Huiying Song

Hydrophilic interaction chromatography (HILIC) in the analysis of antibiotics

This paper presents a general overview of the application of hydrophilic interaction chromatography (HILIC) in the analysis of antibiotics in different sample matrices including pharmaceutical, plasma, serum, fermentation broths, environmental water, animal origin, plant origin, etc. Specific applications of HILIC for analysis of aminoglycosides, β-lactams, tetracyclines and other antibiotics are reviewed. HILIC can be used as a valuable alternative LC mode for separating small polar compounds. Polar samples usually show good solubility in the mobile phase containing some water used in HILIC, which overcomes the drawbacks of the poor solubility often encountered in normal phase LC. HILIC is suitable for analyzing compounds in complex systems that elute near the void in reversed-phase chromatography. Ion-pair reagents are not required in HILIC which makes it convenient to couple with MS hence its increased popularity in recent years. In this review, the retention mechanism in HILIC is briefly discussed and a list of important applications is provided including main experimental conditions and a brief summary of the results. The references provide a comprehensive overview and insight into the application of HILIC in antibiotics analysis.

Extensive database of liquid phase diffusion coefficients of some frequently used test molecules in reversed-phase liquid chromatography and hydrophilic interaction liquid chromatography

Diffusion plays an important role in all aspects of band broadening in chromatography. An accurate knowledge of molecular diffusion coefficients in different mobile phases is therefore crucial in fundamental column performance studies. Correlations available in literature, such as the Wilke-Chang equation, can provide good approximations of molecular diffusion under reversed-phase conditions. However, these correlations have been demonstrated to be less accurate for mobile phases containing a large percentage of acetonitrile, as is the case in hydrophilic interaction liquid chromatography. A database of experimentally measured molecular diffusion coefficients of some 45 polar and apolar compounds that are frequently used as test molecules under hydrophilic interaction liquid chromatography and reversed-phase conditions is therefore presented. Special attention is given to diffusion coefficients of polar compounds obtained in large percentages of acetonitrile (>90%). The effect of the buffer concentration (5-10mM ammonium acetate) on the obtained diffusion coefficients is investigated and is demonstrated to mainly influence the molecular diffusion of charged molecules. Diffusion coefficients are measured using the Taylor-Aris method and hence deduced from the peak broadening of a solute when flowing through a long open tube. The validity of the set-up employed for the measurement of the diffusion coefficients is demonstrated by ruling out the occurrence of longitudinal diffusion, secondary flow interactions and extra-column effects, while it is also shown that radial equilibration in the 15m long capillary is effective.

Evaluation and comparison of the kinetic performance of ultra-high performance liquid chromatography and high-performance liquid chromatography columns in hydrophilic interaction and reversed-phase liquid chromatography conditions

An intrinsic performance comparison is made of the reduction in analysis time that can be obtained when switching from HPLC to UHPLC column formats in HILIC and reversed-phase conditions. A detailed overview of the packing characteristics of both stationary phase types is given first. It is demonstrated that HILIC columns demonstrate higher external porosity values than their reversed-phase counterparts resulting in lower flow resistance values. Column total porosity values determined from the elution time of a small marker molecule are shown to depend strongly on the composition of the mobile phase. To omit errors that might arise from an over- or underestimation of the column void time, all plate height and kinetic plot data are therefore expressed as a function of the interstitial velocity. Although only a limited number of columns are evaluated in this study, it is shown that the column efficiency of the HILIC columns is lower than that of their reversed-phase counterparts, at least for the compounds evaluated here. Despite this lower efficiency, the kinetic performance of both stationary phase types is similar, due to the much lower viscosity of the mobile phases typically used in HILIC conditions. Finally, it is demonstrated that a similar, yet slightly larger reduction in analysis time can be obtained when switching from HPLC column formats to UHPLC formats in HILIC compared to reversed-phase conditions.

Evaluation of the Kinetic Performance Differences between Hydrophilic-Interaction Liquid Chromatography and Reversed-Phase Liquid Chromatography under Conditions of Identical Packing Structure.

