1. Introduction

At present, carbon semiconductor nanomaterials have attracted increasing amount of interest. This is especially true for fullerene derivatives [1], as they are promising candidates for use in novel and high-performance opto-electronic devices [1–5]. Among them, one of the most successful devices is the organic photovoltaic cell (OPV), since the photo-induced transfer of electrons from a conjugated polymer donor to a fullerene acceptor was initially reported [5–7]. Fullerene-based OPVs have attracted much attention due to their numerous advantages. For instance, they are environmentally safe, involve only a low cost to cover a large area, are lightweight materials which are also mechanically flexible, and can be produced at high speeds using solution-coating or printing processes [5–9]. During the past decade, to improve the efficiency levels of OPVs, numerous studies have focused on the development of new materials [10–14] as well as on the optimization of device structures [15–19] with various fabrication processes, such as coating and printing techniques [20]. Thus far, fullerene-based PV layers which rely on the bulk heterojunction (BHJ) concept have shown the best performance with the efficient generation of photocurrent in OPVs [6–20]. In the BHJ layer, an electron acceptor of a fullerene derivative, e.g., phenyl C₆₁-butyric acid methyl ester (PCBM), is thoroughly blended with a photoactive conjugated polymer, e.g., poly(3-hexylthiophene) (P3HT), to form a bicontinuous interpenetrating networks of phase-separated materials with a large interfacial area, leading to efficient exciton dissociation and increased quantum efficiency. As a result, it is now possible to achieve efficient BHJ OPVs based on PCBM and P3HT (P3HT:PCBM) with a power conversion efficiency (PCE) of 3-5% under AM 1.5 (AM = air mass) illumination [9,18]. These BHJ OPVs, however, are associated with drawbacks related to their low carrier extraction stemming from their continuous connections and/or isolated islands of a single donor or acceptor material between the electrodes, which may not contribute to the generation of photocurrent [21]. In addition to the BHJ layers, there has recently been heightened interest in the use of vertical composition gradients [22–25], in which the donor and the acceptor materials are segregated toward the electrodes. This is believed to enhance device performance levels through efficient exciton dissociation due to the thorough intermixing of the components while also making the charge transport process more efficient. Inverted and/or bilayer structures are common means of obtaining effective vertical phase separation [25–29]. In recent reports, using a PCBM electron-acceptor layer with optimization of the morphology and the structure, high PCE levels of 3.8-4.0% were reported for a bilayer OPV with P3HT and PCBM, thereby demonstrating one of the best performances to date for a P3HT/PCBM bilayer OPV [25,30].

Despite the foregoing achievements, however, relatively little progress in high throughput processing has thus far been made in the design of a reliable and simple fabrication process that ensures the formation of flat, uniform layers of fullerene derivatives over large areas, which is particularly important for achieving efficient and reliable device performance levels. During the fabrication of bilayer heterojunction OPVs, thin functional layers, especially for the electron-acceptor layer of the fullerene derivatives, can be prepared by means of vacuum evaporation [23,29] or a wet solution-coating process [21,24,25]. Although bilayer heterojunction OPVs manufactured using a fullerene derivative layer prepared via the vacuum-deposition process have a good performance record [23,29], the vacuum-deposition process itself is relatively complex and expensive. The application of the vacuum-evaporated fullerene derivative layer has therefore been limited, and relatively little progress has been made with regard to the development of low-cost processable electron-acceptor layers, restricting their application across a wide range of possible devices.

Meanwhile, solution-coating processes are also of interest for preparing fullerene derivative layers, as these techniques involve a simple production technique [21,24,30–32]. In laboratory fabrication conditions, most fullerene-based devices are typically fabricated by means of a spin-coating method [24,25,30,31]. For example, PCBM electron-acceptor layers were formed by spin-coating in bilayer OPVs, yielding PCEs of approximately 2.6-3.5% [24,31]. However, although the spin-coating method is simple and convenient, it has several serious disadvantages, including the high stress caused by the spinning motion, the poor uniformity of the coated films, and the large amounts of wasted solution. Further, with spin-coating, the underlying pre-deposited layers may dissolve during the fabrication of multilayer devices [31–34]. These factors make spin-coating unsuitable for the multiple coating of large and homogeneous areas. Thus, solution-processable PCBM layers have not been studied widely with regard to their fabrication and characterization, with the exception of a small number of alternative methods such as spray deposition [21,32]. Using these deposition methods, PCBM layers may be formed on substrates in a controlled fashion, but the performance of the devices is not ideal [21,32]. Thus, with regard to the continued difficulty of controlling the uniformity of large-area PCBM layers, further research on solution deposition techniques is required to achieve the simpler and more reliable fabrication of large-area PCBM active layers.

