Subject: Re: Step size vs. Beam speed in Raith
From: James Conway <jwc@snf.stanford.edu>
Date: Fri, 01 May 2009 14:00:05 -0700
Fri, 01 May 2009 14:00:05 -0700
Good Afternoon Byoungil Lee

Great Question -- soon to be a new FAQ!

Smaller step sizes will give you an improvement in Line Edge roughness 
for high contrast resist but should not necessarily affect ultimate 
writing resolution.
Yes it will increase the beam speed if the dwell time is the same....
If you need a smaller shot density consider using a smaller write field 
size so you can place more shots within the same area dimension. Shot 
grid, not to be confused with the 1 nm design grid, is the write field 
dimension divided by 2n**16 (Two n to the 16th)

With either higher beam speeds or lower settling time you will start to 
see dynamic effects in your EBL writes. Particularly evident in Y 
deflections EBL writes may appear chirped or distorted compared to the 
pattern design.  For Example, a cross bar target might have little tails 
on the beginning or end of the X or Y cross bar features.  There may 
also be orthogonality distortions evident.

Best beam speed is between 4 and 7 mm per second and you will never see 
any dynamic effects in normal exposures with proper WF alignment.
10 mm/sec is the highest recommended beam speed for quality writes.
10 to 15 mm/sec will likely not show defects in normal writes but more 
difficult patterns such as Single Pixel lines may have some odd 
beginning and end features or breaks.
Once you exceed 25 mm / second it all falls apart and a host of defect 
deflection and exposure artifacts may present themselves.

To decrease your beam speed for the same dose to clear you can increase 
your area step, or reduce your total Beam Current (I sub beam) by 
selecting the next smaller aperture size.



here is one example: 



Inner part of a Global Mark in the demo.csf pattern written in 100 nm 
PMMA with a 0.00156 area step versus a 0.016 um area step.  The beam 
speed was ~ 25 mm/sec.  Intra pattern break evident as well as blocks of 
the pattern data wrote out of normal alignment.  There are NO write 
field breaks in this pattern....

Hope this answers your questions well.

James

Byoungil Lee wrote:
> Hi James,
>  
> I have a question regarding the step size and beam speed for the 
> exposure parameters in Raith.
> I know that smaller step size is better for smaller patterns(better 
> resolution from smaller step size).  But, smaller step size increases 
> the beam speed.
> If the beam speed is too high (higher than 10mm/s), would that cause a 
> problem?  If so, what kind of problem will be caused by high beam speed?
> Also, what would be the optimal step size and beam speed for 40-100nm 
> width lines?
>  
> Thank you!!
>  
> -Byoungil


["multipart/related" not shown]
Good Afternoon Byoungil Lee

Great Question -- soon to be a new FAQ!

Smaller step sizes will give you an improvement in Line Edge roughness for high contrast resist but should not necessarily affect ultimate writing resolution.
Yes it will increase the beam speed if the dwell time is the same....
If you need a smaller shot density consider using a smaller write field size so you can place more shots within the same area dimension. Shot grid, not to be confused with the 1 nm design grid, is the write field dimension divided by 2n**16 (Two n to the 16th)

With either higher beam speeds or lower settling time you will start to see dynamic effects in your EBL writes. Particularly evident in Y deflections EBL writes may appear chirped or distorted compared to the
pattern design.  For Example, a cross bar target might have little tails on the beginning or end of the X or Y cross bar features.  There may also be orthogonality distortions evident.

Best beam speed is between 4 and 7 mm per second and you will never see any dynamic effects in normal exposures with proper WF alignment.
10 mm/sec is the highest recommended beam speed for quality writes.
10 to 15 mm/sec will likely not show defects in normal writes but more difficult patterns such as Single Pixel lines may have some odd beginning and end features or breaks.
Once you exceed 25 mm / second it all falls apart and a host of defect deflection and exposure artifacts may present themselves.

To decrease your beam speed for the same dose to clear you can increase your area step, or reduce your total Beam Current (I sub beam) by selecting the next smaller aperture size.



here is one example: 



Inner part of a Global Mark in the demo.csf pattern written in 100 nm PMMA with a 0.00156 area step versus a 0.016 um area step.  The beam speed was ~ 25 mm/sec.  Intra pattern break evident as well as blocks of the pattern data wrote out of normal alignment.  There are NO write field breaks in this pattern....

Hope this answers your questions well.

James

Byoungil Lee wrote:
Hi James,
 
I have a question regarding the step size and beam speed for the exposure parameters in Raith.
I know that smaller step size is better for smaller patterns(better resolution from smaller step size).  But, smaller step size increases the beam speed.
If the beam speed is too high (higher than 10mm/s), would that cause a problem?  If so, what kind of problem will be caused by high beam speed?
Also, what would be the optimal step size and beam speed for 40-100nm width lines?
 
Thank you!!
 
-Byoungil

["image/bmp" not shown]