La2CoMnO6: Magnetic Double Perovskite

• La2CoMnO6 is a magnetic double perovskite with variable crystal symmetries arising from cation ordering and oxygen stoichiometry.
• Its ferromagnetic superexchange yields Curie temperatures from 150 to 230 K, with properties sensitive to antisite disorder and microstructure.
• LCMO shows significant magnetodielectric and magnetostriction effects that couple magnetic order to dielectric response and lattice strain.

La$_2$CoMnO$_6$ (LCMO) is a prototypical magnetic double perovskite of the $A_2BB'O_6$ class in which coupled spin, lattice, dielectric, and microstructural degrees of freedom generate a diverse set of phenomena, including ferromagnetism, magnetodielectricity, magnetostriction, spin–phonon coupling, and exchange-bias-related interface effects. Across ceramics, floating-zone single crystals, nanoparticles, and epitaxial thin films, LCMO has been reported in monoclinic $P2_1/n$, orthorhombic $Pnma$, pseudo-tetragonal, and rhombohedral variants, with the observed symmetry, transition temperatures, and functional responses depending strongly on B-site order, oxygen stoichiometry, antisite disorder, and grain-boundary topology (Silva et al., 2018, Macchiutti et al., 2021, Murthy et al., 2012, Galceran et al., 2014).

1. Crystal chemistry, symmetry, and structural variability

In the fully ordered limit, LCMO belongs to the double-perovskite motif $A_2B'B''O_6$, with La on the $A$ sites and Co/Mn occupying corner-sharing octahedra in a rock-salt pattern. Ideal B-site order is associated with cubic $Fm\bar{3}m$ symmetry, but octahedral tilts lower the symmetry in real samples; room-temperature forms reported for LCMO include monoclinic $P2_1/n$, orthorhombic $Pnma$, and pseudo-tetragonal structures, while rhombohedral phases occur at elevated temperature or in compositionally modified systems (Silva et al., 2018, Silva et al., 2018, Macchiutti et al., 2021, Murthy et al., 2012).

Ceramic LCMO prepared by infrared and Raman studies was refined in monoclinic $_6$0 with lattice parameters near $_6$1 Å, $_6$2 Å, $_6$3 Å, and $_6$4, while a microwave-synthesized polycrystalline sample gave $_6$5 Å, $_6$6 Å, $_6$7 Å, $_6$8, and $_6$9 Å$A_2BB'O_6$0 (Silva et al., 2018, Manikandan et al., 2023). A highly ordered polycrystalline sample used for magnetodielectric work also refined as monoclinic $A_2BB'O_6$1, with $A_2BB'O_6$2 Å, $A_2BB'O_6$3 Å, $A_2BB'O_6$4 Å, $A_2BB'O_6$5 (Murthya et al., 2014). By contrast, floating-zone single-crystal LCMO was refined in orthorhombic $A_2BB'O_6$6 at $A_2BB'O_6$7 K with $A_2BB'O_6$8 Å, $A_2BB'O_6$9 Å, and $P2_1/n$0 Å; the average octahedral geometry was reported as $P2_1/n$1 Å and $P2_1/n$2, with a tolerance factor

$P2_1/n$3

consistent with the $P2_1/n$4 distortion (Macchiutti et al., 2021).

Structural variability is also controlled by oxygen content and epitaxial constraint. In RF-sputtered LCMO$P2_1/n$5 thin films on $P2_1/n$6 SrTiO$P2_1/n$7, reciprocal-space maps identified two crystallographic orientations of the pseudocubic cell: a $P2_1/n$8-axis normal to the surface and a $P2_1/n$9-axis lying in-plane. Their relative fractions evolve with oxygen stoichiometry, and the out-of-plane lattice parameter contracts monotonically as $Pnma$0; for representative films, $Pnma$1 changes from $Pnma$2 Å to $Pnma$3 Å while the pseudocubic cell volume approaches the bulk value (Galceran et al., 2014). The high-temperature phase behavior is likewise nontrivial: far-infrared reflectivity established a first-order structural phase transition near $Pnma$4 K from monoclinic $Pnma$5 to rhombohedral $Pnma$6, with the Co$Pnma$7/Mn$Pnma$8 rock-salt ordering retained but octahedral tilts reduced (Silva et al., 2018).

