Skip To Content

How to deal with ambiguity in an HMBC spectrum? … Part 1

Generally, a 1H-13C HMBC experiment offers a wealth of connectivity information about an unknown structure(s). However, an elucidator may be faced with the issue of ambiguity in assigning a correlation(s) to a 1D resonance(s).

A correlation with an ambiguous assignment may be assessed in one of 5 ways:

-reprocess the HMBC data using a different weighing function to try and narrow down an assignment,

-look for additional information among the other experiments to confirm one assignment over an other,

-ignore the ambiguous correlation and see if there is adequate HMBC data to continue the elucidation,

-make an assumption and assign the correlation to a 1D resonance at the risk of being incorrect, or

-consider multiple assignments at the risk of complicating the elucidation process.

The case below illustrates a mild form of ambiguity in an HMBC spectrum. (A subsequent blog will demonstrate how to deal with a severe case of ambiguity.) The HMBC spectrum exhibits 2 adjacent CH2 resonances, at 41 and 42 ppm, whereby one or both carbons are correlated to the 1H resonance at 2.37 ppm. From afar, the correlation appears to be linked to both carbons. However, upon closer examination of the correlation, the carbon at 42 ppm seems to be the better choice.

Hmbcambiguity_july182008 Hmbcambiguityspec2_july182008

TIP: Zooming-in on a correlation can sometimes help resolve the uncertainty associated with the assignment.

2 Replies to “How to deal with ambiguity in an HMBC spectrum? … Part 1”

  1. There seems to be a small tilt in the pattern. Is this an
    artifact or can be drawn any information from this ´upward
    roof´ ?

  2. Hello hko,
    I had to consult a fellow expert in NMR data acquisition. Until I receive further notice, I will leave the name as anonymous.
    Generally speaking, tilting in the F1 frequency domain is related to some type of evolution going on during the delays in the pulse sequence when 13C chemical shift information is being encoded into the data matrix. The most obvious example are responses that involve fluoro couplings, in which cases the tilts can be significant, ranging up to ~255 Hz, the size of the 1JCF coupling. The other source of tilting in F1 can be due to 1H chemical shift evolution during the duration of the delay. Generally, the F1 digital resolution isn’t high enough to worry about that or to see it.
    The tilt of the proton response is unrelated to what’s going on in the F1 frequency domain of a GHMBC spectrum. Rather, the proton tile is a function of the size of the homonuclear proton-proton coupling constant and the difference between the chemical shifts of the two coupled proton resonances.
    The point is illustrated on the following blog:


Your email address will not be published.