How are MEMS sensors manufactured?

High-volume microelectromechanical system (MEMS) sensors are used in almost any application that requires a sensor input. The most common are pressure sensors, accelerometers, gyroscopes, temperature sensors and magnetometers but the technologies used to make these sensors (and actuators) can be applied to many other sensing needs. Two chemical etching processes are the primary techniques to create or fabricate mechanical structures that can move by flexing or rotating.

Bulk micromachining uses a chemical etchant to dissolve silicon in a preferred direction depending on the orientation of the silicon crystal as shown in the left-hand side of Figure 1. In contrast, the right-hand side of Figure 1 shows surface micromachining that uses sacrificial and added structural layers to create small and intricate structures. Bulk micromachining is typically used to make the diaphragms for pressure sensors and surface micromachining is commonly used for the deflecting structures in accelerometers. Surface micromachining is more compatible with the high-volume CMOS processing used to make semiconductors. While these are two common wet etching MEMS processes, there are others that include LIGA, a high aspect ratio micromachining (HARM) technique, that combines X-ray lithography, electroforming and molding processes. Several dry etching methods, micromilling and laser micromachining techniques are also used to make the mechanical structures in sensors.

Figure 1. Bulk micromachining compared to surface micromachining. Source: Understanding Smart Sensors, Third Edition, 2013, Artech House.

Even smaller nanoelectromechanical or NEMS sensors with more exotic structures for more specific applications are being developed in many industry or academic research labs. Atomic force microscope tips are a primary application and carbon nanotubes are among the nanostructures used in NEMS development.