Achiral SFC: Development of an Orthogonal SFC Method for Mometasone Furoate Impurity Analysis Zhenyu Wang Merck Research Laboratories NJCG 2013
Outline SFC chronicle The journey of a half century Re-birth of SFC Chiral SFC in pharmaceutical industry New opportunity of SFC era Achiral SFC for complex analysis Summary
The chronicle of SFC 1962-1981: Packed Column SFC 1962: Ernst Klesper High pressure GC
The chronicle of SFC 1981-1995: Capillary SFC 1981: M. Novotny, and M.L. Lee. et.al.
1980s: the enthusiasm SFC should exhibit the combined advantages of GC and HPLC. SFC allows the separation of materials which are thermally labile and of much higher molecular weight than is possible using GC. High diffusivities, low viscosities SF CO 2 provide higher flow rates, shorter analysis, and higher efficiency separations.
1990s: the frustrations The extent to which SFC will be useful for relatively polar materials is still an open question and a matter for much future research. A decade debate: Open tubular SFC vs. Packed column SFC The winner: Packed column SFC: user friendly, quantitative injection, and apply to higher polarity compounds
The chronicle of SFC Late 1990s-present: Re-birth of parked column SFC Application: Chiral separation and purification Polar compounds even peptides Hardware: Dedicated SFC stationary phases New generation SFC instrumentation Conception: Green technology Cost effective
Key driver of SFC s Renaissance Chiral separation and purification in Pharm Ref.: Jeff Elleraas, Pfizer
Chiral Prep SFC : Stacked injections 16.5 g in 7 hours 2,100 2,000 1,900 1,800 1,700 1,600 1,500 1,400 1,300 1,200 1,100 1,000 900 800 700 600 500 400 300 200 100 0-100 0 50 100 150 200 250 300 350 400
Chiral SFC for non-chiral compound separation Nobiletin (NOB): anti-inflammatory agent NOB is metabolized by hepatic p450 enzymes yielding 3 -demethyl- NOB and 4 -demethyl-nob No separation of the two metabolites on RPLC OMe OMe MeO O OH MeO MeO OMe OMe O O OMe OMe P450 MeO MeO OMe O OMe O 3 -demethyl-nob OH OMe Nobiletin (NOB) MeO OMe O 4 -demethyl-nob
1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 RPLC separation of 3 -DNOB and 4 -DNOB Z5800503 100 3.87 ELSD An1 7.70e5 No Separation on Reversed Phase LC % 0 Time Mass Spectrum: 3 -DNOB 4 -DNOB
3 -DNOB and 4 -DNOB: Chiral LC vs. Chiral SFC Column: Chiralpak AD HPLC: 40/60 hexane/ethanol, 1.0 ml/min LC Column: Chiralpak AD Rs: 1.7 SFC: 20% MeOH as modifier, 2.0 ml/min, CO 2 100 bar, 30ºC 50 40 30 20 10 0 0 AD0420055.DATA [1] mau 2 4 6 8 10 SFC Rs: 22.5 12 14 16 18 RT [min] 20
Chiral SFC for non-chiral compound separation ID and quantitation of NOB metabolites in mice urine 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0-2 0 AD04200516.DATA [1] mau 1 2 3 4 5 6 7 8 9 4 -Demethyl-NOB (major metabolite) 28.9 µg/ml 10 NOB (parent) 11 12 13 14 15 16 17 18 19 20 21 22 23 RT [min] 3 -Demethyl-NOB (minor metabolite) 24 25 SFC: AD column, 20% MeOH as modifier, 2.0 ml/min, CO 2 100 bar, 30ºC Z. Wang, S. Li, Biomed. Chromatogr., 20 (2006) 1206-1215
The new opportunity of SFC era: Achiral separation of complex analytes SFC achiral method development General practice for method development Mometasone furoate case study SFC achiral method validation Improve method sensitivity SFC sample prep for complex matrix (e.g. bio-fluids, formulated drugs)
Mometasone furoate franchise A highly potent glucocorticoid anti-inflammatory agent The active ingredient of several drug products
Mometasone furoate and its major impurities Cl O HO O O O Cl O Mometasone furoate
Current RPLC method for MF impurity analysis What if I use SFC 0.080 0.075 0.070 0.065 0.060 0.055 0.050 24.825 Sample: MF with spiked impurities Column: Ultrasphere ODS (250 4.6 mm, 5 µm) Mobile phase A: Water/Mobile phase B: Acetonitrile Gradient: 42% B to 52% B in 60 min Flow rate: 1.5 ml/min 0.045 AU 0.040 0.035 8.165 11.390 0.030 0.025 0.020 0.015 18.688 22.500 27.487 32.847 36.929 40.322 42 min! 0.010 0.005 0.000 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.00 36.00 38.00 40.00 42.00 44.00 46.00 48.00 50.00 Minutes *Method slightly modified from USP MF monograph
SFC & HPLC instrument and software All SFC experiments were performed on: TharSFC Method Station Analytical System Solvent selector Column selector Waters 2998 PDA detector Instrument control and data collection: Empower 2 All HPLC experiments were performed on: Alliance 2690 HPLC System equipped with 2996 PDA detector Instrument control and data collection: Empower 2
Achiral SFC method development Fine tuning 2 Screening: P, T, Additive Preliminary Screening: Stationary Phase & Modifier
SFC achiral column screening Most time-consuming step Retention mechanism is not fully understood Normal phase, Reversed phase, Mixed mode Simplify column screening
SFC column selection: QSRR: Quantitative Structure Retention Relationship West, C.; Lesellier, E., J. Chromatogr. A, 2008, 1203, 105.
