Enhanced Light Trapping in Periodic Aluminum Nanorod Arrays as Cavity Resonator

Transcription

1 Enhanced Light Trapping in Periodic Aluminum Nanorod Arrays as Cavity Resonator Rosure B. Abdulrahman, Arif S. Alagoz, Tansel Karabacak Department of Applied Science, University of Arkansas at Little Rock, Little Rock, AR, U.S.A. ABSTRACT Metallic nanostructures can exhibit different optical properties compared to bulk materials mainly depending on their shape, size, and separation. We present the results of an optical modeling study on ordered arrays of aluminum (Al) nanorods with a hexagonal periodic geometry placed on an Al thin film. We used a finite-difference time-domain (FDTD) method to solve the Maxwell's equations and predict the reflectance of the nanorod arrays. The thickness of the base Al film was set to 100 nm, and diameter, height and nanorod center-to-center periodicity were varied. Incident light in the FDTD simulations was an EM-circular polarized plane wave and reflectance profiles were calculated in the wavelength range nm. In addition, we calculated spatial electric field intensity distributions around the nanorods for wavelengths 300, 500, and 700 nm. Our results show that average reflectance of Al nanorods can drop down to as low as ~50%, which is significantly lower than the ~90% reflectance of conventional flat Al film at similar wavelengths. In addition to the overall decrease in reflectance, Al nanorod arrays manifest multiple resonant modes (higher-order modes) indicated by several dips in their reflectance spectrums (i.e. multiple attenuation peaks in their absorption profiles). Positions of these dips in the reflectance spectrum and spatial EM field distribution vary with nanorod height and diameter. Multiple reflectance peaks are explained by cavity resonator effects. Spatial EM field distribution profiles indicate enhanced light trapping among the nanorods, which can be useful especially in optoelectronic and solar cell applications. INTRODUCTION Engineering optical properties of nanomaterials can provide lead to high performance optoelectronic devices. In the specific case of solar cells, nanostructured metals can be used to enhance light trapping in solar cell absorbent layer by using surface plasmon effects [1]. Surface plasmons (SP) are collective excitation of free conduction electrons at metal dielectric interface. When metal is illuminated by light, electromagnetic field couples with surface plasmons and propagate along the surface, which is called surface plasmon polariton (SPP). At nanoscale, these waves are confined on the nanostructure surface and called localized surface plasmon resonances (LSPR). LSPR is very sensitive to nanostructure shape and size [2]. It has been shown that, geometries such as nanoparticles, nanoshell [3] and coaxial holes [4] can utilize plasmonic behavior of metals by introducing field enhancement. Crystalline Al exhibits interband absorption at far infrared wavelengths of 826 nm and 2479 nm [5]. Al nanoparticles show two plasmon resonances at 250 nm and 190 nm due to dipolar field and quadrupolar field, respectively, and exhibit red shift when the particle size is increased [6]. In this paper, we demonstrated enhanced light trapping between hexagonally packed well-ordered Al nanorod arrays by using finite-difference-time-domain (FDTD) method. Hexagonally packed Al nanorods can be fabricated by combination of lithography and etching techniques.

2 FDTD SIMULATION OF METALLIC NANOROD ARRAYS Optical properties of Al nanorod arrays were modeled by using a three-dimensional (3D) FDTD method with the help of a high performance parallel computing cluster. Figure 1 depicts 3D simulation region, light source, monitors and hexagonally packed periodic Al nanorods array on top of 100 nm Al thin film. X- and Y- boundaries of simulated region were set to periodic boundary conditions (PBCs) due to the nanorod array periodicity along these directions. Z- boundaries of simulated region were set to perfect matched layers (PMLs) in order to diminish back reflection. The nanorod array geometrical parameters including periodicity (P), height (H), diameter (D), and separation (i.e. gap) between the rods (P-D) were varied as summarized in Table 1. TE-circular polarized plane wave injected along surface normal towards to nanorod array. Figure 1. Schematics of FDTD simulation model for Al nanorod on top Al thin film; (a) perspective view, (b) top view, and (c) side view. Table 1. Nanorods parameters P (nm) H(nm) P-D=450 nm P-D=250 nm P-D=50 nm , 500, 700 D=50 nm D=250 nm D=450 nm , 500, 700 D=300 nm D=500 nm D=700 nm , 500, 700 D=550 nm D=750 nm D=950 nm RESULTS and DISCUSSION Light reflection and trapping behavior of hexagonally packed Al nanorods on top of Al thin were systematically investigated. Figure 2 shows the effect of nanorod diameter and length on light reflection characteristic of nanorod arrays. Multiple reflection dips (higher order) were observed for different nanorod geometries with ~ 0% transmission due to 100 nm thin film layer at the bottom [7]. Figure 2 shows that average reflectance of Al nanorods can drop down to as low as ~50% at ultraviolet and visible regions, which is significantly lower than the ~90% reflectance of conventional flat Al film at similar wavelengths. In order to investigate the effect of rod-to-rod coupling and periodicity, we varied the rod-rod separation (P-D) and periodicity (P) values, respectively. Both have strong effects on the reflectance profiles as can be seen in Figure 2. However, in order to keep the separation constant, we had to change the rod diameter as well, which can cause changes in the individual nanorod SP response itself. Therefore, the relative

