Albert Theuwissen Reports from IEDM 2011

Albert Theuwissen published a very interesting review of IEDM 2011 papers:

1. Extremely-low noise CMOS image sensor with high saturation capacity, by K. Itonaga (Sony).
2. High performance 300 mm backside illumination technology for continuous pixel shrinkage, by D. Yaung (TSMC).
   "QE values for a 0.9 um pixel were shown : 50 % in blue, 47 % in green and 45 % in red. The pixels were realized in a 65 nm process with a remaining thickness of the silicon equal to 2 um … 4 um. In the case of the 0.9 um pixel, the optical cross-talk is about 4 times as large as in the 1.1 um version."
3. A 1.4 um front-side illuminate image sensor with novel light-guiding structure consisting of stacked lightpipes, by H. Watanabe (Panasonic).
   "QE in green 74 % in comparison with 69 % for the BSI and 43 % for the FSI without stacked lightpipe."
4. Investigation of dark current random telegraph signal in pinned photodiode CMOS image sensors, by V. Goiffon (ISAE).
5. A CMOS compatible Ge-on-Si APD operation in proportional and Geiger modes at infrared wavelengths, by. A. Sammak (TU Delft).
6. Enhanced angle sensitive pixels for light field imaging, by S. Sivaramakrishnan (Cornell University).
7. A 192×108 pixel ToF-3D image sensor with single-tap concentric-gate demodulation pixels in 0.13 um technology, by T.-Y. Lee (Samsung).

Update: Moving EF's comments to the front page:

1. The Sony paper was an embarrassment for Sony. The presenter who claimed significant noise reduction and increased saturation charge could not answer the question as to what the actual value of noise and saturation signal was, saying he was not knowledgeable about the details. This is shameful in a conference like IEDM.

2. TSMC said that they had a lot of particulate problems in wafer bonding leading many manufacturing issues, including wafer distortion during alignment and "breaking bubbles" etc. They said they were able to now reduce particulates to a lower level. This speaker was good about answering questions so a plus for TSMC. He also called the process "TSMC BSI" and not by his co-authors' company Omnivision. Interesting.

3. A very nicely presented paper although the 74% QE number is somewhat hard to accept. If true, it is remarkable. Reminds me that Aptina has also said lightpipes with FSI makes BSI less necessary. Watanabe says at 1.1 um, FSI might be comparable to BSI using the lightpipe technology. He is not sure about 0.9 um. Personally this is on own my short list for WKA nominations.

4. Interesting investigation of blinking pixels. Location of the traps were not determined but there was a lot of interesting statistical data presented.

5. Very interesting results. Too bad we cannot really work on this in the open in the US.

6. Enthusiastically presented student paper.

7. Well, I found this paper interesting, in as much as it is my device but with a life of its own at Samsung. When AT says "pretty good" performance, I think he means "pretty much SOA" performance. Note also that normal two-tap operation throws away 50% of the light (because you also need the quadrature signals) so losing 75% in single-tap is not as bad as it sounds. It is only a ~30% loss in SNR compared to other techniques, but the reduction in FPN makes up for the SNR loss when it comes to determining depth accuracy.