| Summary: | Rectifying and non-rectifying contacts were fabricated in n-type 4H-Silicon Carbide. To improve the characteristics of the fabricated contacts various different surface pre-treatments were used. The impact of these pre-treatments on both contacts types was evaluated, and any improvements recorded. The findings report on the improvements made to the specific contact resistance of nickel contacts fabricated to n-type 4H-SiC epilayers. Ohmic contacts with an average value of specific contact resistances as low as 1.15 x 10<sup>-4</sup> Ω cm<sup>2</sup> following annealing at around 900<sup>o</sup>C were created. In addition nickel Ohmic contacts were created with similarly low specific contact resistance, which required no such annealing, a phenomenon never previously reported. Further investigation was conducted into the reasons behind this finding, and a hypothesis developed. Various different surface preparations were also experimented for use in the formation of Schottky contacts. No improvements were seen over the standard cleaning process however. High power Schottky diodes were fabricated using a single nickel Schottky contact that exhibited reverse breakdown voltages of around 600V. This figure was improved upon through the use of boron implantation as an edge termination but at the detriment of the forward I(V) characteristic. Further development of the diodes, using a multiple metal Schottky contact, yielded breakdown voltages of 1kV without the need for any further edge termination. This value is more than 85% of the theoretical value for reverse breakdown. In addition to the work on metal contacts, investigation was also performed into the use of ion implantation for the purpose of semiconductor doping. A database was developed to allow the prediction of implant profiles for both Boron and Nitrogen into SiC. This prediction compared well to experimental results. The damage created by high temperature annealing of SiC and the possible steps to prevent and repair this damage is also investigated.
|
|---|
