Correlation of Isotropic Shifts and Chemical Shift Anisotropies by Two-Dimensional Fourier-Transform Magic-Angle Hopping NMR Spectroscopy

During the 1960s Andrew and others examined the rapid spinning of a sample about an axis that makes an angle of 54” 44’ with the direction of the static magnetic field (I-LJ in order to remove broadening effects in the NMR spectra of solids (l-3). It was much later when Schaefer and Stejskal (4) applied this approach, magic-angle spinning (MAS), to remove broadening due to chemical shift anisotropy (CSA) in 13C NMR, combining this approach with high-power ‘H decoupling and cross polarization (CP). The resulting levels of resolution and sensitivity obtained with this combination have made the 13C CP-MAS experiment the most widely applied solid state NMR experiment in recent years. As powerful, versatile, and popular as the 13C CP-MAS experiment has become, there remain some characteristics that limit its usefulness in certain types of applications. Technological problems persist in techniques for spinning the sample rapidly, problems that are intensified by the scaling of CSA with increasing magnitude of the static field (H,,), although recent advances show great promise for alleviating these problems (5, 6). Another limitation of the usual CP-MAS 13C experiment is that it eliminates the potentially useful information embodied in the CSA pattern, i.e., independent values of the three principal elements of the shielding tensor, ul, , uz2, and g33. Only the trace, actually (a, I + ~2~ + a&/3, of the shielding tensor survives under MAS. Techniques have been proposed for retrieving CSA information from a MAS experiment (7-1 I); although each of these techniques has merits, each suffers from disadvantages. Introduced here is a two-dimensional (2-D) Fourier transform (FT) technique which presents the isotropic average chemical shift, q = (a, 1 + 622 + ~~~)/3, in one frequency dimension (F,) and the static CSA powder pattern along the other frequency axis (F2). The experiment is carried out using discrete “hops” between evolution segments, rather than continuous sample spinning, and no spinning sidebands are produced. As the detection occurs on a static sample, the signal decays more rapidly than in a normal MAS experiment, and sensitivity suffers correspondingly. Nevertheless, the experiment shows considerable promise, not only for the CSA results it is capable of