Public Lab Research note

Idea For Detecting Specific Emissions.

by amysoyka | July 02, 2014 12:34 02 Jul 12:34 | #10646 | #10646

What I want to do

Try to detect specific elements/molecules using acrylic paint to filter waves of light down a piece of optic fibre & into a light sensor.

Questions and next steps

Why Acrylic Paint? Because it is Translucent - so will hopefully only let light that is the wavelength targeted enter the optic fibre & be detected by the light sensor.

What next? Updating this page with a better explanation of the science involved & then performing experiments to see if it works!


Interesting.. with the oil testing kit we've been using mineral oil. Might that serve a similar function?

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

Reply to this comment...

The purpose of the paint is to act as a filter so that only the exact colour wavelength that matched the colour will be absorbed and trigger the light sensor.

Not sure if you mean to tint the mineral water and use it as a collector, or something else? Could you elaborate please?

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

Reply to this comment...

Ohh I understand. Ok I might get this wrong (I'm still new to the science) but we were using the mineral oil as a sort of base sample. We know how much it fluoresses on its own (not much) and then we can tell our samples on top of it. So when we mix the oils in (which are often solid) we should be able to ID the oil. Someone correct me if I'm wrong.

Reply to this comment...

So, the translucent paint would provide a filter and you'd select the paint color to select the spectral wavelength to detect? ...which is similar to using colored plastic filters or food coloring to let "just one color through" ...

I suspect that the acrylic material has a very broad frequency pass-band -- meaning it will not be very wavelength specific -- meaning it will be difficult to detect one 'blue' from another 'blue' strictly by output intensity -- 'red' might be orange, red and infrared. However, experiments are a good way to find out.

If you have a PLab spectrometer, you could do a simple test. Capture a spectrum using white light from a broad-band source like a halogen bulb. Then, insert your acrylic painted material between the light source and the spectrometer and observer the difference. Is the result a nice sharp peak or a broad curve?

Cheers, Dave

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

Reply to this comment...

hmmm... neat idea. Some paints have aniline dyes, but many higher quality paints have just pure mineral pigments. the spectrum of the reflectivity of the mineral should be obtainable and confirmable with a spectrometer.

Reply to this comment...

If you look at some of Chris Fastie's work on the Infragram technique, he's posted a lot about theater gels as a way to filter certain wavelengths -- and indeed, the cheaper materials are typically not very precise. (

However, what about filtering after the diffraction grating, by making a "mask" of very opaque material (maybe printed on acetate or transparency paper, like Dave Stoft's done: and literally placing it such that it blocks parts of the spectrum once it's been separated out?

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

Reply to this comment...

good work amy, I like the thinking.

I've had the optic fibre tactic on the backburner now for a while, I have the fibre, just need to work on a lens to make it all work. I have used the vernier labquest and spectrovis apparatus with a fibre cable and it adds a lot to the capability of the machine.

I'm not sure that using a filter is that useful in what you are describing though. The great ability of the spectroscope is to act as a filter itself, by spreading out the spectrum of a range of wavelengths spatially. If you want to filter, then you can simply chop off the sides of the display that you don't need. Putting a colour filter in would be needed if you have a simple photometer, say in AAS. The spectrometer gets around this by virtue of its design. Let me know if I have misunderstood your technique or if I'm being obtuse !

well done again


Reply to this comment...

Login to comment.