Innovative Multijunction with movPe and Epitaxy at low-TemperatUre for Solar – IMPETUS

Multijunction solar cells hold by far the highest conversion efficiency for any photovoltaic cell, with a 44% efficiency record under concentration, and offer a viable solution for photovoltaic power plants in high irradiation countries. If the technology trends hold true, III-V cells are poised to attain a significantly larger market share in the upcoming years, especially when associated with concentration devices, with impact in both high power solar plants and medium power off grid autonomous systems, like wireless base stations. The belief in the high potential of III-V multijunction cells has led many leading PV players, such as Panasonic, Sharp, Spectrolab, to put considerable resources in R&D in the past years. But new players have also gained an important role in bringing innovation from laboratories to production; one can mention the outstanding 28.8% of Alta Devices GaAs cells, or the 44% for III-V dilute nitrides from Solar Junction, or even Solexel with 20.1% for transferred thin film crystalline Silicon [1–3]. Although very high efficiencies have already been demonstrated, there is still room for improvement if one takes into account theoretical maximum efficiency for tandem, triple or multiple cells, with perspectives close to 60%. Cost of present multijunction cell technology still represents a major hurdle to large deployment; this is why a breakthrough is still to be done in the field of high efficiency multijunction solar cells. This project introduces two major innovations:
First, a low-temperature (< 250°C) and low-cost epitaxy using plasma enhanced chemical vapor deposition (PECVD) technique for thin film single crystal Si, Ge and Si1-xGex, combine with metalorganic vapour phase epitaxy deposition (MOVPE) for the III-V top cell.
Second, a new architecture of tandem cell, based on an inverted thin-film stack of low-temperature epitaxial Si1-xGex on top of MOVPE grown AlGaAs cell.
High quality Si and Ge epitaxy by low temperature PECVD has already been demonstrated by one of the partner, LPICM [4–8]. And more recently we have demonstrated the epitaxial growth of Si1-xGex layers with low temperature PECVD. Using Si1-xGex as the rear cell material presents two advantages: i) a small lattice mismatch with GaAs and related alloys, ii) a high absorption coefficient compared to silicon iii) possibility to tune the bandgap between Si and Ge by changing the chemical fraction. The choice of AlyGa1-yAs as the top material is justified because it provides the optimum bandgap combination with Si1-xGex (x?0.40, y?0.17), 1.63 eV/0.96 eV, able to reach conversion efficiencies in excess of 47% in a tandem configuration. In addition MOVPE technology is mastered for years by III-V Lab, a joint laboratory of Alcatel-Lucent Bell Labs France, Thales Research & Technology and CEA-LETI.

In this new submission, partners have taken into account the recommendations of the 2012 jury. Originality and advantages of our inverse growth approach is explained in details (high III-V cell quality, diffusions problem minimized, etc.). More flexibility is added with the choice of Si1-xGex material for the bottom cell to optimize both bandgap combination and current matching issue. The question of growth rate and thickness is clearly addressed, as well as the lift off step, and tunnel junction. The role of simulation and its interaction with characterization has been reevaluated. However, the main objective remains the same: to make breakthroughs for low-cost alternative III-V/IV multijunction solar cells.

In addition to the already mentioned LPICM and III-V Lab that are in charge of the process development, LGEP, will be in charge of simulation and modeling of cells in order to support overall cell performance optimization.