1997 William Beaty

Years ago I *made* a big prism from scratch, using polyester casting resin. This material has very good dispersion (makes good rainbows.) For an earlier science museum exhibit, our machinist had milled and polished a Plexiglas prism, but the dispersion wasn't great and the rainbow was narrow. We had to find something better. We ended up using holographic diffraction film such as Edmund Science $8 sheets. But while trying various things, I ended up making a nice large polyester prism which casts good rainbows.

For the prism mold, I took a couple of 4" x 16" plates of glass 1/4" thick, cleaned with alcohol, put a very tiny bit of spray-on silicone mold-release on them, taped them together on one edge to form a V-shaped trough, and plugged the ends of the trough with big chunks of plasticine clay. I arbitrarily picked 45deg as a prism angle, but I think a smaller angle would work OK too. I filled it with water to check for leaks and determine the volume of liquid plastic needed, then emptied and let dry. I saw that the plasticine wasn't holding the trough together by itself very well, so I added lots of duct tape strips across the top.

Next, I pre-coated the glass with highly-catalyzed plastic to give the final product a hard finish. To do this I mixed up a half cup of clear paperweight-making polyester resin, using plenty of hardener, poured it in, tilted the mold around to coat the glass, then poured out the excess. ( Ernst hardware had one-gallon cans of the stuff.) I don't know if this mold-coating step is required, but I think it guarantees a much harder surface for your prism. Thick castings sometimes are sticky, since you cannot use lots of hardener without overheating and warping or cracking them. The more hardener in the plastic, the less sticky and jelly-like it is, but the hotter it gets while hardening.

After the coating in the mold jelled, I mixed up the measured amount of resin with a small amount of hardener (3 drops per ounce? something like that,) and poured it in. Note: to attain a clear casting with no internal "ripples," mix the hardner very well. I mean VERY well, and when done, pour the mix into a second clean container and continue stirring. This eliminates the boundary layer of unmixed plastic which clings to the sides of the first container. Use uncoated paper containers (plastic ones will melt, and waxpaper ones will scrape off flakes.) If the hardener isn't mixed in totally uniformly, there will be swirls of index-variation in the finished plastic block. Also I whacked the container hard onto the table and waited a while for any bubbles to rise, then swept them to the sides with a toothpick, then poured the plastic into the glass mold. I taped the top edges of the trough together with stretched duct tape, since the oily plasticine clay looked like was going to de-adhere from the glass and dump the liquid plastic on my workbench.

After it was nearly hardened, I made a mix of black-dyed plastic with lots of hardener, then poured it on top to form a thin 1/4" layer. This forms one black face on the prism's "bottom." It eliminates that one grungy clear top surface, and removes both the stickyness and internal optical reflections from the prism.

The result worked great. Everything seen through the prism glowed with spectra. The one black side improves the esthetics, if not the optics. I sawed off and sanded/polished the ends where the clay had touched. The end result was about four inches wide and a foot long.

I didn't have a lead-glass prism for comparison, and I never did try actually measuring the dispersion. A slide-projector with a slit- aperature slide made a nice rainbow in the dark.


Fill your mold with water and look through it. This gives you an idea of how the plastic prism will behave. (If you're doing a science project, maybe a water-prism would be lots easier to make than a cast-plastic prism.)

When a 1mw classroom laser was sent through my prism, I noticed that the beam was visible inside the bulk of the plastic. This effect could be harnessed. Instead of a prism, cast a small, cubical "light beam smoke tank" made of solid plastic, with no milk and water to spoil over time. The effect might be enhanced with a VERY small amount of white pigment.

I've noticed that variations in the amount of hardener in the bulk plastic will create variations in refractive index. The plastic shrinks under too much hardener, increasing the index. This effect could be used to build an "artificial mirage," inside a plastic block, by pouring two layers having differing amounts of hardener. If a plastic cube with polished faces was cast in this way, then a laser should bounce off the interface between the layers when directed at a glancing angle. Direct observation should reveal a layer of "silver" when viewed at a low angle. Or, a narrow layer of differing index could be cast within a block of plastic, giving an "optical duct" similar to an optical fiber, and analogous to the "acoustic ducting" which lets whales communicate with each other across the world's oceans.

I wonder if it's possible to harness the above effect to build your own crude GRIN lens (gradient index lens?) If a straight streak of overly-catalysed LIQUID plastic could be put in a large block of partially gelled plastic, the catalyst should diffuse outwards and produce a cylinder with radial gradient index. Maybe this could be done just by dripping in some hardener, or pouring a little highly-catalyzed chilled liquid plastic into a volume of heated, less-catalyzed mix and hoping it sinks and leaves a trail behind? Or dunk a long needle in some hardener, then stab it into a block of partially gelled plastic? Inject it with a syringe? Or somehow make a thin "wire" out of plastic (pump it into a narrow hose?), then when it is hard, dip it in hardener and stick it into a cup of lightly-catalyzed plastic. When it has completely hardened, there might be a gradient-index rod which runs through the larger block.

If all of the above is too complicated, you can also get a "giant prism" effect by using Holographic Diffraction Gratings. These are cheap at Project Star website:

5" plastic spectroscope gratings & filters, 654-0025 $19.95

Created and maintained by Bill Beaty. Mail me at: