5.5 Ferrimagnetism

Fundamentals of ferrimagnetism

Ferrimagnetism is a type of magnetic ordering where magnetic moments on different sublattices align antiparallel, but because those moments are unequal in magnitude, the material retains a net magnetization. It sits conceptually between ferromagnetism (parallel alignment, strong net moment) and antiferromagnetism (antiparallel alignment, zero net moment).

Understanding ferrimagnets matters because they combine useful magnetic properties with practical advantages like high electrical resistivity, making them essential in high-frequency electronics, magnetic storage, and emerging spintronic devices.

Definition and basic concepts

A ferrimagnetic material has two or more magnetic sublattices whose moments point in opposite directions but don't fully cancel. The result is a spontaneous net magnetization that persists below a critical temperature called the Néel temperature ($T_N$).

Key characteristics:

• Antiparallel alignment of unequal magnetic moments across sublattices
• Spontaneous magnetization below $T_N$ without an applied field
• Strong temperature dependence, often including a compensation point where the net moment temporarily drops to zero
• Hysteresis behavior similar to ferromagnets

Comparison with ferromagnetism

Ferrimagnets and ferromagnets share several macroscopic features: both show spontaneous magnetization, hysteresis loops, and domain structures. The differences are structural and subtle but important.

• Ferromagnets have one magnetic sublattice with moments aligned parallel. Ferrimagnets have two or more sublattices with moments aligned antiparallel.
• The critical temperature in ferromagnets is the Curie temperature ($T_C$); in ferrimagnets it's the Néel temperature ($T_N$). Both mark the transition to a paramagnetic state.
• The magnetization-vs-temperature curve for ferrimagnets can be non-monotonic. Some ferrimagnets display a compensation temperature ($T_{comp}$) below $T_N$ where the sublattice magnetizations exactly cancel and the net moment passes through zero. Ferromagnets don't show this behavior.

Historical background

Louis Néel proposed the theory of ferrimagnetism in 1948 to explain anomalous magnetic behavior observed in certain iron oxide compounds (ferrites). These materials clearly weren't simple ferromagnets or antiferromagnets, and Néel's two-sublattice model provided the missing framework.

His contributions to understanding both ferrimagnetism and antiferromagnetism earned him the 1970 Nobel Prize in Physics. The discovery of ferrimagnets opened the door to a whole class of technologically important magnetic oxides used in electronics and telecommunications.

Magnetic ordering in ferrimagnets

The magnetic ordering in ferrimagnets arises from the interplay between distinct magnetic sublattices coupled through exchange interactions. The details of this ordering control the material's Néel temperature, compensation behavior, and magnetic anisotropy.

Sublattice structure

Ferrimagnets contain two or more interpenetrating magnetic sublattices, each occupied by ions carrying different magnetic moments. In spinel ferrites, for example, magnetic cations sit on two crystallographically distinct sites: tetrahedral (A) sites and octahedral (B) sites. The moments on A sites point opposite to those on B sites.

Because the ions on each site differ in species, valence, or number, the sublattice magnetizations are unequal. That imbalance is what produces the net moment.

Exchange interactions

The antiparallel alignment between sublattices is driven by negative exchange interactions. In oxide-based ferrimagnets (like ferrites and garnets), the dominant mechanism is superexchange: magnetic cations interact indirectly through an intervening oxygen anion. Direct exchange can also play a role in metallic systems like rare earth-transition metal alloys.

The strength of these exchange interactions sets the energy scale for magnetic ordering and directly determines $T_N$. Stronger exchange means a higher Néel temperature.

Néel temperature

The Néel temperature ($T_N$) is the critical temperature above which thermal fluctuations destroy the long-range ferrimagnetic order, and the material becomes paramagnetic.

• Analogous to the Curie temperature in ferromagnets
• Determined primarily by the strength of inter-sublattice exchange interactions
• Magnetic anisotropy also influences the sharpness and nature of the transition
• Typical values range widely: ~580°C for magnetite ($\text{Fe}_3\text{O}_4$), while some garnets exceed 300°C

Types of ferrimagnetic materials

Ferrimagnetic materials span a broad range of crystal structures and compositions. The three major families each offer distinct advantages.

