UCL promotors : Xavier Gonze and Gian-Marco Rignanese.
UCL collaborators : Bruno Bertrand, Matteo Giantomassi, Anna Miglio, Samuel Poncé, David Waroquiers and Geoffroy Hautier.
External collaborations : M. Mikami Mitsubishi (Chemical Corp.), M. Coté (U. de Montréal), Martin Stankovski (U. Lund).
Funding : FRIA.
Introduction
Optical functional materials play an ever increasing role in our everyday life as well as in high-end technological applications. Photovoltaic technology, light-emitting devices (LED) and optical fibers all rely on specifically tuned optical functional materials.
Using first-principle techniques, we study such materials at the atomic scale. In optical processes, an electron and a hole are created simultaneously and they interact with one another creating either a bound exciton or an electron-hole pair. An accurate description can be obtained by combining DFT ground state calculations with many-body perturbation theory (MBPT). The equation of motion of the two-particle Green function (the
so-called Bethe-Salpeter equation) must be solved. Fully ab initio Bethe-Salpeter calculations can by now be applied to quite complex systems, like transition metal oxides, semiconductor surfaces, nanotubes, or liquids. The resulting spectra are very accurate, and serve as a benchmark for other computational schemes.
A. Transparent conducting oxides
One of our research projects focuses on transparent conducting oxides (TCO), which are an important class of metal-oxides of great industrial interest. TCO materials show the unusual property of being transparent for the visible part of the electromagnetic spectrum and at the same time being a good electrical conductor. This combination of properties makes that TCOs find many applications, like in solar cells, heated windows, flatscreens, ... Among all technologies, the efficiency of standard Si-based solar cells scores highest with more than 20%. However, due to the present Si shortage, thin film (TF) solar cells become more interesting. Actually, the market share of TF solar cells has grown from 10% in 2006 to 20% in 2010. Due to the strong need for cutting production, energy, and materials costs and for increasing the cell efficiency, a further growth can be expected. This dynamic development has triggered a search for new materials, device designs, and new concepts for enhanced cell efficiency. Doped (post)transition metal oxides or combinations of them are the most common TCO materials, Sn-doped In2O3 being the most popular one. Driven by the increased demand for environmentally friendly TCOs (Cd is toxic) with optimal transparency and electrical conductivity at a low cost (In is scarce and thus very expensive) a big research activity exists to develop and characterize new TCO materials. A good n-type TCO must simultaneously satisfy two requirements: (i) a large optical band gap, as well as a large energy separation between the conduction band minimum and the second conduction band, for transparency; (ii) a low conduction band minimum with respect to the vacuum level, for high dopability, as well as a small effective mass, for good conductivity. Using ab initio techniques the band structure has already been investigated for different TCO materials. "We have conducted (Nature Communications 4, 2292 (2013)) a high-throughput computational search on thousands of binary and ternary oxides and identify several highly promising compounds displaying exceptionally low hole effective masses (up to an order of magnitude lower than state-of-the-art p-type transparent conducting oxides), as well as wide band gaps. In addition to the discovery of specific compounds, we have performed the chemical rationalization of our findings to open new directions, beyond current Cu-based chemistries, for the design and development of future p-type transparent conducting oxides. High-throughput computations are going on, for both n- and p- type materials.
B. Materials for luminescent devices (LED)
In collaboration with Misubishi Chemical Corporation, we study materials for luminescent devices, especially those that can lead to emission of white warm light. Many different hosts, when doped with e.g. Europium or Ytterbium can convert the blue or ultraviolet light created by a LED to green, yellow or orange light. The combination of the different colours creating the desired white light. However, the efficiency of the transformation is often degraded at the moderate temperatures reached in normal operating conditions. We currently explore both the fundamentals aspects of the photoluminescence and our ability to describe them on the basis of first-principles calculations, and the specificities of such luminescence in the case of Barium-Silicon oxynitrides of interest for our industrial partner, see e.g. Phys. Rev. B 88, 075136 (2013).
C. Electronic and optical properties of silica glass
The accurate computation of electronic and optical properties of glasses is also a challenge for our tools, because of the large unit cells that are needed to faithfully represent the amorphous state (in contrast to the crystalline state), as well as to include the presence of impurities or deviation from perfect stoechiometry. This project goes hand by hand with developments in the ABINIT project.
D. New absorber materials for photovoltaics
In photovoltaic cells, the band gap of the absorber material must be chosen so as to maximize the absorbed light energy (it should not be too large, typically smaller than the visible light that is between 1.6 and 3.2 eV) and minimize wasted energy heat (and therefore also not too small). There is in fact an optimum value between 1.2 and 1.4 eV. This is why the most common solar cells (so-called first generation) are composed of silicon (whose band gap is 1.2 eV). It is enriched with doping elements (lead, arsenic, antimony or boron) to transform it into a donor (P-type) or receiver (N-type) electron. However, silicon is not a good absorbent and it takes a fairly thick layer - which has a significant impact on the cost of the cells.
The thin-film photovoltaic technology (so-called second generation) is specifically designed to reduce the cost. It is therefore necessary to find new materials that have a combination of an optimum band gap and good absorption. Cadmium telluride (CdTe) and CIGS (CuIn<sub>x</sub>Ga<sub>1-x</sub>Se<sub>2</sub>) have been used with a yield of 20%. However, these compounds are unlikely to meet the projected solar energy due to the scarcity of tellurium and indium (not to mention the harmful effects of cadmium). In addition to the criteria already mentioned, it is important to find abundant and non-toxic materials. In this regard, the CZTS (Cu<sub>2</sub>ZnSnS<sub>4</sub>) has recently attracted considerable interest due to its high similarity with CIGS. However, its performance is just over 10%. It is therefore crucial to understand the origin of such a low yield.
Several possible causes have been suggested. First, there are at least two stable phases for CZTS - the kesterite and stannite. It has been shown experimentally that they co-exist inducing a disorder. Moreover, it is possible that some defects (such as copper vacancies) have a detrimental effect on the electronic and optical properties. Finally, it is clear that the alignments of bands (including the position of the Fermi level of the anode and cathode) play a crucial role.