Public Lab Research note

Classroom Flame Spectroscopy

by straylight | October 14, 2012 09:23 14 Oct 09:23 | #4406 | #4406

A group of students and I put the PLOTS spectroscope to use observing the spectra from solutions of common metal salts that were excited in the flame of a bunsen burner. We calibrated the spectroscope using a flouro in the classroom and identified lines in the spectra of Fe (III), Ca(II), Na(I), Cu(II) and Ba(II). This was an investigation into whether the spectroscope could be used in an educational setting and how hard it would be to observe and identify elements based on their spectra.

bunsen flame rig We used a modified bunsen burner to excite the metal salt solutions which were sprayed into a 2 litre soft drink bottle which fed the spray into the venturi of the bunsen burner. The spectroscope was simply held in a retort stand level with the flame. The solutions were approximately 1M concentration.

We noticed that copper was particularly hard to see in the spectroscope and a wider single slit at the front of the spectroscope would help by letting more light enter. We experimented with different sized slits and noticed that while we lost a little wavelength resolution, the results were still visually good. We had to re-calibrate each time we moved the slit however.

The slits were made from aluminium foil that was sandwiched between two microscope slides. Many slits were made so it was a simple matter of moving the slide and sticking it down over the main aperture into the spectroscope.

The spectroscope was our first attempt and used a cheap 8MP webcam, 3cm autofocus and cost $5. We used DVD plastic and a spare box covered in contact adhesive. The fittings are only held in place using blu-tak, including the DVD over the lens. We only worked in the visible spectrum.

We made a video of our efforts.

I would imagine extending this activity confidently for a middle school class. I would expect to be able to provide the students with the laptops and spectroscopes and have them calibrate the spectroscopes using the simple CFL calibration that is available in spectral workbench (offline). I would then have the students record the identified lines in the spectra of some know solutions. I would then ask the students to identify an unknown that contained up to two salt solutions that they had already identified.

For senior students, I would envisage using the NIST database to identify unknown solutions.


We noticed that using too greater amount of a solution contaminated the apparatus and the residue took a long time to disappear.

We put the spectroscope quite close to the flame and did on one occasion melt it slightly.

The experiment can be done in a lit classroom, but it is better with the lights off. The spectroscope consistently found stray reflections from the flouro lights. Ideally a black backdrop would be preferable behind the bunsen.


A simple capture button would enable the students to save their spectra for printing in a report. A colour printer would be excellent.

An API being proposed is to have sample spectra displayed on the analysis screen, at the same scale as that being captured by the spectroscope, to give immediate feedback and provide qualitative analysis of the spectra being observed.


My students enjoyed participating in the exercise. It would take some practice and good classroom management to do it with a large group of students. Compared to merely "burning" the chemicals in a bunsen burner and if you are lucky, using a handheld ($60) optical spectroscope to see some lines, this apparatus gives quantitative data that is equal to many classroom spectrometers costing thousands of dollars.


We also observed some discharge tubes and got some marvelous looking spectra from Ne, Hg, Cd and Na. I would imagine setting up an exercise where the students observed the spectra from a sodium lamp and then compared it with an "unknown" sample of common salt captured using the flame rig on the spectroscope.

Stu Williams

I did this Help out by offering feedback! Browse other activities for "spectrometry"

People who did this (0)

None yet. Be the first to post one!


can some kind person explain how to put images inline and how to embed the youtube vid, ta !


Reply to this comment...

Hi Stu - sorry, the best way to do inline images is to upload them as attachments, then use HTML IMG tags, a bit cumbersome. We're working on a smoother/better/more intuitive posting system but it's a bit of a big project.

I was so glad to see this post! The setup looks fantastic. I'm eager to try it out. I got a butane kitchen torch which I hope will be hot enough. Need to buy butane to fill it with. We used a similar type of burner to do sodium tests at LEAFFEST a few weeks ago.

What percent were the solutions?

As for the online database, we're working on a couple different ways to do this from your offline setup. Right now, you can upload raw data as image files at, but you do have to re-calibrate them. I can add a "use recent calibration" checkbox to the image upload form to streamline this if it's interesting to you.

