Unconventional Rashba-Edelstein Effect

• The unconventional Rashba-Edelstein effect is defined by deviations from the standard transverse current-induced spin polarization, incorporating collinear, orbital, and valley responses.
• Key studies reveal that modified band topology and disorder mechanisms can boost inverse spin-charge conversion efficiency by constructive Fermi contour contributions.
• Nonlinear, ultrafast, and electrodynamic regimes further diversify the effect, opening avenues for applications in spintronics and superconducting phase control.

The unconventional Rashba-Edelstein effect denotes a family of charge-to-spin, spin-to-charge, spin-to-orbital, and even electrodynamic conversion phenomena that depart from the textbook Rashba-Edelstein geometry in which an in-plane current or electric field generates a nonequilibrium spin density transverse to the drive. In the conventional interfacial form, Rashba spin-orbit coupling (SOC) locks momentum to an in-plane spin texture, so a shifted distribution produces a transverse spin accumulation. The literature uses “unconventional” for departures in response geometry, band topology, dominant angular-momentum channel, temporal regime, disorder mechanism, or readout modality, including collinear and valley-parallel spin polarization, orbitally dominated responses, nonlinear and ultrafast dynamics, disorder-generated kinetic variants, and superconducting phase-sensitive manifestations (Senapati et al., 2023, Tenzin et al., 2022, Taguchi et al., 2017, Leiva-Montecinos et al., 27 May 2025, Vignale et al., 2015, Shumilin et al., 2023).

1. Canonical Rashba-Edelstein response and the baseline for “unconventionality”

The standard Rashba-Edelstein effect is a current-induced nonequilibrium spin density in a noncentrosymmetric conductor or interface. At a Rashba interface, the Hamiltonian may be written as

$H_R = \frac{\alpha_R}{\hbar}(E_z \times p)\cdot S,$

and the induced spin density obeys

$S = \frac{\alpha_R m}{e\hbar}(E_z \times j_c),$

so the spin polarization is orthogonal to both the interfacial electric field and the charge current (Senapati et al., 2023). In equivalent interfacial notation, the Rashba Hamiltonian is

$$H_\mathrm{R}=\alpha_\mathrm{R}(k\times \hat{e}_\mathrm{z}) \cdot \sigma,$$

which produces the familiar chiral in-plane spin texture and underlies direct charge-to-spin conversion and inverse spin-to-charge conversion (Jungfleisch et al., 2015).

A tensor formulation makes explicit that the effect is a linear-response conversion law rather than a fixed geometric cartoon: $$\delta \mathbf{s}=\chi\,\mathbf{j}^A, \qquad \delta s_j=\chi_{ij} j_i.$$ In the conventional Rashba case, the dominant tensor elements are off-diagonal and antisymmetric, so the induced spin is transverse to the current. This transverse geometry is only one symmetry-allowed realization. The modern literature therefore treats the Rashba-Edelstein effect as a tensorial current-induced polarization problem whose form is set by Brillouin-zone averages of spin and velocity, not by a single local spin texture alone (Tenzin et al., 2022).

Against this baseline, an unconventional Rashba-Edelstein effect is any response that does not reduce to a small, steady-state, spin-only, transverse interfacial polarization. The main deviations fall into several categories: altered response geometry, unconventional Fermi-surface chirality, orbital dominance, nonlinear or ultrafast driving, disorder-enabled transport channels, and nonstandard detection through magnetic resonance, Kerr microscopy, Josephson interference, or bulk optical electrodynamics (Taguchi et al., 2017, Song et al., 2021, Salemi et al., 2019, Vignale et al., 2015, Shibata et al., 2015).

2. Response geometry beyond the transverse Rashba limit

One major unconventional class changes the direction of the induced polarization itself. In gated monolayer transition-metal dichalcogenides, the predicted valley Edelstein effect (VEE) adds a valley-odd spin density parallel to the applied field: $$\langle \bm{s}^{\textrm{VEE}}_{\textrm{v}} \rangle = - e \nu_e \mathcal{C}_\perp(\hat{\bm z}\times \bm E) - \mathrm{v} e \nu_e \mathcal{C}_\parallel \bm E.$$ The first term is the conventional Edelstein contribution, whereas the second is the unconventional one: it is parallel to $\bm E$ and reverses sign between $\mathrm{v}=\pm K$. Its microscopic origin is the coexistence of Rashba SOC and Ising SOC in gated monolayer transition-metal dichalcogenides, which produces a valley-odd spin Berry curvature rather than the orbital Berry curvature of pristine transition-metal dichalcogenides (Taguchi et al., 2017). The effective model,

$$\mathcal{H}_{0,\textrm{v}} = \sum_{\bm{k}} \psi^\dagger_{\textrm{v},\bm{k}} \Bigl[ \varepsilon_{k}\sigma^0 + \mathrm{v}\beta_{\textrm{I}}\sigma^z + \alpha_{\textrm{R}}(k_y\sigma^x-k_x\sigma^y) \Bigr] \psi_{\textrm{v},\bm{k}},$$

shows that the Ising term acts as a valley-dependent Dirac mass, while the Rashba term supplies in-plane spin-momentum locking. Their combination generates the spin-space Berry curvature that converts anomalous transverse drift into a spin density parallel to the driving field (Taguchi et al., 2017).

A broader symmetry analysis shows that the transverse Rashba geometry is not generic. A classification over all 230 nonmagnetic space groups identified 127 space groups in which Rashba-Edelstein response is symmetry-allowed and showed that diagonal tensor elements can produce a collinear Rashba-Edelstein effect, with spin accumulation parallel to current (Tenzin et al., 2022). In this analysis, GeTe and monolayer In$_2$Se$_3$ realize the conventional antisymmetric form $S = \frac{\alpha_R m}{e\hbar}(E_z \times j_c),$0, while tellurium and Pd$S = \frac{\alpha_R m}{e\hbar}(E_z \times j_c),$1Se realize diagonal responses such as $S = \frac{\alpha_R m}{e\hbar}(E_z \times j_c),$2 or $S = \frac{\alpha_R m}{e\hbar}(E_z \times j_c),$3, establishing that collinear conversion is allowed in both chiral and non-chiral crystals (Tenzin et al., 2022).

A central correction to a common oversimplification follows directly from this symmetry work: a Rashba-like local spin texture does not by itself determine a transverse Rashba-Edelstein response. The full Brillouin-zone average, the space-group constraints on $S = \frac{\alpha_R m}{e\hbar}(E_z \times j_c),$4, and the coexistence of Rashba-like, Dresselhaus-like, or Weyl-like textures at different $S = \frac{\alpha_R m}{e\hbar}(E_z \times j_c),$5-points together determine whether the induced spin is transverse, collinear, or oblique (Tenzin et al., 2022).

3. Unconventional Rashba bands and giant inverse or spin-galvanic conversion

A second major unconventional class originates in band topology rather than tensor symmetry. In the conventional Rashba picture, the two spin-split Fermi circles usually contribute with opposite chirality or with opposing velocity structure, which causes partial compensation in inverse Edelstein conversion. Several works instead consider “unconventional Rashba bands” for which the relevant Fermi circles have the same in-plane spin chirality and, in the key regime, the same group-velocity direction (Song et al., 2021, Huang et al., 2024).

