Effects of Hole-Collecting Buffer Layers and Electrodes on the Performance of Flexible Plastic Organic Photovoltaics

Here we report the influences of the sheet resistance (Rsheet) of a hole-collecting electrode (indium tin oxide, ITO) and the conductivity of a hole-collecting buffer layer (poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate), PEDOT:PSS) on the device performance of flexible plastic organic photovoltaic (OPV) devices. The series resistance (R S ) of OPV devices steeply increases with increasing Rsheet of the ITO electrode, which leads to a significant decrease of short-circuit current density (JSC) and fill factor (FF) and power conversion efficiency, while the open-circuit voltage (VOC) was almost constant. By applying high-conductivity PEDOT:PSS, the efficiency of OPV devices with high Rsheet values of 160Ω/◻ and 510Ω/◻ is greatly improved, by a factor of 3.5 and 6.5, respectively. These results indicate that the conductivities of ITO and PEDOT:PSS will become more important to consider for manufacturing large-area flexible plastic OPV modules.


Introduction
Organic photovoltaic (OPV) devices have attracted considerable attention as a promising solution for next-generation solar energy conversion, due to their potential to realize low-cost, flexible, large-area, and simple solution processed photovoltaics [1,2].The power conversion efficiency (PCE) of OPVs has been greatly improved by introducing donoracceptor bulk hetero junction (BHJ) structures, with a PCE of about 5% being achieved for a single junction cell based on blend systems of poly(3-hexylthiophene) (P3HT) and [6,6]-phenyl-C 61 -butyric acid methyl ester (PCBM) [3][4][5].There is still further potential to increase the PCE to above 10% by using low band gap materials and/or tandem cell structures [1,6,7].To date, however, most studies have focused on performance improvement based on an indiumtin-oxide-(ITO-) coated rigid glass substrate so that the advantages of organic-based photovoltaic devices have not been fully exploited.A plastic substrate with an ITO electrode is considered the best candidate to achieve a fully flexible OPV [8].To increase the transparency and conductivity of an ITO thin film, it is necessary to heat the substrate at 200∼300 ∘ C to enhance the ITO crystallization as well as to facilitate diffusion and substitution of tin ions during the ITO deposition process.With a plastic substrate, however, heating beyond that temperature is difficult due to its poor thermal stability, and thus the conductivity of ITO becomes lower than that of glass-based substrates.Probable limiting factor of the PCE for flexible plastic substrate-based OPVs relative to OPVs based on glass substrates are the low conductivity of the ITO electrode and the temperature limitation of the postthermal annealing process of the photoactive layer due to the weak thermal properties of plastic substrates.
Kang et al. reported that reducing  sheet of the ITO electrode led to enhanced device performance by decreasing the   value of OPV devices [9].Therefore, it is imperative to mitigate the high  sheet (or bulk resistance) problem of ITO electrodes in flexible plastic substrate-based OPVs and/or large-area OPV module products.In case of using the same photoactive material and cathode electrode, the conductivity of the hole-collecting buffer layer (HCBL) is another factor related to the   of OPV devices.The most commonly used HCBL material for OPVs is PEDOT:PSS, which plays an important role in organic electronics such as controlling the hole transport conditions, reducing surface roughness of the substrate, and converting nonohmic contacts into ohmic states.Representative commercially available PEDOT:PSS include CLEVIOS P VP AI 4083 (AI4083) and CLEVIOS PH 500 (PH500), with conductivity of 10 −3 S cm −1 and 10∼50 S cm −1 , respectively [10,11].It is well known that a highly conductive PEDOT:PSS would be helpful to enhance PCE by providing fast charge transfer with small resistance as well as improved surface or interface properties.Ko et al. [12] and Hu et al. reported that improved device performance could be attributed to the reduction of the resistance of PEDOT:PSS by doping organic solvents [13].Xiao et al. [14] and Sun et al. suggested that enhanced performance can also be achieved through an increase of light absorption in the active layer by improving light scattering and by obtaining high surface area with solvent-treated PEDOT:PSS [15].Lastly, improvement of interfacial properties and alignment of surface free energy between PEDOT:PSS and the active layer could be another factor to facilitate greater device performance [16][17][18].However, the aforementioned experiments were mostly performed on glass-based low  sheet ITO electrodes (∼10 Ω/◻).Therefore, the influence of the PEDOT:PSS conductivity on the performance of OPVs with a high  sheet ITO electrode warrants further study for understanding delicate changes in final solar module performances when flexible plastic OPV devices are commercialized.
In this work, we investigated the effect of  sheet of ITO electrode on the performance of OPV devices and also assessed whether the high  sheet of an ITO electrode on a flexible substrate could be compensated by using highconductivity PEDOT:PSS in the normal configuration of flexible substrate/ITO/PEDOT:PSS/P3HT:PCBM/LiF/Al.Our results show that the conductivity of ITO and PEDOT:PSS strongly affects devices parameters including short-circuit current density ( SC ), fill factor (FF), and PCE through their influence on the device   .
The OPV devices were fabricated as follows: the patterned ITO substrates were sequentially cleaned by ultrasonic treatment in methanol, de-ionized water, and isopropyl alcohol and dried in a vacuum oven.After being treated with UV-O 3 for 15 min, PEDOT:PSS was immediately spin-coated on the ITO substrate.The coated PEDOT:PSS films were baked on a hot plate for 15 min at 120 ∘ C to give a thickness of 30 nm and transferred to a glove box filled with N 2 gas for the remaining fabrication steps.P3HT and PCBM with weight ratios of 1 : 0.9 were dissolved in chlorobenzene, followed by stirring for 24 h at 50 ∘ C.An active layer consisting of P3HT:PCBM blends was spin-cast on top of the PEDOT:PSS layer and dried on a hot plate for 40 min at 50 ∘ C to give a thickness of 120 nm.A top electrode consisting of a LiF (1 nm) layer and a subsequent Al (100 nm) layer were deposited by thermal evaporation under a vacuum of ∼10 −7 Torr.The deposited Al electrode defines an active area of the device as 0.09 cm 2 .The prepared devices were directly annealed on a hot plate for 15 min at 120 ∘ C.Although the device annealing process is generally performed at 140 ∘ C∼150 ∘ C for glass-based OPVs in order to develop well-ordered active layer morphology and enhance the device efficiency, we conducted the annealing process for all devices including a glass/ITO reference at 120 ∘ C because of device failure of our plastic/ITO-based OPV devices annealed at temperature over 130 ∘ C.

