We report on the effect of arrays of Au nanopillars of controlled size and spacing on the spectral response of a P3HT: PCBM bulk heterojunction solar cell. Prototype nanopillar-patterned devices have nearly the same overall power conversion efficiency as those without nanopillars. The patterned devices do show higher external quantum efficiency and calculated absorption in the wavelength range from approximately 640 nm to 720 nm, where the active layer is not very absorbing. The peak enhancement was approximately 60% at 675 nm. We find evidence that the corresponding resonance involves both localized particle plasmon excitation and multiple reflections/diffraction within the cavity formed by the electrodes. We explore the role of the attenuation coefficient of the active layer on the optical absorption of such an organic photovoltaic device.
©2010 Optical Society of America
Bulk-heterojunction (BHJ) organic photovoltaic (OPV) devices have been the focus of much recent work  due to their potential in enabling affordable solar energy by a simple coating or printing process. The efficiencies of BHJ OPV devices, however, are significantly lower than silicon-based photovoltaic devices. One possible approach toward raising the efficiency of OPV’s is to increase their absorption of solar radiation. This is not the only consideration, however; in a typical BHJ organic solar cell, the optimal thickness of the absorbing layer is determined by a tradeoff: the absorber must be optically thick to absorb a significant fraction of the incident light but the thickness should not be large compared to the carrier collection length. This and other tradeoffs ultimately limit the maximum power conversion efficiency.
Introducing noble metal nanoparticles and thus coupling of light to particle plasmons in thin film photovoltaic absorber layers is emerging as a potential method for enhancing their absorption. Several experiments, calculations, and combined studies have been done in inorganic solar cells, particularly those based on silicon [2–11]. The dielectric properties of organic materials and that of silicon, however, are different. Silicon, an indirect bandgap semiconductor absorbs relatively weakly in much of the visible part of the spectrum, while the organic materials used in OPV devices have significant absorption coefficients in this range. As a result the plasmonic characteristics of noble nanoparticles in organic OPV are quite different. For organic solar cells, many reports to date involving nanoparticles are of simple absorption measurements (i.e. without the presence of electrodes and intermediate layers) or external quantum efficiency (EQE) measurements [12–15]. However, the absorption spectrum of an OPV device is expected to be considerably different in the presence of electrodes, due to reflectance at the interfaces and interference effects . The use of metallic cathodes generally prevents measurement of the absorption of the active layer in functional devices. In addition, in previous reports [12–15,17–19] the relative positions of the NPs are not well defined, making it difficult to identify the role of particle plasmon excitation. A direct comparison of measured  and calculated [18–20] optical spectral responses for a well defined metallic nanostructure configuration in a BHJ OPV has not as yet been fully carried out. In this paper, we present a systematic experimental and numerical study of the effect of periodic Au nanopillar arrays on absorption of poly(3-hexylthiophene) (P3HT) and [6,6]-phenyl C61 butyric acid methyl ester (PCBM) BHJ solar cells.
