2D NMR

2-Dimensional NMR

The example shown on this page is presented at this page as you will see your own unknown.

1. The mass spectrum is always the place to start; you must find out what the substituents are (and generally, what the molecular formula is):

All of your unknowns are disubstituted benzenes, so you know part of the formula is C₆H₄. You want to find the largest nominal mass (based on an understanding of the presence of heavy isotopes), look for characteristic patterns for Cl or Br, examine major fragmentations by differences with the parent ion mass and identify what is present in addition to the aromatic core.

2. The one-dimensional ¹H and ¹³C spectra of 2-chlorotoluene:

To the extent you can, you want to analyze any clearly separated peaks for their coupling constant patters and extract J values. For your unknowns, the most common ones to look for are dd, td, and occasionally ddd. Coupling to fluorine will complicate each peak (note: the presence of an ¹⁹F spectrum in your packet is a big hint that fluorine is present!).

3. The 2D Experiment:

Image courtesy Harvard Medical School.

A 2D COSY (COrrelated SpectroscopY) spectrum. The off-diagonal peaks arise from coupling between protons; the higher the peak, the stronger the coupling (the larger J is). In addition to coupling among the aromatic signals (see expansion, 3 below), we can see some very weak coupling between the methyl signal and an aromatic proton. This is <<1 Hz. The "normal" 1-D spectrum appears along the diagonal.

Top: a "stacked plot" view, emphasizing the 3-dimensionality of the spectrum.
Bottom: a more conventional contour plot; the view is "looking down" on top of the stacked plot.

4. A closeup of the aromatic region showing ¹H-¹H coupling. The ortho and meta couplings are easily detected; some very weak para coupling is barely visible.

5. An HSQC (Heteronuclear Single Quantum Corelation) spectrum showing ¹H-¹³C coupling. Bottom: a closeup of the aromatic region.

5. A final very useful feature is the ability to look at a "slice" of the HSQC spectrum. This shows the multiplet pattern of the proton attached to the carbon we are examining. We use the ¹³C signal to reveal the spectrum of only the ¹H attached to it! This allows us to pull apart overlapping multiplets. Be careful, though, since low digital resolution will prevent us from seeing small splittings (<5 Hz).

6. The final piece of evidence is the HMBC: Heteronuclear Multiple Bond Correlation spectrum. This can show cross peaks for 1-bond, 2-bond, 3-bond and even 4-bond coupling between ¹H and ¹³C, and can be confusing. The 1-bond couplings are easiest to find: they are typically doublets in our unknowns, and the J value depends on the carbon hybridization:
sp³: 125 Hz
sp²: 160-180 Hz
sp: 250 Hz
These should match with what you see in the HSQC.
The most usual HMBC correlations other than these are 3-bond couplings, and you want to start out assuming that cross peaks arise from these. 2- and 4-bond cross peaks are usually weaker and often questionable as to whether they are a "real" peak.

In the overlaid spectrum below, the red peaks are the HSQC peaks, and the black are from the HMBC. You can identify all four 1-bond doublets; the remainder (with one exception--a weak 4-bond peak) are 3-bond couplings. From analyzing this (in conjunction with the COSY and HSQC) you can definitively identify the structure and provide a complete assignment for each line in the spectra.

Back to NMR page

Back to CH 362 Home page

Last updated: 12/17/2014