The relative error caused by reproducibility in two-dimensional correlation spectroscopy

Reproducibility is a very important issue for the construction and interpretation of two-dimensional (2D) correlation spectra. This paper introduced the way to estimate the relative error in 2D correlation spectroscopy, especially focused on the possible source of errors that may arise from the limitation of experimental reproducibility. Mainly two sources of reproducibility are examined. The first is the reproducibility of the performance of experimental apparatus, in which the baseline drift as well as the noise level of spectroscopic measurements are taken into consideration. The maximum intensities of the 2D correlation maps obtained by the measurement of a series of small fluctuations of baselines are typically only in the order of 10−5 and 10−7, respectively, for synchronous or asynchronous spectra, which are so small that their effect can be safely ignored. Meanwhile, the 2D correlation maps based purely on the experimental noise can also be constructed, which provides the threshold above which any correlation can be treated as meaningful. The second source of relative error is caused by the reproducibility of the experimental condition. The most commonly used perturbations for generalized 2D correlation spectroscopy are concentration and temperature. Therefore, the effect of the reproducibility of the concentration of each sample prepared, as well as the small variation of the temperature during the measurement, which may all give rise to relative error, was examined. A quantitative measure of the extent of relative error caused by this kind of reproducibility can also be analyzed by using 2D correlation maps.