Measurements.
The conductivity of the ITO electrode was measured in a unit of Ω/◻ ( sheet ) by the 4-point probe method (CMT-series, Advanced Instrument Technology) [19], and the transmittance was measured with a UV-Vis spectrometer (Lambda 750, PerkinElmer).Photocurrent density-voltage (J-V) measurements were performed with a Keithley model 2400 Source Meter and a Newport 91192 solar simulator system (equipped with a 1 KW xenon arc lamp, Oriel).Light intensity was adjusted to simulated AM1.5 radiation at 100 mW/cm 2 with a Radiant Power Energy Meter (model 70260, Oriel).The   value of OPV was calculated from the inverse slope of the J-V curve at high cell voltages ( = 0).All measurements were carried out under ambient conditions at room temperature.

Results and Discussion
To compare the performance of OPVs according to the conductivity of the ITO electrode and the PEDOT:PSS buffer layer, we fabricated several identically structured devices, but with different substrates of glass/ITO (6 Ω/◻), PEN/ITO (21 Ω/◻), PES/ITO (160 Ω/◻), and PET/ITO (510 Ω/◻) and also with different PEDOT:PSS of AI4083, PH500, as shown in Figure 1 (note the chemical structures of plastic substrates just below the device structures in Figure 1(b)).The conductivities of two different PEDOT:PSS samples are also given in this figure.As shown in Figure 1(c), the OPV devices fabricated with the plastic substrates were bendable and flexible even in small sizes compared to that fabricated with glass substrate.
The optical transmittance spectra of ITO-coated glass and three different ITO-coated plastic substrates are given in Figure 2. The transmittance at the wavelength between 450 nm and 550 nm was slightly lower for the plastic substrates than the glass substrate, whereas it was reversed Glass substrate Flexible substrate between 550 nm and 650 nm for the PEN and PET substrates (note that the P3HT absorption ends at around 650 nm).Interestingly, the PES substrate exhibited a sensitive transmittance response with the wavelength, which is different from other substrates.The current density-voltage (J-V) characteristics of OPV devices under illumination of the simulated solar light (AM1.5, 100 mW/cm 2 ) are shown in Figure 3.All devices show similar open-circuit voltage ( OC ) of about 0.6 V, suggesting good contact between the active layer and the buffer layers.For the devices using AI4083, their  SC , FF, and PCE are, however, significantly decreased, from 6.2 mA/cm 2 , 0.605, and 2.3% in the glass/ITO device to 0.5 mA/cm 2 , 0.139, and 0.04% in the PET/ITO device, respectively.We note that the relatively low PCE of 2.3% in the glass/ITO device was due to the low postannealing temperature at 120 ∘ C after device fabrication, as mentioned in the Experimental section (our normal device has a PCE of over 3% after postannealing treatment at 140∼150 ∘ C, as reported previously [20,21]).
Based on the J-V characteristics, we tried to analyze the main factor to decrease the PCE with increasing substrate  sheet .The efficiency of an OPV is typically influenced by the following relation [22]: where   is the photon absorption efficiency,   is the generation efficiency of excitons,  diff is the exciton diffusion efficiency,  dis is the efficiency of hole-electron separation (exciton dissociation),  tr is the efficiency of carrier transport in the active layer towards the electrodes, and   is the charge collection efficiency at the respective electrodes.We can assume that each device has equal capacity of   ,  diff ,  dis , and  tr , because their device structure including active (P3HT:PCBM) and buffer (AI4083) layers is exactly the same.  and   then remain as responsible factors.From Figure 1, the average optical transmittance of the ITO substrates at the wavelength of 400-700 nm is above 70%.related to   , is within 20%, while their  sheet difference is almost two orders.Thus, the decreasing photovoltaic performance with increasing  sheet of the substrate is not significantly related to the transmittance of the base substrate, but rather to the high  sheet of the ITO electrode.This reflects that the main factor of limitation is the hole collection efficiency at the hole-collecting electrode (  ), not the light absorption difference (  ).Hence we tried to correlate the device performances with the conductivity of each substrate which can represent the hole collection efficiency in Figure 4.
