# How can I figure out which coupling constants match which hydrogen?

I am looking for some guidance assigning the $J$ coupling constants to the corresponding hydrogens. Here is my data: I can obviously see that shift 3.56 is for the methyl group. I can also see that the 4.63 is for the hydrogen adjacent to the methoxy group as it can only couple ($J=\pu{0.8Hz}$) to one hydrogen next to it. Is the approach then to see that the hydrogen at shift 4.04 also has $J=\pu{0.8Hz}$ so this must be adjacent (going clockwise from C of methoxy, next C along)?

• Yes, use the connectivity provided by matching coupling constants. Note that the methylene on the C6 carbon bears a pair of diastereotopic hydrogens, so these will be different. That should also give a hint about which hydrogen is at $\delta$ 3.43. – Zhe Sep 25 '17 at 15:10

## 1 Answer

Coupling constants are mutual. If nuclei $\ce{A}$ and $\ce{B}$ couple and this coupling results in a coupling constant $J_\ce{AB} = \pu{7.4Hz}$ on $\ce{A}$ then $\ce{B}$ will also be split into two signals with a coupling constant $J_\ce{BA} = \pu{7.4Hz}$.

Therefore, you can indeed start at the anomeric hydrogen (which you can identify not only by the coupling pattern but also by its chemical shift) and work your way along the ring, noting whether two hydrogens are in gauche or anti configuration along the way.

• I am confused how to assign the diastereotopic hydrogens to their shifts, please could I have advice on this? – gamma1 Sep 26 '17 at 10:08
• @gamma1 You can’t distinguish the two diastereotopic hydrogens on $\ce{C{6}}$ unless you have some sort of external reference. Typically, one would label them H6.1 and H6.2 or something along those lines. The only small hint you have is the coupling constant which can be used together with the Karpluls curve to determine a dihedral angle. – Jan Sep 26 '17 at 10:56
• So out of 3.79 and 3.98 shifts, I can assign them to either? thank you – gamma1 Sep 26 '17 at 10:57
• @gamma1 Yes, there’s no way to tell. – Jan Sep 26 '17 at 10:59
• As an aside, if you wanted to know the conformation of the molecule, could you use the data to find this do you think? – gamma1 Sep 26 '17 at 11:00