Ferrites vs garnets

Spinel ferrites have the general formula $MFe_2O_4$, where $M$ is a divalent metal ion (e.g., Mn, Ni, Zn, Co). Their key practical advantage is high electrical resistivity, which suppresses eddy current losses and makes them ideal for high-frequency applications.

Garnets have a more complex cubic structure with the formula $R_3Fe_5O_{12}$, where $R$ is a rare earth ion (e.g., Y, Gd). Yttrium iron garnet (YIG) is the most well-known example. Garnets generally have lower saturation magnetization than ferrites but can achieve higher Néel temperatures and exceptionally narrow ferromagnetic resonance linewidths, making them valuable for microwave applications.

Rare earth-transition metal alloys

These alloys combine the localized 4f magnetism of rare earth elements with the itinerant 3d magnetism of transition metals. Examples include GdFeCo, TbFe, and DyCo alloys.

The two sublattices (rare earth and transition metal) have very different temperature dependences, which produces strong compensation effects. This makes these alloys useful for:

• Magneto-optical recording (the compensation point enables thermomagnetic writing)
• Ultrafast all-optical magnetic switching
• Spintronic devices requiring tunable magnetic properties

Other ferrimagnetic compounds

• Hexagonal ferrites (e.g., $BaFe_{12}O_{19}$): very high magnetocrystalline anisotropy, used in permanent magnets and millimeter-wave devices
• Chromium chalcogenides (e.g., $CuCr_2Se_4$): unusual combination of ferrimagnetic order and metallic conductivity
• Prussian blue analogues: molecule-based magnets with chemically tunable magnetic properties
• Metal-organic frameworks: organic-based ferrimagnets being explored for lightweight and flexible magnetic materials

Microscopic origin of ferrimagnetism

The microscopic origin of ferrimagnetism lies in quantum mechanics: the electronic structure of magnetic ions and the exchange pathways connecting them.

Atomic magnetic moments

Magnetic moments in ferrimagnets come from unpaired electrons in partially filled d orbitals (transition metals) or f orbitals (rare earths).

For transition metal ions in crystal fields, orbital angular momentum is often largely quenched, so the moment is dominated by spin. The effective moment is then approximately $\mu_{eff} \approx g\sqrt{S(S+1)}\,\mu_B$, where $S$ is the total spin quantum number and $g \approx 2$.

Rare earth ions, by contrast, retain significant orbital angular momentum because the 4f electrons are shielded from crystal fields by outer shells. Their moments are described by the total angular momentum $J$ via Hund's rules.

Superexchange mechanism

Superexchange is the dominant exchange mechanism in oxide ferrimagnets. Here's how it works:

1. Two magnetic cations (e.g., $Fe^{3+}$) sit on opposite sides of a non-magnetic anion (e.g., $O^{2-}$).
2. Virtual electron transfer occurs: an electron from the oxygen hops to one cation, and the oxygen "borrows" an electron from the other cation.
3. The Pauli exclusion principle and Hund's rules constrain the spin orientations during these virtual hops, favoring antiparallel alignment of the two cations.

The sign and strength of superexchange depend on the bond angle ($M$ - $O$ - $M$) and the d-orbital occupancy of the cations. A 180° bond angle with half-filled orbitals gives the strongest antiferromagnetic coupling (the Goodenough-Kanamori-Anderson rules govern these details).

Molecular field theory

Molecular field theory (also called mean-field theory) treats the exchange interactions experienced by each ion as an effective internal field. For a two-sublattice ferrimagnet, each sublattice feels a molecular field from both its own sublattice and the opposing one.

This approach:

• Predicts the temperature dependence of each sublattice magnetization
• Accounts for the existence of compensation points
• Gives reasonable estimates of $T_N$

Its main limitation is that it ignores fluctuations and short-range correlations, so it becomes inaccurate near phase transitions where critical fluctuations dominate.

Macroscopic properties

The macroscopic magnetic behavior of ferrimagnets follows directly from their sublattice structure and is what makes them useful in devices.

Spontaneous magnetization

Below $T_N$, a ferrimagnet has a net magnetization equal to the difference between its sublattice magnetizations: $M_{net}(T) = |M_A(T) - M_B(T)|$.

