IntroductionIn food analyses, separations with good selectivity accompanied by highlysensitive detection are desired. This can be difficult to achieve with polarcompounds analyzed by reversed-phase liquid chromatography (RPLC) (Figure1A). Hydrophilic Interaction Liquid Chromatography (HILIC) encompasses a widerange of stationary phases that provide suitable retention and selectivity formany classes of polar molecules with highly organic mobile phases (Figure 1Band 2). When approaching RPLC, there are limited retention mechanisms toconsider. Aliphatic hydrocarbon stationary phases (e.g. C18) provide van derWaals interactions between the phase and the analytes, while phenyl-typestationary phases (e.g. Biphenyl) provide additional π-π interactions. HILICphases encompass a wider variety of chemical interactions that are used with ahigh organic (typically buffered) mobile phase consisting primarily of acetonitrileand water. The intermolecular interactions between analytes and stationaryphase can include, but are not limited to, hydrogen bonding, strong/weak cationor anion exchange. The objective of this poster is to introduce the HILICtechnique, discuss frequent problem areas, and help you successfullyincorporate HILIC methods with superficially porous particles into yourlaboratory’s repertoire.

A Perfect Pairing: HILIC and Superficially Porous Particles in Food AnalysesPaul Connolly, Landon Wiest, Dan Li, Joe Konschnik, Ty Kahler; Restek Corporation

Retention Time ReproducibilityThe first step in ensuring success with HILIC methods is understanding that theanalytical column must be properly conditioned before use as well as betweeninjections. Failure to condition and then equilibrate the column between runscan result in the water layer not being fully re-established on the particlesurface (Figure 1B). This can lead to irreproducible retention times. To conditionyour column prior to use, it must be flushed with your initial mobile phase. Forisocratic HILIC methods, at least 50 column volumes (Table 1 contains typicaldimension volumes) should be used. For gradient HILIC methods, at least 10blank injections, using the full gradient, should be performed. Conditioning isalso required when you alter the mobile phase composition or theconcentration of any additives. Use the same number of column volumes orblank injections as you would when conditioning the column for the first timeafter changing the mobile phase. In addition to the initial conditioning of thecolumn with the mobile phase, it is critical that the column be re-equilibratedbetween injections.

SummaryAs with any new technique, successful application depends in large part onunderstanding the differences between the new approach and more familiarprocedures. HILIC methods provide a powerful new way to analyze polarcompounds, but implementation can be challenging if analysts assume it issimilar to reversed-phase LC. Important points to remember:

Confirm your column is well conditioned Re-equilibrate your column sufficiently between injections Match your diluent and initial mobile phase condition as much as possible Understand the effect of pH and buffer strength on your analysis

Hopefully, these guidelines provide a good starting point for establishing yourHILIC methods. With careful and adequate method development you can ensureoptimum performance of your HILIC separation.

Mobile Phase pH Effects Are Analyte Dependent

Figure 4: Effects of Acid Concentration on Retention and Selectivity

Column: Raptor FluoroPhenyl 2.7 µm, 100 mm x 2.1 mm (cat. # 9319A12); Temp.: 40°C; Mobile Phase A: formic acid in water; Mobile Phase B: formic acid in acetonitrile;

Flow: 0.8 mL/min; Gradient (%B): 0.00 min (95% B), 10.00 min (20% B), equilibrate 2 min

HILI

C M

ode

0.2% Formic Acid

0.05% Formic Acid

PATENTS & TRADEMARKSRestek patents and trademarks are the property of Restek Corporation. (See www.restek.com/Patents-Trademarks forfull list.) Other trademarks appearing in Restek literature or on its website are the property of their respective owners.Restek registered trademarks used here are registered in the U.S. and may also be registered in other countries.