Our device architecture is shown in Fig. 1(a) . ITO (thickness 200 nm) was used as high work function anode and a 300 nm thick aluminum (Al) layer was used as a low work function cathode. We defined square arrays of square-cross section Au nanopillar patterns on a ITO-coated glass substrate using e-beam lithography; a scanning electron microscope (SEM) image of a typical NP pattern is shown in Fig. 1(b). Details of the method have been discussed elsewhere . Individual pillars were approximately 180 nm in width, approximately 70 nm in height, and spaced with a period of 540 nm. The total pattern size was 120 μm × 120 μm. The region in which the active organic layer overlaps the anode and cathode was much larger, 2.5 mm × 2.5 mm. We thus created an opaque mask, consisting of a 200 nm thick Au film and 100 nm insulator of aluminum oxide, around the nanopillar patterned area to ensure that the measured photocurrent comes almost entirely from the region of the NP pattern. The oxide layer was added to prevent a short circuit between the Al (top) and ITO (bottom) electrodes. We spun-cast a solution of regio-regular P3HT and PCBM with weight ratio 1:1 in dichlorobenzene (DCB) onto a Au nanopillar (NP) layer, which had been patterned onto a transparent indium-tin-oxide (ITO) coated glass substrate. Immediately after spin-coating, the film was “solvent dried”, i.e. placed in a covered glass container with a small amount of DCB solvent added into the bottom, for 30 min. The typical thickness of the resulting active layer was determined to be approximately 220 nm. These latter two steps were carried out in an inert nitrogen gas atmosphere inside a glove box to minimize photo-oxidative degradation. A TiOx precursor solution prepared by a sol-gel method  was spun-cast onto the P3HT:PCBM composite, resulting in a 30 nm thick layer. To maximize the volume of P3HT/PCBM in which an enhanced field due to particle plasmons within the nanopillars occurs, no PEDOT:PSS layer was used. Finally, a control cell was fabricated on each sample, using the same architecture as for the nanopillar-patterned cells.
The insert in Fig. 2(a) shows the results an of external quantum efficiency measurement (EQE) across the wavelength range from 400 nm – 800 nm. While the overall performance of cells with and without nanopillars in this range is similar, the control sample shows very slightly higher efficiency at wavelengths below ~640 nm. The close correspondence with the shape of calculated |E|2 within the active layer vs. λ (Fig. 2(b)), which is also smaller for the patterned sample, suggests that this is due to limited transmission through the array of nanopillars. Interestingly, however, the nanopillar patterned devices show higher EQE in the wavelength range from ~640 nm to 720 nm.
We now consider the origin of the improved EQE within this narrow band of wavelengths. We recall that EQE (λ) = A(λ) × IQE (λ), where A(λ) is the absorbance of the photoactive layer at a given wavelength and IQE (λ) is the internal quantum efficiency. This latter quantity is related to the conversion of the absorbed photons into free carriers and their subsequent collection at the electrodes , and has been reported to depend on λ due to the wavelength dependence of the relative absorption in P3HT and PCBM . We expect the dominant λ-dependence to come from A(λ), which we can calculate for both the nanopillar-patterned and control devices, given the geometry and optical constants of the individual structures within each device. As a first step we determined the complex refractive index, n + ik, of ITO and P3HT:PCBM by ellipsometry. Optical constants for glass , Au , titanium oxide , and Al  were taken from literature values. We calculated the optical field within our devices via the finite-difference time-domain (FDTD) method [29,30]. Based upon the calculated local field EA(x,y,z), we next find the absorbed optical power per unit volume within the active layer, which is given by
Finally the absorbance A(λ) is calculated by dividing the integration of QA over the volume of the active layer by the incident optical power .
The results of the above-described calculations for devices with and without nanopillar arrays are shown in Fig. 2(b). As seen in the inset, the overall shapes of the calculated absorption spectra for the NP patterned and control devices between 400 nm and 800 nm are also similar, with the control samples showing very slightly higher EQE below ~640 nm. Strikingly, the calculated absorbance is enhanced in the nanopillar patterned device above ~640 nm, in qualitative agreement with the enhancement in the experimentally determined EQE in this range.
To test further whether the increased EQE in this wavelength range results from the optical field enhancement, we calculated the ratio of the field strength for NP device to that for the control device and compared the |E|2 ratio to that for the measured EQE. The shapes of the simulated and experimental ratio curves, shown in Fig. 2(c), are similar; both show a peak at 675 nm wavelength. Furthermore, the magnitude of peak enhancement agrees well in the two cases: 63 % and 60 % increase for the simulated absorption and measured EQE, respectively. These results indicate that within a narrow range of wavelength: (1) the optical field enhancement occurs within the Au nanopillar arrays in the patterned devices, (2) this results in increased absorption within the bulk heterojunction organic layer, leading to (3) higher photocurrent. Near the peak of the measured EQE (i.e. λ ~575 nm, inset Fig. 2(a)) our simulations (Fig. 3(a) ) do not show strong field in the organic layer near the nanopillars.
