Bound-to-Quasi-Bound QWIPs With Random Reflectors

NASA's Jet Propulsion Laboratory, Pasadena, California

Quantum-well infrared photodetectors (QWIPs) that are designed to exploit transitions between bound and quasi-bound electron quantum states and that incorporate random reflectors are undergoing development. Focal-plane arrays of such detectors are also undergoing development, all as part of a continuing effort to increase the responsivities and decrease the noise levels (dark currents) of infrared-imaging devices operating at wavelengths from about 3 to about 18 µ.

Each of These Random Reflectors was fabricated on one pixel of a focal-plane array of QWIPs. In the one on the left, the minimum feature size is 1.25 µ in the one on the right, the minimum feature size is 0.6 µ.

QWIPs have been discussed in numerous prior articles in NASA Tech Briefs. Two articles with particular relevance to the present devices were "Bound-to-Quasi-Bound Quantum-Well Infrared Photodetectors" (NPO-19633), Vol. 22, No. 9 (September 1998), page 54 and "Demonstration of 15 µ 128 x 128 Quantum Well IR Photodetector Imaging Camera" (NPO-19407) Vol. 20, No. 11 (November 1996), page 30. The first-mentioned article discussed, in some detail, the advantage of designing QWIPs to exploit bound-to-quasi-bound transitions to reduce dark currents below those achievable in QWIPs that exploit bound-to-continuum transitions. The second-mentioned article included a passing mention of the use of random reflectors to increase the efficiency of coupling of light into the QWIPs. In the time since the second-mentioned article, more information on the random reflectors has become available, and is presented below.

The light-coupling problem was discussed in yet another prior article; namely, "Cross-Grating Coupling for Focal-Plane Arrays of QWIPs" (NPO-19657), NASA Tech Briefs, Vol. 22, No. 1 (January 1998), page 6a. To recapitulate: (1) The direction through the thicknesses of the quantum wells is parallel to the focal plane; (2) Quantum selection rules allow the detection of only that part of the incident light that is electrically polarized along the direction through the thicknesses of the quantum wells and thus perpendicular to the focal plane; and (3) The light to be detected is incident along directions approximately perpendicular to the focal plane, and thus only a small fraction of it is electrically polarized along the thicknesses of the quantum wells. Prior to the development of the random reflectors, light-coupling efficiency was increased by illuminating QWIPs via facets inclined 45° to the directions through the thicknesses of their quantum wells. However, the 45° coupling scheme is not suitable for two-dimensional imaging arrays of QWIPs. The random-reflector scheme is suitable for two-dimensional arrays.

Many more passes of infrared light inside a QWIP, with a corresponding increase in responsivity over that achievable with a 45° facet, can be obtained by incorporating a randomly roughened reflecting surface on top of the QWIP. The random structure of the reflector prevents the light from being diffracted perpendicularly backward after the second bounce, as happens in the case of a cross-grating coupling like that discussed in the third-mentioned prior article. After each bounce, light is scattered at a different random angle, and the only chance for light to escape from the detector occurs when it is reflected toward the surface within the critical angle of the perpendicular. For a GaAs/air interface, this angle is about 17°, defining a very narrow escape cone for the trapped light.

The QWIP in each pixel of an array according to the present design concept contains a random reflector (see figure) with scattering surfaces at two levels separated by a quarter of the wavelength of interest in GaAs. The area of the top (unetched) level equals the area of the bottom (etched) level. The combination of equal areas and quarter-wavelength separation maximizes the destructive interference of light reflected from the two levels along the perpendicular, thus limiting the leakage of light through the escape cone. This random reflector structure can be fabricated by use of standard photolithography and selective dry etching with CCl2F2. The advantage of photolithography over a completely random fabrication process is the ability to accurately control the sizes of features to preserve pixel-to-pixel uniformity.

This work was done by Sarath Gunapala, John K. Liu, Mani Sundaram, and Jin S. Park of Caltech for NASA'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