🔬 mesoscale physics

Electrically switchable one-dimensional quadrupolar excitons in lateral double heterojunctions

This paper theoretically demonstrates that a lateral WS2-MoS2-WS2 double heterojunction supports electrically switchable one-dimensional excitons, where an external electric field continuously transforms a zero-dipole, quadrupolar ground state into a dipolar state, enabling geometric control over the system's Stark sensitivity.

Original authors: Ryo Kitaura

Published 2026-09-29

📖 7 min read🧠 Deep dive

CC BY 4.0

Original authors: Ryo Kitaura

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 world of modern electronics, the smallest building blocks are often flat, one-atom-thick sheets of material. Scientists have learned to stitch these sheets together side-by-side, creating sharp boundaries where one type of semiconductor meets another. At these boundaries, electrons and the empty spaces they leave behind, called holes, can get stuck together in a pair known as an exciton. Usually, these pairs form right at the edge, with the electron on one side and the hole on the other, creating a tiny electrical separation. This separation acts like a miniature magnet for electricity, known as a dipole, which responds strongly to external electric fields. For years, researchers have been able to control these single-boundary excitons, but a new question has emerged: what happens if you place two of these boundaries very close together, sandwiching a narrow strip of material between them?

A researcher has now simulated exactly this scenario, creating a theoretical model of a narrow strip of molybdenum disulfide sandwiched between two wider sheets of tungsten disulfide. By running detailed computer calculations, they discovered that the two boundaries do not act independently. Instead, the exciton can exist in a state where it is simultaneously shared between both boundaries. In this shared state, the electrical forces from the left and right sides cancel each other out, leaving the pair with no net electrical pull in any direction. However, this balanced state is not static. When the researcher applied an electric field across the strip, they found they could smoothly and continuously push the exciton from this balanced, shared state into a state where it sits firmly on just one side. This transition allows the exciton to change its electrical character from having no pull to having a strong, directional pull, all controlled by the strength of the electric field.

The researcher focused on a strip of material just 1.5 nanometers wide, a size that has already been observed in real laboratory experiments. In their simulations, they found that at zero electric field, the exciton exists as a stable pair with a binding energy of about 104 milli-electron-volts, meaning the electron and hole are held together very tightly. Because the strip is so narrow, the exciton feels the influence of both boundaries at once. This creates two distinct energy levels, separated by a tiny gap of 4.6 milli-electron-volts. One level is slightly lower in energy, and the other is slightly higher. Both of these levels are special because they have no permanent electrical pull, but they do possess a more complex electrical shape called a quadrupole, which arises from the specific way the charges are arranged across the strip.

When an electric field is applied perpendicular to the strip, the situation changes dramatically. The field acts as a switch, gradually tilting the energy landscape. As the field increases, the lower energy state shifts from being a balanced mix of both boundaries to becoming a state where the exciton is almost entirely localized on one side. This transition is smooth and continuous. At a specific field strength of 0.87 volts per micrometer, the exciton is halfway between the balanced state and the localized state. Beyond this point, the exciton behaves like a standard dipole, with a strong electrical pull that grows linearly with the field. The researcher calculated that the exciton can develop an electrical pull as large as 126 Debye, a unit of measurement for electrical separation, which is significantly larger than what is typically seen in similar vertical structures.

The width of the strip plays a crucial role in this behavior. The researcher found that changing the width of the strip by just a fraction of a nanometer dramatically alters how easily the electric field can switch the exciton's state. A wider strip makes the two boundaries less connected, requiring a much weaker electric field to push the exciton to one side. Conversely, a narrower strip keeps the boundaries strongly connected, requiring a stronger field to make the switch. This means that by simply designing the width of the strip, engineers could tune the sensitivity of the device to electric fields without changing the material itself. The simulations also showed that this switching mechanism works for excited states of the exciton as well, suggesting that the entire family of these particles could be controlled in the same way.

One of the most significant findings is that this behavior is not just a theoretical curiosity but a robust physical phenomenon that survives even when the model is tested against more complex scenarios. The researcher checked their results by including higher energy states and different numerical methods, confirming that the lowest energy pair of states is sufficient to describe the main switching behavior. They also verified that the optical properties of these excitons change in a predictable way. At zero field, the exciton on the lower energy level does not absorb light well because its electrical pull is zero. However, as the electric field is applied and the exciton shifts to one side, it suddenly becomes bright and starts absorbing light strongly. This change in brightness, combined with the shift in energy, provides a clear signature that could be measured in a real experiment.

The study establishes a clear path to creating one-dimensional excitons that can be electrically reconfigured. Unlike previous systems where the electrical properties were fixed by the material's vertical stacking, this lateral design allows the electrical response to be tuned by the geometry of the strip and the applied voltage. The researcher noted that while the binding energy of the exciton remains strong regardless of the strip width, the coupling between the two boundaries changes rapidly with width. This decoupling of binding strength from switching sensitivity offers a new degree of freedom for designing optical and electronic devices. The ability to switch between a state with no electrical pull and a state with a large, directional pull using a modest electric field suggests potential applications in ultra-sensitive detectors or fast optical switches.

The work also clarifies the nature of these particles, ruling out the idea that they are simply two separate excitons sitting next to each other. Instead, the calculations show they are a single, coherent quantum object that spans the entire width of the strip. The researcher demonstrated that the transition from the balanced state to the localized state is not a sudden jump but a continuous evolution, where the probability of finding the exciton on one side gradually increases as the field strengthens. This continuous control is a key feature that distinguishes this system from other quantum switches that might snap abruptly from one state to another.

