Info

Proton (ppm)

Figure 11.15

11.4 EXAMPLES OF EDITED, 13C-DECOUPLED HSQC SPECTRA

We have already seen the effect of 13C decoupling on 1D spectrum of a 13C-labeled compound. In 2D HSQC or HMQC spectra, we can just add 13C decoupling (e.g., GARP) during the acquisition of the 1H FID. The wide pair of crosspeaks collapses into a single crosspeak at the 1H chemical shift in F2, with twice the intensity. The main practical consideration is to limit the duty cycle (percent time on) of the 13C decoupling. Duty cycle is defined as AQ/(recycle delay) where AQ (acquisition time) is the length of time for acquiring the FID and the recycle delay is the total time for a single scan or transient, including the relaxation delay, the pulse sequence, and the acquisition of the FID. The duty cycle can be limited by setting a minimum relaxation delay (e.g., 1.0 s) and a maximum (e.g., 220 ms) setting for the acquisition time (Bruker: aq; Varian: at). This way the duty cycle can never exceed 0.22/1.22 = 0.18 or 18%. The acquisition time of the t2 FID depends on the spectral width in F2: td(F2) = 2 x swh x aq (Bruker) or np = 2 x sw x at (Varian). Here swh is used for the Bruker parameter "spectral width in hertz" because the sw parameter refers to spectral width in parts per million on the modern (AMX, DRX) Bruker instruments. After setting the spectral width in F2, the aq/at parameter is checked, and if it exceeds the safe limit (e.g., 220 ms) for 13C decoupling, the number of points (td(F2)/np) is reduced until aq/at is within the limits. As we saw with INEPT and DEPT, refocusing of the initially antiphase transferred coherence is required if we are to apply decoupling: Decoupling of an antiphase coherence puts the positive and negative lines on top of each other, resulting in a signal of zero.

Another very useful technique is to "edit" the HSQC data just like an APT or DEPT spectrum, so that the sign of the crosspeaks gives us information about the number of protons attached to each carbon: CH3 crosspeaks are positive, CH2 crosspeaks are negative, and CH crosspeaks are positive. We saw how CH3 crosspeaks can usually be distinguished by their intensity, so this gives us a complete identification of the type of carbon (quaternary carbons are absent in one-bond correlations). Spectral editing is accomplished just like it is in the APT experiment: by a delay of 1/J during which we have 13C SQC undergoing /-coupling evolution. Starting with in-phase coherence on the X axis, the CH 13 C coherence moves from Sx to -Sx; CH2 coherence moves one full cycle for the outer lines of the triplet and not at all for the inner line, landing back at Sx; and CH3 coherence moves 11 full cycles for the outer peaks and 2 cycle for the inner peaks, ending up in-phase on the -x axis (—Sx). The important thing is that the CH and CH3 coherences are now opposite in sign to the CH2 coherences. The phase is corrected in data processing to make the CH3 and CH crosspeaks positive and the CH2 crosspeaks negative. CH2 crosspeaks are often obvious by their "pairing" (two crosspeaks with different 1H shifts and the same 13 C shift) along horizontal lines for chiral molecules, but with spectral editing even "degenerate" CH2 groups (both protons equivalent or coincidentally having the same chemical shift) are clearly identified as such by their negative sign. This places a heavier burden on the user in data processing: You have to be very light on the first-order phase correction, just like in the processing of DEPT spectra, so that you do not destroy the phase information by phase "correction." Usually some crosspeaks are known to be CH3, CH2, or CH from their chemical shift, intensity, or "pairing", and this knowledge helps with phase correction. Another useful trick is to watch for "waves" in the baseline when phase correcting a row or column of the matrix. Usually if the first-order phase correction is way off, the baseline will take on a slight sinusoidal "wobble," which will have more cycles the farther off the first-order phase correction becomes. Try for a flat baseline and then "touch up" the phase without making large changes to the chemical-shift dependent (first-order) correction.

Figure 11.16 shows the decoupled, edited HSQC spectrum of 3-heptanone. Positive intensity is shown in black and negative intensity in gray. With 13 C decoupling, the cross-peaks now line up in F2 with the "normal" (12C-bound) 1H peaks in the proton spectrum. Crosspeaks are also twice the intensity relative to noise because they are no longer divided into two peaks by the wide 1/CH coupling. The editing feature makes all of the CH2 cross-peaks negative and the CH3 crosspeaks positive. The decoupled, edited HSQC spectrum of cholesterol in CDCl3 is shown in Figure 11.17. Positive contours are shown in black and negative contours are shown in gray. The overall appearance of the spectrum is roughly diagonal from lower left to upper right, with the olefinic position (6 = z) in the lower left, the alcohol position (3 = y) in the center, and the "hydrocarbon" bulk of the molecule in the upper right. The carbons have been labeled according to the 13 C spectrum (Chapter 1, Fig. 1.26), using letters from the most upfield to the most downfield peak. In crowded regions, it is useful to compare expansions of the 13C and DEPT spectra (Fig. 11.18). Although a single peak of twice the intensity is observed at 31.9 ppm in the 13C spectrum of cholesterol, a slightly upfield CH peak (l) and a slightly downfield CH2 peak (m) can be resolved in the DEPT-135 spectrum. In the tight group n-o-p-q in the 13C spectrum, peak p can be identified as a quaternary carbon because it is missing in the DEPT spectrum. Likewise, in the very tight pair t-u (too close to be resolved in the F1 dimension of 2D spectra), we see that the downfield peak u is quaternary. These observations will help us to correctly label the crosspeaks in the HSQC spectrum. In peak lists and assignment tables, use the precise chemical-shift values from the 1D 13 C spectrum if it is available, because the 1D spectrum has much higher resolution.

