MEMS Chamber Pressure Control via Segmented Wafer Sealing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for manufacturing microelectromechanical devices with different operating pressures result in unpredictable pressure levels and chamber dimensions, leading to performance issues and inefficiencies due to premature getter activation and gas saturation.
Innovation Solution
A process involving selective etching of trenches and deposition of anchoring layers to form sealed chambers with controlled pressures, using PVD or PECVD for precise sealing without getters, ensuring accurate pressure control in each chamber.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If getter material is used to reduce chamber pressure, then low pressure is achieved in the chamber, but the getter may saturate and fail to achieve the desired pressure level
Solution Approach 1:
The patent divides the single chamber into multiple sealed chambers, each with its own pressure control system. This allows each chamber to be optimized for specific pressure requirements without relying on a single large getter that may saturate. The segmentation enables independent pressure management for different device types (gyroscopes requiring low pressure, accelerometers requiring higher pressure).
Solution Approach 2:
Different chambers are assigned different pressure levels according to the specific requirements of the devices they contain. Gyroscope chambers are sealed at low pressure (0.1-10 mbar) while accelerometer chambers are sealed at higher pressure (1-1000 mbar). This local optimization ensures each device operates in its optimal pressure environment.
2Quantity of substance
If sufficient getter material is included to prevent saturation, then pressure control is improved, but chamber dimensions must be increased
Solution Approach 1:
Instead of using one large chamber with excessive getter material, the patent segments the volume into multiple smaller chambers. Each chamber contains only the amount of getter needed for its specific volume and pressure requirements, eliminating the need for oversized chambers while preventing getter saturation.
Solution Approach 2:
The patent changes the pressure parameter for different chambers based on device requirements. By sealing chambers at different pressure levels (low pressure for gyroscopes, higher pressure for accelerometers), the system optimizes getter usage in each chamber without requiring uniform over-provisioning across all chambers.
3Reliability
If getter is activated during wafer bonding, then sealing is achieved, but the getter may absorb gas from higher-pressure chambers causing pressure degradation
Solution Approach 1:
The patent uses segmentation to isolate pressure systems. Each chamber is independently sealed with its own adhesion layer, preventing cross-contamination of gases between chambers. This eliminates the cross-talk problem where activated getter in one chamber would absorb gases from other chambers with different pressure requirements.
Solution Approach 2:
The adhesion layer is deposited and activated during the wafer bonding process itself, creating the seal before the chambers are fully enclosed. This preliminary sealing action ensures that each chamber maintains its intended pressure atmosphere without subsequent contamination from getter activation or gas exchange between chambers.
4Ease of manufacture
If trade-off pressure level is used in a single chamber, then device integration is simplified, but performance of individual devices deteriorates
Solution Approach 1:
The patent maintains manufacturing simplicity through wafer-level processing while segmenting the final product into multiple pressure-optimized chambers. Each chamber is formed and sealed on the wafer using standard semiconductor fabrication techniques, then the wafer is diced into individual devices. This approach preserves manufacturing ease while achieving optimal pressure conditions for each device type.
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
Achieves precise and controlled pressure levels in separate chambers, reducing chamber dimensions and eliminating early getter activation risks, enhancing device performance and miniaturization without increasing complexity or cost.
Implementation Method 1
a) depositing an adhesion layer on a microstructure wafer in a physical vapour deposition chamber at the first pressure
Implementation Method 2
b) depositing an adhesion layer on a microstructure wafer in a plasma enhanced chemical vapour deposition chamber at the second pressure
Implementation Method 3
sealing chambers at different pressures using PVD or PECVD for precise sealing without getters
Implementation Method 4
sealing chambers at different pressures using PVD or PECVD for precise sealing without getters
Data Source
Figure 1
Figure 2~4
Figure 5~6
AI summary
A process for manufacturing microelectromechanical devices includes forming a dielectric layer (3) and a structural layer (21) on a substrate (2) of a first semiconductor wafer (1) and forming a first and a second microelectromechanical device (22, 23) in the structural layer (21). The first and second microelectromechanical devices (22, 23) are sealed respectively in a first chamber (32) and in a second chamber (33) at a first pressure (P1). The first chamber (32) is fluidically coupled to an external environment through the substrate (2) and sealed at a second pressure (P2) different from the first pressure (P1). To fluidically couple the first chamber (32) to the outside, there are formed a stop layer (7) between the dielectric layer (3) and the structural layer (21) and a cavity (25) fluidically coupled to the first chamber (32) in the dielectric layer (3). A channel (35) is formed by etching the substrate (2) in a position corresponding to the cavity (25) and the stop layer (7), and the etching of the substrate (2) is ended against the stop layer (7).