Solution-Processed Bulk Heterojunction Solar Cells with Silyl End-Capped Sexithiophene

1 Photo-Electronic Hybrids Research Center, Korea Institute of Science and Technology, Hwarangno 14-gil 5, Seongbuk-gu, Seoul 136-791, Republic of Korea 2Department of Chemistry, Faculty of Science, Kafrelsheikh University, Kafr ElSheikh 33516, Egypt 3 Department of Chemistry, Pusan National University, Jangjeon-dong Kumjeong-gu, Busan 609-735, Republic of Korea 4Department of Material and Life Science, Graduate School of Engineering, Osaka University, ALCA, Japan Science and Technology Agency (JST), Suita, Osaka 565-0871, Japan 5Department of Chemistry and Nanoscience, Ewha Womans University, 52 Ewhayeodae-gil, Seodaemun-gu, Seoul 120-750, Republic of Korea


Introduction
Organic solar cells have received strong attention due to the possibility to realize flexible and low-cost large photovoltaic cells [1][2][3][4][5].Polymer based organic photovoltaic cells (OPVs) demonstrated power conversion efficiencies (PCEs) in excess of 7% by the results of synthesis of low bandgap polymers [6,7].However, the polymers still have high batch-to-batch variations and difficulty in purification, and there have been many efforts to discover solution-processed small molecules.Recently, oligothiophenes have received much attention as donor materials in OPVs because of their high hole mobility, easy multigram synthesis, high purity, and simple chemical modification [8][9][10].Some studies on oligothiophene based solar cells have already been reported using bilayer heterojunction solar cells [11][12][13][14][15].By controlling the film morphology using coevaporation of excess fullerene (C 60 ), bulk heterojunction organic photovoltaic cells consisting of sexithiophene (6T) as a donor and C 60 as an acceptor showed good photovoltaic properties [16].However, the small molecules based on sexithiophene have not been investigated solution-processed OPVs.The low performance of sexithiophene-based solar cells was mainly attributed to the difficulties encountered in fabricating solution-processed bulk heterojunction cells.
In regard to our continuing interest for sexithiophenebased organic solar cells, we report herein analysis of the performance of solution-processed organic solar cells based on ,-functionalized linear sexithiophene: PCBM systems.We fabricated the solution-processed OPVs using two linear sexithiophenes, ,-dihexylsexithiophene (DH-6T) and ,bis(dimethyl-n-octylsilyl)sexithiophene (DSi-6T) as donors and [6,6]-phenyl-C 61 -butyric acid methyl ester (PCBM) as an acceptor (Figure 1).Solution-processed bulk heterojunction (BHJ) solar cells using DH-6T and DSi-6T with PCBM showed good photovoltaic properties in spite of their poor solubility.Although DH-6T has higher hole mobility than DSi-6T, the DSi-6T : PCBM blend cell showed higher hole mobility than DH-6T : PCBM cell.Therefore, DSi-6T cell showed higher device performance than DH-6T cell due to their silyl substitutions, which lead to the increase of the solubility.The higher photovoltaic performance of DSi-6T cell can be explained by the film morphology dependence on the solubility of donor.In addition, utilizing the nanosecond transient absorption spectroscopy in polar media in the visible and near-IR region proved electron-transfer reactions in DH-6T : PCBM and DSi-6T : PCBM mixtures.

Experimental Section
2.1.Instruments.The absorption spectra were measured using a Perkin Elmer Lambda 35 UV-vis spectrometer in spin-coated films at room temperature.The morphology of the DSi-6T : PCBM and DH6T : PCBM was determined using atomic force microscopy (AFM) (Park systems) operating in a noncontact mode, under ambient conditions.Steady-state fluorescence measurements were carried out on a Shimadzu spectrofluorophotometer (RF-5300PC).Phosphorescence spectra were obtained by a SPEX Fluorolog  3 spectrophotometer.Emission spectra in the visible region were detected by using a Hamamatsu Photonics R5509-72 photomultiplier.A deaerated 2-MeTHF solution containing DH-6T and DSi-6T at 77 K was excited at indicated wavelengths.Cyclic voltammograms (CV) and differential pulse voltammograms (DPV) techniques were carried on a BAS CV 50 W Voltammetric Analyzer.A platinum disk electrode was used as working electrode, while a platinum wire served as a counter electrode.SCE electrode was used as a reference electrode.All measurements were carried out in deaerated benzonitrile containing tetra-n-butylammonium hexafluorophosphate (TBAPF 6 ; 0.10 M) as a supporting electrolyte.The scan rate = 50 mV/s.
