The solar cell market is currently dominated by the crystalline silicon technology, monocrystalline or multi-crystalline, a perfectly ordered semiconductor. It will now be seen that solar cells can also be made from silicon thin films which are disordered, usually amorphous. We'll, therefore, be interested in the fundamental properties of disordered semiconductor. This is the subject of this new chapter. The band structure of this material will first be treated. Then, we'll see how we can succeed in doping them, although they are disordered. Next, we focus on the main thin film semiconductor, which is a completely disordered hydrogenated amorphous silicon. Then, the nanocrystalline silicon, which displays an order at the nanometric level. We shall first make some experimental remarks about the crystalline order in order to answer the question, is periodicity a dominant characteristic of materials? First, resistivity of most materials varies little at the melting point. Then, there are gaps in the gassy material. This is the case, for example, of conventional window glass. It is transparent, which means that solar photons pass through it without being absorbed. This means that its band gap is greater than CEV. I show here X-ray spectroscopy measurements; absorption and emission, which make it possible to characterize the density of states of the materials. At the top is the crystalline silicon, below the totally disordered amorphous silicon. Let's recall the principle of the method. Sufficiently, high photon energy will be able to create transitions between the core level on the first available energy band. That is to say, the condition band. By varying the energy of the X-rays, it will be possible to measure the density of states of the condition band. Moreover, if one removes the electron from the core level, the system will relax. This will result in emission of a photon from the last populated band energy, so the valence band. The measurements are presented in this figure. Also, the band structures are different in the amorphous and the crystal. In both cases, there is a band gap. That is to say, an energy gap in which there is no state available. Thus, even if the semiconductor is disorder, it may present a band gap. Let us now turn to doping. Consider, for example, the n-doping of silicon. Recall that in the case of a crystal, if when substitutes a phosphorus atom, for example, to a silicon one without disturbing the lattice crystal, the phosphorus atom having five valence electrons, the fifth electron will be quasi-free. The doping mechanism appears to be a direct consequence of the crystalline order. Now, let's consider in the figure below the disorder semiconductor. In this case, the solid lattice can relax in such a way as to incorporate a phosphorus atom which will be able to share free electrons with the neighboring silicon atoms without any doping effects related to any fifths quasi free electron. This was the experimental situation until 1980s. The existence of a band gap in the amorphous was known but it was thought that they could not be doped. That prevented the preparation of electronic devices from silicon thin films. In 1975, a discovery in Scotland Dundee University in parallel with research article pre-critique on XEROX-Parc in California would break this image. There are many ways of preparing silicon thin films. The method presented here consist in depositing the thin films of silicon from a reactive plasma. More precisely, it consist in introducing a silicon buffer gas and thus the silane, SiH4 into a glow discharge. The silane is dissociated by electrons and thus a film of silicon is deposited on the glass substrate at typically 200 degrees. The main characteristic of this process is to start with a silane, SiH4, atomic hydrogen is then produced from the silane dissociation. The growth of silicon thin films is therefore carried out in the presence of hydrogen valence one, which can possibly saturate the defects of silicon as illustrated in this figure. This plasma methods allows a decrease in the density of defects by several orders of magnitude. It's possible to introduce into the discharge either a phosphorus carrier gas, phosphine or boron carrier gas, diborane B2H6 mixed with the silane. This figure shows the variation of the conductivity as a function of the phosphine or diborane contact is a gaseous mixture. Variation in the conductivity of hydrogenated amorphous silicon of 10 orders of magnitude is obtained, which is considerable. The doping effect n or p inverse demonstrated by phosphorus or boron. This variation in conductivity reveals that phosphorus or boron are incorporated into the disordered amorphous silicon creating a doping effect. Therefore, the crystalline order is not a necessary condition for the existence of the doping effects. Nevertheless, it will be seen herein after that the properties of disordered semiconductors are very different from those of crystalline semiconductor. With regard to the band structure, one of the main characteristic of the disorder is lack of conservation of the k vector. Further description in terms of [inaudible] just disappears. It will also be seen that the disordered semiconductors have states in the band gap related to the presence of defects. Some of the band edge states are localized. That is to say, they do not contribute to the transport of carriers. Also, doping is possible as we have seen before. Doping efficiency is very low in amorphous especially for high doping. One percent or less instead of 100 percent for the crystal. Another drawback for applications such as thin film solar cells is related to the increase of the density of defects with doping. We will, of course, consider the consequences of these characteristics of thin film semiconductor on the properties of electronic devices, in particular solar cell. It will be seen in particular that the efficiency of the cells based on amorphous silicon is two or three times lower than that obtained from crystalline silicon. It will also be seen that the disordered materials exhibit specific phenomena of instability linked using incorporation of hydrogen. Thank you.
