Researchers at KTH Royal Institute of Technology have developed a 3D-printing technique for specialized manufacturing of microelectromechanical systems (MEMS). The approach offers a cost-efficient way to manufacture custom-designed MEMS devices, in small volumes, that could be used as sensors in robotics, navigation, and more. Although MEMS technologies can be efficiently produced at high volumes using large-scale semiconductor manufacturing techniques, the manufacture of MEMS components in small and medium-size batches is challenging, due to the high startup cost of manufacturing process development and device design optimizations. As a result, engineers often must choose between suboptimal off-the-shelf MEMS devices or economically unviable startup costs, according to professor Frank Niklaus, who led the research. “The costs of manufacturing process development and device design optimizations do not scale down for lower production volumes,” he said. Using a two-photon polymerization process in combination with metal evaporation to form strain gauge transducers, the team demonstrated a 3D-printed functional MEMS accelerometer. It characterized the responsivity, resonance frequency, and stability of the accelerometer over time and confirmed its successful operation. The results suggest that the 3D-printing method could enable the efficient manufacture of a variety of custom-designed MEMS devices. The process of two-photon polymerization produces high-resolution objects that are as small as a few hundred nanometers, but that are not capable of sensing. To form the transducing elements, the researchers used a shadow-masking technique that worked in way similar to a stencil. On the 3D-printed structure, the researchers fabricated features with a T-shaped cross-section that functioned like an umbrella. When they deposited metal from a point above the 3D-printed structure, the sides of the T-shaped features, protected by this “umbrella,” were not coated with the metal. The metal on the top of the T was electrically isolated from the rest of the structure. Using this method, the researchers manufactured about 12 custom-designed MEMS accelerometers in just a few hours using commercial manufacturing tools. This 3D-printing method could be used for prototyping MEMS devices and for manufacturing small- and medium-size batches of a few thousand to tens of thousands of MEMS sensors per year in an economical way. “This is something that has not been possible until now, because the startup costs for manufacturing a MEMS product using conventional semiconductor technology are on the order of hundreds of thousands of dollars and the lead times are several months or more,” Niklaus said. “The new capabilities offered by 3D-printed MEMS could result in a new paradigm in MEMS and sensor manufacturing.” The approach could be used for devices that require costly customization, including accelerometers for aircraft as well as vibration sensors for industrial machinery. It could also be applied to a variety of MEMS sensors, such as pressure sensors, gyroscopes, and flow sensors. Other low-volume products that could benefit from the technique include motion and vibration control units for robots, industrial tools, and wind turbines. A 3D-printed MEMS unit is seen next to a coin. A photonics-based 3D-printing approach for MEMS sensors supports the ability to integrate the sensors efficiently and in a cost-effective way for robotics, navigation, and other applications. The first sensor type that the developing team created was an accelerometer. It used a two-photon polymerization process in combination with metal evaporation to form strain gauge transducers to make the sensor. Courtesy of Simone Pagliano. Moreover, 3D printing could enable complex device geometries for new MEMS sensors that are not currently possible to achieve using conventional silicon micromachining. The strategy used by the researchers to selectively functionalize the surfaces of the 3D-printed MEMS structure by integrating shadow-masking elements in combination with directional material deposition is versatile, and facilitated innovative designs and the integration of a variety of transducer elements. The quick turn-around between the design and the fabrication of small batches of 3D-printed MEMS accelerators allowed the researchers to assess the performances of the devices and optimize them in a matter of a few hours. From an industrial perspective, according to the researchers, this could dramatically reduce the startup cost for manufacturing custom MEMS devices for small- and medium-volume applications, compared to standard microfabrication techniques. “Scalability isn’t just an advantage in MEMS production; it’s a necessity,” Niklaus said. “This method would enable fabrication of many kinds of new, customized devices.” The research was published in Nature: Microsystems & Nanoengineering (www.doi.org/10.1038/s41378-022-00440-9).
Vision Sensor