Determination of the association constant of strongly bonded host-guest systems by multivariate regression of infrared spectroscopic data

A new method applying infrared spectroscopy in combination with multivariate regression for the determination of association constants has been developed to overcome the sensitivity problems arising when strongly associating complexes are investigated by NMR or UV–VIS methods. A concentration profile for the complex is derived by correction of multivariate regression data. Subsequent iterative fitting of the corrected data yields the association constant.The regression part is not integrated in the process of the association constant determination itself. Separating data treatment from the actual fitting procedure offers a means of evaluating the quality of the data set and the order of association before the actual calculation of the association constant. Furthermore, the method of regression can be varied and the applicability of the selected method can be assessed from the data-regression output.From simulated data, it is estimated that an association constant range of 102–106M–1 can be determined when measuring at millimolar concentration levels.Examples are shown for halide complexation of a thiourea compound and a urea-substituted resorcinarene cavitand.