[SOLVED] Fabrication : Contact for Poly and Active

[Prashanthanilm]

We use CONTACT to connect Poly to Metal and Diffusion to Metal.
CONTACT width and Height is same for a process.

On the Fabrication side, given that the Diffusion and Gate will not be on same horizontal axis (Correct me if am wrong), how does the same CONTACT usage make the Metal1 to reach Diffusion/Gate respectively.

 

[Serfacy]

Simplifying, the contact can be treated as a hole in the thick dielectric (e.g. di-oxide). During depisiting metal1 layer, the atoms of this metal goes also into such holes and make contact from gate or drain/source area to this metal. Thus contact is the same metal. So it is not difference if underlying layour is gate or diffusion.

 

[Prashanthanilm]

Understand.

We have the CONTACT size fixed (ie.. Hole Size is Fixed).
Also, Gate and Diffusion are not on the same Horizontal Plane during Fabrication.

Now, Metal have to be connected to Poly and Diffusion.
From Fabrication, The Depth of Hole required for Metal to Poly Connection and Metal to Diffusion is different.
So, how does the same CONTACT size help here?

Let me know if the Query is understood.

 

[Serfacy]

The same contact size is defined by dielectric etching process. This process requires specific size in X and Y directions in order to ensure proper quality of that "hole".
The depth of this "hole" rather doesn't matter because etching process has to remove dielectric only. It is chemical or chemical and physical process. And it does not affect underlying layers.

 

[Prashanthanilm]

The depth of this "hole" rather doesn't matter because etching process has to remove dielectric only. It is chemical or chemical and physical process. And it does not affect underlying layers.

I assumed the same as the Depth does not matter of the Contact Hole. But I did not have any evidence.
Can you explain this in bit detail? Any reference would be appreciated.

On your explanation, Diffusion is in bit lower Horizontal plane compared to Poly.
Meaning, Metal to Poly is through CONTACT. Whereas Metal to Diffusion is through (CONTACT + dielectric space which is etched)

Please correct me if I am wrong.

 

[Serfacy]

The contact is the hole etched in the dielectric. So in both cases (poly and diffusion contacts) the process looks the same. In fact the depth of diffusion contact is greater.
But you cannot say "Metal to Poly is through CONTACT. Whereas Metal to Diffusion is through (CONTACT + dielectric space which is etched)" because your CONTACT is
the same etched dielectric. It only has different depth.
You can see in:
S. A. Campbell, "The Science and Engineering of Microelectronic Fabrication"
P. Van Zant, "Microchip Fabrication"

 

[Prashanthanilm]

I understand now. Thank you.

Can you please explain this statement :
"In fact the depth of diffusion contact is greater." . How and Why?
As the Di-electic width will be same over Poly and Diffusion.

 

[Serfacy]

After depositing or grow of dielectrc layer (in fact any layer), this layer is polished. So we can assume that the height (what I called depth) from the top surface of dielectric to the polisilicon layer is smaller than to the diffusion layer. As you noticed "Gate and Diffusion are not on the same Horizontal Plane during Fabrication." and "Diffusion is in bit lower Horizontal plane compared to Poly."

 

[dick_freebird]

It's the process development engineer's job to come up
with an etch recipe that tolerates the different depth
and depth-of-field issues, and stops clean on both
targets (silicided poly and silicided active). It's not a
trivial thing. But you get to just draw the polygons if
they get it right (and if they don't, you won't see the
flow offered).