Automated development of continuous-flow methods

Date

1978

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Analytical methods need to be developed systematically if they are to be reliable and maximally useful; complete development of a method should include optimization of response(s) from the method, followed by characterization of the method in the optimum region using structured experimental designs. The variable-size simplex algorithm represents an efficient optimization strategy which is conceptually and mathematically simple, and which is general enough to be applicable to a wide variety of analytical methods. After the simplex has located an apparent optimum, the surrounding region can be mapped using an appropriate three-level experimental design; analysis of the mapping data can provide a working understanding of the method in the optimum region. The application of automation to the process of analytical methods development can yield benefits in terms of precision, time, and data handling. Automation can be implemented in this context by establishing closed-loop communication between a computer and analytical instrumentation. Interface circuitry has been constructed which allows bi-directional communication between a Hewlett-Packard 9830 minicomputer and a modified Technicon AutoAnalyzer II (continuous-flow clinical analyzer), enabling this system to be used for automated (closed-loop) methods development. The cresolphthalein Complexone method for serum calcium determination was investigated using the modified AutoAnalyzer II under computer control. Simplex optimization of reagent concentrations, followed by response-surface mapping in the region of the optimum, produced a method yielding 8.5% greater calcium sensitivity and 15% lower baseline absorbance than the standard method, with comparable insensitivity to interferences, and only a very slight sacrifice in linearity; a valuable operational understanding of the chemical system was also obtained. A "stopped-flow" method for the kinetic Jaffe determination of creatinine was developed under computer control using the automated continuous-flow system. Simplex optimization was used to find conditions of hydroxide and picrate giving maximum sensitivity to creatinine. A modified central composite experimental design was employed to evaluate creatinine sensitivity and protein, glucose, and acetone interferences as functions of hydroxide and picrate concentrations.

Description

Keywords

Citation