[MUSIC] In the later part of this chapter we'll describe solar cells fabricated from III V semiconductors, and in particular the high-efficiency multi-junction cells. The III V semiconductor most widespread are gallium arsenide and indium gallium, INP. Here we are interested in gallium arsenide, comparing this figure to the silicon absorption coefficient in red, and blue for gallium arsenide. At high energy, about 4 EV, all semiconductor are highly absorbent, as we have seen previously. In contrast for lower energy, silicon absorbs much less than gallium arsenide. This is due to the fact that silicon is indirect gap, which gallium arsenide is direct gap. Moreover gallium arsenide is close to the maximum for homojunction, 1.35 EV, which theoretically provides efficiencies superior to 30%. As a result of the indirect gap, a thickness of 700 microns of silicon is needed to convert the full solar spectrum. We will return to this aspect, since in fact the actual thickness of the cells on the market are of the order of 200 to 300 microns. In the case of gallium arsenide, above the gap the absorption is increased by several orders of magnitude, as compared to silicon. Therefore the thickness of gallium arsenide cells can be reduced at least by an order of magnitude, to allow the conversion of the solar spectrum. Let's open the parentheses on crystalline germanium, element of column four, but well adapted to the III V semiconductors, as will be seen later. The crystalline germanium is somewhat special, strictly speaking what we see in this figure on the left, the crystalline germanium is indirect gap. But the first direct gap, 0.8 EV, is very close to 0.7 EV indirect gap, so absorption will increase rapidly from 0.8 EV. We present here the solar spectrum with indirect gap and direct gap germanium, and the gap of silicon, which also corresponds to the gap of the amorphous germanium. Later we will see that amorphous materials always have larger gaps than crystalline materials. The crystalline germanium therefore allows to extend the conversion of solar photons in the infrared, as compared to crystalline silicon. Furthermore, unlike silicon, germanium lattice parameter is well suited to the growth of III V compounds, such as gallium arsenide. Accordingly, the germanium allows the manufacturing of crystalline multi-junction. I display here an example of a triple junction on germanium substrate, with a special layer of indium gallium arsenide, on indium gallium phosphide. InGaP converts most of visible photons on indium gallium arsenide, and of the visible on the infrared. The multi-junction allows to improve the conversion efficiency. Without concentration on small area, one can thus obtain efficiencies of over 30%. I showed above the VOC upon second voltage increases logarithmically as a function of the flux. And thus VOC increases with concentration, and therefore the conversion efficiency, which is what we see here on the right figure. Depending on the concentration, one can obtain up to 40% efficiency in the case of a triple juncture still on small areas. The maximum efficiency, 41%, is obtained for a concentration of 454. Beyond saturation is observed, probably due to thermal effects. Note that the thermal effects are even more important for narrow band gaps. These types of highly efficient cells, but expensive, is used for space applications. These multi-junction are complex torture. We present here the case of a triple junction consisting of three elementary cells. However when p-n junctions are connected in series, one can create n-p junction, parasitic heterojunction at each interface. As stated in exercise also, we must insert tunnel junction to prevent these parasitic effects, that could lead to opposite photovoltage. So to a set current limitation, current must be conserved throughout the cell. Which means that the thickness of the different materials which go towards the photo generation must be adapted at the multi-junction. It is usually not possible to insert electrodes in the triple junction, because of the epitaxy conditions. We'll see later, that in case of amorphous multi-junction, on the contrary, such electrodes can be inserted. These high-performance multi-junctions, over 40% efficiency on small area, use very expensive materials, germanium III V and so on. They are therefore often combined with concentration techniques in order to reduce the reactive area. Concentration techniques for photovoltaics are the same as for the thermal concentration. The thin plates are based on one dimensional tracking then it, they consist of linear axis parabolic shape that are less expensive. They provide concentration on the order of several lengths. Parabolic structure allow tracking on the sun in zenith and azimuth two dimension. They allow concentration of several hundred, even more. In the remainder of this section, we'll mention the use of recent developments in microelectronics to enhance the development of multi-junction processes. Thank you. [MUSIC]
