The first sample was a chiral molecule modified by attachment of a 2-propynyl group. To be consistent with the presence of the propynyl group the apparent quartet of doublets in the spectrum below was assigned as two doublets of doublets (at 4.37 and 4.32 ppm) due to the two protons attached to the carbon adjacent to the oxygen.
In this assignment, coupling between the geminal protons gives the large splitting of 15.8 Hz, long range coupling to the proton on the far side of the triple bond produces the small 2.2 Hz splitting, and a chemical shift difference of 25 Hz for the two geminal protons leads to strong coupling that distorts the multiplet intensities so that the "outer" lines are reduced in intensity while the "inner" lines are increased.
Two other samples showing strong coupling came from a series of compounds with conjugated double and triple bonds. In one compound the protons on either side of a double bond produced what looked like a quartet of triplets, but was actually two doublets of triplets. Close examination of the multiplets shows that the small coupling on the two right lines is less than the coupling on the two left lines, a hint that this is two resonances showing strong coupling to each other.
The allylic resonances in another compound from this series produced a quite different spectrum. Here a modification in another part of the molecule reduced the chemical shift difference of the allylic protons, enhancing the strong coupling. Now the outer lines have nearly disappeared and the inner lines are nearly overlapping.
Strong coupling even more extreme than the example above is why a methyl group produces what looks like a singlet. Since the chemical shifts of all three protons in a methyl group are identical the scalar coupling between them is much larger than the coupling. The intensity of the outer lines is then reduced to virtually zero, while that of the inner lines is increased. Since the inner lines all overlap, a singlet is obtained.
As strong coupling appears when the chemical shift difference approaches the scalar coupling it is more often seen at lower fields, where the chemical shift range in hertz is less. Still, these examples show that even at 600 MHz strong coupling can occur and should be taken into account when you are trying to assign your spectra.
Acknowledgements
Thanks to Sam Kantonen of the Gilson lab and Xiao Wang of the Molinski lab for use of their spectra.
References
1. Reich, Hans J. "5.9 Second order effects in coupled systems" Structure Determination Using Spectroscopic Methods, University of Wisconsin http://www.chem.wisc.edu/areas/reich/nmr/05-hmr-09-2ndorder.htm
2. Jacobsen, Neil J. "NMR spectroscopy explained", Wiley Interscience 2007
https://books.google.com/books/about/NMR_Spectroscopy_Explained.html?id=KCkiiQ0uefoC
(examples pp 63-69, density matrix explanation of origin pp 481-484)
Always enjoy your blog - and this month two of our natural products appear as 'celebrity molecules'!
ReplyDeleteI have one correction to offer. In the paragraph describing the spin system of your example you write "vicinal protons gives the large splitting", but I think you mean "geminal" coupling. The same substitution should be made later in the paragraph. Also, it's worth pointing out that the chemical shifts of each signal in strongly coupled systems (like the ABX of your propargyl ether example) are not at the arithmetic mean of the outer and inner lines, but at the 'center of gravity', closer to the inner (taller line). Rather than use the formula, I simply tell my students to 'guestimate', based on the relative line intensities, using symmetry as their guide (unless more strong couplings are present - then bets are off!), so I would estimate the two chemical shifts as closer to 4.36 and 4.33 ppm.
As you know, Xiao and had 'fun' trying to interpret the structure of the third example for two reasons: the outer lines were not at first noticed and the 2-proton AB part integrated for only 1.5 H. Xiao thought there was only one H in the signal. Turned out, the apparent T1 for each is about 4.5s (no degassing of sample) due to poor spin lattice relaxation efficiency and partial saturation on rep scans - see our manuscript!
Thanks for the comments Ted. I have corrected my vicinal/geminal mistake. Perhaps I should get you to proofread before posting!
DeleteSince my background is in protein NMR I have not had to deal with strong coupling situations much before. I only read up on it when I was trying to help Sam with his molecule and noticed Xiao's spectra with a similar situation. You are probably correct about the chemical shifts in Sam's compound. The best solution would be to fit the experimental data and obtain the shifts that way.
The formula for accurate chemical shifts of AB systems is as follows: Measure the frequencies (not ppm!) of each of the four lines, labeled here in order '1', '2', '3' and '4', and the AB coupling, J. If ∆v is the difference between chemical shifts of A and B, then (1–3) = (2–4) = square root of [∆v**2 + J**2]. Resonant frequencies of A and B will be located ±∆v from the center of the pattern. Works approximately for ABX if J(AX) and (JBX) are not to large - replace line positions with center-weighted positions of the A and B components.
ReplyDeleteExcellent! Thanks Ted!
ReplyDelete