In a model of two hybridized Rashba doublets, the Hamiltonian

$S = \frac{\alpha_R m}{e\hbar}(E_z \times j_c),$6

reorganizes the dispersions so that, in a specific energy region, two Fermi circles possess the same helical spin texture and the same velocity sense. Under weak spin injection, both contours then contribute with the same sign to the inverse Rashba-Edelstein effect rather than canceling. The conversion efficiency

$S = \frac{\alpha_R m}{e\hbar}(E_z \times j_c),$7

is thereby strongly enhanced, and for the unconventional regime the reported ratio

$S = \frac{\alpha_R m}{e\hbar}(E_z \times j_c),$8

amounts to about an order-of-magnitude improvement over the conventional case (Song et al., 2021). First-principles calculations identify monolayer OsBi$S = \frac{\alpha_R m}{e\hbar}(E_z \times j_c),$9 as a candidate realization, with the relevant energy windows tunable by chemical doping, electrostatic gating, or ionic liquid gating (Song et al., 2021).

A generic four-band model on hexagonal and square lattices reaches a similar conclusion by a different route. There the two split circles again share the same global chirality, and the inverse Edelstein, or spin galvanic, efficiency is enhanced relative to conventional Rashba bands in both potential-impurity and magnetic-impurity settings (Huang et al., 2024). An especially notable distinction is disorder sensitivity: for conventional Rashba bands the efficiency is reported to be insensitive to whether scattering is potential or magnetic, whereas for unconventional bands magnetic impurity scattering further enhances the conversion beyond the potential-scattering case (Huang et al., 2024).

These results shift attention from Rashba strength alone to the organization of spin texture and velocity on the Fermi surface. The enhancement mechanism is not simply “stronger SOC”; it is the constructive addition of multiple Fermi-contour contributions that conventionally compensate (Song et al., 2021, Huang et al., 2024).

4. Orbital, valley, ferroelectric, and antiferromagnetic generalizations

Another fundamental generalization is that the Edelstein response need not be predominantly spin. In noncentrosymmetric antiferromagnets, a generalized Rashba-Edelstein effect generates both spin and orbital polarization, with the induced local magnetization written as

$$H_\mathrm{R}=\alpha_\mathrm{R}(k\times \hat{e}_\mathrm{z}) \cdot \sigma,$$0

First-principles calculations for CuMnAs and Mn$$H_\mathrm{R}=\alpha_\mathrm{R}(k\times \hat{e}_\mathrm{z}) \cdot \sigma,$$1Au show that the orbital Rashba-Edelstein susceptibility can dominate the spin susceptibility and can survive even without SOC (Salemi et al., 2019). In CuMnAs, the orbital staggered response at $$H_\mathrm{R}=\alpha_\mathrm{R}(k\times \hat{e}_\mathrm{z}) \cdot \sigma,$$2 is reported to be about 60 times larger than the spin response for one magnetic configuration, while in Mn$$H_\mathrm{R}=\alpha_\mathrm{R}(k\times \hat{e}_\mathrm{z}) \cdot \sigma,$$3Au the orbital off-diagonal response is about 3 times larger than the spin counterpart (Salemi et al., 2019). The response can be Rashba-like or Dresselhaus-like depending on magnetic configuration, and it includes both staggered in-plane and non-staggered out-of-plane components (Salemi et al., 2019).

Bulk ferroelectric GeTe provides a related but distinct example. There the induced magnetic moment per unit cell is

$$H_\mathrm{R}=\alpha_\mathrm{R}(k\times \hat{e}_\mathrm{z}) \cdot \sigma,$$4

and Wannier-based tight-binding plus semiclassical Boltzmann calculations show that the orbital Edelstein susceptibility is about one order of magnitude larger than the spin susceptibility in the relevant energy window below the valence-band maximum (Leiva-Montecinos et al., 27 May 2025). The orbital moment on the Fermi surface surpasses the spin moment by one order of magnitude, and the orbital Edelstein effect remains largely unaffected in the absence of SOC, unlike the spin contribution (Leiva-Montecinos et al., 27 May 2025). Because GeTe is a ferroelectric Rashba semiconductor, switching the ferroelectric polarization reverses the Rashba texture and therefore the sign of the current-induced spin and orbital response (Leiva-Montecinos et al., 27 May 2025).

