Systematic Optimization of Sample Introduction Parameters for Liquid Chromatography Methods Involving Feed Injection

One of the ongoing challenges faced by those developing liquid chromatography (LC) methods is the need for detection of analytes at trace levels in samples where the sample solvent constitutes a “strong solvent” for the LC separation. For example, it is common to elute analytes concentrated by solid phase extraction (SPE) from the SPE cartridge with a neat organic solvent such as methanol. Direct injection of this sample, especially in the large volumes needed for sensitive detection, can lead to poor peak shapes, including broad or even split peaks. The recent commercial introduction of “feed injection” technology for LC provides a solution to this problem, wherein the sample can be slowly “fed” into a mobile phase composed of a large fraction of “weak solvent” (i.e., water in reversed-phase separations) to mitigate the problems introduced by injecting large volumes of strong solvent. Although this sample introduction technique has been shown to be very powerful, it is does come with the complication that several new parameters must be optimized during method development, in addition to those normally focus on such as flow rate, column length, and particle size. In this presentation we will describe our work on the development of a systematic approach to optimizing these additional parameters with detection sensitivity and chromatographic resolution in mind. Specifically, we will assess the performance of simulation tools designed to faithfully simulate the effects of injecting large volumes of samples in strong solvents on peak shape and resolution, by comparison to experimental results. We will show that these simulation tools can significantly accelerate development of optimal parameters for feed injection, providing rapid access to this powerful sample introduction method without adding much additional method development time.