Scaling Factors and Uncertainties for ab Initio Anharmonic Vibrational Frequencies

Abstract

To predict the vibrational spectra of molecules, ab initio calculations are often used to compute harmonic frequencies, which are usually scaled by empirical factors as an approximate correction for errors in the force constants and for anharmonic effects. Anharmonic computations of fundamental frequencies are becoming increasingly popular. We report scaling factors, along with their associated uncertainties, for anharmonic (second-order perturbation theory) predictions from HF, MP2, and B3LYP calculations using the 6-31G(d) and 6-31+G(d,p) basis sets. Different scaling factors are appropriate for low- and high-frequency vibrations. The method of analysis is based upon the Guide to the Expression of Uncertainty in Measurement, published by the International Organization for Standardization (ISO). The data used are from the Computational Chemistry Comparison and Benchmark Database (CCCBDB), maintained by the National Institute of Standards and Technology, which includes more than 3939 independent vibrations for 358 molecules.

A critical review of statistical calibration/prediction models handling data inconsistency and model inadequacy

“…As discussed by Pernot, 49 it has been used, for instance, in the calibration of the mBEEF density functional, 10,66,67 or, indirectly, in the estimation of the scaling factor uncertainty for ab initio properties statistical correction. [68][69][70] An empirical likelihood is built by scaling the data covariance matrix…”

Structures, binding energies, temperature effects, infrared spectroscopy of [Mg(NH3)n = 1−10]+ clusters from DFT and MP2 investigations

“…As previously shown by Irikura and co‐workers, the uncertainty, σ ( λ ), associated with the scaling factor λ is a critical measure of the confidence and applicability of the scaling scheme. It is defined as …”

Defects in alpha and gamma crystalline nylon6: A computational study

“…The spectra show good agreement between the results of DFTB and DFT-D methods in the high-frequency regions (corresponding to C-H stretching region), to within 1%, but noticeable differences in the area around 1,000-1,500 cm −1 , where both spectra have the same major features but with peak positions differing by up to 4% and with only the CC stretch peak differing by 6.5%. 7 Such a difference is not uncommon, 29,30 for example Quarti et al used a scaling factor of 0.96. 6 See Arabnejad et al 9 for the comparison of the vibrational intensities obtained with our DFT-D setup and experiment.…”