HMQC-TOCSY (3D experiment)
This is a 3D analog of the 2D HMQC-TOCSY, and can resolve connections that are ambiguous due to overlapping peaks. For codeine, this is not necessary because the spectra have minimal overlap, but it serves as an illustration of the experiment and its interpretation.
The first step of the pulse sequence is an HMQC, in which 1H magnetization is transferred to the directly-bonded 13C. 13C magnetization evolves during t1 as a function of 13C chemical shift and 1H-13C coupling. The magnetization is then transferred back to the directly-bonded proton(s). The completion of the sequence is a TOCSY experiment, which transfers magnetization from the "carbon spin-labeled" proton along a chain of coupled protons.
The difference between the 2D and 3D experiments is that in the 2D version, the TOCSY evolution is eliminated, and all peaks related to a given carbon fall along along a single line. In the 3D version, the TOCSY magnetization transfer is "sampled" during the 3rd incremented time delay in the experiment, and the TOCSY information is spread out over the 3rd dimension, so can resolve ambiguities arising from peak overlap in the 2D spectrum.
In the resulting 3D spectrum, the carbon spectrum runs along the Z dimension, perpendicular to the plane of the screen/plot. The data are viewed by stepping through planes of spectra, each plane corresponding to a different 13C chemical shift. At the chemical shift of each protonated 13C, a 2D TOCSY spectrum is observed containing proton peaks for the proton(s) directly bonded to that carbon, and correlated to other protons in its spin system.
The experiment can be performed using different mixing times in the TOCSY sequence. As with TOCSY, the intensity of each cross-peak will vary with mixing time – long range correlations will "grow in" as the mixing time is increased, and shorter range couplings may disappear.
Pairs of vertically-spaced peaks are observed along the horizontal line at the chemical shift of H-9. The peaks occur in pairs due to 13C-1H 1-bond coupling of C-9 to H-9. A pair of peaks is observed at the 1H shifts of (from left to right) Hs 3, 5, 9, 10, OH and 16, and arise from magnetization relayed through the spin system by scalar coupling.
Pairs of vertically-spaced peaks are observed along the horizontal line at the chemical shift of H-10 (4.9 ppm). The TOCSY peaks are to Hs 3, 5, 9, 10, OH and 16 (from left to right). There is an intense vertical noise stripe at 3.8 ppm due to the large OMe singlet.
TOCSY peaks are seen from H-3 to H-5 and H-16.
TOCSY peaks are seen from H-11 to H-18, H-16 and H-18′. Another vertical noise band is seen at 2.4 ppm, due to the intense singlet from N-Me.
TOCSY peaks are seen from both H-18 peaks to each other, and to H-11.
TOCSY peaks are seen to H-3, H-5 and H-11.
Note that there is one other difference in the appearance of the TOCSY spectrum in any given plane, as compared to a 2D 1H-1H TOCSY: The symmetry about the diagonal, normally seen for homonuclear spectra, is absent in the planes of the 3D HMQC-TOCSY.
In a homonuclear experiment such as COSY, the first 90º pulse excites all protons. During the evolution time, magnetization that started out on proton A is transferred to its coupled partner, proton B, and vice versa. A crosspeak appearing on one side of the diagonal is due to proton A magnetization transferring to proton B, and the symmetrical peak on the other side of the diagonal arises from proton B magnetization that transfers to proton A.
In the 3D HMQC-TOCSY sequence, the HMQC part serves as a "filter", meaning that only magnetization that started out on proton A exists at the start of the TOCSY part, and is then transferred to other protons in the spin system. The reverse process, involving transfer of proton B magnetization to proton A, gives contour peaks on a different plane of the 3D spectrum.
The sample is 45 mg codeine in ~ .65 ml CDCl3
512 complex points in direct dimension
128 t1 increments in the 13C dimension
64 t3 increments in the 1H dimension
2 sec. relaxation delay
Total acquisition time: ~ 40 hrs
sine squared window function in the direct dimension with 0 degree phase shift
5 Hz exponential multiplication in the 13C dimension
exponential multiplication and 1x zero-fill in the 1H
magnitude calculation (no phasing is required)
final data size 512 x 128 x 128 = 65 Mbytes
See also: detailed description of 3D processing
Interactive view – Those who have installed the Chime plug-in can view and manipulate the codeine structure in 3D, examine and expand the 1H spectrum, and interactively correlate 1H peaks with H atoms in the structure.
Last updated: 01/22/2003