i-Micronews: SystemPlus publishes a comparison of camera modules in Samsung smartphones. One of the image sensor cross-section pictures has an interesting diagonal line crossing a photodiode array:
Does SystemPlus mention this as a feature of the photodiodes? While it's not something I've been before (not than I'm a micro-sectioning expert), I find it hard to believe this is anything other than an artifact of the sample preparation. It doesn't make sense as an array feature.
Obviously sample preparation. The cut is not vertical but tilted along crystal orientation, and then crosses the DTI in each pixel at a slightly different location. Remember that the pixel has DTI at each of its 4 sides.
The photodiodes are surprisingly deep (compared to the lateral size). is the motivation for this because the area is so small? so to get the probability up that a photon generates a electron? Or is the motivation more to increase the sensitivity in the nir spectral range?
BTW.. to me it is amazing that it is possible to fabricate such devices. All the technical beauty and so many years of inovation, all the magic in the many tools you need to make it happen (e.g. a litho machine to me is something like a miracle). And in the end you can buy a powerful camera for a few $ packaged inside a portable, powerful computer...
The sample was polished in a plane that is not flush in the vertical or horizontal direction. The obvious aspect is the cut across the pixel from the top to bottom. The diagonal slant is also due to the tilt in the horizontal. The horizontal tilt can be looking at the colour filter array pattern of light and dark, which appears to be, for example, RRGG, from the bright-dark elemental contrast in the SEM image. From the left edge to the right edge of the image, the bright CFA elements in the SEM image become more faint.
Does SystemPlus mention this as a feature of the photodiodes? While it's not something I've been before (not than I'm a micro-sectioning expert), I find it hard to believe this is anything other than an artifact of the sample preparation. It doesn't make sense as an array feature.
ReplyDeleteObviously sample preparation. The cut is not vertical but tilted along crystal orientation, and then crosses the DTI in each pixel at a slightly different location. Remember that the pixel has DTI at each of its 4 sides.
ReplyDeleteThe photodiodes are surprisingly deep (compared to the lateral size). is the motivation for this because the area is so small? so to get the probability up that a photon generates a electron? Or is the motivation more to increase the sensitivity in the nir spectral range?
ReplyDeleteBTW.. to me it is amazing that it is possible to fabricate such devices. All the technical beauty and so many years of inovation, all the magic in the many tools you need to make it happen (e.g. a litho machine to me is something like a miracle). And in the end you can buy a powerful camera for a few $ packaged inside a portable, powerful computer...
The sample was polished in a plane that is not flush in the vertical or horizontal direction. The obvious aspect is the cut across the pixel from the top to bottom. The diagonal slant is also due to the tilt in the horizontal. The horizontal tilt can be looking at the colour filter array pattern of light and dark, which appears to be, for example, RRGG, from the bright-dark elemental contrast in the SEM image. From the left edge to the right edge of the image, the bright CFA elements in the SEM image become more faint.
ReplyDelete