Highly efficient dual-plasmon polymer solar cell incorporating Au@SiO₂ core–shell nanorods and Ag nanoparticles

Abstract

Generally, metallic nanoparticles (NPs) are widely employed in polymer solar cells (PSCs) to enhance power conversion efficiency (PCE) via the localized surface plasmon resonance (LSPR) effect. Herein, a significant performance enhancement is demonstrated in inverted PSCs by incorporating two kinds of metallic NPs on the rear side of photoactive layer. The Au nanorods (NRs) capped with ∼2 nm ultrathin SiO₂ shells, with plasmonic absorption peaks of 520 nm for the transverse axis and 615 nm for the longitudinal axis (in deionized water), and 1 nm-ultrathin Ag NPs with an absorption peak of 517 nm, are in sequence spin-coated on a photoactive layer and thermally evaporated onto MoO₃. To the best of our knowledge, this is the first time that Au NRs are directly spin-coated on a photoactive layer. The dual-plasmon device exhibits an improved short-circuit current density (J_SC) of 10.36 mA cm⁻² and an increased fill factor (FF) of 0.64, generating a PCE of 4.11% with a ∼30% enhancement factor compared with that of the reference device. The enhancement is attributed to the light absorption improvement in the photoactive layer over a wide wavelength range of 500–650 nm, induced by the LSPR of Au@SiO₂ NRs and Ag NPs. In addition to the excitation of the LSPR effect, the incorporation of Ag NPs into MoO₃ also contributes to the increase in FF value through enhancing the electric conductivity of MoO₃.

---

1. Introduction

Polymer solar cells (PSCs) are currently gaining a great deal of attention due to their flexibility, low-cost, and simple manufacturing process. Recent developments of photoactive materials and device structures have significantly enhanced power conversion efficiencies (PCEs).^1–4 It is well known that since the carrier mobility is low in most photoactive materials (on the order of 10⁻⁴ cm² V⁻¹ s⁻¹), the photoactive layer must be rather thin (100 nm or less) to achieve efficient carrier extraction. However, such thin photoactive layers lead to a significant loss of incident sunlight with a final low light absorption efficiency and PCE. One of the most promising methods to enhance light harvesting is incorporating metallic nanoparticles (NPs) into devices. Metallic NPs, such as Au and Ag, have special characteristics totally different from their bulk metal, which is well known for the oscillations of the conduction electrons at surfaces of the metallic NPs. When these electrons have a similar frequency with the incident light, the electromagnetic field near the metallic NPs surface can be greatly amplified due to the localized surface plasmon resonances (LSPR) induced by collective oscillations of these electrons,⁵ which is beneficial for improving light harvesting in the photoactive layer and hence increasing the short-circuit current density (J_SC) in solar cells.

There widely reported incorporation of metallic NPs within hole transport layer,^6–10 such as poly(3,4-ethylenedioxythiophene):poly(4-styrenesulfonate) (PEDOT:PSS), or on the top of indium tin oxide (ITO)¹¹ in some papers. Lu et al.¹² reported an about 20% PCE enhancement by blending Au and Ag nanospheres into PEDOT:PSS. Hsiao et al.¹³ synthesized Au NPs with various sizes and shapes, and by incorporating them into PEDOT:PSS they acquired a J_SC of 11.49 mA cm⁻² in poly(3-hexylthiophene):[6,6]-phenyl-C₆₁-butyric acid methyl ester (P3HT:PCBM)-based cell. However, Fung's research work¹⁴ revealed that incorporation of small size Au NPs into PEDOT:PSS is not an ideal approach to enhance device performance because the LSPR-excited near-field around Au NPs mainly distributes laterally along the PEDOT:PSS layer rather than vertically into the photoactive layer, thereby resulting in no significant enhancement of light absorption in the photoactive layer. In addition, there were illustrations that dense NPs in charge transport layer on front side of electrode could lead to the reduction of J_SC.^15,16 Before cast into photoactive layer, a part of sunlight will be lost due to the backward scattering/reflection of metallic NPs, that is to say, sunlight intensity decreases before reaching the photoactive material, resulting in no obvious increase in J_SC.

Some researchers considered the direct mixing of metallic NPs coated by SiO₂ into photoactive solution^17,18 to fully exploit LSPR, thus improve light harvest ability of photoactive layer. Unfortunately, this method may bring an extra problem, that is, metallic NPs can not be well dispersed in PSC-compatible solvent, such as dichlorobenzene. Recently, researchers have reported that metallic nanostructures were incorporated on rear side of electrode of solar cell,^19–21 in which the sunlight would first pass through the photoactive layer before reaching metallic NPs, thus not only the absorption of metallic NPs itself but also the backward scattering of these NPs are beneficial to absorption enhancement in photoactive layer. Xu²² fabricated a series of devices with various thicknesses of thermal-evaporated Ag by inserting them into MoO₃, and got an over 20% increase in PCE with an optimal 1 nm thin Ag. Yao²³ obtained significant increases in J_SC and PCE (over 40%) upon inserting thermal-deposited Au NPs into WO₃ in P3HT:indene-C₆₀ bisadduct (ICBA)-based inverted PSCs.

