Understanding insulating ferromagnetism in LaCoO3 films under tensile strain

This study utilizes density functional theory to reveal that the robust ferromagnetic insulating state in strained LaCoO₃ films arises from a unique ordered array of high-spin and low-spin Co³⁺ ions, where ferromagnetic superexchange interactions mediated by low-spin ions overcome competing antiferromagnetic forces to stabilize the ground state.

The Gist

The Big Picture: Turning a "Sleepy" Material into a "Superhero"

Imagine a material called Lanthanum Cobalt Oxide (LaCoO3). In its natural, bulk form (like a big block of it), this material is a bit of a "sleepy" insulator. It doesn't conduct electricity, and more importantly, it doesn't have a magnetic personality. The atoms inside are all in a "low-energy" state, sitting quietly with their magnetic spins canceled out.

However, scientists have noticed something weird: if you stretch this material out into a very thin film (like a microscopic sheet of plastic wrap) and glue it onto a specific type of crystal substrate, it wakes up! It suddenly becomes ferromagnetic (like a magnet) but still remains an insulator (it still blocks electricity).

This is a rare and valuable combination. Usually, if something is magnetic, it conducts electricity (like iron). If it's an insulator, it's not magnetic (like wood). Having both at once is like finding a car that is both a tank (strong) and a sports car (fast) at the same time. This is the "Holy Grail" for next-generation electronics (spintronics).

The big question was: How does stretching it make this happen?

The Experiment: Stretching the Material

The researchers used a supercomputer to simulate what happens when you stretch this material. They stretched it by about 2.4% (imagine stretching a rubber band just a tiny bit).

They discovered that this stretching forces the Cobalt atoms to change their "mood" or spin state.

• Low Spin (LS): The atoms are quiet, small, and non-magnetic.
• High Spin (HS): The atoms are loud, large, and magnetic.

In the stretched film, the atoms don't just randomly switch. They organize themselves into a very specific, orderly pattern.

The Discovery: The "Columnar" Dance

The most exciting finding is how the atoms arrange themselves. It's not a random mess.

Imagine a dance floor with a grid of dancers (the Cobalt atoms).

1. The Pattern: The dancers form 2x2 squares. Inside each square, you have a mix of "loud" (High Spin) and "quiet" (Low Spin) dancers.
2. The Separation: These 2x2 squares are separated by rows of only "quiet" dancers.
3. The Result: You get columns of mixed dancers, separated by walls of quiet dancers.

The researchers call this the "Ferromagnetic Columnar Model." It's like a city where you have blocks of active neighborhoods (the magnetic columns) separated by quiet parks (the non-magnetic walls).

Why is it Magnetic? The "Bridge" Analogy

Now, why does this specific pattern make the whole thing magnetic?

Usually, magnetic atoms hate being near each other if they are separated by an insulator; they tend to cancel each other out (Antiferromagnetism). But here, the "quiet" atoms (Low Spin) act as magic bridges.

• The 90-Degree Rule: When two "loud" (High Spin) atoms talk to each other through a "quiet" atom at a 90-degree angle, the quiet atom helps them agree to point in the same direction. It's like a mediator who says, "You two are actually on the same team!"
• The 180-Degree Rule: When they talk in a straight line (180 degrees), the quiet atom makes them point in opposite directions.

In this specific stretched pattern, the "90-degree" connections are much more frequent and stronger than the "180-degree" ones. The "teamwork" wins over the "canceling out." The result? The whole column points in the same direction, creating a strong magnet.

Why is it an Insulator? The "Traffic Jam"

You might ask, "If it's magnetic, why doesn't electricity flow?"

Think of electricity as cars trying to drive through a city.

• In a normal metal, the roads are wide and empty.
• In this material, the "High Spin" atoms are like big trucks (they take up more space), and the "Low Spin" atoms are like small cars.
• Because the atoms are arranged in this specific, alternating pattern of big trucks and small cars, the "roads" (electron paths) get jammed. The electrons get stuck in their specific spots and can't flow freely.

So, you have a material where the spins (the direction the atoms are pointing) are all marching in lockstep (Magnetic), but the electrons (the traffic) are stuck in a traffic jam (Insulator).

The Conclusion: Strain is the Key

The paper proves that you don't need to add impurities or remove oxygen (common tricks in chemistry) to get this effect. Stretching the material alone is enough.

By simply pulling the material tight (tensile strain), you force the atoms to rearrange into this special "columnar" dance. This dance creates a perfect balance where the material becomes a magnet without losing its insulating properties.

Why does this matter?
This discovery gives engineers a new "knob" to turn. If we want to build faster, more efficient computers or quantum devices, we can use this stretching technique to create materials that carry pure "spin" information without wasting energy as heat. It's like discovering a new way to build a super-efficient engine just by changing the shape of the pistons.

Technical Summary

1. Problem Statement

Lanthanum cobalt oxide ($LaCoO_3$) is a well-known diamagnetic insulator in its bulk form at low temperatures, with all Cobalt ions ($Co^{3+}$) in a low-spin (LS) state ($S=0$). However, when grown as thin films on substrates inducing epitaxial tensile strain (e.g., $SrTiO_3$), the material exhibits a robust ferromagnetic insulating (FMI) state with Curie temperatures ($T_C$) between 65–80 K.

Despite extensive experimental and computational studies, the microscopic origin of this FMI state remains unresolved. Key questions include:

• How does tensile strain alter the spin-state distribution of $Co^{3+}$ ions?
• What is the specific spatial ordering of these spin states (Low-Spin vs. High-Spin)?
• How can a ferromagnetic state coexist with insulating behavior, given that ferromagnetism in transition metal oxides is often linked to metallic double-exchange mechanisms?
• Previous hypotheses involving oxygen vacancies have been challenged by experiments showing no $Co^{2+}$ ions, suggesting strain alone is the driver.

