Using a single crystal from which nonstoichiometries have been removed, the principal susceptibilities, ${\chi }_{\parallel }$ and ${\chi }_{\perp }$, of CoO have been determined in the temperature range of 4.2 to 350 K, bracketing $T_N\simeq 289$ K. A stress ≃ 210 bars along [001] was applied during measurements and in cooling through $T_N$ (to remove $T$ domains), and the initial $\chi$ was measured for $\overrightarrow{\mathrm{H}}\parallel \mathrm{}\left[001\right]\mathrm{}$ and $\overrightarrow{\mathrm{H}}\perp \mathrm{}\left[001\right]\mathrm{}$ at $\overrightarrow{\mathrm{H}}=250\mathrm{}$ Oe. These measured $\chi$'s are then related to ${\chi }_{\parallel }$ and ${\chi }_{\perp }$ using the known magnetic structure of CoO, from which ${\chi }_{\parallel }$ and ${\chi }_{\perp }$ at various temperatures are determined. Then, using the relation ${\chi }_p=\frac{({\chi }_{\parallel }+2\mathrm{}{\chi }_{\perp })}{3}$, the calculated ${\chi }_p$ (powder) is found to agree well with the measured ${\chi }_p$ at all temperatures, thus providing a consistency check on the procedures used in this work. For $T\to 0$ K, the experimental ${\chi }_{\parallel }\mathrm{}\left(0\right)\mathrm{}\simeq 39.5\times {10}^{-6\mathrm{}}$ ${\mathrm{cm}}^3$/g. This value is about 50% higher than the only available theoretical estimate by Tachiki [although his estimates of ${\chi }_{\perp }$ (0) and $\chi \left(T_N\right)$ are in good agreement with our measured values]. Various sources for this discrepancy are examined, and it is suggested that further theoretical work is needed to explain this discrepancy as well as the observed temperature dependence of ${\chi }_{\parallel }$ and ${\chi }_{\perp }$. A comparison with the recent results of a similar study in MnO is also made.

• Received 3 March 1981

DOI:https://doi.org/10.1103/PhysRevB.24.419