Field-Like Torque in Spintronics

Updated 19 January 2026

Field-like torque (FLT) is an effective transverse magnetic field generated by spin-polarized currents, crucial for controlling magnetization in nanostructures.
It is quantified through metrics such as torque efficiencies and conductivity, enabling precise modeling and engineering of spintronic device dynamics.
FLT plays a key role in frequency tuning, switching efficiency, and oscillator synchronization, making it essential for advanced magnetic memory and microwave applications.

Field-like torque (FLT) is a crucial component of spin-transfer and spin-orbit torque phenomena in magnetic nanostructures. It acts on the magnetization as an effective transverse magnetic field generated by spin-polarized current, distinct from the damping-like (Slonczewski) torque. FLT arises from both interfacial and bulk effects—originating in mechanisms such as transverse spin accumulation, Rashba–Edelstein coupling, spin Hall effect, and spin swapping—and plays a central role in frequency control, switching efficiency, synchronization, and emergent dynamics in spintronic devices, including magnetic tunnel junctions, spin Hall oscillators, and arrays of spin-torque nano-oscillators.

1. Theoretical Foundations and Mathematical Formulation

The magnetization dynamics in ferromagnetic nanostructures is commonly described using the Landau–Lifshitz–Gilbert (LLG) or Landau–Lifshitz–Gilbert–Slonczewski (LLGS) equations, augmented by spin-torque contributions. For a free-layer unit magnetization vector $\mathbf{m}$ subject to a spin polarization direction $\mathbf{p}$ (pinned layer or SOT source), the torque decomposition reads:

$\frac{d\mathbf{m}}{dt} = -\gamma\,\mathbf{m} \times \mathbf{H}_{\mathrm{eff}} + \alpha\,\mathbf{m} \times \frac{d\mathbf{m}}{dt} + a_J\,\mathbf{m} \times (\mathbf{m} \times \mathbf{p}) + b_J\,\mathbf{m} \times \mathbf{p}$

where:

$\gamma$ is the gyromagnetic ratio.
$\mathbf{H}_{\mathrm{eff}}$ is the effective magnetic field (sum of anisotropy, demagnetizing, exchange, and applied fields).
$\alpha$ is the Gilbert damping parameter.
$a_J$ is the damping-like (Slonczewski) torque amplitude, typically proportional to the spin current and polarization efficiency.
$b_J$ is the field-like torque amplitude; $b_J = \beta\,a_J$ defines $\beta$ as the dimensionless FLT-to-damping-like ratio.

In SOT systems, the torques are induced via spin Hall or Rashba–Edelstein effects, and the FLT appears as a term $m \times \sigma$ , where $\sigma$ is the injected spin polarization. For magnetic tunnel junctions and multilayer devices, FLT is generally of the form $m \times p$ , where $p$ is the reference layer magnetization.

The LLGS equation with both torques underpins analytic modeling and micromagnetic simulation across spintronic device classes (Lakshmanan et al., 2021, Krizakova et al., 2022, Taniguchi, 2021).

2. Physical Origins and Microscopic Mechanisms

FLT arises from diverse microscopic processes:

Spin mixing conductance: The imaginary part of the interfacial spin-mixing conductance $\operatorname{Im} G^{\uparrow\downarrow}$ generates an effective transverse field-like torque when a spin current impinges on the ferromagnet/normal-metal interface (Abert et al., 2016).
Spin Hall effect (SHE): In heavy-metal/ferromagnet bilayers, SHE-generated spin currents create both damping-like and field-like torques. The damping-like torque is linked to the real part of $G^{\uparrow\downarrow}$ ; the field-like component often arises from reflection, interfacial spin-orbit coupling, or Rashba–Edelstein effects (Ou et al., 2016, Peterson et al., 2020).
Spin swapping: Amorphous or disordered interfaces, as in sputtered WTe $_{2-x}$ , can enhance FLT via spin swapping processes, which convert longitudinal spin currents to transverse polarization through spin-orbit scattering (Yihong et al., 2020, Peterson et al., 2020).
Ferromagnet-originated Rashba effects: In certain systems, FLT primarily arises from the ferromagnetic layer, rather than the heavy-metal SOT source, with the sign and amplitude tunable via ferromagnet composition or thickness (Zhang et al., 13 Apr 2025).
Device geometry and symmetry: The relative alignment of current, magnetization, and device axes (e.g., type-X vs type-Y SOT geometry) determines the functional impact and efficiency of the FLT (Liu et al., 2021, Taniguchi, 2021).

