_Image courtesy of _
This activity was developed for the Oil Testing Kit Beta Challenge last winter. The basic hypothesis is that different grades (molecular weights) of oil will have significantly different fluorescence spectra, such that an oil type could be distinguished by the maximum intensity wavelength and range of its fluorescence. The purpose of the challenge was to test this hypothesis, and specifically to:
assess whether or not the PL v3 desktop spectrometer (with cuvette holder addition) could collect visibly distinct spectra for five different grades of oil,
assess whether or not spectra of oil samples captured with PL v3 desktop spectrometers were consistent across multiple spectrometers,
people could use their PL v3 desktop spectrometers to categorize unlabeled oil samples.
Our goal was to have at least 20 people replicate this activity and compare their results. Yagiz Sutcu conducted an analysis of the reproducibility of spectra, which I will try to attach in the comments section. The basic results indicated that spectra collected with different spectrometers (and by different people) were not consistent enough to allow for direct comparison. Replicate scans of individual samples using individual spectrometers were fairly consistent, but even that variation was large enough to inhibit proper categorization of certain unlabeled samples. It would be great to increase the number of replications for this activity, however, and to have more thorough documentation of experimental conditions that may have impacted results. Please try to replicate this and add your results for comparison!
The materials needed for this experiment are detailed in the oil testing kit wiki. They include:
In order to best be able to compare results, please include descriptions of the ambient conditions when you collect your spectra. Important notes may include:
Following the procedure detailed below, please be sure to conduct parts A and B summarized here, and parts C and D as much as you are able!
A. Documenting known samples 1. Take triplicate scans of each known sample, and label the scans as stated in the Procedure below. 2. Make a set from your triplicate scans of a given sample. 3. Analyze and post results, as described in the Procedure.
B. Investigating unknown samples 1. Take triplicate scans of each unknown sample, labeling the scans appropriately. 2. Make a set from your triplicate scans. 3. Analyze the results. 4. Compare the results against the suite of known samples you have already analyzed. 5. Post your results, along with your assertion of which kind of oil comprised each unknown sample.
C. Further Investigations 1. Try diluting samples using mineral oil, recording the dilution factor (e.g. 1 mL oil sample + 3 mL mineral oil = 4 mL total, and a dilution factor of 4), and analyzing the diluted sample as described in the procedure above. How dilute can a sample be and still be detectable and discernible? 2. Try mixing oil samples and analyzing the resulting mixtures. Are their spectra additive and discernible as such? 3. Try other fun things!
D. Share your thoughts 1. Provide ongoing feedback about the oil testing kit, including hardware, software, and programatic aspects. 2. Utilize Questions, Research Notes, the Spectrometry google group, and wiki editing to communicate and exchange knowledge.
OTK_(sample#, e.g. 20W50)_a,
oil-testing-kit, and your unknowns with
Verification and trouble-shooting are vital parts of any procedure. Please add more trouble-shooting tips in the comments, since I know this list will be incomplete (and likely insufficient).
To verify that you have constructed the spectrometer correctly, first visually inspect it to ensure that the sides do not bow. Next, follow the steps to ensure you can capture and calibrate a spectra, detailed here.
To verify that you have correctly constructed the oil testing kit, ensure that the cuvette is held snuggly and vertically.
Please add further verification steps in the comments!
Trouble-shooting can be an arduous process. Common problems with spectrometer construction include:
webcam installation at an angle. This interferes with proper spectra collection in Spectral Workbench since the software assesses a horizontal line of pixels, which will be off-axis if the webcam is installed on a diagonal.
webcam connection to USB is disrupted. This connection can be finicky. If you don't see any spectra at all, check to ensure that the cord is still plugged in to your webcam.
ambient light entering spectrometer due to bowing side walls. If the folds you make are not deep enough and precise enough, the spectrometer side walls can bow out and allow for ambient light to enter. This obscures the spectra because it disrupts the angle of incident light (which should only be coming through the narrow slit on the front side), and the spectra will appear blurry.
the angle between the spectrometer and the laser (via the cuvette) is not 90 degrees. Since fluorescence is properly observed at a 90-degree angle so as to avoid measuring incident light, your spectra can be drowned out by incident light if it is closer to 180-degrees, or have very low intensity of it is less than 90-degrees. If you observe a peak in your fluorescence spectra at the wavelength of the incident light (~405 nm in this case), you are likely not at 90-degrees.
the spectrometer slit and fluorescence from cuvette are not aligned by height. If you see no fluorescence spectra, it may be that the height of the spectrometer slit is above the height of the oil sample, such that the slit does not receive any fluorescent light. Try filling up the cuvette with more oil, or using a small platform to lift the height of the cuvette to be aligned with the spectrometer slit.
