[22] In orderto compare the properties, the device with Spiro-OMeTAD asHTM has been used as reference. The concentration of theHTM plays an important role in ssDSCs, which can signicantly

Adv. Energy Mater . 2015, 5, 1401185


Table 1. Summary of optical and electrochemical of the HTMs used inthis study.

HTMs andDye

λ abs [nm]

λ em [nm]

E ox a) [V vs NHE]

E ox b) [V vs NHE]

E 0–0 c) [eV]

X1 353 434 0.72 0.81 3.12

X2 371 429 0.73 0.89 3.06

X3[9] 373 429 0.77 0.89 3.03

X35 370 432 0.75 0.93 3.05

Spiro-OMeTAD

385 424 0.63 0.75 3.05

LEG4[22] 514 717 0.88 1.12 2.04

a) 0.1 M of tetrabutylammoniun hexauorophosphate (n-Bu4 NPF6 ) in DCM as elec-trolyte; Ag/0.01 M AgNO3 electrode (acetonitrile as solvent) as the reference elec-trode; a glassy carbon disk (diameter 3 mm) as the working electrode; a platinumwire as the counter electrode. Scan rates: 50 mV s−1 . All redox potentials werecalibrated vs normal hydrogen electrode (NHE) by the addition of ferrocene. TheconversionE (Fc/Fc+) = 630 mV vs NHE;b) Ionic liquid 1-butyl-1-methylpyrrolidiniumbis(triuoromethyl-sulfonyl) imide as electrolyte; Ag/AgCl (3 M NaCl) as the refer-ence electrode; FTO/TiO2 /HTMs (spin coating with concentration of 5 mg/100 µL

in chlorobenzene) as the working electrode (area: 1 cm2 ); a stainless steel plate asthe counter electrode. Scan rates: 50 mV s−1 . All redox potentials were calibrated

vs normal hydrogen electrode (NHE) by the addition of ferrocene. The conversionE (Fc/Fc+) = 630 mV vs NHE;c) Calculated from the intersection of the normalizedabsorption and emission spectra.

Table 2. The reorganization energy of the different HTMs calculated withthe four-point method based on the adiabatic potential energy surface.

HTMs X1 X2 X3 X35 Spiro-OMeTAD

Reorganization energy (meV) 216 154 119 129 147

8/9/2019 A Enm 201401185

6/12


inuence the device performance. [27] All of the HTMs in thisstudy are well soluble in many common organic solvents, suchas toluene, tetrahydrofuran (THF), dichloromethane (DCM),and chloroform. However, we found that, when higher concen-trations of the X-HTMs were used in the device, inferior perfor-mances were obtained in initial optimization process. Hence,the nal concentrations of the X-HTMs were xed at around110 mg mL −1 after optimizing various parameters of the deviceregarding optimal photovoltaic operation, which was slightlylower than the optimal concentration for the Spiro-OMeTAD-based device (180 mg mL−1 ) for ssDSCs. [20] In addition, to con-rm the thickness of the blocking layer and TiO 2 lm as wellas to check the inltration of HTM in TiO 2 pores, we haverecorded SEM images of X-HTM based devices, as shown inSupporting Information Figure S16. The thickness of blocking

layer has been determined to be around 150 nm, while thethickness of the TiO 2 lm is around 2.0–2.2 µm.

It has been widely demonstrated that lithiumbis(triuoromethanesulfonyl)imide (Li-TFSI) needs to be addedin the HTM to efciently generate photocurrent in ssDSCs. [25,26] Furthermore, Snaith and co-workers have demonstrated thatLi-TFSI is also a strong and stable p-dopant for HTMs, [28] andour previous work has shown that different HTMs need dif-ferent concentration of Li + doping. [8,9] Therefore, we rstly opti-mized the amount of Li-TFSI added into the HTMs solution.The nal optimal Li + concentration for X1 , X2 , X3 [9] and X35 based ssDSCs were 60 mM, 50 mM, 30mM and 30 mM, whilethe optimal concentration of Li-TFSI with Spiro-OMeTAD was20 mM. [29] The J –V curves of the ssDSCs investigated in thisstudy are shown in Figure 4 a; the best efciencies obtainedby the ssDSCs based on these HTMs are depicted in Table 4 .The X3 -based device yielded the highest efciency of up to

5.8% among these HTMs with short-circuit photocurrent den-sity ( J SC ) of 9.52 mA·cm

−2 , open-circuit photovoltage ( V OC ) of0.91 V and ll factor (FF ) of 0.67. Under the optimal condition,the X35 -based device gave a slightly lower efciency of 5.5%,with J SC of 9.81 mA cm

−2 , V OC of 0.89 V and FF of 0.63. Thestandard HTM Spiro-OMeTAD-based device reached a refer-ence efciency of 5.4% with J SC of 9.37 mA cm

−2 , V OC of 0.93 Vand FF of 0.62. For devices made with X1 and X2 , we foundthe J SC , V OC and FF value of 9.47 and 9.79 mA cm

−2 , 0.75 and0.81 V, and 0.62 and 0.63, yielding PCEs of 4.4% and 5.0%,respectively. Interestingly, we have observed that the photocur-rent density of all the devices can reach the same level, around



Figure 3. Potential energy surface ofX3 with different dihedral angles and the congurations of its isomers.

Table 3. Hole mobility and conductivity of the HTMs used in this study.

HTMsa) Hole mobility noneLi+ doping

[cm2 V−1 s−1 ]

Conductivity noneLi+ doping[S cm−1 ]

Conductivity30 mM Li+ [S cm−1 ]

X1 6.19× 10−5 4.95× 10−7 3.43× 10−5

X2 9.82× 10−5 2.75× 10−7 6.96× 10−5

X3 1.47× 10−4 6.98× 10−7 1.99× 10−4

X35 1.34× 10−5 4.47× 10−7 1.20× 10−4

Spiro-OMeTAD 1.67× 10−5 3.54× 10−7 1.37× 10−4

a) The concentrations of the HTMs were the same as in case of photovoltaicdevices.

