Cross-talk in integrated-circuit focal-plane arrays of quantum-well photodetectors (QWIPs) equipped with microlenses would be reduced, according to a proposal, by etching deep trenches into the substrates of these devices. The proposal applies, more specifically, to GaAs-based, back-side-illuminated QWIP arrays with microlenses - devices like those described in the preceding article.

The cross-talk problem in such a device without trenches would arise as follows: The microlenses would be formed by patterning the back side of the substrate, as described in the preceding article. The lenses would focus the incident infrared light into and through sub-pixel-size active device (QWIP) areas. Most of the focused light would not be absorbed by the QWIPs, due to lower quantum efficiency, and would, instead, be scattered from patterned reflective surfaces on the front side. A significant portion of the light scattered in each pixel would travel through the unthinned substrate to neighboring pixels, where some of it would be absorbed, thereby giving rise to cross-talk. The cross-talk-reduction problem would thus become one of preventing the scattered infrared light from traveling through the substrate to neighboring pixels.

The problem could not be solved by thinning the entire substrate to the membrane level because such thinning would make it impossible to achieve the required focal length of the microlenses. The proposal would afford the optical advantage of microlenses, without the optical disadvantage of thinning the entire substrate. Instead of thinning the entire substrate, one would etch the substrate only along the boundaries between neighboring pixels; in other words, one would etch deep trenches in the substrate between microlens/pixel units. Such trenches are shown in the figure of the preceding article. Because of the large difference between the indices of refraction of air and the GaAs substrate, the trenches would be highly effective as optically isolating cavities to reduce cross-talk.

This work was done by Sarath Gunapala, Sumith Bandara, and John Liu of Caltech m forNASA's Jet Propulsion Laboratory. In accordance with Public Law 96-517, the contractor has elected to retain title to this invention. Inquiries concerning rights for its commercial use should be addressed to

Technology Reporting OfficeJPLMail Stop 122-1164800 Oak Grove DrivePasadena, CA 91109(818) 354-2240

Refer to NPO-20311