Continue to Site

Welcome to EDAboard.com

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.

Register Log in

more via overlap of metal

Status
Not open for further replies.
D

dinesh hegde

Member level 2
Joined
Apr 1, 2009
Messages
42
Helped
4
Reputation
8
Reaction score
3
Trophy points
1,288
Activity points
1,528
Why more via overlap of metal is required for thick metals?
As per my understanding via overlap is required because of mask misalignments, but why more for thick metals?

Regards,
Dinesh
 

Thicker metal suffers from more under-etching.
 

Thanks,
But why thicker metal suffers from under-etching?

Regards,
Dinesh
 

Thick metal is usually etched using wet chemicals rather than Plasma due to the time it would take to plasma etch and the fact that the chemistry used for plasma metal etch is not very selective to the resist mask, so a very thick resist would be required which is not really feasible.
The wet chemical etch etches isotropically so it etches sideways the same rate as vertically. So a 2.6um thick metal feature will shring in size by about 2.6um.
So fabs use a combination of wet and plasma etches to reduce the dimensional loss.
 

Colbhaidh said:
...
The wet chemical etch etches isotropically so it etches sideways the same rate as vertically.
So a 2.6um thick metal feature will shring in size by about 2.6um.
So fabs use a combination of wet and plasma etches to reduce the dimensional loss.
Click to expand...
With pure wet (i.e. isotropic) etch, wouldn't it shrink by 2*2.6µm ?
 

One more observation is that via on wider metal also needs a more via overlap of metal. Is this is to take care of EM?

Regards,
Dinesh
 


But in OPC case, a metal with lesser width should get more affected than metal with more width. Then why more overlap required for higher width metal than lesser width metal?
 

dinesh hegde said:
But in OPC case, a metal with lesser width should get more affected than metal with more width.
Click to expand...
I guess it's the other way round. Otherwise all the semiconductor fabs' DR's would be wrong :cry:
 

erikl said:
dinesh hegde said:
But in OPC case, a metal with lesser width should get more affected than metal with more width.
Click to expand...
I guess it's the other way round. Otherwise all the semiconductor fabs' DR's would be wrong :cry:
Click to expand...

For lesser width metals even a small change makes a lot of % change and hence as Dinesh says the metal with lesser width should get affected more.
 

dinesh hegde said:
But in OPC case, a metal with lesser width should get more affected than metal with more width.
Click to expand...

erikl said:
I guess it's the other way round. Otherwise all the semiconductor fabs' DR's would be wrong :cry:
Click to expand...

sandeep_torgal said:
For lesser width metals even a small change makes a lot of % change and hence as Dinesh says the metal with lesser width should get affected more.
Click to expand...
Guys,

sorry, I think you don't understand the OPC algorithms! sandeep_torgal's statement "even a small change makes a lot of % change" is perfectly correct, and this "% change" is a rather good description of how the OPC method works:
The (possible) cutback of edges is - in first order approximation - proportional to the width of the structures. Do you understand what that means?

Why else would the metal-to-via-overlap have to increase for wider metal wires?
 

Since the metal is thicker and teh etch is isotopic, we need to have larger metal overlap for the via... Just imagine a hole being dug in the earth and it being isotropic, we need larger overlap...

Regards,
Sandeep
 

sandeep_torgal said:
Since the metal is thicker and teh etch is isotopic, ...

Regards,
Sandeep
Click to expand...

Hi Sandeep,
the begin of this discussion dealt with overlap dependency on different metal heights, and this was due to isotropic undercut by wet etching. But now we are discussing overlap dependency on different metal widths at constant height, pls. don't mix-up both!

Cheers, erikl
 

If the metals heights are same and the width varying, i still dont understand how the cutback of edges is proportional to the width of the structures... Let me know if you find any document online...

Regards,
Sandeep
 

sandeep_torgal said:
If the metals heights are same and the width varying, i still dont understand how the cutback of edges is proportional to the width of the structures... Let me know if you find any document online...

Regards,
Sandeep
Click to expand...
I tried to give a short explanation on the effect of OPC above; the algorithms behind it are quite complex, and I can't explain them better, nor do I own documents about OPC. If you're interested enough in it, I suggest you search for papers on OPC and study them.
 

Thanks erikl.
Will check and publish if i get some data to substantiate your point.

Regards,
Sandeep
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top