Partial least squares prediction of the first hyperpolarizabilities of donor-acceptor polyenic derivatives

Graphical abstract: PLS regression equations predicts quite well static β values for a large set of donor-acceptor organic molecules, in close agreement with the available experimental data. Display Omitted Highlights: → PLS regression predicts static β values of 35 push-pull organic molecules. → PLS equations show correlation of β with structural-electronic parameters. → PLS regression selects best components of push-bridge-pull nonlinear compounds. → PLS analyses can be routinely used to select novel second-order materials. - Abstract: A partial least squares regression analysis of a large set of donor-acceptor organic molecules was performed to predict the magnitude of their static first hyperpolarizabilities (β's). Polyenes, phenylpolyenes and biphenylpolyenes with augmented chain lengths displayed large β values, in agreement with the available experimental data. The regressors used were the HOMO-LUMO energy gap, the ground-state dipole moment, the HOMO energy AM1 values and the number of π-electrons. The regression equation predicts quite well the static β values for the molecules investigated and can be used to model new organic-based materials with enhanced nonlinear responses.