Final Report Summary - PHOTOQWELL (Photonic optimisation of multiple quantum well structures for single and dual-<br/>junction solar cells) The main goal of this research project was the photonic optimization of strain balanced multiple quantum well (MQW) solar cells for high efficiency photovoltaic devices. These kind of structures had shown their benefits in the previous decade by continuously improving the material quality and the solar cell conversion efficiencies and they offered a unique opportunity to beat the limits achieved by other competing technologies.The high efficiency multi-junction GaInP/GaAs based solar cells makes them the leading technology for space applications and they are becoming increasingly popular in terrestrial concentrator systems. Despite the success of this technology, reaching efficiencies of 41.6% when combined with a third Ge junction, or 44% when combined with a dilute nitride bottom junction – result achieved during the course of this project – there is a fundamental limitation related with the available combination of materials: monolithic, lattice matched solar cells have larger bandgaps that the optimum under typical global and direct spectra. This limitation becomes more apparent with increasing number of junctions pursuing 50% efficient solar cells. MQW solar cells offer a way to overcome such limitation by enabling a flexible tuning of the bandgaps while keeping good crystalline structure and optical and electrical properties. To further improve upon current efficiencies of MQW solar cells and reach the maximum efficiencies, attention must be paid to some of the fundamental processes in the solar cell. The structure of the present, record solar cell based on QWs – a single junction with 28.3% efficiency – has evolved from experimental trial and error, but a more optimal structure can be envisaged for multi-junction devices, guided by theoretical insight into the balance between optical absorption and electrical recombination.Three basic phenomena had to be addressed: matching the QW absorption to the solar spectrum, maximise the photogenerated carrier extraction from the QW and promote photon recycling and radiative coupling between subcells. These general goals constituted the spine of this project. During this project a drift-diffusion – Poisson equation solver was developed able to simulate the transport properties of MQW structures. These equations describe how electrons and holes move within a semiconductor material, where they are lost – recombination processes –, where they are absorbed and, ultimately, what is the current flowing through the device at a given voltage. For any optimisation process of MQW solar cells, obtaining a reliable and fast simulation solver was a key milestone. Additionally, it was necessary to have a means of calculating the amount of light absorbed by the QWs. This was done in close collaboration with other colleges of the Quantum Photovoltaic Group, developing a band structure and absorption calculator – 8-band k•p method – for arbitrary profile and composition QWs. The combination of both solvers allowed to explore the dependence of the performance of the solar cell with the properties of the quantum wells in a way that could not be attempted by experimental trial and error. Several conclusions could be drawn from these simulations. First, QW structures made by multiple layers – in opposition to single layer, square QWs as it had been used so far – offered a potential for higher light absorption by increasing the number of possible energy transitions within the well. Additionally, it was discovered that the position of the QWs inside the solar cell played a major role in the carrier collection efficiency. In particular, it was shown that if the MQW structure was broken into two smaller stacks, leaving a gap in the appropriate location of the cell, recombination was much lower and the current produced by the cell could be up to 6% higher (figure 1). Finally, the strong dependence of the MQW solar cell properties on the mobilities of the carriers across the QWs was shown, a parameter barely known or studied in the past for this system. The impact of the mobilities showed that, for each value, there was an optimum number of QWs beyond of what including more would be detrimental for the solar cell performance. In parallel to this theoretical work and the development of a roadmap for multi-junction solar cell performance in order to reach 50% efficiency, experimental work was carried on with the project partners of the Quantum Photovoltaic group: the Fraunhofer Institute for Solar Energy, in Germany, and the University of Tokyo, in Japan, highlighting the international nature of the research. Incorporating some of the conclusions of the theoretical analysis, a solar cell with QWs - designed to be used in four junction devices - was fabricated, achieving an unprecedented absorption edge for QWs on GaAs of 1.15 eV and outstanding photocurrent, paving the way for the design and fabrication of a world record 4J solar cell with MQW in the short term. The experimental work also led to unexpected and very exciting results. During the analysis of QW samples fabricated under different conditions – in particular under mis-oriented substrates such as those needed in real multi-junction solar cells – it was discovered that they were not really QW, but quantum wires: the thin semiconductor layers had suffer from severe thickness modulation and were forming stripes of triangular section in a particular direction (figure 2). Such quantum wires (QWR) had been observed in other contexts but never expected or consider in solar cells. A detailed analysis of their properties showed that photogenerated carriers survived without recombining for much longer times than in QWs – 10 times longer. This opened two novel approaches for the design of solar cells with nanostructures: using QWRs instead of QWs to enhance carrier collection and using QWRs for novel solar cell concepts, such as intermediate band solar cells. The outcomes of this project provide two major resources for the photovoltaic community worldwide in the near future. One is an invaluable set of computational tools and rules for optimising MQW solar cells that are already guiding the experimental work towards the most efficient designs. The second is a novel type of nanostructures to study new physics and ideas for solar cells, the quantum wires, with exciting and challenging properties that deserve further exploration. Powiązane dokumenty final1-figures.pdf
