A robust multi-syringe system for process flow analysis. Part 3. Time based injection applied to the...

A new software-controlled time-based system for sample or reagent introduction in process flow injection analysis was developed. By using a multi-syringe burette coupled with one multi-port selection valve, the time-based injection of precise known volumes was accomplished. Characteristics and performance of the injection system were studied by injecting an indicator in a buffered carrier. Two multi-syringe time-based injection (MS-TBI) systems were implemented: first, the injection of a sample in a multiple-channel manifold where the sample would sequentially merge and react with different reagents, and second, the sequential injection of several solutions (sample and reagents) into a particular flowing stream. The first system was applied to the spectrophotometric determination of nickel(II) in diluted samples from the acidic nickel ore leaching process, by using ammonium citrate as carrier, a saturated solution of iodine as oxidizing agent and alkaline dimethylglyoxime as chromogenic reagent. The sampling frequency attained was 57 h-1. Determinations on process samples compared well at the 95% confidence level with the reference values obtained by ICP-OES. The second time-based injection system was applied to the speciation of iron. Total iron and iron(II) concentrations were separately and sequentially determined using 1,10-phenanthroline in acetic buffer medium as reagent. The developed manifold allowed the optional use of two different carrier solutions, containing or not containing ascorbic acid, for performing the separate determinations. Also, in the sequential procedure, plugs of reducing carrier were alternatively intercalated before the sample injections used for total iron determinations. Sampling frequencies of 68 injections per hour were routinely used. Accuracy was assessed by analyzing synthetic known mixtures of Fe(III) and Fe(II) standard solutions. Recoveries of 98-100.5% with a maximum relative standard deviation of 3.6% were found. Results obtained for various samples of fertilizers agreed well with those attained by the standard batch procedure.