1. Introduction

Organic photovoltaic (OPV) devices have great potentials for energy-generation applications in recent years due to its many advantages such as low cost, flexibility, light weight, large area, etc [1–5]. So far, polymer solar cells (PSCs) based on conjugated polymers blended with [6,6]-phenyl-C71-butyric acid methyl ester (PC₇₁BM) have been known to achieve power conversion efficiency (PCE) of roughly 10% [6]. On the other hand, small molecule organic semiconductors have attracted increasing interest for preparing OPVs due to the advantages of its well-defined structures, facile synthesis and purification, and generally high V_oc [7]. Up until now, the conventional small molecule solar cells (SM-OPVs) based on a two-dimensional conjugated small molecule material SMPV1 blended with PC₇₁BM, have achieved PCE of 8.0% [8,9].However, both PSCs and SM-OPVs are typically fabricated on indium tin oxide (ITO) substrates, and do not make full use of the processing advantages of organic materials. Therefore, there is an increasing demand for replacing ITO with alternative materials [10]. To achieve to this goal, thin metals, metal gratings, metal oxide thin films, and conducting polymers are being developed [11–17]. Although many research groups have successfully demonstrated the possibility of substituting the ITO with other transparent electrode layers using OPV materials, the conversion efficiency are still lower than what is desired. Studies of efficient ITO-free inverted OPVs are rarely reported. It is noteworthy that, compared to conventional solar cells, inverted solar cells are generally relatively more stable in air [7]. Therefore, the ITO-free inverted SM-OPVs are very promising for future energy-generation applications, especially for outdoor applications. This study demonstrates an efficient ITO-free inverted small molecule solar cells. The ITO transparent electrode layer has been replaced by AZO thin film, which exhibits high transmission (>85%) and low sheet resistance (~30 Ω/sq). To further enhance the light harvesting capability of the ITO-free inverted SM-OPVs based on AZO electrode, a ZnO nanorods interlayer is grown on the AZO layer to increase the absorption of the active layer. By introducing the ZnO nanorods interlayer, the PCE of the devices are improved from 4.0% to around 6.0%. Moreover, in order to optimize the performance, devices with various lengths of ZnO nanorods are fabricated and compared with each other. Furthermore, the experiment results agree well with the simulation result, which optimizes the optical performance of the devices under the condition of maximum optical absorption.

2. Experiments

To prepare the transparent AZO electrode layer, Al-doped ZnO thin films were deposited on glass substrates by PLD technique at 300°C. Initially, the chamber was pumped down to a base pressure of 5 × 10⁻⁶ torr. After that, the sintered Al-doped ZnO target (98 wt% ZnO + 2 wt% Al₂O₃) was ablated in an oxygen background of 40 mtorr, by KrF (248 nm) excimer laser with a fluence of 140 mJ and frequency of 5Hz. Then, the ZnO nanorods were grown on the AZO layer, synthesized using hydrothermal methods. The AZO/glass samples were suspended upside-down in a glass beaker, filled with an aqueous solution containing zinc nitrate hexahydrate (Zn(NO₃)₂·6H₂O, 50 mM), hexamethylenetetramine (HMT, 50 mM), and deionized (DI) water. During the growth process, the samples were baked at 90 °C in the oven for 30, 60, 90 and 120 min as shown in Figs. 1(a)-1(d), respectively. Immediately after the growth period ends, the substrates were rinsed with DI water, blown off by nitrogen gun, and dried on the hot plate at 250 °C for 10 min.

 

Fig. 1 (a) Scanning electron microscopy (SEM) cross-sectional image of ZnO nanorods grown for (a) 30 min, (b) 60 min, (c) 90 min and (d) 120 min.

Download Full Size | PDF

The layer structures of the devices consisted of glass substrate, AZO (110 nm), ZnO nanorods layer, SMPV1:PC₇₁BM (80nm), MoO₃ (5 nm), then Ag (150 nm) as shown in Fig. 2(a). The active layer (SMPV1:PC₇₁BM (1:0.8 by weight)) was spin coated at 2000 rpm, and annealed at 60°C for 2 minutes. The electrical characteristics of the solar cells were measured in glovebox at room temperature. To measure the efficiency of the solar cells, the devices were illuminated at 100 mW/cm² from a 150 W Oriel solar simulator, using an air mass 1.5 global (AM 1.5G) filter to obtain the current density–voltage (J–V) curve. Shown in Fig. 2(b), SEM cross-sectional images of the samples were taken to make sure that the active layer is indeed filled into the ZnO nanorods layer, and covers the well.

 

Fig. 2 (a) Schematic diagrams of the ITO-free inverted small molecules solar cell structures. (b) SEM cross-sectional image of device with ZnO nanorod length of 200 nm.

Download Full Size | PDF

3. Results and discussion

In order to optimize the performance of the ITO-free inverted SM-OPVs, devices with various lengths of ZnO nanorods (the growth times varies) were fabricated. Table 1 summarizes the device performance of ITO-free inverted SM-OPVs, including short circuit current J_sc, open circuit voltage V_oc, fill factor (FF), and the conversion efficiency η. Among the ITO-free inverted SM-OPVs, the device with 200 nm nanorods (sample C) exhibit the best performance. Generally for the devices, J_sc will be enhanced by utilizing the nanorods to collect the light. However, as the length of nanorods grows over 200 nm, the surface roughness undergoes a dramatic increase, as can be observed in Fig. 1. Therefore, the leakage current of the devices will increase, and the FF of the devices will decrease.