2. Methodology

The authors employed Density Functional Theory (DFT) calculations to systematically investigate the magnetic ground state of stoichiometric $LaCoO_3$ under tensile strain equivalent to that imposed by a $SrTiO_3$ (STO) substrate.

• Computational Framework:
  • Software: Vienna Ab initio Simulation Package (VASP).
  • Functional: Meta-GGA (r2SCAN) for structural relaxations; r2SCAN+U for electronic structure (using $U_{eff} = 1$ eV for Co 3d states).
  • Supercell: A large $3 \times 3 \times 1$ pseudocubic supercell was constructed to accommodate complex spin-state ordering and octahedral tilting, avoiding the limitations of smaller cells used in prior studies.
  • Strain Conditions: In-plane lattice constants were fixed to the STO parameter ($a = 3.905$ Å), imposing ~2.4% tensile strain. The out-of-plane lattice constant was fully relaxed.
• Configuration Search:
  • The study considered 95 unique magnetic configurations of $Co^{3+}$ ions (High-Spin [HS] and Low-Spin [LS]) within the supercell, adhering to symmetry constraints and Goodenough-Kanamori rules.
  • Spin states considered: HS ($S=2$, $t_{2g}^4 e_g^2$) and LS ($S=0$, $t_{2g}^6 e_g^0$). Intermediate Spin (IS) was excluded as it was not stabilized by the strain.
• Analysis:
  • Energy Mapping: Total energies were mapped onto an Ising Hamiltonian to extract magnetic exchange coupling parameters ($J_0, J_1, J_2, J_3$).
  • Superexchange Analysis: Detailed perturbation theory (Appendix A) was used to derive the signs and magnitudes of exchange interactions based on orbital overlaps and Hund's coupling.

3. Key Contributions & Results

A. Identification of the Magnetic Ground State

The calculations identified a unique Ferromagnetic Columnar Model as the energetically favored ground state, lower in energy by at least 18.1 meV compared to previously proposed models (e.g., rocksalt ordering).

• Structure: The ground state consists of $2 \times 2$ columns of alternating HS and LS ions arranged in a rocksalt-like pattern within the plane.
• Separation: These ferromagnetic columns are separated by continuous planes of pure LS ions.
• Sequence: This creates a repeating HS–LS–LS sequence along both pseudocubic [100] and [010] directions.
• Magnetization: The configuration yields a net magnetization of 0.88 $\mu_B$ per Co ion, with an HS fraction of ~22.2% (16 out of 72 ions).

B. Structural and Electronic Properties

• Lattice Distortion: The HS $CoO_6$ octahedra exhibit an ~8.5% volume increase compared to bulk, while LS octahedra show minimal expansion. The in-plane Co–O bond lengths for HS ions elongate by ~4.1%.
• Insulating Nature: Electronic structure calculations confirm an insulating gap of 0.98 eV.
  • PDOS Analysis: LS ions show fully occupied $t_{2g}$ and empty $e_g$ orbitals. HS ions show partial $e_g$ occupancy. The system remains insulating because the HS ions are isolated by LS ions, preventing the delocalization required for metallic conductivity.

C. Mechanism of Ferromagnetism (Superexchange)

The study elucidates the competition between antiferromagnetic (AFM) and ferromagnetic (FM) superexchange pathways mediated by diamagnetic LS ions:

1. Intra-column Interactions:
   • $J_0$ (180° path): AFM interaction between HS ions via a single LS ion along the z-axis. Scales as $t_{pd}^8$.
   • $J_1$ (90° path): FM interaction between HS ions via a corner LS ion in the xy-plane. This is driven by Hund's coupling ($J_H$) on the LS ion, which stabilizes parallel spins in orthogonal $e_g$ orbitals.
2. Inter-column Interactions:
   • $J_2$ (180° path): Longer-range AFM interaction between columns.
   • $J_3$ (90° path): FM interaction between columns (L-shaped path).

Key Finding: While 180° paths are AFM, the 90° ferromagnetic interactions ($J_1$ and $J_3$) are sufficiently strong to overcome the competing AFM interactions. The multiplicity of FM neighbors and the specific orbital geometry (involving orthogonal $e_g$ orbitals on LS ions) stabilize the long-range ferromagnetic order.

D. Role of Strain

The study confirms that epitaxial tensile strain alone is sufficient to drive the LS-to-HS transition. Above a critical strain of ~1.8%, the HS state becomes energetically favorable. This eliminates the need to invoke oxygen vacancies or non-stoichiometry to explain the magnetism in stoichiometric films.

4. Significance

• Resolution of a Long-standing Puzzle: The paper provides a definitive microscopic mechanism for the FMI state in strained $LaCoO_3$, resolving the conflict between previous models and experimental observations.
• Strain Engineering: It establishes epitaxial strain as a powerful, controllable knob for inducing spin-state transitions and stabilizing exotic magnetic phases (FMI) without chemical doping.
• Spintronics Applications: By demonstrating a material that is both ferromagnetic and insulating, the work supports the development of next-generation spintronic devices that require pure spin currents with minimal charge flow (e.g., for quantum information processing).
• Theoretical Framework: The derivation of superexchange parameters in complex HS-LS-LS systems offers a robust theoretical framework for understanding magnetism in other correlated oxide heterostructures.

In conclusion, the authors demonstrate that tensile strain induces a specific columnar ordering of High-Spin and Low-Spin states in $LaCoO_3$. This ordering creates a unique superexchange network where 90° ferromagnetic interactions dominate, resulting in a stable ferromagnetic insulator.