Theory of magnetic susceptibility of graphite intercalation compounds

The orbital magnetic susceptibility (${\chi }_{or}^c$) of graphite intercalation compounds is calculated for $\overrightarrow{\mathrm{H}}\parallel \overrightarrow{\mathrm{c}}$ using a tight-binding model for the $\pi$-electron energy bands and the compact expression for ${\chi }_{or}^c$ derived by Fukuyama, which includes both intraband and interband terms. The results are presented as an approximate analytic expression for ${\chi }_{or}^c$ as a function of Fermi level $\mu$ ($0<\mathrm{}\mu \lesssim 3$ eV) and temperature. For $\frac{\mu }{k_{B}T}\ll 1$ the susceptibility is large and diamagnetic (due to interband transitions), while for $k_{B}T\ll \mu \lesssim 3$ eV, ${\chi }_{or}^c$ is paramagnetic (mainly due to intraband transitions) in contrast to the usual Landau-Peierls diamagnetism of conduction electrons in parabolic bands. Furthermore, ${\chi }_{or}^c$ is shown to be a sensitive function of $\mu$ and hence of the conduction-electron charge distribution in each graphite layer. We show that this theory accounts for the main features of the experimental data (stage and temperature dependence of ${\chi }_{or}^c$) and suggest that measurements of the stage dependence of ${\chi }_{or}^c$ can be used to estimate the $c$-axis screening length in graphite intercalation compounds.