Figure 4 presents the  OC ,  SC , FF, PCE, and   changes as a function of  sheet (ITO) of the substrates under illumination of AM1.5, which clearly shows the relations between these parameters.The  OC is almost independent of  sheet of ITO, as mentioned before.However, the  SC and FF values gradually decreased with increasing  sheet of the ITO, resulting in a decrease of the PCE.Moreover, the decreasing trend in  SC , FF, and PCE is closely connected with the increase of   of the devices.The hole collection efficiency at the anode (  ) is usually affected by   , which is also the dominant parameter determining the FF by altering the slope of the J-V curve near the  OC and is a key factor for high-efficiency solar cells [22][23][24][25].  represents the total resistance of the cell and depends on the resistivity of the organic material(s), the metal electrodes, and the metal/organic interface [24].As all devices were fabricated with the same structure and active materials under an identical process, it is reasonable to assume that the resistances of PEDOT:PSS, P3HT:PCBM, and LiF/Al remained constant among devices.Therefore,   is mainly changed with  sheet of the ITO anode in our case.Xue et al. reported that the contribution of  sheet of the ITO electrode to the   of an OPV is more critical in the fabrication of the OPV for a large area of more than 0.01 cm 2 All measurements were performed under illumination of simulated solar light (AM1.5, 100 mW/cm 2 ).[26].Several other reports have described the effect of  sheet on the   and the cell design rule to minimize the   in large area devices [27][28][29][30][31].
As additional indirect evidence for poor hole extraction (collection) due to the high-resistivity ITO, we can see a kink (or S-curve) characteristic in devices with high   such as PET/ITO/AI4083 and PES/ITO/AI4083.This kink causes a negative curvature that can substantially reduce the FF, and therefore the solar cell performance.The kink of the J-V curve under illumination was due to several reasons, such as large interfacial energy gaps at the electrode-active layer  ) interface, unbalanced mobility between holes and electrons [32,33], nonuniform contact or incomplete coverage of the metal over the active layer [34,35], slow charge transfer at one of the electrical contacts of the active layer [36], strong interfacial dipoles, defects, and traps that can create barriers for charge extraction [37], and defect-induced nonefficient extraction of charge leading to buildup of a counter field [38,39].These can be summarized that the kink characteristic is caused by a counterfield created by accumulated space charges from unbalanced charge carrier extraction.In devices incorporating our high-resistivity ITO substrate, we may reasonably suppose that separated holes undergo restricted extraction at the anode and show a kink J-V curve.Next, we checked whether these losses that originate from the high  sheet ITO can be reduced by using a highconductivity buffer layer.The conductivity of PEDOT:PSS can be increased by up to two or three orders of magnitude by adding high-boiling point and/or polar compounds such as diethylene glycol (DEG), ethylene glycol (EG), dimethylsulfoxide (DMSO), sorbitol, and glycerol, due to the phase separation of each PEDOT-rich region and PSS-rich region, as reported in previous works [40][41][42].We used PH500 doped with 5% DMSO (in this report, we call it PH500) to improve the conductivity up to 500 Ω/◻.From Figures 3 and  4, the  SC , FF, and PCE were greatly improved by using a highconductivity PH500 buffer layer.Typically, the effect of highconductivity PH500 is more pronounced with the high  sheet of the ITO electrode.As a result, PCE and   are, respectively, improved by a factor of 3.5 and 1.29 for PES/ITO and 6.5 and 4.25 for PET/ITO (see Table 1).This indicates that the poor conductivity of the ITO electrode can be partly compensated by a high-conductivity buffer layer via the reduction of   .Finally, we can suppose that an ITO electrode with similar or higher resistance than the buffer layer is not as helpful as an ITO-free case to extract hole carriers, because interfacial resistance at inorganic ITO/organic PEDOT:PSS inevitably arises.From this assumption, we fabricated an ITO-free device using PH500 (spin-coating at 2500 rpm, 30 nm, 500 Ω/◻) as an anode.

