Method Development Tools Used in Hydrophilic Interaction Chromatography (HILIC) for the analysis of Polar Basic Pharmaceuticals Min Seok Chang a , Kevin A. Schug b , and Ritu Arora a a: Varian, Inc., 25200 Commercentre Drive, Lake Forest, CA 92630, USA b: Department of Chemistry and Biochemistry, University of Texas at Arlington, Arlington, TX 76019, USA Inspiring Excellence ™ Ionic Strength - Salt Concentration Effect Solvent Strength Effect Abstract The analysis of hydrophilic compounds is a major challenge to any researcher. Different modes of chromatography are employed to gain retention of extremely polar compounds, e.g. ion exchange, ion-pair reversed phase, or polar functionalized chemistries. Each mode suffers from different limitations, be it high ionic strength, incompatibility with MS detection, insufficient retention, or time- consuming sample preparation. HILIC (hydrophilic interaction chromatography) has been drawing a great attention in various industries including pharmaceutical, environmental, and food industry for the separation for hydrophilic compounds. Since HILIC-mode columns are inherently different from conventionally used reversed phase or normal phase columns, a better understanding of the retention mechanism involved and guidelines for method development are essential for their proper use and maintenance. The authors have screened several bonded phases for use in HILIC mode. Out of several options available, we have selected a unique diol chemistry (Varian Prototype HILIC Diol) for investigating different parameters used in method development for the analysis of popular polar basic pharmaceuticals. Selectivity options have been explored using different tools such as organic:aqueous modifier ratios, solvent strength, pH, ionic strength in terms of salt concentration of a single buffer system and comparison of different buffer types. Besides these, selectivity differences between HILIC- mode and C18 columns have also been looked into to demonstrate the benefit of HILIC-mode columns over conventionally used reverse phase column for polar compound analysis. Data will be presented illustrating the effect of different method development options available to an end-user working with polar analytes and diol bonded phases. 5 4.8 4.6 4.4 4.2 4 3.8 3.6 3.4 3.2 3 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 250 200 150 100 50 0 RT [min] 10mM Pursuit XRs Diol 3u_90B_0@4_beta blocker mix_3.DATA mAU 5 4.8 4.6 4.4 4.2 4 3.8 3.6 3.4 3.2 3 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 250 200 150 100 50 0 RT [min] 50mM Pursuit XRs Diol 3u_90B_0@4_beta blocker mix_4.DATA mAU 5 4.8 4.6 4.4 4.2 4 3.8 3.6 3.4 3.2 3 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 250 200 150 100 50 0 RT [min] 100mM Pursuit XRs Diol 3u_90B_0@4_beta blocker mix_4.DATA mAU 5 4.8 4.6 4.4 4.2 4 3.8 3.6 3.4 3.2 3 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 250 200 150 100 50 0 RT [min] 250mM Pursuit XRs Diol 3u_90B_0@4_beta blocker mix_5.DATA mAU Mobile Phase A: 10, 50, 100, 250 mM ammonium formate (pH = 3.0) Column: Prototype HILIC Diol 3 , 100 X 2.0 mm Mobile Phase B: Acetonitrile Flow rate: 0.4 mL/min 90% B Isocratic Detection: 272 nm Flow rate: 0.4 mL/min 1. Acenaphthene (t 0 marker) 2. Pindolol 3. Practolol 4. Atenolol log P 3.92 1.75 0.79 0.16 Conc. 0.2 mg/mL 0.1 mg/mL 0.3 mg/mL 1.1 mg/mL 1. Acenaphthene (t 0 marker) 2. Ascorbic acid 3. Cytosine log P 3.92 -1.85 -1.73 Conc. 0.1 mg/mL 0.7 mg/mL 0.2 mg/mL 6 5.5 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 250 200 150 100 50 0 RT [min] 250mM Pursuit XRs Diol 3u_90_5_5 H2O_0@4_beta blocker mix_4.DATA mAU 5:90:5 = Buffer:ACN:H 2 O 1 2 3 4 6 5.5 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 250 200 150 100 50 0 RT [min] 250mM Pursuit XRs Diol 3u_90_5_5 MeOH_0@4_beta blocker mix_4.DATA mAU 5:90:5 = Buffer:ACN:MeOH 1 2 3 4 6 5.5 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 150 100 50 0 RT [min] 250mM Pursuit XRs Diol 3u_90_5_5 IPA_0@4_beta blocker mix_6.DATA mAU 5:90:5 = Buffer:ACN:IPA 1 2 3 4 Column: Prototype HILIC Diol 3 , 100 X 2.0 mm Mobile Phase A: 50:50 premix of 250 mM ammonium formate (pH = 