Heats of formation of platonic hydrocarbon cages by means of high-level thermochemical procedures

• Published in: Journal of computational chemistry Vol. 37; no. 1; pp. 49 - 58
• Main Authors: Karton, Amir, Schreiner, Peter R., Martin, Jan M. L.
• Format: Journal Article
• Language: English
• Published: United States Blackwell Publishing Ltd 05.01.2016; Wiley Subscription Services, Inc
• Subjects: Atomizing; Benchmarks; Cages; CCSD(T); Chemistry; computational thermochemistry; Correlation; Cubane; Data bases; Density; Density functional theory; explicitly correlated coupled cluster; Heat of formation; Heats (energies); Hydrocarbons; Level (quantity); Mechanics (physics); Parameterization; Performance evaluation; platonic hydrocarbon cages; prismatic hydrocarbon cages; Reference materials; Thermodynamics; platonic hydrocarbon cages; CCSD(T); computational thermochemistry; explicitly correlated coupled cluster; prismatic hydrocarbon cages
• ISSN: 0192-8651; 1096-987X
• DOI: 10.1002/jcc.23963

Hydrocarbon cages are key reference materials for the validation and parameterization of computationally cost‐effective procedures such as density functional theory (DFT), semiempirical molecular orbital theory, and molecular mechanics. We obtain accurate total atomization energies (TAEs) and heats of formation (ΔfH°298) for platonic and prismatic hydrocarbon cages by means of the Wn‐F12 explicitly correlated thermochemical protocols. We consider the following kinetically stable (CH)n polycyclic hydrocarbon cages: (i) platonic hydrocarbons (tetrahedrane, cubane, and dodecahedrane), (ii) prismatic hydrocarbons (triprismane, cubane, and pentaprismane), and (iii) one truncated tetrahedrane (octahedrane). Our best theoretical heat of formation for cubane (144.8 kcal mol−1) suggests that the experimental value adopted by the NIST thermochemical database (142.7 ± 1.2 kcal mol−1) should be revised upwards by ∼2 kcal mol−1. Our best heat of formation for dodecahedrane (20.2 kcal mol−1) suggests that the semiexperimental value (22.4 ± 1 kcal mol−1) should be revised downward by ∼2 kcal mol−1. We use our benchmark Wn‐F12 TAEs to evaluate the performance of a variety of computationally less demanding composite thermochemical procedures. These include the Gaussian‐n (Gn) and the complete basis set (CBS) methods. The CBS‐QB3 and CBS‐APNO procedures show relatively poor performance with root‐mean‐squared deviations (RMSDs) of 4.2 and 2.5 kcal mol−1, respectively. The best performers of the Gn procedures are G4 and G3(MP2)B3 (RMSD = 0.5 and 0.6 kcal mol−1, respectively), while the worst performers are G3 and G4(MP2)‐6X (RMSD = 2.1 and 2.9 kcal mol−1, respectively). Isodesmic and even homodesmotic reactions involving these species are surprisingly challenging targets for DFT computations. © 2015 Wiley Periodicals, Inc. This work determines total atomization energies and heats of formation for platonic and prismatic polycyclic hydrocarbon cages by means of the W1‐F12 and W2‐F12 thermochemical protocols. Using these accurate reference data, the performance of computationally economical theoretical methods (e.g., density functional theory and composite ab initio methods) was evaluated via atomization and bond separation reactions for the calculation of these challenging thermochemical quantities.