Primary screening: Column and Modifier Interaction Plot for Selectivity Data Means 2-PrOH EtOH MeOH Column 1.100 1.075 1.050 Column 2-EP Cyano Si 1.025 1.100 1.075 1.050 Modifier 1.000 Modifier 2-PrOH EtOH MeOH 1.025 1.000 2-EP Cyano Si
Separation on Silica column with Methanol as modifier 0.080 0.070 9.078 0.060 MF 0.050 AU 0.040 0.030 0.020 0.010 8 4.841 5.095 5 7 5.912 1 7.730 6 8.347 8.649 2 3 4 9.546 11.489 0.000 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 Minutes
Achiral SFC method development Step Two Fine tuning 2 Screening: P, T, Additive Preliminary Screening: Stationary Phase & Modifier
Impact of temperature and pressure 0.080 0.070 0.060 9.078 30 ºC, 100 bar 0.050 AU 0.040 0.030 0.020 0.010 4.841 5.095 5.912 7.730 8.347 8.649 9.546 11.489 0.000 0.00 2.00 4.00 6.0 0 8.00 10.00 12.00 14.00 16.00 18.00 20.00 M in u te s 0.070 9.880 35 ºC, 100 bar 0.060 0.050 0.040 AU 0.030 0.020 0.010 5.474 5.792 6.669 8.538 9.244 9.442 10.227 12.343 0.000 0.00 2.00 4.00 6.0 0 8.00 10.00 12.00 14.00 16.00 18.00 20.00 Minutes Silica: 250 4.6 mm, 5 µm, Kromasil Mobile phase: 4 ml/min, CO 2 Gradient: 5% MeOH to 20% MeOH in 15 min
Impact of temperature and pressure (Cont d) 0.055 0.050 0.045 0.040 10.656 40 ºC, 100 bar 0.035 0.030 AU 0.025 0.020 0.015 0.010 0.005 6.019 6.335 7.248 9.250 10.177 13.286 0.000-0.005 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 M in u te s Silica: 250 4.6 mm, 5 µm, Kromasil Mobile phase: 4 ml/min, CO 2 Gradient: 5% MeOH to 20% MeOH in 15 min
Impact of temperature and pressure (Cont d) 0.080 0.070 0.060 9.078 30 ºC, 100 bar 0.050 AU 0.040 0.030 0.020 0.010 4.841 5.095 5.912 7.730 8.347 8.649 9.546 11.489 0.000 0.00 2.00 4.00 6.0 0 8.00 10.00 12.00 14.00 16.00 18.00 20.00 M in u te s 0.100 0.090 0.080 7.875 30 ºC, 150 bar 0.070 0.060 AU 0.050 0.040 0.030 0.020 0.010 4.108 4.339 4.797 6.566 6.605 6.773 7.479 8.483 10.261 0.000 0.00 2.00 4.00 6.0 0 8.00 10.00 12.00 14.00 16.00 18.00 20.00 Minutes Silica: 250 4.6 mm, 5 µm, Kromasil Mobile phase: 4 ml/min, CO 2 Gradient: 5% MeOH to 20% MeOH in 15 min
0.080 0.070 0.060 MF 9.078 SFC 0.050 AU 0.040 0.030 0.020 0.010 8 4.841 5 5.095 7 5.912 1 7.730 6 2 8.347 8.649 3 9.546 4 11.489 0.000 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00 15.00 Minutes 0.080 0.075 0.070 0.065 0.060 0.055 0.050 0.045 24.825 RPLC AU 0.040 0.035 8.165 11.390 0.030 0.025 0.020 0.015 18.688 22.500 27.487 32.847 36.929 40.322 0.010 0.005 0.000 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.00 36.00 38.00 40.00 42.00 44.00 46.00 48.00 50.00 Minutes
Good news for pharmaceutical chromatographers One of the most challenges for us: To develop stability indicating method to monitor DS & DP Most of stability indicating methods are RPLC based How to confirm method s specificity (detection or separation) SFC can be the 2 nd method with true orthogonal selectivity I get new tool
Stability indicating method: RPLC or SFC? Which one? Why don t we use SFC for the primary stability indicating method? But Is SFC sensitive enough?
