Process Chamber Matching Through Input and State Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing chamber matching techniques only focus on matching process outputs across different chambers, which does not necessarily imply that the states of operation of the chambers are matched, leading to decreased yield and throughput.
Innovation Solution
Implementing an improved definition of chamber matching that matches both input recipes and output values across chambers, ensuring that process conditions such as process inputs and variables are also aligned, thereby matching the states of operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If process control is used to tune each chamber independently to match process outputs, then process output matching is improved, but true state matching deteriorates
Solution Approach 1:
The system implements feedback control by continuously monitoring process outputs and adjusting process inputs accordingly. The controller compares actual process outputs with target values and modifies process inputs in subsequent runs to achieve desired outputs, enabling dynamic adaptation and true state matching across chambers.
Solution Approach 2:
The system changes process input parameters (such as temperature, pressure, power) based on observed output deviations. By adjusting these parameters iteratively through feedback, the system achieves both output matching and true state matching, resolving the contradiction between the two goals.
2Manufacturing precision
If chambers are tuned to produce same output with different inputs, then output consistency is improved, but process efficiency deteriorates
Solution Approach 1:
The system optimizes process input parameters to achieve target outputs while considering efficiency metrics. By systematically adjusting parameters such as temperature, pressure, and time, the system finds optimal combinations that both match outputs and maximize process efficiency, preventing wasteful parameter combinations.
Solution Approach 2:
The system replaces trial-and-error mechanical tuning with automated computational optimization. The controller uses algorithms to calculate optimal process inputs based on desired outputs and efficiency constraints, substituting inefficient manual adjustment processes with intelligent automated control.
3Difficulty of detecting and measuring
If only process outputs are matched across chambers, then measurement simplicity is improved, but yield deteriorates
Solution Approach 1:
The system uses feedback from process outputs to infer and control internal chamber states. By monitoring outputs and using this information to adjust inputs, the system indirectly controls unmeasured states that affect yield, achieving both measurement simplicity and high yield through intelligent feedback control.
Solution Approach 2:
The controller acts as an intermediary that translates simple output measurements into comprehensive state control. It uses output data as a mediator to infer internal chamber conditions and adjust inputs accordingly, bridging the gap between simple measurements and complex state matching requirements.
Data Source
AI summary
A method includes: generating composite input parameters using first input parameters of a first processing chamber and second input parameters of a second processing chamber; generating third input parameters by adjusting the first input parameters such that the third input parameters more closely approximate the composite input parameters than the first input parameters; generating fourth input parameters by adjusting the second input parameters such that the fourth input parameters more closely approximate the composite input parameters than the second input parameters; causing fabrication of first semiconductors in the first processing chamber in a third process run based on the third input parameters and corresponding to a lower yield loss than the first process run; and causing fabrication of second semiconductors in the second processing chamber in a fourth process run based on the fourth input parameters and corresponding to a lower yield loss than the second process run.