Atomic Simulation Environment Research Articles

JEDI: A versatile code for strain analysis of molecular and periodic systems under deformation.

Stretching or compression can induce significant energetic, geometric, and spectroscopic changes in materials. To fully exploit these effects in the design of mechano- or piezo-chromic materials, self-healing polymers, and other mechanoresponsive devices, a detailed knowledge about the distribution of mechanical strain in the material is essential. Within the past decade, Judgement of Energy DIstribution (JEDI) analysis has emerged as a useful tool for this purpose. Based on the harmonic approximation, the strain energy in each bond length, bond angle, and dihedral angle of the deformed system is calculated using quantum chemical methods. This allows the identification of the force-bearing scaffold of the system, leading to an understanding of mechanochemical processes at the most fundamental level. Here, we present a publicly available code that generalizes the JEDI analysis, which has previously only been available for isolated molecules. Now, the code has been extended to two- and three-dimensional periodic systems, supramolecular clusters, and substructures of chemical systems under various types of deformation. Due to the implementation of JEDI into the Atomic Simulation Environment, the JEDI analysis can be interfaced with a plethora of program packages that allow the calculation of electronic energies for molecular systems and systems with periodic boundary conditions. The automated generation of a color-coded three-dimensional structure via the Visual Molecular Dynamics program allows insightful visual analyses of the force-bearing scaffold of the strained system.

SMATool: Strength of materials analysis toolkit

The study of the strength of materials is a cornerstone in material science and engineering, playing a critical role in shaping the progress and application of materials in diverse industrial sectors. The strength of a material is meticulously examined to understand the behavior of the material under different stress conditions and environments, thereby guiding material selection and structural design. Herein, we introduce the SMATool, a computational toolkit for the efficient calculation and analysis of material strength at both zero and finite temperatures for 3D, 2D, 1D, and tubular 2D-based nanostructures and nanotubes, as well as 1D nanoribbons. The toolkit is capable of calculating tensile, shear, ultimate, yield, and indentation (Vickers' hardness) strengths in various dimensions, as well as the energy storage capacity. We conducted several calculations both at zero and finite temperatures to validate the accuracy and reliability of the developed software. The results show that the SMATool package provides accurate predictions that align with existing data on material strength. SMATool integrates seamlessly with widely used electronic structure codes like VASP and Quantum Espresso, providing a user-friendly interface catering to academic researchers and industry professionals. SMATool is useful for either the exploration of the strength of materials or high-throughput new material design. SMATool is open-source and available on GitHub at SMATool@github and at Zenodo at 10.5281/zenodo.10780514. Program summaryProgram Title: SMAToolCPC Library link to program files:https://doi.org/10.17632/fnkzfsg7t7.1Licensing provisions: GNU General Public License, version 3Programming language: Python 3Nature of problem: Developing a comprehensive framework to accurately determine the strength of materials is crucial in various engineering and scientific disciplines. This involves developing a robust and efficient computational toolkit to predict the strength of a material under different loading conditions, including tension, compression, shear, and indentation.Solution method:SMATool effectively addresses the challenge of accurately simulating the mechanical properties of materials by leveraging advances in ab initio methods, including density functional theory and molecular dynamics within the atomic simulation environment. The SMATool toolkit can efficiently compute various material strengths, such as tensile, shear, ultimate, yield, and indentation (Vickers' hardness) strengths as well as the corresponding energy storage capacity of the material. It is adept at handling a variety of material dimensions, ranging from 1D nanotubes and nanoribbons to 2D and 3D structures, at both zero and finite temperatures. SMATool currently integrates VASP and Quantum Espresso as the stress calculator and is easily adaptable to use other electronic structure codes.Additional comments including restrictions and unusual features:SMATool offers serial and parallel execution. Depending on the system size, the nature of the calculations (zero or finite temperature), and the computing architecture, the computing time can vary from ∼5 minutes to as high as 18 hours.

Locating Transition States by Variational Reaction Path Optimization with an Energy-Derivative-Free Objective Function.

Locating transition states is essential for understanding molecular reactions. We propose a double-ended transition state search method by revisiting a variational reaction path optimization method known as the MaxFlux method. Although its original purpose is to add temperature effects to reaction paths, we conversely let the temperature approach zero to obtain an asymptotically exact minimum energy path and its corresponding transition state in variational formalism with an energy-derivative-free objective function. Using several numerical techniques to directly optimize the objective function, the present method reliably finds transition states with low computational cost. In particular, only three force evaluations per iteration are sufficient. This is confirmed on a variety of molecular reactions where the nudged elastic band method often fails. The present method is implemented in Python using the Atomic Simulation Environment and is available on GitHub.