8/9/2019 A Enm 201401185

7/12


9.5 mA cm−2 after optimizing the concentration of Li-TFSI. Theincident photon-to-current conversion efciency (IPCE) spectraof these homologous devices were shown in Figure 4 b. ForLEG4 dye, all of these HTMs exhibit the similar IPCE of 75%at the maximum absorption wavelength of 475 nm. However,the open-circuit voltages of the X-HTMs were still lower thanthat of the Spiro-OMeTAD-based device, although the redoxpotential of the X-HTMs is more positive than that of the latter,both in solution and on TiO 2 electrode, which was contrary toour previous expectation. This suggests that other parametersbeyond the oxidation potential might play an important role ofthe V OC in ssDSCs.

It is well known that both liquid-electrolyte DSCs andssDSCs, an enhancement of device performance have been

found upon addition of lithium salts, such as Li-TSFI, LiClO 4 .These additives are known to modulate the energetics of thecell components and hence inuence the dynamics and yieldsof the interfacial charge-transfer reactions. [30] More specically,lithium ions are known to shift the TiO 2 conduction-band edgeto more-positive potentials, and to screen electrostatic interac-tions, resulting in increased charge-injection efciency andreduced recombination, [31] whereas 4-tert-butyl pyridine (t-BP)shifts the conduction-band edge to more negative potentials,leading to an increase of the device V OC .[32] In our case, all ofthe X-HTMs showed higher oxidation potentials than Spiro-OMeTAD but lower open-circuit photovoltage in the devices,

which might be due to the different concentration of Li-TSFI.Therefore, we further fabricated devices with the same concen-tration of Li-TFSI at 20 mM or 30 mM Li-TFSI with the sameconcentration t-BP (200 mM). The J –V curves of the ssDSCswere shown in Figure 5 a,c and all of the photovoltaic param-eters were collected in Table 5 . When the doping concentra-tion of Li-TFSI is 30 mM, X3 based-devices showed the highestpower conversion efciencies of 5.8% among the HTMs, whilethe efciency of X 35 -based device was lower than that of the X3 -based device, but slightly higher than that of the deviceincorporating Spiro-OMeTAD. Meanwhile, the ll factor andphotocurrent of X3 , X35 and Spiro-OMeTAD-based devicesare higher than that of X1 and X2 based ones, which could becaused by the higher conductivity of the former compared tothat of X1 and X2 , due to the fact that the higher conductivityof HTM can decrease the hole transfer resistance in cell. More-over, the V OC of X1- and X2- based devices are much lower thanthat of X3- and X35 -based devices, which might be in virtue ofthe small molecule size of X1 and X2 . Because the small HTMmolecules are not only easily to permeate into the meso-TiO 2

pore, but also easily to permeate into the lacunas between dyemolecules and contact with TiO 2 surface, then the recombina-tion process between injected electrons and oxidized HTMsis faster. Normally, the different lifetime ( τ e ) values of injectedelectrons in the devices based on these HTMs may be different,which can also inuence the obtainable V oc . Figure 5b,d showthe τ e plotted vs V OC from transient photovoltage decay meas-urements of these devices. At a certain V OC value, the τ e valuesfrom X3 -based device was obviously the highest among all ofthe HTMs (indicating the slowest recombination), for the seriesof measurements corresponding to the 30 mM-Li-TFSI concen-tration, while for the measurements related to the 20 mM-Li-TFSI concentration the Spiro-OMeTAD-based device showedthe slowest electron recombination; these results indicate thatthe inuence of different doping ratios of Li-TFSI on the elec-tron recombination is different. Actually, the electron recombi-nation rate of X3 , X35 and Spiro-OMeTAD based devices werevery close both at low and high concentration of Li-TFSI fromthe Figure 5 b,d, which might be due to their similar molecularweight. However, the τ e value for the X1 and X2 -based devicewere much lower than that of X3 , X35 and Spiro-OMeTAD-based devices under the same conditions, which is in goodagreement with our previous hypothesis that the HTM- X1 and X2 with small molecule size result to a facilitated approach tothe TiO 2 interface, leading to a fast recombination process inthe devices. Especially, X1 showed higher oxidation potentialand hole mobility than Spiro-OMeTAD, yet the V OC of the X1 -

based device was still much lower than incorporating Spiro-OMeTAD, which might be due to the molecular size less thanone half of Spiro-OMeTAD. This is a good example to show thata smaller HTM leads to faster electron recombination, givinglower V OC in the device.

Figure 6 a,b show the energy level diagram of the corre-sponding materials and the open-circuit voltage of the devices vsoligomerization number (n, one triphenylamine unit counting 1)for the HTMs. By oligomerization of triphenylamine, the V OC ofthe devices can be signicantly increased with aggrandizing themolecular weight of HTMs. However, we did not observe anycorrelation between the oxidation potential (both in solution and

Adv. Energy Mater . 2015, 5, 1401185


0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.00123456789

1011

C u r r e n

t D e n s

i t y

( m

A / c m

- 2 )

Voltage (V)

X1 X2 X3X35

Spiro-OMeTAD

a)

b)

350 400 450 500 550 600 650 700 7500

10

20

30

40

50

60

70

80

I P C E ( % )

Wavelength (nm)

X1 X2 X3 X35 Spiro-OMeTAD

Figure 4. a) J –V characteristics ofX1 , X2 , X3 ,[9] X35 and Spiro-OMeTADbased ssDSCs measured under 100 mW cm−2 AM1.5G solar intensity.b) Corresponding IPCE spectra of the devices.