3.0) and H 2 O Mobile Phase B: ACN C: organic modifier (MeOH, IPA, and H 2 O) A:B:C = 5:90:5 Flow Rate: 0.4 mL/min Temperature: Ambient Detection: 272 nm Samples: 1. Acenaphthene 2. Pindolol 3. Practolol 4. Atenolol Fig 3. Compounds with positive log P values Fig 4. Compounds with negative log P values Solvent strength in HILIC mode THF < Acetone < ACN < IPA < EtOH < MeOH < H 2 O (H 2 O is the strongest solvent in HILIC mode) Higher ionic strength possibly suppresses electrostatic interactions with the silica, as it removes some secondary interaction modes with the analytes due to a shielding effect. Retention times are lowered, but not too much, especially for the ones that are more hydrophobic (pindolol and practolol) which are less resistant to changes in buffer strength. Interaction with the diol dominates at higher buffer strength. The hydrophilic molecules are induced to partition into a water rich layer at the stationary phase surface that has a higher proportion of salt associated with it. Interactions with the higher concentration of salt in the “immobilized” water layer may cause added retention. Retention time of analytes changes according to the solvent strength of mobile phase. The use of a relatively weak solvent like IPA increases retention of all analytes. Complementary Selectivity to Reverse Phase Prototype HILIC Diol 3 100 X 2.0 mm 90:10 ACN:buffer Isocratic 1073 psi Pursuit XRs C18 3 100 X 2.0 mm 60:40 ACN:buffer Isocratic 2075 psi 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 200 150 100 50 0 RT [min] 100mM Pursuit XRs 3 C18_60B_0@4_atenolol_naphthalene_mix_4.DATA mAU 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 700 600 500 400 300 200 100 0 RT [min] 100mM Pursuit XRs Diol 3u_90B_0@4_atenolol_naphthalene_mix_5.DATA mAU 1 1 2 2 Columns: Prototype HILIC Diol 3 , 100 X 2.0 mm Pursuit XRs C18 3 , 100 X 2.0 mm Mobile Phase A: 100 mM ammonium formate (pH = 3.0) B: Acetonitrile 90% B Isocratic (for HILIC, Prototype HILIC Diol 3 , 100 X 2.0 mm) 60% B Isocratic (for RP, Pursuit XRs C18 3 , 100 X 2.0 mm) Flow rate: 0.4 mL/min Detection: 272 nm Temperature: ambient 1. Naphthalene 2. Atenolol log P 3.30 0.16 Conc. 0.2 mg/mL 0.8 mg/mL Complementary selectivity between reverse phase column and HILIC column Ionic Strength Effect – Different cation / anion effect in buffer ACN vs. MeOH MS Application a) Mobile Phase Effect 7 6.5 6 5.5 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 250 200 150 100 50 0 RT [min] mAU Mobile Phase A: 100 mM ammonium formate (pH = 3.0) 1 2 3 4 7 6.5 6 5.5 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 300 250 200 150 100 50 0 RT [min] mAU Mobile Phase A: Formic acid (pH = 3.0) 1 2 3 4 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 300 250 200 150 100 50 0 RT [min] mAU 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 300 250 200 150 100 50 0 RT [min] mAU Mobile Phase A: 100 mM ammonium formate (pH = 4.0) Mobile Phase A: 100 mM ammonium acetate (pH = 4.0) 1 2 3 4 1 2 3 4 Column: Prototype HILIC Diol 3 , 100 X 2.0 mm Mobile Phase A: Various (Specified in the chromatograms.) 100 mM ammonium acetate (pH = 4.0) 100 mM ammonium formate (pH = 4.0) 100 mM ammonium formate (pH = 3.0) Formic acid buffer (pH = 3.0) Mobile Phase B: Acetonitrile 90% B Isocratic Nicotinamide Pindolol Salbutamol Atenolol Mobile Phase A: 0.1% formic acid in H 2 O Mobile Phase B: ACN Sample solvent: 90:10 ACN:H 2 O (10% aqueous) Data was irreproducible due to low ionic strength of mobile phase. Mobile Phase A: 0.1% formic acid in H 2 O Mobile Phase B: 0.1% formic acid in ACN Sample solvent: 90:10 ACN:H 2 O (10% aqueous) Data was irreproducible due to low ionic strength of mobile phase. Typical MS mobile phase A and B for reverse phase separation Mobile Phase A: 20 mM ammonium formate in H 2 O (pH = 3.0) Mobile Phase B: ACN Sample solvent: 90:10 ACN:H 2 O (10% aqueous) Data was reproducible . Ideal HILIC mobile phase condition on prototype HILIC Diol Salbutamol Pindolol Atenolol Nicotinamide Log P 0.64 1.75 0.16 -0.37 Structure Conc. (ng/mL) 29.4 29.4 117.6 411.8 Sample solvent 90:10 ACN:H 2 O (when mobile phase B was ACN-based) or 90:10 MeOH:H 2 O (when