QMeCha: quantum Monte Carlo package for fermions in embedding environments

arXiv:2511.03439 · physics.chem-ph · Submitted 2025-11-05 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Physics. The day in general physics: instruments, methods and results across the field.

Kai: We're looking at the paper: "QMeCha: quantum Monte Carlo package for fermions in embedding environments".

Mira: Here's the gist of the paper. We present QMeCha (Quantum MeCha /’m"ke/), a novel, modular, and expandable Quantum Monte Carlo (QMC) code package designed to investigate many-body interactions among diverse quantum particles.

Kai: The package is now open access under the CC BY-NC-ND license at github.com/QMeCha.

Mira: QMeCha is engineered to solve Hamiltonians encompassing nuclei and fermions of varying mass and charge, such as electrons and positrons, embedded within an environment composed of classical charges and quantum Drude oscillators.

Kai: The code incorporates various wavefunction approaches for approximating the ground state of this many-particle operator.

Paper summary: Kai: So we're looking at the paper "QMeCha: quantum Monte Carlo package for fermions in embedding environments" again. The main thesis is that they’ve developed a modular and easy to expand Quantum Monte Carlo code specifically designed to study many-body interactions between different types of quantum particles.

Mira: They claim this package is now open access through github dot com QMeCha, and it's built to solve Hamiltonians involving nuclei and fermions of different mass and charge, like electrons or positrons.

Lev: The paper describes the system they are modeling as these particles embedded in an environment made up of classical charges alongside quantum Drude oscillators.

Kai: To find the ground state of this complex many-particle operator, the code offers different wavefunction approaches to approximate that ground state. For fermions, they go beyond just the traditional Slater determinant.

Mira: They include advanced functions like Geminal functions, such as the Pfaffian, and they incorporate explicit correlation terms within Jastrow factors for those fermions.

Lev: And for those classical charges and quantum Drude oscillators, they use different variational ansätze to explicitly capture dispersion, polarization, and electrostatic effects experienced by the nuclear and fermionic subsystem.

Kai: The integration of these wavefunctions is done through developed efficient Variational Monte Carlo methods, specifically VMC, and Diffusion Monte Carlo protocols called DMC. They also have a robust wavefunction optimization procedure featuring correlated sampling.

Mira: The paper applies this package to three primary areas: first, using DMC calculations benchmarked against LNO coupled cluster t methods to generate datasets for describing pocket ligand interactions within proteins.

Lev: Second, they use QMC methods for interacting electron positron systems because they can explicitly describe electrostatic correlation effects between those pairs via wavefunctions incorporating the two particle distances.

Kai: Third, they show the El-QDO embedding method’s ability to recover dynamical correlation between the embedding environment and electronic molecular systems, which is vital for modeling solvation energies and bond energies.

Mira: The total Hamiltonian they describe is split into H O plus H f plus V O f where H O is the QDO Hamiltonian, H f is the standard fermionic Hamiltonian, and V O f contains Coulomb interactions between the embedding potential particles and electronic system particles.

Lev: They construct a total wavefunction as a product of terms: Psi T of r sub f bar r sub d equals Psi f of r sub f bar r sub d times Psi d of r sub d bar times Psi f dot psi dbar, which describes the fermion and drudon parts.

Kai: The coupling function between fermions and QDOs takes an exponential form, recovering the dynamical correlation responsible for dispersion, polarization, and electrostatic effects.

Mira: They also support various trial wavefunctions for electrons and positrons including The Pfaffian, antisymmetrized geminal power AGP, and the Slater determinant for fermions.

Lev: Jastrow factors in these calculations include general cusp functions for fermion nucleus and fermion fermion cusps, along with a dynamical Jastrow factor describing few body correlation effects between fermions in the field of nuclei.

Conclusion: Kai: So the QMeCha package is essentially this new tool we've discussed. It’s a quantum Monte Carlo package for studying many-body interactions where particles are embedded in an environment of classical charges and Drude oscillators.

Mira: The authors present this as a modular system that allows researchers to build up these complex systems piece by piece using different wavefunction approaches. They make it open access on github dot com QMeCha.

Lev: It’s important for the community because it provides a framework where you can try out different numerical methods for tackling these hard problems, like VMC and DMC, which scale reasonably well on HPC facilities.

Kai: The paper shows concrete uses in modeling things like van der Waals interactions in macro molecules using DMC calculations benchmarked against LNO coupled cluster t methods.

