Magneto-Ionic Modulation: Mechanisms & Applications

• Magneto-ionic modulation is a reversible control technique that uses electric-field-driven ionic migration to alter magnetic properties in solid-state materials.
• It combines ionic drift, diffusion, and redox chemistry to achieve persistent changes in magnetization, anisotropy, and exchange bias across various material systems.
• This approach underpins energy-efficient spintronic, neuromorphic, and memory devices, offering tunable, multi-state functionality with high endurance.

Magneto-ionic modulation refers to the reversible control of magnetic properties in solid-state materials through gate-driven ion migration—predominantly oxygen, but also other ionic species (e.g., nitrogen, hydrogen, hydroxyl, boron, carbon, lithium)—under electric field. Unlike purely electrostatic mechanisms such as charge accumulation/depletion (conventional voltage-controlled magnetic anisotropy, VCMA), magneto-ionic effects involve field-activated, non-volatile redistribution of ions or vacancies, enabling direct and persistent alteration of magnetic parameters such as magnetization, anisotropy, exchange bias, and spin–orbit coupling. This process is fundamentally rooted in redox-driven ionic kinetics, drift–diffusion models, and interface engineering. The technology underpins energy-efficient, cyclable, and multi-state functionalities in emerging spintronic, neuromorphic, and memory devices (Rojas et al., 2020).

1. Fundamental Mechanisms: Ion Transport and Magnetic Modulation

Magneto-ionic effects rely on electric-field–driven migration of mobile ions—typically O²⁻ or N³⁻—within a magnetic host or at its interfaces. The fundamental steps are:

• Ionic drift and diffusion: Under an applied field ($E=V/d$), species such as O²⁻ migrate according to the Nernst–Planck equation:

$J_i = -D_i \nabla c_i + \mu_i c_i E$

where $J_i$ is the ionic flux, $D_i$ the diffusion coefficient, and $\mu_i$ the mobility of species $i$. The local concentration $c_i$ changes as:

$$\frac{\partial c_i}{\partial t} + \nabla\cdot J_i = 0$$

• Redox chemistry: Extraction (or insertion) of anionic species (e.g., O²⁻, N³⁻, OH⁻) at reactive interfaces reduces cations (e.g., Co³⁺ or Fe³⁺ to Co⁰, Fe⁰), forming metallic or low-valence clusters and thereby switching between paramagnetic and ferromagnetic phases.
• Interfacial and bulk effects: Migration can be confined to a few atomic layers (especially at high–$k$ oxide or multilayer interfaces) or can propagate as a planar chemical front tens of nanometers deep in the bulk, depending on materials engineering and device structure (Rojas et al., 2020, Gilbert et al., 2016).

Kinetically, the rate of magnetic modulation is governed by the drift velocity $v = \mu E$ and can be described by simple first-order or Avrami-type models: $J_i = -D_i \nabla c_i + \mu_i c_i E$0 where $J_i = -D_i \nabla c_i + \mu_i c_i E$1 is the saturation magnetization and $J_i = -D_i \nabla c_i + \mu_i c_i E$2 a characteristic time-scale set by ionic mobility and field.

2. Materials Systems and Device Architectures

Magneto-ionic modulation has been demonstrated in a variety of materials platforms:

• Oxide systems: Co₃O₄, Co/CoO, Fe–B–O, GdOₓ, HfO₂, ZrO₂ as high mobility ionic reservoirs or solid ionic conductors. Anode/cathode processes allow for room temperature ferromagnetic–paramagnetic interconversion (Rojas et al., 2020, Ma et al., 14 Mar 2025).
• Nitride systems: CoN and FeCoN films show uniform, energetically favorable, and cyclable N³⁻ migration, driven at lower threshold voltages and with enhanced rates over oxides (Rojas et al., 2020, Spasojevic et al., 2024).
• Hydroxide systems: $J_i = -D_i \nabla c_i + \mu_i c_i E$3–Co(OH)₂ supports rapid and reversible OH⁻ migration at low voltages (–2 to –8 V), yielding switchable ferromagnetic clusters (Quintana et al., 2022).
• Solid-state stacks: Pt/Co/Pd with YSZ, HfO₂, or ZrO₂ as ionic conductors enable nanosecond–microsecond switching of interfacial anisotropy via H⁺ or O²⁻ (Lee et al., 2020, Fassatoui et al., 2021).
• Multilayers and superlattices: Magnetic multilayers with PMA (e.g., [Co/Al/Pt]$J_i = -D_i \nabla c_i + \mu_i c_i E$4 stacks) allow non-volatile, layer-by-layer gating of participating magnetic units, modulating collective properties such as AHE amplitude and stripe domain periodicity (Gomes et al., 2023).