mobile phase B was MeOH-based) Column: Prototype HILIC Diol 3 100 X 2.0 mm Mobile phase A: Specified in the chromatograms. Mobile phase B: Specified in the chromatograms. A:B 10:90 Isocratic Flow rate: 0.4 mL/min Detection: Varian 320 LC/MS/MS Temperature: Ambient Inj. Vol.: 20 L THF < Acetone < ACN < IPA < EtOH < MeOH < H 2 O (H 2 O is the strongest solvent in HILIC mode) ACN Based Organic Mobile Phase MeOH Based Organic Mobile Phase Column: Prototype HILIC Diol 3 100 X 2.0 mm Mobile Phase A: 20 mM ammonium formate in H 2 O (pH = 3.0) Mobile Phase B: MeOH Sample solvent: 90:10 ACN:H 2 O (10% aqueous) Data were reproducible. Poor peak shape. Significantly reduced retention due to the presence of a strong organic solvent (MeOH) in mobile phase B. Nicotinamide Pindolol Salbutamol Atenolol Formic acid alone cannot produce good peak shapes for basic analytes due to a) low ionic strength, and b) possible cationic exchange effect with the sorbent (3 - 4). Both basic analytes and silanol groups are ionized under HILIC conditions with 0.1% formic acid. The hydronium cation generated by formic acid is not an effective competing cation for the ionized silanol groups. Therefore, the ion exchange component of the cationic basic analyte is more influential on overall separation performance. Peak shapes are greatly improved if a cation with a higher affinity for the ionized silanols, such as ammonium, is added to the mobile phase. Flow rate: 0.4 mL/min Detection: 272 nm, Temperature: Ambient Samples: 1. Acenaphthene, 2. Pindolol 3. Practolol, 4. Atenolol Comparison with Competitor’s Column a) HILIC-MS Application Columns: Shown in the chromatograms Mobile Phase A: 20 mM ammonium formate in H 2 O (pH = 3.0) Mobile Phase B: ACN A:B 10:90 Isocratic Flow rate: 0.4 mL/min Temperature: Ambient Sample solvent: 90:10 ACN:H 2 O (10% aqueous) Prototype HILIC Diol 3 100 X 2.0 mm 100Å Competitor’s HILIC diol 3 100 X 2.0 mm 200Å Nicotinamide Pindolol Salbutamol Atenolol b) Sample Solvent Effect Columns: Prototype HILIC Diol 3 100 X 2.0 mm Mobile Phase A: 20 mM ammonium formate (pH = 3.0) Mobile Phase B: ACN A:B 10:90 Isocratic Flow rate: 0.4 mL/min Temp: ambient Detection: Varian 320 LC/MS/MS Samples: Salbutamol, Pindolol, Atenolol, Nicotinamide Sample Solvents: Seven different sample solvents were tested and they are specified in the chromatograms. In reverse phase separation, it is desirable to have sample solvent weaker than initial mobile phase conditions. In HILIC separation, it is a must to have sample solvent equal to or weaker than initial mobile phase conditions. Evaporation / reconstitution or dilution with initial mobile phase may be necessary to produce good peak shapes. The more aqueous content in the sample solvent, poorer the peak shapes due to the strong elution strength of water in HILIC mode. ACN Based Sample Solvent Sample solvent 100% ACN, 0% H 2 O Sample solvent 90% ACN, 10% H 2 O Sample solvent 75% ACN, 25% H 2 O Sample solvent 10% ACN, 90% H 2 O Sample solvent 100% MeOH, 0% H 2 O Sample solvent 90% MeOH, 10% H 2 O Sample solvent 10% MeOH, 90% H 2 O Prototype HILIC Diol column showed longer retention and better separation compared to competitor’s HILIC column with diol chemistry MeOH Based Sample Solvent MeOH is too strong to be used as a sample solvent. Poor peak shapes were produced as expected. Conclusion • Varian prototype HILIC Diol showed complementary selectivity to reverse phase column. • Solvent strength in HILIC separation is THF < Acetone < ACN < IPA < EtOH < MeOH < H 2 O and it was confirmed by the data generated by using constant mobile phase composition of solvents differing in solvent strengths. • Ionic strength • Buffer types with different cations / anions • Bigger anion offers longer retention due to stronger ion-pairing effect and induction of greater partitioning of analytes with the immobilized water layer on the stationary phase. Besides, its greater hydrophobicity allows greater interaction of analytes with the stationary phase leading to increased retention. • Formic acid alone cannot provide good peak shapes for the basic compounds due to lack of ionic strength leading