Photonic design principles for ultrahigh-efficiency photovoltaics

Journal name:
Nature Materials
Year published:
Published online

For decades, solar-cell efficiencies have remained below the thermodynamic limits. However, new approaches to light management that systematically minimize thermodynamic losses will enable ultrahigh efficiencies previously considered impossible.

At a glance


  1. Solar-cell characteristics.
    Figure 1: Solar-cell characteristics.

    a, Energy diagram of a single-junction solar cell. Light at an energy ħω (red arrow) creates an excitation from the valence (V) to the conduction (C) band of a semiconductor. After thermalization in the conduction band an electron–hole pair is formed across the bandgap with energy Eg. Light with an energy below the bandgap (purple arrow) is not absorbed. b, Typical current–voltage (IV) characteristics of a solar cell. The short-circuit current Isc is a direct measure of the conversion efficiency from incident photons to electrical current. The open-circuit voltage Voc is determined by the factors described in the main text; it is significantly lower than Eg due to entropic reasons. The maximum-power operating point of the solar cell is indicated by the dashed lines.

  2. Light-management architectures for reaching ultrahigh efficiency.
    Figure 2: Light-management architectures for reaching ultrahigh efficiency.

    a, Three-dimensional parabolic light reflectors direct spontaneous emission back to the disk of the Sun. b, Planar metamaterial light-director structures. c, Mie-scattering surface nanostructure for light trapping. d, Metal–dielectric–metal waveguide or semiconductor–dielectric–semiconductor slot waveguide with enhanced optical density of states to increase the spontaneous emission rate.

  3. Multi-junction solar cells.
    Figure 3: Multi-junction solar cells.

    a, Multi-junction energy diagram. Semiconductors with different bandgaps convert different portions of the solar spectrum to reduce thermalization losses. The quasi-Fermi levels defining the open-circuit voltage are indicated by the horizontal blue dashed lines. The yellow dots represent the electrons. b, Parallel-connected architecture that can be realized using epitaxial liftoff and printing techniques of the semiconductor layers, followed by printing of a micro- or nanophotonic spectrum splitting layer. Each semiconductor layer can be combined with one of the structures in Fig. 2 to reduce entropy losses and these structures can be separately optimized for each semiconductor.

  4. Scalable inexpensive large-area layer transfer and nanofabrication techniques.
    Figure 4: Scalable inexpensive large-area layer transfer and nanofabrication techniques.

    a, Fabrication of ultrathin silicon wafers: hydrogen-ion implantation into a silicon wafer followed by annealing leads to formation of hydrogen bubbles at a well defined depth; the surface silicon layer can subsequently be peeled off and the remaining wafer is polished for re-use. b, Fabrication of ultrathin GaAs layers: AlAs (blue) and GaAs (orange) layers are epitaxially grown onto a GaAs substrate by chemical vapour deposition (CVD). Selective chemical etching removes the AlAs layer, subsequently the GaAs layer is lifted-off. In the lift-off processes in a and b, the thin layer is often laminated to a flexible substrate before it is peeled off. c, Soft-imprint lithography: a patterned rubber stamp (pink) is printed into a sol–gel layer (green) that is spin-coated onto the substrate. On drying of the sol–gel the stamp is removed and the pattern is transferred into the wafer by reactive-ion etching and the mask is removed. Using soft imprinting a spatial resolution of 10 nm is routinely achieved over a full 6” wafer.

  5. Thermodynamic losses in solar-energy conversion.
    Figure 5: Thermodynamic losses in solar-energy conversion.

    The maximum efficiency realized for a conventional single-junction solar cell is 28.3% (indicated in green). Dark blue bars indicate entropy-related losses and light blue bars indicate energy-related losses . The main energy loss is due to thermalization and lack of absorption. The solutions to reducing the entropy- and energy-loss problems are listed in the right-hand column.


  1. Green, M. A., Emery, K., Hishikawa, Y., Warta, W. & Dunlop, E. D. Prog. Photovolt. Res. Appl. 20, 1220 (2012).
  2. Yuen, H. in Renewable Energy and the Environment SRWB3 (OSA Technical Digest, 2011).
  3. Technology Roadmap — Solar Photovoltaic Energy (International Energy Agency, 2010);
  4. Shockley, W. & Queisser, H. J. J. Appl. Phys. 32, 510519 (1961).
  5. Green, M. A. Third-generation Photovoltaics: Advanced Solar Energy Conversion (Springer, 2006).
  6. Martí, A. & Luque, A. Next Generation Photovoltaics: High Efficiency through Full Spectrum Utilization (Institute of Physics, 2003).
  7. Landsberg, P. T. & Tonge, G. J. Appl. Phys. 51, R1R20 (1980).
  8. Ruppel, W. & Würfel, P. IEEE Trans. Electron. Dev. 27, 877882 (1980).
  9. Würfel, P. Physica E 14, 1826 (2002).
  10. Campbell, P. & Green, M. A. IEEE Trans. Elec. Dev. 33, 234239 (1986).
  11. Araújo, G. L. & Martí, A. Solar Energy Mater. Solar Cells 33, 213240 (1994).
  12. Tiedje, T., Yablonovitch, E., Cody, G. D. & Brooks, B. G. IEEE Trans. Electron. Dev. 31, 711716 (1984).
  13. Yablonovitch, E. J. Opt. Soc. Am. 72, 899907 (1982).
  14. Yu, Z. F., Raman, A. & Fan, S. H. Proc. Natl Acad. Sci. USA 107, 1749117496 (2010).
  15. Callahan, D. M., Munday, J. N. & Atwater, H. A. Nano Lett. 12, 214 (2011).
  16. Luque, A. Solar Energy Mater. Solar Cells 23, 152163 (1991).
  17. Atwater, H. A. & Polman, A. Nature Mater. 9, 205213 (2010).
  18. Atwater, J. H. et al. Appl. Phys. Lett. 99, 151113 (2011).
  19. Laux, E., Genet, C., Skauli, T. & Ebbesen, T. W. Nature Photon. 2, 161164 (2008).
  20. Coenen, T., Vesseur, E. J. R. & Polman, A. ACS Nano (2012).
  21. Spinelli, P., Verschuuren, M. A. & Polman, A. Nature Commun. (in the press).
  22. Ferry, V. E. et al. Nano Lett. 11, 42394245 (2011).
  23. Imenes, A. G. & Mills, D. R. Solar Energy Mater. Solar Cells 84, 1969 (2004).
  24. Barnett, A. et al. Prog. Photovolt. Res. Appl. 17, 7583 (2009).
  25. Green, M. A. & Ho-Baille, A. Prog. Photovolt. Res. Appl. 18, 4247 (2010).
  26. Yablonovitch, E., Hwang, D. M., Gmitter, T. J., Florez, L. T. & Harbison, J. P. Appl. Phys. Lett. 56, 24192421 (1990).
  27. Yoon, J. et al. Nature 465, 329333 (2010).

Download references

Author information


  1. Albert Polman is in the FOM Institute AMOLF, Science Park 104, 1098 XG Amsterdam, the Netherlands

  2. Harry A. Atwater is in the California Institute of Technology, Pasadena California 91125, USA

Competing financial interests

The authors declare no competing financial interests.

Corresponding authors

Correspondence to:

Author details

Additional data