Un nuovo metodo per realizzare celle a multigiunzione ad alta efficienza e basso costo combinando gli elementi III-IV-V della tavola periodica

This thesis assesses the potential of a new technological path for combining III-V and IV elements, in order to exploit the band gap engineering possibilities which this integration offers and set up the basis for realizing, at low cost, high efficiency - monolithically thin - InGaP/InGaAs/SiGeSn/Ge solar cells. A suitable model for evaluating the performance of the MJ solar cell is built up, as well as the growth of the SiGeSn alloy and its integration in the III-V structure, by utilizing a high throughput MOVPE equipment. TMM has then been applied in the modelling of a MJ cell structure designed for improving the device voltage (obtained by thinning the solar cell and depositing on the back side a perfect mirror), however, in this case, a numerical instability has been discovered. This problem has been overcome by applying a simplified scattering matrix method (SMM). By using the optical properties of SiGeSn experimentally determined, the SMM has been successfully applied to simulate the performances of InGaP/InGaAs/SiGeSn/Ge solar cells in two and three terminal configuration. For the first time, the results of the investigation concerning the growth of SiGeSn and III-V compounds in the same MOVPE growth chamber are presented. The deposition has concerned different semiconductors currently deposited in the MJ structure: GaAs, AlGaAs, InGaAs, InGaP, Ge, SiGe and SiGeSn. Different precursors have been tested, however, in order to pursue un industrial scale up of the growth process only commercially sources have been selected. The epitaxial layers have been characterized by XRD, SEM, TEM, EDX, AFM, SIMS Ellipsometry and ECV profiling. Two key challenges have been faced: i) the problem of getting adequate tin incorporation in the Ge matrix, avoiding tin precipitation, ii) the necessity to overcome the cross contamination problem among the III-V elements and the IV elements, as they are grown in the same MOVPE equipment. Preliminary deposition of Ge, SiGe and SiGeSn has allowed to address important hardware change in the MOVPE equipment. For the deposition of IV elements, higher growth rates have been obtained by using hydrogen as carrier gas. GeH4 is shown to be more suitable than IBuGe: by using Si2H6 and SnCl4, the growth rate of SiGeSn has been increased by an order of magnitude, from 3 nm/min to 10 nm/min, by replacing the metalorganic source with the hydride one. A further 40% increase in the SiGeSn growth rate has been reached by introducing DeZn during the deposition. This result can be explained by considering that the products of DeZn decomposition (CH3 radicals) can either reduce the hydrogen or the HCl at the wafer surface. SiGeSn containing 3% of Si and 1% of Sn, lattice matched to Ge, with mirror like morphology and high crystallographic quality (without tin precipitates) has been obtained at the growth temperature of 490°C. As expected, by using thick buffer, by reducing the growth temperature and by increasing the growth rate it is possible to strongly reduce the cross contamination problem. In particular, by lowering the growth temperature from 650° to 500°C, the contamination of IV elements in III-V compounds has been reduced from 4-5*1017cm-3 to 6*1016 – 3*1014 cm-3 depending on the substrate. The contamination of III-V elements in IV based materials has been drastically reduced from 1020 cm-3 to 2*1017- 5*1015 cm-3 by using 3 µm thick and 8 µm thick Ge buffer layers, respectively. To demonstrate the effectiveness of the new MOVPE growth approach, GaAs/InGaP/SiGeSn/Ge functional devices have been manufactured. The results here presented constitute a cornerstone for the realization of low cost, high efficiency - monolithically thin - fully MOVPE grown - InGaP/InGaAs/SiGeSn/Ge solar cells and they can also be exploited to monolithically integrate III-V MJ structures on the silicon platform, paving the way toward a very competitive CPV technology.