Mira: They also use it to study electron positron systems and embedding approaches for quantum Monte Carlo, like the El-QDO method, which helps model solvation energies and bond energies.

Lev: Ultimately, it’s about providing a way to simulate these specific many-body problems with a framework that has both theoretical foundation and practical implementation details for running on real hardware.

Matteo Barborini, Jorge Alfonso Charry Martínez, Matej Ditte, Andronikos Leventis, Georgios Kafanas, Alexandre Tkatchenko

University of Luxembourg

physics.chem-ph

Submitted: 2025-11-05

Updated: 2026-01-30

Comments: 29 pages, 9 Figures, 2 Tables

Journal ref: J. Chem. Phys. 164, 062501 (2026)

DOI: 10.1063/5.0311994

Code: https://github.com/QMeCha/QMeCha_tools

License: http://creativecommons.org/licenses/by-nc-nd/4.0/

Importance score: 89/100

The gist: We present QMeCha (Quantum MeCha /’m"ke/), a novel, modular, and expandable Quantum Monte Carlo (QMC) code package designed to investigate many-body interactions among diverse quantum particles.

Key concepts

Quantum Monte Carlo (QMC)
A computational method used to find the lowest energy state of a quantum system. QMeCha uses this technique, specifically Variational Monte Carlo (VMC) and Diffusion Monte Carlo (DMC), to approximate the ground state wavefunction by simulating particle movements in configuration space.
Embedding Approaches
This involves placing the system of interest (like electrons) inside an environment made of other particles, such as classical charges or quantum Drude oscillators. This allows researchers to model how the environment influences the electronic system's properties like solvation energy.
Diffusion Monte Carlo (DMC)
A specific QMC algorithm used by QMeCha to calculate energies and properties. It works by simulating the evolution of a particle governed by a diffusion equation, helping to find highly accurate solutions for many-body problems, even when dealing with fermionic systems.
Jastrow Factors
Mathematical terms used in the trial wavefunctions within QMeCha. These factors are crucial for describing correlations between particles, such as how fermions interact with nuclei or each other, improving the accuracy of the calculated ground state.

Terminology

Summary

We present QMeCha (Quantum MeCha /’m"ke/), a novel, modular, and expandable Quantum Monte Carlo (QMC) code package designed to investigate many-body interactions among diverse quantum particles. The package is now open access under the CC BY-NC-ND license at github. p. 0

Where the paper says it

QMeCha: quantum Monte Carlo package for fermions in embedding environments Matteo Barborini , 1, 2, a) Jorge Alfonso Charry Martínez , 2 Matej Ditte , 2 Andronikos Leventis , 2 Georgios Kafanas , 1 and Alexandre Tkatchenko 2 1)HPC Platform, University of Luxembourg, L-4364 Esch-sur-Alzette, Luxembourg 2)Department of Physics and Materials Science, University of Luxembourg, L-1511 Luxembourg City, Luxembourg (Dated: February 2, 2026) We present the first open access version of the QMeCha (Quantum MeCha /’m"ke/) code, a quantum Monte Carlo (QMC) package developed to study many-body interactions between different types of quantum particles, with a modular and easy-to-expand structure. QMeCha is now available under CC BY-NC-ND license through the repository github.

com/QMeCha. p. 0
Where the paper says it

QMeCha is now available under CC BY-NC-ND license through the repository github. com/QMeCha.

QMeCha is engineered to solve Hamiltonians encompassing nuclei and fermions of varying mass and charge, such as electrons and positrons, embedded within an environment composed of classical charges and quantum Drude oscillators. p. 0
Where the paper says it

The present code has been built to solve the Hamiltonian of a system that can include nuclei and fermions of different mass and charge, e. electrons and positrons, embedded in an environment of classical charges and quantum Drude oscillators.

The code incorporates various wavefunction approaches for approximating the ground state of this many-particle operator. p. 0

Where the paper says it

electrons and positrons, embedded in an environment of classical charges and quantum Drude oscillators. To approximate the ground state of this many-particle operator, the code features different wavefunctions.