We therefore focus our research on a solution-based preparation method for thin PCBM functional layers in ambient atmospheres without the need for a vacuum process. This development can be of key importance with regard to the realization of solution-processable multilayer devices such as highly efficient bilayer OPVs. We herein introduce an alternative approach for the preparation of PCBM active layers on P3HT layers which involves the successive solution deposition of a simple horizontal dip (H-dip) coating process [35–38] to form heterojunction photoactive layers using a pair of orthogonal solvents [24,25,31]. We demonstrate for the first time that H-dip coating produces quite homogeneous and uniform electron-acceptor layers of PCBM by controlling the operating parameters of the coating process and that the process is readily expandable to large-scale heterojunction photoactive layers. It is noted that the H-dipping method used for the PCBM layers can be easily transferable to a roll-to-roll (R2R) coating environment for fast, efficient, large-area, and reproducible fabrication which has the potential for use in future mass-production processes. We also added a surfactant additive to the homogeneous P3HT/PCBM bilayers to modify the interfacial properties of the bilayers and showed that the bilayer-based OPVs with additives exhibited PCEs as high as 3.95%, close to the highest values ever reported for the P3HT/PCBM bilayer heterojunction OPV system [24,25,31]. Moreover, we deposited H-dip-coated PCBM layers onto rubbing-induced oriented P3HT layers and showed that the bilayer OPVs thus fabricated also exhibited improved in-plane anisotropic PV effects with higher device performance levels than those of conventional rubbed BHJ OPVs.

2. Experiment

To fabricate PCBM films on solid substrates, we used PCBM (Nanostructured Carbon Inc.) dissolved in dichloromethane at a concentration of 2.0 wt%. The PCBM was used as received without further purification, and the PCBM solution was mixed thoroughly for 2 hours at room temperature using a magnetic stirrer. As shown in Fig. 1(a), the PCBM solution was coated onto the prepared substrates using the H-dip-coating method [35–38]. On the substrate, a thin PCBM film was deposited using an H-dip coater according to the following process: a small volume of the PCBM solution (5~10 μl) per unit of coating area (1 cm²) was fed into the gap of the cylindrical barrier using a syringe pump (NE-1000, New Era Pump Systems Inc.). The height h₀ of the gap was adjusted vertically using micrometer positioners mounted at the end of the coating barrier, and the coating speed U was controlled using a computer-controlled translation stage (SGSP26-200, Sigma Koki Co., Ltd). After a concave meniscus of the PCBM solution had formed on the substrate, the substrate was transported horizontally such that the meniscus formed by the coating barrier caused the solution to spread evenly on the transporting substrate while maintaining the shape of the meniscus of the solution. The thin PCBM films coated onto the substrates were then kept at room temperature in a dry air environment. For comparison, a reference PCBM layer was also prepared by means of the spin-coating method on substrates at 3,000 rpm for 35 s.

 

Fig. 1 (a) Schematic illustration of the horizontal-dip (H-dip) coating process described herein: a cylindrical coating barrier with radius R, gap height h₀, and coating speed U. The inset shows the chemical structure of the PCBM used. (b) Film thickness h of the H-dip coated PCBM film as a function of the coating speed U for two different gap heights h₀s. The solid curves show the theoretical predictions.

Download Full Size | PDF

The bilayer heterojunction OPVs on glass substrates were fabricated using coating solutions with a sandwich-like configuration of bilayers consisting of electron-donor and electron-acceptor layers. They were built according to the following procedure. An ITO layer (80 nm, 30 Ω /square) on a glass substrate used as a transparent anode was ultrasonically cleaned using a sequence of detergent, deionized water, acetone and isopropyl alcohol for 30 min, after which it was rinsed several times using deionized water and then dried using nitrogen. This was followed by an ultra-violet (UV) ozone treatment (5 min). The cleaned ITO anode was then coated with a poly (3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) hole-collective layer (HCL, thickness: ~40 nm) via the spin-coating of a commercial PEDOT:PSS solution (Clevios PVP AI 4083, H. C. Starck) and baked at 120°C in a vacuum oven for 20 min to extract the residual water solvent. To form the bilayer PV layers, we used orthogonal solvents of 1,2 dichlorobenzene and dichloromethane and prepared a solution of an electron-donor polymer of P3HT (1-material Chemscitech, Inc.) with a concentration of 10 mg/ml in 1,2-dichlorobenzene and a solution of a PCBM electron acceptor at a concentration of 6.4 mg/ml in dichloromethane [24,25,31]. For an electron-donor layer, a P3HT layer was spin-coated onto the PEDOT:PSS HCL and baked at 70°C for 2 min to reduce the residual solvent. The thickness of the P3HT layer was fixed at 65 nm. Sequentially, for an electron-acceptor layer, a PCBM layer was then fabricated by H-dip-coating the PCBM solution on the P3HT electron-donor layer to form bilayer heterojunction PV layers. Then, by thermal evaporation in a vacuum chamber, 1 nm of the Al:Li electron-collecting layer and 50 nm of the Al cathode were subsequently deposited on top of the PV active layers at a rate of 0.5 nm/s under a base pressure of less than 2.0 × 10⁻⁴ Pa. The active area of the fabricated device was 3 × 3 mm². Following the fabrication of the OPVs, the devices were annealed at 150°C for 10 min to induce the crystallization of the PV layer (i.e., post-thermal annealing [24,25,29,31]). For comparative purposes, we also fabricated a reference bilayer OPV using a spin-coated PCBM layer on the P3HT layer (reference 1). Accordingly, reference 1 was fabricated using a method identical to that used for the sample bilayer OPV with the H-dip-coated PCBM layer, apart from the differences in the PCBM layer as described above. For comparative purposes, we also fabricated another reference OPV based on a BHJ PV layer by spin-coating a blended solution of P3HT (1.20 wt%) and PCBM (0.88 wt%), dissolved in 1,2-dichlorobenzene at room temperature for 24 hours, to form a composite film of P3HT and PCBM (a P3HT:PCBM BHJ PV layer) on the PEDOT:PSS HCL (reference 2) [36]. The coated P3HT:PCBM BHJ PV layer had a thickness of about 85 nm. It is also important to note that, apart from the differences in the PV layer described earlier, all of the comparative reference 2 OPVs were fabricated using a method identical to that used to produce the sample bilayer OPV.

