From HoloWiki - A Holography FAQ
Jump to: navigation, search

Notes from Ed Wesly

Here are some notes on photo-resist technology that go against the grain of conventional wisdom that you may have heard or read.

The thickness of the resist is not critical. The original application of making masks for the micro-circuit industry dictated the micron or so thickness, as the resist would be developed all the way down to the substrate so it was laid bare for etching.

A good friend of mine, Manfred Stelter, who runs Process Technologies Inc, is the father of the chrome on glass technique. The resist is exposed, then developed so that the metal is bare in some parts, and then immersed in an acid, so the bare spots are etched, while the areas covered by the coating are not. Hence the term photo-resist, as it resists the acid. That is why the developer is an alkali. I have seen Drano or NaOH used, but I would recommend getting the Shipley developer, as it has some sort of surfactants and who knows what else secret ingredients in it. There are many developers, we used 303A.

Ferric Oxide, or rust is the usual way that the holographic plates are sold. It is put on top of the glass, then the resist on top of it. It could be etched to make a mask, but for our application it is a good anti-halation layer, as its orange color absorbs the blue photons. But we would spray the backs of the ordinary single strength window glass with Krylon Flat Black Acrylic, which could be soaked off with water if the paint didn’t get too dry.

Shipley does not sell direct, and when I typed in, I got this site: which is hard to navigate. We used 1813, diluted with its thinner 1:1. If you are going to electroform the resist, you will also need the stripper to remove the resist from the first generation. You could use the developer, but it might damage the silver.

Originally we got the plates from Towne as John mentions above.

Another source we used was: [url] [/url] as they would coat bigger sizes.

PR plates cost a lot, but they are in the same ballpark as AgX plates nowadays!

We coated our own plates, 12” by 12” became our standard, but we went all the way up to 42” square to make huge gratings! So here is how we would do it, courtesy of Steve Provence, one of the wildest of the wild men in holography ever! He is out of the business, so I don’t think it is a problem. Good luck, Jamonero! (He went into the Prosciutto business!)

The plate must be scrupulously cleaned with distilled H2O and a little Alconox. It must be rinsed with the clean water, no additives, and let to dry completely, absolutely no water.

Because the next step is to treat the surface with HMDS, some sort of silane. I forget the brand and the US distributor of this Japanese product. The directions say fume the plate with it, but not having a big enough fume box, we just spun it on. It dries pretty quickly. But it does evolve ammonia in contact with water, so make sure everything is dry in the spin area.

Then we would spin at record turntable speeds, like 80 rpm. This goes against the usual conventional wisdom, but be don’t need a super thin layer like the electronics industry for etching, since we are just picking up the surface relief. So don’t worry about the fast start and spin like you read. And in regards to a question above, the thickness does not affect the sensitivity. Most of the resist will fly off the plate as it starts spinning, but if you keep the bowl around the spinner clean, you can recycle it! We would pour on the resist, then turn on the spinner.

After spinning, we would let the plate sit on the spinner to dry a few minutes before packaging. Then we would let them sit in their box in the dark in a clean area for a week before shooting. Although the instructions say to bake the plates, this is only to remove residual solvents. A slow cure does the same. You can use the plates sooner than that, but they will be faster.

Boy, do you do need blue photons! We shot with Argon 458 nm for stereograms, as we used Agfa Millimask for the H1, which would be way too scattery at the Krypton 413, which was used for the big gratings. Provence used 413 nm for his objects and 2D/3D jazz, since he shot PR masters. At first we used the 442 nm of the He-Cd for the Dotz! Machine, but would only get about 3000 hours out of the tube before it ran out of Cadmium. This dot-matrix machine was used 24/7, so it was like two $5k tubes a year, so we switched to a 405 nm diode eventually. And the deeper the lambda, the more sensitivity.

Exposure doses were about 1000X more than AgX. Yes, you read that right! Our benchmark was 93 milliJoules compared to the 200 microJoules of MilliMask, and we would bracket over and under that by 1/3 stops to really tune it in. At the beginning we used ratios of 40:1, but then tested and brought it down to 10 or 5 to 1.

Developing times are ridiculously short: 6 to 8 seconds! We would plunge the plate into the developer, diluted 1 part D to 6 parts H2O, and agitate furiously! Then yank it out as quickly as possible, to be plunged into running 18 MegOhm water for a couple of minutes. A hair dryer took care of the water at the end. Some people like to develop under a yellow bulb, but we had automated equipment to handle the exposure angle, so we could gang up multiple exposures on one plate. Play with dilution and times to find the method that suits your style.

If you do dabble in resist, you need to silver the plates to find the best result for embossing. This company sells spray and brush on chemicals, and will take credit cards over the phone as opposed to someplace else that want you to set up an account, etc. [url][/url] We would pick the best expo to the eye, then double it, as the embossing process can be quite lossy.

Another thing that goes against conventional wisdom is that you can shoot Single Beam Reflections with photo-resist! I would never have thought of that, but one of my student workers did it for the heck of it on an unbacked plate, and it worked!

The reason it worked is that the only time the reflection Bragg planes are parallel to the substrate is when both beams are hitting the plate along the normal. But with the usual object behind, reference from above geometry, the fringes are no longer parallel to the substrate but break the surface of the coating. Once silvered, it didn’t look too bad, except for the spurious twin image not unlike a Gabor hologram.

So if anyone is serious about setting up a resist lab, I am available for consulting. PM me if interested. _________________ I want some blue photons!