Scaling Up to Preparative Chromatography

Scaling Up to Preparative Chromatography Dr. Shulamit Levin home page: www.forumsci.co.il/hplc Strategy for Preparative Separation Selection of the appropriate mode of chromatography ptimization of the Separation (Stationary phase, Mobile Phase, Temperature, Additives) ptimization of the throughput (Sample amount: Column verloading) Adsorption Isotherm & Competition 1

Seven Basic Considerations in Choosing HPLC perating Parameters 1) Solubility - Hexane, Chloroform, Methanol, Water (buffer ph), other? ) Molecular Weight - Would GPC be useful in either the analysis or sample prep? 3) Functional Groups - Any ionizable groups? Acidic, Basic, or Neutral? 4) Sample Matrix - What amounts are expected in matrix for either analytical or preparative isolation? 5) Levels in Matrix - What amounts are expected in matrix for either analytical or preparative isolation? 6) Detectability - Any chromophores or fluorophores? Consider Redox or derivatization. Together with point #5, an appropriate detector is chosen. 7) How Do Species Differ - An important clue to manipulate selectivity in the separation, especially if compounds are similar in their structure. High Performance Liquid Chromatography Modes PRINCIPLE F SEPARATIN: ADSRPTIN NRMAL PHASE PARTITIN REVERSED PHASE SLUTES: LIPPHYLIC: ILS, FATS, LIPIDS MST F THE BIMEDICAL SUBSTANCES MBILE PHASE: RGANIC SLVENTS: n-hexane, HEPTANE, CHLRFRM, ALCHLS AQUEUS MIXTURES WITH METHANL, ACETNITRILE AND ADDITIVES (BUFFERS, IN -PAIRS)

High Performance Liquid Chromatography Modes PRINCIPLE F SEPARATIN: IN-EXCHANGE SIZE-EXCLUSIN BI-AFFINITY CHIRALITY SLUTES: INRGANIC INS, ACIDS, BASES PLYMERS, PRTEINS, NUCLEIC ACIDS PRTEINS & ENZYMES ENANTIMERS MBILE PHASE: AQUEUS BUFFERS, INIC SLUTINS AQUEUS BUFFERS R RGANIC SLVENTS AQUEUS BUFFERS AND SPECIAL ADDITIVES AQUEUS R RGANIC SLVENTS Strategy for Preparative Separation: Selection of the appropriate mode of chromatography ptimization of the Separation (Stationary phase, Mobile Phase, Temperature, Additives) ptimization of the throughput (Sample amount: Column verloading) Adsorption Isotherm & Competition 3

PERFRMANCE CRITERIA BY NE PEAK Application RETENTIN FACTR or CAPACITY RATI k' = t R - t 0 t 0 k = C s φ Cm ASYMMETRY FACTR B(10 h) A f = A (10 h) AU 0.00 0.015 0.010 0.005 0.000 height Width (50) Minutes Area: 8501 USP Resolution: 4.48 USP Tailing: 1.11 K Prime: 4.0 Retention Time: 0.74 A B 10h Width(base) 0.710 0.70 0.730 0.740 0.750 0.760 0.770 TAILING FACTR T f = A + B (5 h) A NUMBER F THERETICAL PLATES t R N = 16 ( ) w PERFRMANCE BY TW PEAKS SELECTIVITY FACTR α = k' ( ) k' (1) t R(1) t R() EXPERIMENTAL RESLUTIN Rs = t R () - t R(1 ) 1/ (w 1 + w ) t 0 w 1 w 4

R E S P N S E 0 0 VID VID 1 3 4 ISCRATIC vs GRADIENT 1 3 4 5 9 6 7 8 10 11 5 6 7 8 9 10 11 TIME (MIN.) Gradient enables wide scope of chemical entities General Introduction to Preparative Chromatography 5

HPLC System-Prep Solvent (Mobile Phase) HPLC Column (Packing Material/Stationary Phase) Injector AutoSampler Data Pump Detector Fraction Collector (Purified Analyte(s)) Modern Setup of Preparative Chromatography: Mass and UV directed Separations Makeup Pump MS Autosampler Column Splitters Gradient Pump Purified Fractions PDA Waste Trigger of Fraction Collector by Detector on MS ELSD 6