For fermionic particles, in addition to the conventional Slater determinant, QMeCha supports advanced functions like Geminal functions (e. p. 0
Where the paper says it

To approximate the ground state of this many-particle operator, the code features different wavefunctions. For the fermionic particles, beyond the traditional Slater determinant, QMeCha also includes Geminal functions such as the Pfaffian, and presents different types of explicit correlation terms in the Jastrow factors.

g., Pfaffian) and includes explicit correlation terms within Jastrow factors. p. 0
Where the paper says it

To approximate the ground state of this many-particle operator, the code features different wavefunctions. For the fermionic particles, beyond the traditional Slater determinant, QMeCha also includes Geminal functions such as the Pfaffian, and presents different types of explicit correlation terms in the Jastrow factors.

The classical point charges and quantum Drude oscillators are modeled using different variational ansätze to explicitly capture dispersion, polarization, and electrostatic effects experienced by the nuclear and fermionic subsystem. p. 0
Where the paper says it

For the fermionic particles, beyond the traditional Slater determinant, QMeCha also includes Geminal functions such as the Pfaffian, and presents different types of explicit correlation terms in the Jastrow factors. The classical point charges and quantum Drude oscillators, described through different variational ansätze, are used to model a molecular environment capable of explicitly describing dispersion, polarization, and electrostatic effects experienced by the nuclear and fermionic subsystem.

Integration of these wavefunctions is facilitated through developed efficient Variational Monte Carlo (VMC) and Diffusion Monte Carlo (DMC) protocols, complemented by a robust wavefunction optimization procedure featuring correlated sampling. p. 0
Where the paper says it

The classical point charges and quantum Drude oscillators, described through different variational ansätze, are used to model a molecular environment capable of explicitly describing dispersion, polarization, and electrostatic effects experienced by the nuclear and fermionic subsystem. To integrate these wavefunctions, efficient variational Monte Carlo and diffusion Monte Carlo protocols have been developed, together with a robust wavefunction optimization procedure that features correlated sampling.

The QMeCha package has been applied to three primary application areas detailed in Section VI:

  1. Van der Waals interactions in macro-molecules: DMC calculations, benchmarked against state-of-the-art LNO-CCSD(T) methods, are utilized to generate datasets for describing pocket-ligand interactions within proteins.

  2. Interacting Electron-Positron systems: QMC methods are promising for studying electron–positron bound states because they can explicitly describe electrostatic correlation effects between electron–positron pairs via wavefunctions incorporating the two-particle distances.

  3. Embedding approaches for quantum Monte Carlo: The El-QDO embedding method has demonstrated its ability to recover dynamical correlation between the embedding environment and electronic molecular systems, which is vital for modeling environmental effects on solvation energies, electronic excitations, and bond energies.

The QMeCha code is implemented in Fortran 2008 and supports two levels of parallelization via Message Passing Interface (MPI) and shared-memory Open Multi-Processing (OpenMP). Its general structure involves initialization of the MPI/OpenMP environment, input reading, and calls to modules executing VMC, DMC, or wavefunction optimization procedures. p. 1

Where the paper says it

The full Hamiltonian that comprises the two types of quantum particles, ie fermions and drudons, is integrated over one correlated wavefunction that is constructed as the product of various fermionic terms, such as the Slater determinant, and Jastrow factors. Within the code, the observables are computed through an efficient implementation of the VMC method, and through an improved size-consistent DMC algorithm, that reduces finite time-step errors when computing energy differences35,51,52 .

The intrinsic parallelization ensures near perfect weak scaling with respect to the number of nodes. p. 1
Where the paper says it

First, the stochastic nature of their algorithms renders them highly parallelizable on modern high-performance computing (HPC) facilities. This together with the favourable scaling of the computational cost with respect to the number of particles N in the system, O(N3∼4 ), brings a competitive advantage for large systems when compared to the previously discussed deterministic approaches.

The code supports various trial wavefunctions for electrons and positrons, including The Pfaffian, antisymmetrized geminal power (AGP), and the Slater determinant. p. 7

Where the paper says it

8 QMeCha we follow this last scheme, for which P =   T ↑ G ϕ¯ ↑ −G⊤ T ↓ ϕ¯ ↓ −ϕ¯⊤ ↑ −ϕ¯⊤ ↓ 0   , (25) where T ↑ , T ↓ and G are the matrices defined above, while the elements of the last column (or row) vectors ϕ¯↑ i = ϕ ↑ (r ↑ i ) = X Q q l ↑ qφq(r ↑ i ) i ∈ N ↑ e (26) and ϕ¯↓ i = ϕ ↓ (r ↓ i ) = X Q q l ↓ qφq(r ↓ i ) i ∈ N ↓ e (27) represent the values of molecular orbitals, occupied by the spin up or spin down electrons (or positrons), and depending on the linear coefficients l ↓ q and l ↑ q that are in general different for the two spin populations. The antisymmetrized geminal power To recall the relationship between the Pfaffian wavefunction and antisymmetrized geminal power (AGP) wavefunction95–97 let us consider a closed shell system with an even number of electrons.