The valley Edelstein effect belongs to the same broad expansion of the Edelstein paradigm. It adds a valley-contrasting, Berry-curvature-driven component to the conventional transverse response and thereby converts Rashba-Edelstein physics into a valley-spin effect rather than a purely uniform spin accumulation (Taguchi et al., 2017).

These studies collectively overturn a second widespread simplification: the Edelstein effect is not inherently a spin-only interface phenomenon. It can be bulk, ferroelectrically switchable, staggered over antiferromagnetic sublattices, valley-contrasting, and orbitally dominated (Salemi et al., 2019, Leiva-Montecinos et al., 27 May 2025, Taguchi et al., 2017).

5. Nonlinear, ultrafast, and electrodynamic regimes

The term “unconventional” also marks a breakdown of linear-response transport. In a perfectly clean Rashba two-dimensional electron gas driven strongly enough that the drift velocity becomes comparable to the Fermi velocity, the spin dynamics becomes nonlinear and is governed by Landau-Zener-type evolution rather than a static proportionality to the field (Vignale et al., 2015). The control parameter is

$$H_\mathrm{R}=\alpha_\mathrm{R}(k\times \hat{e}_\mathrm{z}) \cdot \sigma,$$5

For $$H_\mathrm{R}=\alpha_\mathrm{R}(k\times \hat{e}_\mathrm{z}) \cdot \sigma,$$6, the spin follows the effective field adiabatically and the polarization grows monotonically before saturating. For $$H_\mathrm{R}=\alpha_\mathrm{R}(k\times \hat{e}_\mathrm{z}) \cdot \sigma,$$7, the evolution is strongly nonadiabatic and the polarization is progressively reduced and eventually suppressed (Vignale et al., 2015). The Landau-Zener survival probability,

$$H_\mathrm{R}=\alpha_\mathrm{R}(k\times \hat{e}_\mathrm{z}) \cdot \sigma,$$8

captures the momentum-angle dependence of this crossover (Vignale et al., 2015). The same work predicts an inverse nonlinear Edelstein effect driven by a linearly increasing magnetic field and formulates a generalized Onsager reciprocity between the direct and inverse nonlinear conductivities (Vignale et al., 2015).

On femtosecond timescales, the effect becomes intrinsically time dependent and optical rather than dc. A real-space tight-binding simulation for Au(001) driven by a 10 fs laser pulse with $$H_\mathrm{R}=\alpha_\mathrm{R}(k\times \hat{e}_\mathrm{z}) \cdot \sigma,$$9 eV shows nonlinear spin and orbital Edelstein effects whose envelopes differ from the Lorentzian laser pulse, peak after the laser maximum around $$\delta \mathbf{s}=\chi\,\mathbf{j}^A, \qquad \delta s_j=\chi_{ij} j_i.$$0 fs, and exhibit post-pulse beating (Busch et al., 4 May 2025). The same calculations identify accompanying longitudinal spin and orbital currents $$\delta \mathbf{s}=\chi\,\mathbf{j}^A, \qquad \delta s_j=\chi_{ij} j_i.$$1 and $$\delta \mathbf{s}=\chi\,\mathbf{j}^A, \qquad \delta s_j=\chi_{ij} j_i.$$2, as well as transverse laser-induced spin Hall and orbital Hall currents (Busch et al., 4 May 2025). In this regime, the effect is neither steady-state nor perturbatively linear in field amplitude.

Bulk Rashba conductors reveal yet another unconventional regime in which direct and inverse Edelstein effects combine to renormalize the electromagnetic response itself. The long-wavelength conductivity,

