Public Lab Research note


Oil testing kit Beta programme - Scanning the unknown & the not so well known…

by cindy_excites | November 09, 2015 18:07 | 93 views | 4 comments | #12389 | 93 views | 4 comments | #12389 09 Nov 18:07

Read more: unstable.publiclab.org/n/12389


Oil testing kit Beta programme - Scanning the unknown & the not so well known…

Date: 5/Nov/2015

Here I document my experience collecting samples in my neighbourhood with the Public Lab Spectrometer 3.0, which was assembled using these instructions.

Due to time constraints I have only been able to scan three substances. These are: * Mr Muscle glass cleaner * Hardened and compacted residue from outside a car mechanic's workshop * Oil on asphalt

1) Scanning Mr Muscle glass cleaner

IMG_4151_-_Mr_Muscle.JPG

I had not planned to scan for this originally. It came about because in my last experiments, I spilled crude oil over the table and side of my computer. @TedF suggested I try cleaning the spill with window cleaner as it is very volatile. It worked! And I am glad for it - my laptop doesn't smell of crude! So, if any of you ever spill crude on your table or electronic equipment, Mr Muscle can help!

Mr Muscle scan:

Not knowing what the scan would return, I was very surprised to find that the peak was very similar to that of mineral oil. Indeed, the peak of Mr Muscle is at 407 nm. Here is rep a:

Mr_Muscle_-_capture-page.png

Mr_Muscle_-_rep_a.png

And when I went to find similar matches I found that it was similar to @melissan's OTK _Mineral_oil reps b and c:

Mr_Muscle_-_rep_a_-_closest-matches.png

I wonder if @melissan and in general if we were all given the same mineral oil in out OTK packages.

I scanned again for rep b

Mr_Muscle_-_rep_b.png

And when I went to find similar matches I found that it was similar again to @melissan's OTK _Mineral_oil reps b and c but aldo @warren's OTK Test UV laser baseline, and my own MrMuscle rep a. Why if rep a was similar to @melissan's samples did it not return these other similar matches before?

Mr_Muscle_-_rep_c_-_closest-matches.png

The wavelength range of rep b is from 404 to 411 nm with a peak at 407; while rep a also has a peak at 407 nm the range for rep a is from 404 to 413 nm with a less 'smooth' drop of wavelength at the higher end of the spectrum.

Mr Muscle rep c:

Mr_Muscle_-_rep_c.png

With the similar matches showing @melissan's rep c; my previous MrMuscle reps; and my Mineral oil test with attenuator:

Mr_Muscle_-_rep_c_-_closest-matches.png

This rep had the same range as rep a but the peak was not as smooth as the other two reps. Again, as with my scanning of previous samples I switched the cuvette around 180 degrees (which means that if all else remains the same (e.g. spectrometer is solid and stable) rep a and c would always look the same - as shown in the image below). In reality, however, the spectrometer was indeed moved around a little as I opened the cuvette frame lid. These changes/errors could of course be minimised if I make a support base that firmly holds the spectrometer and cuvette frame tightly in place.

Here we can see that rep a and rep c are the closest in shape:

Mr_Muscle_-_all_reps_compared.png

A few reflections: Does this mean that the cuvette is different on each side? And that the spectrometer can actually pick up that difference? Does it matter? I went back to look at my comparisons of previous scanned samples and found that in some case, reps a and c were similar but not in others. I did reuse some of the cuvettes washing them with dish soap and MrMuscle - but I did not keep track of which were which.

2) Scanning hardened and compacted residue from outside a car mechanic's workshop

IMG_4152_-_Black_substance_site.JPG

IMG_4153_-_Black_substance_cuvette.JPG

The main mechanic's workshop was closed when I went to collect my sample but the place where they keep the cars, just up the cul-de-sac was open. A man re-arranging one of the cars came over in the car and asked "What's that?". Crouched, I finished scraping the sample and turned to him. I recognised him from some minutes ago, when I had come to see where I wanted to take the sample from - we had briefly made eye contact from afar. I realised I needed something sturdy to scrape the black residue and collect it into the cuvette.

"This?" I said.

"Are you a scientist?" he asked.

"Ha!" I hesitated - a million possible answers relating to informal science flashing before my eyes.

"Me? No. I'm just doing a... project."

"Are you with Harringey*?" he insisted.

"No. Look, my shoe laces are untied - a council rep wouldn't come out like this and would not be scraping a sample with an old medication tray."

"Oh, ok, so you're not with the council? What are you going to find in that?" he asked.

"Well, I don't know. That's what I'd like to try and see."

"Oil. That's what you'll find. It's just old oil."

*Harringey is our borough; the Harringey council is our local government.

A few reflections: When poking around and behaving out of the ordinary, one might raise suspicion… In this case I did not feel like sharing with him my story about OTK or our cardboard spectrometer. Maybe because he felt threatened by what I was doing but also because I was not sure that what I was doing was going to yield any results that would affect him… something to think about. Did he not ask more pressingly because he did notice my shoes laces were untied - or because I was a woman? or…

I took my collected sample and separated it into two cuvettes. I poured mineral oil and shook the samples for about 10 minutes to try to dissolve the material. Reps a and d correspond to the sample in cuvette 1; reps b and c correspond to cuvette 2.

