Tunnel magnetoresistance in magnetic tunnel junctions with embedded nanoparticles
In this paper, we attempt the theoretical modeling of the magnetic tunnel junctions with embedded magnetic and nonmagnetic nanoparticles (NPs). A few abnormal tunnel magnetoresistance (TMR) effects, observed in related experiments, can be easily simulated within our model: we found, that the suppressed TMR magnitudes and the TMR sign-reversing effect at small voltages are related to the electron momentum states of the NP located inside the insulating layer. All these TMR behaviors can be explained within the tunneling model, where NP is simulated as a quantum well (QW). The coherent (direct) double barrier tunneling is dominating over the single barrier one. The origin of the TMR suppression is the quantized angle transparency for spin polarized electrons being in one of the lowest QW states. The phenomenon was classified as the quantized conductance regime due to restricted geometry.
💡 Research Summary
The paper presents a comprehensive theoretical investigation of magnetic tunnel junctions (MTJs) that incorporate magnetic or non‑magnetic nanoparticles (NPs) within the insulating barrier. The authors aim to explain several anomalous tunnel magnetoresistance (TMR) phenomena reported in recent experiments, namely a pronounced suppression of TMR at low bias and a reversal of the TMR sign near zero voltage. To achieve this, they construct a model in which each NP is treated as a quantum well (QW) sandwiched between two ultra‑thin insulating layers, thereby forming a double‑barrier tunneling system. The key premise is that electron transport proceeds predominantly via coherent (direct) double‑barrier tunneling rather than through a single‑barrier pathway.
In the formalism, spin‑dependent wave vectors are defined for the ferromagnetic electrodes and for the NP region. The NP’s QW imposes quantization on the component of the electron momentum perpendicular to the interfaces (k⊥), yielding discrete energy levels. The transmission probability for an electron incident at angle θ is expressed in the usual form T(θ)=4k₁k₂/(k₁+k₂)², but the values of k₁ and k₂ are now constrained by the QW boundary conditions. When the electron’s longitudinal energy aligns with a QW level, resonant tunneling occurs, dramatically enhancing conductance for that spin channel. Conversely, if the electron occupies the lowest QW state, the allowed incident angles become highly restricted—a phenomenon the authors term “quantized angle transparency.” This restriction reduces the number of available conduction channels and leads to a strong suppression of the overall TMR.
Numerical simulations explore a wide parameter space: NP radii ranging from 0.5 nm to 3 nm, barrier thicknesses of 0.8–1.2 nm, and both magnetic and non‑magnetic NP materials. For non‑magnetic NPs, the TMR curve remains relatively symmetric with respect to bias, but a distinct dip appears when the NP size matches a condition that forces the lowest QW state to dominate transport. In magnetic NPs, the spin‑up and spin‑down QW levels are split, so that at low bias only one spin channel may satisfy the resonance condition. This asymmetry produces a negative TMR (i.e., the antiparallel magnetic configuration conducts better than the parallel one) at voltages below roughly ±10 mV. As the bias increases, higher QW levels become accessible, restoring the conventional positive TMR.
The authors interpret these results as evidence of a “quantized conductance regime” induced by the confined geometry of the double‑barrier system. The quantization of incident angles and the discrete nature of the QW states together create a situation where conductance changes in step‑like fashion as the bias sweeps through resonant conditions. This framework successfully reproduces the experimentally observed TMR suppression and sign reversal without invoking extrinsic mechanisms such as spin‑flip scattering or magnon excitations.
In the discussion, the paper highlights several practical implications. First, the ability to tailor TMR by adjusting NP size, material, and barrier thickness offers a new degree of freedom for spintronic device engineering. Second, the low‑bias TMR sign reversal could be exploited for novel logic functionalities in magnetic random‑access memory (MRAM) where the resistance state can be switched not only by magnetic orientation but also by bias polarity. Finally, the model underscores the importance of quantum confinement effects in nanoscale MTJs, suggesting that future experimental designs should carefully consider the interplay between QW resonances and spin polarization.
In conclusion, the study provides a robust quantum‑well‑based description of double‑barrier tunneling in NP‑embedded MTJs, elucidating the origin of anomalous TMR behavior and opening pathways for the deliberate manipulation of spin‑dependent transport at the nanometer scale.