GPAW: An open Python package for electronic structure calculations.

We review the GPAW open-source Python package for electronic structure calculations. GPAW is based on the projector-augmented wave method and can solve the self-consistent density functional theory (DFT) equations using three different wave-function representations, namely real-space grids, plane waves, and numerical atomic orbitals. The three representations are complementary and mutually independent and can be connected by transformations via the real-space grid. This multi-basis feature renders GPAW highly versatile and unique among similar codes. By virtue of its modular structure, the GPAW code constitutes an ideal platform for the implementation of new features and methodologies. Moreover, it is well integrated with the Atomic Simulation Environment (ASE), providing a flexible and dynamic user interface. In addition to ground-state DFT calculations, GPAW supports many-body GW band structures, optical excitations from the Bethe-Salpeter Equation, variational calculations of excited states in molecules and solids via direct optimization, and real-time propagation of the Kohn-Sham equations within time-dependent DFT. A range of more advanced methods to describe magnetic excitations and non-collinear magnetism in solids are also now available. In addition, GPAW can calculate non-linear optical tensors of solids, charged crystal point defects, and much more. Recently, support for graphics processing unit (GPU) acceleration has been achieved with minor modifications to the GPAW code thanks to the CuPy library. We end the review with an outlook, describing some future plans for GPAW.

MASAP: A package for atomic scattering amplitude in solids

We present the MsSpec Atomic Scattering Amplitude Package (MASAP), composed of a computation program and a graphical interface to generate atomic scattering amplitude (ASA) of an atom, either isolated or embedded in an environment, at any chosen energy of the impinging electron up to ≈15 KeV. The ASA is calculated using an effective, complex optical potential which provides damping effects in the scattering process in a fully relativistic framework. Optionally, scalar relativistic and non-relativistic approximations are also available to assess their applicability to a given problem. In order to describe electron propagation in solids we suggest to replace ASA's based on Plane Waves (PW) scattering with effective ASA's based on curved Spherical Waves (SW) using truncated-overlapped potentials of the Muffin-Tin (MT) type constructed according to the Mattheiss prescription. The graphical user interface generates not only ASA data files providing atomic Differential Cross Sections (DCS) but also files of related quantities such as total Cross Section (CS), both elastic and inelastic, atomic tl-matrices and phase shifts. We found in general that the imaginary part of the optical potential enhances the calculated elastic DCSs in the forward direction compared to the same potential without the imaginary part, a feature related to the optical theorem, but gives rise to a lower intensity at all other directions as expected due to the damping effect of the complex part of the potential. We show calculated differential and transport Cross Sections for aluminum and gold atoms both in isolation and in crystals with the Face-Centered-Cubic (FCC) structure. Program summaryProgram Title: MASAPCPC Library link to program files:https://doi.org/10.17632/gtxp3mn5hs.1Licensing provisions: Apache 2.0Programming language: FORTRAN 77, Python3External routines:•SciPy (https://www.scipy.org)•Atomic Simulation Environment (https://wiki.fysik.dtu.dk/ase/)•h5py (https://www.h5py.org)•wxPython (https://www.wxpython.org)Nature of problem: Calculation of atomic scattering amplitudes of scattered electrons by an embedded atom in a solid including damping and relativistic effects. Comparison between ASA's in an isolated atomic system and a solid. Introduction of an effective ASA for electrons propagating in a solid derived from Multiple Scattering Theory.Solution method: Potential scattering theory. Muffin-tin approximation. Superposition of atomic electron densities obtained by self-consistent Dirac-Fock atomic calculations. Hedin-Lundqvist potential as an optical potential. Fully relativistic theory, with scalar-relativistic and non-relativistic approximations for comparison.Additional comments including restrictions and unusual features: Even though the program can perform calculations at impinging electron energies up to ≈15 Kev, DCS calculations are recommended in the range 50 : ≈1500 eV both in the case of isolated atoms and atoms embedded in solids, since this interval is more suitable for studying the details of the atomic potentials in their environment.In the case of solids, at energies lower than ≈50 eV, the details of the electronic structure might be important, whereas in the case of isolated atoms this range might suffer from the neglect of the polarization potential and coupling to available inelastic channels. In this respect an option to add an empirical form of a polarization potential to investigate its effects is available to the interested user.

