Valérie Depredurand

Degradation and passivation of CuInSe2

The degradation of CuInSe2 absorbers in ambient air is observed by the decay of the quasi-Fermi level splitting under well defined illumination with time. The decay is faster and stronger in absorb ...

Influence of copper excess on the absorber quality of CuInSe2

The compositional dependence of the optoelectronic quality of CuInSe2 thin film absorbers is investigated on single- and poly-crystalline films with varying [Cu]/[In]-ratios. We quantify the quality of the absorbers by the splitting of quasi-Fermi levels, determined by spectral photoluminescence. This quantity determines the maximum achievable open circuit voltage by an absorber. Our results indicate a significant increase of this value for growth under Cu-excess, indicating a decrease of recombination losses. By comparison of the predicted achievable open circuit voltage and the actually measured ones of finished solar cells, we find a huge “un-utilized potential” for the Cu-rich devices.The compositional dependence of the optoelectronic quality of CuInSe2 thin film absorbers is investigated on single- and poly-crystalline films with varying [Cu]/[In]-ratios. We quantify the quality of the absorbers by the splitting of quasi-Fermi levels, determined by spectral photoluminescence. This quantity determines the maximum achievable open circuit voltage by an absorber. Our results indicate a significant increase of this value for growth under Cu-excess, indicating a decrease of recombination losses. By comparison of the predicted achievable open circuit voltage and the actually measured ones of finished solar cells, we find a huge “un-utilized potential” for the Cu-rich devices.

Current loss due to recombination in Cu-rich CuInSe2 solar cells

The absorbers in Cu(In,Ga)Se2 solar cells in general are Cu-poor. However, better transport properties and lower bulk recombination in “Cu-rich” material led us to develop “Cu-rich” CuInSe2 solar cells. We expect higher diffusion lengths and better carrier lifetimes for “Cu-rich” CuInSe2 solar cells, resulting in a higher short circuit current of “Cu-rich” solar cells, compared to Cu-poor ones. However, recent investigations show that the current is lower for absorbers grown under Cu-excess compared to Cu-poor absorbers. Therefore, this work investigates both “Cu-rich” and Cu-poor CuInSe2 absorbers, as well as their resulting cells, in order to understand why the “Cu-rich” CuInSe2 solar cells do not show the expected increase in current. While this contribution gives proof that “Cu-rich” based solar cells in fact do have better carrier collection properties, one limitation of “Cu-rich” devices is a very short space charge width associated with a higher doping level. We suggest tunneling enhanced recombination in the space charge region as the most likely cause of the loss in current. This work shows also that the high doping level of the “Cu-rich” film cannot be decreased by controlling the sodium supply.

The influence of Se pressure on the electronic properties of CuInSe2 grown under Cu-excess

Standard Cu-poor Cu(In,Ga)Se2 solar cell absorbers are usually prepared under high Se excess since the electronic properties of the absorbers are better if prepared under high Se pressure. However, in CuInSe2, grown under Cu-excess, it was found that solar cell properties improve with lowering the Se pressure, mostly because of reduced tunnel contribution to the recombination path. Lower Se pressure during Cu-rich growth leads to increased (112) texture of the absorber films, to better optical film quality, as seen by increased excitonic luminescence and to lower net doping levels, which explains the reduced tunnelling effect. These findings show an opposite trend from the one observed in Cu-poor Cu(In,Ga)Se2.

In-Se surface treatment of Cu-rich grown CuInSe 2

This work focuses on the chalcopyrite CuInSe2 as a model for the more complex but also more widely used thin-film material Cu(In,Ga)Se2. Both materials are characterized by a very broad existence region that allows Cu-poor as well as stoichiometric growth. Although Cu-poor solar cells are more studied and commercially available, Cu-rich CuInSe2 exhibits qualities that make it the superior material. But due to an inherently high doping and interface problems, it has not been possible to take advantage of these. On the other hand it has been shown in previous studies, that forming a Cu-poor surface layer on Cu-rich grown CuInSe2-absorbers can greatly improve the open-circuit voltage of these solar cells. Surface treatments will be discussed, that are comprised of an indium and selenium co-deposition stage with the goal to form the Cu-poor layer by copper migration. They were performed on a new Cu-rich material, which is characterized by a low Se environment during growth. Through this it was possible to reduce the doping level greatly, which results in reliably delivering devices with high currents. Making them excellent candidates for interface optimization, that mainly effects the open-circuit voltage. Thus it became possible to produce high efficiency Cu-rich devices. There is still room for improvement though, as the devices show absorption losses in a wavelength region in accordance with a remainder of InSe on top of the CIS surface. Optimization of the process is a straightforward approach to remove this layer and shows potential for even greater efficiencies. Still the striking point is, that the here presented solar cells, are already as efficient as the Cu-poor devices, that have been published by our group.

Electrical Characterization of Defects in Cu-Rich Grown CuInSe

We study defects in CuInSe2 (CIS) grown under Cu-excess. Samples with different Cu/In and Se/metals flux ratios were characterized by thermal admittance spectroscopy, capacitance-voltage (C-V) measurements, and temperature-dependent current-voltage (IVT) measurements. All samples showed two different capacitance responses, which we attribute to defects with energies around 100 and 220 meV, plus the beginning of an additional step that we attribute to a freeze-out effect. By application of the Meyer-Neldel rule, the parameters of the two defects can be assigned to two different groups, both lying within the energy region of the so-called N1-defect that has been observed for Cu-poor absorbers.

_{\text{2}}

CuInSe2 films for photovoltaic applications grown under Cu-excess have been rarely investigated up to now. In general CuInSe2 solar cells use an overall Cu-poor absorber. In this work we argue that it is valuable to investigate Cu-rich solar cells, since all the basic material properties are better in Cu-rich absorbers. With less defects in the bulk and better transport properties it is somehow intriguing why devices with Cu-rich absorber perform less. We demonstrate that this can be attributed to the too high doping of these films. Such a high native doping leads to tunneling enhanced recombination and interface recombination, strongly affecting the devices performances. We demonstrate different attempts to overcome the problem of doping: at first a Cu-poor surface was grown on the Cu-rich absorbers which enables to decrease the doping in the space charge region, then to directly decrease the doping in the bulk, the influence of sodium content was investigated. Finally, here we show that different selenium activity during the absorber growth enables to decrease the doping of these films and to open thus a way to fully exploit the favorable properties of the Curich CuInSe2 films.