2. Ferromagnetism, competing exchange, and magnetic disorder

The dominant magnetic interaction in ordered LCMO is the ferromagnetic superexchange along Co$Pnma$9–O–Mn$A_2B'B''O_6$0 linkages. In bulk-like polycrystalline samples, this interaction produces Curie temperatures near $A_2B'B''O_6$1–$A_2B'B''O_6$2 K: microwave-synthesized LCMO exhibits $A_2B'B''O_6$3 K, conventionally heated material gives $A_2B'B''O_6$4 K, and Raman/magnetodielectric studies report $A_2B'B''O_6$5 K (Manikandan et al., 2023, Murthya et al., 2014, Silva et al., 2018). Nanoparticles show a lower primary ferromagnetic transition at $A_2B'B''O_6$6 K together with a second transition at $A_2B'B''O_6$7 K, assigned to ordered Co$A_2B'B''O_6$8–O–Mn$A_2B'B''O_6$9 superexchange and disordered Co$A$0–O–Mn$A$1 vibronic superexchange, respectively (Murthy et al., 2012, Murthy et al., 2013). In floating-zone single crystals, the characteristic temperatures are further shifted: LCMO shows $A$2 K associated with Co$A$3–O–Mn$A$4 order and $A$5 K associated with Co$A$6–O–Mn$A$7 short-range order (Macchiutti et al., 2021).

Magnetic moments are correspondingly sample dependent. The single crystal reaches $A$8/f.u. at $A$9 K, close to the ideal $Fm\bar{3}m$0 value with an additional contribution from approximately $Fm\bar{3}m$1 Co$Fm\bar{3}m$2–Mn$Fm\bar{3}m$3 (Macchiutti et al., 2021). Ordered polycrystalline LCMO reaches approximately $Fm\bar{3}m$4/f.u. at $Fm\bar{3}m$5 T and $Fm\bar{3}m$6 K, consistent with low antisite disorder of about $Fm\bar{3}m$7 (Murthya et al., 2014). Microwave-synthesized material gives $Fm\bar{3}m$8/f.u., below the ideal value because of some antisite disorder (Manikandan et al., 2023). Nanoparticles are substantially more disordered: $Fm\bar{3}m$9/f.u. even at $P2_1/n$0 T, with incomplete saturation, mixed Co and Mn valence states from XPS, and clear evidence that antisite disorder and vacancies suppress the moment (Murthy et al., 2012, Murthy et al., 2013).

Antisite disorder modifies the exchange network by creating Co–O–Co and Mn–O–Mn linkages, which introduce antiferromagnetic pathways into a nominally ferromagnetic matrix. Raman spectroscopy has shown that partially ordered LCMO can host a short-range antiferromagnetic interaction around $P2_1/n$1 K that is not detected in the bulk magnetization curve; the relevant anomaly appears instead in phonon frequency and linewidth (Silva et al., 2018). In nanoparticles, further cooling below about $P2_1/n$2 K produces a reentrant cluster-glass state. Its dynamical signature includes a frequency-sensitive ac-susceptibility peak with $P2_1/n$3, power-law critical slowing down with $P2_1/n$4 s, $P2_1/n$5, and $P2_1/n$6 K, together with the absence of a memory effect and only a weak shift under dc bias, consistent with weakly interacting clusters (Murthy et al., 2013).

Oxygen stoichiometry provides an additional control parameter. In thin films, tuning the annealing temperature, oxygen partial pressure, and cooling rate shifts $P2_1/n$7 from about $P2_1/n$8 K in oxygen-poor films to $P2_1/n$9–$Pnma$0 K in oxygen-rich films, reflecting the competition between the intended Co$Pnma$1–Mn$Pnma$2 ferromagnetic superexchange and the disorder-induced pathways introduced by vacancies and antisites (Galceran et al., 2014).

