Si Yun
Khoo
ab,
Hongbin
Yang
a,
Ziming
He
a,
Jianwei
Miao
a,
Kam Chew
Leong
b,
Chang Ming
Li
*c and
Timothy Thatt
Yang Tan
*a
aSchool of Chemical and Biomedical Engineering, Nanyang Technological University, 62 Nanyang Drive, Singapore 637459. E-mail: tytan@ntu.edu.sg; Fax: +65 6794 7553; Tel: +65 6316 8829
bGlobalfoundries Singapore Pte. Ltd, 60 Woodlands Industrial Park D, Street 2, Singapore 738406, Singapore
cInstitute for Clean Energy and Advanced Materials, Southwest University, Chongqing 400715, China. E-mail: ecmli@swu.edu.cn
First published on 1st July 2013
The utilization of metallic nanoparticles is one of the key strategies to improve the performance of photovoltaic devices. In this work, we elucidate the power conversion efficiency (PCE) enhancement mechanism by gold nanoparticles (Au-NPs) through a bilayer anodic buffer structure in polymer solar cells. The results show that the PCE of the device based on a Au-NP:poly(sodium-4-styrenesulfonate)/V2O5 bilayer buffer exhibits a ∼16% enhancement compared with the device without Au-NP. By controlling the density of Au-NPs to minimize plasmonic effects, the Au-NP induced enhancement of charge extraction and crystallinity of the photoactive layer were demonstrated for the first time. Our work indicates that the plasmonic effect may not be the only factor that enhances the PCE of polymer solar cells, while providing new insights into the roles of Au-NPs in performance improvement of a bulk-heterojunction polymer solar cell.
There is no shortage of good studies on the understanding of LSPR in polymer solar cells.20,21 Wu et al. performed an in-depth investigation on the LSPR contribution from Au-NPs towards P3HT:PC61BM based polymer solar cells. The work by Kim et al. demonstrated significant enhancement in external quantum efficiency for polymer solar cells using plasmonic noble metal nanoparticles.17 However, the LSPR effect from Au-NPs might not be the only factor that increases the PCE of polymer solar cells. Heeger et al. exemplified an increase in PCE through light scattering activities caused by incorporated Au-NPs.22
Here, we design a Au-NP/poly(sodium-4-styrenesulfonate)/V2O5 (Au-NP:PSS/V2O5) bilayer anodic buffer to enhance the PCE of polymer solar cells. Our results show that a ∼16% improvement in the PCE was achieved through the utilization of a Au-NP:PSS/V2O5 bilayer anodic buffer. The enhancement mechanisms induced by the Au-NP layer were investigated by analysis of the electrical characteristics of the devices, optical properties and the crystalline structure of the polymer active layers. The current results indicate that the Au-NP:PSS/V2O5 bilayer significantly enhances the device efficiency by improving the crystallinity of the active layer and enhancing the charge extraction of the polymer solar cell.
Fig. 1 (a) The device architecture of the conventional polymer solar cell used in this work, i.e. ITO/Au-NP:PSS/V2O5/P3HT:PC61BM/Al, and (b) the band alignment of the components used in the device. |
The dried ITO glasses were further cleaned with UV ozone plasma for 20 min to remove any residual organic molecules. A solution of 0.2 wt% poly(sodium-4-styrenesulfonate) (PSS) mixed with Au-NPs (denoted as Au-NP:PSS) was first coated onto the ITO/glass substrate at 5000 rpm for 60 s (∼16 nm thick) and heated on a 150 °C hot plate for 30 min. Afterwards, vanadium oxytriisopropoxide was diluted with iso-propanolsolvent at 1:150 volume ratio and spin coated on the patterned ITO/glass substrate at 8000 rpm for 30 s (∼10 nm thick). It is speculated that the possibility of intermixing of Au-NP:PSS with the V2O5 layer would be minimal because the Au-NP:PSS layer was heated on the hotplate first to ensure that the layer was dried before the V2O5 layer was spin coated onto it. The as-prepared sample was maintained at ambient conditions for 1 h for hydrolysis to take place. Then, a 10 mg ml−1 P3HT:PC61BM (both purchased from Sigma-Aldrich) blend (1:1 weight ratio) in chlorobenzene was sonicated for 5 min before spin coating at 800 rpm for 25 s on the as-prepared ITO substrate and capped in a chlorobenzene-rich petri dish for 20 min. The photoactive device structures were completed by thermally evaporating 100 nm Al electrodes in a base pressure of 1.0 × 10−5 Pa with a shadow mask to define the active device area. The fabricated solar cells were annealed at 150 °C for 30 min. For the control devices, a polymer solar cell without the Au-NP:PSS layer was fabricated using the conditions described previously. All the processes were carried out under ambient conditions at room temperature except the aluminum electrode deposit process and the annealing process.
