-V 23 / H ' CH

Figure 11.50

Figure 11.50

that is converted to ZQC/DQC by that pulse is detected. Each FID coming into the receiver is alternately added to and subtracted from the sum in memory. Another way to say this is that the receiver phase is x, —x, x, —x in this scheme, and the desired signals, which are also modulated in this way, will combine and add in memory. The first13 C 90o pulse, which is a trap put there to "spoil" the signal of protons directly bound to 13 C (the one-bond signals) is phase cycled x, x, x, x for each set of four scans. Because the receiver phase cycle is x, -x, x, -x, which does not match this pattern, the signals that arise from 1H magnetization following this pathway are canceled by subtraction and eliminated from the sum FID in memory. We can call this dual-purpose phase cycle a "four step" phase cycle; the number of transients (Bruker: ns, Varian: nt) will have to be a multiple of four in order for the cancelation of undesired signals to work.

Because the 1H chemical-shift evolution during the long (~50 ms) 1/(27) defocusing delay is not refocused as it is in HMQC, HMBC data is normally presented in magnitude mode. To make the data suitable for phase-sensitive presentation, it is necessary to insert spin echoes to allow for refocusing of all unwanted shift evolution. For example, one can insert simultaneous 180o pulses on 1Hand 13C in the center of the long 1/(27) delay, refocusing the 1H chemical-shift evolution (Fig. 11.50). To eliminate the 13C-bound 1H signal a different strategy can be used: a variant of the BIRD sequence called TANGO delivers a 90o pulse to the 13C-bound protons without exciting the 12C-bound protons (Fig. 11.51). The resulting 13C-bound proton coherence is then destroyed with a "spoiler" gradient, which has no effect on the 12C-bound proton magnetization which is on the z axis. As long as we are adding gradients, we can do coherence pathway selection by adding gradients during the two halves of the evolution delay. Now we have a sequence that blocks the one-bond artifacts and eliminates the "streaks" due to 1H coherence that is not correlated to any 13C, all in a single scan. But the gradients during t1 create a new problem: again we can not start with t1 = 0in

0 0

Post a comment