Reaction and Property Prediction
Computational chemistry's payoff is prediction: of how reactions proceed, how fast, and of the structural, electronic, and spectroscopic properties of molecules before they are made or measured.
Definition
The collected aim and methods of computational chemistry directed at predicting reaction behavior and molecular properties for use in interpretation and design.
Scope
Covers the computational characterization of reaction mechanisms and kinetics through transition states and reaction paths, the estimation of molecular properties and biological activity through quantitative structure-property and structure-activity relationships, the treatment of solvent effects with continuum and explicit models, and the prediction of spectra. Applies the methods of the other areas to chemical questions.
Sub-topics
Core questions
- How are reaction mechanisms, barriers, and rates obtained from calculations?
- How are properties and activities predicted from molecular structure?
- How are solvent and environmental effects incorporated?
- How are spectra computed and used to interpret experiment?
Key theories
- Transition-state theory
- Relates reaction rates to the properties of the transition state on the potential energy surface, providing the framework that connects computed barriers to kinetics.
- Structure-property relationships
- Properties and activities of molecules can be predicted from features of their structure, whether through physically computed quantities or statistical models, enabling property-driven design.
Clinical relevance
Predicting reactivity, properties, and spectra computationally guides synthesis planning, catalyst and material design, environmental fate assessment, and the interpretation of experimental measurements across chemistry.
History
Rooted in Eyring's transition-state theory and Hansch's quantitative structure-activity relationships, computational prediction grew with reliable electronic-structure and solvation methods, becoming a routine partner to experiment in chemistry and drug discovery.
Key figures
- Henry Eyring
- Christopher Cramer
- Donald Truhlar
- Corwin Hansch
Related topics
Seminal works
- cramer2004
- jensen2017
Frequently asked questions
- How accurate are computational predictions of reactions?
- Accuracy depends heavily on the method and system; barriers from high-level methods can be very reliable for small molecules, while large or strongly correlated systems carry larger and less predictable uncertainties.
- What kinds of properties can be predicted computationally?
- These include geometries, energies and barriers, dipole moments, acidities, redox potentials, solubilities, spectra, and biological activities, using physical calculations, statistical models, or both.