Content area
Full Text
MEMS sensor technology. Professor Hejun Du
The origins of what we now know as micro-electromechanical system (MEMS) technology can arguably be traced back to 1 April 1954, when a paper by [6] Smith (1954), then at the Bell Telephone Laboratories, was published in Physical Review . This described for the first time certain stress-sensitive effects in silicon and germanium termed piezoresistance. During the mid-1950s, researchers were starting to investigate whether the same technologies that had yielded the transistor, which subsequently revolutionised the fledgling electronics industry, could be applied to sensors. Might not the bulky electromechanical sensors of the day be replaced by small, rugged devices in the same way that the transistor had replaced the thermionic valve? Smith's paper was followed a year later by what is probably the first publication to consider this possibility ([4] Paul and Pearson, 1955) and during the early 1960s, a series of papers from the Honeywell Research Centre and the Bell Labs described the first silicon diaphragm pressure sensors and strain gauges ([5] Pfann and Thurston, 1961; [8] Tufte et al. , 1962). Interest in silicon sensor technology grew dramatically and by the late 1960s a number of pioneering American companies had commercialised the first silicon pressure sensors. These were crude by today's standards (Figure 1 [Figure omitted. See Article Image.]) but in the early 1970s developments in micromachining, as it was then called and improvements to silicon processing led to pressure sensors with non-planar diaphragm geometries which yielded superior performance. These were arguably the first true MEMS sensors.
What exactly do we mean by MEMS? The term is something of a misnomer, as not all so-called MEMS devices are "electromechanical" and few are "systems". Nevertheless, the term is now widely applied to all manner of miniaturised devices, generally 3D microstructures of one sort or another, mostly fabricated from silicon and using techniques which are often derived from the microelectronics industry. These include isotropic and anisotropic etching ("microengineering"), various thin film deposition methods, anodic bonding and the well-known masking and doping techniques employed in IC manufacture. Today, silicon microsensors, "lab-on-a-chip" and micro-TAS (micro-total analytical systems) devices are all referred to as MEMS.
So what progress has been made during the half century that has passed since Smith's ground-breaking paper?...