Purification of 0.1 µmole of 5 HEX labeled 5mer Non-labeled failure sequences Target Product Labeled failure sequences UV 60nm UV 539nm 6 10 Minutes Fountain et al., J. Chromatogr. B, 003, vol. 783, 61-7 Various Dimensions of Preparative Columns 50 30 30 19 7.8 19 7.8 7

Problem Definition: Purifications What quantity of material needs to be isolated? Is the material a major or minor component? Do you need to maintain biological activity? What degree of purity (or specific activity) is required? How will purity or activity be verified? If analyzed sample is to be collected; the above questions must be answered Preparative Chromatography Terminology: Different from Analytical Chromatography Sample Solubility Load - verload Throughput Purity Recovery/Yield from Column Recovery from Fractions Cost of Purification 8

Understanding the Chromatographic Process is Essential for Rational Scaling-up B+A Stationary Phase Mobile phase A B Distribution: K = Cs/Cm B A k' = Elution through the Column t R - t 0 t 0 k = C s φ Cm Chromatogram RETENTIN FACTR = Capacity Factor We know the measurement of k from the analytical work: Retention Factor: k' = t R - t 0 t 0 We use its thermodynamic expression to rationally scale up the separation: CAPACITY RATI k = C s C s V s = = φ K C m C m V m 9

ADSRPTIN ISTHERMS: THE KEY T RATINAL SCALE-UP k Langmuir type C s = C m linear Adsorption Isotherm φ Linear C [ s k = φ ] C m C S Non-Linear practical working limit in the analytical mode C s = a C m 1 + b C m Non-Linear 0 C m, mole / L C s = k φ C m Mechanism of Tailing C s = a C m 1 + b C m Linear C S Non-Linear 0 practical working limit in the analytical mode C m, mole / L 10

Mechanism of Diffuse Front Non-Linear Linear C s = ac α C S Non-Linear practical working limit in the analytical mode 0 C m, mole / L C s = k φ C m Example of the Lab Scale Purification of Single Stranded RNA 11

Mass verload Absorbance 6.5 µg injected Analytical load of 6 µg yields efficient peak shape Minutes 4 6 Preparative load of 5 mg generates mass overload peak shape Absorbance 5 mg injected Note that the back of the peaks of the analytical and prep loads are at the same retention (-------) 4 Minutes 6 Volume verload Absorbance 6.5µg injected in 5 µl of column volume: 0.65 Column Volume: 0.8 ml (800 µl) Wider peaks first observed at low retention Absorbance 0.65 mg injected in 500 µl of column volume: 6.5 4 Minutes 6 Peak position shifts to higher retention in proportion to the injection volume Start of peak remains in the same position unless injected in a weaker solvent Minutes 4 6 1

Scaling Comparison verlaid 8.0e- 7.5e- 7.0e- 6.5e- 6.0e- 5.5e- 5.0e- AU 4.5e- 4.0e- 3.5e- 3.0e-.5e-.0e- 1.5e- 1.0e- 5.0e-3 Time 10.00 10.50 11.00 11.50 1.00 1.50 13.00 13.50 14.00 14.50 15.00 15.50 16.00 Strategy for Preparative Separation Selection of the appropriate mode of chromatography ptimization of the Separation (Stationary phase, Mobile Phase, Temperature, Additives) ptimization of the throughput (Sample amount: Column verloading) Adsorption Isotherm & Competition 13

From Small Scale To Purification Parameters of Scaling Up and ptimization of the verloaded Chromatography Mass Load Injection Volume Flow Rate Gradient/Run-Time Duration Column Dimensions Particle Size Chemistry Chemistry of the Mobile phase 14

