Accurate prediction of the enthalpies of formation for xanthophylls

Abstract

This study investigates the applications of computational approaches in the prediction of enthalpies of formation (ΔH(f)) for C-, H-, and O-containing compounds. Molecular mechanics (MM4) molecular mechanics method, density functional theory (DFT) combined with the atomic equivalent (AE) and group equivalent (GE) schemes, and DFT-based correlation corrected atomization (CCAZ) were used. We emphasized on the application to xanthophylls, C-, H-, and O-containing carotenoids which consist of ∼ 100 atoms and extended π-delocaization systems. Within the training set, MM4 predictions are more accurate than those obtained using AE and GE; however a systematic underestimation was observed in the extended systems. ΔH(f) for the training set molecules predicted by CCAZ combined with DFT are in very good agreement with the G3 results. The average absolute deviations (AADs) of CCAZ combined with B3LYP and MPWB1K are 0.38 and 0.53 kcal/mol compared with the G3 data, and are 0.74 and 0.69 kcal/mol compared with the available experimental data, respectively. Consistency of the CCAZ approach for the selected xanthophylls is revealed by the AAD of 2.68 kcal/mol between B3LYP-CCAZ and MPWB1K-CCAZ.

Atom Pair Contribution Method: Fast and General Procedure To Predict Molecular Formation Enthalpies

“…Accordingly, a wealth of procedures to predict this quantity have been reported, covering a wide spectrum from simple additivity schemes to high-level composite ab initio methods . Between these two ends, other prominent approaches include quantitative structure–property relationship (QSPR) methodologies, − molecular mechanics (MM) force fields, − semiempirical quantum-chemical methods (SQM), − procedures based on Hartree–Fock (HF) or density functional theory (DFT) combined with empirical corrections or isodesmic/isodesmotic reaction schemes, , and a variety of alternative hybrid procedures. − As shown in Figure , these approaches can be loosely classified on the rungs of a Jacob’s ladder according to their computational requirements.…”