Besides the location of metallic NPs, their size is also one of the most important factors for enhancing PSCs performance, because different NPs sizes will lead to variant LSPR absorption spectral ranges. Li²⁴ et al. compared the performances of Au-doped devices with embedding Au nanospheres (NSs) with diameters of 20 and 50 nm into TiO₂, and the device with 50 nm Au NSs exhibited a higher J_SC than that with 20 nm diameter. As one of widely used NPs in PSCs, nanorods (NRs) doping has superior effects compared with NSs, due to their wide LSPR spectra excited by both transverse and longitudinal axes. Moreover, a wide band of plasmon wavelengths along longitudinal axis, ranging from ∼600 to ∼900 nm, can be realized by varying aspect ratio of NRs. Janković¹⁷ et al. incorporated (Au–SiO₂ core–shell structures NRs Au@SiO₂ NRs) into photoactive layer in 2013. Their experimental results proved that the greatest PCE enhancement was observed in wavelength regions where PSC polymer absorbs poorly, instead of the spectral regions of efficient absorption. The similar research result was reported by Hsiao¹³ et al. In order to broaden wavelength range of absorption enhancement, some researchers explored the dual-plasmon structure.^24–27 Li et al.²⁴ fabricated a dual-plasmon device with Au NSs and a Ag nanograting, which were embedded into the active layer and acted as a back electrode, respectively. It is noted that in their research, the manufacturing process of Ag nanograting is relatively complex despite application of the two kinds of metallic NSs enables a broadband absorption. In this paper, we synthesized Au@SiO₂ NRs and spin-coated them onto the P3HT:PCBM layer. The direct spin-coating of Au NRs on photoactive layer, instead of incorporating into hole/electron transport layer, can fully utilize the local field excited by Au NRs. Compared with previous reported plasmon RSCs with metallic NPs with core–shell structures, the SiO₂ shell used in our experiment is as thin as 2 nm, which is enough to prevent from exciton quenching on Au surface. The average aspect ratio of NRs used in our work is 2.1, corresponding to 615 nm of plasmon peak from longitudinal axis, which is in the poor absorption band of P3HT. Employing above Au@SiO₂ NRs together with thermal-deposited Ag NPs (into a MoO₃ hole extraction layer), we fabricated inverted PSCs with dual plasmon effects. In our work, the integration of Au NRs and Ag NPs can effectively broaden LSPR wavelength to nearly cover the whole P3HT absorption spectrum. Optimization on distribution density of Au NRs and Ag NPs realized an enhancement factor of as high as 29.3% in PCE. We also explored how these NPs modulate the internal electric field and influence the carrier transport ability.

2. Experimental

The photoactive layer solution was prepared as follows: the mixed solution consisting of P3HT and PCBM (15 : 15 mg mL⁻¹) in 1,2-dichlorobenzene was stirred at 70 °C overnight in a N₂-filled glove box. The ZnO precursor was obtained by dissolving 65.6 mg zinc acetate dihydrate and 20 μL monoethanolamine in 3 mL ethanol solution, followed by at least 12 hours of stirring at 60 °C.

Au NRs was prepared as literature reported.²⁸ The process of wrapping Au NRs is as follows: NaOH solution (0.1 mM) was added to 5 mL Au NRs solution by drops to adjust its PH value to 10–11. Tetraethylorthosilicate (TEOS) solution (20 vol% in ethanol) was subsequently added to Au NRs solution under gentle stirring with 400 rpm. About 2 nm SiO₂ shell was coated on Au NRs after continuous stirring at overnight.

Fig. 1a shows the absorption spectra of Au NRs in deionized water without and with SiO₂ shells. After capped with 2 nm SiO₂ shells, the NRs show a same LSPR peak for the transverse axis as the naked Au NRs (at 520 nm), while their longitudinal axis resonance peak slightly red shifts from 608 nm to 615 nm. In order to confirm an even coating of SiO₂ on most of Au NRs in our experiment, we measured the transmission electron microscopy (TEM) image of Au@SiO₂ NRs (Fig. 1b). From the statistical results of SiO₂ thicknesses on Au NRs (Fig. 1c), we note that almost all Au NRs are coated with even SiO₂ shells, and most of the shell thicknesses are 2.0 ± 0.1 nm, which can sufficiently prevent the direct contact of the Au NRs from the P3HT:PCBM layer when spin-coating them on the photoactive layer. Such an ultrathin thickness is also beneficial for obtaining a significantly strengthened local-field and effectively extending the enhanced field into the photoactive layer. Before application these NPs into our solar cells, the deionized water consisting of Au@SiO₂ NRs was centrifuged at 8000 rpm for 10 min to remove solvent, and then the Au@SiO₂ NRs were dispersed in ethanol solution with different volume ratios. Note that prior to spin-coating, the ethanol solution containing of Au@SiO₂ NRs should be sonicated for 15 min to ensure a good dispersion.