In additional, to fabricate the polarizing OPVs, alignment of the P3HT layer was done by rubbing the P3HT layer two times unidirectionally (along the x-direction) using a rubbing machine (Jaeil Optical System) with a roller (diameter: 6.0 cm) covered with a cotton velvet material (pile impression: 0.55 mm) at a roller rotation speed of 250 rpm and a substrate moving speed of 125 mm/min at a temperature of 110°C [39].

The optical microscopic morphology of the fabricated PCBM layer was observed using an optical microscopy (alpha300R, Witech GmbH). The surface roughness and topographic properties of the fabricated layers were characterized using atomic force microscopy (AFM, Nanosurf easyscan 2 AFM, Nanosurf AG Switzerland Inc.). The optical properties of the fabricated layers were also investigated using a UV-visible spectroscopy system (8453, Agilent).

The PV performance was measured using a source meter (2400, Keithley) and calibrated using a reference cell (BS-520, Bunkoh-Keiki) under illumination of 100 mW/cm² produced by an AM 1.5G light source (96000 Solar Simulator, Newport). The in-plane PV anisotropy was evaluated under polarized illumination (28.5 mW/cm²) produced by passing light from the solar simulator through a linear polarizer. The average power loss due to the polarizer was corrected. The reported values were averaged from several (at least 4) individual cells. The external quantum efficiency (EQE) spectra were obtained using a measurement system (Oriel® IQE-200 EQE/IQE, Newport). All of the fabricated PV cells were then tested under ambient conditions without encapsulation.

3. Results and discussion

3.1 Horizontal-dip-coated PCBM layers

Figure 1(a) shows a schematic illustration of the H-dip-coating [35–38] method used in this study as a solution-deposition process for the PCBM layers. The use of H-dip-coating allows for the thickness (h) of the coated layer to be increased with an increase in capillary number (Ca = μU/σ) of the coating solution. Here, μ and σ represent the viscosity and surface tension of the solution, respectively, and U is the coating speed. The thickness h of the H-dip-coated layer was described by the associated drag-out problem proposed by Landau and Levich as h = k·Ca^2/3·R_d, for Ca << 1, where R_d is the radius of the associated downstream meniscus and k is a constant of proportionality [35–38,40]. By applying the H-dip-coating process, we could successfully deposit PCBM layers onto the substrates and measure the film thickness. Figure 1(b) shows the film thickness of the H-dip-coated PCBM layers as a function of the coating speed U and gap height h₀. At an h₀ of 0.2 mm, the thickness of the PCBM layer showed a continuous increase from approximately 10 nm to 55 nm with an increase in U from 0.4 cm/s to 2.0 cm/s. Furthermore, when h₀ was increased to 0.3 mm, the thickness of the PCBM layer also showed a further increase as U was increased. These results are in good agreement with the theoretical predictions [solid lines in Fig. 1(b)]. It is therefore clear that H-dip-coating allows for precise control of the nanoscale thickness of solution-processable PCBM films through adjustments of U and h₀ as the controlling parameters.

3.2 Film properties of the H-dip-coated PCBM layers

Figure 2(a) shows optical microscope images of the spin- and H-dip-coated PCBM films (~35 nm-thick) on glass substrates at two different magnification levels ( × 500 and × 1000 zoom ranges). As shown in the images, large inhomogeneities were clearly visible for the spin-coated PCBM film with a large amount of PCBM aggregation over the entire film surface [upper image in Fig. 2(a)]. It should be noted that the dichloromethane solvent used has a low boiling temperature (~40°C); thus, rapid evaporation of the solvent may induce aggregation of the PCBM molecules during the spin-coating process, resulting in poor uniformity. Such an inhomogeneous PCBM film would indicate that the widely used spin-coating process is not suitable for the formation of PCBM film samples. In contrast, interestingly, the optical microscope images of the H-dip-coated PCBM film show almost no spatial variation [lower image in Fig. 2(a)], implying improved uniformity of the H-dip-coated PCBM films. This result is consistent with the findings of recent studies which showed that homogeneous and uniform bilayer and even multilayer organic/inorganic films can be obtained by means of the H-dip-coating process [36,37]. Thus, with H-dip coating process, it is possible to obtain a homogeneous PCBM films with proper nanoscale thicknesses.

