At the beginning of the week I thought that I might have an adequate setup and procedure for identifying the oil contamination in water, via the fluorescence spectrometry, discussed in my last update. On Thursday I tried taking the fluorescent spectrum of an actual oil product for the first time- I used mineral oil. Mineral is not optimal, because it is essentially the most refined form of oil, as opposed to the crude oil I am ultimately attempting to detect. However, it should still contain fluorescing aromatics, based on this paper. Mineral oil should theoretically fluorescence, just with lower intensity, and shifted wavelength (should still be in the visible range though). In any case, I tried to take a spectrum of pure mineral, which in contrast to dilute pollution, should produce a stronger fluorescence. I was unable to pick up any visible fluorescence. Whether this was because my source light was not intense enough, or that it is not fluorescing in the visible spectra I do not know- I will have to do more research. I suspect that fluorescent spectrum may just have too little intensity to be useful. In this case, I think detecting UV adsorption of oil contaminants by looking directly at the UV spectrum may be the most definitive way detect contamination. We just got a new camera that is supposed to be able to detect into the UV range, once Jeff and I remove the UV filter. (On a side note, I was looking into UV pass/Vis block filters, which are normally expensive, but a cheap alternative may be just using the glass of a black light, which is based on Wood's Glass, which blocks generally 420-700 nm, and lets UV through.)
Before taking the filter out of the camera, I decided to build a spectrometer 3.0, and test the new camera with it, to see how much I could improve. Today I put together a set of instructions for making 3.0.
I was able to make a thinner and cleaner slit with card stock this time, using a new method outlined in the instructions. I took a spectrum of a fluorescent bulb. Earlier in the week, I found what the actual spectrum of a fluorescent bulb is, and overlayed it on the spectrum of a fluorescent bulb taken with spectrometer 2.0, and found 5 peaks that line up with the major peaks in the real spectrum.
Then I did the same with the new camera+3.0 spectrum, and got much better results. I count 11 peaks in my spectrum that line up with real peaks in the actual spectrum. I believe this increase in resolution is mostly attributable to the finer slit. Next week I plan to make an even finer slit.
If the new camera is able to take usable photos of the UV spectrum, in the right range, this higher resolution spectrometer would allow me to detect finer variations in UV spectra, allowing for detection of lower concentrations of oil contamination, than I otherwise would be able to.