Fig. 1 (a) The absorption spectra of Au NRs without and with SiO₂ shells dispersed in deionized water. (b) TEM image of Au@SiO₂ NRs with a bar of 100 nm. (c) Statistics on the thickness of SiO₂ shell coating on Au NR.

Devices were fabricated on patterned ITO-coated glass substrates. Before depositing the P3HT:PCBM photoactive layer, the ZnO precursor was spin-coated onto the clean ITO-coated substrate at 1200 rpm for 60 s, followed by a subsequent pyrolysis on a hot plate in air at 150 °C for 30 min to form a ∼28 nm ZnO film. The ZnO-covered substrates were transferred into the glove box to spin-coat a blend layer of P3HT:PCBM with a rotation speed and time of 800 rpm and 30 s.

After a 2 h solvent annealing in glove box at room temperature, a photoactive layer with a thickness of ∼70 nm was formed. Then the ethanol solution containing Au@SiO₂ NRs was spin-coated on the photoactive layer. At last, the devices were transferred into a vacuum chamber to complete thermal evaporation of a 8 nm MoO₃ layer and a 80 nm Ag under 2.0 × 10⁻³ Pa. For the dual-plasmon device, the hole extraction layer and the anode were formed by sequently evaporating the first 4 nm MoO₃ layer, 1 nm Ag, the second 4 nm MoO₃ layer and 80 nm Ag. The dual-plasmon device structure is shown in Fig. 2.

Fig. 2 The structure of dual-plasmon device with Au@SiO₂ NRs and Ag NPs.

The reference device without any metallic NPs was also manufactured for comparison.

3. Results and discussion

3.1. Plasmon devices with Au@SiO₂ NRs on the rear side of P3HT:PCBM

Fig. 3 shows the current density–voltage (J–V) characteristics of the reference device and the plasmon ones with various distribution densities of Au@SiO₂ NRs on the rear side of P3HT:PCBM. Through adjusting concentration of Au@SiO₂ NRs in ethanol solution, the plasmon devices show different increases in J_SC values. The J_SC increases at first with the Au@SiO₂ NRs concentration, and the highest J_SC is achieved with an optimal Au@SiO₂ NRs concentration of 14%. When further increasing the Au@SiO₂ NRs concentration to 20%, J_SC slightly decreases. From the detailed photovoltaic parameters listed in Table 1, we find that the optimized plasmon device exhibits excellent performances with a J_SC of 9.86 mA cm⁻², an open-circuit voltage (V_OC) of 0.62 V, and a fill factor (FF) of 0.61, generating a PCE of 3.75% with an enhancement factor of 17.9%.

Fig. 3 J–V characteristics of the reference device and the plasmon devices with various concentration of Au@SiO₂ NRs.

Table 1 Photovoltaic characteristics for the plasmon devices with various concentrations of Au@SiO₂ NRs

Au@SiO2 NRs concentration	JSC (mA cm^−2)	VOC (V)	FF	PCE (%)	RS (Ω cm^2)
0%	8.28	0.62	0.62	3.18	9.2
7%	8.67	0.62	0.62	3.35	8.8
10%	9.20	0.62	0.62	3.51	9.0
14%	9.86	0.62	0.61	3.75	8.9
20%	9.49	0.61	0.56	3.24	13.1

---

We attribute the apparent improvement of J_SC to absorption enhancement in P3HT:PCBM due to a strong local field excited by the Au@SiO₂ NRs. This is proved to be true by calculating the electric field distribution within the devices using FDTD software. In our simulations, x and z axes are along the longitudinal axis of Au@SiO₂ NRs and the incident direction of sunlight, respectively. In addition, y axis is along the transverse axis of Au@SiO₂ NRs and perpendicular to the incident direction of sunlight (the coordinates of the center of Au@SiO₂ NR are x = 0, y = 0, and z = 0). The periods of Au@SiO₂ NRs on P3HT:PCBM layer along x and y axes are both 664 nm. The simulation results shown in Fig. 4a–c reveal obvious increases in electric field intensities in the plasmon device with Au@SiO₂ NRs. We chose several typical wavelengths, e.g., 518 nm, 553 nm, and 649 nm, to observe changes of local field. At 553 nm, we find a ∼120-fold enhancement from a transverse axis resonance. We simultaneously extracted the electric fieldintensity along z axis (x = 0, y = 0) to directly observe the influence of Au@SiO₂ NRs on the internal electric field. From Fig. 4d, we observe a significant enhancement on field intensity in the vicinity of Au@SiO₂ NRs. We amplified the Fig. 4d with z axis ranging from −81 to −11 nm (the inset of Fig. 4d), which exactly corresponds to the P3HT:PCBM layer. As shown in the inset, the enhancement of electric field intensity is obvious, especially at 649 nm, which is as high as nearly 4 times near Au@SiO₂ NRs, and it is amazing the enhancement almost covers the whole P3HT:PCBM layer. This is helpful to enhancing the light absorption efficiency of the photoactive layer.