 

Fig. 2 (a) Optical microscopy images of PCBM films fabricated on glass by spin- (upper) and H-dip-coating (lower) at two magnification levels ( × 500 and × 1000). (b) Topographical AFM images of the spin- (upper) and H-dip-coated (lower) PCBM films fabricated on P3HT film. (c) Optical transmission spectra of the PCBM films on glass substrates. (d) Photograph of the H-dip-coated PCBM film on glass substrates (as seen on the upper half of the glass).

Download Full Size | PDF

Next, in order to examine the film quality of the fabricated PCBM films, we obtained AFM topographic images of the surfaces of the spin- and H-dip-coated PCBM films (35 nm-thick) on flat 65 nm-thick P3HT films (rms roughness ~3.0 nm). Figure 2(b) shows the AFM images of a 10 × 5 µm² scanned area of the PCBM films. The AFM results clearly show that the topography for the spin-coated PCBM layer was fairly inhomogeneous and rough, whereas that of the H-dip-coated PCBM layer was relatively uniform and smooth. In comparison, the rms surface roughnesses observed for the spin- and H-dip-coated PCBM films were approximately ~4.7 nm and ~3.3 nm, respectively. Moreover, for the H-dip-coated PCBM film, the rms values of the surface roughness were nearly identical at different positions of the investigated films. These microscopic AFM observations indicate that the use of H-dip-coating leads to the formation of high-quality films with good coverage for functional PCBM materials.

The optical transmission spectra of the PCBM films (thickness: ~35 nm) coated onto glass substrates are shown in Fig. 2(c). In the visible range, the H-dip-coated PCBM films on glass substrates showed high transmittance of about 80% at about 510 nm; this is comparable to that (ca. 80%) of the spin-coated film on glass substrates and that of a bare ITO substrate [see also Fig. 2(d)]. Note that the characteristic transmittance spectrum of the ITO substrate is mainly a result of the difference in the refractive index between the ITO layer (~1.90 at 550 nm) and the glass (~1.50 at 550 nm). The resulting smoothness and uniformity of the solution-processable PCBM layers clearly indicate that the functional PCBM layers, produced by H-dip-coating may improve the PV performance of bilayer heterojunction OPVs.

3.3 PV performance of bilayer heterojunction OPVs with the H-dip-coated PCBM layers

Our next investigation focused on the PV performance of the P3HT/PCBM bilayer heterojunction OPVs. The upper panel of Fig. 3(a) shows the device structure of the bilayer OPV used to investigate the effects of different PCBM layers on the PV performance. We fabricated bilayer OPVs with the following device configuration: an ITO anode/ PEDOT:PSS HCL (40 nm) / spin-coated P3HT electron-donor layer (65 nm) / spin- or H-dip-coated PCBM electron-acceptor layer (35 nm) / electron-collecting Al:Li buffer layer (1 nm) / Al cathode (50 nm). Here, for a straightforward comparison with previous results, the hole-collecting buffer PEDOT:PSS layer and the electron donor P3HT layer, but not the PCBM layer, were formed by the conventional spin-coating process. The lower panel in Fig. 3(a) shows representative top-view and cross-sectional SEM images of an H-dip-coated bilayer OPV.

 

Fig. 3 (a) Upper: Schematic device structure used in the P3HT/PCBM bilayer OPVs studied here. Lower: Top-view and cross-sectional SEM images of the H-dip-coated P3HT/PCBM bilayer OPV. (b) UV-vis optical absorption spectra of PCBM layers prepared by H-dip- and spin-coating on the P3HT layer after thermal annealing at 150°C for 10 min. The inset shows the chemical structure of the P3HT donor polymer. (c) Current density-voltage (J-V) characteristics of bilayer OPVs with spin- and H-dip-coated PCBM layers under AM 1.5G illumination conditions and in the dark (inset).

Download Full Size | PDF

For the fabricated P3HT/PCBM bilayers, we investigated the optical absorption properties of the bilayers after thermal annealing at 150°C for 10 min and observed that the PV active layers in the OPVs exhibited a strong optical absorption with a single broad absorption peak centered at approximately 510 nm with a band edge at approximately 660 nm [Fig. 3(b)]. The absorption in the visible region can be attributed mainly to the π-π* transition in the P3HT polymer. Note that the crystalline ordering of P3HT after thermal annealing may be confirmed by the pronounced vibronic absorption peaks (ca. 550 nm and 600 nm) [8]. These optical characteristics of both the spin- and H-dip-coated bilayer PV active layers were found to be similar to each other, despite the differences in the layer uniformities of the active layers.

