Publication | Closed Access
Comparison of the performance of local, gradient-corrected, and hybrid density functional models in predicting infrared intensities
171
Citations
35
References
1998
Year
EngineeringInfrared IntensitiesComputational ChemistryChemistryHybrid FunctionalsSpectra-structure CorrelationData ScienceBiostatisticsInfrared OpticThermal Infrared Remote SensingStatisticsBiophysicsDifferent FunctionalsChemical BondInfrared SpectroscopyNear-infrared SpectroscopyComputational ModelingQuantum ChemistryMolecular ChemistryFunctional Data AnalysisInfrared SensorNatural SciencesSpectroscopyRemote Sensing
Density functional theory has been used to calculate infrared (IR) intensities for a series of molecules (HF, CO, H2O, HCN, CO2, C2H2, H2CO, NH3, C2H4, CH2F2, CH2Cl2, and CH2Br2) in an effort to evaluate relative performance of different functionals. The methods employed in this study comprise most of the popular local, gradient-corrected, and hybrid functionals, namely, S-VWN, S-PL, B-LYP, B-P86, B-PW91, B3-LYP, B3-P86, and B3-PW91. Calculations were carried out using various qualities of split valence basis sets augmented with diffuse and polarization functions, both to determine basis set dependence and to evaluate the limit performance. Computed intensities were compared with results from conventional correlated ab initio methods (MP2 and QCISD). Hybrid functionals give results in closest agreement with QCISD over the other methods surveyed. Local and gradient-corrected methods performed remarkably alike, both are comparable to MP2, and outperform Hartree–Fock. Hartree–Fock intensities can be dramatically improved by scaling, making them similar to MP2 results.
| Year | Citations | |
|---|---|---|
Page 1
Page 1