Reexamination of the role of nitrogen in oxynitrides-fixed charge reduction in the p+-polysilicon gate MOS

The problems associated with the use of p⁺-polysilicon gate MOS have been intensively investigated. Although the utilization of oxynitrides has been considered to be effective for the suppression of the threshold voltage (V_T) deviation in the p⁺-polysilicon gate MOSFETs, the investigation revealed that the p⁺-polysilicon gate MOS exhibits significantly different properties when oxynitrides contain no nitrogen at the oxynitride/substrate interface (MOS interface) than it does with usual oxynitrides which contain nitrogen at the MOS interface. This discrepancy arises because, contrary to what is usually considered to be the case, boron diffused into the substrate is not the origin of the negative fixed charge generated in the p⁺-polysilicon gate MOS structures, which is one of the most important factors influencing VT in those structures. We have found fluorine in the p⁺-polysilicon gate MOS structures even when the polysilicon is doped using boron ion implantation. This is a consequence of the use of BF₃ as a boron source. We propose a model in which fluorine is responsible for the negative fixed charge generation and nitrogen at the MOS interface prevents not only the boron penetration but also the negative fixed charge generation by suppressing the fluorine incorporation into the MOS interface