Sensitivity: key to drug impurity analysis Drug safety and quality control Reporting threshold: 0.05% impurity in Drug Substance ( 2 g/day), ICH Q3A Reporting threshold: 0.1% degradation product in Drug Product ( 1 g/day), ICH Q3B
Why SFC-UV is less sensitive (vs. HPLC-UV) Three main sources of noise: Electronic: noise from detector system Mechanical: BPR, pump Thermal: endothermic process during depressurization Anton, K.et.al. Analusis, 1999, 27, 691 Helmy, R. et.al. Chirality, 2007, 19, 787 Wang, Z, et.al, Am. Pharm. Rev, 2009, 5, 59
How to get better sensitivity on SFC-UV Software filtering: reduce non-wavelength dependent noise 1,2 Hardware modification: reduce mechanical and thermal noise 3 Next generation SFC 1 Chen, R., LC-GC Application Notebook, 2009, Sep 2. Wang, Z, et.al, J. Chromatogr. A, 2011, 1218, 2311 3. Helmy, R. et.al. Chirality, 2007, 19, 787
Reference Wavelength Compensation to increase S/N PDA detector settings Without wavelength compensation With wavelength compensation Improvement in sensitivity Sampling rate Bandwidth Filter constant Peak width S/N a Peak width S/N b (S/N b )/(S/N a ) i 5 2.4 Slow 0.072 67 0.072 151 2.2 ii* 5 2.4 Normal 0.066 44 0.066 139 3.2 iii* 5 3.6 Slow 0.073 57 0.072 163 2.9 iv 5 3.6 Normal 0.066 62 0.066 122 2 v 5 4.8 Slow 0.072 62 0.072 142 2.3 vi 5 4.8 Normal 0.066 54 0.067 115 2.1 vii* 2 2.4 Slow 0.108 62 0.108 251 4 viii 2 2.4 Normal 0.077 71 0.077 170 2.4 ix 2 3.6 Slow 0.108 52 0.108 190 3.7 x 2 3.6 Normal 0.077 47 0.077 159 3.4 xi 2 4.8 Slow 0.109 63 0.108 175 2.8 xii 2 4.8 Normal 0.077 43 0.077 162 3.8 Detection wavelength: 245 nm Compensation wavelength: 400-450 nm
AU Ref. Wavelength Compensation to increase S/N (Cont d) 0.080 0.070 0.060 0.050 0.040 0.030 0.020 ii Sampling rate: 5 Bandwidth: 2.4 Filter cons.: Normal Peak width: 0.066 S/N: 139 0.010 0.000 AU 0.080.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00 15.00 Minutes 0.070 0.060 0.050 0.040 0.030 0.020 S/N x3 increase iii Sampling rate: 5 Bandwidth: 3.6 Filter cons.: Slow Peak width: 0.072 S/N: 163 0.010 0.000 AU 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00 Minutes 0.060 0.050 0.040 0.030 0.020 vii Sampling rate: 2 Bandwidth: 2.4 Filter cons.: Slow Peak width: 0.108 S/N: 251 0.010 0.000 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00 Minutes
Comparison of RPLC and SFC method validation results RP-HPLC method SFC method Sample concentration 0.2 mg/ml 2.0 mg/ml Linearity 0.9999 0.9999 Accuracy Assay level a 99.1% - 100.7% 99.8% -101.6% Impurity level b 96.6% - 115.4% c 88.3% - 104.7% c Precision Assay level a 0.4% 0.7% Impurity level b 1.9% - 5.0% 1.4% - 5.4% Limit of Quantitation 0.05% (or 0.1 µg/ml) 0.05% (or 1.0 µg/ml) a. Six preparations (n = 6) b. Six preparations of spiked individual impurities c. The average recovery of each impurity was reported. Z. Wang, J. Chromatogr. A, 2011, 1218, 2311-2319
Sensitivity is not a major challenge for SFC anymore Can new instrumentation broaden SFC to complex sample analysis? Such as formulated drugs.
Using SFC in complex formulation analysis The Big Fact: SFC is rarely used for drug product analysis Chromatograph used for drug product analysis: RPLC, nearly 100% The Challenge: Drug product sample prep: working for RPLC working for SFC SFC prefers neat organic as sample solvent Direct injection of aqueous containing sample on SFC may result in: - Freezing of aqueous solutions - Precipitation of samples - Deteriorated separations
SFC Sample Preparation Strategies for Drug Products For solid dosage forms: - Dissolve sample in organic solvent - Filter to remove un-dissolved residue - Assay by SFC Extraction robustness need to be assessed For solutions and suspensions: - Direct injection might work (not recommended as 1 st choice) - Use Solid Phase Extraction or Liquid-Liquid Extraction
SPE-SFC: for an aqueous formulation SPE Cartridge Screen Study 120 SFC Recovery (%) (%) 100 80 60 40 20 0 Strata X C18 E C8 Phenyl Methanol DMSO Acetonitrile SDB-L Z. Wang, Am. Pharm. Rev, 2013, 16(3), 28-25
Summary SFC dominated chiral separation and chiral purification The advantages of achiral SFC have not been fully recognized: Orthogonal selectivity Economical Faster SFC friendly sample prep is important for complex analysis (e.g. for formulated drugs)