On Sunday, October 3, 2021 at 4:50:22 PM UTC-6, RichD wrote:
On October 3, M. Roussel wrote:
How do chemists determine the precise locations
of individual atoms within a molecule?
For atomic physics, the only thing we can observe
is spectra. Bombard the sample with radiation,
modulate the temperature, read the spectra. But
we can't micro-photograph and see the locations
of the components.
For instance, fatty acids:
https://tinyurl.com/fatty-acid00
Ian is correct that x-ray diffraction is one of the most satisfying techniques in terms of
getting structural information.
That's a special case, restricted to to crystals.
True, but a lot of things can be crystallized, and in many cases, the structural information is exquisitely precise.
Even a simple proton NMR tells you a lot about how the atoms in a molecular are connected.
Add to that fancier NMR experiments that give you atom-to-atom distances like the nuclear
Overhauser effect, or experiments that give direct connectivity information like INADEQUATE
and suddenly a whole world of structural characterization opens up with the need to crystallize a compound.
What is the output of the NMR, how does the chemist reconstruct the structure of the molecule?
Are they solving Schrodinger's equation?
No. The data you get from NMR can be more or less complicated. In a proton spectrum, you get resonant frequencies, with splitting of peaks that depends on what a given atom's neighbours are (roughly speaking). Then you go through the puzzle-solving
activity of figuring out what the structure must be given the splitting pattern, aided also by the knowledge that certain functional groups reproducibly show up near certain frequencies. The nuclear Overhauser effect gives you atom-to-atom distances, so
again, figuring out the structure is a puzzle-solving activity (often sorted out by a computer program, which reduces the puzzle to an optimization problem). Two-dimensional experiments like INADEQUATE sort the signals along two dimensions. INADEQUATE in
particular gives you connectivity information in addition to the peak splitting. More puzzles to put it together...
Then there is mass spectroscopy. Because typical mass spec experiments break up molecules
into pieces, the masses of the pieces tell you something at least about local connectivity.
Again, do they have mathematical models, using quantum mechanics?
Nope. More puzzles to solve. If you get a piece that has the mass of a methyl group, provided you know that nothing else in your system would have that mass (methyl = 15, so it could be confused with an NH fragment in theory, although most fragmentation
methods wouldn't make that), then you know that somewhere in your molecule there is a methyl group. You would typically start out knowing the empirical formula of your compound (from elemental analysis). You also know something about how molecules break
up under the particular ionization conditions that your mass spec uses. Mass spec doesn't usually give you quite enough to figure out the structure of a compound, but combined with other methods (typically spectroscopic methods), you can often puzzle it
out.
X-ray crystallography is very direct. NMR is, if not direct, sufficiently powerful that you can really nail down the structure if you're willing to do enough experiments. Other methods (mass spec, IR spectroscopy, etc.) tend to give more indirect
information, so you often have to put several of these techniques together to get a structure.
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)