Controlling Pixel Size Improves Infrared Camera Technology
Article Posted On 10/24/2010
For decades, Judson Technologies, a small company located near Lansdale, PA, has been a leading commercial supplier of a wide range of discrete infrared detectors. With the acquisition by Teledyne Technologies at the beginning of 2008, Teledyne Judson Technologies is embarking on an ambitious plan to expand its personnel and its facilities in Pennsylvania and develop high-resolution infrared cameras for both commercial and defense applications, such as residential energy auditors and night-vision goggles.
With partial funding support by PITA, Lehigh University professor James Hwang is leading a project focused on the development of a 1280x1024 high-resolution detector array with 10-micron or smaller pixels for these infrared cameras. The detector material is based on InGaAs, which is a compound semiconductor sensitive to infrared wavelengths between 0.9 to 1.7 microns. InGaAs - grown by IQE, a multinational corporation with a major operation in Bethlehem, PA - is fabricated into detector arrays by Judson. Lehigh has been helping to design, model and characterize the arrays, and together, Lehigh, Judson and IQE will be optimizing the starting material and fabrication process for the best camera performance.
One of the recent achievements of the project is the development of a novel yet convenient technique to precisely monitor and control the pixel size of the detector array. As the pixel size decreases below 10 ?m, the difference between the photolithography mask size and the actual pixel size becomes critical. In InGaAs detectors, this difference is mainly caused by lateral diffusion of the zinc dopants. While the vertical diffusion depth can be inferred from secondary-ion mass spectroscopy, the lateral diffusion width has been difficult to determine. The novel technique developed with this project is based on probing individual pixels in an array and analyzing the parasitic bipolar transistor formed between them to determine the lateral diffusion width, hence, the actual pixel size. The actual pixel size can then be used to separate the area-dependent dark current from the periphery-dependent dark current. The result shows that in the present arrays, the dark current is dominated by the periphery current, unless the pixels are so close that they effectively touch each other. This electrical measurement technique can be conveniently implemented during fabrication to monitor and control the pixel size.