What I want to do

Since the oils in the Oil Testing Kit 5W30 and 20W50 are distinguishable by their fluorescence spectra (see Research Note: http://publiclab.org/notes/gretchengehrke/07-10-2015/oil-testing-kit-discerning-5w30-20w30), I wanted to see if a 1:1 mixture of the two oils would have a fluorescence spectra with bimodal lambda-max values that would align with the lambda-max value of each neat oil.

My attempt and results

I used the gradation on the eye droppers to mix equal parts 5W30 and 20W50 in an independent glass container (a shot glass, empty of course) and then used a new eye dropper to agitate the oils into a homogeneous mixture and transfer the mixture into a new cuvette.

I took a new white light spectra for this session's calibration, and then took the spectra of the mixture, neat 20W50, and neat 5W30. I used the new white light spectra to calibrate the sample spectra, smoothed each spectra, and equalized the peak heights. https://spectralworkbench.org/sets/show/2439

The mixture does indeed have bimodal lambda-max values, and they do align with the lambda-max values of each neat oil. Awesome!

I do wonder whether or not we would be able to discern the end-members of the mixture without direct comparison to neat products in this same analytical session. I doubt it, simply because of peak overlap. I will try comparing spectra across days and different ambient conditions.

Questions and next steps

The next step is to do triplicate analyses to observe replication/precision. I also want to compare spectra of the same samples across days and using different calibrations.

Why I'm interested

Oil Testing Kit -- we want to know the precision and the ability to discern oils and mixtures of oils. Also, we need to know how fluorescence intensity of different samples impacts their mixtures' spectra (i.e. if you have a 1:1 v:v mixture of oils, but one has a much stronger intensity than the other, how could you tell from the resultant mixture spectra what the mixture ratio is?)