Conventional solar devices require an antireflective coating to reduce the reflection loss at the interface between the air and the cell. However, the standard antireflective dielectric coating only works for a limited spectral range and incident angle of the radiation, and so there is a specific need to find materials with better antireflective properties. Nature can provide inspiration for solving many technological and scientific problems, and in this case, the corneas of some night-flying moths have inspired antireflective structures for solar cells.
Moths' eyes are composed of well-aligned nanoscale pillar arrays that reflect very little light at night. These nanostructures collectively function as a dielectric buffer (i.e., an intermediate refractive index) at the air-medium interface. As a result, moths' eyes are antireflective for a broad wavelength range, regardless of the incident angle of radiation. Therefore, structures inspired by moths' corneas are very desirable for collecting solar radiation, which contains a broad spectrum and an incident angle that changes during the day.
The moth-eye concept has already been implemented in various types of solar cells.1, 2 However, it has scarcely been attempted for high-efficiency triple-junction solar cells (TJSCs), even though bio-inspired nanostructures could fully complement the wide spectral absorption range of such devices. There are two major challenges. First, TJSCs are only a few microns thick. Patterning nanoscale structures on such a thin device inevitably results in surface defects, which increase unwanted losses of photogenerated carriers by recombination. Second, a TJSC essentially consists of three series of connected diodes, with each subcell engineered to absorb different portions of the solar spectrum. The output photocurrent is ultimately limited by the smallest current generated by any of the subcells, and so ‘current matching’ is required: current generation must be balanced among subcells by choosing a suitable thickness for each for the desired numbers of absorbed photons.
To surpass these technological barriers, my co-workers and I fabricated motheye-style nanostructures on a silicone nitride (SiNx) protection layer incorporated into a TJSC with subcells of indium gallium phosphide, indium gallium arsenide and germanium. Moreover, we developed a reflectance engineering approach, based on a validated model of subwavelength structures (SWS), to maximize the current output of a TJSC employing bio-inspired nanostructures.3
Our fabrication technique involves spin-coating the front surface dielectric with a monolayer of polystyrene nanospheres, self-assembled into a closely packed hexagonal lattice arrangement that serves as the sacrificial mask. We then employ ‘inductively coupled-plasma reactive-ion etching’ to shrink the nanospheres while etching down the dielectric, resulting in a 2D parabola array. Afterwards, the device is finished with standard semiconductor processing.
Figure 1(a) shows a scanning electron micrograph of the nearly close-packed monolayer arrangement of nanospheres with aperiodic domain structures. Figure 1(b and c) shows top and cross-sectional views of the SWS. Figure 1(d) displays a 5cm germanium wafer with self-assembled polystyrene nanospheres showing uniform grating-like color dispersion. The fabricated solar cell with bio-inspired antireflective structures has a black surface due to suppressed optical reflection for the entire range of visible wavelengths: see Figure 1(e). In contrast, the device with a conventional single-layer antireflective coating (SL-ARC) appears gray: see Figure 1(f).
Figure 1. Scanning electron micrographs of (a) polystyrene nanosphere assembly, with (b) tilted top view and (c) cross-sectional view of fabricated subwavelength structures (SWS). Photographs of (d) nanosphere coverage on a 5cm wafer, and fabricated triple-junction solar cells (TJSCs) with (e) SWS and (f) a conventional single-layer antireflective coating (SL-ARC).
We used angle-resolved reflectance spectroscopy to characterize the optical properties of the fabricated SWS. The measurements use an integrating sphere to capture both the specular and diffusive components of optical reflection from the surface of the control and nanostructured devices. The TJSC with an SL-ARC suffers from a high reflection loss in the UV and near-infrared wavelengths, which further deteriorates for incident angles larger than 60°: see Figure 2(a). In contrast, the reflectance map presented in Figure 2(b) displays the broadband and omnidirectional antireflective properties from the nanotextured surface. Here, a finite reflection loss is due to optical diffraction and scattering, as well as a refractive-index mismatch between the protection layer and top-cell material.
Figure 2. Angle-resolved reflectance spectra for TJSCs with (a) SL-ARC and (b) SWS in the wavelength range of 400nm to 1000nm at an incident angle of 0° to 80°. R: Reflectance.
Our fabricated device with bio-inspired nanostructures achieves a photocurrent and power conversion efficiency of 11.6mA/cm2 and 25.3%, respectively, under the standard ‘AM1.5G’ illumination condition. The photocurrent and efficiency values are superior to the control device due to an alleviated current-matching condition: see Figure 3(a). Moreover, the nanostructured device also exhibits omnidirectional photocurrent enhancement, which is ideal for concentrator operation at large acceptance angles: see Figure 3(b).
Figure 3. (a) Current-voltage characteristics and (b) normalized current density as a function of incident angle under (simulated) standard illumination conditions (AM1.5G) for TJSCs with SL-ARC and SWS. Jsc: Current density under simulated illumination.
Finally, it is also of interest to understand how the SWS structural parameters impact the output photocurrent of a TJSC. We developed a practical design based on a rigorous coupled wave analysis to first engineer the SWS reflectance and subsequently match the current generated from subcells.3 Figure 4 shows the calculated short-circuit current densities of a TJSC with different SWS periods and heights, taking into account the internal response of the top and middle subcells. According to our calculations, the optimized current density should ideally occur for structures with a period of 700nm and heights between 500 and 600nm. Such an optimization scheme allows the design of SWS for general tandem cells.
Figure 4. Calculated short-circuit current densities of a TJSC for different pitches and etch depths of SWS, taking into account the internal response of the top and middle subcells.
In summary, we have successfully incorporated bio-inspired nanostructures into TJSCs. We have also proposed a comprehensive design scheme to customize nanostructures that fully exploits the wide-range absorption of tandem cells. Looking forward, we have identified that the finite reflection loss from the nanotextured surface mainly results from the difference between the refractive index, n, of the SiNx protection layer (n∼1.8) and of the top-cell materials (n∼3.5). In the future, we plan to improve our solar cells by using materials with a higher refractive index than SiNx, such as titanium dioxide or tantalum pentoxide.
Department of Photonics
National Chiao Tung University
Peichen Yu is an associate professor. Her research focuses on the applications of nano-materials and patterning techniques for solar-cell technologies.
1. M.-A. Tsai, P.-C. Tseng, H.-C. Chen, H.-C. Kuo, P. Yu, Enhanced conversion efficiency of a crystalline silicon solar cell with frustum nanorod arrays, Opt. Express
19, p. A28-A34, 2011. doi:10.1364/OE.19.000A28
2. M.-A. Tsai, H.-V. Han, Y.-L. Tsai, P.-C. Tseng, P. Yu, H.-C. Kuo, C.-H. Shen, J.-M. Shieh, S.-H. Lin, Embedded biomimetic nanostructures for enhanced optical absorption in thin-film solar cells, Opt. Express
19, p. A757-A762, 2011. doi:10.1364/OE.19.00A757
3. P. Yu, M.-Y. Chiu, C.-H. Chang, C.-Y. Hong, Y.-L. Tsai, H.-V. Han, Y.-R. Wu, Towards high-efficiency multi-junction solar cells with biologically inspired nanosurfaces, Prog. Photovolt. Res. Appl.
, 2012. doi:10.1002/pip.2259