2014-10

2014-10Edit

Mercredi 22/10

BPC2v4_1 + BPC2v4_2

EBL: Using 30kV and Spot 1 to get around 23 pA. Area step size=4 nm, area dose=300 uC/cm^2

dev: MIBK/IPA=90s, IPA=30s, ODI=15s, MIF726=60s, ODI=60s, Ethanol=15s, N2 dry.

BPC2v4_1  the underctus were not visble, decided to put it again in developper, but for too long. All structures collapsed -> reprocess: spin only PMMA A6

BPC2v4_2 As well no undercut visible so decided to use O2 ashing for 10 sec, 0.2mbar O2, 100 W.
+ MIF726=15s, ODI=60s, Ethanol=15s, N2 dry -> undercut OK

Evap BPC2v4_2:  Ar milling (3mA, 505V), 3.0x10^-7 mbar, -90deg, Ti=30nm @ ? at 0deg

Lift-off BPC2v4_2: 30min in PGRemover, 60 C. worked very well. ODI and blow N2 dry.

Obs BPC2v4_2: Looks fine and all marks are visible:

Jeudi 23/10

BPC2v4_2 step 3, spinning: 

spin TI prime @ 4000rpm 60"
bake 120°C for 1'
spin PMGI SF8 (batch 12060407, exp. 7/1/2013) @ 3000rpm for 60"
bake hot plate setpoint 175°C, 5', under beaker
spin PMMA A6 (from Youri bottle) @ 6000rpm for 60"
bake hot plate setpoint 175°C, 15', under beaker

EBL: 30keV
Alignement is very difficult because bilayer too thick
one can see only the largest marks (on the working area of CPB). However, luckily, 1st alignement was performed only with those marks, so it appears the patterns are well aligned

Evap:

Ar milling (3mA, 505V) 2x10",
Al 15nm @ 1nm/s, 0° (very bad vacuum , 7e-7mb !!!)
Ox 800mb, stay 5' (then pump down to 30mb in 1')
Ti=100nm
Al 60nm @ 1nm/s, +30° (vacuum still bad, ~2e-7)

vendredi 24/10

Lift OK
Patterns are awful! The mask was deformed, or the e-beam was too unstable (problems with e-beam during the last exposures). NB: during observation, we notice the edge of the trilayer was damaged (see SEM and optical)

-> reprocess. Remove Al with MIF726 1'.

BPC2v4_2 step 3, spinning (this time uses MAA EL10 + PMMA A6): 

spin MAA EL10 (batch 14020103, exp. 3/1/2015) @ 2000rpm for 60"
bake hot plate setpoint 180°C, 5', under beaker
spin PMMA A6 (batch 14020103, exp. 3/1/2015) @ 6000rpm for 60"
bake hot plate setpoint 180°C, 17', under beaker

BPC2v5_1
Design avec capas d'overlap
Step 2: gates
EBL 30keV, standard PMGI PMMA development, Ar mill 20" 5mA 500V, evap Al 30nm @ 1nm/s P~7e-7mb

Oxidation development:

Following F. Nguyen  thesis, Al oxidation in RIE, 50cc O2, 10µb, 260W, >1'
+ have to try it before lift-off, otherwise creates a 2DEG in Si (reason why it was abandonned in FN thesis)

witness wafer with same Al evap (wafer standard, P doped Si [100] 1-10ohm.cm + Si thermal 100nm)
Cut in 4 parts
Measure ellipsometer (model Al2O3 -fit param 20nm, +/-99%-, Al 30nm -fit param, +/-30%-, SiO2 100nm, Si):
find between 2 and 9nm depending on how much one let the Al thickness be varied. Uniformity is usually more or less equivalent to thickness. GOF is around 85%

sample 1: RIE on hot Al rod (oven 220°C), 10cc O2, 10µb, 260W (V = 460V) 1'

Now measure 14nm +/-4nm, but the results depend a lot on the fit constraints.
Method not really appropriate.
Maybe we can use the laser in-situ to monitor the growth of Al2O3 ??

