Hydrophilic interaction chromatography (HILIC) was employed to investigate chromatographic behavior of selected flavonoids from their different subgroups differing in polarity. Chromatographic measurements were performed on two different HILIC columns: unmodified silica (Atlantis-HILIC) and zwitterionic sulfoalkylbetaine (SeQuant ZIC-HILIC). Separation parameters such as content and type of organic modifier were studied. On ZIC column retention factors were observed to be inversely proportional to the buffer content in the mobile phase, which is the typical partitioning mechanism. In the case of bare silica column more or less apparent dual retention mechanism was observed, depending on the water component content in the mobile phase. ZIC-HILIC showed better selectivity (in comparison to silica column) with the detection limit of 0.01 mg/L (only for rutin was 0.05 mg/L). Finally, this chromatographic procedure was validated and applied for the determination of some flavonoids in
Hydrophilic interaction chromatography (HILIC) was first introduced by Alpert in 1990 [
Many stationary phases have been applied in HILIC mode [
The objective of this work was to investigate chromatographic behavior of several flavonoids onto two HILIC stationary phases—unmodified silica and zwitterionic sulfobetaine. The compounds used for the experiments differ in polarity. Acetonitrile and methanol were applied as the main components of mobile phases. In acetonitrile or methanol-rich mobile phase, spraying conditions improve and enhance the efficiency of desolvation and ionization in the electrospray ion source, which can provide increasing sensitivity with respect to reverse phase conditions. Finally, the potential of hydrophilic interaction liquid chromatography was used in the analysis of
Water was obtained from a Milli-Q water purification system (Millipore, Bedford, MA, USA). Acetonitrile (ACN) and methanol (MeOH) were of HPLC grade from Merck (Darmstadt, Germany).
The commercial standards of flavonoids were purchased from Sigma (Steinheim, Germany). The stock solutions of these individual compounds were prepared at 100
Chromatographic analysis was performed with the Shimadzu LC system consisting of binary pumps LC20-AD, degasser DGU-20A5, column oven CTO-20AC, autosampler SIL-20AC, and 3200 QTRAP Mass spectrometer (Applied Biosystem/MDS SCIEX). The MS system was equipped with the electrospray ionization source (ESI) operated in negative-ion or in positive mode. ESI conditions were as the following: the capillary temperature 450°C, the curtain gas at 0.3 MPa, the auxiliary gas at 0.3 MPa, and the ionisation mode source voltage of 4.5 kV. Nitrogen was applied as the curtain and auxiliary gas. For each compound the optimum conditions of Multiple Reaction Mode (MRM) were determined in the infusion mode. Standard solutions were infused into the electrospray source via 50
Compounds were separated on two chromatographic columns: Atlantis-HILIC column (100 × 2.1 mm, 3.0
Dried plant of
The
Polarity of studied flavonoids.
Compound | Abbreviation |
|
---|---|---|
Chrysin | CHR | 2.943 |
Apigenin | AP | 2.463 |
Luteolin | LUT | 1.974 |
Genistein | GEN | 2.268 |
Kaempferol | KAM | 2.172 |
Myricetin | MYR | 1.392 |
Rutin | RUT | −1.063 |
Catechin | CAT | −1.370 |
The effect of organic solvent content in the mobile phase on the retention of flavonoids was investigated using acetonitrile and methanol. The study was carried out in the presence of ammonium formate at pH 7.0, which refers to the aqueous portion of the mobile phase. It should be noticed that the apparent pH value of the mobile phase containing high percentage of organic solvent is different from that of the water component [
The retention factor as a function of ACN or MeOH content in the mobile phase. (a) Bare silica column; (b) ZIC column.
The higher content of both applied organic solvent in the mobile phase enhances the retention of studied flavonoids on ZIC-HILIC column. At increasing concentration of acetonitrile, water is adsorbed more strongly on the surface of the polar stationary phase. The more hydrophilic the analytes are, the more the partitioning equilibrium is shifted towards the adsorbed water layer on the stationary phase, and the more analytes are retained. It can be clearly seen on the example of rutin. For this flavonoid the retention factor increases rapidly with the increase of ACN. Its retention time is longer than 120 min, when mobile phase containing 95% of ACN is applied. Such a high value of retention factor for rutin can be shortened when ACN is replaced with MeOH (Figure
The retention time of rutin at different concentration of ammonium formate (95% ACN v/v pH 7).