Fig. 2 AFM topography images of (a) V2O5, (b) Au-NP/V2O5, (c) Au-NP:PSS/V2O5 on the ITO substrate and (d) SEM image of Au-NP:PSS coated ITO/glass substrate. |
The J–V characteristics of the illuminated polymer solar cells with different anodic buffer layers are shown in Fig. 3(a) and the solar cell performance parameters are summarized in Table 1. As displayed in Table 1, the reference device prepared with the V2O5buffer layer (V2O5) exhibits a short-circuit current density (Jsc) of 7.02 mA cm−2, an open-circuit voltage (Voc) of 0.596 V, a fill factor (FF) of 50.5%, and a calculated power conversion efficiency (PCE) of 2.11%. Compared with the reference device, the insertion of Au-NP (Au-NP/V2O5) from as-synthesized Au-NP solution does not enhance the performance of the solar cell effectively. As is evident from the lower Voc and Jsc values, the decrease in device performance could be attributed to a large amount of charge recombination induced by aggregated Au-NPs at interface between the ITO and V2O5 layer. On the other hand, PCE enhancement by ∼16% is observed for the polymer solar cell with Au-NP:PSS and a V2O5 bilayer structure (Au-NP:PSS/V2O5). The PCE improvement mainly originates from the enhancement in the fill factor, which increases from 50.5% to 58.7% for the V2O5buffer and Au-NP:PSS/V2O5 bilayer buffer, respectively. The enhancement in the fill factor indicates an improvement of the charge transport properties induced by the incorporated Au-NP:PSS layer. A similar Jsc value indicates that Au-NPs do not have a pronounced effect on the optical absorption in the photoactive layer, which is confirmed by the IPCE spectra and UV-visible absorption spectra of various buffer layers (Fig. 3(b)).
Fig. 3 (a) Current density–voltage (J–V) characteristics under AM1.5G illumination for photovoltaic measurements and (b) incident photon-to-current efficiency (IPCE) of polymer solar devices with (black) V2O5, (red) Au-NP/V2O5 and (blue) Au-NP:PSS/V2O5buffer layer. Inset: UV-visible absorption spectrum of (black) V2O5, (red) Au-NP/V2O5, and (blue) Au-NP:PSS/V2O5buffer layer. |
Buffer layer | V oc (V) | J sc (mA cm−2) | PCE (%) | FF (%) |
---|---|---|---|---|
a Au-NP/V2O5 indicates that the as-synthesized Au-NP solution was spin-coated onto the ITO substrate before V2O5 layer deposition. b Au-NP:PSS indicates that 0.2 wt% PSS-mixed Au-NP solution was spin-coated onto the ITO substrate before V2O5 layer deposition. | ||||
V2O5 | 0.596 | 7.02 | 2.11 | 50.5 |
Au-NP/V2O5a | 0.562 | 6.08 | 1.57 | 46.1 |
Au-NP:PSSb/V2O5 | 0.596 | 7.01 | 2.45 | 58.7 |
IPCE is commonly used to demonstrate the presence of the plasmonic effect from noble metallic nanoparticles, which could cause significant absorption enhancement at wavelengths ranging from around 450 nm to 600 nm,21 corresponding to the LSPR range of metallic nanoparticles. The IPCE curves of devices with the V2O5buffer and Au-NP:PSS/V2O5 bilayer buffer in Fig. 3(b) show similar shapes and intensity throughout the visible light wavelengths. This result is in good agreement with the similar Jsc values of the two devices, as shown in Table 1. As such, the IPCE results demonstrate that the plasmonic effect is probably absent in the device with the Au-NP layer. In addition, the UV-visible absorption spectra of various buffer layers were examined to investigate the plasmonic contribution from Au-NP layer. As shown in the inset of Fig. 3(b), incorporating Au-NP underneath the V2O5buffer layer has no discernible effect on the absorption spectrum of the buffer layer at wavelengths ranging from 450 nm to 600 nm. Based on the current IPCE and UV-visible absorbance results, we conclude that the insertion of a low density Au-NP:PSS layer does not induce optical absorption enhancement.
It should be noted that in this work, PSS is mixed with Au-NPs to create a more uniform distribution of Au-NPs on the ITO substrate. In order to investigate the effects of PSS on the device performance, a PSS-only device was fabricated in which only PSS was deposited at the interface of ITO substrate and V2O5buffer layer. The concentration of PSS used for spin coating on the ITO substrate was 2 mg ml−1, which is the same as the amount of PSS mixed with the as-prepared Au-NP solution. As shown in Fig. S3,† the PSS-only device shows similar J–V characteristica and PCE as the reference devices. We thus conclude that PSS itself does not exert a discernible effect on the device performance.
