Experimental Design and Response Surface Modeling: A Method Development Application for the Determination of Reduced Inorganic Species in Environmental Samples

Abstract

To confirm the significance of reduced inorganic species in nature, it is important to develop sensitive and selective analytical techniques to detect these species in complex environmental matrices. As a model application, we report on the successful use of fractional factorial and Box-Behnken designs in factor screening, optimization and validation of an on-line flow-injection method for the determination of phosphite [P(+III)] in aqueous samples. Fractional factorial results indicated that the combined KI, KIO3 and ammonium molybdate flow rates, reaction temperature and KIO3 concentration were the most important single effects. The main interactive effects were between flow rate and reaction temperature, and between sample volume and reaction temperature. The Box-Behnken design further optimized the response with results confirming the significant single effects of flow rate and temperature as well as the interactive effects between flow rate and reaction temperature. Overall, the model from the Box-Behnken design predicted critical values as: flow rate = 0.40 mL.min-1, reaction temperature = 47 °C, sample volume = 85 μL, KI reagent concentration = 1.06 g.L-1 and KIO3 reagent concentration = 0.29 g.L-1. P(+III) determinations in spiked ultra-pure water were performed using the predicted optimized values from the Box-Behnken design and compared favorably with experimental results. In addition, the potential use of such methodology in the development of sensitive laboratory and field-based methods for the detection of a suite of reduced inorganic species in complex matrices was discussed.

Keywords: Box-Behnken design, fractional factorial design, optimization, reduced inorganic species, response surface modeling