8/9/2019 A Enm 201401185

8/12


on TiO2 electrode) of the HTMs and the device parameters. Atthis stage we cannot rule out the possibility that the higher oxida-tion potential is a critical factor in determining solid-state deviceperformance. Nevertheless, the transient photovoltage decay datashowed that the smaller molecular size of the HTM induced

faster electron recombination in the device, which indicates thatthe molecule size of the HTM is an important factor as regards itsimpact on the V OC in the device, or even more important than theHOMO level. This result shows a possibility to evaluate more ef-cient HTMs for ssDSCs and perovskite solar cells in the future.

3. Conclusion

We have systematically and deeply investigated a series ofnovel triphenylamine-based HTMs with different conjugationlength, oxidation potential and hole mobility in ssDSCs. The

thermal properties, electrochemical properties, photophysicalproperties and computation of all the HTMs were comprehen-sively studied. The DSC data illustrated that their crystallinetemperature will be lower by increasing the oligomerizationnumber of triphenylamine unit for HTMs, and then they



0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.001234

56789

10

C u r r e n

t D e n s

i t y

( m

A / c m

- 2 )

Voltage (V)

X1 X2 X3 X35 Spi ro-OMeTAD

0.4 0.5 0.6 0.7 0.8

0.01

0.1

τ e

[ s

]

VOC [V]


0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0012345

6789

10

C u r r e n

t D e n s i

t y ( m

A / c m

- 2 )

Voltage (V)


0.4 0.5 0.6 0.7 0.8

0.01

0.1

τ e

[ s

]

VOC [V]


a) b)

c) d)

Figure 5. a) J –V curves of ssDSCs with different HTMs doping with 20 mM Li-TFSI. b) Corresponding lifetime curves of ssDSCs with differentdoping with 20 mM Li-TFSI. c) J –V curves of ssDSCs with different HTMs doping with 30 mM Li-TFSI. d) Corresponding lifetime curves of ssDwith different HTMs doping with 30 mM Li-TFSI.

Table 4. Photovoltaic parameters determined from J –V measurements of ssDSCs doping with different concentration of Li-TFSI based on thesHTMs. Measured under simulated AM1.5G solar irradiance (100 mW cm−2 ).

HTMsa) Li-TSFI[mM]

V OC [V]

J SC [mA cm−2 ]

FF η [%]

X1b) [8] 60 0.725± 0.025 (0.75) 9.26± 0.34 (9.47) 0.56± 0.06 (0.62) 4.0± 0.2 (4.4)

X2c) 50 0.800± 0.015 (0.81) 9.51± 0.42 (9.79) 0.60± 0.04 (0.63) 4.7± 0.3 (5.0)

X3d) [9 ] 30 0.880± 0.035 (0.91) 9.23± 0.45 (9.52) 0.62± 0.05 (0.67) 5.4± 0.4 (5.8)

X35e) 30 0.875± 0.025 (0.89) 9.62± 0.35 (9.81) 0.61± 0.03 (0.63) 5.2± 0.3 (5.5)

Spiro-OMeTADf) 20 0.905± 0.025 (0.93) 9.13± 0.51 (9.37) 0.58± 0.02 (0.62) 5.2± 0.2 (5.4)

a) All of the HTM-based devices using 2.0 µm TiO2 lm in this study;b) 170 mM HTM doped with 200 mM t-BP and 60 mM Li-TFSI;c) 130 mM HTM doped with 200 mMt-BP and 50 mM Li-TFSI;d) 100 mM HTM doped with 200 mM t-BP and 30 mM Li-TFSI;e) 100 mM HTM doped with 200 mM t-BP and 30 mM Li-TFSI;f) 150 mM HTMdoped with 200 mM t-BP and 20 mM Li-TFSI. Note: the values are reported as the average obtained from 8 devices.

8/9/2019 A Enm 201401185

9/12


exhibit better solubility and more pronounced amorphouscharacter. The theoretical study showed that X3 and X35 hadsmaller reorganization energy than Spiro-OMeTAD due tothe larger conjugation length, which implied relative fast holetransport properties.

By optimizing aforementioned parameters of the aboveHTMs in conjunction with ssDSC operation, efciencies of5.8% and 5.5% were achieved by X3 - and X35 -based devices,which were higher than that the efciency of 5.4% obtainedusing the standard HTM Spiro-OMeTAD, thereby makingthese new HTMs promising for preparing high-efciencyssDSCs. More importantly, the transient photovoltage decaydata showed that the open-circuit voltage in ssDSCs was relatedto the molecular size of the HTM. Our results showed that thesmaller molecular size of the HTM induced faster electronrecombination in the device, which is probably due to the smallHTM molecules are easily to permeate into the meso-TiO 2 poreand the lacunas between the dye molecules, approaching tothe TiO 2 surface. These results provide new guidance for themolecular design of efcient HTMs in the future. Meanwhile,the experiment involving different HTMs using different Li-TFSI doping concentrations showed that Li-TFSI was a vitalpart of ssDSCs to achieve the optimal efciency, which providedan effective way to evaluate more HTMs for ssDSCs, and even-

tually materials for perovskite solar cells as well in the future.Further work with a focus on developing larger molecular sizeHTMs in order to suppress the electron recombination processso that to obtain higher open-circuit voltage in ssDSCs is inprogress.

4. Experimental SectionGeneral : Bis(triuoromethane)sulfonimide lithium salt (Li-TSFI,

99.95%) and 4-tert-butylpyridine (t-BP, 96%) were purchased fromAldrich. The t-BP was distilled before using. Chlorobenzene (anhydrou

Adv. Energy Mater . 2015, 5, 1401185


Table 5. Photovoltaic parameters determined from J–V measurements of ssDSCs doping with the same concentration of Li-TFSI based on thesHTMs. Measured under simulated AM1.5G solar irradiance (100 mW cm−2 ).

