Fabrication of three-dimensional architecture for solid...

Details Fabrication of three-dimensional architecture for solid... Fabrication of three-dimensional architecture for solid...

1999-03-30

2001-03-20

Bowers, Charles (Department: 2817)

Semiconductor device manufacturing: process

Making device or circuit responsive to nonelectrical signal

Responsive to corpuscular radiation

Reexamination Certificate

active

06204087

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates to three-dimensional architecture for solid state devices including radiation detectors, as well as fabricating such devices that are active from chip edge-to-edge, that are operable in an avalanche mode, and that may be protected and/or combined with other chips using vapor deposition bonding.
BACKGROUND OF THE INVENTION
Solid state radiation detectors are known in the art, and provide a useful mechanism for detecting radiation. Such detectors have evolved from detectors using surface barrier electrodes, to ion-implanted electrodes.
Descriptions of such detectors and their uses may be found in applicant's article “A Proposed VLSI Pixel Device for Particle Detection”, Nucl. Instr. and Meth. A275, 494 (1989), A342 59-77 (1994), and in U.S. Pat. Nos. 4,593,381, 5,237,197, 5,355,013, 5,461,653 and 5,465,002, among other references.
FIG. 1
depicts such a prior art radiation detector
10
, which is more fully disclosed in U.S. Pat. No. 5,237,197 (in which applicant herein is co-inventor). In
FIG. 1
, a detector array 10 includes a preferably lightly doped P-type charge depletable substrate
20
, having first and second surfaces
30
and
40
spaced-apart by a substrate thickness L of perhaps a few hundred microns. Substrate thicknesses in this range provide good sensitivity for collecting radiation-generated charge from within the substrate, as well as providing acceptable voltage break-down levels, and protection from radiation damage.
Adjacent the first substrate surface
30
voltage-biasable doped well regions
50
of preferably N-type material are formed. Buffer well region
55
is formed of N-type or P-type material, depending upon the nature of the circuitry
60
in this well region. Well regions
50
,
55
preferably are sufficiently highly doped to act as an electrostatic shield for underlying regions of the detection device. Electronics
60
may be fabricated within buffer well region
55
.
Also adjacent first substrate surface
30
and separated from each other by the N-type well regions
50
are formed spaced-apart collection electrodes
70
, preferably made from highly doped P-type material. Preferably the gate lead of one (or more than one) metal-oxide-semiconductor (“MOS”) transistor
80
is coupled to each collection electrode
70
. The lower surface of the substrate includes a preferably heavily doped N-diffusion region
90
, beneath which is an electrode (not shown), and isolation regions
100
. Of course, the conductivity types of the materials used to form detector
10
could be reversed, e.g., substituting P-type for N-type and vice versa.
One collection electrode
70
, its associated MOS device
80
, and indeed the associated underlying semiconductor structure may collectively be termed a “pixel”, and the terms pixel and detector may be used interchangeably. It is seen from
FIG. 1
that P-type collection electrodes
70
and P-type substrate
20
form a plurality of PN diode junctions with the N-type well regions
50
adjacent the first surface.
In practice, a well bias voltage of many volts is coupled between the collection electrode regions and bottom regions and N-doped well regions. The resultant electric fields extend from the second surface
40
toward and to the first surface
30
. The resultant depletion region extends through the perhaps 300 &mgr;m thickness of the substrate, whereupon a plurality of P-I-N diodes are formed by P-type collection regions
60
, intrinsic substrate region
20
, and N-type region
90
.
The biasing causes force lines to emanate from the N-diffusion region
90
through the substrate thickness and focus upon the P-type collection electrodes
70
. Incoming radiation (not shown) releases charge within the substrate, which charge is focused by the force lines and caused to be collected by the electrodes
70
. As noted, N-wells
50
further serve as a Faraday shield for the array of pixels in structure
10
. As noted, well regions
50
,
55
can also serve as areas in which electronics are fabricated. Unfortunately, CMOS electronics that require wells of both dopant conductivity types can present a problem. Such CMOS electronics can be accommodated in the area over the active detection region, providing wells of like-conductivity type as the collection electrodes are implanted completely within wells of the opposite conductivity type. Understandably, if same-type wells were to be formed directly on the depleted silicon substrate, the wells would collect ionization charge on the substrate, which charges would not be collected and detected by the collection electrodes and assorted circuitry.
As described in the U.S. Pat. No. 5,237,197, detector
10
can nonetheless function reasonably well because the wells surrounding the collection electrodes were doped with opposite type dopant and were back-biased relative to the collection electrodes. This configuration caused electric field lines to be directed to carry one sign of ionization charge from the substrate and the well to the collection electrodes. Like-signal wells, needed for CMOS electronics, were placed along the structure edges, beyond the sensitive detection area. Even though only perhaps 10% of upper surface
30
may be covered by collection electrodes, efficiency in the sensitive region is extremely high with more than 99.99% of the radiation-induced charges being collected by electrodes
70
. The collection electrodes preferably are uniformly distributed in a two-dimensional array on the surface, to provide resultant uniform array sensitivity and spatial resolution.
Once the radiation-induced charge has; been collected by the collection electrodes
70
, the transfer to an associated MOS device(s) can be rapid as the distance is now but a few microns. Further, because there is small capacitance (C) at the MOS gate, the charge (q) developed by the incoming X-ray radiation can produce a substantial voltage signal (v), since v≈q/C. Electronics
60
may be used to signal process the charge associated with the MOS devices. For example the collected charge at a MOS gate may be used to modulate readout current caused to flow through the MOS device. Such readout may be made on an addressable row-column basis.
As noted, radiation detection sensitivity for prior art detector
10
can be very high. But radiation-induced charge cannot be detected until it has been collected by the surface-located collection electrodes
70
. Unfortunately the collection or drift path that released charge must traverse before being collected can be very long, e.g., comparable to the few hundred micron full substrate thickness. Of course should some radiation-generated charge happen to be released closer to the collection electrode surface, collection of the charge can occur in a shorter time. In practice, prior art: detectors using two-dimensional electrodes such as shown in
FIG. 1
, or the common silicon strip technology that preceded what is shown in
FIG. 1
, may take upwards of 25 ns for charge-produced signals to return to a baseline level from a peak. Of course, amplifier delays may extend this time even further.
Attempting to reduce radiation detection time by using a thinner substrate is counter productive because thinner substrates have shorter tracks, and therefor less signal charge. Also, thinner materials can break more readily during fabrication. It would also be desirable to provide a detector structure, that is kept small in size in the presence of radiation damage, requiring a smaller voltage magnitude to achieve depletion, while still preventing so-called bulk type-reversals. Finally, in many detection environments it is necessary to continuously refrigerate the detector, even for maintenance.
U.S. Pat. No. 5,889,313 provided a sensitive solid state radiation detector having such improved detection response times, with good voltage breakdown, without requiring excessively high depletion voltages, while still exhibit good radiation damage resistance characteristics. The resultant detector permitted implementation as a monol

Affiliated with

Also associated with

Rating

Fabrication of three-dimensional architecture for solid... does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Fabrication of three-dimensional architecture for solid..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Fabrication of three-dimensional architecture for solid... will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2501243