sample 2: use laser but there is not much of a signal. 2'30 same conditions
ellispometer: 19nm +/- 7nm.... well, could be (at least the change is significant with same fit procedure)

lundi 27/10

Put sample 1 with 1' ox + piece with natural ox in MIF319.
From the beginning, bubbles appear on the no ox. At 2'30, all Al disappeared, leave up to 3'.
The sample 1 with 1' ox has nothing, no bubbles even clos to corners (visually OK -> observe it in SEM)

Now oxidize the true sample BPC2v5_1 with Al rod @ 220°C, and RIE 50cc O2, 10µb, 260W, 1'30
Optically looks as if the resist below has been damaged! (granular)

Lift-off warm aceton 15' + low power 2" US -> OK, except 1 pad which remains on the side of the sample
Followed by 1h remover PG (lunch time) -> OK

step3, spin BPC2v5_1

- spin MAA8.5 EL10 (batch 14020103, exp. 3/1/2015) @ 2000rpm for 45"
NB: vivien recipe works very well for edge removal on small sample:
5" 500rpm 200rpm/s + 45" 2000rpm 1000rpm/s + 10" 8000rpm 4000rpm/s 
- bake setpoint 180°C, 6'
- spin PMMA A6 (batch 14020103, exp. 3/1/2015) @ 4000rpm for 45" (same trick to remove edges)
 NB: spin speed should be 6000 to comply with previous tests on whole wafers
 but not calibrated for small samples with edge removal
- bake hot plate setpoint 180°C, 17'

ellipsometer: 660nm MAA, 320nm PMMA
(Compared to previous results (2013-07): 500nm MAA, 270 PMMA)

step3, expo BPC2v5_1
30keV, 300µC/cm², spot 1 30pA, align on the coarse and fine optical marks but not on Al marks from step 2 (cannot see them).
area step 4nm, dwell 
undercut area step 8nm, dwell 

dev: 1'10 MIBK/IPA + 1' stop IPA
obs: looks well aligned but island and ground patterns might be not separated (?)

old canon

- 30nm Al @ 1nm/s, P_ev = 7e-7 !!!!!
- oxidation, 118mb, 9'40 then pump (@10' is down to 20mb @12' open valve)
- ~1nmAl @ 0.2nm/s (target 0.5nm)
Does not have a program for 0.1nm/s. Launch before rate 0.2 is achieved, to be more precise
But there is a timeout and pop up message, and the shutter does not close automatically!!! 
grrr. So in the end the thickness is close to 1nm instead of target 0.5nm
- oxidation 149mb, 9' then 2' to pump
- Ti pump -> P=1e-7
- 60nm Al @ 1nm/s, P_ev = 3e-7

Lift-off warm aceton + 30" US (hard to remove small pieces on the edges)
obs optical: alignement seems OK, gates seems damaged on res 0 (left) but OK on res1 (right)!!
Does the problem come from residues of MIF under the gates, which is rehydrated and reactivated when processed again ?????

- bake 110°C, 1'
- spin AZ5214E (batch) 4000rpm
- bake 110°C, 1'
- expo 3" @ 9mW/cm2
- bake 125°C (setpoint) 2'30
- flood expo 30"
- dev MF-CD26 30" 19°C

- Ti pump, 30nm @ 0.2nm/s -> P_sas = 3e-8mb (gun off)
- Al 130nm @ 2nm/s, 10°, planetary 16°/s, P_ev ~ 3e-7mb (on 23/04/14 : 2.5e-7mb)
- Ti 20nm @ 0.5nm/s, 10°, planetary 16°/s, P_ev ~ 3.3e-8mb (on 23/04/14 : 2e-8mb)
- Au 10nm @ 0.1nm/s, 10°, planetary 16°/s, P_ev ~ 6.8e-8mb (on 23/04/14 : 4.6e-8mb)

- spin MAA8.5 EL10 (batch 14020103, exp. 3/1/2015) @ 2000rpm for 45"
- bake setpoint 180°C, 6'
- spin PMMA A6 (batch 14020103, exp. 3/1/2015) @ 4000rpm for 45"
- bake hot plate setpoint 180°C, 17'