DSi-6T and DH-6T were excited by a Panther OPO pumped by Nd:YAG laser (Continuum, SLII-10, 4-6 ns fwhm) at  = 440 nm with the powers of 1.5 and 3.0 mJ per pulse.The transient absorption measurements were performed using a continuous xenon lamp (150 W) and an InGaAs-PIN photodiode (Hamamatsu 2949) as a probe light and a detector, respectively.The output from the photodiodes and a photomultiplier tube was recorded with a digitizing oscilloscope (Tektronix, TDS3032, 300 MHz).All measurements were conducted at 298 K.The transient spectra were recorded using fresh solutions in each laser excitation.
2.2.Fabrication of BHJ Cells.BHJ solar cells were fabricated on patterned ITO (indium tin oxide) coated glass substrates as the anode.ITO on glass substrates was sequentially cleaned with isopropyl alcohol, acetone, and isopropyl alcohol in ultrasonic baths.Afterwards, it was dried in oven at 80 ∘ C for 10 min and then treated with UV-ozone for 20 min.The surface of the ITO substrate was modified by spin-coating conducting poly(3,4-ethylenedioxythiophene) : poly(styrenesulfonate) (PEDOT : PSS, H. C. Starck Baytron P VP Al 4083) with a thickness of around 40 nm, followed by baking at 120 ∘ C for 10 min in oven.The molecular structures of the donor (DSi-6T and DH-6T) and acceptor materials (PCBM) and their energy band diagram are shown in Figure 1.DSi-6T was synthesized according to literature procedures [18].DH-6T was purchased from Aldrich Chemical Co. and was used without further purification.DSi-6T : PCBM (1 : 4, 1 : 5, 1 : 6 w/w) and DH-6T : PCBM (1 : 4, 1 : 6 w/w) were blended together and dissolved in chloroform at a total concentration of 20 mg mL −1 .And then DSi-6T : PCBM blended solution was no filtration, but DH-6T : PCBM solution was filtered out through a 0.45 m filter.The active layer was spin-cast at 4000 rpm from the solution of DSi-6T or DH-6T and PCBM (Nano-C) in chloroform.The thickness of the active layers was around 80 nm, as measured with an Alpha-step IQ.A solution containing 0.5 wt% TiO 2 nanoparticles in ethanol was spin-coat at a rate of 4000 rpm on top of the active layer.The synthesis of TiO 2 nanoparticles is described elsewhere [20,21].The cathode consists of 100 nm of aluminum and was thermally evaporated on top of the film through the use of a shadow mask to define an active area of 0.12 cm 2 under a base pressure of 1 × 10 −6 Torr (1 Torr = 133.32Pa).The devices were annealed at 60 ∘ C for 10 min in the thermal evaporator.The current density-voltage curves were obtained using a Keithley 2400 source-measure unit.The photocurrent was measured under illumination using a Newport class-A 100 mW cm −2 solar simulator (AM 1.5 G) and the light intensity was calibrated with an NREL-calibrated Si solar cell with KG-1 filter for approximating one sun light intensity.External quantum efficiency (EQE) was measured in range from 300 to 800 nm using a specially designed EQE system (PV Measurements, Inc.).A 75 W xenon lamp was used as a light source for the generating monochromatic beam.A calibration was performed using a silicon photodiode, which was calibrated using the NIST-calibrated photodiode G425 as a standard, and IPCE values were collected under bias light at a low chopping speed of 10 Hz.  absorption spectra in a dilute chloroform solution, with absorption maxima observed at 440 nm and 442 nm, respectively.The absorption spectra of sexithiophene films exhibited a noteworthy blue shift with respect to that in solution for both compounds, but this shift is more significant for DH-6T than for DSi-6T.While the DH-6T film showed the strongly blue-shifted absorption spectrum with maxima at 362 nm, the DSi-6T film showed an absorption peak at 415 nm.Such feature is attributed to molecular excitons formed as a consequence of the intermolecular interactions in the solid state [22] and is found to be blue-shifted by about 0.6 eV with respect to the characteristic absorption peaks of the isolated molecules.The PCBM film has an absorption band of about 340 nm and significantly greater absorption in the visible region (350-750 nm) compared to the PCBM solution, consistent with the literature [23].The UV-vis absorption spectra of DSi-6T : PCBM (1 : 5 w/w) and DH-6T : PCBM (1 : 6 w/w) blend films showed similar shape due to their same backbone of donor and strong absorption of PCBM.