Next, we investigated the current density-voltage (J-V) characteristics of the fabricated bilayer heterojunction OPVs. Figure 3(c) shows the representative J-V curves under dark and illumination conditions of two OPVs: a bilayer OPV with an H-dip-coated PCBM electron-acceptor layer (sample device) and a bilayer OPV with a spin-coated PCBM layer (reference 1). The dark J-V curves [inset of Fig. 3(c)] show that both OPVs clearly functioned as good diodes with high rectification ratios exceeding 10^4-5 at 1.5 V. However, the small but clear difference in the dark J-V curves depending on the coating method used to form the PCBM layer indicates that there may be some difference in the PV effects. In order to understand the PV effects of the bilayer OPVs, the J-V characteristics were also measured under AM 1.5G illumination (100 mW/cm²). The observed PV characteristics of the devices are also shown in the figure and are summarized in Table 1. For reference 1 with the spin-coated PCBM layer, reasonable PV performance was observed under illumination, with an open-circuit voltage (V_OC) of 0.60 V, a short-circuit current density (J_SC) of 7.52 mA/cm², and a fill factor (FF) of 44.43%, corresponding to a PCE value of 2.00%. This is somewhat lower than those values (~2.7%) in previous studies of bilayer OPVs with an evaporated PCBM active layer [29]. The PV performance of reference 1 mainly originates from the low work function (or energy level) of the PEDOT:PSS HCL-coated ITO anode (~5.0 eV), which is close to the HOMO energy level of P3HT (~5.2 eV) [24,41] and which increases the built-in potential across the cell.

Table 1. PV performances of spin- and H-dip-coated bilayer and BHJ OPVs with and without PTE additives^a

View Table

In contrast, interestingly, for the sample device with the H-dip-coated PCBM layer, improved PV performance was observed under illumination, as shown in the figure, with a V_OC of 0.62 V, a J_SC of 9.40 mA/cm², and a FF of 48.24%, corresponding to a PCE value of 2.89%. This PCE is notably higher than that of reference 1, by 45%, and comparable to previous results of bilayer OPVs with an evaporated PCBM active layer. Because the energy levels of the functional layers for the both sample and reference 1 bilayer OPVs were identical to each other, we mainly attribute this improvement in the PV performance of the sample device not only to the large built-in potential inside the devices but also to the improved electron-transferring properties of the homogeneous PCBM layer on an P3HT layer, resulting in significantly increased FF and J_SC values [see Table 1 and Fig. 3(c)]. It is therefore clear from the above results that the H-dip-coated PCBM electron-acceptor layers are more effective than conventional spin-coated PCBM layers.

In order to verify the difference in the PV effects due to the uniformity of the PCBM electron acceptor layer, we subsequently investigated the PV performance using focused illumination depending on the illumination positions of the two bilayer OPVs with the spin- and H-dip-coated PCBM layers. The photocurrents were observed and averaged at nine random illumination positions on the active areas using a focused laser beam (incident laser wavelength: 633 nm, power: 1 mW, beam size: ~0.6 μm). Figure 4 shows the typical J-V characteristics measured for the two bilayer OPVs investigated. For reference 1 with the spin-coated PCBM layer, the PV effects exhibited significantly large variations of the average PV performance with a V_OC value of 0.46 ± 0.01, a J_SC value of 0.36 ± 0.09 mA/cm², a FF value of 58.32 ± 7.20%, and a PCE level of 1.90 ± 0.44%. This result implies that the PV effects were inhomogeneous depending on the illumination point. In contrast, for the sample OPV with the homogeneous H-dip coated PCBM layer, the PV effects show only slight changes relative to each other, with the following average values: a V_OC of 0.48 ± 0.00 V, a J_SC of 0.46 ± 0.03 mA/cm², a FF of 56.73 ± 2.98%, and PCE of 2.46 ± 0.09%. Thus, it is clear that the H-dip-coated PCBM is more appropriate for efficient and homogeneous PV effects as compared to the spin-coated PCBM layer for bilayer heterojunction OPVs. It may therefore be expected that during the fabrication of efficient bilayer OPVs, H-dip-coated PCBM layers can be used in place of conventional spin-coated PCBM layers.

 

Fig. 4 Representative J-V characteristics of bilayer OPVs with spin-coated (a) and H-dip-coated (b) PCBM layers, showing the highest, lowest, and mean current densities for nine different illumination points on the active layers of the bilayer OPVs by using focused monochromatic laser beam (wavelength: 633 nm, beam size: ~0.6 μm).