Microscopically, FLT is associated with a torque $\mathbf{m} \times \sigma$ , reflecting the transverse response of the local magnetization to an incoming spin accumulation, regardless of whether generated in the bulk via SHE or at an interface via Rashba–Edelstein coupling (Dutta et al., 2021).

3. Quantitative Characteristics and Experimental Measurement

FLT is characterized using several experimentally accessible metrics:

Torque efficiencies ( $\xi_{\rm FL}$ , $\xi_{\rm DL}$ ): Defined via the ratio of effective field per unit current, typically in the range $|\xi_{\rm FL}| \sim 0.01$ –0.3 for various materials (Krizakova et al., 2022, Dutta et al., 2021, Peterson et al., 2020, Zhang et al., 13 Apr 2025).
Torque conductivity ( $\sigma^{\rm eff}_{\rm FL}$ ): Expressed in S/m, capturing the thickness and material dependence of FLT and allowing separation of interfacial and bulk contributions (Dutta et al., 2021). For ultrathin Ir/CoFeB, $\sigma^{\rm FL}_{\rm int} \simeq -5 \times 10^4~{\rm S/m}$ .
Sign and tunability: The sign of FLT (relative to DLT) affects device performance and is determined by the relative orientation of spin accumulation, layer composition, and microscopic process. Notably, FLT can be reversed (e.g., by engineering a Ni wedge in Pt/Ni/CoFeB) and modulated over an order of magnitude by minor adjustments in ferromagnet composition (Zhang et al., 13 Apr 2025).
Measurement methods: FLT is extracted by second-harmonic Hall measurements, spin-torque ferromagnetic resonance (ST-FMR), field-compensated planar Hall methods, real-time resistance tracking, and micromagnetic fitting against activation and switching thresholds (Dutta et al., 2021, Zhang et al., 13 Apr 2025, Krizakova et al., 2022).

Temperature and thickness scans reveal that interfacial FLT can dominate in ultrathin layers, while bulk (SHE-derived) FLT contributions are typically weaker and may change sign or magnitude depending on the ferromagnetic layer (Dutta et al., 2021, Ou et al., 2016, Peterson et al., 2020).

4. Dynamical Impact in Oscillators, Switching, and Synchronization

FLT directly modifies magnetization precession dynamics, critical current thresholds, switching probabilities, and oscillator synchronization properties:

Frequency tuning: FLT acts as an independent "frequency knob" in spin-transfer and spin-orbit torque nano-oscillators (STNOs, SOT NOs), shifting precession frequencies over GHz ranges without altering damping thresholds (Lakshmanan et al., 2021, Taniguchi et al., 2015, Arun et al., 2021, Arun et al., 2023). Analytical expressions and simulations confirm that frequency increases with increasing (appropriately signed) FLT.
Oscillation stability: In perpendicularly magnetized oscillators, FLT is essential for stable large-amplitude precession at zero field—without FLT, the Slonczewski torque overpowers damping and the system cannot sustain a limit cycle. Negative FLT (relative to the in-plane torque) stabilizes oscillations, optimizes cone angle, and enhances output power and $Q$ -factor (Taniguchi et al., 2014, Taniguchi et al., 2015, Arun et al., 12 Jan 2026).
Damping cancellation and limit cycles: Appropriate tuning of current and FLT can cancel effective damping, enabling undamped or weakly damped precessional states (Lakshmanan et al., 2021, Arun et al., 2019).
Synchronization: In arrays of coupled STNOs, FLT greatly facilitates phase locking across multiple elements. For small arrays, moderate FLT alone enables global synchrony; for large arrays, slight field angle tuning plus FLT is required. FLT also increases oscillation frequency and coherent output power, with power scaling as $N^2$ under full synchronization (Arun et al., 2023, Arun et al., 2019).
Switching efficiency and dynamics: FLT can either assist or hinder the switching of magnetization, depending on its sign, magnitude, and interplay with other torques (DLT, DMI). Positive FLT (assisting the hard-axis bias) lowers switching current and activation delays, while negative FLT may slow or block domain-wall propagation. The detailed switching phase space—assisted, adverse, or required FLT—depends on device parameters (DMI, current, anisotropy), as evidenced in both micromagnetic simulation and experiment (Krizakova et al., 2022, Wu et al., 2019, Yoon et al., 2017, Liu et al., 2021).
Suppression of “incubation delay”: In magnetic tunnel junctions, a sizable FLT explains the absence of pre-switching oscillations (incubation delay) by lowering the effective energy barrier and transforming the reversal pathway to a purely thermally activated process (0810.3421).