$$\delta \mathbf{s}=\chi\,\mathbf{j}^A, \qquad \delta s_j=\chi_{ij} j_i.$$3

implies a reduced transverse plasma frequency

$$\delta \mathbf{s}=\chi\,\mathbf{j}^A, \qquad \delta s_j=\chi_{ij} j_i.$$4

For BiTeI, the reported $$\delta \mathbf{s}=\chi\,\mathbf{j}^A, \qquad \delta s_j=\chi_{ij} j_i.$$5 below the upper interband edge gives $$\delta \mathbf{s}=\chi\,\mathbf{j}^A, \qquad \delta s_j=\chi_{ij} j_i.$$6, which produces a hyperbolic electromagnetic medium in the interval $$\delta \mathbf{s}=\chi\,\mathbf{j}^A, \qquad \delta s_j=\chi_{ij} j_i.$$7 (Shibata et al., 2015). With magnetization, the same framework predicts directional dichroism enhanced near interband-transition edges and interprets the optical nonreciprocity in terms of toroidal and quadrupolar moments (Shibata et al., 2015). Here the Rashba-Edelstein mechanism is unconventional because it feeds back into the dielectric tensor rather than appearing only as a local spin density.

Closely related nonlinear transport has also been identified in unconventional Rashba bands through an electric-field-induced second-order anomalous Hall effect controlled by the quantum metric and Berry connection polarizability rather than by ordinary linear Edelstein polarization (Bhattacharya et al., 2024). This is not the standard Rashba-Edelstein effect proper, but it extends the same inversion-broken Rashba framework into a geometric nonlinear Hall response.

6. Interfaces, disorder, and experimental manifestations

Experimental and device work has emphasized that unconventionality may lie not only in the microscopic mechanism but also in the readout channel. At Bi/Ag interfaces, spin-torque ferromagnetic resonance shows that a direct Rashba-Edelstein spin accumulation can drive a neighboring Py layer into resonance. In Py/Ag/Bi trilayers, the oscillatory interface-generated spin accumulation provides a damping-like torque, while the Oersted field supplies the field-like torque, allowing the standard ST-FMR decomposition into symmetric and antisymmetric Lorentzian components (Jungfleisch et al., 2015). A dc-current-induced linewidth modulation of about $$\delta \mathbf{s}=\chi\,\mathbf{j}^A, \qquad \delta s_j=\chi_{ij} j_i.$$8 and an effective spin-Hall-angle-like estimate of about 18% were reported, establishing an interface-only spin-orbit torque source rather than a bulk spin Hall mechanism (Jungfleisch et al., 2015).

Direct optical detection has been demonstrated at nonmagnetic metal/oxide Rashba interfaces. In Cu/Bi$$\delta \mathbf{s}=\chi\,\mathbf{j}^A, \qquad \delta s_j=\chi_{ij} j_i.$$9O$$\langle \bm{s}^{\textrm{VEE}}_{\textrm{v}} \rangle = - e \nu_e \mathcal{C}_\perp(\hat{\bm z}\times \bm E) - \mathrm{v} e \nu_e \mathcal{C}_\parallel \bm E.$$0 and Ag/Bi$$\langle \bm{s}^{\textrm{VEE}}_{\textrm{v}} \rangle = - e \nu_e \mathcal{C}_\perp(\hat{\bm z}\times \bm E) - \mathrm{v} e \nu_e \mathcal{C}_\parallel \bm E.$$1O$$\langle \bm{s}^{\textrm{VEE}}_{\textrm{v}} \rangle = - e \nu_e \mathcal{C}_\perp(\hat{\bm z}\times \bm E) - \mathrm{v} e \nu_e \mathcal{C}_\parallel \bm E.$$2, grazing-incidence structural characterization and time-resolved transverse magneto-optical Kerr effect were used to detect a uniform in-plane spin accumulation oscillating with a 100 MHz electrical drive (Auvray et al., 2018). The Kerr phase reverses between Cu/Bi$$\langle \bm{s}^{\textrm{VEE}}_{\textrm{v}} \rangle = - e \nu_e \mathcal{C}_\perp(\hat{\bm z}\times \bm E) - \mathrm{v} e \nu_e \mathcal{C}_\parallel \bm E.$$3O$$\langle \bm{s}^{\textrm{VEE}}_{\textrm{v}} \rangle = - e \nu_e \mathcal{C}_\perp(\hat{\bm z}\times \bm E) - \mathrm{v} e \nu_e \mathcal{C}_\parallel \bm E.$$4 and Ag/Bi$$\langle \bm{s}^{\textrm{VEE}}_{\textrm{v}} \rangle = - e \nu_e \mathcal{C}_\perp(\hat{\bm z}\times \bm E) - \mathrm{v} e \nu_e \mathcal{C}_\parallel \bm E.$$5O$$\langle \bm{s}^{\textrm{VEE}}_{\textrm{v}} \rangle = - e \nu_e \mathcal{C}_\perp(\hat{\bm z}\times \bm E) - \mathrm{v} e \nu_e \mathcal{C}_\parallel \bm E.$$6, consistent with Rashba parameters

