I wanted to report back from last month's water quality meetup in Somerville, which was graciously hosted by Parts & Crafts and organized by Bryan Bonvallet, Katie Gradowski, and others on the Boston-area mailing list. Attendees were a nice mix of Public Lab regulars and newcomers, many from Parts & Craft's community of students and their parents. The event was on Nov 24 from 12-6pm.

Though there was a variety of work going on, here I'll just write a bit about the spectrometry work we did, as that's what I spent most of my time on.

Oil sample extraction

A group of us jumped right into trying to do some "extractions" -- where we try to use different liquids to dissolve materials we'd like to analyze in a spectrometer. Its common to have a solid or muddy sample with suspected oil contamination from motor or crude oil, but in order to make a dissolved solution of the oil through which we can shine a laser and take a fluorescence spectrum, we need a quick and reliable way to dissolve the oil and separate it from the sediment. (See that post for the full setup.)

I had learned a bit about this at the SETAC conference in Nashville some weeks earlier, and found that while we have been dissolving samples in non-toxic, cheap mineral oil from pharmacies, in the lab, a typical preparation is to dissolve samples using methanol, hexane, or more toxic solvents. I bought some methanol (also called denatured alcohol, and somewhat toxic -- it breaks down in a few days, but it will make you blind if you drink it) which seemed the least bad of the bunch, but we didn't really get around to using it, instead sticking to mineral oil to try to just do some more rigorous repeated tests with different samples, also trying to scan some samples 2 at a time so that the data would be recorded at exactly the same camera settings.

We'll try the methanol another time -- as a stronger solvent, maybe it'll work with some of the coal dust samples people have been having trouble dissolving: http://publiclab.org/notes/eustatic/09-09-2013/nungesser-test-detection-of-coal-or-petroleum-coke-for-flame-spectrometry

Patrick Herron from the Mystic River Watershed Association takes a sample of motor oil from his car.

Anyhow, we went through a bunch of samples and collated them here, including various ones from the BP oil disaster, some from the Gowanus Canal. The basic technique we used (trying something new that could be a bit easier than the repeated dilutions Scott Eustis and I have been doing) was to dip a cotton swab in mineral oil, then rub it against a chunk from a sample, or in one case the dipstick from a car's motor oil reservoir.

Then we dipped the swab in a cuvette filled partially with more mineral oil and kind of stirred it around until it began to dissolve, which was usually visible as a yellowish cloud.

https://spectralworkbench.org/sets/show/745

The really neat part was seeing the pure heavy motor oil spectrum, which wasn't the same preparation (it was just a cuvette of pure DTE Heavy motor oil) have almost exactly the same spectral curve at the beginning and end of all our tests -- it shows that the setup is quite repeatable, and that our worry about the laser's batteries dying was unfounded. The dimness of all the other spectra was not related to the strength of the laser.

Takeaways

We need more concentrated solutions, because the fluorescence wasn't nearly as bright as in the pure motor oil. Dave Stoft has suggested that we also use more laser, or brighter lasers. Both ideas should be tested. Also, sand in some samples made it hard to get just dissolved oil into the cuvette; but the sand does settle out and the bubbles to eventually dissipate (above). Still, not as tidy as we'd like.

Another thing -- the setup I described for blue laser testing can be assembled from the box the Desktop Spectrometry Kit comes in! I believe Cort discovered this:

Other projects

There were a number of other projects going -- including the construction and testing of a thermal fishing bob (written about in this wiki page by Sara Wylie), demonstration and discussion of a DIY potentiostat by Craig Versek and Don Blair and much productive discussion of different water quality sensing approaches.

The event was a lot of fun and I'm certainly looking forward to a more in-depth reprisal in January, hopefully at HackerFarm!