Download Full Size | PDF

3.4 Surfactant additives for improved PV performance levels of bilayer OPVs with H-dip-coated PCBM layers

Next, in order to realize a further improvement in the PV performance of bilayer OPVs with the H-dip-coated PCBM layers, we introduced the surfactant of poly(oxyethylene tridecyl ether) (PTE) [29] into the P3HT/PCBM active layers. The chemical structure of PTE is shown in Fig. 5(a). It may increase the PV performance via a reduction of the charge carrier recombination loss in the OPV [29,42]. Owing to the different hydrophobicities of the donor of P3HT and the acceptor of PCBM, some of the PTE surfactant molecules mixed in the active layers may form interfacial layers at the P3HT/PCBM interfaces, acting as an “interface engineering additive” [29,42]. Due to the low highest-occupied-molecular-orbital (HOMO, −8.1 eV) and high lowest-unoccupied-molecular-orbital (LUMO, −2.1 eV) levels with the dipolar characteristics of PTE [29,42], it is expected that the PTE surfactant at the interface may change the electrical properties of the interfaces between the P3HT and PCBM layers. In this study, we fabricated four types of bilayer PV active layers: one without any additives (P3HT/PCBM), one with additives via the mixing of ca. 0.04 wt% of the PTE surfactant into the P3HT layer (P3HT:PTE/PCBM), one with the PTE-mixed PCBM layer (P3HT/PCBM:PTE), and one with both the PTE-mixed P3HT and PCBM active layers (P3HT:PTE/PCBM:PTE). Note that hereafter, for comparisons with previous results, the upper electron-acceptor layers (PCBM and PCBM:PTE) in bilayer PV active layers were formed by the H-dip-coating process.

 

Fig. 5 Typical absorption spectra (a) and surface morphology images (b) of the PV layers studied. Typical current density-voltage (J-V) curves under illumination and in the dark (inset) (c) and EQE spectra (d) of the OPVs under investigation.

Download Full Size | PDF

The optical characteristics of the bilayers with and without PTE were observed using UV-vis absorption spectroscopy. Figure 5(a) shows the investigated absorption spectra of the P3HT/PCBM, P3HT:PTE/PCBM, P3HT/PCBM:PTE and P3HT:PTE/PCBM:PTE PV layers after a thermal treatment at 150°C for 10 min. As shown in the figure, similar to the P3HT/PCBM PV layer, the absorption spectra of the bilayers were nearly identical with a peak at around 510 nm and pronounced vibronic absorption peaks at around 600 nm in the visible region, clearly indicating that the PTE additives scarcely alter the optical characteristics of the bilayers.

Next, AFM topographic images of the three bilayers including PTE additives were obtained. As shown in Fig. 5(b), the three bilayers with PTE additives are fairly smooth. Compared to the surface morphologies of the P3HT/PCBM layer (rms roughness: 3.3 nm)[Fig. 2(b)], the spin-coated single P3HT (rms roughness: 3.0 nm), and the spin-coated single P3HT:PTE (rms roughness: 4.1 nm), it is clear that a similar surface with some degree of surface roughness was formed on the P3HT:PTE/PCBM layer (rms roughness of 4.1 nm), as shown in the figure. Moreover, similar results were observed on the surfaces of the P3HT/PCBM:PTE layer (rms roughness: 3.4 nm) and the P3HT:PTE/PCBM:PTE layer (rms roughness: 4.1 nm). These findings provide evidence that the addition of a small amount of PTE surfactant into the P3HT/PCBM PV layer has only a very minor effect on the surface morphologies of the active layers.

We also investigated the current flows and the PV effects of bilayer OPVs with the PV layers of P3HT/PCBM, P3HT:PTE/PCBM, P3HT/PCBM:PTE, and P3HT:PTE/PCBM:PTE, as shown in Fig. 5(c). In a dark condition, the bilayer OPVs showed clear and nearly identical diode behavior with high rectification ratios exceeding 10⁴. Under illumination, fairly interesting performance was observed. Table 1 summarizes the PV performance levels of several bilayer OPV devices with various concentrations of PTE additives in the P3HT and PCBM layers. As an example, as shown in the figure, with an appropriate PTE concentration (0.04 wt%), the bilayer OPV with the P3HT:PTE/PCBM PV layers exhibited a V_OC of 0.62 V, a J_SC of 10.49 mA/cm², a FF of 54.07%, and PCE of 3.52%, which was as much as 22% higher than the PCE of the P3HT/PCBM bilayer OPV without any PTE. Similarly, the bilayer OPV with the P3HT/PCBM:PTE PV layers with a PTE concentration of 0.04 wt% showed a V_OC value of 0.62 V, a J_SC of 10.25 mA/cm², a FF of 53.23%, and PCE of 3.40%, which again was up to 18% than that of the P3HT/PCBM bilayer OPV. We attribute these improved PV effects to the increased dissociation efficiency of the charge carrier via the interfacial dipole effects of the PTE molecules at the interface between the P3HT and PCBM layers. These results also indicate that the PTE additives lead to the selective collection of charge carriers at the electrodes. More interestingly, a further improvement of the PV effects was observed in bilayer OPVs in which both the P3HT donor layer and the PCBM acceptor layer were mixed PTE additives, i.e., P3HT:PTE/PCBM:PTE PV layers, which showed a V_OC of 0.63 V, a J_SC of 11.49 mA/cm², a FF of 54.40% and PCE of 3.95%, which was as much as 37% higher than that of P3HT/PCBM bilayer OPVs without PTE. These improved PV effects can be attributed to both the increased dissociation efficiency of electron-hole pairs at the P3HT/PCBM interface and the selective collection of charge carriers at the electrodes [29,42].

