# Spectrum of Chemical with a Glass Prism

1. Sep 11, 2017

### ISamson

Hello,
I was wondering if it is possible to make a definite clear spectrum of a chemical using a glass prism. I have one at home and no matter how hard I try to get an emission or absorption spectrum, it just does not turn out.
Can you help? Do I need to focus the light into one point or such?
Thank you.

2. Sep 11, 2017

To start, you need to make the rays parallel by focusing the rays with a lens onto a narrow slit and then putting this slit at the focal point of a second lens to make the rays parallel. You then pass the rays through the prism, and you focus the emerging rays with a third lens. The pattern should be the spectrum and if there are any spectral lines in the source, they will show up as a single bright line of that color (basically in the shape of the illuminated slit). High quality optics can be helpful, but \$5 magnifying glasses you get at a neighborhood store should work reasonably well for the lenses.

Last edited: Sep 11, 2017
3. Sep 11, 2017

### Staff: Mentor

Also note for simple hardware like the one you are intending to use you need to select a good substance for your spectrum, otherwise you won't be able to see any characteristic signals (or, you won't be able to understand what you are seeing without someone experienced able to comment on the result).

4. Sep 11, 2017

LED's of different colors might make a simple source to test that are quite inexpensive. Also ordinary light bulbs. Otherwise, many of the sources that will contain spectral lines, such as a sodium (Na) arc lamp are more expensive Also a HeNe laser at wavelength $\lambda= 632.8 \, nm$ and other laser pointers could show interesting results as well.

5. Sep 11, 2017

### Staff: Mentor

Sure, you can easily see that red LED spectrum and green LED spectrum are different - but you don't need prism to spot that, so the effect is rather obvious.

What I am aiming at is that when the spectrum is more or less continuous (ie no spectral lines) it will be hard to see differences in intensities, and the result will be rather dull and not that interesting. I remember a "wow effect" when we were shown spectra of noble gases from Plucker tubes back in school. That is rather impossible to reproduce.

6. Sep 11, 2017

The OP is of grade school age (age=12), and I think he is simply trying to get the prism to work for some very simple sources. It's unlikely that he is needing any advanced spectroscopic techniques and/or needing to perform a wavelength calibration. To just get a good rainbow of colors would be a good start. $\\$ The technique in post #2 with the 3 lenses should get him some very good results. A similar kind of optics is also employed in diffraction grating type spectrometers to collimate and then refocus the beam, where spherical or paraboloidal mirrors are often used instead of lenses for the two optical elements inside the spectrometer besides the diffraction grating, which is used instead of a prism.

Last edited: Sep 11, 2017
7. Sep 11, 2017

### Ygggdrasil

8. Sep 11, 2017

### Staff: Mentor

Definitely, and that's not what I am thinking about.

Yes, but try to imagine mindset of an inquisitive teenager. Is seeing a rainbow enough to prove spectroscopy is a valuable technique? It can be easily seen on Om's pictures from the thread Ygg linked to that in most cases the differences between sources are hard to spot. I am afraid most simple sources will be difficult to distinguish. OP will see rainbow here and rainbow there, just a bit darker in places, so the effect will be dull and in no way allowing to determine what the light source is. And if it doesn't allow even qualitative results experiment will be quite disappointing.

Having sources that will give distinctively different spectra would be much better. Having sources with different spectra but difficult to distinguish just by the sight would be even better. It doesn't mean any advanced spectroscopic techniques nor calibration, just well defined bands would do.

9. Sep 11, 2017

### Ygggdrasil

Something like comparing the continuous, blackbody spectrum of an incandescent bulb versus the more discrete spectrum of a fluorescent bulb would be a nice home experiment.

10. Sep 11, 2017

### ISamson

What do you mean as a good substance? What could be one?

11. Sep 11, 2017

### ISamson

What I wanted is to see a spectrum of a chemical and recognise the chemical from the emitted/absorbed lines.

#### Attached Files:

File size:
91.2 KB
Views:
30
• ###### IMG_4994.jpg
File size:
122.9 KB
Views:
30
12. Sep 11, 2017

I think you should get greater spreading of the spectrum than what you saw with the sun by using the technique with the slit and the lenses. You also can rotate the incident angle to optimize the results by putting the prism on a rotating table and turning it. If the prism is made of a glass that has good dispersive properties, you should get quite a lot of spreading of the colors. The one picture of the sun is reasonably good, but the results should get better with the slit and lens technique.