Wfl Python toolkit for creating machine learning interatomic potentials and related atomistic simulation workflows.

Predictive atomistic simulations are increasingly employed for data intensive high throughput studies that take advantage of constantly growing computational resources. To handle the sheer number of individual calculations that are needed in such studies, workflow management packages for atomistic simulations have been developed for a rapidly growing user base. These packages are predominantly designed to handle computationally heavy abinitio calculations, usually with a focus on data provenance and reproducibility. However, in related simulation communities, e.g., the developers of machine learning interatomic potentials (MLIPs), the computational requirements are somewhat different: the types, sizes, and numbers of computational tasks are more diverse and, therefore, require additional ways of parallelization and local or remote execution for optimal efficiency. In this work, we present the atomistic simulation and MLIP fitting workflow management package wfl and Python remote execution package ExPyRe to meet these requirements. With wfl and ExPyRe, versatile atomic simulation environment based workflows that perform diverse procedures can be written. This capability is based on a low-level developer-oriented framework, which can be utilized to construct high level functionality for user-friendly programs. Such high level capabilities to automate machine learning interatomic potential fitting procedures are already incorporated in wfl, which we use to showcase its capabilities in this work. We believe that wfl fills an important niche in several growing simulation communities and will aid the development of efficient custom computational tasks.

Atom-centered machine-learning force field package

In recent years, machine learning algorithms have been widely used for constructing force fields with an accuracy of ab initio methods and the efficiency of classical force fields. Here, we developed a python-based atom-centered machine-learning force field (PyAMFF) package to provide a simple and efficient platform for fitting and using machine learning force fields by implementing an atom-centered neural-network algorithm with Behler-Parrinello symmetry functions as structural fingerprints. The following three features are included in PyAMFF: (1) integrated Fortran modules for fast fingerprint calculations and Python modules for user-friendly integration through scripts and facile extension of future algorithms; (2) a pure Fortran backend to interface with the software, including the long-timescale dynamic simulation package EON, enabling both molecular dynamic simulations and adaptive kinetic Monte Carlo simulations with machine-learning force fields; and (3) integration with the Atomic Simulation Environment package for active learning and ML-based algorithm development. Here, we demonstrate an efficient parallelization of PyAMFF in terms of CPU and memory usage and show that the Fortran-based PyAMFF calculator exhibits a linear scaling relationship with the number of symmetry functions and the system size. Program summaryProgram title: python-based atom-centered machine-learning force field (PyAMFF)CPC Library link to program files:https://doi.org/10.17632/fsn6dkcvrv.1Developer's repository link:https://gitlab.com/pyamff/pyamffLicensing provisions: Apache License, 2.0Nature of problem: Determine an approximate (surrogate) model based upon atomic forces and energies from density functional theory (DFT). With a surrogate model that is less computationally expensive to evaluate than DFT, there can be a rapid exploration of the potential energy surface, accelerated optimization to minima and saddle points, and ultimately, accelerated design of active materials where the kinetics are key to the material function.Solution method: The atomic environments of training data are calculated in terms of Behler-Parrinello fingerprints. These fingerprints are passed to a neural network which is trained to reproduce the energy and force of the training data. A parallel implementation and Fortran backend allow for efficient training and calculation of the resulting surrogate model. Examples of long-time simulations of materials on the surrogate model surfaces are provided.

Оценка энергии формирования вакансий в ОЦК-, ГЦК- и ГПУ-металлах с использованием теории функционала плотности

Introduction. Vacancies are among the crystal lattice defects that have a significant effect on the structural transformations processes during thermal, chemical-thermal, thermomechanical, and other types of alloys treatment. The vacancy formation energy is one of the most important parameters used to describe diffusion processes. An effective approach to its definition is based on the use of the density functional theory (DFT). The main advantage of this method is to carry out computations without any parameters defined empirically. The purpose of the work is to estimate vacancy formation energy of BCC-, FCC- and HCP-metals widely used in mechanical engineering and to compare these findings obtained using various exchange-correlation functionals (GGA and meta-GGA). Computation procedure. The computations were carried out using the projector-augmented wave method using the GPAW code and the atomic simulation environment (ASE). The Perdew-Burke-Ernzerhof, MGGAC and rMGGAC functionals were used. The wave functions were described by plane waves within simulations. Vacancies formation energy was evaluated using supercells approach with a size 3 × 3 × 3. Computations were carried out for BCC-metals (Li, Na, K, V, Cr, Fe, Rb, Nb, Mo, Cs, Ta, W), FCC-metals (Al, Ni, Cu, Rh, Pd, Ag, Ir, Pt, Au, Pb, Co) and HCP-metals (Be, Ti, Zr, Mg, Sc, Zn, Y, Ru, Cd, Hf, Os, Co, Re). Results and discussion. A comparison of the defined vacancy formation energies indicates the validity of the following ratio of values: . The values obtained using the open source GPAW code are characterized by the same patterns as for widely spread commercially distributed program VASP. It was revealed that the use of the PBE and MGGAC functionals leads to a slight deviation relative to the experimentally determined vacancies formation energy in contrast to the computations using rMGGAC.

