Modifying conventional high-performance liquid chromatography systems to achieve fast separations with Fused-Core columns: A case study

Abstract

The theoretical increase in performance from the use of high efficiency columns with conventional HPLC equipment is generally not observed due to the design limitations of such equipment, particularly with respect to extra-column dispersion (ECD). This study examines the impact of ECD from a Waters Alliance 2695 system on the performance of 2.7 μm HALO ® C 18 Fused-Core superficially porous particle columns of various dimensions. The Alliance system was re-configured in different ways to reduce extra-column volume (ECV) and the ECD determined in each case as a function of flow rate up to a maximum of 2 mL/min. The results obtained showed a progressive decrease in ECD as the ECV was reduced, irrespective of the flow rate employed. However, this decrease in ECD was less than theoretically expected for the lower ECV configurations. The inability to reduce the actual extra-column dispersion further was attributed to additional dispersion associated with the design/volume of the auto-injector. This was confirmed by making sample injections with a low dispersion manual injection valve, instead of auto-injection, for the two lowest ECV configurations studied. In each case, the measured and predicted ECD values were in good agreement. The auto-injector module is an integral part of the Alliance 2695 instrument and cannot be easily modified. However, even with autosampler injection, for a 3 mm ID × 100 mm Fused-Core ® column approximately 70% of the maximum plate count (∼84% of the resolution or more) could still be obtained in isocratic separations for solutes with k ≥ ∼4.5 when using the lowest ECV configuration. This study also highlights some of the problems inherent in trying to measure accurately the true extra-column dispersion of a chromatographic system and compares the results obtained to those theoretically predicted. Using this same lowest volume instrument configuration, two real-world pharmaceutical methods were scaled to separations that are ∼3–3.5-fold faster, while still maintaining comparable data quality (resolution and signal-to-noise ratios).