The characteristics of a thermal detector, such as sensitivity, response time, and saturation power (or energy resolution), are functions of the thermal conductance of the detector to its cryogenic environment. The thermal conductance is specified to achieve a tradeoff among the highest sensitivity, allowed response time, and the desired saturation energy or power budget for the particular application. It is essential to achieve the design thermal conductance (within an acceptable variance) after a thermal detector has been fabricated. Otherwise, the detector will fail to achieve its desired functionality. In addition, the formation of a multi-pixel imaging array becomes difficult and costly when the design thermal conductance is not achieved with high post-fabrication yield.

The control of thermal conductance in prior art is achieved with long and narrow dielectric beams that support the thermal detector. The thermal conductance can be reduced to the desired value by decreasing the width-to-length ratio of the beam or beams under the assumption that diffusive scattering is controlled during processing. The control of thermal conductance in prior art is non-optimal. As the beam width-to-length ratio decreases, the physics of the thermal conductance becomes increasingly sensitive to the exact surface conditions (e.g., roughness) of the dielectric beam used. The lithographic processing of the detector must be controlled to a precision that is difficult and typically not readily achievable in practice. The yield of the thermal conductance can be low because of unpredictability in the resulting lithographically produced surfaces. As a result, an engineering approach is taken where many width-to-length beam ratios are fabricated and tested in order to explore the parameter space. This step is time-consuming and costly in realizing sensors for a multi-pixel imaging array. Beams with lengths greater than 1 mm and widths less than 0.0005 mm have been fabricated as candidates for very sensitive space-based cryogenic thermal detectors using this basic diffusive approach for conductance definition.

This innovation, conductance definition via ballistic thermal transport, targets the design and fabrication of sensitive cryogenic thermal detectors. These detectors are currently the state-of-the-art for the detection of astronomical radiation ranging from x-rays to sub-millimeter wavelength. For a focal plane with thousands of detector units, the conductance, G, must be uniform across the array. With this innovation, the absolute value of G and its variability can be predicted prior to fabrication, and is insensitive to fabrication details. The mode of operation for thermal detector arrays that utilize this innovation is not affected by this implementation choice. The innovation allows greater control over the signal frequency for antenna- and absorber-coupled sensor designs widely employed at microwave through sub-millimeter wavelengths.

This innovation makes use of the low-temperature thermal properties of dielectric films (e.g., micro-machined single-crystal silicon) to control the conductance of thermal detectors to their cryogenic environment via ballistic transport. The thermal conductance is controlled with knowledge of a film’s material properties and geometry in a manner that is insensitive to details of the lithographic processing that may affect the supporting dielectric structures. The thermal conductance is thus defined with high accuracy across a wafer, and from wafer-to-wafer when thermal detectors across several wafers are fabricated. Thermal detectors that are sensitive to x-rays, optical, far-infrared, and sub-millimeter wavelengths can potentially benefit from the underlying innovation, as it enables the formation of multi-pixel imaging arrays that comprise highly uniform detectors across the array, a characteristic that has been difficult to achieve with other methods of thermal conductance control.

This work was done by David T. Chuss, Kevin L. Denis, Samuel H. Moseley, and Edward J. Wollack of Goddard Space Flight Center; and Karwan Rostem of Johns Hopkins University. GSC-16923-1