ICOSAHEDRAL C60 REVISITED: AN AROMATIC MOLECULE WITH A VANISHINGLY SMALL RING CURRENT MAGNETIC SUSCEPTIBILITY.

In a previous publication in this journal we conclude that: "Although planar aromatic hydrocarbons and graphite have provided appropriate points of analogy for the pi-electron energy structure of C sub 60, it is clear that the magnetic response of Ci sub 60 is unlike that of any other molecule yet encountered". In an accompanying publication in this issue of Nature, Fowler, Lazzeretti and Zanasi (FLZ), report ab initio calculations on the magnetic susceptibility of icosahedral C sub 60. Their best calculation gives a magnetic susceptibility of -3916 cgs ppm for the molecule, whereas an extrapolation scheme leads to a value of -357 cgs ppm. On this basis FLZ conclude that C sub 60 possesses "... the strong diamagnetism expected of an aromatic system". The validity of this conclusion is addressed in the present note. The first task is to estimate the local atom and bond contributions to the diamagnetism of C sub 60. The value obtained for chi(local-C sub 60) (-291 cgs ppm) is in reasonable agreement with the extrapolated value obtained by FLZ for the total magnetic susceptibility of C sub 60! The pi-electron ring current magnetic susceptibility is then obtained for the following models: chi(pi-graphitic-C sub 60)=- 4977 cgs ppm (based on graphite); chi(pi-C sub 60-Pauling)=- 1681 cgs ppm (based on Pauling free electron model); chi(pi- C sub 60-benzenoid)=274 cgs ppm (based on the 20 benezene rings implicit in the C sub 60 structure); chi(pi-C sub 60-London)=+9 to - 123 (based on our previous London calculation as a function of relative bond strengths). When combined with the local contribution [chi(local- C sub 60)], the London calculation leads to chi(total-C sub (60-) London) = -282 to -414 cgs ppm. Considering the imponderables introduced in the two treatments, the agreement between the FLZ results and our previous London calculations on sub C60 must be considered remarkable. Reference to the models presented above makes clear that the London and FLZ calculations lead to a pi-electron ring current magnetic susceptibility for C sub 60 which is incompatible with its description as a normal aromatic molecule. We stand by our previous conclusions and look forward to the experimental measurement of the magnetic susceptibility of C sub 60.