Results and Discussion
Fluorescence spectra of the singlet-excited state of DH-6T and DSi-6T in benzonitrile exhibited emission bands at 514 and 520 nm, respectively, from which the energy of the singlet state of DH-6T and DSi-6T were estimated as 2.41 and 2.39 eV, respectively (See Supporting Information, Figure S1 of Supplementary Material available online at http://dx.doi.org/10.1155/2013/843615).Fluorescence quantum yields of DH-6T and DSi-6T was determined as 0.13 and 0.36, respectively.Fluorescence lifetimes of the singlet states of DH-6T and DSi-6T were determined as 1.0 and International Journal of Photoenergy 1.8 ns, respectively.In the case of the DH-6T/PCBM and DSi-6T/PCBM mixtures, it was found that the fluorescence intensities and the lifetimes are quite close to those of the DH-6T and DSi-6T, respectively.This observation suggests the intermolecular electron transfer between the DH-6T and DSi-6T with PCBM, but not the intramolecular electron transfer.
The redox potentials of the examined DH-6T, DSi-6T and PCBM have been examined by using cyclic voltammetry (CV) and differential pulse voltammetry (DPV) techniques to evaluate the driving forces for the electron transfer (−Δ et  ).The first reduction potential ( red ) of the PCBM was located at -856 mV versus Ag/AgNO 3 , while the oxidation potentials ( ox ) of DH-6T and DSi-6T were located at 732 and 600 mV versus Ag/AgNO 3 (see Supporting Information, Figure S2-S5).The redox potential measurements of the DH-6T/PCBM and DSi-6T/PCBM mixtures did not show significant interaction suggesting no interaction in the ground state.The redox potentials of the examined DH-6T, DSi-6T, and PCBM suggest their potentials as promising materials in the photovoltaic cells.The feasibility of the electron-transfer process via the triplet-excited states is controlled by the free energy change (Δ et  ), which can be expressed by the Rehm-Weller relation [24] (1): where  ox is the first oxidation potential of the DH-6T and DSi-6T,  red is the first reduction potential of PCBM,  T is the triplet energy of DH-6T (1.78 eV) and DSi-6T (1.79 eV) (we found from phosphorescence measurements in deaerated benzonitrile that energy level of the triplet DH-6T and DSi-6T at 1.78 and 1.79 eV, resp.), and  c is the Coulomb energy term (approximately 0.06 eV in the polar benzonitrile) [25].The free energy change (Δ et  ) values via the triplet DH-6T and DSi-6T were estimated as −0.05 and −0.17 eV, respectively.The negative Δ et  values suggest that the quenching process should be close to the diffusioncontrolled limit ( diff ) [26].
