Multivariate standardisation for non-linear calibration range in the chemiluminescence determination of chromium

Multivariate standardisation is proposed for the successful chemiluminescence determination of chromium based on luminol-hydrogen peroxide reaction. In an extended concentration range, non-linear calibration model is needed. The studied instrumental situations were different detection cells, instrum...

Full description

Saved in:
Bibliographic Details
Published in:Talanta (Oxford) 2007-05, Vol.72 (3), p.1004-1012
Main Authors: Tortajada-Genaro, L.A., Campíns-Falcó, P.
Format: Article
Language:English
Subjects:
Analytical chemistry
Chemical and thermal methods
Chemiluminescence
Chemistry
Chromium
Exact sciences and technology
General, instrumentation
Multivariate standardisation
Non-linear calibration models
Piecewise direct standardisation
Citations: Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Multivariate standardisation is proposed for the successful chemiluminescence determination of chromium based on luminol-hydrogen peroxide reaction. In an extended concentration range, non-linear calibration model is needed. The studied instrumental situations were different detection cells, instruments, assemblies, time and their possible combinations. Chemiluminescence kinetic registers have been transferred using piecewise direct standardisation (PDS) method. The optimisation of transfer parameters has been carried out based on the prediction residual error criteria. Non-linear principal component regression (NL-PCR) and non-linear partial least square regression (NL-PLS) were chosen for modelling the relationship signal-concentration of transferred registers. Good accuracy and precision were obtained for water samples. The concentrations of chromium were statistically in agreement with reference method values and with recovery studies. Therefore, it is possible to transfer chemiluminescence curves without loosing ability of prediction, even the presence of a non-linear behaviour.
ISSN:0039-9140
1873-3573
DOI:10.1016/j.talanta.2006.12.033