Finding the atomic makeup of a substance

[PhiPhenomenon]

Hey all,

This is actually my first post on this forum but I have been reading it for a few months now.

Anyway, I'm sure the title of this thread might be confusing but this question has been bugging me for years now and I'm sure I'm missing something quite elementary. My question has to do with how exactly do people find the atomic makeup of an unknown substance and actually know enough to talk about it's bond structure.

For instance, I'm just imagining we have a bucket of ethenol (C2H4O1). I don't know what it is exactly and for all I know it could be just ethanol (C2H6O1) with a single sigma bond between the two carbons. How would I figure this out? For that matter, how do people know for certain that it is made up of carbon, oxygen and hydrogen?

 

[DDTea]

Besides many "wet" chemical tests that can be used to qualitatively identify functional groups and certain atoms (e.g., by precipitation, formation of colored derivatives, etc.), there is an entire field known as spectroscopy which encompasses a huuuuge range of test methods. To name a few: atomic emission spectroscopy, atomic absorption spectroscopy, UV-Vis, Fluoroescence/Chemiluminescence techniques, Mass spectrometry (this in itself is a huge field), Nuclear Magnetic Resonance (besides MS, this is the most powerful technique for determining organic structures), Infrared, Microwave, Raman, X-Ray diffraction, X-Ray Fluorescence, X-Ray Photoelectron Spectroscopy...

the list is massive and each technique gives certain information about a molecule. If I wanted to determine its molar mass, I would use Mass Spectrometry with a non-destructive ionization method. If I wanted to determine bonding patters, I would use MS with a destructive ionization pattern or better yet, something called tandem MS. Alternately, there is NMR, but that can be a finicky technique sometimes (but otherwise, it's EXTREMELY powerful, especially 2D or 3D pulse NMR experiments). For quick tests to look for functional groups, Infrared or UV-Vis would be a good technique. UV Vis could also tell me something about the orbital energies of a molecule (which is useful when looking at inorganic species).

Seriously, there are books and theses written about this and I can't even hope to touch on all the techniques! For Ethanol, though, I would perform a proton NMR experiment. It's very easy to analyze--you'd see 3 proton signals at different "chemical shifts." The peak height of the signal tells you about their relative ratios, so you'd see 3:2:1 relative peak heights. The specific chemical shift tells you the amount of electronic shielding against the applied magnetic field that is around each proton, and correspondingly, if it's attached to any heteroatoms, electron-donating, or electron-withdrawing groups.

It's a cool field :)

 

[PhiPhenomenon]

Wow, I actually understand that. Thank you! :)

I've been reading books on the subject for a while now. I'm actually fresh out of school with a BS in psychology (bleh!) and trying to transition into neuroscience.