No abstract

Theoretical principles and experimental details of the centerband-only detection of exchange ͑CODEX͒ nuclear magnetic resonance ͑NMR͒ experiment for characterizing slow segmental dynamics in solids are described. The experiment, which is performed under magic-angle spinning, employs recoupling of the chemical-shift anisotropy before and after a long mixing time during which molecular reorientations may occur. By an analysis in terms of the difference tensor of the chemical shifts before and after the mixing time, the dependence on the reorientation angle is obtained analytically for uniaxial interactions, and a relation to two-dimensional exchange NMR patterns is established; the same theory can also be applied for analyzing stimulated-echo and pure-exchange NMR data. A favorable linear dependence is derived generally for small rotations, which makes the experiment suitable for detecting small-amplitude motions. Quantification is excellent because the peaks are narrow and intense, unlike the broad powder or sideband spectra that are characteristic of all previous NMR experiments for probing slow segmental rotations. We also introduce and demonstrate a four-time CODEX experiment that yields information previously obtained only in 3D ͑three-dimensional͒ and reduced 4D ͑four-dimensional͒ exchange NMR experiments, such as the number of orientational sites accessible to the mobile groups. Chemical-shift anisotropies required in the CODEX analysis of motional amplitudes can be estimated using a closely related chemical-shift recoupling experiment. The implementation of total suppression of sidebands before detection is also explained. The experiments are demonstrated on dimethylsulfone, isotactic polypropylene, and poly͑methyl methacrylate͒, PMMA. In isotactic poly͑1-butene͒, the signals of the amorphous and interfacial regions have been observed selectively by using pure-exchange CODEX near the glass transition. The four-time CODEX experiment confirms that in the ␤-relaxation process of glassy PMMA, fewer than half of the sidegroups perform jumps between two orientations.

The quantification of quantum correlations (other than entanglement) usually entails labored numerical optimization procedures also demanding quantum state tomographic methods. Thus it is interesting to have a laboratory friendly witness for the nature of correlations. In this Letter we report a direct experimental implementation of such a witness in a room temperature nuclear magnetic resonance system. In our experiment the nature of correlations is revealed by performing only few local magnetization measurements. We also compared the witness results with those for the symmetric quantum discord and we obtained a fairly good agreement.