Yu-Chiang Chao*ab,
Fu-Min Zhana and
Husan-De Lia
aDepartment of Physics, Chung Yuan Christian University, Chung-Li 32023, Taiwan. E-mail: ycchao@cycu.edu.tw
bCenter for Nanotechnology, Chung Yuan Christian University, Chung-Li 32023, Taiwan
First published on 11th July 2014
In this study, a wet chemical method of etching indium-tin-oxide (ITO) nanorods from commercially available ITO-coated glass is proposed and its applicability to polymer solar cells (PSCs) is validated. The ITO nanorods can be fabricated within a few minutes using the proposed method. The ITO-nanorod-based PSCs show 67% and 46% improvement in short-circuit current density and power conversion efficiency, respectively, compared with devices based on flat ITO electrodes. The enhanced performance is related to the anti-reflection and light-scattering properties of ITO nanorods, which enhance incident light intensity and light-trapping properties.
Innovative PSC structures have attracted great attention because their light-trapping characteristic efficiently enhances light absorption.4 Nanoimprint lithography,5–7 laser interference lithography,8,9 and mechanical rubbing10 have been used to create nanostructures on hole-transporting and photoactive layers of PSCs, and notable improvements in PCE have been reported. In addition to the texturing of organic layers, incorporating inorganic nanostructures, such as zinc oxide and ITO nanorods, on electrodes is also of great benefit to PSCs because of the anti-reflective11,12 and light trapping13,14 characteristics of the nanostructures. Further, ITO nanorods lead to significant improvement in PCE.15–18 Because ITOs cannot be easily nanostructured by etching,19 ITO nanorods are conventionally fabricated using molecular beam epitaxy,11 oblique angle deposition,15–18,20,21 plasma etching,22 catalyst-mediated vapor–liquid–solid method,23 or radio frequency magnetron sputtering deposition.24 However, because these fabrication procedures are expensive, the fundamental understanding and development of ITO-nanorods-based PSCs and other optoelectronic devices were hampered. Moreover, the height of ITO nanorods cannot be precisely controlled in most of these methods. Therefore, a relatively easy ITO nanorods fabrication procedure that can be used in the production of different optoelectronic devices must be developed.
In this study, we show that an ITO film on a commercially available ITO-coated glass substrate can be etched into ITO nanorods through wet chemical etching within a few minutes. The ITO nanorods are selectively formed in regions exposed to phosphoric acid by covering the remaining area of the ITO electrodes with an Al protective layer. The transmittance and the light-scattering capability are shown to be enhanced. Because the ITO nanorods enhance incident light intensity and trapped light in the photoactive layer, the short-circuit current density (JSC) and the PCE of the ITO-nanorods-based PSCs were improved by 67% and 46%, respectively.
A Keithley 2400 source measure unit was used for device characterization, and the current density–voltage curves were plotted. Photocurrent measurements were performed under simulated air mass 1.5 global irradiation at an intensity of 100 mW cm−2 using a xenon lamp-based solar simulator (Oriel 96000, 150 W). A field emission scanning electron microscope (FESEM, JEOL JSM-7600F) was used for scanning electron microscope (SEM) measurements. Dark-field microscopy images were recorded using an optical microscope with an Epiplan 50×/0.7 HD objective (E990, Carl Zeiss). The transmittance and absorbance were measured with a microspectrometer (SD1200-LS-HA, StreamOptics Co.).
The optical properties of the ITO nanorods were investigated using transmittance spectroscopy and dark-field microscopy. The bare ITO film exhibits transmittance peaks at 480 nm and 710 nm owing to Fabry–Pérot resonance (Fig. 2a).25 The transmittance of the etched substrates varies slightly with etching time for wavelengths below 500 nm. However, a remarkable enhancement in transmittance can be observed in a wavelength range of 500 nm to 1000 nm. For wavelengths below 800 nm, maximum transmittance was achieved in 10 min. Therefore, the formation of ITO nanorods results in increase of transmittance, which is an essential characteristic of an anti-reflective coating.26 The effective medium theory considering a layer comprising two mediums with indices n1 and n2 and volume fractions f1 and f2 has been used to investigate the anti-reflective characteristics of nanostructures, and reduced reflection and enhanced transmission were reported.27 Moreover, the dark-field microscopy image of the ITO nanorods with 10 min etching time (Fig. 2d) is brighter than that of the bare ITO film (Fig. 2b). Because dark-field microscopy includes only scattered light, the brighter image of the ITO nanorods proves the light-scattering nature of the ITO nanorods. These optical characteristics of the ITO nanorods such as high transmittance and the light-scattering nature enhance the performance of PSCs. Because of the adequate nanorod morphology, the excellent light transmittance, and the light-scattering property, the etching time for ITO nanorods formation was selected to be 10 min during the fabrication of PSCs.
