Managing sample introduction problems in hydrophilic interaction liquid chromatography

Practical examination of flow rate effects and influence of the stationary phase water layer on peak shape and retention in hydrophilic interaction liquid chromatography

The effect of flow velocity on retention and peak shape of neutral, acidic and basic probe compounds was studied using seven different UHPLC hydrophilic interaction chromatography (HILIC) columns. Surprisingly on some columns, the retention factor k was found to vary somewhat with flow velocity, due to the combined effects of pressure and of frictional heating on retention. The selectivity of different HILIC columns was much greater than typically found with RP columns. The volume of the water layer on the HILIC columns was measured using the toluene exclusion procedure. For the neutral solute uridine, a good correlation was found between the volume of the water layer and retention, indicating the likely domination of a partition mechanism. For the ionogenic solutes, the correlation was generally poor, due to the presence of strong additional mechanisms such as ionic retention and repulsion. Reduced Van Deemter plots for uridine showed a negative correlation between the reduced b coefficient and the volume of the water layer, which can be attributed to reduced surface diffusion in this viscous layer. Once again, the behaviour of ionic solutes was complex on some columns making detailed interpretation difficult.

Development of a column switching for direct online enrichment and separation of polar and nonpolar analytes from aqueous matrices

Trace substances in surface waters may threaten health and pose a risk for the aquatic environment. Moreover, separation and detection by instrumental analysis is challenging due to the low concentration and the wide range of polarities. Separation of polar and nonpolar analytes can be achieved by using stationary phases with different selectivity. Lower limits of detection of trace substances can be obtained by offline enrichment on solid phase materials. However, these practices require substantial effort and are time consuming and costly. Therefore, in this study, a column switching was developed to enrich and separate both polar and nonpolar analytes by an on-column large volume injection of aqueous samples. The column switching can significantly reduce the effort and time for analyzing trace substances without compromising on separation and detection. A reversed phase (RP) column is used to trap the nonpolar analytes. The polar analytes are enriched on a porous graphitized carbon column (PGC) coupled serially behind the RP column. A novel valve switching system is implemented to enable elution of the nonpolar analytes from the RP column and, subsequently, elution of polar analytes from the PGC column and separation on a hydrophilic interaction liquid chromatography (HILIC) column. To enable separation of polar analytes dissolved in an aqueous matrix by HILIC, the water plug that is flushed from the PGC column is diluted by dosing organic solvent directly upstream of the HILIC column. The developed method was tested by applying target analysis and non-target screening, highlighting the advantage to effectively separate and detect both polar and nonpolar compounds in a single chromatographic run. In the target analysis, the analytes, with a logD at pH 3 ranging from -2.8 to + 4.5, could be enriched and separated. Besides the 965 features in the RP phase, 572 features from real wastewater were observed in the HILIC phase which would otherwise elute in the void time in conventional one-dimensional RP methods.

Improved sample introduction approach in hydrophilic interaction liquid chromatography to avoid breakthrough of proteins

When therapeutic proteins are analysed under hydrophilic interaction liquid chromatography (HILIC) conditions, there is an inherent mismatch between the sample diluent (proteins must be solubilised in aqueous media) and the mobile phase, which is mostly composed of aprotic solvent (acetonitrile). This difference in eluent strength between sample diluent and mobile phase is responsible for severe analyte breakthrough and peak distortion. As demonstrated with therapeutic proteins of different sizes (insulin of 6 kDa, anakinra of 17 kDa and rituximab subunits of 25 and 50 kDa), only very small volumes of 0.1-0.2 µL can be injected without breakthrough effects, when performing rapid analysis on short HILIC columns of 20-50 mm, leading to poor sensitivity. In order to avoid the undesired effect of the strong sample diluent, a special injection program should be preferred. This consists in the addition and automatic injection of a defined volume of weak solvent (acetonitrile) along with the sample to increase retention factors during sample loading. Various injection programs were tested, including the addition of a pre-injection or post-injection or both (bracketed injection) of acetonitrile plugs. Several weak to strong injection solvent ratios of 1:1, 1:2, 1:4 and 1:10 were tested. Our work proves that the addition of a pre-plug solvent with a weak vs. strong injection solvent ratio of 1:10 is a valuable strategy to inject relatively large volumes of proteins in HILIC, regardless of column dimensions, thus maximising sensitivity. No peak deformation or breakthrough was observed under these conditions. However, it is important to note that peak broadening (40 % larger peaks) was observed when the injection program increased the injection solvent ratio from 1:1 to 1:10. Finally, this strategy was applied to a wide range of therapeutic mAb products with different physico-chemical properties. In all cases, relatively large volumes can be successfully injected onto small volume HILIC columns using a purely aqueous sample diluent, as long as an appropriate (weak) solvent pre-injection is applied.

Practical study on the impact of injection conditions in gradient elution mode for the analysis of therapeutic proteins when using very short columns

The impact of injected sample volume on apparent efficiency has been studied for very short columns in a systematic way. For large molecules such as therapeutic proteins, it was found that relatively large volumes can be injected onto ultra-short RPLC and IEX columns (i.e. L < 50 mm) without significantly affecting the quality of the separation. This favourable behavior is due to the on-off elution mechanism of large molecules and to the fact that therapeutic protein samples are formulated in aqueous-based media, which is the weakest solvent in RPLC and IEX. Therefore, their peak is strongly focused at the column inlet even when large volume is injected, and pre-column peak dispersion is compensated. However, ultra-short HILIC columns do not seem to be favorable, as they require for very low injection volume to avoid detrimental peak splitting and breakthrough effects. Such peak distortion is related to the inherent solvent mismatch between sample diluent (aqueous) and mobile phase strength (highly organic in HILIC). When studying mass load, the ranking of the elution modes was the same, and the largest relative mass could be injected onto IEX columns (as large as 10% sample to sorbent mass), without affecting the separation quality.