For comparison, we also fabricated and investigated BHJ OPVs (P3HT:PCBM), and compared them to the bilayer OPVs studied here. In this case, the BHJ OPV device configuration investigated was as follows: ITO / PEDOT:PSS / spin-coated P3HT:PCBM (without PTE) or P3HT:PCBM:PTE (with PTE) / Al:Li / Al. The observed device performance of the BHJ OPVs is also summarized in Table 1, and the representative photo J-V curves are shown in Fig. 5(c). For both the BHJ OPVs, we observed high device performance levels, with PCE values of 3.57% for the P3HT:PCBM BHJ OPV, and 4.13% for the P3HT:PCBM:PTE BHJ OPV, which are also comparable to the results of earlier studies [8,9,18]. From the comparison with such high PV effects of BHJ OPVs, it is clear that the PV performance levels of the homogeneous bilayer OPVs with PTE (P3HT:PTE/PCBM:PTE) were greatly improved compared to those of a conventional spin-coated bilayer device, with the PTE-doped bilayer OPVs even comparable to their counterpart BHJ OPVs.

The improved PV performance of the homogeneous bilayer OPVs with the PTE additives was also confirmed by observing the external quantum efficiency (EQE) spectra. Figure 5(d) shows the representative EQE spectra of several OPV devices studied here. The responses of bilayer devices are similar to those of BHJ devices in the region from 400 to 650 nm. The observed EQE spectra are also similar to the absorption spectral shape of the PV layers [Fig. 5(a)], indicating that the photocurrents are generated mainly by dissociated excitons produced by absorbed photons in P3HT. It can also be observed that the EQE values are consistent with the variations in J_SC for the OPVs with various PV layers. The maximum EQE value was 62.6% at 500 nm for the P3HT:PTE/PCBM:PTE bilayer OPV, which is higher than that (~50.5%) of the P3HT/PCBM bilayer OPV. Moreover, this EQE value for the P3HT:PTE/PCBM:PTE bilayer OPV is comparable to (or slightly lower than) those of P3HT:PCBM BHJ OPVs (61.2%) and P3HT:PCBM:PTE BHJ OPVs (68.9%). These PV results clearly indicate that photo-generated electrons in the interface-engineered bilayer PV active layers can efficiently be transported through the homogeneous H-dip-coated PCBM layers into the Al cathode. Although the underlying mechanism behind this is not yet clear, we attribute this improvement largely to the formation of sufficiently homogeneous functional layers of PCBM with PTE, which fill any pin-holes and overcome other shorting effects, resulting in increased internal resistance and therefore greater overall efficiency. It is noteworthy that when the PTE additives were introduced into the spin-coated bilayer OPVs, the effect of the PTE additives was insignificant on the PV performance of the spin-coated OPVs, possibly due to the irregularity and inhomogeneity of the top spin-coated PCBM layer on the P3HT layer, despite the introduction of PTE additives in the P3HT/PCBM bilayers. More detailed information pertaining to the device performance of the spin-coated OPVs will be reported elsewhere.

3.5 Anisotropic PV performance of polarizing bilayer OPVs with H-dip-coated PCBM layers

Finally, in order to verify the device performance capabilities of the bilayer OPVs produced using the H-dip-coated PCBM layer, we also investigated the anisotropic PV effects in the bilayer OPVs, i.e., polarizing OPVs [39]. To obtain polarizing PV effects in the bilayer OPVs, we fabricated an oriented P3HT polymer film as the anisotropic electron donor layer in our bilayer OPVs via a mechanical rubbing method [39]. Thus, the polarizing bilayer OPV has the following configuration: ITO / PEDOT:PSS / rubbed P3HT:PTE / H-dip-coated PCBM:PTE / Al:Li / Al. For comparison, we also fabricated rubbed BHJ OPVs with the configuration of ITO / PEDOT:PSS / rubbed P3HT:PCBM:PTE / Al:Li / Al.

The polarized absorption spectra of the rubbed bilayer and BHJ PV layers [Fig. 6(a)] also exhibited strong absorption at ca. 510 nm with pronounced vibronic absorption peaks due to the intermolecular interactions of the P3HT. Moreover, both the rubbed bilayer and the rubbed BHJ layer show much higher absorption (A_||) levels for incident light polarized parallel to the direction of rubbing (x-direction) than those (A_┴) for incident light polarized perpendicular to the x-direction, showing that the P3HT was unidirectionally aligned along the x-direction [39]. The observed dichroic ratio (DR = A_||/ A_┴) of the rubbed bilayer was 4.13 at 590 nm, which was higher than that (3.36) of the rubbed BHJ PV layer and than those (~1.61) of the rubbed BHJ PV layers in the previous work [39]. This result indicated that the rubbing-induced anisotropy of the rubbed bilayer is higher than that of the rubbed BHJ layer.

