Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
Highly scaled bulk CMOS technologies make use of high energy implants to form the deep retrograde well profile needed for latch-up protection and suppression of lateral punch-through. During the implant process, atoms can scatter laterally from the edge of the photoresist mask and become embedded in the silicon surface in the vicinity of the well edge. The result is a well surface concentration that changes with lateral distance from the mask edge, over the range of 1um or more. This lateral non-uniforimity in well doping causes the MOSFET threshold voltages and other electrical characteristics to vary with the distance of the transitor to the well-edge. This phenomenon is commonly known as the well proximity effect (WPE).
During N-well implantion process, atoms can scatter laterally from the edge of the photoresist mask and become embedded in the silicon surface in the vicinity of the well edge. This will change channel dopping hence the threshold voltage.
ok.
you may be right.
But threshold voltage is a function of depletion region.
Could you please explain how depletion region is going to increase so that threshold voltage variation takes place.?
Also just wondering, but is WPE modeled and included after parasitic extraction in an extracted view? Does the extracted view take into account well distance?
Autodoping is a possibility. Could be implant scatter or could be in-tube during
implant activation if this is done without a capping oxide layer. Dopants will
diffuse laterally and vertically during subsequent thermal processing. This
affects net body doping and can produce a context-dependent VT shift away
from nominal, which is a modeling nightmare.
Also just wondering, but is WPE modeled and included after parasitic extraction in an extracted view? Does the extracted view take into account well distance?
Depending on the kit. In out company kit we have WPE enforced as a DRC rule to make sure you have suitable spacing, so as we have to meet it is not extracted part. it is easier to avoid it.
If you are using a dual well process, then where there isn't Nwell (or Native) you will have Pwell implant.
Near the edges of Pwell you will get wpe effects.
I am still not clear about how nmos is getting affected in well proximity effect. Only the silicon area is exposed where nwell implantation is to be created. Other areas will be covered with photoresist, which will be removed by etching after well implantation is over. So the chance of atoms getting bombarded and hence scattered on the areas where pmos resides is nil.
Please explain more clearly how nmos is getting affected in well proximity effect.
You sure seem certain about stuff you have no direct and
detailed knowledge of. If the foundry sees an effect then
all of your theories are of no consequence.
As just one possibility, consider that (a) PR sidewalls after
development are not dead vertical, and (b) implants often
are shot at tilt-and-rotate planetaries meaning (c) there
is substantial X/Y scatter of implant tails into areas not
limited to the incoming mask geometry. Those scatter-tails
have an extent that has to be respected if you want to
have consistency (even if you purport to not care, the
foundry has an interest in not letting your ignorance hurt
you - or them, when you start whining and lawyering up
about obscure circuit level problems).
Discard your assumption that as-drawn equals as-printed,
siacard your assumption that there is an ideal Z axis shot,
and obey the groundrules. You can argue with us, you
can argue with the foundry, but that all has no useful
outcome.
I am still not clear about how nmos is getting affected in well proximity effect. Only the silicon area is exposed where nwell implantation is to be created. Other areas will be covered with photoresist, which will be removed by etching after well implantation is over. So the chance of atoms getting bombarded and hence scattered on the areas where pmos resides is nil.
Please explain more clearly how nmos is getting affected in well proximity effect.
I had no intention to give an argument. Actually I learned about the fabrication steps, and here seeing some replies I couldn't make a bridge between them. That's why I asked the question.
ok.
you may be right.
But threshold voltage is a function of depletion region.
Could you please explain how depletion region is going to increase so that threshold voltage variation takes place.?
Hi,
the depletion region depends on channel doping.. so threshold voltage is more dominantly depends on doping concentration..
Because the doping level is the determinant of a mos's threshold voltage.. At the time of fabrication MOS is having native threshold voltage... But we use doping to make that voltage according to our specification and requirement..
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