Because the two sublattices generally have different temperature dependences, the net magnetization curve can be non-monotonic. In materials where the sublattice magnetizations cross (e.g., GdFeCo around 220 K), a compensation temperature $T_{comp}$ exists where $M_{net} = 0$. Above and below $T_{comp}$, the net moment is nonzero but dominated by different sublattices.

Magnetic susceptibility

The magnetic susceptibility $\chi$ of a ferrimagnet measures its magnetization response to an applied field. It shows complex temperature dependence:

• Below $T_N$: the material is spontaneously magnetized, so $\chi$ is governed by domain processes
• Near $T_N$: $\chi$ typically peaks as the material transitions to the paramagnetic state
• Above $T_N$: $\chi$ follows a modified Curie-Weiss law (see below)

Susceptibility can also be anisotropic, reflecting the crystal symmetry and magnetocrystalline anisotropy of the material.

Curie-Weiss law for ferrimagnets

Above $T_N$, the susceptibility of a ferrimagnet follows a hyperbolic form rather than the simple Curie-Weiss law of ferromagnets. The general expression is:

$\chi = \frac{C}{T - \theta}$

where $C$ is the Curie constant and $\theta$ is the Weiss temperature. For ferrimagnets, $\theta$ can be positive or negative depending on the relative strengths of intra- and inter-sublattice exchange.

In practice, the susceptibility above $T_N$ often shows a curved $1/\chi$ vs. $T$ plot (concave upward), unlike the straight line expected for simple ferromagnets. This curvature is a signature of competing sublattice interactions. Deviations from Curie-Weiss behavior close to $T_N$ indicate short-range magnetic order persisting above the transition.

Domain structure in ferrimagnets

Like ferromagnets, ferrimagnets break into magnetic domains to reduce their total energy. Domain structure and dynamics directly control the material's response to applied fields.

Domain wall formation

Domains form to minimize magnetostatic energy (the energy stored in the stray field outside the material). Within each domain, the magnetization points in a uniform direction; neighboring domains have different orientations.

The boundaries between domains are domain walls. Two main types exist:

• Bloch walls: the magnetization rotates through the plane of the wall (common in bulk materials)
• Néel walls: the magnetization rotates within the plane of the wall (favored in thin films)

Domain wall width is set by the competition between exchange stiffness (which favors wide walls) and magnetic anisotropy (which favors narrow walls). Typical wall widths range from tens to hundreds of nanometers.

Magnetization processes

When you apply an external field to a ferrimagnet, magnetization proceeds through two main mechanisms:

1. Domain wall motion: walls shift so that domains aligned with the field grow at the expense of unfavorably oriented domains. This dominates at low fields.
2. Domain rotation: within each domain, the magnetization rotates toward the field direction. This dominates at higher fields approaching saturation.

Defects, grain boundaries, and inclusions can pin domain walls, requiring additional field energy to move them. This pinning is a major source of coercivity and hysteresis.

Hysteresis behavior

Ferrimagnets display hysteresis loops characterized by three key parameters:

• Saturation magnetization ($M_s$): the maximum magnetization when all moments are aligned
• Remanence ($M_r$): the magnetization remaining after the field is removed
• Coercivity ($H_c$): the reverse field needed to reduce the magnetization to zero

The shape of the hysteresis loop depends on domain structure, anisotropy, and defect density. The area enclosed by the loop equals the energy dissipated per magnetization cycle, which matters for both power loss (in transformer cores) and permanent magnet performance.

Applications of ferrimagnetic materials

The combination of strong magnetization, high resistivity, and tunable properties makes ferrimagnets indispensable in many technologies.

Magnetic storage devices

• Ferrite cores were the basis of computer memory from the 1950s through the 1970s, with each tiny toroid storing one bit
• Garnet films enabled magnetic bubble memory, where cylindrical domains represented data bits
• Thin-film ferrimagnets (especially rare earth-transition metal alloys) are explored for next-generation recording media
• Emerging concepts include racetrack memory, where data is encoded in domain walls moving along nanowires

Microwave devices

Ferrites are the material of choice for many microwave components because their high resistivity minimizes eddy current losses at GHz frequencies.

