This Small Business Innovation Research Phase II (SBIR) project will develop a microfabricated subwavelength antireflective structure (SWARS) for use with a MEMS infrared detector to form and infrared camera. The SWARS structure was prototyped in Phase I and shown to allow greater than 90% of incident radiation in the 8-12 ìm portion of the IR spectrum to pass, thus performing better than standard antireflective (AR) multilayer coatings which presently perform this function. These AR coatings are notoriously unreliable, as the thick films tend to delaminate during the processing and packaging of the IR device. In this Phase II project, the SWARS devices will be mated with a thermal light valve (TLV) to make the IR camera.
If successful the proposed approach may be used to produce MEMS devices for a broad range of IR applications, including gas sensors, IR beacons and IR thermographers. Because of their superior transmission properties and the robustness of the design to temperature fluctuations SWARS structures will be used in devices which must operate over a wide range of temperatures, and withstand virtually any operating or processing temperature. Such applications include equipment for factory floor inspections, power grid monitoring, maritime navigation, and security monitoring, in addition to fire fighting and first response.
This Small Business Innovation Research Phase II SBIR project was to manufacture a microfabricated subwavelength antireflective structure (SWARS) for use with MEMS infrared devices. The SWARS structure was prototyped in Phase I, and found to be manufacturable in Phase II. Although the SWARS structure was fabricated successfully and achieved all target metrics, it became clear that the inability of the rest of the target device to withstand the high processing temperatures required for achieving vacuum would limit the commercial success of the technology. We then turned our attention to developing a bonding process for vacuum devices that did not require these high temperatures. The first attempt to lower the bonding temperature but achieve high vacuum used a metal alloy bond. We found that the metal layers outgassed significantly, and low vacuum levels coud not be achieved. We then tried an Au-Au thermocompression bond. This approach yielded very low vacuum levels as shown in the Figure below. As shown above, we have achieved high vacuum (<~ 20mTorr) at temperatures (<~250C) that can be readily integrated into sensitive infrared devices such as focal plane arrays for night vision systems. Certainly yield improvements are warranted, but a major milestone has been reached. Further improvements will be pursued aggressively in the coming months using IR&D funding. As a means to simplify the process for low temperature high vacuum wafer level packaging, IMT has also pursued a getter deposition process where a blanket film of getter material is deposit over the entire wafer. However this simplification necessitates that the getter material passes from inside to outside of each die, raising the question of whether significant diffusion of atmospheric gases laterally through the getter layer can lead to vacuum degradation over extended time periods. However the average vacuum level achieved on this wafer is very poor. This thermistance level indicates a pressure of 700 – 800 mTorr, while our entitlement for this process is < 5mTorr. A likely explanation is that the diffusion of gases through the getter is very rapid when the wafer and getter are hot. Because the bonded pair is unloaded from the bonder at a temperature of ~ 50C, significant diffusion may have raised the die pressure to the 700 mTorr range. Thereafter the wafer stack remained at 20C and the pressure remained stable. In the future this possibility will be more fully studied by thermal cycling the wafers and remeasuring. In conclusion, this simplification remains questionable and will not be integrated into any program until it is better understood.