Fig. 4 The simulated electric field distribution in the devices with Au@SiO₂ NRs incorporated on the rear side of P3HT:PCBM at (a) 518 nm, (b) 553 nm, and (c) 649 nm using FDTD software. (d) Variances of the electric field intensity along z axis at x = 0 and y = 0 in the devices without/with Au@SiO₂ NRs. x and z axes are along the longitudinal axis of Au@SiO₂ NR and the incident direction of sun light, respectively. The inset in (d) is the amplified image with z axis ranging from −81 to −11 nm.

Fig. 5 shows the scanning electronic microscopy (SEM) images by spin-coating ethanol solution with various concentrations of Au@SiO₂ NRs on P3HT:PCBM layer. The results indicate that the distribution density increases with the concentration of Au@SiO₂ NRs. Since the dense metallic NPs can excite more intense local field via LSPR's coupling generated by adjacent Au NRs,^29–31 we deduce a higher density of Au@SiO₂ NRs (e.g., 14%) helps to achieve a higher J_SC value than those with 7% and 10% concentrations. However, we notice a slight reduction in J_SC accompanied with an obvious decrease in FF as the concentration of Au@SiO₂ NRs increases to 20%. From Fig. 5d, we observe apparent aggregations of Au@SiO₂ NRs, which may be the reason for the decline in performance. The aggregations lead to the reduction of dispersed Au@SiO₂ NRs density, thus causes the weakened local field intensity excited by LSPR. Additionally, it is well known that a good contact between photoactive layer and carrier transport layer is beneficial to enhancing the charge collection efficiency, however, the size of Au cluster in Fig. 5d is more than 200 nm, which is large enough to cause a deteriorated contact between P3HT:PCBM and MoO₃, resulting in a decreased FF value.

Fig. 5 The SEM images of Au@SiO₂ NRs spin-coated on the P3HT:PCBM surface with concentrations of (a) 7%, (b) 10%, (c) 14%, and (d) 20%.

Besides LSPR effect, the electrical properties introduced by Au@SiO₂ NRs incorporation may influence the device performances. Wang et al.³² investigated the carrier mobilities in photoactive layer without/with metallic NPs incorporation, and made a conclusion that the insertion of metallic NPs can introduce dopant states within the bandgap of polymer material, which is helpful to enhance carrier mobility. On the other hand, high NPs concentration will not favour carrier transport because it can cause the changes of morphology of photoactive layer.³³ In this work, we speculate that the reduced charge transport efficiency of the device with 20% NRs concentration leads to reduced FF due to its high NRs concentration. To prove our analysis, we calculated the series resistance (R_S) values of these devices, which is defined by the reciprocal of slope of J–V curve at V = V_OC. The results reveal that the R_S value obviously increases from 9.2 Ω cm² for the reference device to 13.1 Ω cm² for the 20% Au@SiO₂ NRs one, while the values for other plasmon devices with 7%, 10%, and 14% concentrations are almost same as that for the reference one. So we deduce that both the aggregation of Au@SiO₂ NRs and the reduction of charge transport efficiency caused by high NRs concentration lead to the increase in R_S for the device with 20% Au@SiO₂ NRs concentration, thereby decrease in FF.

3.2. Plasmon devices with Au@SiO₂ NRs on the front side of P3HT:PCBM

Incorporating metallic NPs on front electrode was often reported in previous literatures. To compare with devices in Section 3.1 that incorporates Au@SiO₂ NRs on the rear side of P3HT:PCBM, we fabricated our present device by embedding Au@SiO₂ NRs on the front side of P3HT:PCBM with the structure of ITO/ZnO/Au@SiO₂ NRs/P3HT:PCBM/MoO₃/Ag. As shown in Fig. 6, the J_SC values of these two groups of devices have a similar trend with the increase in Au@SiO₂ NRs concentration, however, for the devices with Au@SiO₂ NRs incorporated on ZnO (the front side of P3HT:PCBM), the Au@SiO₂ NRs concentration corresponding to the highest J_SC and PCE is 10%, rather than 14%. When the Au@SiO₂ NRs concentration is enhanced from 10% to 14%, the J_SC slightly decreases from 9.38 mA cm⁻² to 9.04 mA cm⁻².

Fig. 6 The J_SC and PCE values for the devices incorporating the Au@SiO₂ NRs on the rear and front sides of P3HT:PCBM with 0–20% Au@SiO₂ NRs concentration.