Fortnet, a software package for training Behler-Parrinello neural networks

A new, open source, parallel, stand-alone software package (Fortnet) has been developed, which implements Behler-Parrinello neural networks. It covers the entire workflow from feature generation to the evaluation of generated potentials, coupled with higher-level analysis such as the analytic calculation of atomic forces. The functionality of the software package is demonstrated by driving the training for the fitted correction functions of the density functional tight binding (DFTB) method, which are commonly used to compensate the inaccuracies resulting from the DFTB approximations to the Kohn-Sham Hamiltonian. The usual two-body form of those correction functions limits the transferability of the parametrizations between very different structural environments. The recently introduced DFTB+ANN approach strives to lift these limitations by combining DFTB with a near-sighted artificial neural network (ANN). After investigating various approaches, we have found the combination of DFTB with an ANN acting on-top of some baseline correction functions (delta learning) the most accurate one. It allowed to introduce many-body corrections on top of two-body parametrizations, while excellent transferability to chemical environments with deviating energetics could be demonstrated. Program summaryProgram title: FortnetCPC Library link to program files:https://doi.org/10.17632/sjg3n9vr8p.1Developer's repository link:https://github.com/vanderhe/fortnetCode Ocean capsule:https://codeocean.com/capsule/3992747Licensing provisions: LGPLProgramming language: Fortran, PythonExternal routines/libraries: MPI, BLAS/LAPACK, HDF5, DFTB+Supplementary material: See supplementary material for exemplary Human-friendly Structured Data (HSD) input listings, as well as the basic usage of the Fortformat Python layer for generating datasets and extracting results.Nature of problem: Semi-empirical quantum mechanical methods like density functional tight binding (DFTB) rely on fitting empirical energy correction terms, often represented by two-body potentials, to ab initio references. Hereby empirical, beyond-pairwise contributions are inevitably incorporated and therefore inadequately covered by a purely two-body description.Solution method: The new, open source, parallel, stand-alone software package Fortnet provides a powerful, yet accessible tool to construct many-body correction terms by resorting to high-dimensional neural networks of Behler-Parrinello type. Fortnet is characterized by its modern infrastructure, complementing the landscape of available implementations by a robust combination of Fortran and Python based code.Additional comments including restrictions and unusual features: Fortnet's core is supplemented by two additional projects that are BSD 2-clause licensed, namely fortnet-python [2], a collection of Python based tools for generating compatible datasets and extracting results, and fortnet-ase [3], an interface to the Atomic Simulation Environment (ASE) [1]. Both projects are available via the Python Package Index (PyPI). The interaction of all components is explained in cookbook-like recipes (see: https://fortnet.readthedocs.io/en/latest/), meant to guide new users, while learning about various basic and more advanced features by using comprehensible examples with physical reference.

The performance of ANI-ML potentials for ligand-n(H2 O) interaction energies and estimation of hydration free energies from end-point MD simulations.

Here, we investigate the performance of "Accurate NeurAl networK engINe for Molecular Energies" (ANI), trained on small organic compounds, on bulk systems including non-covalent interactions and applicability to estimate solvation (hydration) free energies using the interaction between the ligand and explicit solvent (water) from single-step MD simulations. The method is adopted from ANI using the Atomic Simulation Environment (ASE) and predicts the non-covalent interaction energies at the accuracy of wb97x/6-31G(d) level by a simple linear scaling for the conformations sampled by molecular dynamics (MD) simulations of ligand-n(H2 O) systems. For the first time, we test ANI potentials' abilities to reproduce solvation free energies using linear interaction energy (LIE) formulism by modifying the original LIE equation. Our results on ~250 different complexes show that the method can be accurate and have a correlation of R2 = 0.88-0.89 (MAE <1.0kcal/mol) to the experimental solvation free energies, outperforming current end-state methods. Moreover, it is competitive to other conventional free energy methods such as FEP and BAR with 15-20 × fold reduced computational cost.

