Liquid chromatography above 20,000 PSI

Abstract

Continued improvements in HPLC have led to faster and more efficient separations than previously possible. One important aspect of these improvements has been the increase in instrument operating pressure and the advent of ultrahigh pressure LC (UHPLC). Commercial instrumentation is now capable of up to ~20 kpsi, allowing fast and efficient separations with 5-15 cm columns packed with sub-2 μm particles. Home-built instruments have demonstrated the benefits of even further increases in instrument pressure. The focus of this review is on recent advancements and applications in liquid chromatography above 20 kpsi. We outline the theory and advantages of higher pressure and discuss instrument hardware and design capable of withstanding 20 kpsi or greater. We also overview column packing procedures and stationary phase considerations for HPLC above 20 kpsi, and lastly highlight a few recent applicatioob pressure instruments for the analysis of complex mixtures.

Keywords: Column packing; Long columns; Omics; Small particles; Ultrahigh pressure LC.

Fig. 1.

Kinetic plot illustrating the effect of pressure, particle size, and analysis time on separation efficiency using packed column liquid chromatography. The blue and green traces are what is achievable with 0.5 μm particles at 50 and 12 kpsi, respectively. The black and orange traces are what is achievable with 1.7 μm particles at 50 and 12 kpsi, respectively. Different column dead times are shown as diagonal lines for clarity.

Fig. 2.

Schematic of different gradient instrument designs (A and B). Design A is from Ref. [28] and operable up to 70 kpsi, however lacked automation, produced nonlinear gradients, and required flow splitting to produce nanoliter per minute flow rates. Design B is from Ref. [30] and is fully automated, splitless, and produces linear gradients. Gradient loading and sample injection are performed with a commercial UPLC and stored on a gradient loop, with subsequent ultrahigh pressure separation performed with a pneumatic amplifier pump. The pressure limit was ~45 kpsi. Illustrations of the fittings for each design (C and D) show differences in bolt size and tubing connection leading to the differences in pressure limit of the two systems. Adapted with permission from Refs. [28,30]. Copyright (1999) American Chemical Society.

Fig. 3.

Reduced van Deemter plots illustrating the effect of slurry solvents with 3 mg/mL concentration on chromatographic performance for 1.1 μm superficially porous C18 particles packed in ~12 cm × 30 μm capillaries (A). The dashed line is performance expected from theory. Adapted with permission from Ref. [47]. Reduced van Deemter plots of 75 μm i.d. capillaries packed with 1.3 μm fully porous C18 particles shows the difficulty in packing longer columns (e.g. > 40 cm) with these small particles (B). A 100 cm column was packed and subsequently cut in to three sections with the last packed sections (center and inlet) exhibiting poor performance and dominating the overall column performance (100 cm - before cutting), illustrating large axial heterogeneities. Data from Reising et al., 2016 is a comparison of a single 34 cm column with the same 1.3 μm particles showing similar performance to the outlet of the 100 cm column, indicating minimal axial heterogeneities and good column packing for shorter columns under these conditions. Adapted with permission from Ref. [65].

Fig. 4.

Effect of column packing pressure on various proteomic experiments using 30 cm × 75 μm columns with fully porous 1.7 μm C18 particles. For the 11 kpsi case, columns were packed at ~1 kpsi and subsequently flushed at 11 kpsi. For the >20 kpsi case, columns were packed at either 20 or 30 kpsi. All columns were operated around 11 kpsi on a commercial capillary LC-MS system. Adapted with permission from Ref. [71]. Copyright (2016) American Chemical Society.

Fig. 5.

Separation of intact, global proteoforms in S. oneidensis lysate on a 120 cm × 100 μm column packed with 3.6 μm C4 core-shell particles and operated at 14 kpsi. Effluent was connected to an Exactive mass spectrometer. A peak capacity of 450 was achieved in an 800-min gradient and ~900 proteoforms were identified. Adapted with permission from Ref. [78].