Publication Date:
2011-04-08
Description:
Isomerization energies for hexenes (C 6 H 12 ) were evaluated with ab initio (Hartree–Fock (HF), MP2, SCS-MP2, and CCSD(T)) and several density functional approximation (DFA) methods. CCSD(T)/6-311+G(2d,p) energies were taken as a benchmark standard. The HF method incorrectly predicts that monosubstituted alkenes are more stable than multiply-substituted alkenes. DFAs generally predict the correct stability trends of alkenes (mono-, 〈 di-, 〈 tri-, 〈 tetra-substituted alkenes) but errors in popular functionals, such as B3LYP, can be as large as errors found for alkane hydrocarbon thermochemistries. Some of the HF error is traced back to deficiencies in modeling 1,3-geminal and 1,4-vicinal alkyl–alkyl group interactions, called vinylbranches, and changes in C C and C H bond types (sp 3 –sp 2 C C to sp 3 –sp 3 C C and sp 3 C H to sp 2 C H). The latter is shown to be more significant. Comparison of CCSD(T) energies of trans -2-butene with 2-methylpropylene and cis -2-butene suggests that geminal vinylbranches are stabilizing while vicinal vinylbranches are destabilizing. B3LYP and other DFAs have much smaller errors than HF theory due to inclusion of correlation energy that better reproduces bond type changes. Copyright © 2011 John Wiley & Sons, Ltd. Accurate estimation of electron correlation is necessary to qualitatively and quantitatively reproduce alkene isomerization energies. Hartree–Fock and density functional approximation errors arise from C C and C H bond type changes, geminal and vicinal 1,3-alkyl–alkyl vinylbranch interactions, and alkane protobranch interactions.
Print ISSN:
0894-3230
Electronic ISSN:
1099-1395
Topics:
Chemistry and Pharmacology
,
Physics
Permalink