13. Sep 12, 2017

### DrDu

When I was a boy I used a slit made from two blades for shaving, colimated it with a lens from a photo camera and looked at the spectrum with some prismatics.

14. Sep 12, 2017

### Ygggdrasil

I have published data from a microscope that contains a slit constructed from two razor blades :)

15. Sep 12, 2017

### ISamson

16. Sep 12, 2017

### ISamson

How could that work? Similar to a prism?

17. Sep 12, 2017

### OmCheeto

Thanks! This is a really interesting problem, but........

I don't think I could do what you are talking about.

Emission lines are easy, if you're just using light bulbs. (Black compact fluorescents are the most fun.)
But absorption lines from a random chemical? Um, not in a hundred years could I do that.
And emission lines from a random chemical? That would require either burning the chemical or heating it up to astronomical temperatures. This is also beyond my capabilities.
Well, ok, I could burn the chemicals, but I might burn my house down in the process, so I'm not going to try it.

ps. Here are the comparative results of my experiment I just finished a few minutes ago:
Equipment:
Light source: black compact fluorescent lamp (mercury vapor)
plastic prism
1000 lines/mm diffraction grating​

mercury vapor emission lines through the prism

mercury vapor emission lines via diffraction grating​

My prism wasn't able to cast enough light onto a white screen, so I was forced to photograph it directly at the prism, losing two of my spectral lines. And the ones that are present, are not very well defined.

If you look through wiki's entry on the history of spectroscopy,

Early 19th Century (1800 - 1829)
In 1802, William Hyde Wollaston built a spectrometer, improving on Newton's model, that included a lens to focus the Sun’s spectrum on a screen. Upon use, Wollaston realized that the colors were not spread uniformly, but instead had missing patches of colors, which appeared as dark bands in the sun's spectrum. At the time, Wollaston believed these lines to be natural boundaries between the colors, but this hypothesis was later ruled out in 1815 by Fraunhofer's work.

Joseph von Fraunhofer made a significant experimental leap forward by replacing a prism with a diffraction grating as the source of wavelength dispersion.

you'll see that technologically, prisms were inferior to diffraction grating. So I would be surprised if anyone could do what you are asking, with even the best spectrums in the world.

This is just a guess of course, as I'm not really a scientist.

18. Sep 12, 2017

### DrDu

No, you need the prism, too.

The slit has to be in the focal plane of the objective.

19. Sep 12, 2017

### Ygggdrasil

Sure: https://www.nature.com/nsmb/journal/v17/n12/full/nsmb.1937.html
If you can't access the paper at the journal's site, it's also available here: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3058889/

Basically, I was doing ratiometric imaging to use FRET as a way to measure nanometer-scale distance changes between two fluorescent labels in an enzyme complex during catalysis. I used a dichroic mirror to split the red light from my sample along one optical path and the green light along another optical path. The slit was there to crop the image so that the two channels would fit onto two halves of my camera chip, so I could image both colors simultaneously.

20. Sep 13, 2017

### Andy Resnick

I can add a few details to this thread. First, it's true that gratings have higher dispersions than prisms, and so gratings are the preferred technology. You could try using the reflection off a DVD to obtain pretty good spectra.

Regarding the OP, the quick answer is 'probably not'. The more detailed answer depends on what chemical(s) you are interested in. While atomic absorption lines are indeed in the visible part of the spectrum, molecular absorption bands (what you need for chemical identification) are in the infrared, with the usual wavelength range being from about 3 microns out to about 100 microns- chemical identification comes from locating the vibrational absorption bands (https://en.wikipedia.org/wiki/Infrared_spectroscopy). Besides the fact that you can't see in the infrared, the optical dispersion of your glass prism in that waveband is not the same as the dispersion in the visible.

Given that this is a homebrewed curiosity-driven activity, I tend to agree with @Charles Link and @OmCheeto, get some LEDs or light bulbs and play around. Spectrometer design is fairly simple- http://www.lcas-astronomy.org/articles/images/specscp1.gif is a basic schematic. Post your results!