Modifying the solar spectrum to enhance silicon solar cell efficiency—An overview of available materials
Introduction
Spectrum modification is a well researched topic in physics and in chemistry and has been applied, for example to infrared quantum counter (IRQC) [1] or efficient lamp phosphors [2]. It is also one of the Third Generation concepts suggested to overcome the classical efficiency limit of silicon solar cells [3]. These concepts show extreme promise. While the classical efficiency limit is currently estimated to be 29% [4], detailed-balance calculations show that this could improve to approximately 37% [5], [6] using spectrum modification at one sun. In this paper we concentrate on down-conversion [7], photoluminescence [8] and up-conversion [9] and assess the potential of different systems for application to silicon solar cells.
There are three losses in a silicon solar cell that spectrum modification can reduce. The first of these is thermalization, which occur when an electron–hole pair with energy greater than the band gap of silicon (, ) is created and the excess energy is lost as heat because the electron (and hole) relax to the conduction (and valence) band edges. Thermalization losses can be reduced by using down-conversion whereby, for example, a photon with twice the energy of the band gap is converted into two photons with exactly the energy of the band gap. The second loss mechanism is imperfect collection due to recombination close to or at the surface. Since high energy photons are absorbed in this region they are more likely to be affected and the result is a reduced spectral response at shorter wavelengths. This loss can be reduced by using photoluminescence, whereby photons are shifted into an energy range where the cell has a higher spectral response. For a typical, industrial solar cell, photoluminescence would be beneficial if wavelengths shorter than approximately 500 nm could be shifted into the range 500–1000 nm. The third loss mechanism is transmission, which occurs because photons with energy less than the band gap of silicon are not absorbed. Transmission losses can be reduced by using up-conversion whereby two or more low energy photons combine to create one higher energy photon. Use of these concepts is illustrated in Fig. 1, which shows photon flux as a function of wavelength.
Modifying the incoming spectrum can be realized by incorporating layers above (for down-conversion and photoluminescence) or below (for up-conversion) existing solar cells made with established high-efficiency processes. No modification of the active layer is needed, as is the case with the impurity photovoltaic effect (IPV) [10] and the intermediate band solar cell [11]. Neither a complicated structure such as that used for tandem solar cells [12] or quantum well solar cells [13] is necessary.
The first application of up-conversion on solar cells was reported by Gibart et al. [14] in 1995. On a GaAs cell they applied a vitroceramic co-doped with trivalent erbium and trivalent ytterbium , and saw a response of the cell at an energy of 1.391 eV (the band gap of GaAs is 1.43 eV) under excitation of . The first application of an up-converter to a silicon solar cell was reported by Shalav et al. [15] in 2003. The up-converter consisted of doped with and was located on the rear side of a bifacial cell. This led to a response of the cell under excitation at 1500 nm.
Photoluminescent materials have also been applied to solar cells. This was done by van Sark using CdSe quantum dots in a transparent matrix [16]. At this stage no improvement in the solar cell response was evident. Application of photoluminescent silicon nano-crystals in a matrix was done by Švrček et al. [17], which led to enhanced spectral response at shorter wavelengths.
Section snippets
Conversion to lower energies
There are two possibilities to use the high energy part of the solar spectrum ( or 350–550 nm) more efficiently: photoluminescence and down-conversion. The two techniques are distinguished by their quantum efficiencies. For photoluminescence, the quantum efficiency is always less than or equal to one, whereas for down-conversion the quantum efficiency exceeds one (more than one photon is emitted for each incoming photon) provided non-radiative losses can be prevented.
Much of the early
Conversion to higher energies
For a solar cell, photons with energy less than the band gap are not absorbed. This low energy part of the spectrum can be accessed by using up-conversion processes. The basic principle of up-conversion is shown in Fig. 5a. Two or more incoming photons react with the up-converter, which emits at least one photon with higher energy than the incoming photons.
Up-conversion is investigated for application as infrared indicator cards, bio-lables [40] or for three-dimensional displays [41]. Since the
Conclusions
In this paper we presented different materials that are able to modify the wavelength of incoming light and assess their potential for application to silicon solar cells. Wavelength modification can be achieved using down-conversion, photoluminescence and up-conversion and is seen in a broad range of materials.
Down-converting systems that have been studied to date seem unlikely to be suitable for application to terrestrial silicon solar cells because the excitation energies required are too
Acknowledgments
This work has been carried out in the framework of the Crystal Clear Integrated Project. The EC is gratefully acknowledged for financial support under contract number SES6-CT_2003-502583.
References (117)
J. Alloys Compd.
(1995)- et al.
Thin Solid Films
(2004) - et al.
J. Lumin.
(1974) - et al.
J. Lumin.
(1974) - et al.
Opt. Commun.
(2001) - et al.
J. Lumin.
(1997) - et al.
A. Meijerink, J. Lumin.
(2000) - et al.
A. Meijerink, J. Alloys Compd.
(2000) - et al.
J. Lumin.
(2001) - et al.
J. Lumin.
(1991)