Scaling Mass Load and Injection Volume Mass Load Proportional to column volume Limited by solubility in mobile phase Injection volume Approximately proportional to column volume Approximately proportional to both length and crosssectional area Most strongly dependent on sample solvent Scale Up Equations - Mass V M prep = M anal V prep anal M prep = mass injected on prep column M anal = mass injected on analytical column V prep = volume of prep column M anal = volume of analytical column 15

M M M L L d d 1 1 1 L = M 1 L is Mass is is is is is Mass Length Length 1 on Diameter of Diameter d where : d 1 on Column 1 of Scaling Equations Mass Column of of Column 1 Column Column 1 Column (Analytical) (Prep) Mass Scalability - XTerra Prep MS C 18 Sample: 3 mg load XTerra MS C18 5 µ particle, 4.6x50mm 1.8 ml /min 70 nm 13B/min over 4 min 3mg Components: xicillin Cloxicillin Dicloxicillin Mobile phase: A: Water with 0.1 Formic Acid B: ACN with 0.1 Formic Acid Sample: 51 mg load XTerra Prep MS C18 5 µ particle, 19x50mm 31mL /min 70 nm 13B/min over 4 min 51mg 17x Sample: 18 mg load XTerra Prep MS C18 5 µ particle, 30x50 mm 77mL /min 70 nm 13B/min over 4 min 18mg 43x 1.00.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 Time 16

Mass Scalability - XTerra Prep MS C 18 4.6x50mm 3mg 19x50mm 51mg 17x 30x50 mm 18mg 43x 1.00.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 Time R (cm) R L (cm) V (ml) Scaling Factor Flow Mass Load 0.30 0.053 5 0.83 1.8 3 0.950 0.903 5 14.18 17 30.7 51 1.5.50 5 35.34 43 76.6 18 Buspirone Impurities II 0.0 0.10 AU 4.8 mg Structure of Buspirone N (CH ) 4 N N N 0.00 0.0 0.10 AU 0.00 0.0 10.00 0.00 30.00 Minutes 9.6 mg 10.00 0.00 30.00 Minutes 19. mg Column: SymmetryPrep C18, 7µm (3.9 x 150) mm Mobile Phase: 8 acetonitrile / 7 0.18 TETA-MeCH ph 7.0 Flow Rate: 1.0 ml/min Detection: UV at 360 nm Sample: 1 mg/ml of Buspirone Injection: from 0.4 to 1.6 ml N 0.10 AU 0.00 El Fallah 10.00 0.00 30.00 Minutes 17

Approximate Mass Loading Capacity Many factors affect the mass capacity of preparative columns. The listed capacities represent an average estimate Mass Capacity (mg) Diameter (mm) Length (mm) 3.9 4.6 7.8 8 10 19 0 5 30 40 47 50 50 3 8 15 45 50 110 310 4 5 15 15 5 90 155 5 400 60 150 6 8 5 40 135 150 335 930 00 30 310 795 50 10 13 40 60 5 50 560 1550 300 1 16 45 50 75 70 300 470 670 1195 1650 1860 Reasonable 1.0 1.4 4.0 4. 6.6 4 7 4 60 105 145 164 Flow Rate(ml/min) Reasonable Injection Volume( µ l) 15 0 60 65 350 390 610 880 1565 160 450 Parameters of Scaling Up and ptimization of the verloaded Chromatography Mass Load Injection Volume Flow Rate Gradient/Run-Time Duration Column Dimensions Particle Size Chemistry Chemistry of the Mobile phase 18

Scale Up of Flow Rate * Same Linear Velocity to maintain N and Rs and Retention Time - Van Deemter Scaling Equations Flow Rate F F F d d 1 1 d = F 1 d is Flow 1 where : Rate for Column 1 is Flow Rate Column is Diameter of Column 1 is Diameter of Column (Prep) (Analytical) 19