Table 1. Device Performance of Inverted SM-OPVs Various Lengths of ZnO Nanorods

View Table

Figure 3(a) shows the J–V curves of samples A, B, C and D, and the device parameters are listed in Table 1. The thickness of the each layer is determined by SEM cross-sectional image as shown in Fig. 2(b), and will be used to calculate the optical absorption of the devices. The series resistance (R_s) and sheet resistance (R_sh) can be calculated by the slope of the curve, at V = 1 and V = 0, respectively [18]. The increase in the leakage current will result mainly in decreased R_sh, with sample D showing the smallest R_sh, as can be seen in Fig. 3(a). As a result, the FF of sample D exhibits the lowest value, of 36%.

 

Fig. 3 (a) J-V characteristics of devices with various lengths of 0 nm (sample A), 50 nm (sample B), 200 nm (sample C) and 350 nm (sample D). (b) The EQE spectra of sample A, sample B, sample C and sample D.

Download Full Size | PDF

The external quantum efficiency (EQE) of samples A, B, C and D are also measured, shown in Fig. 3(b). From the EQE spectra, the J_sc of the devices can be calculated [10], 8.84, 10.13, 10.79, and 7.44 mA/cm², for samples A, B, C, and D respectively. The slight difference between the J_sc from the EQE and PCE results are due to the difference in the light source used in the measurement. Nonetheless, the J_sc from EQE measurement and PCE results still exhibit similar trends [10].

To further clarify the function of the ZnO nanorods, the simulated optical absorption spectrum of the devices of various nanorod lengths are calculated by rigorous coupled wave analysis (RCWA) method, as shown in Fig. 4. The refractive index and extinction coefficient of each layer are measured in advance using an ellipsometer. To calculate the optical transmission and reflection of the random ZnO nanorods, a SEM top view image of the nanorods is taken to identify the position of each ZnO nanorod, and is used to simulate the optical effect.

 

Fig. 4 Simulated absorption spectra of the devices with the various lengths of ZnO nanorods.

Download Full Size | PDF

In Fig. 4, the absorption of devices are enhanced as the length of ZnO nanorods increases. The absorption of the devices enhances as the length of the ZnO nanorods increases. The appearance of these light trapping features, at wavelengths of visible light from 400 to 800 nm, are due to the improved antireflection effect [19–21]. However, the absorption of device almost saturates as the length of ZnO nanorods grows to about 200 nm. Consequently, the length of 200 nm can be seen as a sort of threshold, and we define it as the optical saturation length (OSL) of the devices. Since no further optical enhancement are observed when the length of nanorods exceeds 200 nm, the devices with the length of 200 nm should exhibit the best performance, assuming that the electrical properties have little influences. In Table 1, the experimental results match well with the simulated results, as sample C (200 nm) also exhibit the highest J_sc, of 11.62 mA/cm².

By introducing the nanorods layer, both the J_sc and the FF are improved. The nanorods are able to collect the incident light for the active layer, and increase the carrier collection of electron transfer layer (ETL) at the same time. To understand the role the of ZnO nanorods in ETL, the mobility of AZO, with and without the nanorods, are measured by Hall measurements (Ecopia-HMS 5000) using the van der Pauw method at room temperature [22]. For all samples, the mobility of the ETL is increased by introducing the nanorods layer. For the case of 200 nm nanorods, the mobility of the ETL is improved from 3.025 to 4.637 cm²v⁻¹s⁻¹.

In the past, highly efficient polymer solar cells are easily achieved by tuning the thickness of the active layer. However unlike polymer solar cells, SM-OPVs generally have shorter diffusion length, which greatly limits the thickness of the active layer. As a result, utilizing the nano-structure or nanorods to improve carrier collection or enhance the light harvesting is necessary. Although the nanorods structure is useful for the light harvesting, the orientation of the nanorods layer is very important. The surface roughness of the nanorods layer have a great influence on the active layer, since the active layer of SM-OPVs is generally thinner than 100 nm [8,9]. To deposit well-oriented nanorods, the crystallization of the seed layer (AZO) is vital [23]. In this work, the AZO films are prepared by PLD technique under high temperature and the films exhibit high crystallization, implying that the transmission of the AZO film or the orientation of the following growth layer (nanorods layer) can be better, than AZO films prepared by sol-gel or sputtering method [23]. Moreover, by adjusting the laser power, pulse frequency, and doping concentration of the AZO target, the AZO thin film can maintain high crystallization and relatively low sheet resistance (~30 Ω/sq). Hence, ITO-free inverted SM-OPVs can be fabricated by replacing the ITO with AZO thin film. By introducing the well oriented ZnO nanorods interlayer, the PCE of the devices can be further improved from 4.03 to 6.6%.

4. Conclusion

In conclusion, ITO-free inverted SMPV1:PC₇₁BM solar cells with high quality AZO transparent electrodes are fabricated. The AZO thin film prepared by PLD not only exhibits high transmission (>85%), but also low sheet resistance (~30 Ω/sq). The PCE of devices based on AZO electrode can reach around 4%, and by introducing the ZnO nanorods interlayer, both light-harvesting capability and carrier collection ability of the devices can be further enhanced, and the PCE of the devices can be further improved. The experiment results also match well with the simulated results, and the PCE of ITO-free inverted small molecule solar cell with ZnO nanorods can reach 6.6%.