Fin-shaped Structure Width Control via Mandrel Spacing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for fabricating fin-shaped structures in semiconductor devices often result in variations in widths across different regions, such as logic and memory regions, affecting transistor performance due to differences in pattern densities.
Innovation Solution
A method involving the formation of mandrels with specific spacings and cover layers to control the etching process, allowing for the formation of spacers that transfer a consistent pattern to the substrate, ensuring fin-shaped structures have equal widths across regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional photolithography processes are used to define fin-shaped structures, then the manufacturing process is simple, but the fin-shaped structures cannot be directly defined with precise control
Solution Approach 1:
The fabrication process is divided into multiple stages: forming mandrels with different spacings in different regions, selectively removing mandrels based on region type (logic vs memory), and forming spacers only where needed. This segmentation allows precise control of fin widths in different regions while managing process complexity through systematic division of steps.
Solution Approach 2:
Mandrels are pre-formed with different spacings in different regions before the actual fin formation process. The cover layer is also formed in advance to protect certain mandrels. These preliminary actions enable subsequent selective etching and spacer formation to achieve precise fin width control without requiring complex real-time adjustments.
2Manufacturing precision
If SIT process is used to define fin-shaped structures, then the structures can be formed, but width variations occur across different regions due to different pattern densities
Solution Approach 1:
The patent applies different mandrel spacings to different regions: wider spacing in logic regions and narrower spacing in memory regions. The cover layer selectively protects mandrels in specific regions. This local differentiation ensures that each region receives the appropriate pattern density treatment, achieving consistent fin widths tailored to regional requirements rather than applying a uniform approach.
Solution Approach 2:
The mandrel spacing parameter is changed across different regions to compensate for pattern density variations. By adjusting the spacing between mandrels (a key geometric parameter), the process achieves consistent fin widths despite differences in regional pattern densities, directly addressing the width variation problem.
3Ease of manufacture
If uniform mandrel spacing is used across all regions, then the fabrication process is simple, but fin-shaped structures exhibit width variations in different regions
Solution Approach 1:
Instead of uniform mandrel spacing, the patent implements local quality by setting different mandrel spacings for different regions (wider in logic regions, narrower in memory regions). This local differentiation compensates for varying pattern densities and achieves uniform fin widths across all regions, resolving the contradiction between process simplicity and width uniformity.
Solution Approach 2:
The patent introduces a regional dimension to the mandrel spacing parameter, transforming it from a single uniform value to a spatially varying parameter. By considering the regional dimension (logic vs memory regions) and adjusting spacing accordingly, the process achieves width uniformity while maintaining reasonable fabrication simplicity through systematic regional classification.
Data Source
AI summary
A method for fabricating a patterned structure of a semiconductor device includes: forming first mandrels and second mandrels on a substrate, wherein a first spacing is defined between the two adjacent first mandrels and a second spacing is defined between the two adjacent second mandrels, the first spacing being wider than the second spacing; forming a cover layer to cover the first mandrels while exposing the second mandrels; etching the cover layer and the second mandrels; removing the cover layer; concurrently forming first spacers on the sides of the first mandrels and a second spacers on the sides of the second mandrels after removing the cover layer; and transferring a layout of the first and second spacers to the substrate so as to form fin-shaped structures.