Gradient HMBC spectra

Since the C NMR spectrum of the unknown compound is not more congested than its H spectrum, HMBC is the experiment of choice to establish the longer range C-H correlation networks. Standard and gradient HMBC spectra of the 15-mg sample of the unknown compound are shown in Figure 7-18, and the data from these contour plots are summarized in Table 8-4. 

Figure 7-18 HMBC spectrum (left) and gradient HMBC spectrum (right) of T-2 toxin.

On the other hand, Vch coherences which survive the BIRD filter have evolved as HxCz and HyCz AP coherences. They will be therefore labelled together with long-range coupling coherences and would give rise to unwanted [fc i residual peaks in the HMBC spectrum. In the course of the filter delay 8, fortunately, these HxCz and HyCz AP coherences evolve back for the most part into IP coherences Hx and Hy. They are therefore not transformed into multiple quantum coherence by the 90°z 13C pulse and are filtered out by the selection gradients Gi, Gz, and G3. 

While the 2/,3/-HMBC experiment, also known as STAR-HMBC, has undeniably its merits, it also suffers from a severe sensitivity penalty which results from the extended pulse sequence and additional delays, pulses and gradients as compared to the standard HMBC sequence. For instance, the 2/,3/-HMBC spectrum shown in Figure 21 has been recorded using 64 scans, for a total experimental time of 184 min, while for obtaining approximately the same signal-to-noise ratio, the corresponding HMBC spectrum could be recorded with only 8 scans and 23 min. 

Fig. 4. Gradient-selected HMBC spectrum of compound 11 in CDCI3.

Fig. 15. Comparison of HMBC spectra for a 20 gg sample of retrorsine (3) dissolved in 150 pL rf4-metlianol in a sealed 3 mm NMR tube. The data shown in both panels are 8 Hz optimized non-gHMBC spectra. The spectrum shown in Panel A was acquired in 15 h using a 5 mm 500 MHz cryogenic gradient inverse triple resonance. Almost all of the expected resonances are observed when these data are compared to those for a 700 pg sample of 3 shown in Fig. 2. In contrast, the spectrum shown in Panel B, which was acquired with identical conditions using a 3 mm gradient inverse triple resonance probe, shows the most prominent responses in the spectrum and only a relatively small number of the other responses expected. For a sample of this size to yield a useful HMBC spectrum, it would be necessary to acquire data for a weekend when using a conventional 3 mm NMR gradient inverse-detection NMR probe.

Load the file ch5815cfg. Run a simulation of the g-BIRDR-HMBC experiment for the given 13CHH spin system. In 2D WINNMR use the Slice panel button to extract row 124. Save the ID spectrum. Replace the current pulse program with the file ghmbc2.seq and repeat the simulation using the standard gradient-HMBC experiment with a low pass filter. 

Figure 25 H2BC spectrum recorded on cyclosporine (left), compared with a standard gradient-selected HMBC (right).

Fig. 5. HMQC spectrum of a 0.55 pmol sample of cryptolepine (1) dissolved in 30 pL d6-DMSO. The data were acquired in a 25.5 h experiment using a 1.7 mm gradient triple resonance submicro-NMR probe at 600 MHz. The minor correlation responses in the spectrum designated by arrows are an 8% ( 12pg) impurity of cryptolepine A-oxide present in the sample. All of the impurity responses were visible in a 12 h contour plot. Full HMBC data consistent with the structure were acquired for the impurity in 55 h. (Reprinted with permission from Ref. 12. Copyright 1998, American Chemical Society and American Society of Pharmacognosy.)...

Figure 6.25. The HMBC long-range correlation spectrum of 6.2 recorded with A = 60 ms and with gradient selection. The sequence used the low-pass J-filter (Section 6.4.1) to attenuate breakthrough from one-bond correlations (which appear with Jch doublet structure along fj (arrowed)). IK ta data points were collected for 256 ti increments of 8 transients each and the data processed with unshifted sine-bells in both dimensions, followed by magnitude calculation. After zero-filling once in ti the digital resolution was 4 and 80 Hz/pt in f2 and fi respectively.

The structure and composition of Chocarome pyrazine was elucidated by a combination of GC, MS and NMR analyses (Figures 7 and 8 and tables III - VI). The substitution pattern of 3,5-dimethyl-2-isobutylpyrazine and 3,6-dimethyl-2-isobutyl-pyrazine was elucidated by a combination of NMR methods, especially in mixtures by gradient selected H, N HMBC experiments at natural abundance level (I /). Since the N-NMR spectrum of pyrazines was proven to be a powerful tool for their structural elucidation 12, IS) it is used to study these type of molecules. Tables V and VI details the N-NMR chemical shift of both nitrogen atoms of each isomer ... 

When collected in a phase-sensitive mode, HMBC cross peaks are found to have a mixed phase character. That is, we cannot phase HMBC cross peaks so that they are purely absorptive. The use of pulsed field gradients for the purpose of coherence selection in the HMBC experiment (gHMBC) renders a nonphase-sensitive 2-D data set. This latter method is generally preferred because phasing of the spectrum is not required. 

Figure 4 (A) Two-dimensional gradient selected HMBC NMR spectrum of 2-benzyl-1-butyl-2,3-dihydro-imidazo[1,5-c]...

---