In a series of closely related compounds, it may occur that the change in total (diaplus para-) magnetic shielding of atomic nuclei is nearly that of the local paramagnetic term, because of cancellation effects involving nonlocal dia-and paramagnetic contributions. This offers a justification for relationships between nuclear magnetic resonance shifts and local atomic populations, which are occasionally observed. It is important, however, to consider the type (2s, 2p, , or x) of electrons which are responsible for the variations in atomic charges. Correlations between UC NMR shifts and atomic populations of aromatic compounds, for example, should not be interpreted in terms of x electrons only, because the slope of shift vs. x charge (i.e., the ~160-ppm/electron value which is usually invoked) does not describe an intrinsic effect of x charges on magnetic shielding but accounts for the fact that a and x charges vary in opposite directions in this class of compounds. The explicit consideration of the inverse variations of a and x charges, where appropriate, offers an explanation for the observation that charge-shift correlations can have positive or negative slopes. It appears, indeed, that an increase in total electronic population is accompanied (i) by a high-field shift when the electron enrichment results from a gain in x charge prevailing over the concurrent loss in electrons (aromatic and vinyl C, carbonyl O atoms) or (ii) by a downfield shift when the increase in charge is dictated by that of the population (sp1 23 C, carbonyl C, and dialkyl ether O atoms).

Benzenoid hydrocarbons were examined using a bond energy scheme featuring the role of atomic charges. The latter were conveniently deduced from appropriate correlations between theoretical results and 13C nuclear magnetic resonance shifts. Atomization energies calculated in this manner agree with their experimental counterparts to within 0.36 kcal mol-1 (average deviation). It appears that benzenoid hydrocarbons can be efficiently described in terms of local charge density properties. In the absence of any distinctive specific feature characterizing benzenoids, this particular description of chemical bonds ultimately results in a unifying genealogy smoothly relating to one another the various possible types of CC and CH bonds which are formed by sp2 and sp3 carbons.

The solid-state structures of the compounds (1 ) (SeS2+) (Sb2F42+) (Sb2F5+) (SbF8-),, (2) (TeS2+) (SbF6-)2, and (3) (Se42+)(AIC14-)2 are reported. The first compound was initially isolated from the reaction of a 1 : 3 sulphurselenium alloy with a SbF5-S02 solution and the second by the reaction of a mixture of tellurium and germanium, with a SbF,-S02 solution. They have also been prepared by the direct oxidation of Se and Te with SbF, in SO2. (1) are monoclinic, space group P2,/c, with a = 15.739(3), b = 13.498(2), c = 17.040(4) A, p = 92.26(2)', and Z = 4. The dark red plates of (2) are triclinic, space group AT, with a = 5.700(2), b = 16.252(6). c = 8.076(2) A, cc = 100.56(3), a = 102.67(3), y = 97.47(3)', and Z = 2. Compound (3) is orthorhombic, space group Pbarn, with a = 13.245(3), b = 13.223(3), c = 9.266(2) k and Z = 4. The structures of compounds (1 ) and (2) have been solved by direct methods and compound (3) by a Patterson function. They were refined by least squares to final agreement indices of R = 0.052 (R' = 0.062), 0.051 (0.064), and 0.042 (0.046) for 3 634, 709, and 91 2 observed reflections respectively. In these three compounds, the approximately square-planar chalcogen cations Se42+ and TeS2+ were found to have crystallographic inversion symmetry with average Se-Se and Te-Te distances of 2.260(4) ( l ) , 2.688(3) (2), and 2.286(2) 8, (3). These values are close to the distances observed in the other examples of these cations. The packing of these and other examples of the M42.+ cations, the anion-cation charge-transfer interactions, and the stereochemistry of the fluorine contacts to the antimony(1li) atoms in the Sb2F42+ and Sb2Fs+ ions and related species are discussed. Crystals of

No abstract

No abstract