🔬 mesoscale physics

Two Exceptional-Point Regimes in the Spinterface Model of Chirality-Induced Spin Selectivity

This paper identifies two distinct, mutually exclusive near-exceptional-point regimes in the spinterface model of chirality-induced spin selectivity—termed "gold-level" and "chain-level"—which offer fit-free diagnostics for the underlying mechanism and establish CISS junctions as a solid-state platform for non-Hermitian spectral physics.

Original authors: Seif Alwan

Published 2026-10-28

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Seif Alwan

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the microscopic world of molecular electronics, scientists are trying to understand how tiny wires made of twisted, spiral-shaped molecules can act as one-way valves for electron spin. This phenomenon, known as chirality-induced spin selectivity, suggests that the handedness of a molecule—whether it twists left or right—can filter electrons based on their intrinsic magnetic orientation. For years, two main ideas have competed to explain this: one suggests the molecule itself creates a magnetic field that filters the electrons, while the other argues that the effect happens only at the interface where the molecule touches a metal electrode. To settle this, researchers need a way to distinguish between these mechanisms, looking for a unique fingerprint that only one of them could possibly leave behind.

A researcher at Uppsala University has now identified such a fingerprint by looking for a specific kind of singularity in the behavior of these molecular junctions. In physics, when a system becomes so complex that its internal rules break down and two distinct states merge into one, it reaches a point called an exceptional point. The researcher found that the model describing the interface between a gold electrode and a chiral molecule supports two very different types of these merging points. Crucially, they discovered that these two types cannot be mixed; they are mutually exclusive. One type happens at the gold electrode, driven by the metal's interaction with the molecule, while the other happens deep within the molecular chain itself, driven by the molecule's own internal structure. By proving that these two scenarios are algebraically distinct and cannot be blended, the researcher established a clear classification system that could help experimentalists determine exactly where and how spin selectivity is occurring in real devices.

The study reveals that the first type of merging point occurs right at the junction where the gold atom meets the molecule. Here, the two possible spin states of the gold atom blend together due to a specific kind of interaction mediated by the electrode. When the researcher tuned the strength of this interaction, they observed a dramatic shift in how electricity flowed. The system transitioned from a state where spin filtering was weak and symmetric to a state where the flow became purely dependent on the molecule's twist. Most strikingly, they found a specific setting where the ordinary, non-chiral part of the current vanished completely, leaving only the chiral signal. This "extinction point" acts as a definitive test: if an experiment can find a setting where the non-chiral signal disappears while the chiral one remains, it confirms that the effect is driven by the interface mechanism described in this model.

The second type of merging point is found further inside the molecular chain. Unlike the gold interface, which acts like a single, localized event, this phenomenon involves the entire chain of atoms working together. The researcher simulated chains of varying lengths and found that the strength of this effect does not simply grow forever as the chain gets longer. Instead, it follows a precise pattern: the effect grows rapidly as the chain lengthens, reaches a peak at a specific length, and then begins to fade. This peak occurs when the length of the chain matches a specific "coherence length," a measure of how far the quantum influence of the molecule's twist can travel before it gets scrambled. This finding suggests that the molecule's ability to filter spins is a collective effort of its atoms, but only up to a certain size. If the chain is too long, the atoms become too delocalized, and the special filtering effect breaks down.

By mapping out these two distinct regimes, the researcher provided a new way to diagnose the inner workings of molecular junctions without needing to guess the details of the materials involved. They showed that the distance between the theoretical prediction of where these points should occur and where they are actually observed in a simulation tells you exactly how spread out the electron states are. For the gold interface, the states are tightly packed, while for the molecular chain, they are spread out over many atoms. This difference in "dressing" or spreading serves as a built-in ruler for the mechanism. The work confirms that chirality-induced spin selectivity is not just a simple magnetic effect but a complex interplay of quantum states that can be tuned, measured, and classified. It offers a solid foundation for future experiments, suggesting that by measuring how the current changes with chain length or by looking for the specific point where non-chiral signals vanish, scientists can finally pin down the true origin of this mysterious spin-filtering effect.

Technical Summary: Two Exceptional-Point Regimes in the Spinterface Model of Chirality-Induced Spin Selectivity

Problem Statement
The paper addresses the theoretical characterization of Chirality-Induced Spin Selectivity (CISS) within the spinterface model, specifically investigating the behavior of the system when the effective non-Hermitian Hamiltonian approaches a singularity known as an Exceptional Point (EP). While CISS is a physically contested effect with competing structural and interfacial explanations, this work asks what occurs when the non-Hermitian structure responsible for reciprocity breaking becomes singular. The author aims to identify distinct EP regimes within the spinterface model, derive their algebraic conditions, and determine if these singularities offer unique, mechanism-specific signatures that can distinguish the spinterface mechanism from competing theories.

Methodology
The study utilizes the fully quantum formulation of the spinterface model within the Non-Equilibrium Green's Function (NEGF) framework.

