The electric-field gradient tensor at the iron nucleus in ferrous (Fe 2+ ) compounds is investigated. One sees that under the combined action of axial and rhombic crystalline fields and the spin-orbit interaction, the ferrous-ion ( b D,3d & )d e states produce a large, temperature-dependent, contribution to the electric-field gradient tensor. It is found that this direct contribution is diminished by that from the lattice itself (the second-order axial and rhombic components of the crystalline field), as well as Sternheimer polarization effects and covalency. The results of this investigation are then applied to Mossbauer results in FeSiF 6 -6H 2 0 to obtain an estimate of the electric quadrupole moment of Fe 57w (0.29±0.02b), which is in agreement with that from ferric (Fe 3+ ) studies. Finally, estimates also based upon Mossbauer measurements, are made of the d t energy splittings in the ferrous compounds, FeSiF 6 -6H 2 0, FeS0 4 -7H 2 0, FeC 2 04-2H 2 0, Fe(NH 4 S0 4 )2-6H 2 0, FeS0 4 , FeCl 2 -4H 2 0, and FeF 2 . . H. M. Foley, R. M. charge distribution of the aspherical 3d-"valence'' electron belonging to the ferrous ion ( 5 Z),3d 6 ). The subscript "lat" refers to the charge distribution of the neighboring ions in the crystalline lattice. The Sternheimer factors, 7-8 (1 -R) and (I-Y^), are added to correct for the polarization of the ferric-like ( 6 S,3d 5 ) core by the EFG of the valence and lattice charge distributions.The most important contribution to q, in Eq. (3) is # va i, which has an absolute value of 9 (4/7)(lA 3 )3d for the free ion, neglecting the spin-orbit interaction. However, this term, itself, is reduced from the free-ion value by the crystalline field at finite temperatures (Sec. II), the spin-orbit interaction (Sec. Ill), and covalency effects (Sec. V). In addition, the factors, (1 -R), and (1-y^giat (Sec. IV), further reduce the effect of this term. These arguments also hold for the quantity, ??vai<7vai> in Eq. (4), which is present when the valence charge distribution is not symmetric about the z axis. Thus, when all the contributions to the EFG are simultaneously treated, one obtains a significantly larger value for the quadrupole moment, Q, than if one were to use the free-ion approximation (Sec. VI). At the same time it is also possible to make reasonable estimates of the crystal-field splittings in several compounds (Sec. VII) based upon the Mossbauer data of Fig. 1. II. PRIMARY EFFECT OF THE CRYSTALLINE FIELDThe primary effect of the crystalline field is to lift the fivefold spatial degeneracy of the b D state of the freeferrous ion, splitting it into a series of orbital states, Of energies, e n , which may each produce different EFG tensors at the position of the nucleus. (Each of these orbital states will still have fivefold spin degeneracy.) Assuming the thermal transition times between Sternheimer, and D. Tvcko, Phys. Rev. 93, 734 (1954); R. M. Sternheimer and H. M. Foley, ibid. 102, 731 (1956); R. M. Sternheimer, ibid. 84, 244 (1951); 95, 736 (1954); 105, 158 (1957). 8 A.

The extended x-rayabsorption fine structure (EXAFS) above the Cu and Br E edges in CuBr has been measured at 72, 210, and 295 K. The first-shell contributions were isolated by Fourier filtering the EXAFS data, and the resulting single-shell data were analyzed using the ratio method. It is found that the third and fourth cumulants of the nearest-neighbor distance distribution are quite significant even at 210 K and cannot be ignored when making least-squares fits to amplitude ratios and phase differences. At 295 K we obtain values of (0.37+0.05))(10 A' and (0.72+0. 12))&10 A for the third and fourth cumulants, respectively, as well as an anharmonic contribution of 0 (0. 16+0.12))&10 A to the mean-square relative displacement. The data cannot be explained by a disorder model in which a Cu ion is allowed to occupy four equivalent off-center sites as well as its ideal position; the non-Gaussian contribution must be due to intrinsic anharmonicity. Interpretation of the results in terms of an anharmonic effective single-particle potential for the Cu ions indicates that the fourth-order potential term is just as important as the third and cannot be neglected. This model gives a reasonable description of the temperature dependence of the second, third, and fourth cumulants.

No abstract

No abstract

The local structure of the perovskite KNbO 3 at 77 and 300 K under high pressure, up to 15.8 GPa, has been investigated using the x-ray-absorption fine-structure ͑XAFS͒ technique. It is found that the local-structure symmetry does not change even though the average macroscopic crystal symmetry changes several times over this same temperature and pressure range. The existence of different local and macroscopic structures means that the local distortions from the average structure are disordered. Other unexpected results obtained from the XAFS measurements are evidence for a decreasing Nb-Nb displacement correlation length and a destabilization of the oxygen octahedra with increasing pressure.