The process is to use the LED flash of a smartphone camera, a bit of a CDr or DVDr as a diffraction gradient, the camera, and the processing power on the phone. You measure distilled water, and then the sample and compare the two to see the differences.
The drawings are a quick render of the OpenScad file. The half-circle is a pivot that will hold the diffraction gradient so I can play with angles. I also need to add a spot for the test tube somewhere along the thin light slot.
The whole thing will be parametric, and customizable to any given smart phone.
Well, this is the plan. I am going to sleep on it a night, see what I dream up on it, and maybe send it to my printer tomorrow.