HTMsa) Li-TSFI[mM]

V OC [V]

J SC [mA cm−2 ]

FF η [%]

X1b) [8] 20 0.765± 0.45 (0.80) 7.56± 0.66 (8.17) 0.50± 0.05 (0.55) 3.1± 0.5 (3.6)

30 0.755± 0.35 (0.79) 8.11± 0.58 (8.64) 0.53± 0.04. (0.56) 3.2± 0.6 (3.8)

X2c) 20 0.845± 0.25 (0.87) 8.36± 0.29 (8.54) 0.56± 0.03 (0.58) 3.8± 0.5 (4.3)

30 0.825± 0.35 (0.86) 8.95± 0.43 (9.30) 0.57± 0.02 (0.58) 4.2± 0.4 (4.6)

X3d) [9 ] 20 0.905± 0.025 (0.93) 8.91± 0.25 (9.15) 0.56± 0.05 (0.61) 4.8± 0.4 (5.2)

30 0.880± 0.035 (0.91) 9.23± 0.45 (9.52) 0.62± 0.05 (0.67) 5.4± 0.4 (5.8)

X35e) 20 0.865± 0.045 (0.90) 8.89± 0.37 (9.24) 0.55± 0.07 (0.62) 4.9± 0.3 (5.2)

30 0.875± 0.025 (0.89) 9.62± 0.35 (9.81) 0.61± 0.03 (0.63) 5.2± 0.3 (5.5)

Spiro-OMeTADf) 20 0.905± 0.025 (0.93) 9.13± 0.51 (9.37) 0.58± 0.02 (0.62) 5.2± 0.2 (5.4)

30 0.875± 0.015 (0.89) 9.28± 0.41 (9.50) 0.56± 0.06 (0.61) 4.8± 0.4 (5.2)

a) All of the HTM based devices using 2.0 µm TiO2 lm in this study;b) 170 mM HTM doped with 200 mM t-BP;c) 140 mM HTM doped with 200 mM t-BP;d) 100 mM HTMdoped with 200 mM t-BP;e) 100 mM HTM doped with 200 mM t-BP;f) 150 mM HTM doped with 200 mM t-BP. Note: the values are reported as the average obtained fro8 devices.

Figure 6. a) Energy level diagram of the corresponding materials usedin our devices: grey line measured in DCM, black line measured on TiO2 electrode. b) Open-circuit voltage of the devices vs oligomerizationnumber (n ) of the HTMs with different Li-TFSI doping concentrations.

8/9/2019 A Enm 201401185

10/12

© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1401185 (10 of 12) wileyonlinelibrary.com Adv. Energy Mater.2015, 5, 1401185


99.8%) and acetonitrile (anhydrous 99.8%) was used as received. DyeLEG4 was provided by Dyenamo AB. Solvents and other chemicals arealso commercial available and used as received unless specially stated.Compound1 was synthesized according to literature procedure.[17] TheHTMX3 and intermediateTPA-Br were synthesized according to ourprevious published procedure.[9] Chromatography was performed usingsilica gel 60Å (35–63 µm). NMR spectra were recorded on a BrukerAVANCE 500 MHz spectrometer.

Synthesis of N4,N4,N4′ ,N4′ -tetrakis(4-methoxyphenyl)-[1,1′ -biphenyl]-4,4′ -diamine (X1 ) : A mixture of 4,4′-dibromo-1,1′-biphenyl (312 mg,1.00 mmol), Pd(OAc)2 (9 mg, 0.04 mmol), tri-tert-butyl phosphine(8 mg, 0.04 mmol), bis(4-methoxyphenyl)amine (458 mg, 2.00 mmol),and sodium tert-butoxide (240 mg, 2.50 mmol) was placed in a Schlenkask. The ask was subjected to three vacuum/nitrogen rell cyclesin order to remove water and oxygen. Anhydrous toluene (10 mL)was added, and the mixture was stirred overnight at 100°C or untilthe reaction was complete by TLC analysis. After cooling, the reactionwas quenched by water, and then followed by product extraction withethyl acetate. The organic layer was dried over anhydrous Mg2 SO4 andevaporated under vacuum. The collected residue was further puried bysilica gel column chromatography (hexane/EtOAc,v/v, 5:1) to giveX1 aspale yellow solid (578 mg, yield 95%).1 H NMR (d6 -DMSO, 500 MHz,298 K),δ (ppm): 7.40 (d, 4 H, J= 10.0 Hz), 7.01 (d, 8 H, J= 10.0 Hz),

6.90 (d, 8 H, J= 5.0 Hz), 6.80 (d, 4 H, J

= 5.0 Hz), 3.73 (s, 12 H);

13

CNMR (C6 D6 , 500 MHz, 298 K),δ (ppm): 156.42, 148.16, 141.71, 133.96,127.55, 126.81, 122.03, 115.16, 55.04.

Synthesis of N4-(4′ -(bis(4-methoxyphenyl)amino)-[1,1′ -biphenyl]-4-yl)-N4,N4′ ,N4′ -tris-(4-methoxy phenyl)-[1,1′ -biphenyl]-4,4′ -diamine (X2 ) :4′-bromo-N,N-bis(4-methoxyphenyl) -[1,1′-biphenyl]-4-amine (TPA-Br ) (921 mg, 2.00 mmol), 4-methoxyaniline (112 mg, 0.91 mmol),sodium tert-butoxide (217 mg, 2.28 mmol), palladium diacetate(8 mg, 0.04 mmol), and tri-tert-butyl phosphine (8 mg, 0.04 mmol)were transferred to a 50 mL Schlenk ask. The ask was subjected tothree vacuum/nitrogen rell cycles in order to remove water and oxygen.Anhydrous toluene (10 mL) was added, and the mixture was stirred for10 h at 100°C or until the reaction was complete by TLC analysis. Theproduct was collected by extraction with addition of EtOAc and water,followed by a brine solution. The collected organic layer was dried withanhydrous Mg2 SO4 , ltered, and concentrated by a rotary evaporator.The collected residue was further puried by column chromatography(hexane/EtOAc, v/v, 2:1) to giveX2 as pale yellow solid (746 mg, yield90%).1 H NMR (d6 -DMSO, 500 MHz, 298 K):δ (ppm): 7.48 (d, 4 H, J= 5.0 Hz), 7.44 (d, 4 H, J= 5.0 Hz), 7.08 (d, 2 H, J= 10.0 Hz), 7.03 (d, 8 H,J = 10.0 Hz), 6.98 (d, 4 H, J= 10.0 Hz), 6.95 (d, 2 H, J= 10.0 Hz), 6.91(d, 8 H, J= 10.0 Hz), 6.81 (d, 4 H, J= 5.0 Hz), 3.75 (s, 3H), 3.74 (s, 12H); 13 C NMR (C6 D6 , 500 MHz, 298 K),δ (ppm): 156.89, 156.48, 148.33,147.38, 141.64, 141.18, 135.11, 133.71, 128.35, 127.72, 127.65, 126.89,123.79, 121.89, 115.28, 115.18. 55.05. HR-MS (ESI) m/z: [M]+ calcd for881.3829; found, 881.3870.

Synthesis of N4,N4-bis(4′ -(bis(4-methoxyphenyl)amino)-[1,1′ -biphenyl]-4-yl)-N4′ ,N4′ -bis(4-methoxypheny l)-[1,1′ -biphenyl]-4,4′ -diamine (X35 ) :Under inert atmosphere, 5 mL K3 PO4 water solution (0.5 M) and 10 mLdry THF were added to the mixture of tris(4-bromophenyl)amine (0.24 g,0.5 mmol), 4-methoxy-N-(4-methoxyphenyl)-N-(4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl) phenyl)aniline (1 ) (0.863 g, 2.0 mmol) andXphos-Pd (40.0 mg, 0.05 mmol). The mixture was stirred at 40°C for3 h. Then the solvent was removed by vacuum distillation. EtOAc wasadded to solve the residue, and the solution was washed with saturatedNaCl aqueous solution and water sequentially. After being dried byanhydrous Na2 SO4 , the solution was ltered and the solvent wascompletely removed by arotary evaporator. The raw product was puriedby column chromatography on silica gel using mixed solvent of hexane/EtOAc (v/v, 2:1) as eluent. Final product was obtained as pale yellowsolid (0.31 g, 55%).1 H NMR (d6 -DMSO, 500 MHz, 298 K):δ (ppm):7.53 (d, 6 H, J= 10.0 Hz), 7.46 (d, 6 H, J= 10.0 Hz), 7.08 (d, 6 H, J= 5.0 Hz), 7.04 (d, 12 H, J= 10.0 Hz), 6.92 (d, 12 H, J= 10.0 Hz), 6.81 (d,6 H, J= 10.0 Hz), 3.74 (s, 18 H);13 C NMR (C6 D6 , 500 MHz, 298 K),δ (ppm): 157.05, 148.99, 147.46, 142.13, 136.47, 134.08, 128.50, 128.28,

127.49, 125.56, 122.34, 115.73, 55.59. HR-MS (ESI) m/z: [M]+ calcd for1154.4982; found, 1154.4915.

Differential Scanning Calorimetry (DSC) : DSC analysis of thematerials was performed on Metter-Toledo DSC 820 to determine theglass transition temperature and degree of crystalline. Around 5 mg ofeach sample was enclosed into standard 70 µL aluminum cups. Thetemperature was raised 10°C min−1 from 0 °C to 300°C and thendecreased to 0°C with the same rate. The whole temperature programwas performed under a nitrogen gas ow of 80 ml·min−1 .

Optical Characterization : UV–Vis absorption spectra were recordedon a Lambda 750 UV–Vis spectrophotometer. The uorescencespectra of dye solutions were recorded on a Cary Eclipse uorescencespectrophotometer. All samples were measured in a 1 cm cell at roomtemperature at a concentration of 10−5 M in dichloromethane.

Electrochemical Measurements : Electrochemical experiments wereperformed with a CH Instruments electrochemical workstation (model660A) using a conventional three-electrode electrochemical cell. Adichloromethane solution containing 0.1 M of tetrabutylammoniunhexauorophosphate (n-Bu4 NPF6 ) was introduced as electrolyte in athree-electrode cell, where an Ag/0.01 M AgNO3 electrode (acetonitrileas solvent) was used as the reference electrode and a glassy carbondisk (diameter 3 mm) as the working electrode, a platinum wire as thecounter electrode. The cyclovoltammetric scan rates were 50 mV/s. All

redox potentials were calibrated vs normal hydrogen electrode (NHE) bythe addition of ferrocene. The conversion E(Fc/Fc+) = 630 mV vs NHE.In electrochemical experiments on TiO2 electrode, the measurementswere carried out at room temperature and ionic liquid 1-Butyl-1-methylpyrrolidinium bis(triuoromethylsulfonyl) imide as electrolyte.In a three-electrode cell, where the Ag/AgCl (3 M NaCl) was used asthe reference electrode, FTO/TiO2 /hole conductor by spin coatingwith concentration of 5 mg/100 µL (chlorobenzene as solvent) as theworking electrode (area: 1 cm2 ) and a stainless steel plate as the counterelectrode. Scan rates was 50 mV s−1 . Each measurement was calibratedwith ferrocene.

