Transport and magnetic properties of polycrystalline ${\mathrm{La}}_2{\mathrm{CoMnO}}_{6-\delta },$ $0<~\delta <~0.05,$ have revealed the existence of two distinguishable monoclinic ferromagnetic phases separated by a two-phase domain $0.02<~\delta <~0.05$ and a pseudotetragonal $(c/a<\mathrm{}\sqrt{2})$ phase with $\delta >~0.05$ that was prepared at 600 °C. A nearly oxygen-stoichiometric sample having a magnetization $M(5\mathrm{}\mathrm{K},50\mathrm{kOe})=5.78\mathrm{}{\mu }_B/\mathrm{f}.\mathrm{u}.\mathrm{}$ with a Curie temperature $T_c\approx 226\mathrm{K}$ is identified as atomically ordered ${\mathrm{La}}_2{\mathrm{Co}}^{2+}{\mathrm{Mn}}^{4+}{\mathrm{O}}_6$ containing about 1.8% antiferromagnetic spins at antisites. This ferromagnetic phase is an n-type polaronic conductor that progressively traps mobile electrons at the oxygen vacancies that introduced them on lowering the temperature. Although the x-ray-diffraction pattern can be indexed in orthorhombic (Pbnm) or monoclinic ${(P2\mathrm{}}_1/n)$ symmetry with $\beta \approx 90°,$ atomic order identifies the space group as ${P2\mathrm{}}_1/n.\mathrm{}$ A second monoclinic, ferromagnetic phase with $T_c<150\mathrm{}\mathrm{K}$ and $\delta \approx 0.05$ has a large, positive thermoelectric power that increases progressively with decreasing temperature. Quenching a $\delta \approx 0.02$ sample from 1350 °C into liquid ${\mathrm{N}}_2$ gave a single phase with $T_c=134\mathrm{}\mathrm{K}$ and $\delta \approx \mathrm{0.05.}$ A sample with $\delta >~0.05$ that was synthesized at 600 °C was pseudotetragonal $(c/a<\mathrm{}\sqrt{2})$ and had a paramagnetic Weiss constant $\theta <T_c\approx 225\mathrm{K}$ as well as a significantly smaller magnetization, but its magnetization curve $M\left(T\right)\mathrm{}$ showed no evidence of spin-glass behavior; its large, positive thermoelectric power was characteristic of polaronic conduction without trapping of mobile charge carriers at lower temperatures. Interpretation of the two phases with $\delta \approx 0.05$ is based on the hypothesis that introduction of high-spin ${\mathrm{Mn}}^{3+}$ by the oxygen vacancies creates around it additional ${\mathrm{Mn}}^{3+}$ and intermediate-spin ${\mathrm{Co}}^{3+}$ at neighboring sites; the resulting gain in elastic energy from cooperative, dynamic Jahn-Teller deformations at these ions must be sufficient to overcome the cost of about 0.2 eV for the electron transfer from a ${\mathrm{Co}}^{2+}\mathrm{ion}$ to a ${\mathrm{Mn}}^{4+}\mathrm{ion}.$

• Received 16 May 2002

DOI:https://doi.org/10.1103/PhysRevB.67.014401