International Journal of Photoenergy
Figure 5 and Table 1 show that the ITO-free device has not only equivalent or higher  SC but also a high FF with an improved kink curve compared with devices made on PET/ITO or PES/ITO.This indicates that the interface property between ITO and PEDOT:PSS is also another important consideration when developing high-efficiency large-area OPVs, as discussed in many reports [16,17,24].

Conclusions
Eight different OPV devices using four different types of ITO substrates have been fabricated to investigate the influence of the hole-collecting electrode (ITO) and buffer layers (PEDOT:PSS).The optical transmittance of flexible plastic substrates was slightly lower than that of the glass substrate.The J-V curve shape became worse as the ITO conductivity of substrates increased irrespective of the kind of PEDOT:PSS, indicating significant effect of the  sheet of the ITO electrode in different substrates on the performance of OPV devices.However, the high-conductivity buffer layer (PH500) resulted in better device performances than the low-conductivity buffer layer (AI4083), which supports that the conductivity of hole-collecting buffer layer can compensate for the increase of   in OPV devices with a high  sheet ITO electrode.The present result showed that the high conductivities of holecollecting electrode (ITO) and buffer layer (PEDOT:PSS) are together beneficial to realize an efficient OPV devices by increasing the FF as well as the  SC through reduction of the   of devices.Therefore, new transparent electrode and buffer layer materials with high conductivity should be key technological components for successful commercialization of OPV modules.

Figure 1 :
Figure 1: (a) Chemical structure of materials used in this work, (b) device structures with four different ITO substrates (glass and flexible plastics) and two different buffer layers (AI4083 and PH500) (note the chemical structures of plastic substrates), and (c) photographs of OPV devices fabricated using glass (left) and flexible plastic (right) substrates.

Figure 4 :
Figure 4: (a) Open-circuit voltage ( OC ) and short-circuit current density ( SC ), (b) FF and PCE, and (c) series resistance (  ) as a function of the sheet resistance ( sheet ) of the ITO electrode on the substrates.

Table 1 :
Summary of device parameters for organic photovoltaic devices with four different ITO substrates (glass and flexible plastics) and two different buffer layers (AI4083 and PH500).  is calculated from the inverse slope of - curve at high cell voltages ( = 0).b The numbers in parentheses denote the ratio of PH500 to AI4083 (for  OC ,  SC , FF, and PCE) or the ratio of AI4083 to PH500 (for   ). a