Jastrow factors include general cusp functions for fermion-nucleus and fermion-fermion cusps, alongside a dynamical Jastrow factor describing few-body correlation effects between fermions in the field of nuclei. p. 8
Where the paper says it

General cusp functions The Jastrow Jc(¯r) term that is used to describe the many particles’ cusp conditions is written as the linear combination Jc(x¯) = X Nf i=1 X Nn a=1 f1b(riq) + X Nf j>i=1 f2b(rij ) (36) of one-body terms, describing the fermion-nucleus cusp, and the two-body terms, describing the fermion-fermion cusps, depending respectively on the distance between a fermion and a nucleus, ria = |ri − Ra|, and the distance between two fermions rij = |ri − rj |. For two general charged particles with finite masses, interacting via the Coulomb potential, Kato’s cusp conditions112,113 are written in the form 1 ⟨Ψ⟩ ∂ ⟨Ψ⟩ ∂rij rij=0 = 2qiqj 3 ± 1 mimj mi + mj = Γ (37) where in the factor 1 3±1 the + sign holds for indistinguishable particles and the − holds for distinguishable ones.

The QDO wavefunction can be constructed either from an exact solution of the QDO model or as a Hartree product involving molecular orbitals. p. 9
Where the paper says it

An alternative approach to the QDO wavefunction, introduced in ref. 49 and is built as the Hartree product of Ψd(¯r d ; R¯ O ) = Y NO i=1 ϕi(r d i ), (43) where ϕi(r d i ) = X Q q=1 α i qφq(r d i ) (44) are molecular orbitals constructed on a basis of atomic Gaussian (around the QDO’s center) or Slater or Mixed type of orbitals (Section III F).

The coupling function between fermions and QDOs takes an exponential form, recovering the dynamical correlation responsible for dispersion, polarization, and electrostatic effects. p. 9
Where the paper says it

45 has the purpose of recovering the dynamical correlation between the electrons and drudons. This factor is chosen to be always positive, since it recovers correlation that is responsible for dispersion, polarization, and electrostatic effects that change the nodal structure of the fermionic sub-system only indirectly, through the fermionic wavefunction described in Sections III A, III B and III C.

QMC methods implemented include VMC (stochastic integration over the energy functional using Metropolis-Hastings sampling) and DMC (a fixed-node diffusion Monte Carlo algorithm inspired by Umrigar et al. p. 1

Where the paper says it

The long list of modern QMC methods can be divided into two main categories. The first category includes the methods that are based on the stochastic sampling of the system’s degrees of freedom, such as the particles’ coordinates, and includes the most famous variational Monte Carlo (VMC)26,27, diffusion Monte Carlo (DMC)28 and Lattice-Regularized diffusion Monte Carlo (LRMDC)30 methods.

, employing the FixedNode (FN) approximation to address the fermionic sign problem). p. 0
Where the paper says it

INTRODUCTION The solution of the many-body Schrödinger equation1 for molecular or periodic systems of atomic nuclei and electrons, has been one of the main challenges for theoretical physicists and chemists since its conceptualization. The daunting task of constructing an analytical solution to the problem of interacting nuclei and electrons led to the development of numerous models to tackle its solution at least in an approximate manner.

Optimization methods include Stochastic Reconfiguration (SR), which utilizes generalized forces derived from a stochastic differential equation, and correlated sampling techniques. p. 1
Where the paper says it

The long list of modern QMC methods can be divided into two main categories. The first category includes the methods that are based on the stochastic sampling of the system’s degrees of freedom, such as the particles’ coordinates, and includes the most famous variational Monte Carlo (VMC)26,27, diffusion Monte Carlo (DMC)28 and Lattice-Regularized diffusion Monte Carlo (LRMDC)30 methods.

The code supports external point charges as classical embedding, where interactions are screened using a damping function to mitigate divergences. p. 0

Where the paper says it

External point charges as classical embedding 4 E. Screening of the interactions 5 F.

It also accommodates atomic pseudopotentials for core electrons and an external electric polarization field. p. 0
Where the paper says it

Atomic pseudopotentials 5 G. External electric polarization field 5 III.