To gain a better understanding of the performance enhancement by the addition of the Au-NP:PSS layer, the internal resistance of the polymer solar cell was extracted using an equivalent circuit model (inset of Fig. 4(a)) by fitting the experimental results under dark conditions (Fig. 4(a)) using a modified vertical optimization method.28–30 From the equivalent circuit, the current density (J) versus applied voltage (V) can be described using the generalized Shockley equation.31
(1) |
Fig. 4 (a) Dark current response–voltage (J–V) characteristics of polymer solar devices with (black) V2O5, (red) Au-NP/V2O5 and (blue) Au-NP:PSS/V2O5buffer layers and (b) current–voltage characteristics of hole-only devices with (black) V2O5, (red) Au-NP/V2O5 and (blue) Au-NP:PSS/V2O5buffer layers under dark conditions. The inset in (a) shows the equivalent circuit model used to obtain the values of the circuit elements. |
The extracted parameters summarized in Table 2 show that the reference cell (P3HT:PC61BM polymer solar cell with V2O5buffer layer) has low Rp (700 Ω cm2), high Rs (1.2 Ω cm2) and high Js (0.8 μA cm−2). On the other hand, the polymer solar cell with the Au-NP:PSS/V2O5 bilayer buffer has Rp increased by a factor of 3 (from 700 Ω cm2 to 2500 Ω cm2), and Rs reduced by a factor of 2 (from 1.2 Ω cm2 to 0.6 Ω cm2) compared with the reference cell. The improvement in the internal resistance is in agreement with the higher fill factor, which could be ascribed to the reduction of the charge transport resistance (evident from the reduction in Rs) resulting from better electrical conductivity due to the more uniformly distributed Au-NP/V2O5 bilayer buffer. In addition, higher recombination resistance (evident from the increase in Rp) minimizes the occurrence of charge recombination and thus leads to higher PCE.
Buffer layer | J s (μA cm−2) | R s (Ω cm2) | R p (Ω cm2) | n |
---|---|---|---|---|
a Au-NP/V2O5 indicates that the as-synthesized Au-NP solution was spin-coated onto the ITO substrate before V2O5 layer deposition. b Au-NP:PSS indicates that 0.2 wt% PSS-mixed Au-NP solution was spin-coated onto the ITO substrate before V2O5 layer deposition. | ||||
V2O5 | 0.8 | 1.2 | 700 | 1.8 |
Au-NP/V2O5a | 0.62 | 1.3 | 200 | 1.8 |
Au-NP:PSSb/V2O5 | 0.15 | 0.60 | 2500 | 1.9 |
To further evaluate the charge transport properties across the anodic buffer layer/P3HT interface, hole-only devices with different anodic buffer layers, namely ITO/V2O5/P3HT/Au, ITO/Au-NP/V2O5/P3HT/Au and ITO/Au-NP:PSS/V2O5/P3HT/Au, were fabricated. The J–V characteristics of the hole-only devices for various anodic buffer layers are shown in Fig. 4(b). A curved J–V line is observed for the hole-only device with the Au-NP/V2O5 bilayer buffer indicating that the performance of the polymer solar cell is limited by the space-charged limited current, which is in agreement with the lower fill factor, Voc and PCE. Linear and symmetric J–V relationships were observed for the P3HT hole-only devices with V2O5 and Au-NP:PSS/V2O5 bilayer anodic buffer which implies that the V2O5 and Au-NP:PSS/V2O5buffer layers provide true ohmic contact for the polymer solar cell at the anodic buffer and polymer active interface,32 in which the ohmic contact could account for the higher Voc of the devices. Moreover, the current density of the hole-only device with the Au-NP:PSS/V2O5 bilayer buffer is two times higher than that of the reference cell, which implies a smaller series resistance of the Au-NP:PSS/V2O5 device. These results illustrate that the insertion of the Au-NP:PSS layer effectively enhances the electrical conductivity of the polymer solar cell.
