Photoresist bilayer (S18xx or AZ5214E) + LOL 2000Photoresist bilayer (S18xx or AZ5214E) + LOL 2000Edit
Sommaire
Inspired from Chalmers university
To achieve good lifoff, it is necessary to have a clear separation between the deposited film on the top of the resist surface vs. the film on the substrate surface in the opened pattern areas. This can be done by controlling the resist edge profile and giving it a negative slope, usually by modifying the resist surface by chemical treatment which makes the top part develop slower than the bottom part. This does not always work if the exposure is not 'sharp', i.e. if the image is blurred because the mask/resist gap is too large. Another way of ensuring the film clearence is to use an image reversal process which turns a positive slope into negative by reversing 'everything'.
A third way is to use a double layer resist system, much like we do in e-beam lithography, where the bottom layer geometry gets a positive bias compared to the top imaging layer either by higher sensitivity to the exposure dose, or just by a higher dissolution rate in the developer.
The LOL-2000 Lift-Off Layer a material is of the latter type, a non-photosensitive material which dissolves in photoresist developer in a controlled way. It is placed under the normal photoresist. After the photoresist is fully developed and the dissolution of the photoresist stops, the developer continues to dissolve the LOL-2000 layer in the open areas and further in under the resist edge, producing the clearence necessary for liftoff.(See the example graph and SEM pictures below)
In order to get reproducible undercut, fixed conditions should be used, both in exposure and development time (se the caveat above). Note that our previous standard developer MF322 should NOT be used here, as it is too strong and the liftoff layer dissolves too fast. Use MF319 instead. Experience shows that using the MF319 developer should require you to increase your exposure time slightly compared to that of using MF322. Using the conditions indicated below, an sideways undercut distance of 0.3 - 0.5 um is produced. If less undercut is desired, a higher LOL baking temperature (130 - 180 degrees C) should be used. The dissolution rate of LOL-2000 is dependent of baking time and temperature, graphs will be presented here later.
S-1813 example process
- Clean substrate
- Spin spacer layer LOL-2000 @ 3000 rpm for 200nm thickness
- Bake 140 C on hot plate for 5 minutes
- Spin resist layer S-1813 @ 4000 rpm for 1.3 um thickness
- Bake 110 C on hot plate for 2 minutes
- Expose through pattern mask,
- 45s @ 2.45 mW/cm2 (110 mJ/cm^2 nominally)at our 'new' UV250 mask aligner, intensity measured using the 400nm probe.
- 15 s @ 10 mW/cm2 (150 mJ/cm^2 nominally)at our 'old' UV400 mask aligner, intensity measured using the 400nm probe.
- Develop in MF319 for 30 sec, CAREFUL agitation
- Rinse in DI water
- Blow dry CAREFULLY.
- Ash at 50W 250mTorr Oxygen for 30 s
- Deposit thin film
- Lift off in acetone or NMP (Remover 1165)
- If you do liftoff in acetone, do a final clean step in photoresist developer , NMP stripper, or oxygern plasma afterwards. LOL-2000 does not dissolve well in acetone.
- Rinse in IPA and blow dry.
The conditions above should be seen only as a guideline, modifications of exposure and development time will be necessary for each application dependent on substrate reflectivity, etc. There may also be differences when exposing using the UV250 or the UV400 mask aligner as they have different spectrums in the 365 - 436 nm range.
AZ5214E example process
- Clean substrate
- Spin primer TI prime @ 4000rpm
- Spin spacer layer LOL-2000 @ 1500 rpm for 250nm thickness
- Bake 130°C on hot plate for 10 minutes -> theoretical dissolution rate of 30nm/s in MF319
- Spin resist layer AZ5214E 4000 rpm for 1.4 um thickness
- Bake 110 C on hot plate for 1 minutes
- Expose through pattern mask,
- 3s @ 10 mW/cm2 i-line
- Bake 125°C (setpoint) 2'30 for crosslink
- Flood exposure 30" @ 10mW/cm2
- Develop in MIF726 (or MFCD26, or AZ351B 1:4) for 40 sec to 1min, CAREFUL agitation
- Rinse in DI water
- Blow dry CAREFULLY.
- Ash at 50W 250mTorr Oxygen for 30 s
- Deposit thin film
- Lift off in acetone or NMP (Remover 1165)
- If you do liftoff in acetone, do a final clean step in photoresist developer , NMP stripper, or oxygern plasma afterwards. LOL-2000 does not dissolve well in acetone.
- Rinse in IPA and blow dry.
Measured (under optical microscope) undercut versus time in MIF726 @ 20°C for previous recipe:
| undercut (nm) | dev time |
| 800 | 40 |
| 1600 | 50 |
| 2000 | 60 |
Predicted possible problems:
- The LOL adheres very badly onto Si compared to SiO2. One should use a primer such as TI prime (HMDS?)
- If you develop too long, you may lose parts of your pattern if you have small isolated features.
- Another effect of too long development is that the undercut may be too deep and resist will sag down, and the clearance gap between the substrate and the top resist will disappear. Then you are back to your old process with photoresist only! No liftoff improvements!
- The LOL-2000 does not dissolve well in acetone. This is probably a deliberate design feature to avoid intermixing when applying the photoresist on top of the LOL layer. You can do liftoff in acetone anyway, but you need to do a final cleanup step in photoresist developer, NMP (remover 1165) or oxygen plasma afterwards.
- The LOL-2000 thickness is upward limited to about 200nm@2000rpm spin speed, and this puts a limit to the deposited film thickness to the same order. We have not tested to coat several layers of LOL-2000, it may be a possibility.
- Do NOT postbake. It will destroy the liftoff profile.
- Was this page helpful?
- Mots clés (Modifier les mots clés)
- Liens sur cette page
Fichiers 1
| Fichier | Taille | Date | Attaché par | |||
|---|---|---|---|---|---|---|
 | LOL-1000LOL-2000 datasheet.pdf Aucune description | 40.05 Ko | 15:55, 13 Nov 2013 | Helene_Le_Sueur | Actions | |
| Images 0 | ||
|---|---|---|
| Aucune image à afficher dans la galerie |