• Circulators and isolators: route microwave signals in one direction using the non-reciprocal properties of magnetized ferrites
• Phase shifters and tunable filters: exploit the field-dependent permeability of ferrites
• Hexagonal ferrites: extend ferrite device operation into the millimeter-wave range (30-300 GHz)

YIG (yttrium iron garnet) is particularly valued for its extremely narrow ferromagnetic resonance linewidth, enabling high-quality-factor microwave filters.

Magnetic sensors

• Fluxgate magnetometers use ferrimagnetic cores to detect weak magnetic fields (down to ~0.1 nT), widely used in navigation and geophysical surveys
• Magnetoimpedance sensors based on amorphous ferrimagnetic alloys offer high sensitivity at low cost
• Giant magnetoresistance (GMR) devices can incorporate ferrimagnetic layers for field sensing
• Biomedical applications include detection of biomagnetic fields from the heart and brain

Experimental techniques

Several complementary techniques are used to characterize ferrimagnetic materials, probing everything from atomic-scale spin arrangements to bulk magnetic response.

Neutron diffraction

Neutrons carry a magnetic moment, making them uniquely suited to probe magnetic structure. In a neutron diffraction experiment on a ferrimagnet, you can:

• Determine the crystal and magnetic structure simultaneously
• Measure individual sublattice magnetizations and their temperature dependence
• Detect magnetic phase transitions and spin reorientation events
• Study magnetic excitations (magnons) through inelastic neutron scattering

This technique is especially powerful because X-rays interact primarily with charge, not spin, so neutrons provide magnetic information that X-rays cannot.

Mössbauer spectroscopy

Mössbauer spectroscopy exploits the recoil-free emission and absorption of gamma rays by nuclei (most commonly $^{57}Fe$) to probe local magnetic environments.

• The hyperfine field at the nucleus reveals the local magnetic moment and its orientation
• Different crystallographic sites (e.g., tetrahedral vs. octahedral in spinel ferrites) produce distinct spectral components, allowing you to study each site independently
• Temperature-dependent measurements track magnetic phase transitions and spin dynamics
• Sensitivity to valence state helps identify cation distributions in mixed ferrites

Magnetometry methods

• Vibrating Sample Magnetometer (VSM): measures bulk magnetization by detecting the voltage induced as a magnetized sample vibrates near pickup coils. Good for hysteresis loops and temperature-dependent magnetization.
• SQUID magnetometry: uses superconducting quantum interference to achieve extremely high sensitivity (~$10^{-8}$ emu), essential for studying weak signals or small samples.
• Torque magnetometry: measures the torque on a single crystal in a field, directly probing magnetocrystalline anisotropy constants.
• Magnetic force microscopy (MFM): images domain structure with ~50 nm spatial resolution by scanning a magnetic tip over the sample surface.

Theoretical models

Theoretical models provide the framework for interpreting experiments and predicting the behavior of new ferrimagnetic materials.

Néel's theory

Néel's two-sublattice molecular field theory is the foundational model of ferrimagnetism. It assigns each sublattice its own molecular field coefficient and solves for the temperature-dependent magnetization of each sublattice self-consistently.

Key predictions:

• The net magnetization $M(T)$ can show a variety of curve shapes depending on the ratio of sublattice moments and their exchange constants
• A compensation point exists when the two sublattice magnetizations cross at some $T < T_N$
• Néel classified the possible $M(T)$ curves into several types (N-type, Q-type, etc.) based on whether compensation occurs

Yafet-Kittel model

Néel's theory assumes collinear (parallel or antiparallel) spin arrangements. The Yafet-Kittel model relaxes this assumption by allowing canted (non-collinear) spin configurations within a sublattice.

This extension is necessary when competing exchange interactions (e.g., strong B-B interactions in spinel ferrites) frustrate simple collinear order. The Yafet-Kittel angle quantifies the degree of canting. Experimentally, canting reduces the net magnetization below the value predicted by Néel's collinear theory, which explains anomalously low moments observed in certain Zn-substituted ferrites.

Mean-field approximations

Mean-field theory replaces the many-body exchange problem with an effective average field acting on each spin. Beyond Néel's specific two-sublattice version, mean-field approaches are used to:

• Calculate thermodynamic quantities like specific heat and magnetic entropy
• Predict critical exponents near $T_N$ (though the predicted values are approximate; real systems show different critical behavior due to fluctuations)
• Provide starting points for more sophisticated treatments (spin-wave theory, Monte Carlo simulations)

The main shortcoming is the neglect of spatial fluctuations and short-range correlations, which become important near phase transitions and in low-dimensional systems.