3. Dielectric response and magnetodielectricity

The intrinsic dielectric response of LCMO is modest when extracted from lattice-dynamical spectroscopy. Infrared-reflectivity analysis of a Pechini-derived ceramic fitted the spectrum with $Pnma$3 TO–LO oscillator pairs and obtained

$Pnma$4

with $Pnma$5 and $Pnma$6, giving

$Pnma$7

That work concluded that the commonly reported colossal dielectric constant is extrinsic, arising from Maxwell–Wagner interfacial polarization rather than an additional intrinsic low-frequency polar mode (Silva et al., 2014). A temperature-dependent far-infrared study of partially ordered LCMO similarly found a modest intrinsic static dielectric constant, with $Pnma$8, $Pnma$9, and $_6$00 at room temperature (Silva et al., 2018).

The magnetodielectric response is commonly quantified as

$_6$01

In a highly ordered polycrystalline sample, the MD effect reaches approximately $_6$02 at $_6$03 kHz under $_6$04 T near $_6$05 K, and at $_6$06 kHz the peak is approximately $_6$07 near $_6$08 K (Murthya et al., 2014). In LCMO nanoparticles, two MD maxima track the two magnetic transitions: under $_6$09 T, the peak near $_6$10 is about $_6$11 at $_6$12 kHz and about $_6$13 at $_6$14 kHz, while the peak near $_6$15 is about $_6$16 at $_6$17 kHz and about $_6$18 at $_6$19 kHz (Murthy et al., 2012). These values establish that LCMO can sustain sizeable field-tunable dielectric anomalies over a broad frequency range, but they also show strong sensitivity to frequency and microstructure.

Impedance spectroscopy resolves the competition between intrinsic and extrinsic contributions. In nanoparticles, Nyquist plots are fitted by two RC elements in series,

$_6$20

with grain-boundary and bulk arcs separating above about $_6$21 K. The extracted capacitances $_6$22 and $_6$23 show anomalies at both dielectric relaxations, while the resistive components do not show corresponding steps at the magnetic ordering temperatures; under $_6$24 T, both relaxations shift to higher frequency (Murthy et al., 2012). In ordered polycrystals, impedance under magnetic field shows that the high-frequency bulk arc decreases under $_6$25 T, supporting an intrinsic magnetodielectric component, whereas the low-frequency response correlates with grain-boundary Maxwell–Wagner polarization and magnetoresistance (Murthya et al., 2014).

Magnetic disorder weakens the intrinsic component. Sr doping at the A site, which introduces hole carriers and increases antisite disorder, transforms ordered monoclinic LCMO into a rhombohedral $_6$26 phase, suppresses the intrinsic low-temperature permittivity from about $_6$27 to about $_6$28, and brings the MD and magnetoresistance magnitudes into close correspondence at about $_6$29, indicating predominantly extrinsic behavior. In the ordered sample, by contrast, MD at $_6$30 kHz exceeds the corresponding magnetoresistance, consistent with a substantial intrinsic contribution attributed to spin–lattice coupling and asymmetric hopping (Murthya et al., 2014).

4. Phonons, spin–phonon coupling, and the high-temperature structural transition

LCMO has a rich vibrational spectrum that is strongly coupled to both magnetism and structure. At room temperature, Raman spectroscopy of monoclinic $_6$31 material shows a strong symmetric-stretching $_6$32 mode near $_6$33 cm$_6$34 and a combined bending/anti-stretching feature near $_6$35–$_6$36 cm$_6$37, along with lower-intensity modes near $_6$38 and $_6$39 cm$_6$40 and overtone or combination bands around $_6$41 and $_6$42 cm$_6$43 (Murthya et al., 2014, Silva et al., 2018). The temperature dependence of the symmetric stretching mode departs from a purely anharmonic trend below the ferromagnetic transition. In ordered polycrystals, it softens anomalously by about $_6$44 cm$_6$45 below $_6$46 K, and its linewidth exhibits a kink at the same temperature; the phonon renormalization was analyzed using the Balkanski form

$_6$47

together with the mean-field spin–phonon scaling

$_6$48

which supports intrinsic spin–lattice coupling as a source of the dielectric anomaly (Murthya et al., 2014).