Valerophenone Impurities 0.003 0.00 a 5 µm, 3.9 x 150 a: 0.70 ml/min Structure of Valerophenone AU 0.001 0.000 CH 3 (CH ) 3 C 5.00 10.00 15.00 Minutes 0.003 0.00 AU 0.001 b 7 µm, 7.8 x 300 b: 5.60 ml/min Column: a: Symmetry C18 5 µm (3.9x150) mm b: Symmetry Prep C18 7 µm (7.8x300) mm Mobile Phase: 60 acetonitrile / 40 water Sample: 4 mg/ml Valerophenone solution a: µl injection b: 800 µl injection Detection: UV at 340 nm 0.000 El Fallah 5.00 10.00 15.00 Minutes Example: Calculations 0.003 0.00 AU 0.001 0.000 0.003 0.00 AU 0.001 a b 5 µm, 3.9 x 150 a: 0.70 ml/min 5.00 Minutes 10.00 15.00 7 µm, 7.8 x 300 b: 5.60 ml/min 0.000 El Fallah 5.00 Minutes 10.00 15.00 R (cm) R^ L (cm) V (ml) Flow Scaling (ml/min) 0.195 0.038 15 1.79 Factor 0.7 0.390 0.15 30 14.34 8 5.6 Column Volume = π x R x L 0

Column Volume to Run-Time Volume Relationship 5 x50 mm 5 x50 mm Column Volume: (πr L~ ) 5 ml Column Volume: (πr L~ ) 15 ml 5 min Run-Time at Flow = ml/min 5 min Run Time at Flow = 50 ml/min Run-Time Volume: Analytical 5 min x 1 ml/min = 50 ml No. of Column Volumes: Run-Time Vol/Column Vol 50/ 5 = 10 Run-Time Volume: Preparative 5 min x 5 ml/min = 150 ml No. of Column Volumes: Run-Time Vol/Column Vol 150 / 15 15 = 10 Valerophenone Impurities 0.003 0.00 a 5 µm, 3.9 x 150 a: 0.70 ml/min Structure of Valerophenone AU 0.001 0.000 CH 3 (CH ) 3 C 5.00 10.00 15.00 Minutes 0.003 0.00 AU 0.001 b 7 µm, 7.8 x 300 b: 5.60 ml/min Column: a: Symmetry C18 5 µm (3.9x150) mm b: Symmetry Prep C18 7 µm (7.8x300) mm Mobile Phase: 60 acetonitrile / 40 water Sample: 4 mg/ml Valerophenone solution a: µl injection b: 800 µl injection Detection: UV at 340 nm 0.000 El Fallah 5.00 10.00 15.00 Minutes

Example: Calculations 0.003 0.00 AU 0.001 0.000 0.003 0.00 AU 0.001 a b 5 µm, 3.9 x 150 a: 0.70 ml/min 5.00 Minutes 10.00 15.00 7 µm, 7.8 x 300 b: 5.60 ml/min 0.000 El Fallah 5.00 Minutes 10.00 15.00 R (cm) R^ L (cm) V (ml) Scaling Flow Run-Time (Min) Column Volumes 0.195 0.038 15 1.79 Factor 0.7 15 5.9 0.390 0.15 30 14.34 8 5.6 15 5.9 Column Volume to Gradient Volume Relationship 5 x50 mm 5 x50 mm Column Volume: (πr L~ ) 5 ml Column Volume: (πr L~ ) 15 ml 5 min Gradient at Flow = ml/min 5 min Gradient at Flow = 50 ml/min Gradient Volume: Analytical 5 min x ml/min = 50 ml No. of Column Volumes: Gradient Vol/Column Vol 50/ 5 = 10 Gradient Volume: Preparative 5 min x 50 ml/min = 150 ml No. of Column Volumes: Gradient Vol/Column Vol 150 / 15 15 = 10 3

Scale Up Equations - Gradient V prep t prep = tanal V anal F F anal prep t prep = run time with prep column t anal = run time with analytical column V prep = volume of prep column V anal = volume of analytical column F prep = prep flow rate F anal = analytical flow rate Example: Scale-up Gradient SunFire TM C 18 4.6x50 mm Total Mass Load: 5 mg 1 3 Flow Rate: 1.4 ml/min Gradient: Time Profile (min) A B 0.0 95 5 1.0 95 5 6.0 30 70 7.0 30 70 Injection Volume: 0 µl - Time 1.00.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 SunFire TM C 18 19x50mm Total Mass Load: 85mg 1 3 Flow Rate: 3.9 ml/min Gradient: Time Profile (min) A B 0.00 95 5 1.8 95 5 6.8 30 70 7.8 30 70 Injection Volume: 340 µl - Time 1.00.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 Compounds: Metoprolol (50 mg/ml), xprenolol (5 mg/ml), Propranolol (5 mg/ml) 4