Ferrimagnetism in nanostructures

When ferrimagnetic materials are reduced to nanometer dimensions, new physics emerges from finite-size effects, surface contributions, and modified exchange interactions.

Size effects on magnetic properties

As particle size decreases:

• The surface-to-volume ratio increases dramatically, and surface spins (which have fewer neighbors) experience weaker exchange. This can lead to a disordered surface "shell" surrounding an ordered core.
• Magnetic anisotropy changes because surface anisotropy and shape anisotropy become comparable to or larger than bulk magnetocrystalline anisotropy.
• The domain structure simplifies. Below a critical size (~tens of nm for most ferrites), particles become single-domain, meaning the entire particle is uniformly magnetized.
• In ultra-small nanoparticles (<5 nm), quantum confinement can modify the electronic structure and magnetic moments.

Nanoparticles vs thin films

Nanoparticles show size-dependent magnetic properties including a blocking temperature ($T_B$) below which the moment is stable and above which it fluctuates (superparamagnetic regime). Coercivity peaks at an intermediate size and drops to zero for very small particles.

Thin films exhibit thickness-dependent properties. Epitaxial strain from the substrate can modify both the magnetic anisotropy and $T_N$. Interface effects in multilayer structures can induce exchange bias or perpendicular magnetic anisotropy.

When nanoparticles are packed closely, inter-particle exchange coupling and dipolar interactions lead to collective magnetic behavior (superspin glass states, for example).

Superparamagnetism in ferrimagnets

Superparamagnetism occurs in single-domain nanoparticles when thermal energy ($k_BT$) exceeds the magnetic anisotropy energy barrier ($KV$, where $K$ is the anisotropy constant and $V$ is the particle volume).

Above the blocking temperature:

• The particle's magnetic moment fluctuates rapidly between easy-axis directions
• The ensemble shows zero coercivity and zero remanence
• The magnetization curve follows a Langevin function (no hysteresis)

The relaxation time follows the Néel-Arrhenius law: $\tau = \tau_0 \exp\left(\frac{KV}{k_BT}\right)$, where $\tau_0 \sim 10^{-9}$ to $10^{-12}$ s. This time-dependent behavior is critical for applications like magnetic hyperthermia therapy and contrast agents in MRI.

Emerging research areas

Ferrimagnetism remains an active frontier in condensed matter physics, with several directions attracting significant attention.

Multiferroic ferrimagnets

Multiferroics are materials that simultaneously exhibit more than one type of ferroic order (e.g., ferrimagnetic + ferroelectric). The coupling between magnetic and electric order parameters, called magnetoelectric coupling, enables control of magnetization with electric fields and vice versa.

Examples include certain rare earth manganites (e.g., $TbMnO_3$) and Z-type hexaferrites. Potential applications include electrically writable magnetic memory, magnetoelectric sensors, and tunable microwave devices.

Spintronics applications

Ferrimagnets offer several advantages over ferromagnets for spintronic devices:

• Faster domain wall motion: near the compensation point, domain walls in ferrimagnets can move at velocities exceeding 1 km/s, much faster than in ferromagnets
• Reduced stray fields: the partial cancellation of sublattice moments minimizes dipolar interactions between neighboring devices
• Ultrafast switching: all-optical switching of magnetization has been demonstrated in GdFeCo alloys on picosecond timescales using femtosecond laser pulses
• Spin-orbit torque switching: compensated ferrimagnets can be switched efficiently using current-induced spin-orbit torques

These properties make ferrimagnets promising for high-density, high-speed magnetic memory and logic.

Quantum ferrimagnets

At the frontier of fundamental research, quantum ferrimagnets explore regimes where quantum fluctuations play a dominant role:

• Spin chains and ladders with alternating spin magnitudes (e.g., $S = 1$ and $S = 1/2$) exhibit gapped or gapless excitation spectra depending on the topology
• Topological magnon bands have been predicted and observed in certain ferrimagnetic structures, analogous to topological electronic bands
• Low-dimensional ferrimagnetic systems serve as platforms for studying quantum entanglement, magnon transport, and potential applications in quantum sensing