Computational Details : The geometry optimization and electronicstructure of the molecules were calculated at B3LYP/6–31g* level byusing Gaussian 09 program.[33]

Mobility Measurements : Charge transport in the HTMs has beeninvestigated according to literature.[18,19] Tin-doped indium oxide (ITO)coated glass substrates were cleaned by using acetone and detergent.The substrates were then treated by TL-1, which is a mixture of water,ammonia (25%), and hydrogen peroxide (28%) (5:1:1 by volume).A 40 nm thick PEDOT: PSS layer was spin-coated onto the substrates,which were then annealed at 120°C for 30 min in air. The substrateswere then transferred into a glovebox for further fabrication steps. TheHTMs were dissolved in anhydrous chlorobenzene at 70°C with asolution concentration of 10 mg mL−1 . This solution was spin-coatedat 2000 rpm to yield lms. The thicknesses of the lms are measuredby using a Dektak 6M prolometer. 10 nm of molybdenum trioxidewas then evaporated onto the active layer under high vacuum (lessthan 10−6 mbar). Finally, aluminum contact, 90 nm, has been appliedvia evaporation through a shadow mask. J –V characteristics of thedevices have been measured with a Keithley 2400 Source-Measure unit,interfaced with a computer. Device characterization was carried out in air.

Conductivity Measurements : Glass substrates without conductive layerwere carefully cleaned in ultrasonic baths of detergents, deionized water,acetone and ethanol successively. Remaining organic residues wereremoved with 10 min by airbrushing. A thin layer of nanoporous TiO2 wascoated on the glass substrates by spin-coating with a diluted TiO2 paste(Dyesol DSL 18NR-T) with terpineol (1:3, mass ratio). The thickness othe lm is ca. 500 nm, as measured with a DekTak prolometer. Aftersintering the TiO2 lm on a hotplate at 500°C for 30 min, the lm wascooled to room temperature, before it was subsequently depositedby spin-coating of a solution of HTM in chlorobenzene, whereas theconcentrations were the same as in case of photovoltaic devices. J –V characteristics were recorded on a Keithley 2400 SemiconductorCharacterization System. Device fabrication was carried out in a glovebox under nitrogen atmosphere.

8/9/2019 A Enm 201401185

11/12

© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim (11 of 12)1401185wileyonlinelibrary.com Adv. Energy Mater . 2015, 5, 1401185


Cross-Sectional Scanning Electron Microscopy : Field emission scanningelectron microscopy (FESEM) images were obtained with the JEOLSU-7000F microscope operating at an accelerating voltage of 15 kV.

Device Fabrication : Fluorine-doped tin-oxide (FTO) coated glasssubstrates (Pilkington TEC15) were patterned by etching with zincpowder and 2 M hydrochloric acid. The substrates were carefullycleaned in ultrasonic baths of detergents, deionized water, acetone andethanol successively. The remaining organic residues were removedwith 10 min by airbrush. A compact TiO2 blocking layer was depositedonto the surface of a pre-cleaned FTO substrate by spray pyrolysis ona hotplate at 450°C using an airbrush. The solution used in the spraypyrolysis was 0.2 M Ti-isopropoxide, 2 M acetylacetone in isopropanol.In all electrode preparations 10 spray cycles were used as standardparameter. Nanoporous TiO2 lms were coated on the compact TiO2 layer by screen-printing of a diluted TiO2 paste (Dyesol DSL 18NR-T)with terpineol (2:1, mass ratio). The thickness of the lm is ca. 2.0 µm,as measured with a DekTak prolometer. After sintering the TiO2 lmon a hotplate at 500°C for 30 min, the lm was cooled to roomtemperature and immersed in 0.02 M aqueous TiCl4 at 70°C for 30 min.The lm was then rinsed by deionized water and then annealed ona hotplate at 500°C for 30 min. After cooling to 90°C, the lm wasimmersed for 2 h in 0.1 mM solution of LEG4 dissolved in tert-butanoland acetonitrile (1:1), and then the sensitized electrodes were rinsed by

ethanol and dried. Subsequently, the chlorobenzene solution containingthe corresponding HTM and additives was applied to form the HTM lmby leaving the solution to penetrate into the sensitized electrode for 30 sand then followed by spin-coating for 30 s with 2000 rpm. Afterwards,the cells were left in air overnight in the dark (humidity below 10%),then a 200 nm thick Ag back contact was deposited onto the organicsemiconductor by thermal evaporation in a vacuum chamber (Leica EMMED020) with a base pressure of about 10−6 bar, to complete the devicefabrication. The device fabrication was carried out in a glove box undernitrogen atmosphere.

Device Characterization : Current–voltage characteristics were recordedby applying an external potential bias to the cell while recording thegenerated photocurrent with a Keithley model 2400 digital sourcemeter. The light source was a 300 W collimated xenon lamp (Newport)calibrated with the light intensity to 100 mW·cm−2 at AM 1.5 G solar lightcondition by a certied silicon solar cell (Fraunhofer ISE). IPCE spectrawere recorded on a computer-controlled setup comprised of a xenonlamp (Spectral Products ASB-XE-175), a monochromator (SpectralProducts CM110) and a Keithley multimeter (Model 2700). The setupwas calibrated with a certied silicon solar cell (Fraunhofer ISE) priorto measurements. Electron lifetime was measured by the custom-made‘‘toolbox setup′′ using a green-light-emitting diode (Luxeon K2 star 5 W,λ max = 530 nm) as light source. All ssDSC samples were illuminated fromthe glass side with an aperture area of 0.20 cm2 (0.4 × 0.5 cm2 ). Theprepared ssDSC samples were masked during the measurement with anaperture area of 0.20 cm2 (0.4× 0.5 cm2 ) exposed under illumination.