The El-QDO embedding method enables the electronic system, described by its Hamiltonian, to be embedded in an environment of Coulomb-coupled charged harmonic oscillators (QDOs) and point charges, which is critical for modeling solvation energies and bond energy variations. p. 1
Where the paper says it

48 to efficiently capture the quantum correlation effects between an environment modelled through QDOs and point charges67–69 and a molecular subsystem of interest. In this application we have focused on the evaluation of solvation energies, excitation energies and bond energy variations of a set of dimers in water clusters.

The computational cost associated with the El-QDO method is negligible compared to integrating the electronic degrees of freedom. p. 1
Where the paper says it

First, the stochastic nature of their algorithms renders them highly parallelizable on modern high-performance computing (HPC) facilities. This together with the favourable scaling of the computational cost with respect to the number of particles N in the system, O(N3∼4 ), brings a competitive advantage for large systems when compared to the previously discussed deterministic approaches.

The QMeCha code has been employed to compute interaction energies of large molecules via van der Waals (vdW) interactions, study electron-positron metastable interactions in molecules as an efficient alternative to deterministic methods like MBPT and CI, and characterize changes in molecular bonding and excitation within embedding environments built from point charges and QDOs. p. 1

Where the paper says it

The first application focuses on the description of van der Waals (vdW) interactions in macromolecules53, where the DMC calculations are used as state-of-the-art references together with the LNO-CCSD(T)54–57 method, to construct a dataset for the description of pocket-ligand interactions in proteins. The second application is related to positronic chemistry, an emerging field dedicated to the investigation of the still vastly unexplored bound metastable states of positrons with atomic and molecular systems, before electron-positron pair annihilation58–60 .

The two reference methods used are FN-DMC28 and CCSD(T)20. p. 1
Where the paper says it

Finally, QMeCha has been used to tackle three main sets of applications reported in Section VI. The first application focuses on the description of van der Waals (vdW) interactions in macromolecules53, where the DMC calculations are used as state-of-the-art references together with the LNO-CCSD(T)54–57 method, to construct a dataset for the description of pocket-ligand interactions in proteins.

A significant discrepancy has been noted between binding energies estimated by CCSD(T) on single Slater determinants derived from HF or DFT molecular orbitals and those obtained by FN-DMC. p. 7
Where the paper says it

In this situation, we obtain a reduced geminal matrix that corresponds to the matrix product G∗ = S ↑⊤S ↓ between the square matrices S ↑ and S ↓ containing respectively the values of the Ne/2 molecular orbitals computed on the N↑ e = Ne/2 spin up and N↓ e = Ne/2 spin-down electrons. This reduced geminal matrix corresponds to the Slater determinant wavefunction written as the product of the two determinants of S ↑ and S ↓ ψe(x¯ e ) = det h S ↑⊤S ↓ i = det h S ↑ i det h S ↓ i .

The total Hamiltonian described is H hat tot = H hat O + H hat f + V O,f, where H hat O is the QDO Hamiltonian, H hat f is the standard fermionic Hamiltonian, and V O,f contains Coulomb interactions between embedding potential particles and electronic system particles. p. 3

Where the paper says it

3, Hˆ f is the standard fermionic Hamiltonian defined in eq. 1 and VO,f contains the Coulomb interactions between the particles of the embedding potential and those of the electronic system VO,f = X Nf i=1 X NO j=1 qiρj ri − RO j − qiρj ri − r d j !

The total wavefunction is constructed as a product of terms: psi T(r bar f, r bar d) = psi f(r bar f) psi d(r bar d) psi f,d(r bar f, r bar d). p. 1
Where the paper says it

In the years that followed, the code progressively grew to describe interactions between fermions of different charge and mass, such as positrons, muons and anti-muons, interacting with molecules46,47, and to include a QMC embedding procedure in which a molecular system of nuclei and fermions can be immersed in an environment approximated by point charges and quantum Drude oscillators (QDOs)48–50. The full Hamiltonian that comprises the two types of quantum particles, ie fermions and drudons, is integrated over one correlated wavefunction that is constructed as the product of various fermionic terms, such as the Slater determinant, and Jastrow factors.

Observables are computed efficiently using VMC and an improved size-consistent DMC algorithm. p. 1
Where the paper says it

The full Hamiltonian that comprises the two types of quantum particles, ie fermions and drudons, is integrated over one correlated wavefunction that is constructed as the product of various fermionic terms, such as the Slater determinant, and Jastrow factors. Within the code, the observables are computed through an efficient implementation of the VMC method, and through an improved size-consistent DMC algorithm, that reduces finite time-step errors when computing energy differences35,51,52 .