Generally, the bare silica packing consistently delivered lower retention factors. The U-shaped curves are obtained when solute retention factors are plotted versus the ACN content (Figure
Since hydrophilic interactions are one of the main mechanisms which governs the retention in the HILIC mode, the hydrophilic properties of a compound should determine its behavior. We examined the relationships between
The relationship between
The elution order of tested flavonoids agrees with their polarity when ZIC column was used. Replacing ACN with MeOH does not result in a change of this retention order. These observations may suggest that retention of analytes is highly dependent on their polarity. According to HILIC theory, the partition of analytes between water enriched layer and less polar eluent is the main retention mechanism. However other interactions such as hydrogen bonding, dipole-dipole, or electrostatic interactions can affect the separation. For ZIC column the correlation coefficients of the relationship
In the case of silica column, the elution order obtained for both organic solvents differs from the theoretical one. The biggest changes were observed when MeOH was applied. The retention order is as follows: KAT ~ GEN < AP < RUT < KAM < MYR < CHR < LUT. The most polar compound catechin (based on the values of
The comparison of HILIC and RP mode was also carried out as reversed-phase mode of the most popular chromatographic mode for flavonoids analysis. Two RP columns were employed—typical fully porous Luna C18 (10 cm × 4.6 mm × 5 Am) and core shell Kinetex C18 (100 × 2.1 mm × 2.6 Am). The composition of the mobile phase in RP mode was ACN/H2O (20/80, v/v) to obtain similar retention factors. Figure
Extracted ion chromatograms of luteolin on HILIC and RP columns under isocratic elution. Eluent for HILIC column-ACN/H2O (95/5%, v/v), for Luna-C18 ACN/water (45/55%, v/v), and for C18: ACN/water (20/80%, v/v).
The ZIC-HILIC column showed higher retention of the studied analytes; thus, it was applied for quantification of flavonoids. The binary mobile phase composed of acetonitrile and ammonium formate in the gradient mode starting from 98% (v/v) of ACN has been applied to elute all unknown compounds present in natural samples.
The validation method was performed evaluating linearity of the working range, the limit of detection. The linearity of the working range was evaluated by the construction of calibration curves using linear least squares regression. The calibration curve of six points in triplicate was established in the concentration range of 0.5–25 mg L−1. The correlation coefficients (
Validation parameters for flavonoids analysis. Conditions with ACN/ammonium formate as an eluent.
LOD |
LOQ |
Slope |
| |
---|---|---|---|---|
Apigenin | 0,01 | 0,03 | 109168 | 0,983 |
Genistein | 0,01 | 0,03 | 2730000 | 0,994 |
Hesperetin | 0,01 | 0,03 | 60385 | 0,985 |
Quercetrin | 0,01 | 0,03 | 448000 | 0,990 |
Luteolin | 0,01 | 0,03 | 1530000 | 0,993 |
Myricetin | 0,01 | 0,03 | 128100 | 0,990 |
Naringenin | 0,01 | 0,03 | 7314000 | 0,998 |
Rutin | 0,05 | 0,06 | 48669 | 0,980 |
Figure
The content of flavonoids in
Compound | Content |
---|---|
Genistein | 58.7 ± 1.22 |
Luteolin | 67.5 ± 1.42 |
Naringenin | 53.0 ± 2.60 |
Quercetin | 34.6 ± 1.58 |
Myricetin | 3.86 ± 0.17 |
Apigenin | 3.00 ± 0.15 |
Quercitrin | 1.25 ± 0.06 |
ESI-MS spectra of selected compounds in the
To check the accuracy of the method, recovery experiments were performed. Suitable amounts of each detected compound were added to the sample and then analyzed, using proposed method. The obtained recoveries were in ranged from 81 to 95%. The example chromatograms of apigenin and naringenin are presented in Figure
The recovery of apigenin and naringenin in
HILIC chromatography can be an alternative separation technique in the analysis of flavonoids. Its potential can be shown mainly when mass spectrometry is applied for the detection. In such case main attention is paid to the limit of detection but not to good separation of the compounds. HILIC mode offers lower limit of detection in comparison to the reversed mode. It is mainly due to the high content of organic solvent in the mobile phase, which enhances the ionisation in the ion source of mass spectrometer. The potential of this technique can be applied at the sample preparation step, mainly when SPE is considered to be used. Analytes can be eluted from SPE column using organic solvent and then directly injected to HPLC system. All these advantages compensate the asymmetry of the peaks and weaknesses of flavonoids separation.
In this study the effect of organic solvent type in the mobile phase was tested. ACN and MeOH were applied for the separation of flavonoids performed on Atlantis-HILIC and ZIC-HILIC columns. Using MeOH as a mobile phase component results in decreasing of the retention time for more polar compounds. On the other hand it enhances the retention of less polar analytes. The retention factors for flavonoids increase rapidly with the increase of acetonitrile content. The high organic content of the mobile phase enhances the ionisation efficiency in the ion source and simultaneously enhances the sensitivity of detector.
Compared to other chromatographic modes, HILIC could offer also several significant benefits. The highly organic content provides also lower back-pressure to enable fast separations under higher flow rates or to use columns with small particles. HILIC mode is preferred for RP solid phase extracts obtained in a sample preparation step as no evaporation or reconstruction steps are needed. However, HILIC is more influenced by the composition of sample diluent and the retention characteristics are less predictable due to the more complex retention mechanism.
The authors declare that they have no competing interests.
The authors would like to thank the Structural Research Laboratory (SRL) at the Department of Chemistry of University of Warsaw for making possible HPLC-MS measurements. SRL has been established with financial support from European Regional Development Fund in the Sectorial Operational Programme “Improvement of the competitiveness of Enterprises, years 2004-2005” Project no. WPK_1/1.4.3./1/2004/72/72/165/2005/U.