Accurate Binding Free Energy Method from End-State MD Simulations.

Herein, we introduce a new strategy to estimate binding free energies using end-state molecular dynamics simulation trajectories. The method is adopted from linear interaction energy (LIE) and ANI-2x neural network potentials (machine learning) for the atomic simulation environment (ASE). It predicts the single-point interaction energies between ligand–protein and ligand–solvent pairs at the accuracy of the wb97x/6-31G* level for the conformational space that is sampled by molecular dynamics (MD) simulations. Our results on 54 protein–ligand complexes show that the method can be accurate and have a correlation of R = 0.87–0.88 to the experimental binding free energies, outperforming current end-state methods with reduced computational cost. The method also allows us to compare BFEs of ligands with different scaffolds. The code is available free of charge (documentation and test files) at https://github.com/otayfuroglu/deepQM.

Ab initio metadynamics determination of temperature-dependent free-energy landscape in ultrasmall silver clusters.

Ab initio metadynamics enables the extraction of free-energy landscapes having the accuracy of first-principles electronic structure methods. We introduce an interface between the PLUMED code that computes free-energy landscapes and enhanced-sampling algorithms and the Atomic Simulation Environment (ASE) module, which includes several ab initio electronic structure codes. The interface is validated with a Lennard-Jones cluster free-energy landscape calculation by averaging multiple short metadynamics trajectories. We use this interface and analysis to estimate the free-energy landscape of Ag5 and Ag6 clusters at 10, 100, and 300K with the radius of gyration and coordination number as collective variables, finding at most tens of meV in error. Relative free-energy differences between the planar and non-planar isomers of both clusters decrease with temperature in agreement with previously proposed stabilization of non-planar isomers. Interestingly, we find that Ag6 is the smallest silver cluster where entropic effects at room temperature boost the non-planar isomer probability to a competing state. The new ASE-PLUMED interface enables simulating nanosystem electronic properties under more realistic temperature-dependent conditions.

Two-Dimensional Metal–Organic Frameworks with Unique Oriented Layers for Oxygen Reduction Reaction: Tailoring the Activity through Exposed Crystal Facets

Two-Dimensional Metal–Organic Frameworks with Unique Oriented Layers for Oxygen Reduction Reaction: Tailoring the Activity through Exposed Crystal Facets

Gpaw-tools – higher-level user interaction scripts for GPAW calculations and interatomic potential based structure optimization

GPAW is a density-functional theory (DFT) code based on the projector-augmented wave (PAW) method written mainly in Python. It has high integration with atomic simulation environment (ASE) software and high interoperability with the Python language. However, this high interoperability requires Python language knowledge and therefore limits the user numbers of GPAW worldwide. For an end-user, who is not interested in the Python language itself and interested in pure material science, gpaw-tools scripts are free cross-platform Python scripts that aim to be a user interface (UI) and a Tcl/Tk based graphical UI (GUI) software. With gpaw-tools, the end-user can calculate, view, and save density of states (DOS), band structure, charge densities, and optical properties of the investigated structure without the need for Python language knowledge. The gpaw-tools scripts are written in Python language and distributed with an MIT license.

Spread-balanced Wannier functions: Robust and automatable orbital localization

We introduce a new type of Wannier functions (WFs) obtained by minimizing the conventional spread functional with a penalty term proportional to the variance of the spread distribution. This modified Wannierisation scheme is less prone to produce ineffective solutions featuring one or several poorly localized orbitals, making it well suited for complex systems or high-throughput applications. Furthermore, we propose an automatable protocol for selecting the initial guess and determine the optimal number of bands (or equivalently WFs) for the localization algorithm. The improved performance and robustness of the approach is demonstrated for a diverse set of test systems including the NV center in diamond, metal slabs with atomic adsorbates, spontaneous polarization of ferroelectrics and 30 inorganic monolayer materials comprising both metals and semiconductors. The methods are implemented in Python as part of the Atomic Simulation Environment (ASE).

Davis Computational Spectroscopy Workflow-From Structure to Spectra.

