Publication costs assisted by the National Science FoundationIt is generally accepted that an isolated methyl group such as that in toluene is essentially a free rotor ( = O kcal for this sixfold barrier), while the adjacent methyls in o-xylene have a threefold barrier to rotation of about 2 kcal/mol.lr2 This is supported by pmr and cmr (carbon-13 magnetic resonance) spin-lattice relaxation studies3b4 which demonstrate that the protona and carbon-134 relaxations of a lone methyl group have substantial contributions from spin-rotation effects, while carbon-13 relaxations of the rotationally restricted methyl groups in o-xylene are dominated by the dipolar mechanism.4In the 1,2,&trisubstituted methylbenzenes, hemimellitene (1,2,3) and isodurene (1,2,3,5), an additional structural feature arises which affects the rotation rates of the methyl at C-2. According to previous results2 there are two equivalent low-energy conformers for the 2-methyl groups which give rise to a sixfold rotation barrier. These conformations are demonstrated for the 1,2,3-trimethyl interactions as H .
I fIThe 1-CH3 in conformer I1 and the 3-CH3 in conformer I are in minimum energy configurations relative to the 2-CH3 and have threefold energy profiles with 1.5-2 kcal/mol barriers for rotation.'n2 The 1-CHa in conformer I and the 3-CH3 in conformer I1 also are governed by a threefold barrier to rotation but now the magnitude exceeds 2 kcal/mol by a considerable amount due to the orientation of the 2-methyl. On the other hand, the 2-methyl will possess the same steric energy in both forms because of symmetry giving rise to the expected sixfold barrier. As sixfold barriers to rotation are generally much lower than threefold barriers, one may expect the 2-methyl carbon relaxation to exhibit spin-rotational effects. Kuhlmann, et a1.,6 predict significant carbon-13 spin-rotation effects for methyls with barriers 5 400 cal/mol. On the other hand, the 1 and 3 methyls could be sufficiently restricted in their rotation that the dipolar relaxation mechanism should again dominate.The cmr relaxation data given in Table I for hemimellitene and isodurene substantiate the above anticipated results. Both the 2-C& in hemimellitene and the 2-and 5-CH3's in isodurene have substantial contributions (Tlo = 20 sec, and Tlo = 21 and 24 sec, respectively) to their relaxations from mechanisms other than dipolar processes. Decreases in these values of Ti0 with temperature confirm that these contributions are from spin rotation mechanisms. On the other hand, the relaxations of the 1 and 3 methyls are dominated by dipolar processes (TID = 14 sec, 2'10 = 98 sec for hemimellitene; T~D = 12 sec, 2'10 2 100 sec for isodurene). The nuclear Overhauser effect (NOE) data in Table I were used to separate dipolar contributions, T~D , from the proton-decoupled relaxation times, T I (7 = ~T~/ T x , ) ;~ the contributions from all relaxation processes, 2'10, other than dipolar were then determined from the reciprocal additivity relationship of relaxation times (1/T1 = l/TID + l/T10)....