Here is the 'dissolved' sample before vigorous shaking and separation into two cuvettes:

IMG_4156_-_Cuvette_black_subs.JPG

i) Cuvette 1 Here are the scans from cuvette 1. Rep a (which is the first rep) for cuvette 1:

Black_substance_rep_a_-_capture-page.png

Black_substance_rep_a.png

The closest matches for this rep include my mineral oil sample scans as well as @melissan's reps for mineral oil:

Black_substance_rep_a_-_closest-matches.png

Rep d (which is the second rep) for cuvette 1. Again, for this rep, I took the cuvette out of the frame and turned it 180 degrees:

Black_substance_rep_d.png

Interestingly, the closest matches for this rep include @melissan's rep c for mineral oil but also all of my Mr Muscle reps, and my oil on asphalt test (which I document below)…

Black_substance_rep_d_-_closest-matches.png

ii) Cuvette 2

Here are the scans from cuvette 2. This is rep a (which is the first rep) for cuvette 2:

Black_substance_rep_b_-_capture-page.png

Black_substance_rep_b.png

The closest matches for this rep included @cbreuer's Gowanus Canal sediment #2; @duds_bc unknown specimen 1; and UV on limunifore white laminate board by @lievenv:

Black_substance_rep_b_-_closest-matches.png

Rep c (which is the second rep) for cuvette 2. Again, for this rep, I took the cuvette out of the frame and turned it 180 degrees:

Black_substance_rep_c.png

The closest matches for this rep included @jeffh's Supposedly ultraviolet flashlight; @miriamarchem8's Nitrogen #3; and @lievenv's UV on red paper:

Black_substance_rep_c_-_closest-matches.png

And here are the four reps compared.

Black_substance_-_compare-all-reps.png

Clearly, I did not shake the material well enough in my cuvette and the material was not dissolved and spread evenly.

What I would have done differently: 1) Crushed the powdery material even more finely and dissolved it for longer; 2) made that drill-powered centrifuge…

3) Scanning oil on asphalt

There was a convenient (and also inconvenient) oil mark on the asphalt right in front of our house:

IMG_4154_-_Oil_asphalt.JPG

I stood there in front of it thinking about the best way to get a sample of it and then remembered seeing a picture in the Public Lab website of someone mopping up some oil with a q-tip. I then went back into the house to gather the appropriate materials and equipment for data collection: latex gloves, cuvettes filled with mineral oil, q-tips, and zip-lock bag.

To collect the oil sample I pressed the cotton bud onto the asphalt - especially, on the protruding gravel to avoid getting dirt and debris. I repeated this for each of the two cotton buds of two q-tips for a total of 4 'samples'.

This was good in principle. Back in the 'lab' I carefully removed from the buds any debris I didn’t manage to avoid. I then dipped each bud into mineral oil squeezed the tip against the side of the cuvette. A nice droplet of partially dissolved oil slid down into the cuvette, which I had partially filled with mineral oil. I did this with three of the four buds but the third was lost to the bottom of paper cup with old tea… Not the best lab after all!

Sample of asphalt oil 'dissolved' in mineral oil:

IMG_4155_-_Cuvette_oil.JPG

Due to time constraints I only did one rep of my oil on asphalt, so I have yet to finish this properly. For now here is rep a:

Oil-on-asphalt_rep_a_-_capture-page.png

Oil-on-asphalt_rep_a.png

The closest matches found were @melissan's mineral oil rep b and all of my Mr Muscle's reps…

Oil-on-asphalt_rep_a_-_closest-matches.png

Reflections: What I take home from this exercise is that my spectrometer needs to be held more tightly together so that there is consistency in the readings - e.g. the cuvette should always be in the same spot - tightly pressed against the spectrometer's slit. This, I found, made a great difference in the readings - more than whether that cuvette frame lid was on or off. Also, the type of cuvette USED would make a difference (as I briefly discuss in my technical issues research note [ HYPERLINK HERE ]). And finally, more samples at higher concentrations and more reps. Also, has anyone found any other solvent for this kind of sampling? But I also think I might not have had enough concentration of my substance for the spectrometer to capture it. How could this higher resolution be achieved?

This is all for now! Hope to continue some more work on this in the near future!

Related research notes:

Oil testing kit Beta programme - Package content

Oil testing kit Beta programme - Assembly: cuvette frame

Oil testing kit Beta programme - Assembly: spectrometer

Oil testing kit Beta programme - Calibration

Oil testing kit Beta programme - Scanning OTK samples

Oil testing kit Beta programme - Technical adjustments (hack!)


4 Comments

Hi, Cindy - so, any of the ones where you just have a sharp spike around 405 nm just mean that there was not any fluorescence, or perhaps just none bright enough for the camera to detect. So they are just matching other spectra where there was no fluorescence, like the mineral oil, which we chose specifically because it does not fluoresce (a 'known negative result').

The one with the 'wobbles' around it is interesting. It looks a little bit like a Raman spectrum, but I really don't know much about that --

Maybe @stoft or @flatCat or @dusjagr, or someone else can tell us more about if this has anything to do with Raman Stokes or anti-Stokes lines (image from here):

RamanStat.gif

All three reps of Cuvette 2 of residue had those lines? And can you share a photo of Cuvette 1 and 2 next to one another? What was the difference between the two?

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


Yes, I can post a picture of two cuvettes when I get back to the UK (after Dec. 8th). And yes, I'd like to learn more about the Raman scattering. Are you suggesting that the light was scattered - that there was a change in energy of a molecule due to a transition to another (usually higher) energy level - and that is what was captured and that it is because of the laser used?

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


fantastic work!

Reply to this comment...


@warren raman scattering is usually a very weak effect, and not easily seen. The stokes and anti stokes lines are very small compared to the main fluorescence peak since very few molecules are in excited states at room temperature.

I'm not sure what causes the 405 nm laser light to spread out in the ghostly effects like it is in cuvette 2. Could that be bloom from overloading the CCD/CMOS photon detectors in the webcam? I think I remember hearing that if one area of the sensor is overloaded, that can bleed over to adjacent wells in the sensor and appear as signals in neighboring wavelengths...?

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


Login to comment.