3 contribution of periodicity effect in our results is not clear and we plan to investigate this in more detail. Almost all the nanorod array geometries show significantly reduced reflectance profiles with SP dip positions and magnitudes strongly changing with the nanorod height, diameter, separation, and periodicity. The only exception is the nanorods with P=500 nm and D=50 nm, which has the rod-rod distance P-D=450 nm has a similar reflectance behavior to that of conventional Al thin film at VIS and NIR. However, even this geometry shows reduced reflectance at UV and extend to NIR for longer nanorods (blue curves in Figure 2 (a, d and g)). Figure 2. FDTD simulated UV-VIS-NIR reflectance spectrums of Al nanorod arrays placed on top of a 100 nm Al thin film are shown for different rod height (H), diameter (D), periodicity (P), and rod-rod separation (i.e. gaps, P-D) values. It seems that at such larger rod-rod separation conditions, the structure behaves like a rough surface and might have scattered only the UV wavelength. However, this behavior immediately changes when D is increased to 250 nm and 450 nm for the same periodicity of P=500 nm (blue lines in Figure 2(a, b, and c)), where the rod-rod distance decrease and present higher order longitudinal plasmon modes with much lower overall reflectance values compared to Al film (red dashed lines. This is believed to originate from enhanced rod-rod coupling with stronger electric field between more closely spaced nanorods. Overall reflectance values are the smallest

4 for long nanorods of 700 nm height and smaller separations (P-D) of 250 nm and 50 nm. Reflectance at ultraviolet region is quite sensitive to nanorod height for nanorods of P=750 nm & D=300 nm, and P=1000 nm & D=550 nm (Figure 2(a, d and g)). This sensitivity decreases for P=750 nm & D= 500 nm and P=1000 nm & D=750 nm (Figure 2(b, e and h)), and almost diminished at P=750 nm & D=700 nm and P=1000 nm & D=950 nm (Figure 2(c, f and i)). At the visible and near infrared region several strong plasmonic dips are observed for P=750 nm and P=100 nm.300 nm and 700 nm height nanorods showed two strong longitudinal plasmonic dip shifting to red with increasing nanorod diameter, which is believed to be due the enhanced SP response along the sides of nanorod walls. 500 nm height nanorods showed one strong plasmonic dip with P=750 nm. For the same nanorod height, increasing nanorod diameter exhibit red shift in resonance modes which can be attributed to retardation effect [6]. Reflectance results indicate that Al nanorods arrays and underlying thin film together can form a plasmonic nanocavity/nanogap structure and can enhance light trapping at some specific wavelengths. Since nanoparticles do not form clear cavity geometry, this effect is limited to only inter-particle distances shorter than 10 nm [8]. On the other hand, our results show that hexagonally packed nanorods can exhibit cavity-induced LSPR at nanorod-nanorod gaps larger than 50 nm even up to 450 nm. Within the cavities among nanorods standing waves can form which in turn can lead to enhance light trapping. If the gaps are filled with a light absorbing photosensitive materials as in the case of solar cells, this can lead to enhanced light absorption and improved photo-generation of charge carries. In order to further investigate the light trapping behavior in detail, electric field intensity distribution around the Al nanorods were calculated for nanorods of different diameter D = 300 nm, 500nm, 700 nm and constant height H = 700 for light wavelengths 300 nm, 500 nm and 700 nm. Figure 3. FDTD simulation results of cross-sectional electric field intensity distribution for H = 700 nm long Al nanorod arrays of different diameters D = 300, 500, and 700 nm calculated at different wavelength values of λ = 300, 500, and 700 nm. There is an underlying Al thin film of 100 nm thickness. Inserts show top-view field intensity calculated at the half of nanorod height.