Our FDTD simulations of the near field-squared indicate that the resonance seen in Fig. 2(c) is of mixed nature. Figure 3(b) displays the |E|2 image for a cross-section of a nanopillar for 675 nm wavelength incident radiation. High field intensity occurs both at the corners of the nanopillars and further into the organic layer and substrate, in the form of intense cloverleaf-shaped lobes. Fig. 3(c) shows the results of an additional FDTD simulation, in which we substitute for the dielectric function of the nanopillars values corresponding to Fe , for which the imaginary component dominates the real component. This effectively suppresses excitation of the localized particle plasmon, as evidenced by the absence of intense fields at the corners. The extended lobes however remain, suggesting that they are due to multiple reflections within the cavity formed by the nanopillars and the top electrode and/or diffraction. Fig. 4 shows the results of increasing the thickness of the P3HT:PCBM layer, and thus the height of the cavity on the near field-squared. Fig. 4(a) and Fig. 4(b) show |E|2 integrated over the entire organic layer volume for Au and Fe nanopillars, respectively. In this case the lobes dominate the integration; the magnitudes are similar, and both show a red shift with increasing thickness. Fig. 4(c) and Fig. 4(d) show the result of integrating over a narrow shell, 20 nm thick, around the nanopillars; again a red shift occurs for the Au case. In the case of the Fe nanopillars the integration yields nearly zero due to the suppression of the particle plasmon by the large imaginary component of the dielectric function.
Finally we investigated whether the overall efficiency would be increased for a different active layer whose absorption peak (red curve, Fig. 2(b)) coincided with the peak in the wavelength dependence of the enhanced local field near the nanopillars. We checked this by increasing the attenuation coefficient of the active layer at the wavelength corresponding to the plasmon resonance, and calculated the absorption of the active layer with and without NPs at the resonance wavelength (675 nm) with kA varying from 0 to 0.8, bracketing the measured value of 0.03. Example |E|2 images at a series of increasing values of kA are shown in Fig. 5 (a); the high field from the nanoparticles is effectively suppressed by the higher attenuation within the active layer. Interestingly, the difference between the calculated absorption within the active layer with and without nanopillars, shown in Fig. 5(b), reaches a maximum at a value of ~0.03. Figure 5(c) (although the ratio of the absorbance with and without nanopillars, shown in the inset decreases monotonically with increasing kA). For further increases in kA, the enhancement decreases, and for kA larger than ~0.2 there is no enhancement. Our calculation indicates that the measured kA (~0.03) for P3HT:PCBM is already close to optimum for maximum enhancement of absorption by Au nanopillars for the resonance we observe.
In summary, we find that while our prototype Au nanopillar-patterned devices show nearly the same overall power conversion efficiency as those without nanopillars, the patterned devices do show higher external quantum efficiency in a narrow wavelength range where the active layer absorption, however, is relatively low, from approximately 640 nm to 720 nm, with a peak of enhancement of about 60% at 675 nm. Our calculated variation of the local electric field squared with wavelength within the active layer follows that of the measured external quantum efficiency; this modeling can thus reliably be used in further optimization of the nanostructural pattern parameters and optical properties of individual photovoltaic components. We also find evidence that this resonance is of mixed nature, with contributions both from plasmon excitation and multiple reflections/diffraction within the cavity formed by the nanopillars and top electrode. Finally, our calculations indicate that the measured low value of attenuation coefficient kA of the P3HT:PCBM active layer in the wavelength range of the observed mixed resonance is already close to the optimum value for achieving maximum absorption enhancement, and that if kA were larger, which would be expected to increase the overall power conversion efficiency, then the resonance-field enhancement would be extinguished.