Supporting InformationSupporting Information is available from the Wiley Online Library or

from the author.

AcknowledgementsThis work was nancially supported by the Swedish Research Council,the Swedish Energy Agency, the Knut and Alice Wallenberg Foundation,the National Natural Science Foundation of China (21120102036,91233201), the National Basic Research Program of China (973 program,2014CB239402), the National Natural Science Foundation of China(No.21403133) and the China Scholarship Council (CSC). The authorswould like to thank Jianghua Zhao (Dalian University of Technology) forrecording the HR-MS spectra and Weifeng Zhao (KTH Royal Institute of

Technology) for DSC measurements. The authors also greatly thank ErikGabrielsson (KTH) and Jing Huang (KTH) for their helpful discussions anYan Hao (Uppsala University) for her kind help in laboratory experiments

Received: July 16, 2014Revised: August 27, 2014

Published online: September 22, 2014

[1] B. O′Regan, M. Gratzel, Nature 1991 , 353 , 737.[2] A. Yella, H.-W. Lee, H. N. Tsao, C. Yi, A. K. Chandiran,

M. K. Nazeeruddin, E. W.-G. Diau, C.-Y. Yeh, S. M. Zakeeruddin,M. Grätzel, Science 2011 , 334 , 629.

[3] a) A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo, H. Pettersson, Chem.Rev. 2010 , 110 , 6595; b) H. Tian, L. Sun, J. Mater. Chem. 2011 , 21 ,10592; c) M. Wang, C. Gratzel, S. M. Zakeeruddin, M. Gratzel,Energy Environ. Sci. 2012 , 5 , 9394; d) C.-Y. Hsu, Y.-C. Chen, R. Y.-Y.Lin, K.-C. Ho, J. T. Lin, Phys. Chem. Chem. Phys. 2012 , 14 , 14099.

[4] U. Bach, D. Lupo, P. Comte, J. E. Moser, F. Weissortel, J. Salbeck,H. Spreitzer, M. Gratzel, Nature 1998 , 395 , 583.

[5] a) J. Burschka, A. Dualeh, F. Kessler, E. Baranoff, N.-L. Cevey-Ha,C. Yi, M. K. Nazeeruddin, M. Grätzel, J. Am. Chem. Soc. 2011 , 133 ,18042; b) N. Cai, S.-J. Moon, L. Cevey-Ha, T. Moehl, R. Humphry-Baker, P. Wang, S. M. Zakeeruddin, M. Grätzel, Nano Lett. 2011 , 11 ,1452.

[6] a) M. M. Lee, J. Teuscher, T. Miyasaka, T. N. Murakami, H. J. Snaith,Science 2012 , 338 , 643; b) M. Liu, M. B. Johnston, H. J. Snaith,Nature 2013 , 501 , 395.

[7] a) S. A. Haque, T. Park, A. B. Holmes, J. R. Durrant, ChemPhy-sChem 2003 , 4 , 89; b) E. M. J. Johansson, L. Yang, E. Gabrielsson,P. W. Lohse, G. Boschloo, L. Sun, A. Hagfeldt, J. Phys. Chem. C 2012 , 116 , 18070.

[8] L. Yang, B. Xu, D. Bi, H. Tian, G. Boschloo, L. Sun, A. Hagfeldt,E. M. J. Johansson, J. Am. Chem. Soc. 2013 , 135 , 7378.

[9] B. Xu, H. Tian, D. Bi, E. Gabrielsson, E. M. J. Johansson,G. Boschloo, A. Hagfeldt, L. Sun, J. Mater. Chem. A 2013 , 1 , 14467.

[10] C. T. Weisspfennig, M. M. Lee, J. Teuscher, P. Docampo,S. D. Stranks, H. J. Joyce, H. Bergmann, I. Bruder, D. V. Kondratuk,M. B. Johnston, H. J. Snaith, L. M. Herz, J. Phys. Chem. C 2013 , 117 ,19850.

[11] a) I. Chung, B. Lee, J. He, R. P. H. Chang, M. G. Kanatzidis, Nature 2012 , 485 , 486; b) G. Puckyte, B. Schmaltz, A. Tomkeviciene,M. Degbia, J. V. Grazulevicius, H. Melhem, J. Bouclé, F. Tran-Van, J. Power Sources 2013 , 233 , 86; c) C. S. Karthikeyan, M. Thelakkat,Inorg. Chim. Acta 2008 , 361 , 635; d) H. Wang, J. Li, F. Gong,G. Zhou, Z.-S. Wang, J. Am. Chem. Soc. 2013 , 135 , 12627; e) D. Bi,L. Yang, G. Boschloo, A. Hagfeldt, E. M. J. Johansson, J. Phys.Chem. Lett. 2013 , 4 , 1532; f) T. Park, S. A. Haque, R. J. Potter,A. B. Holmes, J. R. Durrant, Chem. Commun. 2003 , 2878;g) I. Y. Song, S.-H. Park, J. Lim, Y. S. Kwon, T. Park, Chem. Commun. 2011 , 47 , 10395.

[12] J. E. Kroeze, N. Hirata, L. Schmidt-Mende, C. Orizu, S. D. Ogier,

K. Carr, M. Grätzel, J. R. Durrant, Adv. Funct. Mater. 2006 , 16 ,1832.[13] T. Leijtens, I. K. Ding, T. Giovenzana, J. T. Bloking, M. D. McGehee,

A. Sellinger, ACS Nano 2012 , 6 , 1455.[14] J. H. Heo, S. H. Im, J. H. Noh, T. N. Mandal, C.-S. Lim, J. A. Chang,

Y. H. Lee, H.-J. Kim, A. Sarkar, K. NazeeruddinMd, M. Gratzel,S. I. Seok, Nat. Photonics 2013 , 7 , 486.

[15] M. Planells, A. Abate, D. J. Hollman, S. D. Stranks, V. Bharti,J. Gaur, D. Mohanty, S. Chand, H. J. Snaith, N. Robertson, J. Mater.Chem. A 2013 , 1 , 6949.