The effects of Au-NPs on the photoactive layer were further investigated by UV-visiblespectrometry, XRD and tapping-mode AFM. Fig. 5(a) shows the absorption spectra of the P3HT:PC61BM active layer on V2O5 with or without the insertion of Au-NPs, while Fig. 5(b) displays the change of the UV-visible absorbance spectrum of the active layer after the insertion of Au-NPs, in which the change is calculated by subtracting the absorption spectrum of the photoactive layer after Au-NP insertion from the absorption spectrum of photoactive layer with the V2O5buffer layer only. The possibility of an increase in the UV-visible absorption spectrum induced by plasmonic contributions from the incorporated Au-NPs is eliminated, as the curve of the UV-visible absorbance spectrum of the as-synthesized Au-NP solution (inset of Fig. 5(b)) is different from the change in the UV-visible absorption spectrum after the insertion of Au-NPs (as shown in Fig. 5(b)). The UV-visible absorption spectra of the active layer with Au-NP/V2O5 and Au-NP:PSS/V2O5 bilayer anodic buffer are overlapping, in which vibronic structure with peaks at 530 nm and 610 nm, and a gradually increasing and broadening peak located at around 500 nm is observed in the absorption spectra of the active layer with Au-NP/V2O5 and Au-NP:PSS/V2O5 bilayer anodic buffer, as shown in Fig. 5(a). The increase in the intensity of the absorption peak at 530 nm indicates the enhancement in intrinsic π–π* transitions of P3HT and the peak located at 610 nm corresponds to the enhancement of the 0–0 vibronic transition of planarized P3HT, which indicates strong interchain-interlayer interactions among the P3HT chains and good polymer ordering in the blend films.33–38 The XRD patterns of various P3HT:PC61BM active layers further confirms the results from the UV-visible measurements. The peak at 2θ ≈ 5.6° shown in Fig. 5(c) corresponds to the interchain spacing in P3HT associated with interdigitated alkyl chains.38–40 In comparison with that of the polymer solar cell with the V2O5buffer layer only, a higher intensity is observed for the peaks at 2θ ≈ 5.6° for the P3HT:PC61BM active layer on the Au-NP:PSS/V2O5buffer, which may imply a higher degree of P3HT crystallinity in the P3HT:PC61BM active layer.
Fig. 5 (a) UV-visible absorption spectra of P3HT:PC61BM polymer solar cells with (black) V2O5, (red) Au-NP/V2O5, and (blue) Au-NP:PSS/V2O5 respectively, (b) absorption change (ΔAbs) of polymer solar cells after incorporating Au-NPs (black line) and Au-NP:PSS (red line) in the devices' anodic buffer layer and (c) XRD spectra of polymer solar cell with (black) V2O5, (red) Au-NP/V2O5, and (blue) Au-NP:PSS/V2O5. (The peak at 2θ = 22° is from ITO). |
The effect of Au-NPs on the morphology of the photoactive layer was further examined with AFM. The morphology was obtained from complete devices by detaching the Al electrode using scotch tape.6,38,41 As shown in Fig. 6, the surface roughness of the photoactive layer increased dramatically and the root mean square roughness increased from 0.580 nm to 4.124 nm for devices with the V2O5buffer and Au-NP:PSS/V2O5 bilayer buffer, respectively. The increase in the photoactive layer surface roughness could be considered a signature of polymer self-organization resulting from enhanced ordered structure formation in the thin film, which has been independently reported in various articles.6,38,42 This observation suggests that the Au-NP:PSS layer induced higher P3HT crystallinity and enhanced the contact interaction with the Al electrode,38 which led to the reduction of the device's charge transfer resistance and thus resulted in the increase of the fill factor. Besides the height profile, the phase diagrams of the active layers from the V2O5 and Au-NP:PSS/V2O5 bilayer buffer also show significant differences. As shown in Fig. 6(a), the phase diagram of the photoactive layer with the V2O5buffer shows an homogenous phase with a lack of phase contrast, which indicates more amorphous characteristics of the as-cast P3HT and PC61BM.43 On the contrary, larger phase contrast (as displayed in Fig. 6(b)) is observed for the active layer with the Au-NP:PSS/V2O5 bilayer. This indicates the formation of good crystallinity and an interfusion network within the bulk heterojunction region, which are beneficial for exciton dissociation. We suggest that the protrusion point, which is caused by the insertion of the Au-NP:PSS layer, induced the alignment of P3HT within the phase-separated networks, hence resulting in the improvement of the charge transport to the electrodes.12,29,44
Fig. 6 Tapping-mode AFM images of the P3HT:PC61BM active layer (without aluminium electrode): (left) 3D height image and (right) 2D phase diagram of the active layer with (a) V2O5 and (b) Au-NP:PSS/V2O5 as the anodic buffer layer. |
Footnote |
† Electronic supplementary information (ESI) available: AFM topography of blank ITO substrate, supplementary figures of J–V curves for the PSS/V2O5 bilayer structure devices and table summarizing the device performance with the PSS/V2O5 bilayer structure. See DOI: 10.1039/c3tc30956h |
This journal is © The Royal Society of Chemistry 2013 |