High Resolution, High Fill Factor A-SI:H Sensor Arrays for Optical Imaging
- PDF / 1,233,678 Bytes
- 6 Pages / 384.12 x 617.04 pts Page_size
- 77 Downloads / 161 Views
Standard Pixel Design Phosphor Data=ý ý Data Gate Contact Contact
a
ensor
ESensor
Metal
SData Contact
High Fill Factor Pixel Design Phosphor EV Va
ITO
Sensor Data Contact
Metal Gate Contact
N- Sen so0r uData Contact
Figure 1.Comparison of standard pixel layout [2] (top row) with high fill factor pixel design. Cross sections and top views of the pixel layout are shown.
809
Mat. Res. Soc. Symp. Proc. Vol. 557 © 1999 Materials Research Society
estimated from a given set of design rules using the formula: f (P - g)(P - 92) - ATFT() 2
p based on the pitch p, design rules for lost space between adjacent sensors, g, and g2 (horizontal and vertical), and TFT area AITFT. An illustration of the vanishing fill factor is shown in figure 2, which assumes a 30-40 pm gap between pixels and 200 ptm 2 area for the TFT. Recently a 127 pm pitch imager achieved a 57% fill factor [3] and a 97 pm pitch imager achieved 45% fill factor [4], as expected from these design rules. Our approach for improving the fill factor is to extend the sensor into a continuous layer on the top of the array. Such a 3-dimensional structure is shown in the second row of figure 1. This pixel is expected to have light sensitivity across the entire surface. Previous work towards achieving this structure includes using a continuous n', i, and p+ contact [5]. This structure is observed to have substantial blooming, with charge spreading to adjacent pixels in a fraction of a
second. The spreading is modeled as a resistive coupling through the continuous bottom p' contact. In another approach [6], a 3-dimensional sensor structure is fabricated with the sensor on top of the addressing circuitry. The sensor is then patterned with grooves parallel to data and gate lines, forming mesa-isolated structures. While this approach eliminates the conduction between pixels, it complicates the processing by requiting additional passivation and metal layers, and a bias connection must cross the sensor. Fill factor is again limited by design rules for the etch process and bias connection width, but improves on the standard mesa-isolated design. With this paper. we have adopted a patterned bottom metal contact and n' layer. The sensor then uses continuous i and p+ layers to form the sensor, followed by an ITO (indium tin oxide, a transparent conductor) film for applying the bias voltage. The equivalent circuit for the pixel and the readout electronics are unchanged. The fabrication process is modified to include a passivation layer to isolate the data lines from the sensor contact. The process is designed for compatibility with large-area processing equipment. In this approach, the pixel fill factor will be at least the fill factor of the n' contact. Depending on biasing and a-Si:H quality, considerable charge should also be collected from the gaps between n+ contacts. The effective fill factor is expected to be somewhere between the dotted and dashed-dotted lines of figure 2, for typical design rules expected in the large-area process. The behavior of the sensor
Data Loading...