The other information obtained from Fig. 6 is that the highest J_SC value for the device with Au@SiO₂ NRs on the rear side of P3HT:PCBM is higher than that on ZnO, which is attributed to the following two possible factors. On one hand, unlike the device with Au@SiO₂ NRs incorporated on the rear side of P3HT:PCBM, the backward scattering/reflection from Au@SiO₂ NRs may cause a loss of part incident light intensity in the device with Au@SiO₂ NRs on the front side of P3HT:PCBM, because the sunlight first passes through Au@SiO₂ NRs before reaching P3HT:PCBM. It means as the density of Au@SiO₂ NRs on ZnO increases, the reduction of sunlight intensity into P3HT:PCBM due to scattering/reflectionof NRs would dominate compared with absorption improvement in P3HT:PCBM aroused by the LSPR, which finally leads to absorption decrease in P3HT:PCBM. To prove our analysis, we measured the transmission spectra of ITO/ZnO/Au@SiO₂ NRs with various NRs concentrations. As shown in Fig. 7, compared with the ITO/ZnO film without Au@SiO₂ NRs, the transmittivity has a very slight reduction when the Au@SiO₂ NRs concentration is 7% or 10%, while it apparently decreases in the absorption band of P3HT with Au@SiO₂ NRs concentration increasing to 14%. This means a significant attenuation of incident sunlight intensity in P3HT:PCBM layer, thereby leading to a decrease in absorption in photoactive material. On the other hand, it has been demonstrated when the blended solution consisting of P3HT and PCBM is spin-coated on substrate, these two kinds of components are not evenly distributed in the film, instead the P3HT concentration is higher on the rear side, because of its lower surface energy than PCBM.^34,35 It should be noted that P3HT, rather than PCBM is mainly responsible for absorbing sunlight in P3HT:PCBM-based photovoltaic devices, thus more P3HT component near the Au@SiO₂ NRs is more helpful to light absorption improvement aroused by LSPR in our devices with metallic NPs incorporated on rear side of the P3HT:PCBM layer.

Fig. 7 The transmission spectra of ITO/ZnO/Au@SiO₂ NRs with various concentrations of Au@SiO₂ NRs on ZnO.

3.3. Dual-plasmon device incorporating Au@SiO₂ NRs and Ag NPs

Fig. 8 displays photovoltaic performances of the reference device, the optimized mono-plasmon device incorporating Au@SiO₂ NRs on the rear side of P3HT:PCBM, and the dual-plasmon device embedding Au@SiO₂ NRs on the rear side of P3HT:PCBM and Ag NPs into MoO₃. The dual-plasmon device structure is shown in Fig. 2, in which the Ag NPs are obtained by thermally evaporating 1 nm Ag. All devices have an 8 nm MoO₃ thickness, and the dual-plasmon device has the 1 nm Ag inserted into the center of MoO₃. From Table 2 we can see the best performance is achieved in the dual-plasmon device among the three kinds of devices, and the enhancement factor of PCE is as high as 29.3% compared with that in the reference device. The trends in incident photon to conversion efficiency (IPCE) curves shown in Fig. 8b are consistent with those in J_SC values, indicating that LSPR excited by Au and Ag particles indeed improved the absorption of photoactive layer, thereby photocurrent. The J_SC values for reference, mono-plasmon, and dual-plasmon devices, calculated from IPCE curves, are 7.94 mA cm⁻², 9.32 mA cm⁻², and 9.98 mA cm⁻² respectively. We note these values are slightly smaller than those listed in Table 2. It is possibly caused by the greater width of light spot (2.5 mm) from the monochromator in IPCE measurement than that of cell area defined by overlap of anode and cathode (2.2 mm). Nevertheless, the variation trends of J_SC by either measurement or calculation are highly consistent.

Fig. 8 (a) J–V characteristics and (b) IPCE curves of the reference device, the optimized mono-plasmon device incorporating Au@SiO₂ NRs on the rear side of P3HT:PCBM, and the dual-plasmon device embedding Au@SiO₂ NRs on the rear side of P3HT:PCBM and Ag NPs into MoO₃.

Table 2 Photovoltaic characteristics for the reference device, the optimized mono-plasmon device, and the dual-plasmon device

Device	JSC (mA cm^−2)	VOC (V)	FF	PCE (%)	RS (Ω cm^2)
Reference	8.28	0.62	0.62	3.18	9.2
Mono-plasmon	9.86	0.62	0.61	3.73	8.9
Dual-plasmon	10.36	0.62	0.64	4.11	7.9

