The objective of this innovation was to develop a methodology of fabricating thermopile detectors using standard semiconductor fabrication techniques. The goal was to develop a fabrication process that minimized the roughening of the Si legs during patterning of the metallic couples, and to enable delineation of the Si legs without the use of highly toxic or carcinogenic chemicals. Another key requirement was at least 50% optical absorbance across the spectral band.
Prior techniques to pattern metallic thin films on thin Si membranes typically involved etchants that roughen the Si itself. Furthermore, most methods used to delineate Si legs involved the use of potassium hydroxide, which is highly toxic, or trichloroethylene, which is carcinogenic. Typical means to achieve high absorption include the use of Bi thin film, which has poor adhesion to Si, and gold black, which is very difficult to delineate.
This innovation involves a fabrication methodology for realizing a silicon-leg isolated thermopile detector. The detector consists of one or more sets of Bi-Cr couples. The detector is designed to operate between 170 and 300 K in the 14-to-400-micron spectral band, and may be used for thermal mapping of outer planet targets (e.g., Jupiter and its moons). Functional operation of the process involved performing the actual fabrication inside a Class 100 cleanroom.
Alternate embodiments of the innovation would include the use of thermopile materials different from the Bi and Cr used; this is a generic process that can be used for a wide variety of different thermopiles and other thermal detectors. This process resulted in the development of TiN thin film absorbers, which have been demonstrated to have >50% absorption over a 14-to-400-micron spectral range; the development of a TiN etching process that does not etch the Si membrane; and the development of a Si leg delineation process, which does not involve the use of highly toxic or carcinogenic chemicals.
The highly absorbent thin film can be lithographically defined, and is thermally and mechanically robust. Analysis of the innovation’s capabilities was evaluated by considering the device yield. A device yield >99% was achieved using this fabrication process.