MAGIC-ANGLE-TURNING EXPERIMENTS FOR MEASURING CHEMICAL-SHIFT-TENSOR PRINCIPAL VALUES IN POWDERED SOLIDS

The magic-angle-turning (MAT) technique introduced by Gan employs slow (approximately 30 Hz) rotation of a powdered sample at the magic angle, in concert with pulses synchronized to 1/3 of the rotor period, to obtain isotropic-shift information in one dimension of a 2D spectrum. The other dimension displays a slow-spinning-sideband powder pattern which, at the low rotor frequencies employed, resembles the stationary-sample powder pattern. The MAT method is very effective for measuring chemical-shift principal values in compounds where spectral overlap precludes the use of 1D methods. Previous MAT implementations are reviewed, and it is shown how a new phase-corrected MAT (PHORMAT) pulse sequence overcomes many of their limitations. This new pulse sequence produces a spinning-sideband-free isotropic-shift spectrum directly as a projection onto the evolution axis with no spectral shearing. Only two purging operations are employed, resulting in a higher signal-to-noise ratio. Pure absorption-absorption-phased 2D spectra are produced. Flat 2D baseplanes result from an echo sequence which delays acquisition until after probe ring down and receiver recovery. The technique used for synchronizing the pulses to 1/3 the rotor period without relying on absolute rotor-frequency stability is described. The PHORMAT spectrum of methyl-alpha-D-glucopyranoside is presented. The data are analyzed with an emphasis on the quantitative accuracy of the experiment in measuring chemical-shift-tensor principal values and determining the relative number of spins of each type present. The FID data from the spectrometer acquisition are fitted with numerical simulations that employ a banded-matrix method for calculating spinning-sideband amplitudes. The chemical-shift principal values. measured in methyl-a-D-glucopyranoside with the PHORMAT method, are compared with those from a single-crystal determination of the full chemical-shift tensors. The two measurements differ by an RMS-average distance of only 0.57 ppm. (C) 1995 Academic Press, Inc.