Lösungsprozessierbare Silazanpräkursoren für Gate-Dielektrika

In the display industry transistors are used, whose dielectric layers are produced through energy-intensive processes such as CVD or thermal oxidation. Alternatively to these energy-intensive processes dielectric layers can be realized via solution based processing of precursors. In the solution based processing of dielectric thin films suitable precursors are homogeneously applied onto a substrate. Here a chemical reaction is initiated in order to produce the dielectric layer. However, the current achievable highest quality layers of these techniques have leakage current characteristics, which are not yet competitive with the established procedures. The present work "Lösungsprozessierbare Silazanpräkursoren for gate dielectrics" is located in this field. The work is intended to help understand the chemistry of the precursors and to understand the influence of these chemicals to the dielectric layers. The examined precursor materials are based on silazanes. In the first step it is investigated whether the C:N-ratios in silazane based precursors have an influence on the film formation process of the dielectric layers or not. With decreasing C:N-ratios of the precursors, the C content of the layers got reduced. A higher C-content in the layers is interpreted as an indication of improved crosslinking in the layers. By improving the crosslinking it is possible to generate better layer morphologies. Thus it is possible to decrease the leakage currents of the layers at field strengths of 1 MV / cm from about 1 * 10-4 A / cm², at elevated C-content in the layers, to about 1 * 10-8 A / cm², at reduced C-content in the layers.To synthesize the precursors ammonolysis of chlorosilanes was carryed out. The reduction of the C:N-ratios of the precursors can be achieved by maximizing the chlorine content in the chlorosilanes. This optimization approach is limited to the use of three chlorine atoms per silicon atom. The second step is therefore the attempt to get the access to precursors based on the ammonolysis of trichlorosilanes with varying organic silicon substituents. It is investigated, to what extent an alternative H, alkyl or aryl substituents and varying reaction temperatures affect the ammonolysis of trichlorosilanes. Analytics on the ammonolysis products of C9H4SiCl3 show a not yet desribed chlorine containing composite. Presumably, this composites are always formed as methastabile intermediates during the ammonolysis of trichlorosilanes. For some substituents, such as tBu-groups, those intermediates can be stabilized. Analytics on the ammonolysis products of C6H5SiCl3 show the not yet described molecule bis(dichlorophenylsilyl)amine. In the next step the impact of the implementation of thermally instable substituents and varying annealing temperatures on the carbon content of the layers on the film formation process and on the electrical properties of the layers is investigated. It can be found that the C-content of the layers can be varied by these parameters. However, the implementation of thermally unstable tBu-Si substituents leads to mechanical stress inside the films during the conversion process and to correspondingly deteriorated film quality. Consequently the leakage currents of these layers are increased compared to layers without implemented unstable substituents. Depending on the transition temperature the layers show leakage currents from 6.2·10-5 A/ cm² over 1.7·10-5 A/ cm² to 2.2·10-7 A/ cm² at 1 MV / cm which is higher than the leakage current of layers based on Me silazane precursors (about 1·10-8 A/ cm²). In the next step it is examined if it is possible to realize a workable solution-based transistor. For that the silazane precursor (5) is used to create dielectric layers. With slight variations in the carbon and nitrogen content this precursor corresponds approximately to HMSQ with the formula Si6(NH)9(CH3)6 or CH3Si(NH)1.5. Dielectric layers based on this material showed leakage currents of about 1.0·10-8 A / cm² at 1 MV / cm. The material is therefore very promising. In this chapter it is shown that it is possible to produce a functional transistor. However, this transistor is not competitive compared to a transistor whose dielectric layer was produced by classic oxidation. On the one hand the transistor shows mobility values of 0.3 cm²/ Vs. A classically produced transistor shows 1.0 cm²/ Vs with the same semiconductor. On the other hand, a shift in the onset voltage of about 4.8 volts can be detected. This is attributed to an increased trap density within the dielectric.