- - 1 Scale-up Gradient SunFire TM C 18 4.6x50 mm Total Mass Load: 5 mg Time 1.00.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 1 3 3 SunFire TM C 18 19x50mm Total Mass Load: 85mg Time 1.00.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 Compounds: Metoprolol mg/ml), xprenolol mg/ml), Propranolol mg/ml) (50 (5 (5 (Addition of 0.8 min to compensate for instrument s delay volume) Flow Rate: 1.4 ml/min Gradient: Time Profile (min) A B 0.0 95 5 1.0 95 5 6.0 30 70 7.0 30 70 Injection Volume: 0 µl Flow Rate: 3.9 ml/min Gradient: Time Profile (min) A B 0.00 95 5 1.8 95 5 6.8 30 70 7.8 30 70 Injection Volume: 340 µl R (cm) R L (cm) V (ml) Scaling Factor Flow 0.30 0.053 5 0.83 1.4 0.950 0.903 5 14.18 17 3.9 Prep Calculator 5

Scale Up Column s Dimensions Dimensions Mass capacity is proportional to column volume Elution volume is proportional to column volume Shorter columns preferred for speed and pressure Longer columns preferred for resolution Effect of Column Length Efficiency increases with length Backpressure increases with length Run time increases with length Cost increases with length 7

Effect of Column Length on Capacity Absorbance Length = 50 mm Load = 1 mg Length = 150 mm Load = 36 mg As can be seen in this example, there is a linear relationship between column length and loading capacity - 3X increase in column length generates a >3X increase in loading capacity Disadvantage of long columns: - higher pressure for equal run time (9x at 3x increase in length) - higher pressure at equal velocity and longer run time 0 Minutes 0 Effect of Column Length XTerra MS C 18, BD, 5 µm 19 X mm -6 Time.00 4.00 6.00 8.00 10.00 1.00 14.00 16.00 19 X 50 mm -6 1.00.00 3.00 4.00 5.00 6.00 7.00 8.00 Time As column length decreases, resolution between the minor impurity and main peak decreases. Loading capacity will also decrease. 8

XTerra Prep Columns for Speed 40 Reduction in Retention and Peak Volume Peak # 1 Peak # Peak # 3 13 ml 9 ml 13 ml Peak # 1 Peak # Peak # 3 8 ml 6 ml 7 ml Scaling to Smaller Columns Allows: Faster chromatography Peak volume reduction Less expensive column Depending on the application, how far is it possible to downsize? Plate-to-plate mapping, injecting from a 96 well plate and collecting fractions in another 96 well plate, analytical size columns are suitable for this work 9

Load Considerations Mass Capacity Proportional to column volume Injection Volume Scale in proportion to the column volume Usually larger for gradient separations Cannot be considered separately from sample solvent Sample Solvent Good solvent needed to load high sample concentrations Good sample solvents degrade chromatography Good chromatographic diluents give low sample concentrations Parameters of Scaling Up and ptimization of the verloaded Chromatography Mass Load Injection Volume Flow Rate Gradient/Run-Time Duration Column Dimensions Particle Size Chemistry Chemistry of the Mobile phase 30

Effect of Particle Size on Capacity Column: 4.6 x 50 mm DS-A 10Å 15 µm 0 mg Load As particle size increases, resolution of the minor impurities decreases ( ) Absorbance 5 µm 0 mg Load Well resolved components not affected by change in particle size 50 µm 0 mg Load 30 40 50 Minutes Effect of Particle Size XTerra MS C 18, BD, 19 X mm 5 µm 15 mg of buspirone -6 Time.00 4.00 6.00 8.00 10.00 1.00 14.00 16.00 10 µm 15 mg of buspirone -6 Time.00 4.00 6.00 8.00 10.00 1.00 14.00 16.00 As particle size increases, resolution between the minor impurity and the main peak decreases. 31

