Thus the HSQC data confirm that, compared to Pristimerin, LGJC3 has one additional CH3 group (a methoxy) and one oxygenated quaternary sp3-hybridized carbon that corresponds to a sp2-hybridized carbon in Pristimerin.

The only upfield (sp3-hybridized without oxygen) CH group in the spectrum (Fig. 11.56) is the positive peak c18/h18. This would correspond to the aliphatic CH group C-18/H-18 at the D/E ring juncture in the Pristimerin structure. In the mid- and downfield regions (Fig. 11.57), we see the methoxy CH3 groups 20 and 21 and the olefinic CH groups 24, 25, and 26, all as positive crosspeaks.

The primary value of the HSQC spectrum is that it provides us with an accurate list of all 1H and 13C chemical shifts. This would not be possible from 1D data alone. This list (Table 11.1, columns 1, 2, and 4) is then our reference for assigning correlations from other 2D spectra: HMBC (2-3 bond C to H), DQF-COSY (2-3 bond, occasionally longer, H to H), and ROESY (< 5 A in space, H to H).We can use this list by sorting it into two lists: one in order of1H chemical shift and one in order of13 C chemical shift. If, for example, we have an HMBC correlation with F1 = 30.9 ppm, we would look down the list of13C shifts until we find something close to this value: c8. This is easy if the list is sorted by 13C shift, and we can then look to the "neighbors" in the list to see if there is any ambiguity in the assignment. If there are other carbons very near to 30.9 ppm (e.g., c7 at 30.6 ppm), we will have to consider them as possible assignments as well; for example, we might assign the crosspeak "c8/c7." Keep in mind that the resolution of 2D spectra is considerably lower than that of

1D spectra—especially in the F\ dimension. Two peaks may be clearly resolved in the 13 C spectrum but the corresponding crosspeaks in the HMBC spectrum may be too close to be distinguished. Another way to use the HSQC spectrum is as a graphical "list" of chemical shifts: many modern 2D data processing software packages allow the user to display more than one 2D spectrum and align them so that a crosshair can be used to precisely compare peak positions in the horizontal (F2) and vertical (F1) dimensions. To get a feel for this capability a number of the figures below show a portion of the HSQC spectrum above or to the side of the 2D spectrum (HMBC, COSY, ROESY, etc.) of interest so that the peaks can be visually aligned. This makes it clear when there is overlap or near overlap so we can determine if an assignment is unique (only one resonance at that chemical shift) or if we need to put down multiple possible assignments. Accurate referencing is essential and all spectra need to be acquired with the same temperature, solvent and concentration.

The assignment list (Table 11.1) includes lots of other information that will be discussed as we come to the other correlation spectra. Carbon shifts come initially from the HSQC but are then replaced by the more accurate values from the 1D13C spectrum. Multiplicity, where indicated other than "m" (multiplet) or "s" (singlet), comes from 1D slices of the HSQC peaks along F2, so it only picks up the large (>4 Hz) couplings. The only exceptions are the olefinic protons and h8a and h6a, which are resolved in the 1D *H spectrum (Fig. 11.54).

11.10.4 2D HMBC spectrum (Fig. 11.58)

The figure is an expansion of the lower right corner of the full spectrum. Note that quaternary carbons will show up in the HMBC spectrum, as long as there is a hydrogen within two or three bonds of the carbon. Using our two chemical-shift lists, the HMBC crosspeak at 5h = 3.30, SC = 86.6 (Fig. 11.58 center left) can be assigned to h21 and c23, indicating that the CH3O group (ether methoxy) is connected to the oxygenated quaternary carbon (three bonds from the CH3O hydrogen to the quaternary carbon):

This quaternary carbon (c23) also correlates to the singlet methyl group h3 (1.53 ppm), which correlates to two other downfield quaternary carbons: c27 (Fig. 11.58, center right) and c32 (Table 1). One of the singlet methyl groups of Pristimerin (C-23) is unique in that it is bound to an sp2-hybridized carbon and should give a XH peak around 2.1 ppm (compare to acetone, toluene, etc.). None of the carbon-bound singlet methyl groups of LGJC3 gives a XH peak downfield of h3 (1.53 ppm), so it appears that the "missing" sp2-hybridized carbon in LGJC3 is derived from C-4 of Pristimerin (Fig. 11.53), which was changed into the sp3-hybridized c23. This is consistent with the downfield quaternary carbons expected in the A ring (c27, c32 = C-3, C-5) that are correlated to the singlet methyl group C-23/c3.

For a more systematic approach to solving the puzzle, a large number of these correlations was tabulated and are included in the chemical-shift table from the HSQC spectrum (Table 11.1, column 6). Using the HMBC data we can now begin to construct structural pieces of the molecule. Since we have a very good idea of the general type of structure, this is an easy process, but for the sake of illustration we will proceed at first without considering the Primisterin skeleton, as if we were dealing with a true unknown. The 1D 1H spectrum (Fig. 11.54) shows eight singlet methyl groups, six of them in the upfield region. These are excellent "handles" to start our analysis because they give very strong HMBC

INVERSE HETERONUCLEAR 2D EXPERIMENTS: HSQC, HMQC, AND HMBC Table 11.1. Chemical shift assignments for LGJC3 in CDCl3

No. 1H (ppm) Mult. 13C (ppm) Type HMBC Carbons COSY Protons
0 0

Post a comment