The code is available under CC BY-NC-ND license at github. p. 0

Where the paper says it

QMeCha: quantum Monte Carlo package for fermions in embedding environments Matteo Barborini , 1, 2, a) Jorge Alfonso Charry Martínez , 2 Matej Ditte , 2 Andronikos Leventis , 2 Georgios Kafanas , 1 and Alexandre Tkatchenko 2 1)HPC Platform, University of Luxembourg, L-4364 Esch-sur-Alzette, Luxembourg 2)Department of Physics and Materials Science, University of Luxembourg, L-1511 Luxembourg City, Luxembourg (Dated: February 2, 2026) We present the first open access version of the QMeCha (Quantum MeCha /’m"ke/) code, a quantum Monte Carlo (QMC) package developed to study many-body interactions between different types of quantum particles, with a modular and easy-to-expand structure. QMeCha is now available under CC BY-NC-ND license through the repository github.

com/QMeCha, with the repository containing tools for wavefunction file initialization and the open version of the code at github. p. 0
Where the paper says it

QMeCha: quantum Monte Carlo package for fermions in embedding environments Matteo Barborini , 1, 2, a) Jorge Alfonso Charry Martínez , 2 Matej Ditte , 2 Andronikos Leventis , 2 Georgios Kafanas , 1 and Alexandre Tkatchenko 2 1)HPC Platform, University of Luxembourg, L-4364 Esch-sur-Alzette, Luxembourg 2)Department of Physics and Materials Science, University of Luxembourg, L-1511 Luxembourg City, Luxembourg (Dated: February 2, 2026) We present the first open access version of the QMeCha (Quantum MeCha /’m"ke/) code, a quantum Monte Carlo (QMC) package developed to study many-body interactions between different types of quantum particles, with a modular and easy-to-expand structure. QMeCha is now available under CC BY-NC-ND license through the repository github.

com/QMeCha/QMeCha_code. The authors acknowledge financial support from grants and computational resources, including those from the Luxembourg National Research Fund and Argonne Leadership Computing Facility, with calculations executed on facilities such as MeluXina HPC. p. 1
Where the paper says it

First, the stochastic nature of their algorithms renders them highly parallelizable on modern high-performance computing (HPC) facilities. This together with the favourable scaling of the computational cost with respect to the number of particles N in the system, O(N3∼4 ), brings a competitive advantage for large systems when compared to the previously discussed deterministic approaches.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved AI can perform accurate calculations for van der Waals interactions in large macro-molecules by using QMC methods like FN-DMC and comparing them against gold standard methods like CCSD(T).

  2. Improved AI can explore new positronic complexes by integrating explicitly correlated wavefunctions, allowing it to model positronic molecules in terms of electron-positron orbitals constructed on the eigenfunctions of positronium bound states.

  3. Improved AI can accurately describe how the environment affects solvation energies, electronic excitations, and bond energies by using the El-QDO embedding method to recover dynamical correlation between the embedding environment and molecular subsystems.

  4. Improved AI can efficiently handle complex many-body systems by utilizing quantum drudonic degrees of freedom in environments, which reduces computational cost compared to treating all degrees of freedom explicitly.

Abstract

We present the first open access version of the QMeCha (Quantum MeCha) code, a quantum Monte Carlo (QMC) package developed to study many-body interactions between different types of quantum particles, with a modular and easy-to-expand structure. The present code has been built to solve the Hamiltonian of a system that can include nuclei and fermions of different mass and charge, e.g. electrons and positrons, embedded in an environment of classical charges and quantum Drude oscillators. To approximate the ground state of this many-particle operator, the code features different wavefunctions. For the fermionic particles, beyond the traditional Slater determinant, QMeCha also includes Geminal functions such as the Pfaffian, and presents different types of explicit correlation terms in the Jastrow factors. The classical point charges and quantum Drude oscillators, described through different variational ansätze, are used to model a molecular environment capable of explicitly describing dispersion, polarization, and electrostatic effects experienced by the nuclear and fermionic subsystem. To integrate these wavefunctions, efficient variational Monte Carlo and diffusion Monte Carlo protocols have been developed, together with a robust wavefunction optimization procedure that features correlated sampling.

Sources

Related papers