Parameters of Scaling Up and ptimization of the verloaded Chromatography Mass Load Injection Volume Flow Rate Gradient/Run-Time Duration Column Dimensions Particle Size Chemistry Chemistry of the Mobile phase Prochlorperazine: Effect of Loading Capacity on the Separation of Impurities Column: 0.003 a: Kromasil a: SymmetryPrep C18 7µm (4.6x150) mm b: Kromasil C18 7µm (4.6x150) mm 0.00 0.001 AU 0.000 1 0.00 10.00 0.00 30.00 Minutes Mobile Phase: 75 methanol / 5 0 mm phosphate buffer ph 7.0 Flow Rate: 1 ml/min Detection: UV at 54 nm Sample: 10 µl of 0.97 mg/ml solution 0.003 0.00 0.001 AU 0.000 1 b: SymmetryPrep CH N 3 N CHCH CH N CH S H CH S H Cl 0.00 10.00 0.00 30.00 Minutes S 3

Impact of Column Chemistry Conditions: Columns: 4.6 x 150 mm, 5 µm Mobile Phase A: 0.1 HCH in H Mobile Phase B: 0.1 HCH in ACN Flow Rate: 1.4 ml/min Gradient: Time Profile (min) A B 0.0 70 30 10.0 10 90 15.0 10 90 Injection Volume: 10.0 µl Sample Diluent: DMS Temperature: 30 o C Detection: UV @ 54 nm Instrument: Alliance 695 with 996 PDA N NH CH 3 H H N xacillin H S CH 3 CH 3 Cl H N NH CH 3 Cloxacillin H S N H CH 3 CH 3 Cl Cl H H NH S N N CH 3 H Dicloxacillin CH 3 CH 3 All C 18 Columns are NT the Same - - -3 SunFire C 18 Rs:.1 W 1/ : 0.4 Peak splitting Luna () C 18 Rs: 1.31 W 1/ : 0.39 Kromasil C 18 Hypersil Gold Rs: N/A W 1/ : N/A Rs: 0.71 W 1/ : 0.90 0.4 mg Total Load -3.00 4.00 6.00 8.00 10.00 1.00 14.00 16.00 18.00 0.00 Zorbax C 18 Rs: N/A W 1/ : N/A Time -.00 4.00 6.00 8.00 10.00 1.00 14.00 16.00 18.00 0.00 Time 33

Parameters of Scaling Up and ptimization of the verloaded Chromatography Mass Load Injection Volume Flow Rate Gradient/Run-Time Duration Column Dimensions Particle Size Chemistry Chemistry of the Mobile phase Chemistry of the Sample Solution: Maximizing Sample Load Mobile phase solvent choice Buffer the mobile phase Control of chromatographic ph conditions Reduces breakthrough of sample Use volatile buffers Reduces sample handling and energy required for evaporation Minimizes contamination of purified fraction Minimizes risk of target decomposition Mass spectrometry compatible Mobile phase ph impacts load Low ph maximizes acid load High ph maximizes base load 34

Sample Solvent Affects Resolution and Peak Shape Chromatography run at ph 3.8 Sample solvent ACN Sample solvent DMS Sample solvent IPA Sample solvent MeH Sample solvent THF Poor resolution DMS in high aqueous mobile phases generates a high viscosity pressure spike (observed post injection) resulting in poor resolution of the chromatographic peaks. All weaknesses in HPLC system will become evident. Increasing Load in DMS: Mass verload Column: XTerra MS C 18, 4.6 50 mm, 5 µm.mobile phase A: water + 50 mm formic acid; mobile phase B: acetonitrile + 50 mm formic acid. Flow rate: 1.8 ml/min. Gradient: 35-65 B in 5 min. UV: 80 nm. 5.4 mg Load Peak begins to front 5.7 mg Load Very bad fronting 7.5 mg Load Breakthrough 1.5 mg Load When DMS is used as the diluent, mass loading is 5.4 mg. 35