to a cation-exchange effect. • Salt concentration • Retention time of analytes with positive log P values tends to decrease with increasing salt concentrations. • Retention time of analytes with negative log P values tends to increase with increasing salt concentrations. • MS compatiblility of prototype HILIC Diol: • Mobile phase • MeOH is a strong solvent in HILIC separation, hence, should be avoided due to drastic reduction in retention times observed. • For better peak shapes and reproducibility, addition of salt (~ 20mM) to the mobile phase is needed. • Sample solvent • Sample solvent must be equal to or weaker than initial mobile phase conditions for good peak shapes. • MeOH is not recommended to be used as a sample solvent since it is a strong solvent in HILIC separation and can generate poor peak shapes. • Competitive data indicate prototype HILIC Diol yields a) longer retention & better separation and b) better reproducibility possibly due to difference in pore size (prototype HILIC Diol is 100Å while competitor’s HILIC with diol chemistry is 200Å.) Acknowledgement • Dr. Huqun Liu, Varian, for synthesis of a very competitive Prototype HILIC Diol phase • Samuel H. Yang, University of Texas at Arlington, in providing MS application of amino acids • William Hudson, Varian, for help in generating MS sensitivity data • Hema Chauhan, Varian, for help in screening of columns for HILIC applications Introduction HILIC columns can provide Improved retention of polar analytes that would be hard to retain on RP columns (1) Complementary selectivity to reversed-phase Improved MS sensitivity due to high organic content of mobile phase (1) Increased sample throughput, as direct injection of SPE eluates is possible without solvent evaporation and reconstitution Higher flow rates and fast separations due to low viscosity eluents Unique benefits for preparative chromatography for purifying samples with poor water solubility A good alternative for compounds with bad carryover in RP, the high organic eluents help with carryover issues To fully utilize the benefits of using HILIC columns, proper method development guidance and related maintenance are necessary. Varian scientists have developed dedicated method development tools for HILIC applications with regards to the following items: Organic / aqueous mobile phase ratios Ionic strength in mobile phase in terms of salt concentration types of buffer counter ions Solvent strength with different organic mobile phase modifications ACN vs. MeOH Mobile phase Sample solvent b) Reproducibility Test under Isocratic Condition (48-hr Immersion Test) Columns: Prototype HILIC Diol 3 m 100 X 2.0 mm 100Å Competitor’s HILC diol 3 m 100 X 2.0 mm 200Å Mobile Phase A: 10mM ammonium acetate Mobile Phase B: ACN 90% B Isocratic for 10 min Flow rate: 0.2 mL/min Temperature: Ambient Detection: 220 nm Sample: Mixture of neutral, basic, and acidic compounds Columns were stored in mobile phase for 48 hrs after first analysis. After being stored for 48 hrs in mobile phase, the same analysis was performed for inter-day immersion reproducibility. 1. Acenaphthene 2. Cytosine 3. Ibuprofen log P 3.92 -1.73 3.97 Structure Prototype HILIC Diol Competitor HILIC Diol Competitor’s HILIC (diol) 3 m 100 X 2.0 mm Retention change: -4.6 % Prototype HILIC Diol 3 m 100 X 2.0 mm Retention change: 2.8 % Day 1 1 2 3 Day 1 1 2 3 Day 3 1 2 3 Day 3 1 2 3 Fig 1. Complementary selectivity between HILIC and reverse phase columns (Unlike reverse phase, non-polar compounds elute at t 0 while polar compounds are retained on prototype HILIC Diol.) Column: Prototype HILIC Diol, 3 , 100 X 2.0 mm Mobile Phase A: 100 mM ammonium formate (pH = 3.0) Mobile Phase B: Acetonitrile Flow rate: 0.4 mL/min Detection: 270 nm Temperature: Ambient Samples 1. Pindolol 2. Practolol 3. Atenolol As the aqueous content of the mobile phase is increased, the observed retention of the analytes decreases due to the strong elution strength of water in HILIC mode. 7 6.5 6 5.5 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 350 300 250 200 150 100 50 0 RT [min] Pursuit XRs Diol 3u_85B_0@4_pindolol_practolol_atenolol_mix1_1.DATA