Hamiltonian: The system consists of a gold atom (Au) with a local magnetic moment coupled to a chiral Rashba molecular chain of length $L$ , connected to non-magnetic leads. The Hamiltonian includes molecular hopping, gold-molecule coupling, lead coupling, and exchange/spin-orbit interactions.
Effective Non-Hermitian Hamiltonian: In the wide-band limit, the retarded self-energies from the leads render the effective Hamiltonian ( $H_{\text{eff}}$ ) non-Hermitian. The eigenvalues of $H_{\text{eff}}$ control transport observables.
EP Analysis: The author projects $H_{\text{eff}}$ onto two-dimensional subspaces to analyze EP conditions. They derive the discriminant for eigenvalue coalescence, distinguishing between Hermitian off-diagonal couplings (Scheme A) and anti-Hermitian off-diagonal couplings (Scheme B).
Diagnostics:
- Eigenpair Tracking: Specific eigenstates are tracked using "gold character" (for Au-level) or "Mol1 weight" (for chain-level) to ensure the correct pair is analyzed as parameters vary.
- Petermann Factor ( $K$ ): Calculated using biorthogonal left and right eigenvectors to detect the divergence associated with EPs. The paper notes that in "near-EPs" (dressed by the rest of the spectrum), $K$ reaches a finite peak rather than diverging to infinity.
- Transport Decomposition: Spin currents are decomposed into even (achiral) and odd (chiral) parts with respect to the chirality angle $\phi$ to analyze reciprocity breaking.

Key Contributions and Results

1. Two Algebraically Distinct EP Regimes
The paper establishes that the spinterface model supports exactly two distinct near-EP regimes, which cannot be mixed due to an algebraic obstruction:

Gold-Level Near-EP (Scheme B): Involves the coalescence of the two gold spin states. This regime requires spin-symmetric dissipative broadening and an anti-Hermitian, electrode-mediated spin flip. The EP condition is met when the anti-Hermitian coupling matches half the diagonal energy splitting ( $| \tilde{\alpha} | = |\Delta \tilde{\varepsilon}|/2$ ).
Chain-Level Near-EP (Scheme A): Involves the coalescence of the two molecular eigenstates with the largest weight on the first molecular site (Mol1). This regime requires spin-dependent dissipative broadening and a Hermitian, on-site Rashba spin flip. The EP condition is met when the Hermitian coupling matches a quarter of the broadening difference ( $| \tilde{\alpha} | = |\Delta \tilde{\gamma}|/4$ ).
Algebraic Obstruction: Mixing the diagonal cancellation of one scheme with the off-diagonal structure of the other yields a strictly negative discriminant, precluding the existence of an EP on any parameter path. This makes the gold/chain classification exhaustive.

2. Gold-Level Near-EP Signatures

Spectral Features: At the gold near-EP, the two gold-like eigenvalues approach a minimum separation (residual gap $\approx 0.19$ eV) due to dressing by the molecular chain. The Petermann factor peaks at $K \approx 6.7$ .
Chirality-Selective Transport: The anti-Hermitian spin-flip coupling is chirality-selective. As the coupling ( $g_{sf}$ $g_{s f}$ ) is tuned, the system undergoes three crossover points:
1. Onset: Near the EP ( $g_{sf} \approx 0.51$ eV), chirality begins to dominate.
2. First Parity Crossing: At $g_{sf} \approx 0.86$ eV, the chiral and achiral contributions are equal.
3. Complete Achiral Extinction: At $g_{sf} \approx 1.63$ eV, the achiral spin-valve signal vanishes entirely, leaving a purely chiral transport response. This extinction point is identified as a robust, fit-free diagnostic of the mechanism.

3. Chain-Level Near-EP and Collective Scaling

Collective Nature: The chain-level near-EP is a collective phenomenon where the coalescing pair is delocalized over the chain.
Dressing Factor: The observed EP location is shifted by a factor of $\approx 4.08$ relative to the bare prediction, directly measuring the delocalization of the coalescing pair (contrast with the gold level, where the shift is only $\approx 5.9\%$ ).
Petermann Scaling: The Petermann factor $K$ grows as a power law ( $K \sim L^\beta$ ) with chain length $L$ in the distinct-state regime. For reference parameters, $\beta \approx 2.43$ .
Bounded Growth: The growth is not monotonic; $K$ peaks at a chain length $L^* \approx 14$ (comparable to the Rashba coherence length $\ell_{coh} \approx 20$ ) and decreases for longer chains as the pair becomes too delocalized to support coherent coalescence.
Rashba Dependence: The exponent $\beta$ is non-monotonic with respect to the Rashba strength ( $\lambda_R$ ). It peaks at intermediate $\lambda_R$ (where coherent enhancement balances the breakdown of the Scheme-A condition) and becomes negative at high $\lambda_R$ , where the near-EP regime is destroyed.

Significance and Claims
The paper claims to identify CISS junctions as a solid-state platform for non-Hermitian spectral physics. The primary significance lies in providing fit-free diagnostics for the spinterface mechanism:

The ratio of observed to bare EP locations measures the delocalization of the coalescing pair.
The existence of an achiral extinction point in the gold regime serves as a direct test of the chirality-selective mechanism.
The collective Petermann scaling (power-law growth followed by a peak) and the non-monotonic dependence of the exponent on Rashba strength offer unique signatures of the collective coherence in the molecular chain.

The author emphasizes that these results are derived within the specific constraints of the spinterface model. They note that the chain-level analysis requires a weakened Rashba amplitude to be reachable within the current model parameters, framing it as a controlled test of the collective mechanism rather than a direct prediction for a specific physical molecule without further parameter tuning. The work suggests that observing these spectral and transport signatures (e.g., via shot noise measurements of the Fano factor) could experimentally distinguish the spinterface mechanism from competing structural explanations of CISS.

More like this