Ab Initio Thermochemistry Beyond Chemical Accuracy for First-and Second-Row Compounds

Martin, Jan M. L.

doi:10.1007/978-94-011-4671-5_17

Cited by 15 publications

(13 citation statements)

References 152 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By separately fitting the self consistent field ͑SCF͒ and correlation energies, Martin has been able to achieve a ⑀ MAD of 0.12 kcal/mol for 15 first row molecules. 95 However, it was not clear if scalar relativistic corrections were considered. Given the size of the ⌬E SR correction, the potential residual error in the CCSD͑T͒ treatment and the uncertainty in the experimental measurements, a mean absolute deviation of less than 0.5 kcal/mol for larger collections of molecules will be hard to maintain.…”

Section: Discussionmentioning

confidence: 99%

Re-examination of atomization energies for the Gaussian-2 set of molecules

Feller

Peterson

1999

The Journal of Chemical Physics

270

268

View full text Add to dashboard Cite

Atomization energies were computed for 73 molecules, many of them chosen from the GAUSSIAN-2 and G2/97 test sets. A composite theoretical approach was adopted which incorporated estimated complete basis set binding energies based on frozen core coupled cluster theory with a quasiperturbative treatment of triple excitations and three corrections: (1) a coupled cluster core/valence correction; (2) a configuration interaction scalar relativistic correction; and (3) an atomic spin-orbital correction. A fourth correction, corresponding to more extensive correlation recovery via coupled cluster theory with an approximate treatment of quadruple excitations, was examined in a limited number of cases. For the molecules and basis sets considered in this study, failure to consider any of these contributions to the atomization energy can introduce errors on the order of 1–2 kcal/mol. Although some cancellation of error is common, it is by no means universal and cannot be relied upon for high accuracy. With the largest available basis sets (including, in some cases, up through aug-cc-pV6Z), the mean absolute deviation with respect to experiment was found to lie in the 0.7–0.8 kcal/mol range, neglecting the effects of higher order excitations. Worst case errors were 2–3 kcal/mol. Several complete basis set extrapolations were tested with regard to their effectiveness at improving agreement with experiment, but the statistical difference among the various approaches was small.

show abstract

Section: Discussionmentioning

confidence: 99%

Re-examination of atomization energies for the Gaussian-2 set of molecules

Feller

Peterson

1999

The Journal of Chemical Physics

270

268

View full text Add to dashboard Cite

show abstract

“…67 A similar approach, albeit without the compilation of such an extensive database, has been pursued by several other laboratories around the world. [68][69][70][71][72][73][74] In the present work, we apply our composite approach based on CCSD(T)/CBS to determine the heats of formation of seven small organic molecules. The availability of reliable experimental data for CO, CO 2 , and HC(O)OH allows them to serve as convenient benchmarks of the accuracy that might be expected for HOCO, HCO 2 , and HC(O)OOH, for which no reliable experimental measurements are available.…”

Section: 4 Kcal/molmentioning

confidence: 99%

“…The performance of our approach, as measured by its ability to reproduce reliable experimental atomization energies, has been benchmarked with the help of the information on 273 molecules contained in the Environmental and Molecular Sciences Laboratory Computational Results Database . A similar approach, albeit without the compilation of such an extensive database, has been pursued by several other laboratories around the world. − …”

Section: Introductionmentioning

confidence: 99%

Coupled Cluster Theory Determination of the Heats of Formation of Combustion-Related Compounds: CO, HCO, CO₂, HCO₂, HOCO, HC(O)OH, and HC(O)OOH

2003

View full text Add to dashboard Cite

Coupled cluster theory through quasiperturbative, connected triple excitations was used to obtain optimized structures, harmonic vibrational frequencies, and heats of formation for seven small molecules important to hydrocarbon oxidation. For the three systems possessing reliable experimental heats of formation, the level of agreement between theory and experiment was excellent. To achieve this level of agreement and to simultaneously minimize the theoretical uncertainty, it was necessary to apply large correlation consistent basis sets (through septuple ζ in some cases) followed by a number of small, but nonnegligible, energetic corrections. For CO, ΔH f 0(0 K) = −27.0 ± 0.2 (theory) versus −27.20 ± 0.04 kcal/mol (expt). For CO2, ΔH f 0(0 K) = −93.7 ± 0.2 (theory) versus −93.97 ± 0.01 kcal/mol (expt). For HC(O)OH (formic acid), ΔH f 0(0 K) = −88.9 ± 0.4 (theory) versus −88.7 ± 0.1 kcal/mol (expt). For HCO, the experimental and theoretical values are in near perfect agreement, with ΔH f 0(0 K) = 10.4 ± 0.2 (theory) versus 10.3 ± 2 kcal/mol (expt), although this may be somewhat fortuitous because the experimental value has a large uncertainty. For trans-HOCO, we predict a value of ΔH f 0(0 K) = −43.9 ± 0.5 kcal/mol, compared to the revised photoionization value of ≥−45.8 ± 0.7 kcal/mol. Theory, however, is in good agreement with the possible experimental value of −42.7 ± 0.9 kcal/mol suggested in the same photoionization experimental analysis. For HCO2, theory predicts a value of ΔH f 0(0 K) = −29.3 ± 0.4 versus −30 ± 3 kcal/mol for a recent negative-ion photoelectron measurement. For HC(O)OOH, in which case no experimental data exists, ΔH f 0(0 K) = −65.6 ± 0.6 kcal/mol. trans-HOCO is only slightly bound (1.1 kcal/mol) with respect to the H + CO2 asymptote. HCO2 is 15.7 kcal/mol higher in energy than trans-HOCO and lies above the H + CO2 asymptote by 14.6 kcal/mol. It is only bound with respect to the OH + CO asymptote by 9.0 kcal/mol. Three widely used parametrized methods (G2, G3, and CBS-Q) were compared to the best coupled cluster heats of formation and found to differ by up to 3.2 kcal/mol.