By photoexcitation of DSi-6T in deaerated benzonitrile using 440 nm laser photolysis, the transient absorption spectrum immediately after the laser pulse exhibited only an absorption band at 700 nm, which assigned to the triplet-excited state of DH-6T ( 3 DH-6T * ) (See Supporting Information; Figure S6).By fitting the decay profile of 3 DSi-6T * , the decay rate constant was found to be 6.60 × 10 4 s −1 , from which the lifetime of 3 DSi-6T * was estimated as 15.2 s.By photoexcitation of DSi-6T in the presence of PCBM [0.01-0.10mM] in Ar-saturated benzonitrile using 440 nm laser photolysis, the transient spectra exhibit the characteristic band of 3 DSi-6T * at 700 nm.With its decay, the concomitant rises of the DSi-6T radical cation (DSi-6T •+ ) at 800 and 1500 nm and the PCBM radical anion (PCBM •− ) at 1000 nm were observed (Figure 3(a) and Figure S7).These observations show clear evidence of occurrence of intermolecular electron transfer from the triplet-excited state of DSi-6T to PCBM.In oxygen-saturated solutions, an intermolecular energy transfer from 3 DSi-6T * to oxygen emerges, suppressing the electron-transfer process.Similar  electron-transfer features were recorded in the case of DH-6T/PCBM mixture in deaerated benzonitrile (See supporting information, Figure S8).
A more detailed picture of the kinetic is shown in Figure 3(b), where the rate constant of the electron-transfer process ( et ) was evaluated by monitoring the formation of the DSi-6T •+ as function of the concentrations of PCBM.The formation of DSi-6T •+ was fitted with clean first-order kinetics; each rate constant is referred to ( 1st ).The linear concentration dependence of the observed  1st values gives the  et , which is calculated as 2.24 × 10 9 M −1 s −1 , which is near the diffusion-controlled limit ( diff = 5.6 × 10 9 M −1 s −1 ) in benzonitrile [27].Similarly, the  et value of 3 DH-6T * /PCBM mixture was found to be 2.65 × 10 9 M −1 s −1 , which is slightly larger than that of 3 DSi-6T * /PCBM mixture (See supporting information, Figure S9).

Photovoltaic Properties of DSi-6T/PCBM and DH-6T/ PCBM.
For investigating the photovoltaic properties of sexithiophenes, OPVs were fabricated with the configuration of ITO/PEDOT : PSS/Donor (DSi-6T or DH-6T) : PCBM/Al.DSi-6T : PCBM and DH-6T : PCBM were blended together and dissolved in chloroform at a total concentration of 20 mg/mL.After Al deposition, the devices were thermally annealed at 60 ∘ C for 10 min.Figures 4 and 5 show the current density-voltage (J-V) curves of the OPV devices with DSi-6T : PCBM and DH-6T : PCBM using different blended ratios under AM 1.5 illumination, 100 mW cm −2 .Table 1 summarizes the photovoltaic performance of the functionalized  sexithiophene based solar cells.The sexithiophene DSi-6T was mixed with PCBM to form bulk heterojunction active layers, for use as a donor with PCBM as an acceptor.The weight ratios of the donor to acceptor were 1 : 4, 1 : 5, and 1 : 6.For comparison, the OPV devices based on DH-6T : PCBM were also fabricated with commercial DH-6T as a donor and the weight ratio of 1 : 4 and 1 : 6.The short circuit current density ( sc ) values of the DSi-6T : PCBM BHJ solar cells increase as the weight ratio of donor to acceptor increases.However, the weight ratio of donor to accepter could not be larger than 1 : 4, due to low solubility of sexithiophenes.The photovoltaic performance was observed on DSi-6T : PCBM 1 : 5 (w/w) blend cell with an open circuit voltage ( oc ) of 0.60 V, a  sc of 1.29 mA/cm 2 , a fill factor (FF) of 42%, and power conversion efficiency of 0.32%.Under the same conditions, DH-6T : PCBM 1 : 6 (w/w) blend cell showed a  oc of 0.63 V, a  sc of 0.60 mA/cm 2 , a FF of 44%, and PCE of 0.17%.The  sc values of the DH-6T : PCBM cells were constant with 0.60 mA/cm 2 regardless of the weight ratio of DH-6T to acceptor.This result could be ascribed to the saturated solution of DH-6T : PCBM (1 : 6 w/w).In fact, a factor strongly affecting the  sc of DSi-6T : PCBM cells is the filtration of the active solution.The  sc of solar cell with the filtration of DSi-6T : PCBM solutions is 20% lower than that of solar cell without filtration.The most DSi-6T donors remained in the filter due to their strong - interaction.However, the DH-6T cells without filtration of active solutions showed the poor photovoltaic performance.