We describe an automated workflow that connects a series of atomic simulation tools to investigate the relationship between atomic structure, lattice dynamics, materials properties, and inelastic neutron scattering (INS) spectra. Starting from the atomic simulation environment (ASE) as an interface, we demonstrate the use of a selection of calculators, including density functional theory (DFT) and density functional tight binding (DFTB), to optimize the structures and calculate interatomic force constants. We present the use of our workflow to compute the phonon frequencies and eigenvectors, which are required to accurately simulate the INS spectra in crystalline solids like diamond and graphite as well as molecular solids like rubrene. We have also implemented a machine-learning force field based on Chebyshev polynomials called the Chebyshev interaction model for efficient simulation (ChIMES) to improve the accuracy of the DFTB simulations. We then explore the transferability of our DFTB/ChIMES models by comparing simulations derived from different training sets. We show that DFTB/ChIMES demonstrates ∼100× reduction in computational expense while retaining most of the accuracy of DFT as well as yielding high accuracy for different materials outside of our training sets. The DFTB/ChIMES method within the workflow expands the possibilities to use simulations to accurately predict materials properties of increasingly complex structures that would be unfeasible with ab initio methods.

Energy contour exploration with potentiostatic kinematics

We introduce a method of exploring potential energy contours (PECs) in complex dynamical systems based on potentiostatic kinematics wherein the systems are evolved with minimal changes to their potential energy. We construct a simple iterative algorithm for performing potentiostatic kinematics, which uses an estimate curvature to predict new configuration-space coordinates on the PEC and a potentiostat term component to correct for errors in prediction. Our methods are then applied to atomic structure models using an interatomic potential for energy and force evaluations as would commonly be invoked in a molecular dynamics simulation. Using several model systems, we assess the stability and accuracy of the method on different hyperparameters in the implementation of the potentiostatic kinematics. Our implementation is open source and available within the atomic simulation environment package.

DJMol: An open-source modeling platform for computational chemistry and materials science with a Python interpreter.

We present a modular and extendable software suite, DJMol, for performing molecular simulations and it is demonstrated with DFTB+, Siesta, Atomic Simulation Environment, and OpenMD codes. It supports many of the standard features of an integrated development environment and consists of a structure builder and viewer, which could be connected with these electronic structure codes along with a set of data analyzers. This program comprises Java and Python modules and its libraries to carry out a different set of modeling tasks in materials science and chemistry. By adopting a Python interpreter into the software, a range of scriptable Python codes, such as Pymatgen can be incorporated into this programmable modeling platform. DJMol, through its common application programming interface (API), supports multiple modeling codes in the backend and several post-processing tools. It benefits an experienced user by increasing efficiency, while a nonexpert user by easy to use API.

Atomic Simulation Recipes: A Python framework and library for automated workflows

The Atomic Simulation Recipes (ASR) is an open source Python framework for working with atomistic materials simulations in an efficient and sustainable way that is ideally suited for high-throughput projects. Central to ASR is the concept of a Recipe: a high-level Python script that performs a well defined simulation task robustly and accurately while keeping track of the data provenance. The ASR leverages the functionality of the Atomic Simulation Environment (ASE) to interface with external simulation codes and attain a high abstraction level. We provide a library of Recipes for common simulation tasks employing density functional theory and many-body perturbation schemes. These Recipes utilize the GPAW electronic structure code, but may be adapted to other simulation codes with an ASE interface. Being independent objects with automatic data provenance control, Recipes can be freely combined through Python scripting giving maximal freedom for users to build advanced workflows. ASR also implements a command line interface that can be used to run Recipes and inspect results. The ASR Migration module helps users maintain their data while the Database and App modules makes it possible to create local databases and present them as customized web pages.

Machine-learning accelerated geometry optimization in molecular simulation.

Geometry optimization is an important part of both computational materials and surface science because it is the path to finding ground state atomic structures and reaction pathways. These properties are used in the estimation of thermodynamic and kinetic properties of molecular and crystal structures. This process is slow at the quantum level of theory because it involves an iterative calculation of forces using quantum chemical codes such as density functional theory (DFT), which are computationally expensive and which limit the speed of the optimization algorithms. It would be highly advantageous to accelerate this process because then one could do either the same amount of work in less time or more work in the same time. In this work, we provide a neural network (NN) ensemble based active learning method to accelerate the local geometry optimization for multiple configurations simultaneously. We illustrate the acceleration on several case studies including bare metal surfaces, surfaces with adsorbates, and nudged elastic band for two reactions. In all cases, the accelerated method requires fewer DFT calculations than the standard method. In addition, we provide an Atomic Simulation Environment (ASE)-optimizer Python package to make the usage of the NN ensemble active learning for geometry optimization easier.