---

The SEM image of 1 nm thermally evaporated Ag on MoO₃ (Fig. 9a) shows it forms Ag clusters with size of ∼20 nm and an average spacing of 38 nm, instead of a uniform film. From Fig. 9b, we observe that inserting Ag NPs into MoO₃ exhibits a plasmonic absorption peak at 517 nm with a wide full width at half maximum of 135 nm, almost covering the whole light absorption region of P3HT. The LSPR effect of Ag NPs has been demonstrated in previous literatures.^36–38 Therefore, we attribute the further J_SC enhancement in the dual-plasmon device to the improved absorption in the photoactive layer due to the excitation of LSPR effect by Ag NPs. To prove our deduction, we simulated the electric field intensity along the incident sunlight direction (z axis) in dual-plasmon device. The cross-section schematic structure of the dual-plasmon device on xz plane is shown in Fig. 10a, in which the directions of x, y, and z axes are consistent with those in the mono-plasmon device in Section 3.1, and the space periods of Ag NPs in the MoO₃ layer, deduced from the SEM image, are 38 nm both along x and y axes in our simulation. Fig. 10b reveals the electric field intensities of the mono-plasmon device and the dual-plasmon device along z axis and cross the center of Ag NP close to Au@SiO₂ NRs (x = 38 nm). As shown in Fig. 10b, the secondary peak values of electric field intensity can be observed at z = −8 nm in the dual-plasmon device, which is attributed to the excitation by Ag NPs. The additive Ag NPs bring a more significant increase in electric field in the whole active layer. Fig. 10c shows the ratios of electric field intensity for dual-plasmon device to mono-plasmon device over the whole P3HT:PCBM layer (the z axis coordinate ranges from −11 nm to −81 nm). We find that all values are higher than 1 in the visible range with a maximum ratio of as high as 1.28 at 649 nm. Above data indicate that the combination of the Au@SiO₂ NRs and the Ag NPs induces a stronger local field in P3HT:PCBM, therefore, efficiently improving the absorption efficiency of the photoactive material.

Fig. 9 (a) The SEM image of 1 nm Ag evaporated on Au@SiO₂/MoO₃ (4 nm). (b) The absorption spectra of 8 nm MoO₃ and MoO₃ (4 nm)/1 nm Ag/MoO₃ (4 nm).

Fig. 10 (a) The cross-section schematic of the dual-plasmon device on xz plane. (b) Variances of the electric field intensity along z axis at x = 38 nm in the mono- and dual-plasmon devices at 518 nm, 553 nm, and 649 nm. The Ag NP is half elliptical spherical with a thickness of 1 nm. (c) The electric field intensity ratios of dual-plasmon device to mono-plasmon one (the coordinate on z axis ranges from −11 nm to −81 nm, over the entire P3HT:PCBM layer).

The enhanced sunlight harvest by incorporating Au@SiO₂ NRs and Ag NPs is further confirmed by the comparison of photoluminescence (PL) spectra for multilayer films of P3HT:PCBM/MoO₃, P3HT:PCBM/Au@SiO₂ NRs/MoO₃, and P3HT:PCBM/Au@SiO₂ NRs/MoO₃/Ag NPs/MoO₃. As shown in Fig. 11a, the P3HT:PCBM/Au@SiO₂ NRs/MoO₃/Ag NPs/MoO₃ structure exhibits the highest PL intensity among three multilayer films. We attribute this enhancement to the excitation of LSPR from both Au@SiO₂ NRs and Ag NPs, which increases the light absorption in photoactive layer and produce more excitons, as demonstrated by the measured absorption spectra in Fig. 11b. A 2 nm SiO₂ shell and a 4 nm MoO₃ interlayer that powerfully prevent the exciton quenching at the surface of Au NRs and Ag NPs help to enhance the PL intensity.

Fig. 11 The PL (a) and absorption (b) spectra of multilayer films of P3HT:PCBM/MoO₃, P3HT:PCBM/Au@SiO₂ NRs/MoO₃, and P3HT:PCBM/Au@SiO₂ NRs/MoO₃/Ag NPs/MoO₃. Here, we use a 470 nm excitation source.

In addition to J_SC, we also observed that FF has an obvious enhancement in the dual-plasmon device, which may be due to changes of the electronic conductivity of MoO₃ by the insertion of 1 nm Ag. To prove this analysis, we fabricated the hole-only devices without and with Ag NPs (the device structures are ITO/PEDOT:PSS/P3HT/MoO₃/Ag and ITO/PEDOT:PSS/P3HT/MoO₃/1 nm Ag/MoO₃/Ag). The J–V characteristics in Fig. 12 show that the current density in the device with Ag NPs is higher than that without Ag NPs, which implies a smaller series resistance.³⁹ The enhanced charge transfer capability of inorganic transition metal oxides by incorporating Metallic NPs has been confirmed by Zhang,⁴⁰ therefore, we concluded that the insertion of Ag NPs into MoO₃ may help to improve hole extraction, resulting in enhanced FF. It is further demonstrated by calculating R_S value in the dual-plasmon device, which is lower than that in the devices without Ag NPs. From the analysis above, we make a conclusion that besides the enhancement of optical absorption, the improvement in charge transfer helps to improving the cell performance due to the incorporation of Ag NPs.