5 As shown in the Figure 3 s top-view inserts, electric field intensity distribution follows the hexagonal symmetry of Al nanorods. Moreover, side view images of Figure 3show the high electric field intensity regions among the gaps of nanorods that further indicate the cavityinduced surface plasmon effects. In other words, light trapping occurs among nanorods arrays, which leads to enhance absorption by the metal and much reduced reflectance values as shown above. In addition, images in Figure 3show the signature of interference pattern of incoming and reflected light. Detailed interference pattern changes with changing the diameter at each wavelength. One could ask the actual contribution of cavities associated with the nanorod array geometry to the light trapping results we observed. In order to distinguish this geometrical effect from nanostructure size effects, we carried out FDTD simulations for a single Al nanorod on a base 100 nm thick Al film as shown in Figure 4(a). Figure 4. (a) Schematics of FDTD simulation model for a single Al nanorod on top Al thin film. (b) Simulated UV-VIS-NIR reflectance spectrum of a flat 100 nm thick Al thin film, single nanorod, and nanorod arrays of H = 700 nm and D = 300 nm on top of 100 nm Al thin film. (c) Cross-sectional electric field intensity distribution images at λ = 288 (I), 404 (II), 649 (III) and 707 (IV) nm, and at non-plasmonic resonance wavelength of 1400 nm (V). Figure 4(b) compares the reflectance profile of a single nanorod of height H = 700 nm and diameter = 300 nm to the reflectance of nanorod arrays of same rod H and D (with periodicity 750 nm), and to that of conventional flat Al film of 100 nm thickness. While single Al nanorod showed less reflection (e.g. ~80% at UV and visible regions) and higher order LSPR modes compared to thin film (~90% reflectance), Al nanorod arrays exhibit much lower reflectance values (~50% reflectance) and increased depth of the plasmonic resonance dips [9]. In order to demonstrate the field intensity in the plasmonic resonance, we calculated the electric field distribution at some plasmonic wavelengths and compared with a non-plasmonic resonance wavelength that was chosen to be 1400 nm. Figure 4(c) shows electric field intensity for

6 resonance wavelengths of 288 nm (I), 404 nm (II), 649 nm (III) and 707 nm (IV) are much stronger among the nanorod gaps compared to that of the non-plasmonic wavelength of 1400 nm (V). Furthermore, III and IV reveal stronger confinement for electric field compare to I and II, which can lead to superior light trapping property at those plasmonic wavelengths. CONCLUSIONS In this study, we demonstrated that metallic nanorod arrays can provide enhanced light trapping property, which seems to be mainly originating from cavity-induced surface plasmon effects. Gaps among nanorods act as the cavity while their smaller dimensions introduce surface plasmon effects. Cavities significantly amplify the surface plasmon response and lead to significant confinement of electric field intensity of the reflected light. This translates into superior light trapping, which eventually leads to enhance absorption by the metal and significantly reduced reflectance values. However, there might be also contributions from the individual nanorods SP response as we change the geometrical parameters of the structure. Relative contributions of these parameters are not clear at the moment and needs further detailed studies. Metallic nanorods arrays can be used in optoelectronic and solar cell devices that can result in higher optical absorption at the active absorber layer due to the their enhanced light trapping property. We are currently conducting experiments to investigate the applications of these nanorod arrays in solar cell devices. ACKNOWLEDGMENTS This work was supported by NASA under the grant number NNX09AW22A. REFERENCES 1. V.E. Ferry, A. Polman, H.A. Atwater: Acs Nano 12, (2011) 2. S.A. Maier: Plasmonics: Fundamentals and Applications (springer ) pp K. Tanabe: J. Phys. Chem. C 112, (2008) 4. H.A. Atwater, A. Polman: Nature Materials 9, (2010) 5. A J Hughes,D Jones and A H Lettington: J. Phys. c (Solid St. Phys.) 2, (1969) 6. Y. Ekinci, H.H. Solak, J.F. Loffler: J. Appl. Phys. 104, (2008) 7. Z. Liu, Y. Wang, J. Yao, H. Lee, W. Srituravanich, X. Zhang: Nano Lett.9, (2009) 8. K. Ueno, H. Misawa: Bull. Chem. Soc. Jpn. 85, (2012) 9. M.A. Sefunc, A.K. Okyay, H.V. Demir: Appl. Phys. Lett. 98, (2011)