2534
COMMUNICATIONS TO THE
EDITOR
Vol. 75
types however there is no sign of convergence a t n = 2, which explains the failure of the OnsagerFuoss theory for calcium chloride and lanthanum chloride. For potassium chloride the terms beyond A? are negligible; for calcium chloride a t 0.005 molar the values are: A, = -0.019, A2 = 4-0.037, A3 = -0.016, A4 = 4-0.016, A5 = -0.010, . (the units being cm.’ set.-' X lo-;). For lanthanum chloride a t 0.003 molar they are: AI = -0.033, 02 = 4-0.074, As = -0.071, A3 = f0.093, A5 = -0.096, . . . . -1fuller account of this work will appear shortly.
intensity with interval 27 oersteds-more than twice the interval observed in liquid solutionsare seen. When the crystal is rotated by 90 degrees around the two-fold axis from the above orientation, the interval shrinks to Six oersteds. At most other orientations the spectrum is complex, consisting of four or five lines. An explanation of the details of the dependence of spectrum on orientation of the crystal awaits completion of the crystal structure determination.4 Nevertheless, sufficient information is available to permit the drawing of certain conclusions concerning the distribution of the unpaired electron in the molecule. Interpretation of the experimental results deCHEMISTRY DEPARTMENT UNIVERSITYOF W.A. R. €€. STOKES pends on the fact that the hyperfine interaction SEDLANDS, WESTERNAUSTRALIA consists of two parts, one isotropic, the other RECEIVED APRIL21, 1953 anisotropic. The former depends only on the nonvanishing component of the w3ve function at the nucleus, the latter only on the non-spherically ANISOTROPY IN THE PARAMAGNETIC RESONANCE symmetric components distant from the nucleus. SPECTRUM OF PEROXYLAMINE DISULFONATE ION1 If the electronic wave function of a free radical is expanded in atomic functions only s components Sir: contribute to the isotropic interaction, non s In an earlier report2 of the hyperfine splitting in components to the anisotropic interaction. When the paramagnetic resonance absarption spectrum the frequency of molecular tumbling is high comof (C6H5)3Cl3attention was calM to the fact that pared with the frequency associated with the hyperobservations in liquid solutions fail to reveal im- fine interaction, only the isotropic part of the portant features of tbe phenoosemn. We have splitting is observed, the non-isotropic part aversince observed the paramagnetic resonome spec- aging t o zero.5J6 trum of the free radical ion peroxylamine d i s d o Thus the observed splitting in liquid solutions of nate (SO(SO,),=) in dilute solid solution in single (GH5)3C13demonstrates the existence of methyl crystals of diamagnetic potassium hydroxylamine carbon 2s component in the electronic function.* disulfonate. In liquid solution the ion exhibits Experiments in single crystals are required for an absorption spectrum of three eqllmally spaced estimation of the non “s” components. In the equal iBtensity lines. The three lines resalt from peroxylamine disulfonate, both nitrogen 2s and 2p the hyperfine interaction between the nitrogen components are important. Their amplitudes will nucleus (spin one) and the e l e c t r ~ n . ~The interval be estimated a t another time. between adjacent lines is 13 oersteds. In the sin& WASHIKGTO\UXIVERSITY S I WEISSMAN crystals (containing about 0.23 mole per cent. of t h ST LOUIS,MISSOURI DOROTHYBAUFILL paramagnetic ion) the spectrum is highly anisoRECEIVED APRIL 24, 1953 tropic. The crystals, whose structwe is under -(4) The magnetic observations suggest that Two orientations 01 the investigation by PPofessor Lindsay Helmholz peroxylamine disulfooate ion with respeet to the crystal axis occur. and Mr. Merton Brooks, are monoclinic. When (5)See for instance, A Abragam and M H L Pryce, Proc R o y the external magnetic field is perpendicular to $he S o c , 205, 1935 (195I), N Bloembergen, E M Purcell and R V two-fold axis of the crystal and makes an angle Pound, P h y s R e v , 71, 466 (19471, H S Gutowsky, D W McCall and of about 45’ with the “c” axis, three lines of equal C P Sllchter, J Chew P h y s , 21, 279 (1953) (1) Assisted by the joint program of 0 N R and A E C (2) S. I. Weissman and J C Sowden, THIS J O U R N A I , 76, 303
(1953) (3) G E. Pake, J Townsend and S I Weissman, P h y s Rev, 86, 682 (1952), J. Townsend, S. I Weissman and G E Pake, tbrd , 89, 606 (1953)
(6) In the cases here described the hyperfine frequency is in the neighborhood of 109 sec - 1 . (7) The result is readily demonstrated for the case in which the external magnetic field is strong and spin orbit interaction is small It is probably correct in the absence of these restrictions. (8) The square of the dmplitude of this component is probably between 0 05 and 0 10
