Upload
vanhuong
View
226
Download
3
Embed Size (px)
Citation preview
Bernhardt L. Trout
Director, Novartis-MIT Center for Continuous Manufacturing
Professor, Department of Chemical Engineering, MIT
Enabling Technologies for the Continuous
Manufacturing of APIs: Continuous
Crystallization
2
Crystallization Process Development
Process Goals
• Purity
• Yield
• Average Size and Size Distribution
• Correct Polymorph or Pseudopolymorph
• Shape
3
Past Current > 2020
Disconnected process steps
Quality by Design
Process steps and their impact understood
Blue Sky Vision: Continuous Manufacturing
Seamlessly integrated and
well characterized processes
Road Map for Pharmaceutical Manufacturing Paradigm shifts in manufacturing and quality envisioned
Traditional
Manufacturing
4
Novartis-MIT Blue Sky Vision Integrated Continuous Manufacturing: A radical transformation
the ultra LEAN Manufacturing
From start of chemical synthesis through final pharmaceutical dosage form
MIT Massachusetts Institute of Technology
5
Our Definition of “Continuous” (ultra QbD)
Flow
Integration (end to end)
Systems approach
Integrated control strategy
6
Novartis-MIT Center for Continuous
Manufacturing
Phase 1 June 2007-May 2012
Phase 2 June 2012—June 2017
60 MIT students and post-docs
20 Novartis staff and 12 MIT Professors
from Chemical Engineering, Chemistry,
and Mechanical Engineering
Most research performed in each
professor’s lab
Dedicated facility for translational work.
7
MIT Continuous Manufacturing Facility
8
MIT Pilot Plant
Courtesy of NVS-MIT
Center
1
2 3 4 5
6
VIDEO
9
Tablets Produced in Integrated Process
10
Batch Versus Continuous Crystallization
Continuous Crystallization
• Built in flexibility for control
• Does not necessarily discharge at equilibrium
conditions
• Lower cost
• CSD is broad
• Polymorph Control?
Batch Crystallization
• Cleaning done between each batch
• Simplicity of equipment
• Narrower size distribution
• Higher cost
11
Novartis-MIT Center for Continuous Manufacturing
Important task: demonstration of end-to-end continuous manufacturing platform
Used to gain experience about integration and control
Involves multiple reactions, workup steps, crystallizations
Each crystallization presented unique challenges
12
Example 1: Optimization via Modeling of MIT Pilot Plant MSMPR
Courtesy of NVS-MIT
Center
1
2 3 4 5
6
VIDEO
13
Courtesy of NVS-MIT
Center
1
2 3 4 5
6
VIDEO
Example 1: Optimization via Modeling of MIT Pilot Plant MSMPR
14
Courtesy of NVS-MIT
Center
1
2 3 4 5
6
VIDEO
Example 1: Optimization via Modeling of MIT Pilot Plant MSMPR
15
Model for Multistage MSMPR
iiivsiT CCdLnLkM 1
3
g
s
sg
C
CCkG
b
s
sb
C
CCkB
0
Population Balance: Conservation equation for the number of crystals in a population
Mass Balance
Crystal Growth
Nucleation
ni: population density at stage i
ti: residence time at stage i
L: crystal size
Gi: crystal growth rate at stage i
B0: nucleation rate
MT i: Suspension density at stage i
C: steady state solute concentration
s: crystal density
kv : volume shape factor
Cs: equilibrium concentration
kg, g, kb, b: model parameters to
be estimated
1 iii
ii nndL
dnGt (1)
(2)
(3)
(4)
16
Kinetics Parameter Estimation
a) Convert the experimental values of CSD into Population
Density
b) Find the values of kg, g, kb, and b that minimize the objective
function F :
Fmax
0
2
exp )(min L
calc Lnn
Subject to equations (1) to (4)
where = [kg, g, kb , b ]
17
Modeling Work for Crystallization of Intermediate
Modeled how changing temperature and residence time of each stage affects purity and yield
200
300
400
500
600
70
75
80
85
90
95
200
3 0 0
4 0 0
5 0 0
6 0 0
Yie
ld,
%
Secon
d Stag
e Resi
dence
Time, m
inFirst Stage residence Time, min
71.40
74.18
76.95
79.73
82.50
85.28
88.05
90.83
