zero-quantum coherences, these can be efficiently suppressed by The NOESY experiment (1, 2) is one of the fundamental and most often performed 2D NMR experiments for applica-the pulsed field gradient in the mixing time, whereas the field gradient will not affect the frequency-labeled z magnetization tions in organic chemistry and in biomolecular research. In the latter application, overnight and even longer measurements are which is the working principle of the NOESY scheme. In addition, axial peaks which are caused by spins which have relaxed often required due to the low concentration of proteins or oligonucleotides. A NOESY experiment for application in organic during t 1 and which are made transverse again by the second 90Њ pulse are also suppressed. chemistry to determine stereochemical assignments typically runs about two hours, depending on the relaxation and mixing
The idea of placing a pulsed field gradient within the mixing period seems to be straightforward and has indeed been used time (3). This rather long time is not due to the sensitivity of current NMR instruments but because an eight-step phase cycle in more complicated 3D sequences (10) and in selective 1D versions of NOE difference spectroscopy (11). It was even ( 4) is required to suppress unwanted coherences, such as double-quantum contributions, COSY-like signals, and axial peaks. mentioned before the introduction of self-shielded gradient coils in the first demonstration of the EXSY experiment (12) In recent years, the other standard sequences used in structural elucidation for organic chemistry, such as COSY (5) or HMQC and applied for the HOESY sequence (13). Surprisingly, to the best of our knowledge, it has not been mentioned or demon-( 6) have experienced a significant reduction in experiment time with the introduction of pulsed field gradients (7). Today a strated for the basic homonuclear 2D NOESY sequence, probably because the developers of modern NMR need many tran-gradient-selected COSY experiment or a HMQC-type inverse C,H correlation of typically 20-50 mg of organic material and sients for the usual protein work. Best results were achieved when the gradient pulse was on during the entire mixing time. a molecular weight on the order of 500 Da can be performed in 5 to 10 minutes (3). Compared with this, the corresponding This is probably due to the integrated gradient strength which can be reached and to the suppression of stray-field effects NOESY measurement time is inconvenient, especially since it is desirable to perform more than one NOESY experiment using during the mixing time ( 14). In principle, our method could be performed using no phase cycle at all; however, in practice, different mixing times. Within this context, Ko ¨ck and Griesinger recently published a FAST NOESY method (8), reducing the it seems to be advantageous to use a two-step phase cycle, similar to many other gradient-selected experiments (3, 7). relaxation waiting time. However, their procedure requires an extra measurement of these relaxation times and special data
As an example, we show in Fig. 2a a standard NOESY spectrum of a 10% solution of strychnine processing so that this method does not seem to have gained wide application in service laboratories.
In this Communication, we wish to demonstrate that by an extremely easy modification of the standard experiment, the usual eight-step phase cycle of the experiment can effectively be reduced to two steps leading to the desired fourfold reduction in experiment time. We replace the NOESY mixing time t m simply by a weak pulsed field gradient, which acts during the entire exchange period as shown in Fig. 1.
For a two-spin system, it can be shown with the productoperator formalism (9) that the second 90Њ pulse of the NOESY sequence creates the desired frequency-labeled z magnetization, but in addition double-and zero-quantum coherences and antiphase coherences leading to COSY-type signals. Despite the