Separating Chemical Shift and Quadrupolar Anisotropies via Multiple Quantum NMR Spectroscopy,
J. T. Ash, N. M. Trease, and P. J. Grandinetti
It is well known that the chemical shift interaction in nuclear magnetic resonance spectroscopy offers a wealth of information about structure and dynamics at the molecular level. A measurement of the full chemical shift tensor, for which liquid NMR spectra provide only a measurement of the trace, is an extremely valuable but difficult measurement to perform. The most common approach is to measure the NMR frequency anisotropy in solid samples. This approach, however, is extremely complicated for the NMR active isotopes that also experience strong quadrupolar couplings. The presence of both interactions makes the extraction of NMR parameters highly problematic, as an anisotropic lineshape must be fit for all chemical shift and quadrupolar coupling tensor parameters simultaneously, i.e., nearly a 10 parameter fit.
The COASTER (Correlation Of Anisotropies Separated Through Echo Refocusing) experiment is a triple to single quantum correlation with a coherence transfer pathway of p = 0 → +3 → -1 while the sample is spun at 70.12°. This pathway refocuses the second-rank quadrupolar anisotropy and chemical shift at different times in the 2D experiment. The COASTER experiment generates a
chemical shift dimension, free from anisotropic quadrupolar interactions, and a
quadrupolar dimension, free from chemical shift interactions. This separation of anisotropies results in 1D spectral projections that are dependent only upon the tensor eigenvalues of each interaction, allowing for more precise determination of both tensor parameters. Additionally, the 2D spectrum can be analyzed separately to obtain the relative orientation between the two tensors.
Comparison of simulated 2D COASTER spectra showing the effect of the changing quadrupolar coupling and chemical shift asymmetry parameters in the case where the quadrupolar coupling and chemical shift tensors have the same principal axis systems. Other simulation parameters included I=3/2, ω0=100 MHz, Cq=3 MHz, σ =0 ppm and ζcs=33 ppm. The one-dimensional projections onto the quadrupolar anisotropy axis, ω2'(Q), are the same for each ηq value. Similarly, the one-dimensional projections onto the chemical shift anisotropy axis, ω1'(CSA), is the same for each ηcs value.
Comparison of simulated COASTER spectra showing the effect of the relative orientation on the two-dimensional spectrum. Other simulation parameters included I=3/2, ω0=100 MHz, Cq=3 MHz, ηq=0.25, σ =0 ppm, ζcs=33 ppm, and ηcs=0.5. Again, note that the projection onto each axis remains unchanged as the relative orientation of the quadrupolar coupling and chemical shift tensors change.