Today was my first day of "real" work on my research. I am working towards the basic goals I outlined in my 1st week plan Before I can actauly try to obtain any quantitative data, I needed to find a way to take consistent photographs of spectrum. This was difficult with the current design for two reasons: the camera was not rigidly mounted to the spectrometer, and there was alot of light pollution in the device. I build a simple spectrometer 1.0 with the tube being a roll of dark card stock, the same diameter as the camera lens assembly. This did mostly solve the problem of light pollution, but the assembly was still not rigid, so I could not take reproducible pictures. I did manage to take a picture of the spectra of an incandescent and a fluorescent bulb. I then made spectrometer 2.0 which I designed more precisely. The whole device is 12 in long, 8 before bend, 4 after, and 1 inch square. Making square allowed me to easily intersect the whole cross section with a square cd piece. This design was rigid, and almost no light pollution, meaning I was able to take high quality, and more reproducible photos. Here are photos of incandescent and fluorescent bulb spectra. All photos

I was surprised at how crisp the spectra I was getting were, despite only using slits cut in card stock as apertures, rather than razor blades. All in all, these spectra were produced with a <1 square foot of black card stock, black electrical, scotch, and packing tape, scissors, and a camera.

My next move will be to see how accurate these spectra are by taking spectra of known colors, and comparing. The grand scheme of this research is to be able to detect pollution in water, due to various contaminants, such as crude oil. One promising piece of information I found was that crude oil, which when dissolved in water may be transparent to visible light, contains PAHs which will adsorb UV light and fluoresce in the visible spectrum. I also have found a number of useful papers on the subject: http://iopscience.iop.org/2041-8205/712/1/L16/fulltext http://www.turnerdesigns.com/t2/doc/appnotes/S-0079.pdf http://costperformance.org/monitoring/pdf/uvfluorescence_2.pdf

If you go here you can see the spectrum analysis of the spectrometer 2.0 spectra

One idea I had to test the feasibility of sensing visible fluorescence due to UV excitation is to spectrum a solution of laundry detergent in UV light, as detergents contain fluorescing compounds to make your clothes look brighter.