For the development of OPVs with increased PCE and lifetime, OPVs with the TiO 2 interfacial layer between the active layer and the Al electrode were fabricated together with typical OPVs without TiO 2 .According to the literature, a Open circuit voltage ( oc ), short circuit current density ( sc ), fill factor (FF), and PCE at DSi-6T and DH-6T as electron donors: PCBM at weight ratio.The device performance was consistent and reproducible.The active area is 0.12 cm 2 .
the TiO 2 interfacial layer improved the PCEs and reduced the sensitivity of such devices to oxygen and water vapour [21,28].As shown in Figure 4, the FF of DSi-6T : PCBM cells with TiO 2 layer is enhanced up to 0.55 which is 30% higher than that of the solar cell.The best performance was observed on DSi-6T : PCBM 1 : 5 (w/w) blend cell with  oc of 0.63 V,  sc of 1.34 mA/cm 2 , FF of 55%, and power conversion efficiency of 0.44%.Also the FF of DH-6T cells is exactly 27% higher than that of solar cells.The DH-6T : PCBM 1 : 6 (w/w) blend cell with TiO 2 layer exhibited a  oc of 0.65 V,  sc of 0.57 mA/cm 2 , FF of 56%, and PCE of 0.21% (Figure 5).The incorporation of solution processed TiO 2 interfacial layer in the sexithiophene : PCBM BHJ devices significantly enhances FF, mainly due to the reduced charge recombination near active layer/Al interface.Figure 6 shows the external quantum efficiency (EQE) plot for the DSi-6T : PCBM (1 : 5 w/w) and DH-6T : PCBM (1 : 6 w/w) cells.The sexithiophene-based cells absorb the solar light in narrow visible range and show low EQEs due to a small amount of donor by limited solubility.For the DH-6T : PCBM solar cell, the maximum EQE is 11.3% at 340 nm that is caused by strong absorption of PCBM.When the DSi-6T was used, the DSi-6T : PCBM solar cell shows a higher EQE in the all wavelength, with a maximum of 16.1% at 410 nm.The absorption of the DSi-6T donor was increased in the EQE spectrum to be different from the UV-vis spectrum.It is expected that horizontal intermolecular packing due to the bulky silyl side chains of DSi-6T may increase the charge transfer at the interfaces, so the EQE increases.