Non-Ionic Loading: Acids XTerra MS C 18 Breakthrough Cl Cl N NH Dicloxacillin @ 50 mg/ml ph = H H S CH3 CH3 N CH3 H DMS 1.5 mg Load -1.3X higher loading Nice peak shape DMS + 50 mm FA Formic acid was added to DMS directly, then use this acidified DMS to dissolve sample 1.5 mg Load -1 0.5 0.50 0.75 1.00 1.5 1.50 1.75.00.5.50.75 3.00 3.5 3.50 3.75 4.00 4.5 4.50 4.75 5.00 5.5 Time When acidic sample is loaded in its non-ionic state, mass loading is up to 1.5 mg on XTerra MS C 18 column. Non-Ionic Loading: Bases XTerra MS C 8 Econazole @ 00 mg/ml ph = 11-0 Shoulder appears USP Tailing: 1.60 DMS 6 mg Load Nice peak shape USP Tailing: 0.78 DMS + 50 mm NH 4 H 6 mg Load -0 0.5 0.50 0.75 1.00 1.5 1.50 1.75.00.5.50.75 3.00 3.5 3.50 3.75 4.00 4.5 4.50 4.75 Time When a basic sample is loaded in its non-ionic state, peak tailing improves dramatically. 36

Loading a Base under Acidic Mobile Phase Conditions Column: XTerra MS C 18, 4.6 50 mm, 5 µm.mobile phase A: water + 50 mm formic acid; mobile phase B: acetonitrile + 50 mm formic acid. Flow rate: 1.8 ml/min. Gradient: 5-95 B in 5 min. UV: 80 nm. Breakthrough DMS 13.5 mg DMS+ NH 4 H 13.5 mg - - - DMS 10.5 mg 1.00.00 3.00 4.00 5.00 Time 6.00 Amitriptyline HCl 1.00.00 3.00 4.00 5.00 Time 6.00 When NH 4 H is added to the DMS, no breakthrough was observed. 1.8X high mass load by adding base to the diluent. - DMS+ NH 4 H 18 mg Choice of Strong Solvent (Acetonitrile vs. Methanol) Column: XTerra MS C 18, 4.6 50 mm, 5 µm.mobile phase A: water + 50 mm formic acid; mobile phase B: acetonitrile/methanol + 50 mm formic acid. Flow rate: 1.8 ml/min. Gradient: 35-65 B in 5 min. UV: 54 nm. Acetonitrile, ph Methanol, ph Cl Cl H H NH S N CH3 N H Dicloxacillin CH3 CH3 5.4 mg 6 mg 1.00.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 1.00.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 Applying the same gradient with both organic solvents, sample elutes later with methanol than with acetonitrile. Changing the organic solvent may improve peak shape due to the additional interaction with the sample. 37

Use of Buffers Sample: Propranolol 5 mm Buffer, ph 7 DI water, ph 7 DI water, ph 7 4.00 3.60 t r = 3.1 4.00 3.60 4.00 3.60 3.0 3.0 3.0.80.40.00.80.40.00 Breakthrough @ t 0.80.40.00 t r = 3.15 1.60 1.60 1.60 1.0 1.0 1.0 0.80 0.80 0.80 0.40 0.40 0.40 0.00 1.00 3.00 5.00 7.00 0.00 1.00 3.00 5.00 7.00 0.00 1.00 3.00 5.00 7.00 0 ACN 3 mg 0 ACN 3 mg 7 ACN 0.5 mg Buffered mobile phases enhance retention and mass loading. There is a high risk of breakthrough and retention loss leading to recovery problems when buffers are left out! Modifier Impact on Load AU.00 1.00 0.1 Formic Acid 0.3 mg Total Mass Analytes: 1. xacillin. Cloxacillin 3. Dicloxacillin 0.00 1.00.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 1.00 13.00 14.00 15.00 Minutes AU 3.00.00 0.1 TFA 0.96 mg Total Mass 1.00 0.00 1.00.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 1.00 13.00 14.00 15.00 Minutes As sample mass / volume loading increases, strong ph shifts can occur, affecting chromatographic peak shapes. 38

