Damping-Like Torque Conductivity

• Damping-like torque conductivity is the measure of converting in-plane charge current into a damping-like spin-orbit torque in FM or FiM layers.
• It is extracted using techniques like ST-FMR and harmonic Hall measurements, linking spin Hall angles with charge conductivity and interfacial effects.
• Its material-dependent magnitude and sign are critical for engineering efficient spintronic devices, impacting switching speeds and power dissipation.

Damping-like torque conductivity quantifies the efficiency of converting an in-plane charge current into a damping-like spin-orbit torque (SOT) exerted on a ferromagnetic (FM) or ferrimagnetic (FiM) layer. The concept arises at the intersection of spin Hall and anomalous Hall effects, spin pumping, and interfacial symmetry-breaking mechanisms, and is central to optimizing the power dissipation and switching speeds in spintronic devices.

1. Definitions and Central Equations

Damping-like torque conductivity, denoted $\sigma_{\rm DL}$, is defined as the ratio of the damping-like spin current density (or torque per unit magnetization) to the applied electric field. In layer geometries, the canonical form is

$$\sigma_{\rm DL} = \frac{2e}{\hbar}\, \mu_0 M_s t_{\rm eff}\, \frac{H_{\rm DL}}{E}\,,$$

where $H_{\rm DL}$ is the effective damping-like field, $M_s$ is the FM/FiM saturation magnetization, $t_{\rm eff}$ is the magnetic thickness (either the spin-current source or detector layer), and $E$ is the applied electric field (Damas et al., 2022).

Alternatively, $\sigma_{\rm DL}$ can be expressed via the spin Hall angle $\theta_{\rm DL}$ and the layer charge conductivity $\sigma_{\rm charge}$: $$\sigma_{\rm DL} = \theta_{\rm DL} \sigma_{\rm charge}, \quad \sigma_{\rm charge} = \frac{1}{\rho_{\rm layer}},$$ linking the microscopic charge-spin conversion efficiency to experimentally accessible quantities, e.g., $\rho_{\rm GdFeCo}$ for GdFeCo ferrimagnet-based systems (Damas et al., 2022).

2. Measurement Protocols and Extraction Strategies

Typical extraction of $\sigma_{\rm DL}$ leverages modulation of the ferromagnetic resonance linewidth with a superimposed dc bias current in an ST-FMR experiment: $$\frac{\partial \Delta H}{\partial i_{\rm dc}} = -\left[ \frac{\gamma}{2H_{\rm res} + M_{\rm eff}} \right] \frac{1}{W t_{\rm GdFeCo}} \frac{\hbar}{2e \mu_0 M_s t_{\rm NiFe}} \sin \phi_H (\theta_{\rm DL}^{\rm SAHE} + \theta_{\rm DL}^{\rm SHE})$$ where known geometry ($W$, $t_{\rm GdFeCo}$, $t_{\rm NiFe}$), magnetization parameters, and resonance field $H_{\rm res}$ allow direct extraction of the spin Hall angles and thus the conductivity (Damas et al., 2022).

The approach is not limited to ferrimagnets; analogous expressions and analysis pipelines are used for Pt/Co, Ir/CoFeB, and alloy systems (e.g., PdPt), typically via either harmonic Hall measurements, loop-shift analysis, or vector network analyzer FMR, with necessary corrections for spin backflow, spin memory loss (SML), and shunt effects (Nguyen et al., 2015, Berger et al., 2017, Dutta et al., 2021, Zhu et al., 2019, Berger et al., 2016).

3. Microscopic Origins and Symmetry Considerations

Damping-like torque conductivity receives contributions from both bulk and interfacial mechanisms:

• Spin Hall Effect (SHE): Conversion of longitudinal charge current into transverse spin current within spin–orbit coupled layers; the dominant origin in heavy metals such as Pt, Ir, or PdPt (Berger et al., 2017, Zhu et al., 2019).
• Spin Anomalous Hall Effect (SAHE): Arises in ferrimagnets due to sublattice magnetizations and their interplay with SOC (Damas et al., 2022).
• Orbital Hall Effect (OHE): Injection of pure orbital current, converted into spin torque at interfaces with strong SOC, especially in textured Ru/Pt stacks (Das et al., 23 Jul 2025).
• Interfacial Berry curvature and Rashba–Edelstein effect: Important in vdW/FM systems (e.g., GeTe/Py), where charge transfer shifts Fermi level into high Berry-curvature bands, maximizing $\sigma_{\rm DL}$ (Bangar et al., 18 Jan 2026).

Symmetry governs the tensorial structure of $\sigma_{\rm DL}$. In tetragonal conductors (IrO$_2$), only three bulk tensor elements are independent, but arbitrary orientation/rotation predicts the anti-damping torque for any film cut by application of the bulk tensor and rotation matrix: $$\sigma^{\rm DL}_{ij} \rightarrow R \sigma^{\rm DL} R^T$$ where $R$ is the rotation matrix into the film axes (Patton et al., 2024).