Fig. 12 J–V characteristics of the hole-only devices without and with Ag NPs. The device structures are ITO/PEDOT:PSS/P3HT/MoO₃/Ag and ITO/PEDOT:PSS/P3HT/MoO₃/Ag NPs/MoO₃/Ag.

4. Conclusions

We demonstrated a noticeable performance enhancement for P3HT:PCBM-based inverted PSCs by incorporating Au@SiO₂ NRs and Ag NPs on the rear side of the P3HT:PCBM active layer. Our analysis reveals that the combination of two kinds of metallic NSs can effectively enhance the light absorption in the whole absorption band of P3HT:PCBM due to the strengthened near-field excited by LSPR, thereby increasing the current density. Besides arousing the LSPR effect, the incorporation of Ag NPs into MoO₃ has a positive influence on reducing the R_S of the device, resulting in an enhanced FF value. Due to the increases in J_SC and FF, a PCE of 4.11% was achieved in our dual-plasmon device, corresponding to an enhancement factor of 29.3% compared with that in the reference device.

Acknowledgements

The authors acknowledge financial support from the National Foundation for Science and Technology Development (973 project, Grant No. 2015CB932202), the National Natural Science Foundation of China (Grant No. 61274065, 61505086, 51173081, 61136003, and BZ2010043), the NSF of the Higher Education Institutions of Jiangsu Province (Grant No. SJ209003 and 11KJD510003), the Natural Science Foundation of Jiangsu Province (Grant No. BM2012010), Science Fund for Distinguished Young Scholars of Jiangsu Province of China (BK20160039), the Priority Academic Program Development of Jiangsu Higher Education Institutions (Grant No. YX030001), the Jiangsu National Synergetic Innovation Center for Advanced Materials, the Synergetic Innovation Center for Organic Electronics and Information Displays, and the Natural Science Foundation of Jiangsu Province (Grant No. BK20141424).