Device architectures frequently exploit buffer layers (e.g., TiN, SiO₂) to control electric field homogeneity and enhance ionic drift, with corresponding large improvements in both modulation amplitude and speed (Rojas et al., 2020). Conducting underlayers (e.g., TiN) can generate fields distributed uniformly across thick films ($J_i = -D_i \nabla c_i + \mu_i c_i E$5 nm), enabling nearly order-of-magnitude increases in $J_i = -D_i \nabla c_i + \mu_i c_i E$6 and drift rates.

3. Quantitative Modulation of Magnetic Properties

Magneto-ionic processes modulate several principal magnetic observables:

Parameter	Typical Modulation	Representative System	Reference
Saturation magnetization	$J_i = -D_i \nabla c_i + \mu_i c_i E$7 up to $J_i = -D_i \nabla c_i + \mu_i c_i E$8 emu/cm³	Co₃O₄ w/ TiN buffer	(Rojas et al., 2020)
Modulation rate	$J_i = -D_i \nabla c_i + \mu_i c_i E$9 $J_i$0 emu/cm³·h	CoN, Fe–C, Co₃O₄	(Rojas et al., 2020, Tan et al., 14 Mar 2025)
Coercivity ($J_i$1)	Up to 25× increase	Fe–C	(Tan et al., 14 Mar 2025)
Exchange bias ($J_i$2)	Up to 35% enhancement, multi-cycle	Gd/NiCoO, MnN/CoFe	(Murray et al., 2021, Jensen et al., 2023)
Anisotropy ($J_i$3)	Quenched/recovered by $J_i$4 J/m³	Co/Pt/YSZ; W/CoFeB/MgO/HfO₂	(Lee et al., 2020, Chen et al., 25 Feb 2025)
Endurance (cycles)	$J_i$5–$J_i$6	CoN, YSZ, GdOx systems	(Rojas et al., 2020, Lee et al., 2020)
Switching time ($J_i$7)	ms–min (tunable by $J_i$8, $J_i$9)	ZrO₂, YSZ, CoFeB/MgO	(Lee et al., 2020, Fassatoui et al., 2021)

Non-volatility is the hallmark: reversed electric fields restore the initial state, while repeated cycling establishes endurance suitable for device applications (e.g., $D_i$0 cycles for YSZ, GdOx, CoN).

4. Kinetic Models and Scaling Laws

Three principal kinetic regimes are established:

• Linear drift regime: Ionic speed $D_i$1, with $D_i$2.
• Exponential field-enhancement: At high $D_i$3, drift velocities and oxidation rates follow exponential scaling as per Cabrera–Mott:

$D_i$4

with $D_i$5 the activation energy (typically $D_i$6–$D_i$7 eV for O²⁻, lower for N³⁻), $D_i$8 a material-dependent characteristic field, and $D_i$9 the attempt frequency (Fassatoui et al., 2021).

• Logarithmic time-growth: Spontaneous re-oxidation follows $\mu_i$0, consistent with logarithmic Cabrera–Mott kinetics (notably observed in field-off relaxation) (Fassatoui et al., 2021).

Switching timescales ($\mu_i$1) can be tuned over five decades by engineering $\mu_i$2, film thickness, and choice of conductor: ms–μs by shrinking oxide thickness or using higher-mobility conductors (e.g., YSZ versus GdOx) (Lee et al., 2020).

5. Device Applications and Performance Trade-offs

Magneto-ionic modulation enables a spectrum of energy-efficient, addressable spintronic, memory, and computation platforms:

• MEMS/actuators: Moderate-speed, all-voltage control of magnetic state, high cyclability (e.g., $\mu_i$3), and modest switching rates fulfill the requirements for microresonators and sensors (Rojas et al., 2020).
• Neuromorphic computing: Ionic-driven plasticity with analog, nonvolatile weight updates, ultralow write energy ($\mu_i$4 attojoules/bit), and time constants emulating biological synapses (Rojas et al., 2020). Demonstrated implementations include reservoir nodes using voltage-controlled domain nucleation and skyrmion/stripe manipulation (Das et al., 2024).
• Spintronics and logic: Non-volatile, local tuning of $\mu_i$5, $\mu_i$6, and DMI allows field-free control of race-track bit lines, MRAM programmable states, and reconfigurable domain-wall/stripe domain periodicity (Balan et al., 2022, Gomes et al., 2023).
• Exchange bias programming: Field-cooling and voltage protocols can “write” and “erase” interfacial exchange-bias in FM/AFM heterostructures, enabling robust, multi-cycle memory states (Murray et al., 2021, Jensen et al., 2023).