Second, I've been toying with the idea of making an "offline folder" -- just all the HTML pages and Javascripts in a folder, which you can open and run on your machine, as if it were a website. This ought to work, and all I have to do is add a "save data locally" button so it doesn't rely on the online uploading function. Then you'd be able to upload the images as described above. The main advantage here is that the interface of the web-based version is much smoother and nicer than that of the older Java based version.

I'll check back in once that progresses a bit. Great video! Thanks!

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

Reply to this comment...

I put together a macro that saves the image offline. Now I need to make a macro system on! (in the "Examples" section)


Reply to this comment...

OK, i got inspired and wrote a basic macro "sandbox console". You can now write simple macros as described at:

One that works right now is:

setup: function() {$W.canvas.toDataURL(),'_newtab').focus()}

which saves an image of the current spectral data waterfall.

Reply to this comment...

This is amazing!

I am also interested in this for the classroom and I have been trying this out myself and had limited success getting enough light from the flame into my phone camera to get a nice response. I think I will just have to find some webcams/laptops and try this setup instead.

One question however...I noticed that the dvd-r had a purple tint to it instead of the normal colorless polycarbonate. The dye is usually purple and I am wondering if you cleaned the DVD off before trying this (isopropanol/ethanol works well). Regardless your lines were quite nice.

Reply to this comment...

"I can add a "use recent calibration" checkbox to the image upload form to streamline this if it's interesting to you."

yes, that would be great. At the moment we can capture a spectra onlne, but unless we have the Hg lines on it, I didn't see way of working out the wavelength of the lines it shows. I thought it was just me being a numbskull and not working out how to save a calibration properly :)

"One question however...I noticed that the dvd-r had a purple tint to it instead of the normal colorless polycarbonate. The dye is usually purple and I am wondering if you cleaned the DVD off before trying this (isopropanol/ethanol works well). Regardless your lines were quite nice."

good point ! a fine observation. You are correct, I didn't clean the DVD plastic at all. I tried to take a slab of plastic without too much dye on it. I'll now go back and clean it, hopefully it will let a bit more light through as well. Thanks for that. I really should read all the instructions LoL.

thanks for the info on images Jeff. I'll get back and edit the note and replace the thumbnail links with the actual images. This is my first time playing with both google groups and this format.


Reply to this comment...

I should add, we have done this experiment using the old turntable spectroscopes using the flame thrower apparatus with good results. We used a proper diffraction grating. It takes some calculation to extract the wavelength of the spectrum from first principles. This method using the PLOTS software is so much better for this purpose.

We've also used a commercial datalogger (vernier labquest) and spectrovis (their digital spectrometer) which also requires the use of logger pro software. All up that set up cost in the order of $1500 (Au). To do it for less than $10 is simply amazing and enables us to use a class set of spectroscopes that the students can construct themselves.

One useful addition that spectrovis has, at extra cost, is an optical fibre connection. This simply fits into the spectrometer where the cuvettes normally sit. We used his to in order to observe flame spectra. I'm thinking of grabbing some optical fibre and seeing if we couldn't craft up a device that we could simply point at the flame or for reflection spectroscopy at, say, a leaf. Feed the image in through the fibre to the slit and it would add versatility to an already awesome piece of kit.


Reply to this comment...

Great work, now I can make a flame specroscope. Love the idea for the apperture. One thing I am wondering. Its hard to see how you got the bunsen into the 2l bottle.i assume it was cut and then paste d. Cheers Brian @brian

Reply to this comment...

thanks Brian,

if you cut a big hole in the side of the bottle and pass the bunsen in through that. Then feed the gas hose in through a smaller hole in the neck of the bottle. I use insulation tape to hold the bunsen in place around the top of the bunsen and top of the bottle. The salt solution is then sprayed through the big hole, underneath the bunsen.

no photoshopping ! :)

the other correction I have to add is to not clean the DVD plastic grating. The lines are imprinted on the purple dye, cleaning it off all but destroys the spectrum.


Reply to this comment...

Thanks for finally writing about >Reply to comment | publiclaboratory.

org <Loved it!

Reply to this comment...

Hi! Scott Eustis and I took some elements from your great activity and made a draft of an activity template for others to use, which also relates flame spectroscopy to monitoring natural gas flares:

Reply to this comment...

Login to comment.