Hi there. I've been working on fabrication of 100nm-width aluminum junctions by E-beam lithography and E-beam deposition.

My questions is that there is always some dark shadows in the EBL-exposed undercut area. (See the 1st figure for SEM image.)I suppose it could be the resist residue from incomplete development. However, as I've been tried to use O2 plasma to remove these shadows, it seems the descum doesn't really help. (see the 2nd figure).

The O2 plasma descum is done right after development and before deposition.

I've been tried many times. Normally the paramter is like 550mtorr (pure O2) , power 7W/18W/30W/50W. Usually it takes 1 minutes but i've tried 5 or 10 minutes also.

I'm pretty sure the plasma does something because when the ashing time is 5 minutes or so, the patterned resist get deformed severely. However, nothing really changes at the dark area.

Any suggestion or comment for this?

Thank you in advance.

P.S.

My recipe is as follows:

Spin coating

1. O2 plasma cleaning 18W, 550mtorr (pure O2) for 1 mins

2. Spin coating cocpolymer(MMA) 8.5 EL6 500

spin 500 rpm for 20s, then 2000 rpm 30s. bake 125 degree C 10min.

3. Spin coating PMMA 950k

500 rpm for 20s, then 5000 rpm 30s. bake 125 degree C 5 min.

E-beam lithography

50kV 100pA ~300uC/cm^2

Development

1. MIBK:IPA=1:3 (one drop on the chip) for 90s

2. immerse in IPA for 40s

3. Dry blow by N2

Descum (as mentioned above.)

E-beam deposition

1. Al 25 degree 0.3nm/s

2. introduce some O2 if oxidation barrier is needed.

3. Al -25 degree 0.3nm/s

Lift off

1. Immerse in Acetone for ~1.5hrs

2. Rinse by Acetone, then rinse by IPA

3. Dry blow by N2