F. G. HERRING, J. H. HWANG, AND W. C. LIN
2086
Electron Spin Resonance of X-Ray Irradiated Polycrystalline and Single-Crystal Sodium Hexafluoroantimonate
by F. G. Herring, J. H. Hwang, and W. C. Lin Department of Chemistry, The University of British Columbia, Vancouver 8, British Columbia, Canada (Received November 16, 1966)
The esr spectra of X-ray irradiated polycrystalline and single-crystal NaSbFe were studied. The hyperfine structure was shown to be due to Sb121with I = "2 and Sb123with I = 7/2. The coupling tensor A for Sb121was found to be axially symmetric with All = 74.6 gauss and A, = 60.8 gauss and the g tensor was also found to be axially symmetric with gll = 2.018 and gL = 2.011. The ratio of the coupling-tensor components of Sb121to that of Sblz3was found to agree with the ratio calculated from the accepted values of the nuclear magnetic moments. No hyperfine splittings due to the fluorine atoms were observed. The species was suggested to be either SbF6- or SbFs2-.
Introduction Morton reported' an esr study of irradiated NH4PFe crystals and concluded that the radical formed was PR. The esr spectra showed large hyperfine interactions due to both the phosphorus and the fluorine nuclei. Moreover, the radical was found to rotate freely in the lattice for the spectra of an irradiated single crystal were found to be the same for all orientations and also the same as the spectrum of irradiated powder. We have found that when NaSbFs was irradiated with Xrays, a paramagnetic species was formed which was of quite a different type from the corresponding phosphorus complex. Most significant was the fact that the fluorine splitting was found to be absent. In addition, the radical was found not to be rotating but rather held rigidly in a fixed orientation compatible with the symmetry of the host crystal. In what follows, we shall present the experimental results and shall try to interpret the spectra as being due to either SbPs- or SbFe2-. Both radical species represent antimony complexes with a formal valency of four. The former species could be formed from the SbFeion by the loss of one fluorine atom while the latter could be formed by the capture of an electron. At present, it is quite difficult to decide which assignment is the correct one but, as will be shown later, qualitative considerations seem to favor slightly the former choice. The Journal of PhysicCrl Chemistry
Experimental Section The powder sample of sodium hexafluoroantimonate, NaSbFe, was from Alfa Inorganics, Inc., Beverly, Mass., without further purification. The single crystals were prepared by recrystallization from solutions containing hexafluoroantimonic acid, the latter being also obtained from Alfa Inorganics. Cubic crystals of approximately 1.5 mm to the edge were obtained. There were some minor contradictions in the description of the crystal structure in the 1iteratu1-e.~~~ Unfortunately our esr work did not help to decide which is the more correct since, in general, the orientation of the radical formed does not necessarily have to be exactly the same as that of the parent molecule. In our work, we arbitrarily called three perpendicular edges of the cube the a, b, and c axes and obtained esr spectra with the external magnetic field lying, in turn, in the ab, bc, and ca planes. The irradiation technique and the esr spectrometer used in this work were the same as those described previ~usly.~The measurements were carried out at room temperature. Results None of the spectra showed any splittings due to the (1) J. R. Morton, Can. J. Phys., 41, 706d (1963). (2) N. Schrewelius, 2. Anorg. Allgem. Chem., 238, 245 (1938). (3) G.Tenfer, A& Cryst., 9 , 539 (1956). (4) W.C.Lin and C. A. McDowell, MOL Phys., 7,223 (1964).
2087
ESR OF X-RAYIRRADIATED SODIUM HEXAFLUOROANTIMONATE
ij k
Here, the subscripts i,j , k = a, b, c, cSij is the Kronecker delta, Ai5and gN are components of the A and g tensors, respectively, I , are the direction cosines of H, and G = 2gn&H0/gePewith Hobeing the field corresponding to the center of the spectrum (this approximation is justified when G (Atj- G&j)&}”’
+
(3)
with the first term as the zero-order term and the last two terms as the first-order term and with S being quantized in the direction of H * gand I being quantized in the direction of the resultant field (S-A - g&,H). Although the second-order perturbation is on the order of 1 gauss, it was not taken into account, for the observed hyperfine splitting was taken to be the average across the entire spectrum and the second-order effect increases the splitting on one end of the spectrum and decreases it on the other end. The second-order effect on the g shift was calculated to be 0.001 and was within the experimental error. The A and the g- tensors thus calculated are given below: for Sb121 A =
[
65.34 4.74 4.52
4.74 65.41 4.59
4.59 f 0 . 5 gauss 65.39 4.521
2.0134
0.0024 2.0133 0.0024
0.0025 0.0024 2 * 0127
.*. 36.19 *..
36.45
0.0025 for Sb12a
36.33
... ...
A=[
[
2.0113
g =
...
...
... 2.0105
...
1
* 0.0015
::: ] * 0.5gauss 1::
]*
0.0015
2.0085
For reasons previously mentioned, the A and the g tensors for Sb12a could not be completely determined. Thus, only the diagonal elements are given in order to compare with the corresponding elements for SblZ1. For the same reasons, the principal values of A and g could only be calculated for the isotope species Sb121. These results are given in Table I. Note: Only one set of the relative signs of the direction cosines is given corresponding to one molecular site. The respective direction cosines corresponding to the reVolume 71 s Number 7 June 1067
F. G. HERRING, J. H. HWANG,AND W. C. LIN
2088