In summary, the researcher has mapped out how a single, narrow strip of material can host an exciton that behaves like a tunable electrical switch. By using computer simulations, they showed that the interplay between two nearby boundaries creates a unique state that can be manipulated with an electric field. The findings provide a theoretical blueprint for creating devices where light and electricity interact in highly controllable ways, driven by the simple geometric design of the material. The results suggest that by mastering the width of these nanometer-scale strips, scientists can create a new class of materials where the electrical and optical properties are not fixed by nature but are instead designed and controlled by human engineering.

Problem Statement
Lateral heterostructures, formed by joining different monolayer semiconductors side-by-side, host charge-transfer (CT) excitons where electrons and holes are confined to opposite materials but bound across the interface. While single-interface theory establishes that these excitons possess large in-plane dipoles, a fundamental question remains: can two neighboring, oppositely oriented CT configurations at a double heterojunction form a coherent exciton with a distinct electric response? Specifically, the paper investigates whether such a system can support a one-dimensional quadrupolar exciton—a state with zero permanent dipole but a non-zero electric quadrupole moment—that can be electrically switched into a dipolar state. This addresses a gap in understanding how geometric control (strip width) and electric bias can manipulate multipolar states in one dimension, complementing existing work on vertical heterostructures.

Methodology
The author employs a spatially resolved, effective-mass two-particle Hamiltonian to model a lateral double heterojunction consisting of a central MoS $_2$ strip embedded within a WS $_2$ monolayer (a WS $_2$ -MoS $_2$ -WS $_2$ structure).

Hamiltonian: The model treats the electron and hole as a two-body system with center-of-mass and relative coordinates. It incorporates band-offset potentials (confining the electron to the MoS $_2$ strip and the hole to the WS $_2$ regions), a screened Coulomb interaction (Rytova-Keldysh potential), and an external transverse electric field.
Numerical Solution: The author solves the reduced Hamiltonian using a basis expansion method. The wavefunction is expanded in terms of eigenfunctions of an auxiliary homogeneous screened-Coulomb problem, with coefficients dependent on the transverse center-of-mass coordinate. This approach captures the coupling between the transverse center-of-mass motion and internal relative motion, which is crucial for describing the deformation of the exciton across the interfaces.
Parameters: The model uses effective masses and band offsets derived from literature (e.g., Kormanyos et al., Huang et al.) and a phenomenological screening parameter. The strip width ( $L$ ) and interface smoothness ( $w$ ) are treated as geometric variables.

Key Contributions and Results

Existence of Quadrupolar Excitons: At zero electric field, the coupling between the left-interface ( $|L\rangle$ ) and right-interface ( $|R\rangle$ ) CT configurations produces two split eigenstates: a lower even-parity state and an upper odd-parity state. Both states possess zero permanent dipole due to left-right symmetry but exhibit a non-zero electric quadrupole moment.
Electric Switching Mechanism: The application of a transverse electric field continuously biases the system. As the field increases, the lower state evolves from a quadrupolar superposition (exhibiting a quadratic Stark shift) into a predominantly single-interface dipolar configuration (exhibiting an approximately linear Stark shift). The crossover occurs at a characteristic field $F_c = J/(ed)$ , where $J$ is the inter-interface coupling energy and $d$ is the dipole matrix element.
Geometric Control via Strip Width: The study demonstrates that the strip width ( $L$ $L$ ) is a powerful tuning parameter.
- Coupling vs. Binding: Reducing the strip width significantly increases the inter-interface coupling ( $J$ ) and the energy splitting between the even and odd states, while the exciton binding energy remains substantial (approx. 100 meV).
- Stark Sensitivity: The low-field Stark sensitivity (differential polarizability) is inversely proportional to the coupling $J$ . Narrower strips increase $J$ , thereby reducing the low-field sensitivity but increasing the spectral separation. Conversely, wider strips reduce $J$ , enhancing the Stark sensitivity. For a representative 1.5-nm strip, the calculated differential dipole response is approximately 146 D/(V/µm), which is significantly larger than that of representative vertical quadrupolar exciton systems.
Optical Signatures: The transition from quadrupolar to dipolar character is accompanied by a redistribution of optical transition strength. The even-derived branch is bright at zero field, while the odd-derived branch is dark due to destructive interference. As the field increases, the odd branch acquires optical strength ("brightening") as the states mix and parity is broken.
Higher-Order States: The study identifies an excited CT pair (CT2) with a larger spatial extent and weaker coupling than the ground state (CT1). While the lowest doublet (CT1) adequately describes the energy and dipole response in weak fields, higher-state mixing is necessary to accurately describe the redistribution of optical transition strengths in the excited spectrum.

Significance and Claims
The paper establishes a continuum-model route to realizing electrically reconfigurable one-dimensional quadrupolar excitons. It claims that lateral double junctions offer a distinct platform compared to vertical heterostructures because the hybridization scale is controlled by a directly designable in-plane length (the strip width) rather than interlayer tunneling.

The author asserts that the combination of an avoided crossing, a field-induced dipole approaching the single-interface value, the redistribution of optical strength, and the systematic dependence of the gap on interface separation provides a robust set of spectroscopic signatures for experimental detection. They note that nanometer-wide lateral strips have already been fabricated, and resolving the predicted doublet requires optical linewidths smaller than the meV-scale splitting. The work positions these structures as a platform for geometrically tunable and electrically switchable multipolar excitons, with potential implications for quantum information processing via a charge-configuration pseudospin, though the paper limits its discussion of future applications to this theoretical possibility without proposing specific experimental protocols.

More like this