References

1. L. Huo, T. Liu, X. Sun, Y. Cai, A. J. Heeger and Y. Sun, Adv. Mater., 2015, 27, 2938–2944.
2. J. You, C. C. Chen, Z. Hong, K. Yoshimura, K. Ohya, R. Xu, S. Ye, J. Gao, G. Li and Y. Yang, Adv. Mater., 2013, 25, 3973–3978.
3. H. Zhou, Y. Zhang, C. K. Mai, S. D. Collins, G. C. Bazan, T. Q. Nguyen and A. J. Heeger, Adv. Mater., 2015, 27, 1767–1773.
4. S. H. Liao, H. J. Jhuo, P. N. Yeh, Y. S. Cheng, Y. L. Li, Y. H. Lee, S. Sharma and S. A. Chen, Sci. Rep., 2014, 4, 6813.
5. W. L. Barnes, A. Dereux and T. W. Ebbesen, Nature, 2003, 424, 824–830.
6. A. Y. Mahmoud, J. Zhang, D. Ma, R. Izquierdo and V. V. Truong, Org. Electron., 2013, 13, 3102–3107.
7. L. Peng, Y. Mei, S. Chen, Y. Zhang, J. Hao and L. Deng, Chin. Phys. B, 2015, 24, 115202.
8. S.-W. Baek, J. Noh, C.-H. Lee, B. S. Kim, M.-K. Seo and J.-Y. Lee, Sci. Rep., 2013, 3, 1726.
9. M. Krassas, G. Kakavelakis, M. M. Stylianakis, N. Vaenas, E. Stratakis and E. Kymakis, RSC Adv., 2015, 5, 71704–71708.
10. X. Ren, J. Cheng, S. Zhang, X. Li, T. Rao, L. Huo, J. Hou and W. C. H. Choy, Small, 2016 DOI:10.1002/smll.201601949.
11. H. L. Gao, X. W. Zhang, Z. G. Yin, S. G. Zhang, J. H. Meng and X. Liu, J. Appl. Phys., 2013, 114, 163102.
12. L. Lu, Z. Luo, T. Xu and L. Yu, Nano Lett., 2013, 13, 59–64.
13. Y. S. Hsiao, S. Charan, F. Y. Wu, F. C. Chien, C. W. Chu, P. Chen and F. C. Chen, J. Phys. Chem. C, 2012, 116, 20731–20737.
14. D. D. S. Fung, L. Qiao, W. C. H. Choy, C. Wang, W. E. I. Sha, F. Xie and S. He, J. Mater. Chem. C, 2011, 21, 16349–16356.
15. F. C. Chen, J. L. Wu, C. L. Lee, Y. Hong, C. H. Kuo and M. H. Huang, Appl. Phys. Lett., 2009, 95, 013305.
16. J. H. Lee, J. H. Park, J. S. Kim, D. Y. Lee and K. Cho, Org. Electron., 2009, 10, 416–420.
17. V. Janković, Y. M. Yang, J. You, L. Dou, Y. Liu, P. Cheung, J. P. Chang and Y. Yang, ACS Nano, 2013, 7, 3815–3822.
18. L.-W. Jang, H. Park, S.-H. Lee, A. Y. Polyakov, R. Khan, J.-K. Yang and I. H. Lee, Opt. Express, 2015, 23, 211–218.
19. A. Williamson, É. Mcclean, D. Leipold and D. Zerulla, Appl. Phys. Lett., 2011, 99, 093307.
20. N. Kalfagiannis, P. G. Karagiannidis, C. Pitsalidis, N. T. Panagiotopoulos, C. Gravalidis, S. Kassavetis, P. Patsalas and S. Logothetidis, Sol. Energy Mater. Sol. Cells, 2012, 104, 165–172.
21. G. Kakavelakis, I. Vangelidis, A. H. Jungemann, A. G. Kanaras, E. Lidorikis, E. Stratakis and E. Kymakis, Adv. Energy Mater., 2016, 6, 1501640.
22. P. Xu, L. Shen, F. Meng, J. Zhang, W. Xie and W. Yu, Appl. Phys. Lett., 2013, 102, 123301.
23. M. Yao, X. Jia, Y. Liu, W. Guo, L. Shen and S. Ruan, ACS Appl. Mater. Interfaces, 2015, 7, 18866–18871.
24. X. Li, W. C. H. Choy, L. Huo, F. Xie, W. E. I. Sha, B. Ding, X. Guo, Y. Li, J. Hou, J. You and Y. Yang, Adv. Mater., 2012, 24, 3046–3052.
25. X. Li, W. C. H. Choy, H. Lu, W. E. I. Sha and A. H. P. Ho, Adv. Funct. Mater., 2013, 23, 2728–2735.
26. L. Zhou, J. Yang, Y. Zhou, S.-T. Han, L. Huang, Y. Yan and V. A. L. Roy, Phys. Status Solidi RRL, 2015, 9, 115–119.
27. X. Li, X. Ren, F. Xie, Y. Zhang, T. Xu, B. Wei and W. C. H. Choy, Adv. Opt. Mater., 2015, 3, 1220–1231.
28. R. Zhang, Y. Zhou, L. Peng, X. Li, S. Chen, X. Feng, Y. Guan and W. Huang, Sci. Rep., 2016, 6, 25036.
29. Y. Zhang, J. Hao, X. Li, S. Chen, L. Wang, X. Li and W. Huang, J. Mater. Chem. C, 2014, 2, 9303–9310.
30. D. Kozanoglu, D. H. Apaydin, A. Cirpan and E. N. Esenturk, Org. Electron., 2013, 14, 1720–1727.
31. J. Zhu, M. Xue, R. Hoekstra, F. Xiu, B. Zeng and K. Wang, Nanoscale, 2012, 4, 1978–1981.
32. C. C. D. Wang, W. C. H. Choy, C. Duan, D. D. S. Fung, W. E. I. Sha, F.-X. Xie, F. Huang and Y. Cao, J. Mater. Chem., 2012, 22, 1206–1211.
33. W. C. H. Choy, W. K. Chan and Y. Yuan, Adv. Mater., 2014, 26, 5368–5399.
34. Z. Xu, L. M. Chen, G. Yang, C. H. Huang, J. Hou, Y. Wu, G. Li, C. S. Hsu and Y. Yang, Adv. Funct. Mater., 2009, 19, 1227–1234.
35. A. Orimo, K. Masuda, S. Honda, H. Benten, S. Ito, H. Ohkita and H. Tsuji, Appl. Phys. Lett., 2010, 96, 043305.
36. M. F. Xu, X. Z. Zhu, X. B. Shi, J. Liang, Y. Jin, Z. K. Wang and L. S. Liao, ACS Appl. Mater. Interfaces, 2013, 5, 2935–2942.
37. C. Ma, W. Qin, X. Xu, M. Li, X. Han, L. Yang, S. Yin, J. Wei and F. Zhang, Sol. Energy Mater. Sol. Cells, 2013, 109, 227–232.
38. W. Shen, J. Tang, R. Yang, H. Cong, X. Bao, Y. Wang, X. Wang, Z. Huang, J. Liu, L. Huang, J. Jiao, Q. Xu, W. Chen and L. A. Belfiore, RSC Adv., 2014, 4, 4379–4386.
39. S. Y. Khoo, H. Yang, Z. He, J. Miao, K. C. Leong, C. M. Li and T. T. Y. Tan, J. Mater. Chem. C, 2013, 1, 5402–5409.
40. D. Zhang, W. C. H. Choy, F. Xie, W. E. I. Sha, X. Li, B. Ding, K. Zhang, F. Huang and Y. Cao, Adv. Funct. Mater., 2013, 23, 4255–4261.

---

Footnote

† These authors contributed equally.

---