4. Representative Magnitudes Across Materials

Damping-like torque conductivity is a material-dependent parameter and varies both with layer thickness and device architecture. Representative values (all in $\Omega^{-1}$m$^{-1}$):

Material System	$\sigma_{\rm DL}$	Notable Features
GdFeCo/Cu/NiFe	$-8.6\times 10^4$	Negative sign from SAHE dominance (Damas et al., 2022)
Ir/CoFeB, Ir/Co	$(1.1{-}1.4)\times10^5$	Bulk-driven, moderate efficiency (Dutta et al., 2021)
Pt/Co (thick)	$1.1 \times 10^6$	Benchmark heavy metal (Nguyen et al., 2015)
Pd$_{0.25}$Pt$_{0.75}$/FM	$1.05\times 10^6$	High efficiency, low resistivity (Zhu et al., 2019)
GeTe/Py (vdW/FM)	$-1.25\times 10^5$	Highest reported for FM/vdW (Bangar et al., 18 Jan 2026)
Ru/NiW/Co/Pt	$0.81\times 10^5$	Texture-boosted OHE/SHE (Das et al., 23 Jul 2025)
Py/Pt (bilayers)	$0.24\times 10^5$	Highly interface-dependent (Berger et al., 2016)

The sign convention—positive for Pt-like SHE, negative for GdFeCo—reflects fundamental differences in band structure, sublattice magnetization, and the relative weights of SAHE vs. SHE (Damas et al., 2022).

5. Thickness Dependence and Spin Diffusion Length

The evolution of $\sigma_{\rm DL}$ with spin-current-generating layer thickness captures the interplay between interfacial and bulk contributions. The typical scaling is

$\sigma_{\rm DL}(t) \sim \sigma^{\rm int}_{\rm DL} + \sigma^{\rm bulk}_{\rm DL} [1 - \sech(t/\lambda_{\rm sf})],$

where $\sigma^{\rm int}$ is the interfacial component (can be sizable in TMD/FM or in systems such as CoPt), $\sigma^{\rm bulk}$ is the bulk-generated spin Hall (or orbital Hall) current, and $\lambda_{\rm sf}$ is the spin diffusion length, extracted experimentally via fitting thickness dependencies or multidimensional parameter searches (Nguyen et al., 2015, Dutta et al., 2021, Berger et al., 2017, Zhu et al., 2019).

Spin diffusion lengths, e.g., $\lambda_{\rm sf}\simeq 1.6$ nm for Ir, $4.2$ nm for Pt (corrected for SML), underpin device scaling laws and optimal layer selection. In real devices, SML and interface transparency further reduce the effective torque delivered to the FM layer.

6. Magnetization, Angular Dependence, and Torkance Tensor

The torque symmetry is encoded in the angular dependence. Damping-like torque acts as a dissipative contribution (modifying effective Gilbert damping) and, formally, exhibits tensorial structure: $$\mathbf{T}_{\rm DL} = \tau_{\rm DL}\, \mathbf{m} \times (\mathbf{m} \times \hat{\sigma}),$$ where $\mathbf{m}$ is the unit magnetization vector, $\hat{\sigma}$ is the spin polarization axis, and $\tau_{\rm DL}$ the amplitude (Damas et al., 2022, Manchon et al., 2020).

For devices employing bulk crystalline conductors, the full spin torque conductivity tensor can be measured and rotated to predict torques in arbitrary orientations, as recently demonstrated for epitaxial IrO$_2$ (Patton et al., 2024). In tight-binding models, the torque per unit field is computed via Kubo linear response theory, explicitly revealing both interfacial and bulk contributions and substantial angular anisotropy (Manchon et al., 2020).

7. Materials, Interface Engineering, and Device Implications

The magnitude and sign of $\sigma_{\rm DL}$ serve as primary material-selection criteria for spin-based logic and memory technologies. Key approaches to optimize $\sigma_{\rm DL}$ include:

• Alloying and compositional tuning: E.g., Pd$_{x}$Pt$_{1-x}$ achieves high $\theta_{\rm SH}$ at low $\rho$, reducing Joule heating (Zhu et al., 2019).
• Crystal texture engineering: Using seed layers (NiW) to enforce hcp(002) texture in Ru significantly boosts orbital Hall contribution and, thereby, $\sigma_{\rm DL}$, with thermal stability (Das et al., 23 Jul 2025).
• Interfacial charge transfer: As shown in GeTe/Py, charge transfer shifts Fermi level to high Berry curvature regimes, amplifying both SHE and OHE (Bangar et al., 18 Jan 2026).
• Stacking order and interface transparency: Py/Pt vs. Pt/Py reveals a four-fold discrepancy, attributed to spin-mixing conductance and spin memory loss (Berger et al., 2016).
• Ferrimagnetic and chemically disordered FMs: Unconventional SAHE dominates in GdFeCo, with possibility of sign inversion, and self-torque is possible in CoPt single layers via internal spin Hall current (Damas et al., 2022, Zhu et al., 2020).

Device-level manifestations include efficient electrical switching of PMA elements for MRAM, tunable Dzyaloshinskii-Moriya interaction for skyrmion manipulation, and reduced critical currents for auto-oscillation and field-free magnetic reversal.

---

Damping-like torque conductivity is an intensive, multifaceted figure of merit, reflecting underlying materials physics, microscopic conversion mechanisms, and interface engineering. Advances in its measurement, enhancement, and understanding directly translate to next-generation applications in energy-efficient, high-performance spin-based information processing (Damas et al., 2022, Dutta et al., 2021, Berger et al., 2017, Nguyen et al., 2015, Zhu et al., 2019, Das et al., 23 Jul 2025, Bangar et al., 18 Jan 2026, Patton et al., 2024, Berger et al., 2016, Manchon et al., 2020, Zhu et al., 2020).