2015-11 (new design)

2015-11 (new design)Edit

mercredi 4/11

step 1- Laser lithography coil Resonator + transmission line

Wafer Si standard, with 500nm SiO2
- 100nm (?????) Nb sputtered by Chloe Rolland (to check thickness with her)
- resist spin:

bake wafer 1' @ 110°C (solvent evaporation)
spin S1813 (batch x, per xx/xx/xxxx) 60" @ 4krpm
bake 2' @ 110°C

- Litho laser (IEF)

Expo defocus 1900, energy 50
Dev MF319 30" @ 19.0°C, rinse ODI 10" beaker

- etch RIE (SPEC)

RIE CF4/Ar 20/10, P=50µb, 50W.
After 3'30 signal is minimum, etch seems to be over
Could not find a proper place to put the laser because the chips are centered on the wafer,
and laser is on the side.
As I don't know the actual Nb thickness, I put as usual, 4'20 !

step 2- Laser Lithography BCB Bridge

Wafer Si standard, with 500nm SiO2 for doses tests

AP3000 - 30" @ 6000rpm ; bake 1min @ 100°C
BCB4024-40 - 30" @ 5000rpm ; bake 2min30sec @80°C

real sample

AP3000 - 3" @ 1000rpm acc 300rpm/s; 60" @ 6000rpm ; bake 1' @ 110°C
BCB4024-40 - 5" @ 500rpm acc 250rpm/s; 60"@5000rpm ; bake 2'30 @80°C

BCB has 10% sensitivity at 440nm compared to 365nm. So we need to increase a lot the dose.
Choose NRJ = 40 (with 30% + 10% filters) and focus = 1700 (to account for BCB thickness: focus at about half the BCB height)
On dose test, it is completely screwed !! (I took pictures but forgot to save, and they were erased before I could save them)
* 1st result is that we have to develop for very long time before the BCB disappears on the whole wafer -> DS3000 is most probably spoiled
* 2nd result is that a wide area close to the exposed patterns was burned, and then does not want to go in the DS3000 even after 15' at 32°C!!! BCB cannot be exposed with laser at 440nm, because it will reticulate a large area around (probably heat effect)
* 3rd result is that when I sonicate it for 2', almost everything goes away. (still adhesion problems!!)

vendredi 06/11

We decide to make a mask containing the 3 steps in a row. Center chip is step 1, top chip is step 2, bottom chip is step 3.
We choose the standard dose for IEF for Cr masks (though not tested for very small patterns on head 4mm)

NRJ = 40 (with 30% + 10% filters)
focus = 1900
40" dev MIF319: patterns of the coil were too exposed!
Again there is a problem of horizontal wires being more exposed than vertical ones.
Note the scan direction is vertical (with scanning width of 20µm that can be clearly seen on the pictures as bands of uniform dose)

- the dose is not uniform (very from one passage -20µm width- to another)
- for horizontal patterns, the shutters needs to open-close-open-close... very fast.
Maybe by reducing the scanning speed we could achieve correct result.

Anyways, decide to etch the mask: 1'20 in Cr etch

lundi 09/11

BCB dose tests on Si/SiO2 wafers

1st wafer: (T0)

- Spin 2 times (!)

AP3000 - 30" @ 6000rpm ; bake 1min @ 100°C
BCB4024-40 - 30" @ 5000rpm + add BCB and respin ; bake 2min30sec @90°C

Leandro noticed adding BCB helped removing the bubbles; but it increases thickness which we don't want.
Actually to ensure no bubbles, should gently spread BCB from pipette all over the wafer.
- expo 6" @ 13mW/cm2
- bake 30" @ 65°C
- dev 4' @ 30°C -> stil BCB all over;
- add dev 5' @ 31.5°C -> BCB went away from wafer surface, patterns are here
Obs: Seems to have a large spread (below, hole patterns are 20µm nominally)
Left focus on top, right focus on bottom (larger patterns)

2nd wafer (T1):

- Spin 1 time, same parameters as above
- Measure thickness at ellipsometer: 5700nm (! BCB is also old !!)
- expo 4" @ 13mW/cm2
- bake 30" @ 65°C
- dev 4'30 @ 32°C; observe: stil BCB all over;
- add dev 5' @ 32°C; observe: BCB went away from wafer surface
 

mardi 10/11

Observation at SEM of T0 and T1: cleave and sputter 20" Au-W

In fact, the wafers were underdosed and overdevelopped

3rd wafer (T2):

- Spin 1 time, same parameters as above
- expo 10" @ 13mW/cm2
- bake 30" @ 65°C
- dev 6' @ 32°C;

4" @ 13mW/cm2 + 9'30 DS300 32°C     6" @ 13mW/cm2 + 9'30 DS3000 31.5°C   10" @ 13mW/cm2 + 6' DS3000 31.5°C

T2 with 3' more development (edges slightly smaller)

jeudi 12/11

Baked T0 and T1 (the same samples that were observed) 1h @ 200°C
Observation in XL40 (max observation angle is 75°)

Gold-W on top was deformed. Shape is rounder but cannot really state if contact could be done. As well, we cannot tell for sure if it is that the BCB spread out (larger on the bottom) or if it restrained on the top due to solvent evaporation.
We will now bake the other part of the cleaved samples T0 and T1. See better the rounding on patterns in the sample center

T0 small bridge:   before bake  (obs ange 85°)                                           after bake (obs ange 75°)

Now we take the other part of the cleaved samples.
To ease comparison, left unbaked, right baked. Top T1 (4"), bottom T0 (6")

vendredi 13/11 - lundi 16/11

BCB etch rates with O₂ : SF₆  90:10
- 1st sample BCB-E2bak-01
BCB 5krpm, exposed 6" @ 13mW/cm2, baked 1h @ 200°C -> 5390nm

TO BE COMPLETED

mardi 17/11

Test of etching on lithografy made on BCB;

BCB 4024-40 was spined at 5000rpm for 30sec and then pre-baked at 80deg for 3min; Lithography of resonator was exposed for 6sec and developed 9min 30 sec in recup developer. The wafer was baked at 200deg for one hour; The wafer was then cut in three pieces, one of which was etched in SF6:O2 (10:90 using 5cc and 45cc at P=250ubar -plasma off- and V=70V) for 30 sec (rate is 5-6nm/s).

mercredi 18/11

Observation of result on SEM; 

 The result was observed in SEM and compared with a piece of waffer which followed the same process but was not etched;

First try to do lithografy of second step on BCB spined on Nb resonator; This was not possible due to inadecuate mask (inversed);

jeudi 19/11

Transfer of IEF mask into home made mask; AZ1518 exposed 7 sec (91 mJ/cm2 ) in hard contact then developed in AZ400K:H2O 1:4 during 40 sec approx; Cr etched in Cr etchant during 40 sec and then removed resine in Remover 1165 during several minutes; Cleaned with abundant water;

Two trys of litho were performed without succes; Very hard to get good contrast;

vendredi 20/11

Now we have adjusted the dose, bake, and etch
-> Proceed with second step on test chip

Unfortunately the spin was bad so I decide to remove it.
Try first AZ400K 15' + 30" us

Then remover 1165 15' + 30" us

Still a bit on the biggest pads of resist -> again 15' in remover 1165