I have this passage from a Mr Maloberti related to the topic of Layout of Capacitors and I don't know how to interpret it:

The second-order effect must be taken into accout when precise capacitors have to be designed.

  1. fabrication inaccuracies - when this error corrensponds to a gradient , it is first-order effect is cancelled by matched elements that are arranged in a common centroid fashion. Placing the capacitors in parallel guarantees that a gradient in the oxide thickness either in the x or y direction does not affect the capacitor matching.

-what is the gradient? and what is the connection between the oxide thickness and it?

  1. the undercut effect - what is this effect?
  2. fringing effects at the boundary of the structures (the electric field at the end of the structure is not uniform).

-can I see this in a simulation with Cadence?

  • is it important to check the ratio tox/W and tox/L to be smaller than 1/500?

1 Answer 1


The gradient (in this context) is the small but uniform change in oxide thickness across the surface of a wafer. If you have two capacitors formed from the same oxide layer you should lay them out so that their long axes are parallel. This will cause the gradient to affect both capacitors in about the same way, so the ratio of capacitance values is more constant. It's usually the ratio of resistance or capacitance values that matters rather than absolute value.

The "undercut effect" could be referring to how the edge profile of an oxide layer may not be perfectly perpendicular to the wafer surface. This can happen during RIE (reactive ion etch) if the ion source is not directly over the point on the wafer that is being etched.

Fringing effects are due to the finite, non-zero, thickness of structures. Again, it's the edge of the structure that is important. Two metal wires on the same metal layer will have fringing capacitance between their sides (edges) and a parallel plate capacitance to layers above and below. As the spacing between structures becomes small the fringing capacitance can be as significant as the parallel plate capacitance. This is also why you care about the ratio of oxide thickness to length and width...if this ratio is small then the fringing effects become more significant. In general, estimating and modeling the fringing effects is more difficult so we like to create designs that are relatively insensitive to these effects.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.