[16] a) X. Liu, W. Zhang, S. Uchida, L. Cai, B. Liu, S. Ramakrishna, Adv.Mater. 2010 , 22 , E150; b) L. Yang, J. Zhang, Y. Shen, B.-W. Park,D. Bi, L. Häggman, E. M. J. Johansson, G. Boschloo, A. Hagfeldt,

8/9/2019 A Enm 201401185

12/12

© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1401185 (12 of 12) wileyonlinelibrary.com Adv. Energy Mater.2015, 5, 1401185


N. Vlachopoulos, A. Snedden, L. Kloo, A. Jarboui, A. Chams,C. Perruchot, M. Jouini, J. Phys. Chem. Lett. 2013 , 4 , 4026.

[17] L. Chen, B. Zhang, Y. Cheng, Z. Xie, L. Wang, X. Jing, F. Wang, Adv.Funct. Mater. 2010 , 20 , 3143.

[18] F. Gao, J. Wang, J. C. Blakesley, I. Hwang, Z. Li, N. C. Greenham, Adv. Energy Mater. 2012 , 2 , 956.

[19] C. Goh, R. J. Kline, M. D. McGehee, E. N. Kadnikova,

J. M. J. Fréchet, Appl. Phys. Lett. 2005 , 86.[20] I. K. Ding, N. Tétreault, J. Brillet, B. E. Hardin, E. H. Smith,S. J. Rosenthal, F. Sauvage, M. Grätzel, M. D. McGehee, Adv. Funct.Mater. 2009 , 19 , 2431.

[21] E. T. Seo, R. F. Nelson, J. M. Fritsch, L. S. Marcoux, D. W. Leedy,R. N. Adams, J. Am. Chem. Soc. 1966 , 88 , 3498.

[22] E. Gabrielsson, H. Ellis, S. Feldt, H. Tian, G. Boschloo, A. Hagfeldt,L. Sun, Adv. Energy Mater. 2013 , 3 , 1647.

[23] B. E. Hardin, H. J. Snaith, M. D. McGehee, Nat. Photonics 2012 , 6 , 162.[24] U. B. Cappel, E. A. Gibson, A. Hagfeldt, G. Boschloo, J. Phys. Chem.

C 2009 , 113 , 6275.[25] H. J. Snaith, M. Gratzel, Appl. Phys. Lett. 2006 , 89 , 262114.[26] A. Abate, D. J. Hollman, J. Teuscher, S. Pathak, R. Avolio,

G. D′Errico, G. Vitiello, S. Fantacci, H. J. Snaith, J. Am. Chem. Soc. 2013 , 135 , 13538.

[27] a) P. Docampo, A. Hey, S. Guldin, R. Gunning, U. Steiner,H. J. Snaith, Adv. Funct. Mater. 2012 , 2012; b) L. Schmidt-Mende,M. Grätzel, Thin Solid Films 2006 , 500 , 296; c) J. Melas-Kyriazi,I. K. Ding, A. Marchioro, A. Punzi, B. E. Hardin, G. F. Burkhard,N. Tétreault, M. Grätzel, J.-E. Moser, M. D. McGehee, Adv. EnergyMater. 2011 , 1 , 407.

[28] A. Abate, T. Leijtens, S. Pathak, J. Teuscher, R. Avolio, M. E. Errico,J. Kirkpatrik, J. M. Ball, P. Docampo, I. McPherson, H. J. Snaith,Phys. Chem. Chem. Phys. 2013 , 15 , 2572.

[29] a) D.-Y. Chen, W.-H. Tseng, S.-P. Liang, C.-I. Wu, C.-W. Hsu, Y. Chi,W.-Y. Hung, P.-T. Chou, Phys. Chem. Chem. Phys. 2012 , 14 , 11689;

b) A. Dualeh, F. De Angelis, S. Fantacci, T. Moehl, C. Yi, F. Kessler,E. Baranoff, M. K. Nazeeruddin, M. Grätzel, J. Phys. Chem. C 2011 , 116 , 1572; c) X. Jiang, K. M. Karlsson, E. Gabrielsson,E. M. J. Johansson, M. Quintana, M. Karlsson, L. Sun, G. Boschloo,A. Hagfeldt, Adv. Funct. Mater. 2011 , 21 , 2944.

[30] a) M. K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker,E. Mueller, P. Liska, N. Vlachopoulos, M. Graetzel, J. Am. Chem.

Soc. 1993 , 115 , 6382; b) S. A. Haque, E. Palomares, B. M. Cho,A. N. M. Green, N. Hirata, D. R. Klug, J. R. Durrant, J. Am. Chem.Soc. 2005 , 127 , 3456; c) S. A. Haque, T. Park, C. Xu, S. Koops,N. Schulte, R. J. Potter, A. B. Holmes, J. R. Durrant, Adv. Funct.Mater. 2004 , 14 , 435.

[31] B. Enright, D. Fitzmaurice, J. Phys. Chem. 1996 , 100 ,1027.

[32] G. Schlichthörl, S. Y. Huang, J. Sprague, A. J. Frank, J. Phys. Chem.B 1997 , 101 , 8141.

[33] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai,

T. Vreven, J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M. Bearpark,J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi,J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox,J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts,R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli,J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski,G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich,A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski,D. J. Fox, Gaussian 09, Revision a. 02, Gaussian, Inc., Wallingford,CT 2009.

A Enm 201401185

Documents