Proton

Figure 11.16

Figure 11.16

Proton (ppm) Figure 11.17

Figure 11.18

The methyl crosspeaks a, b, c, e, and f (Fig. 11.17) have such high intensity that they appear with "holes" in the center. This is because we only display a small number of contours starting with the contour threshold (in this case 10) with a constant ratio between intensity levels (in this case 1.25). So the highest contour level is 7.5 times more intense than the threshold (1.25 to the 9th power) and any intensity higher than that is not shown, leaving "holes" in the center of the most intense crosspeaks. Other peaks with positive intensity (black) include the olefinic (z) and alcohol (y) CH groups (insets) and six other CH peaks (i, l, n, v, w, and x). The general appearance is roughly diagonal from the lower left to the upper right, with the notable exception of crosspeaks v, w, and x, which are about 20 ppm below (downfield in 13C shift) the other "hydrocarbon" CH groups i, l, and n. This is an example of the sensitivity of 13C shifts to steric crowding: These are the CH positions next to the bridgehead quaternary carbons in the steroid framework (positions 9, 14, and 17). Their steric environment is similar to a neopentyl position (e.g., neopentyl alcohol HO-CH2-C(CH3)3) that is known to be extremely hindered in organic chemistry (Fig. 11.19). The other three upfield CH groups in cholesterol are relatively open positions (8, 20, and 25), and their HSQC crosspeaks fall in with the rest of the "pack" (i, l, and n).

For most of the CH2 groups, we see two distinct chemical shifts, leading to two negative crosspeaks lying on a horizontal line at the 13C shift in F\. In some cases (e.g., q and s) the difference is quite dramatic, nearly 1 ppm difference in 1H chemical shift for the geminal pair. These large separations usually result from a highly anisotropic environment in a rigid molecule due to a nearby unsaturation. For a number of CH2 groups (k, m, q, and s) we can see a doublet homonuclear splitting on the left side and a triplet pattern on the right side. This is a common pattern in steroids and triterpenes due to the rigid cyclohexane chair structure. In 2D spectra the F2 dimension gives higher resolution than the Fi dimension because we typically acquire 512-2048 complex pairs in t2 (direct detection of the FID) and only 256-375 complex pairs (512-750 FIDs) for the best spectra in t1 (indirect dimension). But resolution in F2 is still poor compared to a 1D spectrum (typically 8192 or 16,384 complex pairs), so we will not see J couplings smaller than 7 Hz resolved. This has the advantage of simplifying the 1H spectrum: We only see the large couplings, and these are generally the geminal (13-16 Hz) and anti vicinal (9-15 Hz) couplings on saturated carbons. The doublet-triplet pattern is the result of one equatorial (one large coupling—geminal) and one axial (two large couplings—geminal and anti vicinal) proton on the same carbon. We saw this with the H4ax-H4eq pair in cholesterol in Chapter 8 (Fig. 8.35), which can be assigned to peak t in the HSQC due to its unique 1H chemical shift. Even in cyclohexane itself the equatorial positions are downfield of the axial positions by about 0.5 ppm, so it is not surprising that in these six examples (j, k, m, q, s, and t) the equatorial proton is downfield of the axial proton. There are exceptions to this doublet-triplet pattern when the carbon has axial protons on both sides: for C11 we predict a "doublet" for the equatorial proton but a "quartet" (geminal plus two anti vicinal couplings to H12ax and H9) for the axial proton.

Figure 11.20 shows three F2 slices of the HSQC spectrum, at the F1 shifts of carbons j, q, and s. For q and s we see the classic doublet-triplet pattern, but for j there is a broad,

Proton (ppm) Figure 11.20

unresolved signal downfield and a quartet pattern upfield. The slices are impressive because they show good resolution in F\ between j and i (13C shift difference 0.23 ppm), and between r and s (difference of 0.27 ppm), with none of the i or r patterns "leaking" into the F2 slices at j and s, respectively. The "quartet" pattern in slice j has been seen before in the F2 slice of the ROESY spectrum (Chapter 10, Fig. 10.24, top, assigned to 16^) at the F1 shift of methyl-18. All three of these slices have negative intensity because of the editing feature (1/1/CH delay just before mixing) that turns all CH2 crosspeaks upside down.

Some of the CH2 crosspeaks are degenerate, meaning that the two protons have the same *H chemical shift (d and r). This can be a coincidence, but it is more likely to happen in a flexible chain, so we would suspect carbons 22-24 in cholesterol, although 22 is less likely because it is next to a chiral center.

Relaxation Audio Sounds Dusk At The Oasis

Relaxation Audio Sounds Dusk At The Oasis

This is an audio all about guiding you to relaxation. This is a Relaxation Audio Sounds with sounds from Dusk At The Oasis.

Get My Free MP3 Audio


Post a comment