In addition to the importance of the sample diluent, the pH of the mobile phasecan also affect chromatographic performance. In fact, method development andevaluation are most important in this area because the effect of pH on analytecharge state varies based on each compound’s pKa. With HILIC methods, thehigh percentage of organic solvent in the mobile phase raises the effective pH,therefore the actual eluent pH may behave 1 - 1.5 units higher than is observedin reversed-phase separations. The charge state of the column should also beconsidered. For example, bare silica has a pKa between 3.8 and 4.5, therefore,pH changes the charge of the silica surface. Under very acidic conditions thesilica surface is neutral, while under more moderate pH conditions, the surfaceis ionized (negatively charged). This negative surface charge is excellent forretaining analytes with a positive charge. For this reason, if your analyte haspositively charged groups (e.g. amines), it may retain well on a Raptor HILIC-Sicolumn (Figure 4).

Buffer Choice Effects on Chromatographic Performance

Buffers are used to keep the eluent pH constant. They can also significantlyimprove peak shape and assist in separating analytes of interest frominterferences. Many HILIC separations use a mass spectrometer for detection, sovolatile buffers like ammonium formate and ammonium acetate are verycommon. The buffer concentration in HILIC separations is important to considerfor two main reasons:

1) High organic content of the mobile phase may cause buffer salts toprecipitate in your column or system resulting in your instrument goingoffline for maintenance.

2) Positively charged buffer ions compete with target analytes for retention,therefore, if the buffer concentration is high, analyte retention can bereduced.

10 mM is a good recommended starting point for buffer concentration in youraqueous and organic mobile phases. Make sure that both mobile phases arebuffered equally to keep the ionic strength constant during a gradient. This willassist in maintaining consistent MS detector response.Importance of Mobile Phase Diluent

Similar to reversed-phase separations the injection solvent (diluent) makes a bigdifference in peak shape, especially when it is very different from the initialmobile phase conditions at the start of your method. The same is true in HILICmethods. It is critical the injection solvent be as similar to the initial mobilephase conditions as possible. In HILIC separations these are typically high inorganic content.

Figure 3 illustrates the impact of solvent mismatching on peak shape. If thesample is prepared in 100% aqueous diluent, peak shape is poor for diquat, bothcompounds elute early, and the signal is relatively low compared to the injectionwhere the sample was dissolved in mobile phase B (75% acetonitrile). Bymatching the injection solvent to the initial mobile phase conditions, you getbetter peak shape, increased retention, and higher sensitivity.

Figure 3: Separation of 1) Diquat and 2) Paraquat on Raptor 2.7 µm HILIC-Si

Column: Raptor HILIC-Si (cat. # 9310A52), 50 mm x 2.1 mm ID, Temp.: 45 °CSample: Paraquat & diquat calibration mix (cat.# 32437), 50 ng/mL; 5 µL Inj. vol.Mobile Phase

A: Water, 50 mM ammonium formate, 0.5% formic acidB: 25:75 Water:Acetonitrile, 50 mM ammonium formate, 0.5 % formic acid

Flow: 0.6 mL/min; Gradient (%B): 0.00 min (100% B), 4.00 min (65% B), equilibrate 3 min

1. Atenolol2. Phenytoin3. Trenbolone4. Testosterone

5. Nortriptyline6. Amitriptyline7. Celecoxib

Figure 1: Retention Trends of HILIC and Reversed-Phase Modes

B)

A)

Figure 2: HILIC Retention Profile of 4-Methylimidazole (Isocratic)

Table 1: Column Volumes (mL) Based Upon Raptor HILIC-Si ColumnDimensions

It is recommended to equilibrate with at least 10 column volumes starting whenthe gradient program returns to initial conditions. With isocratic separations, atleast 10 column volumes, starting from the retention time of the last peak,should be flushed through the column for proper equilibration between runs.For example, a 2.1 mm x 100 mm column would require at least 2 mL of startingmobile phase to completely re-equilibrate your column. The number of columnvolumes required for complete re-equilibration can be very analyte dependent.It is strongly recommend to investigate retention time reproducibility duringmethod development.

Column Length

Column ID 30 mm 50 mm 100 mm 150 mm

2.1 mm 0.05 0.1 0.2 0.3

3.0 mm -- 0.2 0.4 0.5

4.6 mm -- 0.4 0.8 1.2