To investigate the charge transporting property of the OPV devices, we measured hole mobility of the donor: PCBM blend layers with the space-charge limited-current (SCLC) model.Hole-only devices were fabricated with a device configuration of ITO/PEDOT : PSS/Donor : PCBM/MoO 3 /Al, because high work function of molybdenum oxide (MoO 3 ) blocks the injection of electrons from the Al cathode.The blend ratios of the Donor: PCBM layers were 1 : 5 (w/w)  The hole transport through the active layers is limited by the space charge and the SCLC is described by where  is the current density,  0 is the permittivity of vacuum,   is the dielectric constant of the material (assumed to be 3),  ℎ is the mobility,  is the applied voltage ( applied ) corrected from built-in voltage ( bi ) arising from difference in the work function of the contacts and voltage drop (  ) due to the series resistance of the electrodes, and  is the thickness of the blend layer [29].Equation ( 1) is valid when the mobility Figure 7: (a) ln ( 3 / 2 ) versus  1/2 curves according to the SCLC equation ( 2).(b)  1/2 versus  curves showing the electric-field-dependence of hole mobilities at highly biased region.Dotted and solid lines are calculated curves with (1) and ( 2), respectively.is field independent.It has been found that the behaviour of electric field dependence of the mobility can be described by an empirical rule;  ℎ =  ℎ0 exp( 1/2 ), where  ℎ0 is the hole mobility at zero field,  is field dependence prefactor, and  is the electric field.Therefore, the electric field dependent SCLC model can be expressed by The zero-field hole mobility  ℎ0 and the prefatory  of DSi-6T : PCBM (1 : 5 w/w) and DH-6T : PCBM (1 : 6 w/w) blend films were obtained from (2) as shown in Figure 7(a) ( ℎ0 = 7.7 × 10 −6 cm 2 /Vs,  = 3.3 × 10 −4 m 0.5 V −0.5 for DSi-6T : PCBM, and  ℎ0 = 1.6 × 10 −6 cm 2 /Vs,  = 4.5 × 10 −4 m 0.5 V −0.5 for DH-6T : PCBM).The low  of DSi-6T : PCBM indicates low energetic disorder related to the interaction of each hopping charge in randomly oriented dipoles [30,31].Figure 7(b) shows the SCLC fitting of DSi-6T : PCBM (1 : 5 w/w) and DH-6T : PCBM (1 : 6 w/w) blend films according to (1) and (2).It is observed that the SCLC calculated from the field independent model deviates from the experimental data at the high voltage region.In actual operating devices, the active materials are in the effective bias of ∼1 V (corresponding to the energy level difference of HOMO and LUMO of an electron donor and acceptor) at the short circuit conditions.Taking this into consideration, we estimated the field dependent hole mobility of the active blend layers at 1 V;  ℎ = 2.7 × 10 −5 cm 2 /Vs for DSi-6T : PCBM, and  ℎ = 8.8 × 10 −6 cm 2 /Vs for DH-6T : PCBM.The higher hole mobility of the DSi-6T : PCBM blend film than that of the DH-6T : PCBM film is a key property of the enhanced photovoltaic performance.It seems that the higher hole mobility of DSi-6T : PCBM blend layer is inconsistent with the pure donor's field-effect mobility.However, it should be noted that the mobility in organic thin-film transistors strongly depends on the molecular orientation due to the highly anisotropic charge carrier mobility.In the OPV device configuration, the DSi-6T : PCBM BHJ layer yielded the better vertical charge transport than the DH-6T : PCBM due to the improved solubility of DSi-6T.
To study the thermal annealing effect of this system, the morphologies of sexithiophene: PCBM films before and after annealing at 60 ∘ C were studied by atomic force microscopy (AFM) (Figure 8).The morphological studies indicate the more amorphous nature of DSi-6T : PCBM compared to DH-6T : PCBM films.The root-mean-square (rms) roughness is 47, 38, and 23 nm for DSi-6T : PCBM 1 : 4, 1 : 5, and 1 : 6 (w/w), respectively, and 7.1 nm for DH-6T : PCBM 1 : 6 (w/w).In DSi-6T : PCBM films, the weight ratio donor to acceptor increased as the roughness of films decreased.After thermal annealing, the roughness was 37, 29, and 14 nm for DSi-6T : PCBM 1 : 4, 1 : 5, 1 : 6 (w/w), respectively, and 7.6 nm for DH-6T : PCBM 1 : 6 (w/w).For the solar cell based on the DSi-6T : PCBM cell, the rms roughness of films after thermal annealing at 60 ∘ C was significantly reduced as compared to that of films.In the case of DH-6T : PCBM cell, the morphological change after annealing was slightly observed.Unfortunately, all films were quite rough compared to the polymer solar cells.The inferior morphologies of sexithiophene based BHJ solar cells may be related to the solubility of donors, which is expected to influence the photovoltaic performance.For a comparison, the DSi-6T : PCBM cell was thermally annealed at 135 ∘ C which was the temperature employed for the P3HT : PCBM cell.When a film was annealed at a high temperature, the decrease in