INPUT
The material example is nitrogen-doped graphene. The energy of a configuration depends not only on how many nitrogen atoms are present but also on how those atoms are arranged.
AMAZON WEB SERVICES
Thermodynamic Sampling of Disordered Materials with an Analog Hamiltonian Rydberg Simulator
Researchers use a neutral-atom Rydberg processor available through Amazon Braket to sample low-energy configurations of nitrogen-doped graphene and study finite-temperature material behavior.
THE IDEA
Can a quantum processor sample material configurations?
Disordered materials can have an enormous number of possible atomic configurations. At a finite temperature, the goal is not necessarily to find only the lowest-energy configuration; instead, one wants samples distributed according to thermodynamic weights.
The researchers map an energy model derived from classical density functional theory onto a programmable Rydberg-atom Hamiltonian. In the demonstration, carbon and nitrogen occupancy is represented by the two-level states of neutral atoms.
The QuEra Aquila device, available through Amazon Braket, is used as an analog quantum sampler for the resulting model.
P(z,T) ∝ exp(−E(z)/kBT)
MAPPING
From graphene energetics to Rydberg atoms
The material example is nitrogen-doped graphene. The energy of a configuration depends not only on how many nitrogen atoms are present but also on how those atoms are arranged.
The researchers encode these interactions into distance-dependent Rydberg interactions. Laser detuning controls the energetic bias associated with excitation, while programmable atom spacing controls interaction strengths.
Because the physical hardware cannot directly reproduce the full material energy scale, the study introduces a rescaling procedure that connects the hardware sampling distribution to an effective material temperature.
METHOD
The researchers encode these interactions into distance-dependent Rydberg interactions. Laser detuning controls the energetic bias associated with excitation, while programmable atom spacing controls interaction strengths.
OUTPUT
Because the physical hardware cannot directly reproduce the full material energy scale, the study introduces a rescaling procedure that connects the hardware sampling distribution to an effective material temperature.
SAMPLING
Why imperfect annealing can be useful
An ideal adiabatic quantum annealer would aim for a ground state. The experiment instead uses the fact that a real analog Rydberg device operates for finite time and interacts weakly with its environment.
These non-idealities can make the device behave like a heuristic sampler that preferentially produces low-energy configurations. That behavior is aligned with the thermodynamic sampling problem.
The study calibrates an effective temperature and investigates how programmable parameters alter the resulting distribution.
INPUT
An ideal adiabatic quantum annealer would aim for a ground state. The experiment instead uses the fact that a real analog Rydberg device operates for finite time and interacts weakly with its environment.
METHOD
These non-idealities can make the device behave like a heuristic sampler that preferentially produces low-energy configurations. That behavior is aligned with the thermodynamic sampling problem.
OUTPUT
The study calibrates an effective temperature and investigates how programmable parameters alter the resulting distribution.
RESULTS
28 sites, 78 sites and temperature control
The work first validates the mapping on a 28-site graphene nanoflake, where exhaustive enumeration provides a classical reference.
The researchers then benchmark a 78-site system against classical Monte Carlo sampling and observe preferential sampling of low-energy configurations.
The study also demonstrates temperature tuning through programmable atom spacing, providing a route to changing the effective sampling behavior of the quantum device.
WHERE IT COULD MATTER
Potential applications
The approach is relevant to materials whose finite-temperature properties depend on many interacting configurations, including disordered alloys and doped materials.
A quantum sampler could become useful when direct enumeration of configurations becomes too expensive and when the quantities of interest depend on an ensemble rather than a single optimum.
The work is a proof of concept: scaling to larger, more realistic models and demonstrating a useful end-to-end advantage remain important research questions.
QCAASNET TAKEAWAY
What should you take away from this research?
The core idea
The key shift is to treat an analog quantum processor as a thermodynamic sampler rather than only as a ground-state optimizer.
Why it matters
The experiment connects DFT-derived material models to programmable neutral-atom hardware and uses effective-temperature calibration to interpret the samples.
What remains open
The result shows one concrete route by which quantum hardware could interact with materials modeling without requiring the device to return a perfect ground state every time.
This research should be read in the context of current quantum hardware and hybrid computing constraints. The results described here are research demonstrations, models or workflows rather than evidence that quantum processors have already replaced classical methods for general-purpose workloads.
ORIGINAL RESEARCH
Read the original research
Camino et al., Thermodynamic sampling of materials using neutral-atom quantum computers, npj Quantum Information (2026).