Related Experiment Video
Updated: Aug 6, 2026
07:31
Realistic Membrane Modeling Using Complex Lipid Mixtures in Simulation Studies
Published on: September 1, 2023
Quantum machine learning for predicting properties of van der Waals bilayers
Chandra Chowdhury1, Livia Giordano1
1Department of Materials Science, University of Milano-Bicocca, Via Cozzi 55, 20125 Milano, Italy.
The Journal of Chemical Physics
|July 17, 2026
Summary
Quantum machine learning (QML) offers a new approach to predict electronic properties of van der Waals bilayers. This study benchmarks QML models, showing potential for materials discovery despite computational challenges.
Area of Science:
- Condensed Matter Physics
- Quantum Computing
- Materials Science
Background:
- Layering 2D materials into van der Waals (vdW) bilayers creates novel quantum states and tunable electronic properties.
- Exploring the vast configuration space of vdW bilayers is computationally intensive, driving the need for advanced methods like quantum machine learning (QML).
Purpose of the Study:
- To present a quantum-enhanced machine learning framework for predicting bandgaps of vdW bilayers using quantum simulators.
- To evaluate the performance of different quantum feature maps for classification and regression tasks in materials property prediction.
Main Methods:
- Conventional feature selection was performed to identify key descriptors for electronic properties.
- Electronic properties were encoded into quantum states using parameterized quantum circuits and three feature maps (ZFeatureMap, ZZFeatureMap, PauliFeatureMap).
- Quantum Support Vector Regressor (QSVR) and variational quantum regressor models were trained and benchmarked against classical Support Vector Regression (SVR).
Main Results:
- Task-dependent performance was observed for quantum feature maps; ZFeatureMap excelled in classification, while PauliFeatureMap performed best for regression (QSVR).
- QML models achieved significant predictive accuracy for bandgaps of 1850 semiconducting vdW bilayers, even with limited training data.
- Quantum classifiers showed competitive performance against classical Support Vector Classifiers (SVC), while classical SVR yielded the lowest regression RMSE.
Conclusions:
- The study demonstrates the feasibility of a simulator-based QML workflow for predicting vdW bilayer properties.
- Feature map selection is crucial and should be optimized for specific tasks (classification vs. regression).
- This work serves as an exploratory benchmark for hybrid classical-quantum models in vdW materials, rather than proving a definitive quantum advantage.
Related Concept Videos
01:10Van der Waals Equation
The ideal gas law is an approximation that works well at high temperatures and low pressures. The van der Waals equation of state (named after the Dutch physicist Johannes van der Waals, 1837−1923) improves it by considering two factors.
First, the attractive forces between molecules, which are stronger at higher densities and reduce the pressure, are considered by adding to the pressure a term equal to the square of the molar density multiplied by a positive coefficient a. Second, the volume...
First, the attractive forces between molecules, which are stronger at higher densities and reduce the pressure, are considered by adding to the pressure a term equal to the square of the molar density multiplied by a positive coefficient a. Second, the volume...
01:24Van der Waals Interactions
Atoms and molecules interact with each other through intermolecular forces. These electrostatic forces arise from attractive or repulsive interactions between particles with permanent, partial, or temporary charges. The intermolecular forces between neutral atoms and molecules are ion–dipole, dipole–dipole, and dispersion forces, collectively known as van der Waals forces.Polar molecules have a partial positive charge on one end and a partial negative charge on the other end of the molecule,...
01:26The Van der Waals Equation
The ideal gas law is based on two simplifying assumptions: first, that there are no intermolecular attractions between gas molecules, and second, that the volume occupied by the molecules themselves is negligible compared with the volume of the container. However, these assumptions don't hold up under all conditions - specifically, at high pressures and low temperatures, as gas tends to deviate from ideal gas behavior.The van der Waals equation is an enhanced version of the ideal gas law,...
04:01Real Gases: Effects of Intermolecular Forces and Molecular Volume Deriving Van der Waals Equation
Thus far, the ideal gas law, PV = nRT, has been applied to a variety of different types of problems, ranging from reaction stoichiometry and empirical and molecular formula problems to determining the density and molar mass of a gas. However, the behavior of a gas is often non-ideal, meaning that the observed relationships between its pressure, volume, and temperature are not accurately described by the gas laws.
01:27Electrostatic Boundary Conditions in Dielectrics
When an electric field passes from one homogeneous medium to another, crossing the boundary between the two mediums imparts a discontinuity in the electric field. This results in electrostatic boundary conditions that depend on the type of mediums the field propagates through.
Consider a case where both the mediums across a boundary are two different dielectric materials. Recall that the electric field and electric displacement are proportional and related through the material's permittivity.
Consider a case where both the mediums across a boundary are two different dielectric materials. Recall that the electric field and electric displacement are proportional and related through the material's permittivity.
02:56Intermolecular Forces and Physical Properties