Systematic sequences of well-balanced Gaussian basis sets

1994 ◽  
Vol 72 (7) ◽  
pp. 1741-1752 ◽  
Author(s):  
Mariusz Klobukowski

Systematic sequences of Gaussian basis sets with controlled energy errors for the studies of convergence of the basis set effects on the computed results of molecular calculations are presented. The basis sets were prepared for atoms H–Ne and Al–Ar and may be used in systematic ab initio ad limitem studies of molecular structure and properties.

2004 ◽  
Vol 59 (10) ◽  
pp. 1153-1176 ◽  
Author(s):  
Thomas Zuschneid ◽  
Holger Fischer ◽  
Thomas Handel ◽  
Klaus Albert ◽  
Günter Häfelinger

AbstractHigh-resolution gas phase measurements of 1H NMR spectra at 400 MHz and atmospheric pressure of seven small hydrocarbons are presented. The developed new method and the experimental set-up are described. Ab initio GIAO MO calculations of 1H and 13C NMR absolute shieldings on the HF, MP2 and B3LYP levels using 25 standard gaussian basis sets are reported for these hydrocarbons, based on experimental re distances. The measured gas phase 1H chemical shifts have been converted to an absolute σ0 shielding scale by use of the literature shielding of methane. These and gas phase 13C literature values have been transferred with literature ZPV data to estimated σeexp shieldings which are used to evaluate the basis set dependence of the calculated σe shieldings utilizing linear least squares regressions. Exponential extrapolations of Dunning basis set calculations allow the determination of basis set limits for 1H and 13C shieldings. 1H and 13C chemical shifts have been derived from the HF calculated shieldings with shieldings of TMS which has been geometry optimized and GIAO calculated in each basis. Standard deviations (esd) as low as 0.09 ppm for 1H and 0.76 ppm for 13C calculations have been obtained.The statistically best basis set for simultaneous calculation of 1H and 13C absolute shieldings or relative shifts is 6-311G* within the HF and B3LYP methods. Aiming for highest accuracy and precision, 1H and 13C have to be treated separately. In this case, best results are obtained using MP2/6-311G** or higher for 1H shieldings and MP2/cc-pVTZ for 13C shieldings.


1984 ◽  
Vol 39 (5) ◽  
pp. 495-498
Author(s):  
V. Renugopalakrishnan ◽  
R. Walter

An ab initio molecular orbital technique was used to investigate the rotational barrier about the disulphide bridge in dimethyl disulphide. Various minimal and extended basis sets were used in the calculations. The chosen minimal basis set was the STO-3G set, and the extended basis sets were the STO 4-31G set, the Dunning and Hay set consisting of contracted Gaussian basis sets: [2s], [3s, 2p] and [6s, 4p] for H, C, and S atoms, and the Dunning and Hay basis set augmented with a d-type function on S atoms. The total energy was calculated as a function of the torsion angle about the disulphide bond. The barrier to rotation about this bond was found to be two-fold in nature, in accordance with previous findings. The heights of the barriers were observed to depend upon the basis set and input geometry. For our particular choice of basis sets and input geometry, the calculated value of the eis and trans barriers ranged from 12.68 to 16.49 kcal/mol and from 6.23 to 8 kcal/mol, respectively. Inclusion of a d-type function in the basis sets was found to result in better agreement between the calculated and experimental values, thereby emphasizing the need for considering 3d orbitals of sulphur in MO calculations


1973 ◽  
Vol 26 (7) ◽  
pp. 1381 ◽  
Author(s):  
FR Burden ◽  
BT Hart

A quantitative assessment of the importance of basis set variation at the minimal basis level of accuracy has been carried out. A number of basis sets, using Gaussian lobe functions, were tested by calculating the energy and one-electron properties for a number of molecules and the results used to select an optimum sized accurate basis set. This basis set consisted of a five-component 1s-Gaussian type orbital (GTO), a three-component 2s-GTO and a three-component 2p-GTO for each first row atom, and a three-component 1s-GTO for hydrogen. The accuracy of wave functions calculated with the minimal basis sets was very sensitive to changes in the 2p-GTO. ��� A Gaussian lobe basis set of optimum size for use with large molecules is given. The results of testing this set by calculating several one-electron properties for ten molecules, together with confidence limits, are also presented.


2018 ◽  
Author(s):  
Danilo Carmona ◽  
David Contreras ◽  
Oscar A. Douglas-Gallardo ◽  
Stefan Vogt-Geisse ◽  
Pablo Jaque ◽  
...  

The Fenton reaction plays a central role in many chemical and biological processes and has various applications as e.g. water remediation. The reaction consists of the iron-catalyzed homolytic cleavage of the oxygen-oxygen bond in the hydrogen peroxide molecule and the reduction of the hydroxyl radical. Here, we study these two elementary steps with high-level ab-initio calculations at the complete basis set limit and address the performance of different DFT methods following a specific classification based on the Jacob´s ladder in combination with various Pople's basis sets. Ab-initio calculations at the complete basis set limit are in agreement to experimental reference data and identified a significant contribution of the electron correlation energy to the bond dissociation energy (BDE) of the oxygen-oxygen bond in hydrogen peroxide and the electron affinity (EA) of the hydroxyl radical. The studied DFT methods were able to reproduce the ab-initio reference values, although no functional was particularly better for both reactions. The inclusion of HF exchange in the DFT functionals lead in most cases to larger deviations, which might be related to the poor description of the two reactions by the HF method. Considering the computational cost, DFT methods provide better BDE and EA values than HF and post--HF methods with an almost MP2 or CCSD level of accuracy. However, no systematic general prediction of the error based on the employed functional could be established and no systematic improvement with increasing the size in the Pople's basis set was found, although for BDE values certain systematic basis set dependence was observed. Moreover, the quality of the hydrogen peroxide, hydroxyl radical and hydroxyl anion structures obtained from these functionals was compared to experimental reference data. In general, bond lengths were well reproduced and the error in the angles were between one and two degrees with some systematic trend with the basis sets. From our results we conclude that DFT methods present a computationally less expensive alternative to describe the two elementary steps of the Fenton reaction. However, choice of approximated functionals and basis sets must be carefully done and the provided benchmark allows a systematic validation of the electronic structure method to be employed


2001 ◽  
Vol 115 (8) ◽  
pp. 3561-3565 ◽  
Author(s):  
Toshikatsu Koga ◽  
Hiroshi Tatewaki ◽  
Osamu Matsuoka

Sign in / Sign up

Export Citation Format

Share Document