Not much to report from this last week, unfortunately. I printed a test enclosure for the lenses and a slide holder, then proceeded to test different exposure times with the photoresist.
The goal was to find the shortest time on the projector so that I could evaluate if the projector is sufficient, or if I need to start working on integrating the laser.
It was all sidetracked by problems with the photoresist. Or rather, developing the resist. Many consumer resists are developed in an alkaline solution, ranging from the relatively mild sodium carbonate through sodium metasilicate and ending at potassium or sodium hydroxide.
I assumed this random photoresist from China would be similar. I assumed incorrectly :)
Alkaline solutions do appear to loosen the resist, and allow it to slough off in big sheets when agitated manually. But the resist doesn't dissolve, and doesn't maintain any reasonable resolution.
I've tried several different concentrations of sodium carbonate, metasilicate and NaOH. I've tried manually agitating, sloshing, dunking in alternating developer/distilled water baths, pipetting forcefully. I even tried a few where the solution was heated to ~70C. No dice.
The eBay listing is not terrible helpful either:
Organic solvents do, indeed, remove uncured resist very well. Acetone will aggressively dissolve it, as will less-volatile solvents like lacquer thinner and MEK. The problem is that they the cured sections are visibly disturbed by the solvent. When viewed under a microscope, the cured resist becomes somewhat clumpy, with frayed edges and pieces starting to lift off.
This appears somewhat standard for negative resists. The literature says that the semiconductor industry moved away from negative resists in part because of their lower resolution. Cured resist remains somewhat soluble in the developer, which causes some accidental removal as well as “bloating” the cured section.
So I may just be seeing an extreme version (due to inferior quality resist) of what the industry already knows. In any case, I doubt I can achieve any kind of resolution with an organic solvent as developer and this particular resist.
So I'm back to hunting for a better resist. I've continued enquiring with smaller semiconductor supply houses if I can get a small order / sample of resist, but no luck so far.
In the mean time, I have two more resists on the way:
ER-71, a negative photoresist from Datak. PC-197, a positive novolac photoresist from Injectorall I also purchased the “official” developer for each resist, just to rule out any variables. I can experiment with recreating the developers later if it works. I'm particularly excited for the PC-197 because it's a positive resist that appears to be chemically similar to semiconductor-grade resists (e.g. probably a DNQ-novolac resin). It is, unfortunately, very expensive.
So that's where I'm at. While waiting for new supplies to arrive, I think I'm going to start modeling out an enclosure that incorporates the laser. The projector light seems sufficient to develop, but only after long exposure times (15+ min).