A protocol using trifluoroacetic acid at a temperature of 60 °C is developed for the adequate removal of the stationary phase of reversed-phase liquid chromatography (RPLC) columns. This procedure allows for studying the same column first under RPLC and subsequently under hydrophilic-interaction liquid chromatography (HILIC) conditions to isolate intrinsic differences between mass transfer properties in HILIC and RPLC from differences in packing quality. The established procedure allows for a complete removal of the stationary phase (confirmed by retention studies and thermogravimetry analyses) while leaving the structure of the packing unaffected (witnessed by an unchanged external porosity and pressure drop). Accurate plate height analysis comparing compounds at the same zone retention factor indicates a significant difference in reduced c-term (typically 40-80% larger under HILIC conditions), despite the columns otherwise being identical. Correcting for the known contributions of longitudinal diffusion (b-term) and mass transfer (cm- and cs-term) to focus on band broadening originating from eddy dispersion, similar strong differences are observed (differences of some h = 0.3 up to 1.2). These findings show that the interior structure and retention mechanism of the particles have a very strong effect on the observed eddy dispersion, a factor typically ascribed to phenomena occurring outside the particles. This also implies that comparing the quality of packings of different particle types is virtually impossible without the availability of a sound model to correct for the intraparticle effect on the observed eddy dispersion.

Methodologies to determine b-term coefficients revisited

The accuracy of the longitudinal diffusion term (b-term) plays a vital role in the study of mass transfer mechanisms in high performance liquid chromatography (HPLC). In this study, three commonly used methodologies (peak parking; fitting of an experimental van Deemter curve; and the so-called dynamic method) for the determination of the b-term constant were investigated in detail. The three methods were compared based on their mutual agreement, the intra- and inter-day variation of the obtained values and the time required to measure them. Whereas the dynamic method was found to be plagued by impractically long waiting times and concomitant baseline variations compromising accurate measurements of the band broadening, the two other methods lead to very similar b-values, i.e., well within the 1% RSD inter-day variation typically marking both methods in the present study. The best way to study the agreement of the peak parking and plate height fitting method is in a plot of h.ν versus ν, providing a much better zoom on the b-term region of the van Deemter curve than the customarily employed h versus ν-curve and hence allowing to identify any anomalous measurement values (usually related to measurements with a long experimentation time). Verifying the mutual agreement between both methods is proposed here as an additional accuracy check of the obtained data.

Development and validation of LC methods for the separation of misoprostol related substances and diastereoisomers

Misoprostol is a synthetic prostaglandin E1 analogue which is mainly used for prevention and treatment of gastric ulcers, but also for abortion due to its labour inducing effect. Misoprostol exists as a mixture of diastereoisomers (1:1) and has several related impurities owing to its instability at higher temperatures and moisture. A simple and robust reversed phase liquid chromatographic (RPLC) method is described for the separation of the related substances and a normal phase (NP) LC method for the separation of misoprostol diastereoisomers. The RPLC method was performed using an Ascentis Express C18 (150 mm × 4.6 mm, 5 μm) column kept at 35 °C. The mobile phase was a gradient mixture of mobile phase A (ACN-H2O-MeOH, 28:69:3 v/v/v) and mobile phase B (ACN-H2O-MeOH, 47:50:3 v/v/v) eluted at a flow rate of 1.5 mL/min. UV detection was performed at 200 nm. The NPLC method was undertaken by using an XBridge bare silica (150 mm × 2.1 mm, 3.5 μm) column at 35 °C. The mobile phase contained 1-propanol-heptane-TFA (4:96:0.1%, v/v/v), pumped at a flow rate of 0.5 mL/min. UV detection was performed at 205 nm. This LC method can properly separate the two diastereoisomers (Rs > 2) within an analysis time of less than 20 min. Both methods were validated according to the ICH guidelines. Furthermore, these new LC methods have been successfully applied for purity control and diastereoisomers ratio determination of misoprostol bulk drug, tablets and dispersion.

Relevance and Assessment of Molecular Diffusion Coefficients in Liquid Chromatography

Molecular diffusion plays an important role in high-performance liquid chromatography, especially in fundamental column performance studies. An accurate knowledge of the molecular diffusion coefficients (Dm) of compounds selected for column evaluation is therefore crucial. In this review, a general overview is presented of the advantages and drawbacks of correlation-based and experimental methods that can be employed to determine molecular diffusion coefficients. The former include the Wilke–Chang, Scheibel, Reddy–Doraiswamy, Lusis–Ratcliff and Hayduk–Laudie equations, and other empirical correlations based on the Wilke–Chang equation. It is discussed how the association factor (ψ) that is required in several of these correlations can be obtained from the solubility parameter (δ). Frequently used experimental methods include the light scattering, nuclear magnetic resonance, peak parking and Taylor–Aris method, and methods employing microfluidic devices. The principles of these experimental methods are elucidated in detail. Moreover, the influence of several parameters, such as solute characteristics, solvent viscosity, temperature and pressure on the molecular diffusion coefficient is described.