US 6696314

Based on the authoritative patent text provided and the available search results, here is my summary. Note: my searches did not return any CAFC 2026 docket entry or active litigation referencing this patent number; the only direct hit was the Google Patents record (mirroring the text you provided) and USPTO PTACTS petition documents where the number appears incidentally in search histories for other matters, not as a litigated patent.

US Patent 6,696,314 B2 — Summary

Title: CMOS imager and method of formation
Patent/Publication No.: US6696314B2 (application No. 10/404,147; pre-grant publication US20030176009A1, published 2003-09-18)
Inventor: Howard E. Rhodes
Original Assignee: Micron Technology, Inc.
Current Assignee (as listed by Google Patents): Aptina Imaging Corporation (assignment from Micron to Aptina recorded 2016-12-06, effective 2008-09-26; inventor-to-Micron assignment recorded 2016-11-09, effective 2001-08-28)
Filing Date: April 2, 2003 (divisional of U.S. Application 10/073,136, filed Feb. 13, 2002, which is a divisional of U.S. Application 09/941,556, filed Aug. 30, 2001, now US 6,504,196 B1)
Priority Date: August 30, 2001
Issue Date: February 24, 2004
Legal Status: Expired – Lifetime (anticipated expiration date listed as 2021-08-30; maintenance fees paid through year 12 in 2015)

Abstract: "A CMOS imager having an epitaxial layer formed below pixel sensor cells is disclosed. An epitaxial layer is formed between a semiconductor substrate and a photosensitive region to improve the cross-talk between pixel cells. The thickness of the epitaxial layer is optimized so that the collection of signal carriers by the photosensitive region is maximized."

Technology in brief: The patent addresses quantum-efficiency loss and cross-talk in CMOS imagers caused by photogenerated carriers diffusing out of the photosite into the substrate. The solution is to grow a doped epitaxial layer (typically p-type, about 10,000–150,000 Å, preferably ~60,000 Å) between the substrate and the pixel well. The increased dopant concentration at the epitaxial-layer/substrate interface creates an electric field that reflects signal carriers back toward the photosite. The layer thickness can be tailored by color wavelength (e.g., ~80,000 Å for red, ~10,000 Å for blue) to maximize collection and minimize cross-talk.

Plain-Language Overview of the Independent Claims

The patent has 25 claims; the independent claims are 1, 13, and 21 (all method claims).

Claim 1 – Method of forming a pixel cell. Covers forming a pixel cell by: (a) growing an epitaxial layer of a first conductivity type (e.g., p-type) over a substrate; (b) forming a retrograde well of the first conductivity type in that epitaxial layer (dopant concentration lowest at the surface, highest at the well bottom); (c) forming a photosensitive region in the well; (d) forming a photosensor (e.g., photogate) on the upper surface of the photosensitive region to control charge collection; and (e) forming a floating diffusion region of the opposite (second) conductivity type in the well to receive transferred charge.

Claim 13 – Method of forming a pixel array. Covers forming an array by: (a) forming an epitaxial layer of a first conductivity type over a substrate; (b) forming a well of the first conductivity type in the epitaxial layer; and (c) forming a plurality of pixel sensor cells in the well, where each cell has a photosensitive region, a photosensor on it for sensing charges of a particular color wavelength, and a floating diffusion region of a second conductivity type. This is the array-level analog of claim 1, with a color-wavelength-specific photosensor per cell.

Claim 21 – Method of forming a color pixel cell. Covers forming a color pixel cell by: (a) forming at least three epitaxial layers of a first conductivity type in a substrate, where the layers have different thicknesses from one another; (b) forming a photosensitive region in each of the epitaxial layers; (c) forming a photosensor on each photosensitive region for receiving charges of a particular color wavelength; and (d) forming a floating diffusion region of the second conductivity type in each layer. The differently-thick epitaxial layers are tuned to different wavelengths (red/green/blue), each with its own reflective barrier and readout node.

Dependent-claim highlights: Claims 2–6 and 15–19 specify thermal growth at ~900–1200° C and thickness ranges (10,000–150,000 Å; preferably 30,000–100,000 Å; most preferably ~60,000 Å). Claims 7–8, 14, 22–23 specify p-type/n-type conductivity combinations. Claims 9 and 20 specify ion implantation for the well. Claims 10–11 and 24 add a transfer gate (photogate embodiment). Claims 12 and 25 add a reset transistor whose source is the floating diffusion region.

Uncertainty Notes

• CAFC 2026 docket: I found no evidence of a 2026 Federal Circuit appeal or any active litigation involving US 6,696,314. The patent is shown as expired (term ended ~2021), so post-expiry litigation is unlikely. If you need definitive confirmation, the USPTO's Patent Center / PTAB and CAFC PACER docket searches would be the authoritative sources.
• Assignee: Google Patents' "current assignee" field (Aptina Imaging Corp) is flagged as potentially inaccurate; the recorded assignments confirm Micron → Aptina.