particle morphology hints impurity rejection capability during API - - PowerPoint PPT Presentation

Revealing shapes – particle morphology hints impurity rejection capability during API crystallization

Jochen Schöll

Technobis webinar on 24 SEP 2020

WAG

XLab

MSD Werthenstein BioPharma in Schachen Lucerne

3

“Medicine is for the people. It is not for the profits.” George W. Merck (1950)

Main tasks in pharma crystallization

4

Process:

• Solvent system
• Temperature
• Addition rates
• etc…

Typical development challenge: limited amount of API

5

and 3 ways to make most out of it: Miniaturization Parallelization Data-rich experimentation

Approaches in our lab

6

Miniaturization Parallelization Data-rich experimentation

Unmilled API

Crystallization issues during 1st GMP delivery (1.5 kg API)

7

• Solvent switch crystallization

(EtOAC → Toluene) successfully rejected impurities but failed controlling solid form

• Additional recrystallization from

MeOH/Water was needed to control the crystal form

• Final micronization step via

spiral jet milling yielded 1.5 kg API

Quantiles [µm] x10 1.4 x50 4.1 x90 8.1

Jet-milled API

Quantiles [µm] x10 40.2 x50 78.0 x90 152.6

Solvent (pre-)selection with HT data

8

Solvent Selection Average [mg/ml] Solvent Selection Average [mg/ml] Solvent Selection Average [mg/ml] Acetone 90.14 50:50 EtOH:H2O 0.46 2:1 DMF:H2O 8.24 MeCN 50.41 25:75 EtOH:H2O 0.06 1:1 DMF:H2O 1.10 EtOH 9.87 8:1 EtOH:H2O 10.20 1:1 DMF:MeOH

>100

H2O 0.00 2:1 EtOH:H2O 3.33 2:1 DMAc:H2O

9.06

Toluene 3.87 95:5 IPA:H2O 9.45 1:1 DMAc:H2O 0.91 THF >100 90:10 IPA:H2O 10.27 2:1 NMP:H2O

16.93

Heptane 0.00 80:20 IPA:H2O 9.01 1:1 NMP:H2O 1.38 IPAc

22.46

50:50 IPA:H2O 0.12 1:1 Toluene:MeOH 52.46 DMF >100 25:75 IPA:H2O 0.15 1:1 Toluene:MeCN

73.57

NMP >100 2:1 IPA:H2O 4.75 1:1 Toluene:EtOH

31.25

MeOH 17.95 8:1 IPA:H2O 10.61 1:1 Toluene:IPA 22.83 2-methylTHF 60.72 95:5 THF:H2O >100 1:1 IPAc:MeOH

37.36

DCM 49.77 90:10 THF:H2O >100 1:1 MeCN:EtOH 54.20 DMAc >100 80:20 THF:H2O >100 1:1 2-methylTHF:Heptane 1.40 DMSO >100 50:50 THF:H2O 0.69 1:1 THF:MTBE

32.77

EtOAc 49.83 25:75 THF:H2O 0.09 1:1 Toluene:MTBE 5.50 IPA 5.90 2:1 THF:H2O 36.07 1:1 IPAc:MTBE

13.46

MTBE 3.58 8:1 THF:H2O >100 1:1 IPA:MTBE

7.86

2-butanol 4.59 90:10 Acetone:H2O >100 2:1 Acetone:Water 60.48 MEtOAc 76.10 80:20 Acetone:H2O 0.11 1:1 Acetone:Water 15.85 n-propanol 6.85 50:50 Acetone:H2O 0.07 1:1 EtOH:Heptane 8.32 Hexane 0.00 25:75 Acetone:H2O 0.07 1:1 THF:Heptane 2.85 90:10 MeCN:H2O 68.43 2:1 Acetone:H2O 0.08 2:1 THF:Heptane 0.30 80:20 MeCN:H2O 45.79 8:1 Acetone:H2O 49.13 9:1 THF:Heptane 81.30 50:50 MeCN:H2O 0.84 95:5 MeOH:H2O 10.17 1:1 IPA:Heptane 3.63 25:75 MeCN:H2O 0.08 90:10 MeOH:H2O 8.75 1:1 Toluene:Heptane 0.37 95:5 EtOH:H2O 11.79 50:50 MeOH:H2O 0.17 2:1 IPAc:Heptane 7.81 90:10 EtOH:H2O 11.59 2:1 MeOH:H2O 0.98 1:1 IPAc:Heptane 3.01 80:20 EtOH:H2O 7.50 8:1 MeOH:H2O 8.74 1:1 MTBE:Heptane 0.22 1:1 DCM:Heptane 0.00

• High-throughput solubility

data provides a first set of potential solvent systems: ➢ Good solvents ➢ Good antisolvents ➢ Risky systems

API with two known forms

9

• Both, polar and apolar solvent

systems successfully generated the desired Form 2 at lab scale

• Seeded process with

supersaturation control expected to yield desired form

Form 2 Form 1 EtOAc/Heptane MeCN/Water

First results at 300 mg scale

10

• Particle morphoplogy changes in different solvents– potential for different impurity rejection?

Apolar solvent systems Polar solvent system

EtOAc/Heptane IPAc/Heptane MeCN/Water

Crystallization and impurities in the literature

11

Dichloromethane Ethanol

Development work for 2nd GMP delivery

12

Apolar solvent systems Polar solvent system

EtOAc/Heptane 2-MeTHF/Heptane MeCN/Water EtOAc/Heptane 2MeTHF/Heptane MeCN/Water API purity 99.9% LCAP 99.9% LCAP <99.0% LCAP Form control X ✓ ✓ Color rejection insufficient insufficient insufficient

Data-rich experiment at 3 g scale – EtOAc/Heptane

13

Yield ≈ 96 %

• Vol. prod. ≈ 62 g/L

Cycle time ≈ 18 h

Form impurity (rods)

Temperature

Crust formation

14

Hot reactor walls induce “creeping”: Cool reactor walls avoid “creeping”, but still on probes & stirrer:

Activated carbon treatment & rex in 2MeTHF/Heptane

15

Pure API Crude API

Crystallization + Activated carbon

Successful 2nd GMP delivery (27 kg API)

16

• Improved chemistry reduced RM

cost by 40x

• New API crystallization process
• Eliminates the need for recrystallization

due to improved impurity rejection

• Ensures robust API form control (lower

seed mass was compensated by extended seed age)

• Rejects color using AC
• Delivered 27 kg API with a higher

crystallization yield of +13%

Quantiles [µm] x10 1.4 x50 4.1 x90 8.1

Jet-milled API

Potential to eliminate jet milling?

17

• Unseeded high-shear

reverse additions have been performed in MeCN/Water and EtOAc/Heptane API crystallization process

• Polar and apolar systems

yielded mix of forms

• Ongoing work with

seeded reverse addition

Summary

18

Thank you!

• Different particle morphologies can indicate impurity rejection

differences (lower aspect ratio correlates with higher purity)

• Combination of miniaturization and data-rich experiments

allowed for an optimized process design regarding impurity rejection, form control, de-coloring, and process yield

• Further work on bottom-up process with form control ongoing

Acknowledgements

• Erica Schwalm
• Eric Ashley
• Eric Sirota
• Siwei Zhang
• Yonggang Chen

Q&A