ANR PROJECT « ICEMAN » (2019-2023):
Emerging solar cells
Classical solar cells do not take advantage of all the available solar energy because one part of the solar spectrum is not absorbed (> 35% for GaAs) and a part of the absorbed energy it is dissipated in heat before any carrier collection (> 40% in c-Si). Single junction solar cells efficiencies are thus limited by the so-called Shockley-Queisser limit.
We therefore study two types of emerging solar cells which take advantage of nanostructuration to go beyond this limit.
1. Hot carrier solar cells
The hot carrier concept aims at reducing the thermalization (high voltage) while obtaining a high absorption (high current) [R.T. Ross and A. J. Nosik, J. Appl. Phys. 53, 3813 (1982) https://doi.org/10.1063/1.331124].
Quantum electronic transport models developed in the NQS group are well adapted to study the interplay between these mechanisms of thermalization, absorption and transport at the nanoscale. That line constitutes the basis of our contribution to the ANR project ICEMAN which targets the realization of hot carrier solar cells using semiconductor heterostructures.
We follow two approaches:
The first one is to use our quantum numerical models to understand how an energy selective contact changes the electronic distribution of carriers in the absorber. We show a hot carrier effect when the contact is well chosen. The choice of the contact is thus crucial as can be seen in Fig. 1. Yet, that first approach does not consider neither electron-electron interaction, nor optical-phonon-acoustic phonon interaction.
Therefore, a second approach was built from quasi-analytical modelling. In that model, optical phonons form an intermediate sub-system that interacts following different relaxation times with hot electrons and cold acoustic phonons both considered at-equilibrium. That second approach permits us to determine the so-called thermalization power required to maintain electrons at a temperature higher than room temperature, and thus to investigate the power reduction induced by confinement in the quantum well, as shown Fig. 2 a) and b).
Fig. 1: Valence and conductions bands edges versus position in the ultrathin InGaAs solar cell (12 nm), a) with a non-selective contact, b) with a selective contact with E_n=0.94 eV and, c) with E_n=1.02 eV. In all cases the reservoirs are in InP and contacts between the absorber and the n-reservoir are made with barrier of AlGaAsSb. In cases of selective contact, we assume a QD between the two barriers. The Fermi level in p-reservoir is μp=0 eV, while in n-reservoir μn=0.7 eV. Colors represent the local density of states of electrons [N. Cavassilas, I. Makhfudz, A.-M. Daré, M. Lannoo, G. Dangoisse, M. Bescond and F. Michelini, submitted (2022)].
Fig. 2: Thermalization power as a function of the thickness of quantum well absorber: a) blue curves when the optical phonon population is assumed at equilibrium, b) red curves when it is assumed out-of-equilibrium. Minus and plus symbols are for the 3D case without and without screening effect, while empty and filled circles are for the 2D case without and without screening effect, respectively. Insets zoom in 2D cases [I. Makhfudz, N. Cavassilas, M. Giteau, H. Esmaielpour, D. Suchet, A-M. Daré, and F. Michelini, submitted (2022)].
2. Intermediate band solar cells
Intermediate band solar cells are based on a concept which allows in theory to go beyond the Shockley-Queisser limit. The idea is to consider states in the band gap, allowing to absorb photons with an energy lower than the band gap, which may lead to a current increase. Moreover, these states generate an intermediate band and must not be in equilibrium with either the bands of conduction or valence. In such an equilibrium situation, this intermediate band would only have the role of reducing the gap (which may increase of the current but reduce the potential).
In order to make an intermediate band we can use the ground states of quantum dots or quantum wells. However, the difficulty is to avoid the thermalization of the photogenerated electrons in these states. As a consequence, the intermediate and the conduction bands are in equilibrium. To avoid this regime, the nanostructure should be isolated from conduction band which may induce a detrimental current reduction.
To better understand this trade-off, we have developed a quantum analytical model. This allows us to understand the possibility to obtain an efficient intermediate band cell. We have to consider that the transit rate between the intermediate band and the excited state is the same as the one between the excited state and the valence band. To do so, a tunnel barrier between the excited state and the conduction band has been introduced. This barrier must be adapted to the External Radiative Efficiency (ERE) of each considered device or material (Fig.1).
Fig. 1: Maximum power generated by the intermediate band versus the thickness of the tunnel barrier isolating the nanostructure constituting such an intermediate band. This power is calculated for three External Radiative Efficiency (ERE): 2.10-5 (black), 2.10-3(red) and 0.1 (blue). In the case of close to radiative limit, a thick tunnel barrier enables to largely increase the power. The optimal thickness Lopt can be estimated with an analytical formula [N. Cavassilas, D. Suchet, A. Delamarre, J.-F. Guillemoles, F. Michelini, M. Bescond, and M. Lannoo, Phys. Rev. Appl., 13, 044035 (2020). https://doi.org/10.1103/PhysRevApplied.13.044035].