Solvent Extraction with Electrokinetic Multicommutation Analysis System

• Published in: Instrumentation science & technology Vol. 31; no. 3; pp. 269 - 282
• Main Authors: Li, Long-Quan, Fang, Li, He, You-zhao
• Format: Journal Article
• Language: English
• Published: Taylor & Francis Group 07.01.2003
• Subjects: Electro-osmotic pump; Electrokinetic flow analysis; Multicommutation; Solvent extraction
• ISSN: 1073-9149; 1525-6030
• DOI: 10.1081/CI-120022654

The feasibility of solvent extraction with an electrokinetic flow analysis (EKFA) system should be investigated further because of its non-aqueous condition. As an example, the pre-concentration and determination of copper, including on-line color reaction, solvent extraction, and spectrometric determination were carried out by an electrokinetic multicommutation analysis system in this work. The proposed system consisted of one electroosmotic pump, four solenoid valves which were controlled by a personal computer, and a spectrometric detector. With the electroosmotic pump, sample and reagent solutions were aspirated into the system's color reaction section in small plugs, alternately, and were reacted sufficiently. Then, a plug of chloroform was inserted into the colored zone. The proposed system carried out solvent extraction by moving the sandwiched zone back and forth along its short extraction coil. After phase separation with a T-shape separator, organic phase was determined in a flow detection cell of the spectrophotometer. The electroosmotic pump is a liquid delivery device with the merits of non-pulsed and stable flow rate, simple pump structure, high driving efficiency, and facile flow control. Ethanol solution, a relatively high permittivity and innocuous solvent, was adopted as the pump carrier. In this paper, carrier additives were investigated further in order to improve the pump stability and efficiency (3.9 mL min −1  mA −1 ). The detection limit for copper was 0.03 mg L −1 (K = 3, n = 11) and its recoveries were in the range of 92-96% with a relative standard deviation of 5.2% (n = 3). The proposed method can provide a throughput of 20 samples per hour.