Beryllium-9 NMR study of solid bis(2,4-pentanedionato-O,O′)beryllium and theoretical studies of 9Be electric field gradient and chemical shielding tensors.: First evidence for anisotropic beryllium shielding

Despite the favorable NMR properties of Be-9 (I = 3/2), NMR spectroscopy of this nucleus in the solid state remains comparatively unexplored, perhaps owing to the extreme toxicity of beryllium and its compounds. We present here an integrated experimental and theoretical study of the Be chemical shielding (CS) and electric field gradient(EFG) tensors in bis(2,4-pentanedionato-O,O')beryllium [Be(acac)(2)]. Interpretation of the Be-9 NMR data was facilitated by crystal X-ray diffraction results, which indicate two crystallographically unique sites (Onuma, S.; Shibata, S. Acta Crystallogr. 1985, C41, 1181). Beryllium-9 NMR spectra acquired at 4.7 and 9.4 T for magic;angle spinning (MAS) and stationary samples have been fitted in order to extract the nuclear quadrupole coupling constant (C-Q), asymmetry parameter (eta), and isotropic chemical shift (delta(iso)). The best-fit nuclear quadrupole parameters for the two sites were determined to be C-Q(1) = -294 +/- 4 kHz, eta(1) = 0.11 +/- 0.04; C-Q(2) = -300 +/- 4 kHz, eta(2) = 0.15 +/- 0.02. Our analyses of the stationary samples also reveal a definite anisotropy in the beryllium CS tensor and allow us to place upper and lower limits on the spans of 7 and 3 ppm. This is the first evidence for anisotropic shielding in beryllium. Ab initio calculations of the beryllium CS tensors in Be(acac)(2) at the RHF level indicate spans ranging from 7 to 9 ppm; this represents a substantial fraction of the total known chemical shift range for Be (<50 ppm). The calculated Cps are also in good agreement with the experimental results. To put the Be(acac)(2) results in context, calculations of the beryllium CS tensors for a series of compounds encompassing the known range of Be-9 chemical shifts are also presented. The calculations are in outstanding accord with experimental data from the literature. On the basis of calculations for Linear molecules, it is shown that the assumption that the Be-9 chemical shift is governed essentially by the diamagnetic term is erroneous. For some of these molecules, the calculated Be CS tensor spans are greater than the total known chemical shift range.