 

Fig. 6 (a) Polarized absorption spectra of the rubbed bilayers and the rubbed BHJ PV layers for incident light polarized parallel and perpendicular to the rubbing direction of the PV layers. Polarizing current density-voltage curves (b) and EQE spectra (c) of the rubbed bilayer OPVs and the rubbed BHJ OPVs for incident linearly polarized light.

Download Full Size | PDF

We also observed the PV performance capabilities of the fabricated devices. When tested under an unpolarized AM 1.5G illumination condition, the rubbed bilayer OPV reached a PCE value of 2.75%, with a J_SC of 8.14 mA/cm², a V_OC of 0.63 V, and a FF of 53.36%. Note that this device performance of the rubbed bilayer OPV is better than that of the rubbed BHJ OPV with a PCE value of 2.52%, despite the reduction of the performance after the rubbing process, which may have been primarily caused by the increased interface resistance between the rubbed layer and the adjacent layer [39]. We then investigated the polarization-dependent J-V characteristics of the rubbed bilayer and BHJ OPVs under polarized light illumination (28.5 mW/cm²), as shown in Fig. 6(b). The rubbed bilayer OPVs showed a maximum PCE value of 2.89% for incident light polarized parallel to the aligned (x-) direction (PCE_||), and a minimum PCE value of 1.59% for the incident light polarized perpendicular to the x-direction (PCE_┴), making the anisotropy of the PCE (PCE_||/PCE_┴) as high as 1.81. It is also noteworthy that the observed anisotropy (PCE_||/PCE_┴ = 1.81) of the PCE of the bilayer device is clearly higher than that of the rubbed BHJ OPVs, showing a PCE_|| value of 2.61% and a PCE_┴ value of 1.70% with a PCE_||/PCE_┴ value of approximately 1.54. It is also higher than those (1.22-1.42) of the rubbed BHJ OPVs in the previous work [39]. This result clearly indicated that the rubbing-induced PV anisotropy of the H-dip-coated bilayer OPV can be considerably higher than that of the conventional BHJ OPV.

The polarized PV performance capabilities of the rubbed bilayer OPVs were also confirmed by measuring the polarizing EQE spectra, as shown in Fig. 6(c). It can be seen from the spectra that the EQE values are consistent with the variations in J_SC for both the rubbed bilayer and the rubbed BHJ OPVs under polarized light illumination. For light polarized parallel to the x-direction, the maximum EQE (~10.9% at 510 nm) of the bilayer OPV was higher than that (~9.9%) of the BHJ OPV, while for light polarized perpendicular to the x-direction, the minimum EQE of 6.1% of the bilayer OPV was less than that (6.9%) the BHJ OPV. These results provide clear evidence of the strong anisotropic PV activity in the bilayer OPV with an oriented P3HT donor layer and H-dip-coated PCBM acceptor layers doped with PTE additives. We therefore believe there is a clear case for using such anisotropic PV activity in new energy-harvesting polarizing opto-electrical devices.

All of the results described above clearly demonstrate that H-dip-coated PCBM layers with PTE additives show considerable promise as a potential alternative to the conventional solution-coated PCBM layers for the fabrication of homogeneous, uniform, and highly efficient bilayer OPVs. It should also be noted that by applying the concept utilized with the H-dip-coating method to other coating methods such as self-metered slot-die coating [43], a combined coating method may be further developed for solution processing in high-throughput manufacturing areas such as R2R production, which would enable the rapid processing of low-cost bilayer OPVs. Furthermore, this method can be applied to fabricate other fullerene-based thin-film devices which are sensitive to the uniformity of the functional film.

4. Summary

In summary, we have investigated a solution-processable PCBM layer fabricated by the H-dip coating process to improve the PV effects of bilayer OPVs. It was shown that homogeneous and uniform depositions of PCBM layers can be achieved by controlling the meniscuses of the PCBM solutions applied during the H-dip coating process. It was also shown that the PCE (~2.89%) of P3HT/PCBM bilayer OPVs fabricated with the H-dip-coated PCBM layer was higher than that (~2.00%) when the conventional spin-coated PCBM layer was used. Moreover, by introducing a PTE surfactant as an interface-engineering additive, P3HT/PCBM bilayer OPVs with an H-dip-coated PCBM layer showed a PCE value of 3.95%, which is comparable to those of conventional spin-coated BHJ devices (3.57-4.13%), demonstrating that H-dip-coating with PTE surfactants is an effective means of fabricating bilayer OPVs. Furthermore, it was also demonstrated that the bilayer OPV with an H-dip-coated PCBM layer on an aligned P3HT layer exhibited improved in-plane anisotropic PV effects. These results clearly demonstrate that the use of the H-dipping process with the PTE additive may in the future be extended to conventional mass-produced slot-die and slit-die coatings and that the homogeneous PCBM layers produced by this novel process provide a promising foundation for the future development of various efficient, all-solution, and/or polarization-dependent opto-electrical devices.