Plasma Stage Zone Model Generation for Temperature Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In plasma processing, such as plasma etching, the temperature of substrates like semiconductor wafers can exceed desired levels due to heat introduced from the plasma, making it challenging to control temperature distribution effectively using conventional heater systems.
Innovation Solution
A model generation apparatus that acquires measurement data from a plasma processing apparatus with a stage divided into zones, each with a heater, and uses this data to generate a prediction model that accounts for heat transfer between zones and the impact of power factor changes, allowing for precise control of temperature distribution by adjusting the power flowing to each heater zone.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Temperature
If conventional heater systems are used to control substrate temperature in plasma processing, then temperature distribution can be controlled to some extent, but the substrate temperature exceeds desired levels due to heat introduced from the plasma
Solution Approach 1:
The stage surface is divided into multiple temperature zones (e.g., 9 zones) with independent heaters for each zone. This segmentation allows localized temperature control to compensate for non-uniform heat distribution from plasma, enabling precise control of substrate temperature in different regions despite plasma-induced heating.
Solution Approach 2:
Each zone on the stage has independent temperature control capabilities with dedicated heaters and temperature sensors. This local quality approach allows different regions to be controlled at different temperatures, matching the non-uniform heat input from plasma and achieving uniform substrate temperature distribution across the entire substrate surface.
2Measurement precision
If temperature control maps are generated by measuring temperature distribution under multiple process conditions, then accurate temperature control is achieved, but the time required for map generation becomes excessively long
Solution Approach 1:
Temperature control maps are pre-calculated and stored in a database for various process conditions before actual plasma processing. During plasma processing, the system simply retrieves the appropriate pre-calculated map data based on current process parameters, avoiding the need to perform time-consuming measurements and calculations in real-time, thus reducing map generation time from days to hours.
Solution Approach 2:
Instead of performing actual temperature measurements under multiple process conditions to generate control maps, the system uses simulation or pre-measured data to create representative map data that copies the essential temperature distribution characteristics. This allows accurate temperature control without repeating lengthy measurement procedures for each new process condition.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach reduces the time required to generate temperature control maps, from days to hours, by reducing the number of measurement patterns needed, enabling more efficient and accurate temperature control during plasma processing.
Implementation Method 1
a heater capable of independently controlling the temperature of each zone is embedded in the corresponding zone
Implementation Method 2
a temperature regulator provided in the stage to regulate a temperature of the entire stage
Implementation Method 3
assuming that heat with a heat quantity proportional to a temperature difference between adjacent zones moves between the zones, heat with a heat quantity proportional to a temperature difference between the temperature regulator and each zone moves between the temperature regulator and the zone
Data Source
AI summary
When a plasma processing apparatus changes processing parameters of a plasma processing that include at least a temperature of a stage and a temperature of each zone obtained by dividing a placing surface of the stage into multiple patterns, and measures the temperature of each zone and a supply current to a hater in a state where the temperature is stabilized, an acquisition unit acquires the measurement data. The generator generates a prediction model using the measurement data, assuming that heat with heat quantity proportional to a temperature difference between adjacent zones moves therebetween, heat with heat quantity proportional to a temperature difference between the stage and each zone moves therebetween, heat with heat quantity calculated from the supply current to the heater of each zone is input to the zone, and quantity of heat input and quantity of heat output in each zone are consistent.