For completeness, although the focus of our work is on the external quantum efficiency, we include in Table 1 the measured open circuit voltages, Voc, short circuit currents, Jsc, and fill factors, FF, for our nanopillar patterned device (Table 1, column (c)), a control device (Table 1, column (b)), and a device without nanopillars, but including a PEDOT:PSS layer (Table 1, column (a)). Clearly the absence of this layer causes a large reduction in the overall efficiency. The measurements were done by measuring the output current as a function of bias voltage during illumination of a simulated solar illumination source with an air mass 1.5 Global (AM 1.5 G) spectrum and with the input intensity 95 mW/cm2.
We thank Ben Palmer for allowing us access to the e-beam lithography system used in fabricating the nanopillar arrays, Dong Hun Park for providing us access and assistance in using an ellipsometry system, and Victor Yun for fabrication of shadow masks. We are grateful to acknowledge the use of TEMPEST FDTD software, provided by Professor A. Neureuther of the University of California at Berkeley.
References and links
1. G. Dennler, M. C. Scharber, and C. J. Brabec, “Polymer-Fullerene Bulk-Heterojunction Solar Cells,” Adv. Mater. 21(13), 1323–1338 (2009). [CrossRef]
3. C. Hagglund, M. Zach, G. Petersson, and B. Kasemo, “Electromagnetic coupling of light into a silicon solar cell by nanodisk plasmons,” Appl. Phys. Lett. 92(5), 053110 (2008). [CrossRef]
4. C. Hagglund, M. Zach, and B. Kasemo, “Enhanced charge carrier generation in dye sensitized solar cells by nanoparticle plasmons,” Appl. Phys. Lett. 92(1), 013113 (2008). [CrossRef]
5. V. E. Ferry, M. A. Verschuuren, H. Li, R. E. I. Schropp, H. A. Atwater, and A. Polman, “Improved red-response in thin film a-Si:H solar cells with soft-imprinted plasmonic back reflectors,” Appl. Phys. Lett. 95(18), 183503 (2009). [CrossRef]
6. D. Derkacs, S. H. Lim, P. Matheu, W. Mar, and E. T. Yu, “Improved performance of amorphous silicon solar cells via scattering from surface plasmon polaritons in nearby metallic nanoparticles,” Appl. Phys. Lett. 89(9), 093103 (2006). [CrossRef]
7. K. R. Catchpole and A. Polman, “Design principles for particle plasmon enhanced solar cells,” Appl. Phys. Lett. 93(19), 191113 (2008). [CrossRef]
8. F. J. Beck, S. Mokkapati, A. Polman, and K. R. Catchpole, “Asymmetry in photocurrent enhancement by plasmonic nanoparticle arrays located on the front or on the rear of solar cells,” Appl. Phys. Lett. 96(3), 033113 (2010). [CrossRef]
10. S. Pillai, K. R. Catchpole, T. Trupke, and M. A. Green, “Surface plasmon enhanced silicon solar cells,” J. Appl. Phys. 101(9), 093105 (2007). [CrossRef]
11. R. A. Pala, J. White, E. Barnard, J. Liu, and M. L. Brongersma, “Design of Plasmonic Thin-Film Solar Cells with Broadband Absorption Enhancements,” Adv. Mater. 21(34), 3504–3509 (2009). [CrossRef]
12. J. H. Lee, J. H. Park, J. S. Kim, D. Y. Lee, and K. Cho, “High efficiency polymer solar cells with wet deposited plasmonic gold nanodots,” Org. Electron. 10(3), 416–420 (2009). [CrossRef]
13. S. S. Kim, S. I. Na, J. Jo, D. Y. Kim, and Y. C. Nah, “Plasmon enhanced performance of organic solar cells using electrodeposited Ag nanoparticles,” Appl. Phys. Lett. 93(7), 073307 (2008). [CrossRef]
