Semiconductor Data Line Layout for Contact Pattern Formation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional semiconductor memory devices face challenges in maintaining sufficient spacing between pad extension portions of word lines during the photo-etching process, especially when the line width is reduced, which complicates the formation of contact patterns.
Innovation Solution
The semiconductor device employs a data block with M parallel and sequentially arranged data lines, divided into two decoder regions, where the first data line group includes even and odd-numbered lines extending to one decoder, and the second group includes odd and even-numbered lines extending to the opposite decoder, with pad extension portions oriented in opposite directions to minimize spacing and allow for efficient contact pattern formation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of moving object
If the line width of word lines is reduced to increase integration density, then the spacing between pad extension portions is reduced, but it becomes difficult to maintain sufficient spacing for contact pattern formation
Solution Approach 1:
The word lines are divided into two separate groups: even-numbered word lines and odd-numbered word lines. Each group is routed to a separate decoder region (first decoder region and second decoder region respectively). This segmentation allows the pad extension portions of adjacent word lines to be positioned in different spatial locations, thereby maintaining sufficient spacing for contact pattern formation while enabling reduced line width for higher integration density.
Solution Approach 2:
The patent introduces a new spatial dimension by creating separate decoder regions positioned at different locations (e.g., opposite sides of the data block). By routing even and odd word lines to different decoder regions, the pad extension portions are separated in the planar dimension, allowing reduced line width without compromising contact pattern formation spacing requirements.
2Area of moving object
If the pitch between bit lines and word lines is narrowed to increase capacity, then the layout complexity increases, but providing two row decoders and two column decoders on opposite sides helps manage this complexity
Solution Approach 1:
The decoder structure is segmented into four separate decoder regions: first row decoder region, second row decoder region, first column decoder region, and second column decoder region. These are positioned on opposite sides of the data block, with even and odd word lines routed to different row decoder regions. This segmentation distributes the routing complexity across multiple regions, making the overall layout more manageable while enabling narrower pitch between bit and word lines for increased capacity.
3Ease of operation
If even and odd word lines are alternately connected to left and right decoder regions, then a larger interval between pad extension portions is obtained, but this increases the spacing requirement
Solution Approach 1:
Instead of connecting adjacent word lines (even and odd) to opposite decoder regions as in conventional designs, the patent inverts this approach by connecting even-numbered word lines to the first decoder region and odd-numbered word lines to the second decoder region. This inversion allows adjacent word lines to be routed to the same decoder region, minimizing the interval between their pad extension portions while still providing sufficient spacing for contact pattern formation through the systematic routing pattern.
Data Source
AI summary
In one aspect, a semiconductor device is provided which includes a data block including M parallel and sequentially arranged data lines numbered {0, 1, 2, . . . n, n+1, . . . , m−1, m}, where M, n and m are positive integers, and where n<m, and M=m+1, and a first decoder region and a second decoder region respectively located on opposite sides of the data block. A first data line group among the M data lines extend to the first decoder region from the data block, and a second data line group among the M data lines extend to the second decoder region from the data block. The first data line group includes even numbered data lines among the data lines {0, 1, 2, . . . n}, and odd numbered data lines among the data lines {n+1, . . . m−1, m}, and the second data line group includes odd numbered data lines among the data lines {0, 1, 2, . . . n}, and even numbered data lines among the data lines {n+1, . . . m−1, m}.