FLT magnitude, sign, and ratio to DLT are thus critical parameters for high-coherence oscillators, robust low-power switching, and scalable synchronized arrays (Arun et al., 2019, Taniguchi et al., 2014, Arun et al., 2023).

5. Material Dependence, Engineering, and Device Applications

FLT's efficacy and characteristics are highly sensitive to material choice, layer thickness, interfacial quality, and chemical engineering:

Interface engineering: Maximizing interfacial spin–orbit coupling, e.g., via heavy-metal choice (Ir, W, Ta), interface insertion layers, or oxidation, can tune FLT over a wide range (Dutta et al., 2021, Yihong et al., 2020).
Structural phase and disorder: Amorphous heavy metals and chalcogenides, such as WTe $_{2-x}$ , can enhance FLT via spin swapping; spin Hall–like origins also contribute depending on crystalline order (Yihong et al., 2020, Peterson et al., 2020).
Device geometry: Type-X SOT-MRAM geometries (easy axis collinear with current) benefit significantly from positive FLT, realizing sub-10 ns deterministic switching with lower current than canonical type-Y geometries. Optimization of aspect ratio, canting angle, and array layout can further exploit FLT for efficiency (Liu et al., 2021, Krizakova et al., 2022).
Application domains: FLT is leveraged for GHz-range and rapidly tunable microwave sources (oscillators), neuromorphic computing elements (frequency–phase encoding), high-density multi-level synapses (via spatially variable FLT), and ultra-fast in-plane SOT-MRAM (incubation-free sub-nanosecond switching) (Arun et al., 2023, Zhang et al., 13 Apr 2025, Taniguchi, 2021).

The engineering of FLT thus provides new avenues for low-power, high-speed magnetic memory and signal generation, with design levers including layer materials, stack sequence, bias fields, and device topology.

6. Controversies, Optimization Regimes, and Limits

FLT is neither universally beneficial nor universally adverse—its effects can be constructive or destructive depending on the signs, device configuration, and operational regime:

Beneficial regimes: Moderate, appropriately signed FLT can (i) extend operational windows for deterministic switching, (ii) boost oscillator frequency and output, (iii) maximize synchronization stability and array power, (iv) suppress switching delay, and (v) enable field-free or assistive switching (Wu et al., 2019, Arun et al., 2023, Taniguchi et al., 2015, 0810.3421).
Adverse regimes: Excessively large or misaligned FLT can block switching, destabilize oscillations, or lead to reduced frequency and premature switching failure. Notably, in SOT switching involving domain-wall-mediated reversal, an FLT of the “wrong” sign (opposite to the DLT) can increase the probability of back-switching due to domain-wall reflection at device edges (Wu et al., 2019, Yoon et al., 2017, Liu et al., 2021).
Material selection: Mapping out the regions in the parameter space—anisotropy, DMI, current density, FLT ratio—is key for device optimization, ensuring FLT is neither so small as to be ineffective nor so large as to induce instability (Wu et al., 2019, Lakshmanan et al., 2021).
Temperature effects: Interfacial FLT often exhibits strong temperature dependence, with contributions from spin-flip scattering at FM/oxide interfaces vanishing at low $T$ , while intrinsic (bulk SHE-derived) FLT remains relatively invariant (Ou et al., 2016, Peterson et al., 2020).

Optimal device performance requires precise control of FLT magnitude and sign, through both materials design and circuit layout, to harness its positive effects and counteract potentially destabilizing influences.

7. Summary Table: Key Effects of FLT Across Device Types

Device Type	Role of FLT	Optimal Regime / Observed Effects
STNO / SOT Oscillator	Frequency increase, Q-factor boost, damped-to-undamped transition	Negative FLT (per Slonczewski convention), modest ratio (β~0.1–0.3) for large-amplitude, stable precession (Taniguchi et al., 2015, Taniguchi et al., 2014, Arun et al., 12 Jan 2026)
MTJ/SOT Switching	Lowers critical current, assists/facilitates deterministic switching (if positive and collinear with assist field), may block or hinder for excessive or negative ratios	β~0.3–1.0, sign must align with desired field direction, mapped to device operation window (Krizakova et al., 2022, Liu et al., 2021, Wu et al., 2019)
Oscillator Arrays	Enables/sustains synchronization, boosts collective output power (∝N²), increases oscillation frequency	β~0.1–0.6, moderate field angle as needed for large arrays; excessive β may destabilize individual oscillators (Arun et al., 2023, Arun et al., 2019)
Incubation Delay (MTJ)	Suppresses pre-switching oscillations, enables field-activated switching pathway	Sizable b_J (quadratic in V), matches experimental incubation delay (0810.3421)