Please add more trouble-shooting tips in the comments!
Yagiz started a great discussion of the results of the Oil Testing Kit Beta Program with his report, attached below. Please read and comment, and discuss your own experience and ideas too!
Follow related tags:
spectrometry oil-testing-kit activity:experiment
The second section of the new DIY Oil Testing Booklet, which we’re highlighting this week, is focused on Working with Communities.
In this section, you will find four workshops ready to help you lead a group through:
Check out the workshops here:
To highlight #1:
The Design an Experiment workshop can be used for anyone who is interested in conducting experiments which make use of the scientific method. It will help you work through understanding the capabilities and limitations of data. Participants in this workshop will be able to learn about the elements needed for good experimental design, identify important points of designing a clear experiment, draft questions and transform them into hypotheses, and explored the concept of proof versus likelihood.
This section of the Oil Testing Booklet also explores the nuances of working with both online and offline community members. It provides material on outreach strategies for those working at different stages in projects. It explains how Public Lab run an online program for people to helped identify and to show the ability and limitations of the Oil Testing Kit through replication.
For those interested in community tool development, another section in this chapter explores the new concept of “Open open hardware” -- a reference to the fact that many “open hardware” projects are developed in private and only published openly upon completion. By contrast, the process we’ve proposed and begun to adopt is one where the goals includes things like:
- low barrier to entry for new contributors
- predictable revision timeline
- regular iteration and feedback on proposed changes to help them get prepared for the next release due date
- a single, consistent, versioned, "baseline" design for the project, emphasizing simplicity & low cost, but upon which advanced mods may be made
Learn more about the workshops, the programs and ideas explored in this project with the Oil Testing Kit Booklet:
Order here ($10 paperback)
Follow related tags:
spectrometer community workshops blog
Hello! I have uploaded a template version of the book design for the DIY Oil Testing Toolkit. It's a basic Indesign template. You can print it using a self-publishing service like Blurb or take it to a printer and get it spiral bound!
The template is for people and organizers who want to collaborate with others and help spread open-source tools for environmental exploration and investigation. The graphic elements in this work are release by the Public Lab and licensed under a Creative Commons Attribution-Share Alike 3.0 License.
The template was created by SuperCommunity (super-community.org) for the Public Lab. We’d love to hear how you are using the template and helping others with your work. Please email us email@example.com or firstname.lastname@example.org
Follow related tags:
spectrometer community workshops oil-testing-kit
Hello all -- today we’re announcing a booklet @gretchen, @stevie, @liz and I put together in December and January on the state of the Oil Testing Kit program, and more broadly on developing open collaborative projects. It has lots of details on oil testing in particular, but also covers topics like working with communities, cultivating online collaboration on hardware, and environmental regulatory issues. It’s full title is DIY Oil Testing: Progress toward community oil pollution analysis
The entire booklet has been posted as a series of wiki pages which you can browse, as well as refine, correct, and edit, here.
To quote from the introduction:
We’re writing this for multiple audiences, who have varied interests and experiences with oil and spectrometry. We’re hoping that everyone finds information useful to them within this document and that the case studies, event models, and analysis on our process will be helpful to anyone who seeks to:
- join and further Public Lab’s oil testing program
- understand the history and current state of the Oil Testing Kit
- conduct experiments using DIY spectroscopy
- develop their own community technology development project
- advocate for accessible oil testing
The booklet itself is available for preorder here, beautifully designed by @mlamadrid (see more of her work here):
Order now ($10 paperback)
We’ll be highlighting one of the four chapters each week, and adding more insights and comments we’ve made since the time of writing.
