A method now computes single-particle Green’s functions by extending a technique previously used for ground state calculations; these describe how electrons move through materials and reveal key spectral properties like quasiparticle residue and Hubbard bands. The approach reconstructs information about electron behaviour using a shallow circuit design on quantum devices requiring only short computational steps. A computational technique originally designed for determining ground states has been adapted to calculate single-particle Green’s functions, describing how electrons behave within materials.

This refined method efficiently reconstructs information about electron movement using relatively simple quantum circuits, reducing demands on computing power. Consequently, more complex material properties can be understood through improved modelling utilising emerging quantum computers, representing advancement in linking theoretical predictions with current hardware limitations. A computational technique used for determining ground states now calculates single-particle Green’s functions; these describe how electrons behave within materials, akin to tracking an individual electron’s movement through a crowded space revealing its energy and direction.

The new approach efficiently reconstructs information about electron behaviour using relatively simple quantum circuits, reducing demands on computing power and enabling more complex material properties to be understood via improved modelling utilising emerging quantum computers. The team applied this method to simulate the behaviour of correlated materials, those where interactions between electrons are significant, which often defy traditional calculation methods due to the exponential growth in complexity as system size increases.

Like zooming into one specific atom within a large material before scaling up simulation, researchers used dynamical mean-field theory (DMFT) alongside their technique; DMFT maps a lattice onto an impurity interacting with a self-consistently determined bath.

Reduced Hilbert space access unlocks accurate modelling of strong electron correlations

a small fraction of the full Hilbert space accurately recovered key features within electron behaviour; previously this required accessing exponentially larger computational spaces limiting simulations to very simple systems. Applying a new method to simulate an Anderson impurity model, representing electrons interacting with surrounding material, successfully reproduced spectral characteristics across interaction strengths spanning transitions between metallic and insulating states. Deeper probing into strongly correlated materials is now possible where traditional methods struggle due to complexity arising from numerous interactions between electrons.

The team at Center for Computation and Technology, Louisiana State University has demonstrated that their novel computational technique simulates electron behaviour using less than one percent of resources needed by standard approaches. Improved accuracy in dynamical mean-field theory (DMFT) calculations resulted from incorporating more ‘bath sites’, which represent the interactions with surrounding material; this represents a step forward, enabling study of materials where many interacting electrons make calculations extremely difficult. Despite these promising results, the method currently confirms features only for systems exhibiting particle-hole symmetry and its applicability to all strongly correlated materials remains unproven.

Single-particle Green’s functions were successfully calculated extending an initial ground state determination technique, describing how electrons move through materials and revealing important properties like energy and direction of travel. Employing Krylov subspaces, focused areas within all possible quantum system states, alongside classical calculations allowed reconstruction of electron behaviour using simplified circuits on quantum devices. Consequently, accurate results were achieved by examining just a small fraction of total potential states, markedly reducing demands on computing power compared with conventional methods.

A route towards simulating complex materials currently intractable for conventional techniques is now potentially available; dynamical mean-field theory (DMFT), a key technique in condensed matter physics, often requires immense computational resources to accurately represent electron interactions within these substances. This initial demonstration relies upon an Anderson impurity model exhibiting particle-hole symmetry, a simplification where electrons and ‘holes’, representing missing electrons, behave identically.

While real materials exhibit far greater complexity in their electron behaviour, successfully recreating key features of material properties using fewer computational steps than traditional methods represents significant progress nonetheless. The approach could eventually enable more detailed simulations of complex substances on emerging quantum computers, surpassing current limitations and accelerating materials’ discovery.

Single-particle Green’s functions were calculated by extending a method initially used for ground state determination, providing insight into how electrons move within materials. Utilising Krylov subspaces alongside classical computation allowed researchers to reconstruct these behaviours with simplified circuits on quantum devices, achieving accurate results while examining only a fraction of all possible system states. This work demonstrates the potential to simulate electron interactions in systems studied via dynamical mean-field theory using fewer computational resources than previously required. The authors suggest this may allow larger numbers of bath sites to be modelled on near-term quantum hardware.

👉 More information
🗞 Green’s Functions from Sample-based Krylov Quantum Diagonalization: An Impurity Solver for Dynamical Mean-Field Theory
✍️ Jay Patel, Chakradhar Rangi and Ka-Ming Tam
🧠 ArXiv: https://arxiv.org/abs/2609.09147

More like this