Quantitative resolution of spectroscopic systems using computer assisted target factor analysis (CAT)

Factor analysis (FA) is widely applicable in analytical chemistry. For suit able data structures, the generality of the approach is a strong point of F A. Using target testing procedures, a wealth of specific informations on th e components in a chemical system can be extracted without stating a specif ic model. Target testing, however, is time-consuming and often heuristic. I n the following, application of a computer-assisted target factor analysis (CAT) algorithm is described. CAT estimates a rotation matrix that transfor ms abstract factors into physically meaningful informations by using genera l constraints, i.e. non-negativity of absorptions and/or concentrations. Th us, the rotation step in factor analysis is automated and allows applicatio n of target factor analysis to situations where none or only very limited i nformation on the real factors and their respective contributions is availa ble. Thus, a target testing procedure of guessing physically meaningful fac tors and iteratively adapting these factors is performed automatically. Nev ertheless, CAT is not a black box procedure. The relative importance of dif ferent optimization constraints is balanced interactively. CAT can be appli ed to all situations where factor analysis can be used. CAT is demonstrated using UV-Vis absorption spectra of a Nd(III) polyoxometalate cryptate syst em and the system U(VI)-CO2-H2O as examples.