Related Experiment Video
Updated: Nov 11, 2025

Computation of Atmospheric Concentrations of Molecular Clusters from ab initio Thermochemistry
Published on: April 8, 2020
Arbitrary-Order Derivatives of Quantum Chemical Methods via Automatic Differentiation
Adam S Abbott1, Boyi Z Abbott1, Justin M Turney1
1Center for Computational Quantum Chemistry, University of Georgia, Athens, Georgia 30602, United States.
This study introduces a new method for calculating exact nuclear coordinate derivatives in quantum chemistry. The approach uses automatic differentiation, overcoming limitations of numerical methods and complex analytic formulas.
Area of Science:
- Computational Chemistry
- Quantum Chemistry
- Theoretical Chemistry
Background:
- Calculating higher-order nuclear coordinate derivatives of electronic energies is crucial for understanding molecular properties and reaction dynamics.
- Traditional methods rely on numerical differentiation, which suffers from computational instability, or complex analytic derivative formulas that are difficult to derive for advanced quantum chemistry methods.
Purpose of the Study:
- To present a general and robust methodology for obtaining arbitrary-order nuclear coordinate derivatives of electronic energies.
- To bypass the limitations of numerical differentiation and the complexity of deriving analytic formulas for energy derivatives.
Main Methods:
- Leveraging modern automatic differentiation (AD) software to compute exact nuclear coordinate derivatives.
- Implementing the AD scheme in a freely available, open-source software package.
- Applying the methodology to various quantum chemistry methods, including Hartree-Fock, Møller-Plesset perturbation theory (MP2), and coupled cluster theory with single, double, and perturbative triple excitations [CCSD(T)].
Main Results:
- Demonstrated the ability to obtain exact arbitrary-order nuclear derivatives for electronic energies across different quantum chemistry methods.
- Successfully computed up to sextic derivatives for various test systems, showcasing the method's broad applicability.
- Achieved results that were previously unobtainable by exact means due to the lack of higher-order derivative formulas.
Conclusions:
- The developed automatic differentiation methodology provides a general, stable, and efficient approach for calculating high-order nuclear derivatives.
- This innovation significantly advances computational capabilities in quantum chemistry, enabling new avenues for theoretical studies.
- The open-source implementation facilitates wider adoption and further development of exact derivative calculations.
More Related Videos
Related Concept Videos
Second Derivatives and Laplace Operator
Consider a scalar function. The curl of its...
Maxwell's Thermodynamic Relations
All thermodynamic potentials are exact differentials. Therefore, their second-order...
Fundamental Mathematical Principles in Pharmacokinetics: Calculus and Graphs
On the other hand, integral calculus focuses on...
The Integrated Rate Law: The Dependence of Concentration on Time
The Quantum-Mechanical Model of an Atom
Fundamental Mathematical Principles in Pharmacokinetics: Rate and Order of Reaction
Pharmacokinetic reactions...

