STANDARDIZATION OF 2ND-ORDER INSTRUMENTS

A method is presented to standardize two-dimensional responses (e.g., GC/MS, LC/UV) measured on multiple instruments or on a single instrument under different operational conditions. This second-order standardization method proceeds by calculating two banded diagonal transformation matrices, using the responses of a common standard sample, to simultaneously correct for the response channel shift and intensity variations in both dimensions or orders. Different from first-order standardization, these two transformation matrices must be estimated from a set of simultaneous nonlinear equations via the Gauss-Newton method. The effects of noise and transformation matrix bandwidth on the standardization performance are studied through computer simulation. When tested with experimental LC/UV data, the proposed standardization method can reduce the variation between two runs from 0.15-0.20 to 0.025-0.03 AU. From both the computer simulation and experimental data study, it is found that the design of the standard sample is critical for the parameter estimation and response standardization.