Semiconductor structure and forming method thereof

Abstract

The invention provides a semiconductor structure and a forming method thereof, and the method comprises the steps: providing a semiconductor substrate, and forming a first mask layer on the semiconductor substrate; forming discrete first sacrificial layer patterns with at least two widths on the first mask layer; forming first side walls on the side walls of the first sacrificial layer patterns, and removing the first sacrificial layer patterns to form discrete first side walls with at least two intervals; transferring the pattern of the first side wall to the first mask layer to form a patterned first mask layer, and removing the first side wall; and etching a part of the semiconductor substrate by taking the patterned first mask layer as a mask to form fin parts with at least two intervals. According to the formation method of the invention, non-uniformly distributed fins can be obtained.

Description

technical field [0001] The present application relates to the field of semiconductor technology, in particular to a semiconductor structure and a method for forming the same. Background technique [0002] As the size of the fin semiconductor structure shrinks to 14nm or below, it is shortened by the self-aligned double patterning (SADP, Self-Aligned Double Patterning) process or the self-aligned quadruple patterning (SAQP, Self-Aligned Quadruple Pattern) process. Fin pitch (Fin pitch) to increase device density. However, for some special designs, such as Static Random-Access Memory (SRAM, Static Random-Access Memory), non-uniform fin spacing needs to be introduced to obtain a process window. [0003] At present, non-uniform fin spacing can be obtained by sidewall merging, but there are many process problems in this method, for example, the number of required mask layers is large, the process is complicated, and it is prone to insufficient sidewall merging Phenomenon, the r...

AI Technical Summary

Problems solved by technology

But for some special designs, such as static random access memory (SRAM, Static Random-Access Memory), it is necessary to introduce non-uniform fin pitch to obtain process window

[0003] At present, non-uniform fin spacing can be obtained by sidewall merging, but there are many process problems in this method, for example, the number of required mask layers is large, the process is complicated, and it is prone to insufficient sidewall merging Phenomenon, the removal of the mask layer will affect the line edge roughness (LER, Line Edge Roughness), and it is difficult to control the line width and line spacing at the same time

Images