Previous studies attributed the improved performance of the ITO-nanorods-based PSCs to the reduced charge carrier transport distance, increased charge collection efficiency, and differences in transmission spectra.15–18 Because the ITO nanorods protrude into the photoactive layer, it is not easy to identify the major contribution to the enhanced PCE. Therefore, three device architectures were fabricated in this study for comparison (Fig. 3a). Devices B and C contain ITO nanorods and are compared with Device A having a planar ITO film (left panel of Fig. 3a). The gaps between the ITO nanorods in Device B are filled with PEDOT–PSS, which is a hole-transporting material, (middle panel of Fig. 3a) whereas the PEDOT–PSS coating is uniformly covered on the ITO nanorods in Device C (right panel of Fig. 3a). The PEDOT–PSS was blended with isopropanol and Triton X-100 for improving wetting on the ITO nanorods. The filling of the gaps between the ITO nanorods (Fig. 3c) and the coverage of the ITO nanorods surface (Fig. 3e) by the blend solution can be controlled by adjusting the spin-coating speed between 500 rpm and 6000 rpm. When the spin-coating rate is 2500 rpm, the PEDOT–PSS blend is partly filled in the gaps and partly on the surface of the ITO nanorods (Fig. 3d). Spin coating of such a blend solution on a planar ITO film also results in a good wetting property (Fig. 3b). No remarkable protrusions can be found on the flat PEDOT–PSS surfaces (Fig. 3b and c). The influence of the PEDOT–PSS coating on the optical properties of the ITO nanorods was then investigated. As the spin-coating speed of the PEDOT–PSS is reduced, the transmittance decreases (Fig. 3f); hence, the brightness of the dark-field microscopy image decreases (Fig. 2d–f). The dark-field microscopy image of the PEDOT–PSS coated ITO film is slightly brighter than that of the bare ITO film because the PEDOT–PSS coated ITO film exhibits a higher transmittance in the wavelength range of 540–700 nm (Fig. 2b and c).
A blend of P3HT and PCBM was used for depositing the photoactive layer. The blend solution was deposited not only on the substrates with flat PEDOT–PSS surfaces but also on the PEDOT–PSS coated (conformal) ITO nanorods (Fig. S5†). Good polymer infiltration is achieved, and no air gaps can be observed (Fig. 3g and S6†). Therefore, the structures of Devices A, B, and C shown in Fig. 3a can be realized. The thickness of the photoactive layer can be easily controlled by varying the spin-coating speed, and it can be determined by estimating the distance between the top of the PEDOT–PSS layer and the cathode (Devices A and B) or the distance between the crest of the PEDOT–PSS covered ITO nanorod and the cathode (Device C). It should be noted that the amount of photoactive material in Device C is underestimated because it is also present in the valleys between the PEDOT–PSS coated (conformal) ITO nanorods.
The characteristics of Devices A, B, and C with different photoactive layer thicknesses were investigated and compared (Fig. 4 and S7† and Table 1). The photovoltaic parameters of the devices are summarized in Table 1, and the two representative J–V curves of Devices A, B, and C with similar thicknesses (∼140 nm and ∼180 nm) are shown in Fig. 4a. The maximum averaged values of JSC and PCE of Device A with thickness of 220 nm are 6.34 mA cm−2 and 2.10%, respectively (Table 1 and Fig. S7a†). These results are consistent with the theoretical calculations of ref. 28 that reported an oscillating behavior with two maxima at 80 nm and 220 nm, considering optical reflection and interference within a multilayer device. The oscillatory nature of JSC and PCE is related to the oscillatory behavior of the number of photons absorbed in the active layer.28 For Device B, both JSC and PCE decrease as the thickness of the photoactive layer decreases from 180 nm to 130 nm (Table 1 and Fig. S7b†). Because Device B contains ITO nanorods with light-scattering characteristics, the model considering optical interference within a multilayer device may not be suitable for Device B. We suppose that, at a constant voltage, the lower JSC value and lower PCE of Device B with a thicker photoactive layer may be attributed to a smaller electric field, which results in an increase in recombination rate and a reduction in exciton dissociation rate. The fill factor (FF) also