We report proton Zeeman relaxation rates in solid 1,4-di-isopropylbenzene (DIB) and in 1,3,5-tri-isopropylbenzene (TIB) at Larmor frequencies of 8.5 and 53 MHz. The temperature ranged from 110 to 300 K for DIB and from 85 to 240 K for TIB. In DIB the free induction decay contains both a liquid-like and a solid-like component at higher temperatures, and the solid-like component relaxes non-exponentially. In TIB the free induction decay is typical of molecular solids with reorienting methyl groups and the relaxation is always exponential. The model used to fit the temperature and frequency dependence of the Zeeman relaxation rates R in both moiecules suggests the occurrence of methyl reorientation only in the vicinity of and at temperatures below the observed maximum in R. Isopropyl reorientation does not occur except perhaps at the highest temperatures in TIB, where an additional motion begins to dominate the relaxation. These results are compared with the t-butyl analogues of these molecules in which the methyl and the t-butyl groups reorient at the same rate.

We have measured proton Zeeman relaxation rates R in the 2,4-and 2,5-isomers of di-tertbutylhydroxybenzene (DTHB) in the solid state. R was measured as a function of temperature T at proton Larmor frequencies of (j)/21T = 8.50,22.5, and 53.0 MHz. The Tranges were from 78 K to just below the melting points of 2,4-and 2,5-DTHB, 385 and 323 K, respectively. The 2,5-DTHB R vs T and (j) can be interpreted qualitatively in terms of three Bloembergen-PurcellPound (BPP) spectral densities, one for each of the three types of rotors in the molecule. The quantitative agreement is poor but a good fit is obtained using either a Davidson-Cole (DC) or Frolich spectral density, still preserving the three rotor types. The implications of this are discussed. The BPP and DC spectral densities fail completely in interpreting R vs T and (j) for 2,4-DTHBwhereas good quantitative fits are obtained using a Frolich spectral density. The distributions of activation energies characterizing the three rotor types are so wide for the Frolich spectral density fit of the 2,4-DTHB data that the individual rotor types lose their identity.