SiC MOSFET 소자에서 금속 게이트 전극의 이용

Metal Gate Electrode in SiC MOSFET

  • 방욱 (한국전기연구원 전력반도체그룹) ;
  • 송근호 (한국전기연구원 전력반도체그룹) ;
  • 김남균 (한국전기연구원 전력반도체그룹) ;
  • 김상철 (한국전기연구원 전력반도체그룹) ;
  • 서길수 (한국전기연구원 전력반도체그룹) ;
  • 김형우 (한국전기연구원 전력반도체그룹) ;
  • 김은동 (한국전기연구원 전력반도체그룹)
  • 발행 : 2002.07.01

초록

Self-aligned MOSFETS using a polysilicon gate are widely fabricated in silicon technology. The polysilicon layer acts as a mask for the source and drain implants and does as gate electrode in the final product. However, the usage of polysilicon gate as a self-aligned mask is restricted in fabricating SiC MOSFETS since the following processes such as dopant activation, ohmic contacts are done at the very high temperature to attack the stability of the polysilicon layer. A metal instead of polysilicon can be used as a gate material and even can be used for ohmic contact to source region of SiC MOSFETS, which may reduce the number of the fabrication processes. Co-formation process of metal-source/drain ohmic contact and gate has been examined in the 4H-SiC based vertical power MOSFET At low bias region (<20V), increment of leakage current after RTA was detected. However, the amount of leakage current increment was less than a few tens of ph. The interface trap densities calculated from high-low frequency C-V curves do not show any difference between w/ RTA and w/o RTA. From the C-V characteristic curves, equivalent oxide thickness was calculated. The calculated thickness was 55 and 62nm for w/o RTA and w/ RTA, respectively. During the annealing, oxidation and silicidation of Ni can be occurred. Even though refractory nature of Ni, 950$^{\circ}C$ is high enough to oxidize it. Ni reacts with silicon and oxygen from SiO$_2$ 1ayer and form Ni-silicide and Ni-oxide, respectively. These extra layers result in the change of capacitance of whole oxide layer and the leakage current

키워드