mAU 85% ACN 15% Buffer Isocratic 3 1 2 7 6.5 6 5.5 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 250 200 150 100 50 0 RT [min] Pursuit XRs Diol 3u_90B_0@4_pindolol_practolol_atenolol_mix2_3.DATA mAU 90% ACN 10% Buffer Isocratic 1 2 3 7 6.5 6 5.5 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 180 160 140 120 100 80 60 40 20 0 RT [min] Pursuit XRs Diol 3u_95B_0@4_pindolol_practolol_atenolol_mix2.DATA mAU 95% ACN 5% Buffer Isocratic 1 2 3 Organic / Aqueous Modifier Ratios Fig 2. Effect of organic / aqueous mobile phase ratio in retention time Fig 5. Anion effect in buffer under same pH condition Fig 6. Cation effect in buffer under same pH condition Longer retention times were observed when ammonium acetate was used compared to ammonium formate, which can be explained by the fact that a) acetate, being more hydrophobic than formate, makes the analytes interact with diol phase more, leading to longer retention, and b) acetate can have a stronger ion-pairing effect, facilitating increased partitioning with the immobilized water layer, resulting in increased retention. Fig 7. Solvent strength effect with different organic phase modifiers Fig 8. MS chromatograms with ACN as organic mobile phase (Sample solvent conditions for a), b), and c) are given on the right.) Fig 9. MS chromatograms with MeOH as organic mobile phase (Compared to Fig 8. c), drastically reduced retention times and poor peak shapes were observed with MeOH as organic mobile phase) a) b) c) Fig 10. MS chromatograms under ideal HILIC mobile phase conditions with samples in ACN-based solvents Fig 11. MS chromatograms under ideal HILIC mobile phase conditions with samples in MeOH-based solvents Fig 12. MS chromatogram comparison with Prototype HILIC Diol and competitor’s HILIC column with diol chemistry References 1. Effect of stationary phase chemistry on selectivity of pharmaceuticals, pesticides, and oligosaccharides in HILIC separations, Pittcon Poster #: 600 – 5P, Ritu Arora, Richard Robinson, Min Seok Chang, and Eugene Chang, Mar 2009 2. Nguyen, H.P., Schug, K.A. J. Sep. Sci. 2008, 31, 1465-1480. 3. Hydrophilic interaction chromatography, David McCalley, Chromatographyonline.com, April 2008 4. Deleterious Effects of Formic Acid without Salt Additives on the HILIC Analysis of Basic Compounds, Phenomenex application note TN-1040 Fig 14. Histogram showing retention time shift from reproducibility test for 48-hr immersion Fig 13. Chromatograms of day 1 and 3 for reproducibility test 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Time (min.) Intensity / 10 6 Cys [M + H] + 122.0 m/z CSA [M - H] - 152.0 m/z Prototype HILIC Diol R = 0.9908 k 1 = 1.672 k 2 = 2.259 HILIC-MS Application of Amino Acids Fig 15. HILIC-MS application of cysteine and cysteine sulfinic acid on prototype HILIC Diol Column: Prototype HILIC Diol 3 100 X 2.0 mm Mobile phase A: 20 mM ammonium acetate Mobile phase B: ACN + 0.5% acetic acid A:B 25:75 Isocratic Flow rate: 0.2 mL/min Temperature: Ambient Samples: 1. Cysteine (1 mM) 2. Cysteine Sulfinic Acid (200 M) Injection volume: 10 l 10 mM NH4COOH (1 mM in flow) 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 50 mM NH4COOH (5 mM in flow) 100 mM NH4COOH (10 mM in flow) 250 mM NH4COOH (25 mM in flow) 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 500 400 300 200 100 0 RT [min] 10mM Pursuit XRs Diol 3u_90B_0@4_immersion mix_3.DATA mAU 1 2 3 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 500 400 300 200 100 0 RT [min] 50mM Pursuit XRs Diol 3u_90B_0@4_immersion mix_3.DATA mAU 1 2 3 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 500 400 300 200 100 0 RT [min] 100mM Pursuit XRs Diol 3u_90B_0@4_immersion mix_3.DATA mAU 1 2 3 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 350 300 250 200 150 100 50 0 RT [min] 250mM Pursuit XRs Diol 3u_90B_0@4_immersion mix_12.DATA mAU 1 2 3 10 mM NH4COOH (1 mM in flow) 50 mM NH4COOH (5 mM in flow) 100 mM NH4COOH (10 mM in flow) 250 mM NH4COOH (25 mM in flow) Formic acid alone cannot provide sufficient ionic strength for polar basic compounds, and addition of salt (~ 20mM) is needed for good peak shapes and reproducible chromatography.