7. Comments

The sequence starts with the BIRD sandwich (see Exp. 6.14) and this is adjusted to suppress the signals of all protons bonded to C-12. It appears tempting, and possible in principle, to omit this BIRD sandwich and instead achieve the following @, filter by gradient selection. However, a gradient half-filter would only pass half of the magnetization, and in this application, where sensitivity is important and considerable time has to be spent for good resolution, the traditional BIRD scheme proved advantageous; however, compare Experiment 12.13.

Thus, after the BIRD delay, we find an HMQC-type sequence (see Exp. 10.13) which however, contains no ¢, evolution period for carbon, but fixed delays d5. Due to the proton-carbon coupling H,C doublets are generated. After the pulse p10, which transfers the magnetization back to protons, a TOCSY spin-lock starts for each part of these H,C doublets. Thus the TOCSY cross-peaks generated are split in F, by the 'J(C,H) coupling constant of the starting proton, and these two signals have an offset in F,. This offset is caused by the long-range C,H spin coupling constant and can be measured providing that the 2D spectrum is phase-sensitive and has sufficient digital resolution. A variant, described in reference [3], provides an additional suppression of The method has an obvious limitation, because all carbon atoms, including those residing between a relevant carbon and its long-range-coupled proton, must have at- tached protons for a successful TOCSY transfer.

8. Own Observations