This week, we want to highlight the first chapter, Questions, Motivations, & Best Practices. In this chapter, we wanted to recognize that questions are the start of any research process, and that people with different backgrounds and skillsets can all ask good questions. We also wanted to acknowledge that questions breed more questions, and that we haven’t come close to answering them all.
What we have done is tried to clearly state what we do and do not know, what are likelihoods vs. knowns vs. complete unknowns. See for example this question:
How much do different tests cost?
According to the Surfrider group, which sent many oil samples to different labs over the course of the BP oil spill response, “When we were doing our oil study, we ran batches of ten samples using the 8272 modified solids GCMS method @ $295/sample so, $2995 a set. This was with the understanding that we would be running a lot of samples, so may not reflect the usual pricing.” This is how the per-sample cost broke down from the lab analysis for a lab contracted to run samples in batches:
- $340 for PAHs including alkylated homologues
- $1000 for petroleum biomarkers
- $160 for high resolution GC/FID
- $2 for sample disposal
- $700 for data interpretation and reporting
Total cost per sample would be $2202.00
Costs for oil testing vary widely, to as much as $2000 a sample for dispersant testing (according to Surfrider) and as as much as $10,000 for testing with PACE labs according to Scott Eustis (@eustatic) of the Gulf Restoration Network.
Thanks to everyone who contributed to this book in so many ways, from our community partners to the Oil Test Kit beta testers, to our fellows and organizers and staff, and to the Knight Foundation, who provided key funding for the project. We hope you enjoy this and help us take the next steps!
(Also -- we'll be sending out copies of this everyone who was directly involved in the project, so please expect an email!)
Follow related tags:
spectrometer publication oil-testing-kit barnstar:basic
I am not sure if this word even exists in the English language but well... I browsed this forum and found some notes by "stoft" about rigidness of the PLab spectrometer and proposals how to make the setup more stable. Because I had some stability problems, too, when I put together my PLab Spec. 3.0 I found the proposals and the comments very interesting. One point in this discussion was that results should be repeatable (even for different users with different spectrometers) so that for instance an oil pollution spectra database could make any sense at all.
As I said I am not a scientist, but this discussion urged me to find out how repeatable the results of my own spectrometer were. So I took my table lamp (IKEA style) and aimed it at the slit of my spectrometer and collected 4 spectra within 13 minutes. There was no additional light from any other source and I did not change anything in the path of the light. (I also calibrated them all with the same calibration spectrum.) Here are the results:
In the first picture I stitched the four spectra together in Photoshop, lined them up and marked the peak wavelengths and the peak intensities of the R-,G-, and B-channels. The "waterfall" image shows that the light intensity of the bulb or of the camera sensor perception was not constant. (When I find out how to switch off automatic gain control of the camera I might repeat this experiment.)
Then I took the CSV files and looked for exact wavelengths of the peak intensities and marked them with their respective colours (R, G and B).
For anyone interested in the whole data, here is the CSV for all channels:
So... what do we make of it? Well, the blue peaks happen at 461 and 470 nm. Thats 9 nm difference. Greens 512 to 518 nm, 6 nm difference. Reds about 579 to 589 nm, 10 nm difference. Average peaks 512 to 518 just like the greens.
As I said: I did not change the path of light. The only ways such differences could occur, could be a) a different temperature of the lamp over time (perhaps because of small changes in electrical current) b) an automatic gain control issue concerning the webcam c) both effects overlaying
Comments or further experiments welcome.
Follow related tags:
spectrometer oil-testing-kit barnstar:empiricism repeatability
Two samples tested today, both prepared with a standard regent recipe of 0.1M NaOH and a 20 percent solution of ethanol (200 proof)
each sample @ 3ml was dissolved in 5ml of regent, both samples allowed to separate naturally through gravity. The top opaque layer was removed using a disposable pipette, samples then transfered to cuvettes (1.5ml capacity each.)
Sample #1- 2 stroke multi-mix motor oil
Sample #2- 87 Octane gasoline
Spectrometer used is Plab spectrometer kit 2.5
Processing software used - Spekwin32 version 1.72.0
slit width - 0.09mm
DVD grating - 8.5 G
Light source - Prolight 1W UV LED [400nm-410nm]
Plots were done by me on excel.
Follow related tags:
spectrometer oil-testing-kit gasoline ethanol