show abstract

“…The results are collected in Table . It is seen that, after extrapolation to infinite basis set, the result for the enthalpy of formation of HC(O)OH is only 1.2 kJ/mol apart from the experimental value. Correcting for this error the enthalpy of formation obtained from reaction 5 for HC(O)SH with R = Me, we obtained a value of −123.1 kJ/mol, essentially in agreement with the B3LYP calculation.…”

Section: Resultsmentioning

confidence: 74%

Density Functional Computational Thermochemistry: Isomerization of Sulfine and Its Enthalpy of Formation

et al. 2001

View full text Add to dashboard Cite

Density functional (DFT), second-order perturbation theory (MP2), and coupled-cluster [(CCSD(T)] calculations using Pople's basis sets up to 6-311++G(3df,2pd) and Dunning's correlation consistent basis sets have been employed to determine the enthalpy of formation of sulfine, CH2SO, 1, using the isodesmic reaction CH2S + SO2 ⇌ CH2SO + SO. Previous calculations showed an inconsistency between the enthalpy of formation obtained using this methodology, Δf H o 298.15(1) = −52 ± 10 kJ/mol, and the value obtained employing the isomerization reaction CH2SO (1) ⇌ HC(O)SH (2) if Benson's estimate for the enthalpy of formation of isomer 2 (thioformic acid) was employed. This result was particularly vexing, since the computed enthalpy of formation of 1 was reasonably in agreement with Benson's own estimate. In this paper we extended our previous study using the reactions HC(O)−XH + RH ⇌ H2CO + R−XH with R = H, Me, Et, Pr, and i-Pr. X was either sulfur, to obtain the enthalpy of formation of 2, or oxygen, to assess the errors to be expected in the use of these reactions for the evaluation of Δf H°. The result, Δf H o 298.15(2) = −121 ± 8 kJ/mol, arrived at after a critical assessment of B3LYP, MP2, and CCDSD(T) results, is in complete agreement with the value of −126 ± 4 kJ/mol estimated by Benson. This implies that the isomerization reaction cannot be employed for the determination of the enthalpy of formation of sulfine. We ascribe this inadequacy to the errors introduced due to the change in the oxidation state of sulfur.

show abstract

Ab Initio Thermochemistry Beyond Chemical Accuracy for First-and Second-Row Compounds

Cited by 15 publications

References 152 publications

Re-examination of atomization energies for the Gaussian-2 set of molecules

Re-examination of atomization energies for the Gaussian-2 set of molecules

Coupled Cluster Theory Determination of the Heats of Formation of Combustion-Related Compounds: CO, HCO, CO₂, HCO₂, HOCO, HC(O)OH, and HC(O)OOH

Density Functional Computational Thermochemistry: Isomerization of Sulfine and Its Enthalpy of Formation

Contact Info

Product

Resources

About

Ab Initio Thermochemistry Beyond Chemical Accuracy for First-and Second-Row Compounds

Cited by 15 publications

References 152 publications

Re-examination of atomization energies for the Gaussian-2 set of molecules

Re-examination of atomization energies for the Gaussian-2 set of molecules

Coupled Cluster Theory Determination of the Heats of Formation of Combustion-Related Compounds: CO, HCO, CO2, HCO2, HOCO, HC(O)OH, and HC(O)OOH

Density Functional Computational Thermochemistry: Isomerization of Sulfine and Its Enthalpy of Formation

Contact Info

Product

Resources

About

Coupled Cluster Theory Determination of the Heats of Formation of Combustion-Related Compounds: CO, HCO, CO₂, HCO₂, HOCO, HC(O)OH, and HC(O)OOH