Raman spectroscopy is also sensitive to hidden short-range magnetism. In partially ordered LCMO, the same symmetric-stretching mode near $_6$49 cm$_6$50 shows not only the expected anomaly at $_6$51 K but also a second discontinuity in both frequency and full width at half maximum at $_6$52 K. This lower-temperature anomaly was assigned to short-range antiferromagnetic exchange associated with antisite disorder, and it does not appear in the bulk magnetization curve, underscoring the local sensitivity of Raman probes (Silva et al., 2018).

The high-temperature structural transition is established most directly by far-infrared spectroscopy. Reflectivity measurements from $_6$53 to $_6$54 K show that near $_6$55 K the number of resolved phonon bands drops abruptly from $_6$56 to $_6$57, matching the reduction expected when the symmetry changes from monoclinic $_6$58 to rhombohedral $_6$59 (Silva et al., 2018). Pronounced kinks appear in the temperature dependences of both TO and LO phonon frequencies, and damping constants exhibit sharp peaks near the transition. Among the dominant dielectric modes, mode #2 softens by about $_6$60 cm$_6$61 and mode #3 by about $_6$62 cm$_6$63 on heating toward the transition, while mode #4 shows step-like softening and a divergent linewidth (Silva et al., 2018).

The anharmonicity is unusually strong. Grüneisen parameters extracted from the thermal evolution of selected modes are reported as $_6$64, $_6$65, $_6$66, $_6$67, and $_6$68, well beyond a simple quasi-harmonic picture (Silva et al., 2018). A plausible implication is that the same lattice degrees of freedom that drive the first-order structural transition also provide an efficient channel for magnetoelastic and magnetodielectric coupling at lower temperature.

5. Magnetostriction and magnetoelastic-electronic coupling

LCMO exhibits giant negative magnetostriction in the ferromagnetic state. In microwave-synthesized polycrystalline material, the longitudinal magnetostriction is defined as

$_6$69

and at $_6$70 K it remains unsaturated up to $_6$71 kOe, where $_6$72, or $_6$73 ppm (Manikandan et al., 2023). A conventionally synthesized comparison sample reaches $_6$74 ppm under the same conditions, showing that microwave processing preserves the effect while reducing the active reaction time to about $_6$75 min, with single-phase pellets forming within $_6$76 min of heating (Manikandan et al., 2023). The magnitude decreases monotonically with increasing temperature and becomes negligible above $_6$77 K (Manikandan et al., 2023).

The proposed microscopic origin is the single-ion spin–orbit mechanism on Co$_6$78. In this picture, trigonal distortion further splits the Co $_6$79 manifold, leaving a doublet that carries orbital angular momentum; spin–orbit coupling then rotates the orbital moment with the spin and produces strain along the applied field. Mn$_6$80 contributes negligibly to the orbital moment, and x-ray magnetic dichroism gives $_6$81–$_6$82 in bulk LCMO (Manikandan et al., 2023). This mechanism distinguishes LCMO from conventional ferromagnetic manganites, whose magnetostriction is both smaller and of opposite sign under the cited comparison (Manikandan et al., 2023).

A later field-dependent structural and spectroscopic analysis connected magnetostriction directly to the magnetodielectric effect. At $_6$83 K, orthorhombic $_6$84 LCMO shows anisotropic strains under $_6$85 T,

$_6$86

with $_6$87, $_6$88, and $_6$89, yielding

$_6$90

Simultaneously, the Co–O–Mn angles straighten from $_6$91, $_6$92, $_6$93 to $_6$94, $_6$95, and $_6$96 (Boldrin et al., 9 Sep 2025).

Electronic-structure measurements reinforce that picture. Charge-transfer-multiplet fits to Co $_6$97-edge x-ray absorption spectroscopy showed that the ligand-to-metal charge-transfer energy decreases from $_6$98 eV at $_6$99 K, $A_2BB'O_6$00 T to $A_2BB'O_6$01 eV at $A_2BB'O_6$02 K, $A_2BB'O_6$03 T, while the Co–O hopping amplitudes increase from $A_2BB'O_6$04 eV to $A_2BB'O_6$05 eV and from $A_2BB'O_6$06 eV to $A_2BB'O_6$07 eV (Boldrin et al., 9 Sep 2025). Special-quasirandom-structure DFT similarly found enhanced Co–O covalency under field. The reported phenomenological description writes

$A_2BB'O_6$08

leading to

$A_2BB'O_6$09

with $A_2BB'O_6$10 (Boldrin et al., 9 Sep 2025). In this formulation, the chain $A_2BB'O_6$11 increased covalency $A_2BB'O_6$12 provides a direct route from magnetic field to dielectric response.

