Hi Dilek
About the choice of the PR: I just went for one which has a sufficient energy absorption within the visible UV range still tolerated by the inexpensive DMD chip chosen. Now that you pointed out the issue with TMAH I realize that my choice of PR wasn't the best. Can you suggest alternatives which do not require super toxic and dangerous developers but still work within the specified UV range of 420nm - 450nm? I'm aware how dangerous TMAH is, it even etches silicon, which I observed myself while doing some tests for the ultra low tech process flow variant, where I used it to etch the trench isolation. The absorption wave length was basically the only criteria which led me to suggest this specific resists, but as I've said, if we could collect a list of alternative resists delivering sufficient results in the above mentioned range, we HAVE TO put it onto our wiki page as well.
About the certainty of being able to reproduce 25 um (featuresize 50um -> lambda 25um): In order to be sure, we've got to build a prototype and test it :-)
Cheers, David
On Friday, August 13, 2021 8:03:33 AM WEST Dilek Isik Akcakaya wrote:
Hi All, David,
I see that you selected a greyscale PR for photolithography, it uses TMAH based developers I do not know the range of other developers it can make use of. Can we list them? This series of PR are foreign to me. I used Shipley, AZ etc. You must not buy and use TMAH based developers at home labs due to safety reasons it is extremely toxic! Second chemical I will beware of is HF in the same way in a home environment.
Second, What is the reason for going for a grayscale PR and not a regular one?
Third, The UV range depends on the PR we will be using, the smaller the wavelength, more precise will be the features if we can use the right optical system of course (De Broglies eqn.). Because I am new to the project please let me know why we decide don this set of PRs.
As per UV range of the DLP once again this depends on the PRs to be used and because DLP works by reflection, you may loose power due to reflectance, absorption and other optical phenomena therefore I think we need to be as precise as possible with our selections.
Let me point out: The rail system precision+ DLP resolution (not the pixel size) as an end result of multiple selections such as good focus etc+ PR thickness+ Development time + Light source wavelength + overlay reproducibility = approximately OUR RESOLUTION.
With current selections are we sure we will be able to reproduce 50 um features and have a good overlay if we need a second layer alignment on top?
Thank you,
Sent with ProtonMail Secure Email.
‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
David Lanzendörfer leviathan@libresilicon.com 11 Ağustos 2021 Çarşamba
saat 12:51 tarihinde yazdı:
Hi Dilek, hi list
Sorry again about the downtime of the infrastructure during the day yesterday.
The infrastructure of the data center I've got my rack server in was down.
Now to the topic:
I've started writing down the BOM and I'm not sure whether 420 nm UV already
has too much energy, although the data sheet of the DMD chip says it's ok to
be operated until 400nm.
Here's the BOM:
https://redmine.libresilicon.com/projects/maskless-lithography/wiki
It would be cool to get feedback on that from Texas Instruments.
Cheers
David
(__/)
(='.'=) This is Ninja Bunny.
(")_(")
Copy and paste Bunny into your
signature to help him gain world domination