$$\langle \bm{s}^{\textrm{VEE}}_{\textrm{v}} \rangle = - e \nu_e \mathcal{C}_\perp(\hat{\bm z}\times \bm E) - \mathrm{v} e \nu_e \mathcal{C}_\parallel \bm E.$$7

and the accumulation is reported to be spatially uniform across different probe spots, unlike the sign-changing thickness profile expected from a bulk spin Hall signal (Auvray et al., 2018).

A more indirect but distinctive readout uses superconducting phase sensitivity. In Nb-(Pt/Cu)-Nb planar Josephson junctions, a quasiparticle current at the Pt/Cu Rashba interface generates a nonequilibrium interfacial spin moment that shifts the center of the Fraunhofer pattern to a nonzero field value (Senapati et al., 2023). For a representative shift of about 10 mT and $$\langle \bm{s}^{\textrm{VEE}}_{\textrm{v}} \rangle = - e \nu_e \mathcal{C}_\perp(\hat{\bm z}\times \bm E) - \mathrm{v} e \nu_e \mathcal{C}_\parallel \bm E.$$8, the inferred interfacial quasiparticle current is of order $$\langle \bm{s}^{\textrm{VEE}}_{\textrm{v}} \rangle = - e \nu_e \mathcal{C}_\perp(\hat{\bm z}\times \bm E) - \mathrm{v} e \nu_e \mathcal{C}_\parallel \bm E.$$9 (Senapati et al., 2023). The effect is interpreted not as ferromagnetic hysteresis but as screening-current locking to the Rashba-Edelstein-induced spin moment, suggesting a $\bm E$0-like phase bias and a route to phase batteries in superconducting circuits (Senapati et al., 2023).

Disorder further broadens the phenomenology. At a disordered metal-insulator boundary, skew scattering and interference between multiple reflection paths generate a kinetic Rashba-Edelstein effect in which the spin accumulation extends over a length scale of order the mean free path $\bm E$1, rather than only a few monolayers as at a clean interface (Shumilin et al., 2023). In the same framework, higher-order spin-orbit effects produce an interface spin-Hall effect that persists already within the Born approximation, unlike the conventional bulk spin-Hall effect (Shumilin et al., 2023). This reverses the usual assumption that disorder only suppresses Rashba-Edelstein physics: in intermediate-thickness devices with $\bm E$2, interface disorder can become the dominant source of spin density and spin current (Shumilin et al., 2023).

Taken together, these experiments and models show that unconventional Rashba-Edelstein physics is not a single anomaly but a broad redefinition of charge-spin conversion in inversion-broken systems. The induced polarization need not be transverse, spin-only, interface-localized, linearly driven, or read out through standard spin-charge conversion. It may instead be valley-odd and parallel to the field, collinear by symmetry, constructively enhanced by unconventional Rashba bands, dominated by orbital angular momentum, dynamically reshaped by strong driving or femtosecond coherence, extended by disorder, or transduced into optical, magnetic-resonance, Josephson-phase, or bulk-electrodynamic observables (Taguchi et al., 2017, Tenzin et al., 2022, Song et al., 2021, Salemi et al., 2019, Busch et al., 4 May 2025, Shumilin et al., 2023).