Mobile Phase ph Effect: Loading of Bases XBridge C 18 1, 3 Adjusted gradient to keep the same k values. X5 in y axis Bases 1. Nordoxepin. Doxepin 3. Amitriptyline 0.1 mg ph.3-1 Time 0.00 1.00.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 1.00 13.00 14.00 1 1 3 3 X0 in y axis -8 Time 0.00 1.00.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 1.00 13.00 14.00-0 Time 0.00 1.00.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 1.00 13.00 14.00 0.1 mg 0.5 mg ph 7 ph 10 High ph for basic analytes results in the best loadability, as well as the best retention and separation.* *: Uwe D. Neue et al. J. Chromatography A. Vol. 1030, 13-134, 004. Adjusted gradient to keep the same k values. XBridge TM C 18 1 3 Mobile Phase ph Effect: Loading of Acids Acids 1. xacillin. Cloxacillin 3. Dicloxacillin 0.8 mg ph.3-0 Time 0.00 0.50 1.00 1.50.00.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 7.50 8.00 8.50 9.00 9.50 10.00 1 3 0.3 mg ph 7 0 Time 0.00 0.50 1.00 1.50.00.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 7.50 8.00 8.50 9.00 9.50 10.00 1 3 0. mg ph 10-0 Time 0.00 0.50 1.00 1.50.00.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 7.50 8.00 8.50 9.00 9.50 10.00 Low ph for acidic analytes results in the highest loadability, as well as the best retention and separation.* *: Uwe D. Neue et al. J. Chromatography A. Vol. 1030, 13-134, 004. 39

Loadability with Varied ph Flurbiprofen Amitriptyline ph 4.80 mg 10.5 mg ph 11.4 mg 13.5 mg -1 1.00.00 3.00 4.00 5.00 1.00.00 3.00 4.00 5.00 6.00 7.00 Increased retention for targets in their non-ionic state. Utilizing ph to Manipulate Target Retention ph is the most powerful tool to manipulate retention of ionizable compounds Elution rder (Standard Gradient Mode) early eluting peaks can be isolated more quickly later eluting peaks can be purified at higher loadings and elute in higher concentration organic solvent Basic compounds chromatographed in their unionized state (ph units above pka) are retained longer and exhibit excellent peak shape A popular high ph volatile buffer is 0.1 NH4H 40

Example: Basic Test Compounds CH 3 N CH 3 H N CH 3 CH 3 Diphenhydramine xybutynin C(CH 3 ) 3 N H H Terfenadine Peak shape and retention comparison: Basic Compounds at Low and High ph AU AU 0.50 0.45 0.40 0.35 0.30 0.5 0.0 0.15 0.10 0.05 0.00 0.00 1.00.00 3.00 4.00 5.00 Minutes 6.00 7.00 8.00 9.00 10.00 11.00 0.50 0.45 0.40 0.35 0.30 0.5 0.0 0.15 0.10 0.05 XTerra Prep MSC 18 19 x 50 mm, 5 µm ph 3.8 ph 10 0.00 0.00 1.00.00 3.00 4.00 5.00 Minutes 6.00 7.00 8.00 9.00 10.00 11.00 1 1 Analytes: 1. Diphenhydramine (.5 mg/ml) 3 mg Load. Terfenadine (0.15 mg/ml) 0.18 mg Load 41

Scale-up Strategy - Summary 1. Define the problem Find the chromatographic mode.. Develop and optimize the separation Increas selectivity > 1.5 3. Maximize throughput Measure adsorption isotherm. 4. Increase sample mass and volume to the maximum while meeting purity objectives. Examine the interferences 5. Determine recovery Examine residuals on the column 6. Scale up to desired column size to meet throughput/load objectives. Keep the flow rate and sample load ratio 7. Pool fractions of comparable purity and rerun if necessary. 8. Check fraction purity using analytical column. 4