14. T. H. Reilly, J. van de Lagemaat, R. C. Tenent, A. J. Morfa, and K. L. Rowlen, “Surface-plasmon enhanced transparent electrodes in organic photovoltaics,” Appl. Phys. Lett. 92(24), 243304 (2008). [CrossRef]
15. M. Westphalen, U. Kreibig, J. Rostalski, H. Luth, and D. Meissner, “Metal cluster enhanced organic solar cells,” Sol. Energy Mater. Sol. Cells 61(1), 97–105 (2000). [CrossRef]
16. Y. Kim, M. Ballarotto, D. Park, M. Du, W. Cao, C. H. Lee, W. N. Herman, and D. B. Romero, “Interface effects on the external quantum efficiency of organic bulk heterojunction photodetectors,” Appl. Phys. Lett. 91(19), 193510 (2007). [CrossRef]
17. A. J. Morfa, K. L. Rowlen, T. H. Reilly, M. J. Romero, and J. van de Lagemaat, “Plasmon-enhanced solar energy conversion in organic bulk heterojunction photovoltaics,” Appl. Phys. Lett. 92(1), 013504 (2008). [CrossRef]
18. B. P. Rand, P. Peumans, and S. R. Forrest, “Long-range absorption enhancement in organic tandem thin-film solar cells containing silver nanoclusters,” J. Appl. Phys. 96(12), 7519–7526 (2004). [CrossRef]
19. D. Duche, P. Torchio, L. Escoubas, F. Monestier, J. J. Simon, F. Flory, and G. Mathian, “Improving light absorption in organic solar cells by plasmonic contribution,” Sol. Energy Mater. Sol. Cells 93(8), 1377–1382 (2009). [CrossRef]
20. H. H. Shen, P. Bienstman, and B. Maes, “Plasmonic absorption enhancement in organic solar cells with thin active layers,” J. Appl. Phys. 106(7), 073109 (2009). [CrossRef]
21. T. D. Corrigan, S. H. Guo, H. Szmacinski, and R. J. Phaneuf, “Systematic study of the size and spacing dependence of Ag nanoparticle enhanced fluorescence using electron-beam lithography,” Appl. Phys. Lett. 88(10), 101112 (2006). [CrossRef]
22. K. Lee, J. Y. Kim, S. H. Park, S. H. Kim, S. Cho, and A. J. Heeger, “Air-stable polymer electronic devices,” Adv. Mater. 19(18), 2445–2449 (2007). [CrossRef]
23. P. Schilinsky, C. Waldauf, and C. J. Brabec, “Recombination and loss analysis in polythiophene based bulk heterojunction photodetectors,” Appl. Phys. Lett. 81(20), 3885 (2002). [CrossRef]
24. G. F. Burkhard, E. T. Hoke, S. R. Scully, and M. D. McGehee, “Incomplete exciton harvesting from fullerenes in bulk heterojunction solar cells,” Nano Lett. 9(12), 4037–4041 (2009). [CrossRef] [PubMed]
25. I. Zudans, W. R. Heineman, and C. J. Seliskar, “In situ measurements of chemical sensor film dynamics by spectroscopic ellipsometry. Three case studies,” Thin Solid Films 455-456, 710–715 (2004). [CrossRef]
26. P. B. Johnson and R. W. Christy, “Optical-Constants of Noble-Metals,” Phys. Rev. B 6(12), 4370–4379 (1972). [CrossRef]
28. D. Y. Smith, E. Shiles, and M. Inokuti, “The optical properties of metallic Aluminum, ” in Handbook of Optical Constants of Solids, D. Palik, ed., (Academic Press, Orlando, 1985), pp. 369–406.
29. K. S. Yee, “Numerical solutions of initial boundary value problems involving Maxwell's equations in isotropic media,” IEEE Trans. Antenn. Propag. 14(3), 302–307 (1966). [CrossRef]
30. A. Neureuther, TEMPEST FDTD software developed by Univ. of California at Berkeley.
31. P. B. Johnson and R. W. Christy, “Optical-Constants of Transition-Metals - Ti, V, Cr, Mn, Fe, Co, Ni, and Pd,” Phys. Rev. B 9(12), 5056–5070 (1974). [CrossRef]