suffers from such drawbacks and is the lowest for the thickest Devices A and B (Table 1). Nevertheless, most of the FFs of Device B are also less than 50% whereas most of those of Device A are greater than 50%. Further, the performance of Device B is better than that of Device A when both the Devices have similar photoactive layer thickness (Fig. 4a). The average values of JSC and PCE of Device B with a 180 nm photoactive layer are 9.70 mA cm−2 and 2.61%, respectively, which is an improvement of 67% and 35%, respectively, when compared with Device A. The JSC and PCE of Device B with a 140 nm photoactive layer are also enhanced by 59% and 46%, respectively, when compared with Device A. The superior performance of Device B is attributed to the increased light absorption (Fig. 4b) resulting from the enhanced transmittance and light-scattering characteristics of the ITO nanorods. Moreover, the light-trapping effect inside the photoactive layer may be further strengthened because the light reflected from the cathode can cause scattered reflection from the nanorods.19 Regarding Device C, as the photoactive layer thickness decreases, the JSC and PCE first decrease and then increase (Table 1 and Fig. S7c†). Similar to those of Device B, most of the FFs of Device C are less than 50% except for the one corresponding to the thinnest photoactive layer. The JSC and PCE of Device C with a 150 nm photoactive layer are enhanced by 44% and 17%, respectively, when compared with Device A (Fig. 4a and Table 1). These superior characteristics can also be attributed to the increased transmittance and the light-scattering characteristics of the ITO nanorods. However, the performance of Device C is inferior to that of Device B, even through Device C appears to gain more benefits, such as better charge collection and reduced charge transport distance, from its device structure.15,16,18 The poorer performance of Device C is attributed to the lower open-circuit voltage (VOC) and shunt resistance (RSH) than those of Device B. The VOC and RSH of Device C are approximately 0.57 V and 104 to 106 ohm cm2, respectively, whereas those of Device B (and Device A) are approximately 0.62 V and ∼106 ohm cm2, respectively. In addition, RSH is determined from the J–V characteristics measured under dark conditions.29 The smaller value of VOC can be attributed to the incomplete coverage of PEDOT–PSS on the surface of the ITO nanorods. In the absence of the PEDOT–PSS coating between the anode and the photoactive layer, a non-ohmic contact that may decrease VOC is formed.30 The smaller value of RSH is due to the electrode proximity effects15,18 and the nonuniform current distribution around the nanorods.15 In summary, for devices with similar thicknesses, though the FFs of Devices B and C are lower than that of Device A and the VOC of Device C is lower than those of Devices A and B, the PCEs of Device B and Cs can be easily improved by a larger value of JSC thus resulting in higher PCEs than that of Device A. Because the performance of Device B is superior to that of Device C, the enhanced incident light intensity and the light-trapping characteristic are believed to be the most important advantages of ITO nanorods.
Device type | Thickness [nm] | JSC [mA cm−2] | VOC [V] | FF [%] | PCE [%] |
---|---|---|---|---|---|
A | 260 | 5.82 (5.97) | 0.609 (0.611) | 51 (54) | 1.83 (1.94) |
220 | 6.34 (6.64) | 0.616 (0.620) | 54 (56) | 2.10 (2.16) | |
210 | 5.79 (5.93) | 0.613 (0.615) | 57 (57) | 2.01 (2.07) | |
190 | 5.80 (5.91) | 0.611 (0.614) | 54 (56) | 1.93 (2.04) | |
140 | 4.20 (4.61) | 0.601 (0.607) | 54 (56) | 1.35 (1.49) | |
B | 180 | 9.70 (10.19) | 0.617 (0.625) | 44 (46) | 2.61 (2.88) |
170 | 8.68 (8.81) | 0.621 (0.626) | 49 (52) | 2.62 (2.80) | |
160 | 8.17 (8.53) | 0.618 (0.627) | 45 (48) | 2.25 (2.45) | |
140 | 6.69 (6.91) | 0.615 (0.618) | 48 (49) | 1.97 (2.07) | |
130 | 6.11 (6.43) | 0.626 (0.636) | 53 (56) | 2.05 (2.30) | |
C | 160 | 8.24 (8.56) | 0.573 (0.581) | 47 (49) | 2.20 (2.24) |
150 | 6.05 (6.30) | 0.540 (0.548) | 49 (53) | 1.59 (1.61) | |
120 | 6.41 (6.77) | 0.572 (0.581) | 44 (50) | 1.61 (1.94) | |
110 | 6.31 (6.61) | 0.565 (0.573) | 46 (47) | 1.64 (1.74) | |
100 | 6.34 (6.51) | 0.581 (0.595) | 53 (55) | 1.93 (2.06) |
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra03119a |
This journal is © The Royal Society of Chemistry 2014 |