Cross-Point Memory Arrays Using Two-Mask Damascene Patterning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The development of cross-point memory cell architectures faces challenges such as mask misalignment during patterning steps and difficulties with reactive memory cell materials, particularly with noble metals like platinum and silver, which are hard to pattern due to their non-reactivity.
Innovation Solution
A two-mask damascene scheme is employed for forming cross-point memory cells, utilizing a sacrificial material to pattern noble metals and integrate them with pitch-multiplication technologies, allowing for the formation of highly integrated memory with feature sizes less than 20 nanometers, thereby addressing alignment and reactivity issues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If noble metals (platinum, silver) are used in electrodes to contact reactive oxide memory cell materials, then the non-reactivity and stability are improved, but the ease of patterning deteriorates
Solution Approach 1:
The patent introduces a separate mask layer as an intermediary material that enables the patterning of noble metals. The mask layer is deposited over the noble metal electrode material, patterned using standard photolithography techniques, and then used to transfer the pattern to the underlying noble metal through selective removal or etching. This intermediary approach allows the non-reactive noble metals to be patterned effectively without compromising their chemical stability at the electrode-memory cell material interface.
2Manufacturing precision
If multiple separate masking steps are used to pattern electrodes and memory cell material, then the manufacturing precision can be maintained, but the device complexity and fabrication time increase
Solution Approach 1:
The patent combines the patterning of electrodes and memory cell material into a single integrated masking step. The mask layer is configured to simultaneously define the patterns for both the electrode material and the memory cell material during one deposition and patterning cycle. This merging approach maintains alignment precision because both features are patterned in the same step with the same mask, eliminating cumulative alignment errors from multiple separate steps while reducing overall fabrication complexity.
Solution Approach 2:
The mask layer serves multiple functions: it acts as a pattern definition layer for both electrodes and memory cell material, serves as a protective layer during subsequent processing steps, and enables the transfer of patterns to different underlying layers with different material properties. This multi-functionality reduces the total number of separate masking operations needed while maintaining the precision required for cross-point memory fabrication.
Data Source
AI summary
Some embodiments include methods of forming memory cells. A series of rails is formed to include bottom electrode contact material. Sacrificial material is patterned into a series of lines that cross the series of rails. A pattern of the series of lines is transferred into the bottom electrode contact material. At least a portion of the sacrificial material is subsequently replaced with top electrode material. Some embodiments include memory arrays that contain a second series of electrically conductive lines crossing a first series of electrically conductive lines. Memory cells are at locations where the electrically conductive lines of the second series overlap the electrically conductive lines of the first series. First and second memory cell materials are within the memory cell locations. The first memory cell material is configured as planar sheets and the second memory cell material is configured as upwardly-opening containers.