formation of double-diffused emitter for power semiconductors - using processes based on sidewall spacers or controlled isotropic under-cutting of diffusion windows

摘要

The emitter area to be formed consists of a central, deep, highly doped region (5) and a lightly doped, lateral region (4). It is constructed by 2 separate diffusions through 2 self-aligned windows. In the first process the larger window is opened in a masking layer (22), pref. polysilicon, pref. separated from teh substrate (14) by an isolation layer (16). This is then covered by a layer (17), pref. a low temp. deposited SiO2 or spun-on polyimide, which is etched back anisotropically to leave spacers at the edges of the window, which define the diffusion window for the deep diffusion. After this diffusion the spacers are etched away isotropically using an oxygen-plasma in the case of a polyimide layer. In the alternative process a masking layer (22), pref. separated from teh substrate (14) by an insulating layer (16), is etched, pref. using SF6 with CFC13, through an etchmask layer (19), pref. photoresist. The window etched is the diffusion window for the deep diffusion (5). Then, with teh mask (19) still in place, the layer (22) is undercut, pref. using SF6 only, defining the larger window. After removal of the mask (19) the shallow diffusion (4) is made. USE/ADVANTAGE - The processes define selfaligned diffusion areas, avoiding the yield and performance reductions in current devices resulting from misalignment. It allows lower values of emitter series resistance. The processes are used in the mfr. of insulated gate bipolar transistors (IGBT) and mos controlled thyristors (MCT).