Computational Quantum Chemistry Techniques

Computational quantum chemistry encompasses a suite of theoretical and numerical methods for predicting electronic structure, molecular properties and reaction mechanisms from first principles. Central approaches include wavefunction-based methods, such as Hartree–Fock theory and post-Hartree–Fock techniques (for example coupled cluster and multireference configuration interaction), which explicitly construct electronic wavefunctions to capture electron correlation. Density functional theory offers an alternative by expressing the energy in terms of the electron density and exchange-correlation functionals, enabling routine treatment of larger systems. Semi-empirical methods introduce parameterisation to simplify the Hamiltonian, while composite schemes combine calculations at multiple levels of theory to balance accuracy and efficiency. More recently, machine-learning potentials trained on extensive quantum chemical data sets have emerged to accelerate the prediction of molecular energies, forces and properties with near-ab initio accuracy. Hybrid multiscale techniques couple quantum mechanical regions to classical or continuum models, facilitating studies of chemical reactivity in complex environments. Collectively, these computational tools underpin advances in catalyst design, materials discovery and the interpretation of spectroscopic experiments by providing molecular-scale insight that complements laboratory investigations.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

Large-scale data infrastructures have transformed method development and model training. A recent extension of a benchmark data set introduced energies for a set of small organic molecules calculated with dozens of density functionals and basis sets, alongside reaction energies and bond-change annotations, thus supporting multitask learning and the development of delta-learning approaches. An independent review has surveyed the landscape of quantum chemistry databases for machine-learning potentials, analysing the levels of theory employed, chemical diversity and FAIR-compliance efforts, and calling for standardised, user-friendly platforms to ensure long-term interoperability. In parallel, a web-accessible repository of over 170 000 density-functional theory calculations covering reactive, open-shell and charged species has been integrated into a broader materials project, providing structural, electronic, vibrational and thermodynamic properties via a RESTful API. These developments illustrate a concerted move towards comprehensive, interconnected data resources that underpin innovation in both traditional electronic-structure methods and data-driven approaches.

Computational Quantum Chemistry Techniques publication trend

The graph below shows the total number of articles in computational quantum chemistry techniques across all publications each year (not limited to Nature Index journals).

Technical terms

Density functional theory (DFT): A quantum mechanical method that expresses total energy as a functional of the electron density, using approximate exchange-correlation functionals to account for many-body effects.

Wavefunction: A mathematical function describing the quantum state of electrons in a molecule, from which observable properties are derived via the Schrödinger equation.

Ab initio methods: First-principles approaches that solve the electronic Schrödinger equation without empirical parameters, including Hartree–Fock and post-Hartree–Fock techniques that systematically improve electron correlation.

Basis set: A set of mathematical functions used to expand molecular orbitals in quantum chemical calculations, with larger or more flexible basis sets yielding higher accuracy at increased computational cost.

Machine-learning potentials: Data-driven models trained on quantum chemical results to predict energies, forces and properties efficiently, bridging the gap between accuracy and computational speed.

References

1. MultiXC-QM9: Large dataset of molecular and reaction energies from multi-level quantum chemical methods. Scientific Data (2023).
2. A database of molecular properties integrated in the Materials Project. Digital Discovery (2023).
3. Molecular quantum chemical data sets and databases for machine learning potentials. Machine Learning: Science and Technology (2024).