6. Microstructure, exchange bias, thin-film control, and technological contexts

A central microstructural issue in LCMO is the difference between polycrystalline and single-crystalline exchange-bias behavior. Floating-zone single crystals of LCMO, La$A_2BB'O_6$13Ca$A_2BB'O_6$14CoMnO$A_2BB'O_6$15, and La$A_2BB'O_6$16Sr$A_2BB'O_6$17CoMnO$A_2BB'O_6$18 are all ferromagnetic, but none shows zero-field-cooled exchange bias or conventional exchange bias. For single-crystal LCMO at $A_2BB'O_6$19 K, the hysteresis loop is symmetric with $A_2BB'O_6$20 Oe, $A_2BB'O_6$21 Oe, $A_2BB'O_6$22 Oe, and

$A_2BB'O_6$23

and no conventional exchange bias appears even after field cooling in $A_2BB'O_6$24 T (Macchiutti et al., 2021). The absence is attributed to the lack of grain boundaries and the absence of spin-glass-like dynamics, as supported by frequency-independent ac susceptibility in the single crystals (Macchiutti et al., 2021).

This result sharply contrasts with polycrystalline Ca- and Sr-doped analogues, where spontaneous zero-field-cooled exchange bias has been associated with a low-temperature spin-glass-like cluster phase and grain-boundary interfaces that provide pinned-moment regions (Macchiutti et al., 2021). The comparison establishes that exchange bias in this family is not an intrinsic property of the ideal bulk crystal alone. Rather, it depends on interfacial exchange coupling generated by microstructure and glassy magnetism. This also clarifies a recurrent misconception in the LCMO literature: large functional responses are not always signatures of a homogeneous bulk phase, since both the colossal dielectric constant and spontaneous exchange bias have strong extrinsic components under the conditions where they are most prominent (Silva et al., 2014, Macchiutti et al., 2021).

Epitaxial films provide a separate route to microstructural engineering. In RF-sputtered LCMO$A_2BB'O_6$25 on $A_2BB'O_6$26 SrTiO$A_2BB'O_6$27, careful control of in situ annealing conditions tunes both oxygen uptake and crystallographic orientation. Oxygen-poor films with $A_2BB'O_6$28 show $A_2BB'O_6$29–$A_2BB'O_6$30 K and predominantly $A_2BB'O_6$31-in-plane orientation, whereas oxygen-rich films with $A_2BB'O_6$32 recover $A_2BB'O_6$33–$A_2BB'O_6$34 K and a single $A_2BB'O_6$35-out-of-plane orientation; mixed-orientation films display a characteristic magnetic “biloop” (Galceran et al., 2014). The same study links oxygen-vacancy compensation to partial Mn$A_2BB'O_6$36Mn$A_2BB'O_6$37 reduction, bond-angle changes, and altered coercive fields, identifying oxygen stoichiometry as a practical control parameter for oxide spin-filter barriers (Galceran et al., 2014).

The reported application contexts reflect this microstructural sensitivity. The single-crystal study emphasizes spintronic relevance and the role of interface engineering for exchange-bias devices and double-perovskite heterostructures (Macchiutti et al., 2021). Infrared work identifies LCMO as appropriate for microwave devices and circuitry, with $A_2BB'O_6$38 THz and moderate intrinsic permittivity (Silva et al., 2014). Nanoparticle measurements point to magnetic-field sensors, tunable capacitors in cryogenic electronics, and spintronics devices exploiting dielectric control by magnetic order (Murthy et al., 2012). Across these contexts, the recurring materials lesson is that LCMO functionality is jointly determined by cation order, oxygen stoichiometry, local disorder, and interface topology rather than by composition alone.