Design must balance crucial trade-offs:

• Speed vs. reliability: Higher $\mu_i$7 or thinner films accelerate switching, but risk irreversible ion loss (e.g., O₂ bubble evolution) and structural damage (Rojas et al., 2020).
• Magnitude vs. energy: Larger $\mu_i$8 or $\mu_i$9 swings require higher voltages or longer pulses, increasing energy cost.
• Geometry and layer stack: Uniform fields via conducting buffers maximize usable volume and rates; interface/oxide choice (thickness, defect density) tunes endurance and selectivity.

6. Ion Species and Multi-ion Magneto-Ionics

Recent advances have extended magneto-ionic concepts beyond classic O²⁻ systems:

• Nitrogen magneto-ionics: Uniform, plane-wave migration fronts and low barrier (E$i$0 = 1.14 eV vs. 1.54 eV for O) facilitate lower turn-on voltages and higher rates, with cyclable, non-volatile switching well-suited for endurance-critical applications (Rojas et al., 2020, Spasojevic et al., 2024). Nitride-based exchange-bias programming and wireless magneto-ionics (via bipolar electrochemistry) also utilize N³⁻ carriers (Jensen et al., 2023, Ma et al., 2023).
• Hydroxide (OH⁻) and hydrogen (H⁺): Deliver ultralow voltage thresholds (–2 V for $i$1–Co(OH)₂) and rapid kinetics. Magneto-ionic devices based on H⁺ in YSZ exhibit sub-ms switching at room temperature (Quintana et al., 2022, Lee et al., 2020).
• Multi-ion systems and charge-transfer engineering: Fe–B–O and Fe–C heterostructures support simultaneous, field-driven migration of multiple cationic (Fe, B) and anionic (O, C) species. Multi-ion magneto-ionics offers programmable, multi-level control of $i$2, $i$3, and localized phase transformations; charge-transfer effects underpin these dynamics (Ma et al., 14 Mar 2025, Tan et al., 14 Mar 2025).

A central finding is that the electronegativity and ionic radius of the active species directly dictate the threshold voltage, rate, and reversibility of magnetic modulation. Systems with lower activation barriers and weaker bonding (e.g., N vs. O) offer superior energy efficiency and device endurance.

7. Theoretical Models and Scaling Outlook

The evolution of magneto-ionic theory couples ionic migration models (Nernst–Planck, Cabrera–Mott) with magnetic free energy landscapes:

• DFT and ab-initio models: Capture orbital hybridization, SOC, and anisotropy switching as a function of local ion coordination, e.g., Fe/O and Fe/HfO₂ configurations (Pietro et al., 2022).
• Micromagnetics: $i$4, $i$5, DMI, coercivity, and exchange bias are explicitly parametrized as a function of local ionic concentrations, redox fronts, and voltage protocols (Gomes et al., 2023, Balan et al., 2022).
• Multi-state and analog control: In topologically structured elements (e.g., FeCoN nanodots), the planar nature of ionic migration enables true analog—rather than digital—tuning of optical and magnetic observables (e.g., vortex-state nucleation, coercivity) (Spasojevic et al., 2024).

The scaling prospects are dictated by material choices, ion mobility, and device patterns. For instance, reducing oxide thickness to nanometer scale can bring switching into the sub-microsecond regime (Lee et al., 2020), while patterned electrodes and solid ionic conductors promise spatially resolved, reconfigurable functionality suitable for large-scale, low-power spintronic, neuromorphic, and quantum systems.

---

Key References:

• Boosting room temperature magneto-ionics in Co₃O₄ (Rojas et al., 2020)
• Fast magneto-ionic switching of interface anisotropy using yttria-stabilized zirconia gate oxide (Lee et al., 2020)
• Nitrogen magneto-ionics (Rojas et al., 2020)
• Hydroxide-based magneto-ionics: electric-field control of reversible paramagnetic-to-ferromagnetic switch in $i$6-Co(OH)₂ films (Quintana et al., 2022)
• Control of the magnetic anisotropy in multi-repeat Pt/Co/Al heterostructures using magneto-ionic gating (Gomes et al., 2023)
• Magneto-Ionic Vortices: Voltage-Reconfigurable Swirling-Spin Analog-Memory Nanomagnets (Spasojevic et al., 2024)
• Carbon magneto-ionics: Control of magnetism through voltage-driven carbon transport (Tan et al., 14 Mar 2025)
• Charge-transfer-mediated boron magneto-ionics: Towards voltage-driven multi-ion transport (Ma et al., 14 Mar 2025)
• Tuning the dynamics of chiral domain walls of ferrimagnetic films with the magneto-ionic effect (Balan et al., 2022)