93.60
Yie
ld %
Pu
rity
%
100.0
97.6 70
95
18
ALISKIREN (SPP-100) SALT C12(FUMARIC ACID)
+
C11 (SPP-100 FREE BASE)
a) HCl-gas, iPr2O
b) NaOH, Me-THF
OO
ONH2
OHHN NH2
O O
C11 92%
a) HCl-gas, iPr2O
b) NaOH, Me-THF
OO
ONH2
OHHN NH2
O O
C11 92%
. 1/2
Example 2: Tight Control of Continuous Aliskiren Reactive Crystallization
19
UV control of fumaric acid addition
C13 crystallization is very sensitive to fumaric acid/C11 ratio
0.00%
1.00%
2.00%
3.00%
4.00%
5.00%
6.00%
0.3 0.5 0.7 0.9 1.1 1.3
C13 c
on
cn
etr
ati
on
in
th
e m
oth
er
liq
uo
r, m
ass/m
ass
Molar Ratio of fumaric acid/C11
ALISKIREN (SPP-100) SALT C12(FUMARIC ACID)
+
C11 (SPP-100 FREE BASE)
a) HCl-gas, iPr2O
b) NaOH, Me-THF
OO
ONH2
OHHN NH2
O O
C11 92%
a) HCl-gas, iPr2O
b) NaOH, Me-THF
OO
ONH2
OHHN NH2
O O
C11 92%
. 1/2
% C
13
(m
as
s/m
as
s)
Molar Ratio of fumaric acid/C11
20
Feed forward control
Dilute drug
substance
Water
absorption
column
Fumaric
acid feed
UV
flowcell
Reactive crystallization
Filter & wash
Solvent dilution
Density
flowcell
Drum dryer
Extrusion/ molding
Silicon dioxide
Vacuum dryer
Go P5
B
A
FR
F
21
UV control of fumaric acid addition
Tightly controlled Fumaric Acid addition by the control system can reach high yield with satisfactory crystal properties.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.5 1 1.5 2 2.5
FA
/C1
1 m
ola
r ra
tion,
Yie
ld
Time, day
Vessel 1_FA/C11 molar Ratio
Vessel 1_Yield
Vessel 2_FA/C11 molar Ratio
Vessel 2_Yield
fum
ari
c a
cid
/C11
; y
ield
Time, days
22
Starting from Isolated Freebase to final API
Continuous process- effectively telescoped into 2 steps(15 L pilot-plant scale crystallizer), 8 hours total residence time
First stage: simultaneous reaction and crystallization
Second stage: further growth and nucleation
Process Yield of 97%
Conti. Process results in significantly shorter processing time compared to batch
Example 2: Summary
There are limited studies on polymorphism in continuous crystallization
Polymorphism impacts product bioavailability and manufacturability
There are fundamental differences when moving from batch to continuous crystallization process:
23
● Residence time distribution
● Steady state vs. Equilibrium
● Secondary nucleation
● Seeding Efficacy
Example 3: Control of Polymorphism in Continuous Crystallization
● No solvent mediated transformation
Polymorphism: α form (metastable), β form (stable)
Solvent: Water
Characterization: polymorph ratio (Raman, XRD), solute concentration (FTIR)
| Business Use Only 24
Case study: L-glutamic acid
α form β form
Studying the possibility to control polymorphism via manipulating stage temperature and residence time
• Residence time = 60 min, Temp = 25°C α form specific
• Polymorph specific MSMPR observed for the first time
| Business Use Only 25
*Mass Ratio= α form/ (α form + β form)
pure α form
Polymorph specific MSMPR achieved
| Business Use Only
Effect of temperature and residence time on polymorphism:
26
T=25°C T=45°C
τ (min) 30 60 90 120 60 120
ML conc.
(g/kg solvent) 28.60 22.36 19.74 18.24 35.32 26.41
Polymorphism α form α form α form α form β form β form
Metastable Form Stable Form
Low Temp High Temp
Short RT Long RT
Control of steady state polymorphism
| Business Use Only
Experimental Design:
Single stage MSMPR (T = 25˚C, τ = 60min) the unseeded steady state contains pure α form
β polymorph seed added during startup:
Seeding is unable to alter the steady state polymorphism, the seeds are removed in several residence time
27
Seed type β form β form β form
Seed mass 5% to MTeq 50% to MTeq 100% to MTeq
Final form α form α form α form
τwashout 4τ 7τ 9τ
Efficacy of seeding on Polymorphism
28
Example 4: Heterogeneous Crystallization on Patterned Excipients
28 | Confidential
Chemical/
Biological
Synthesis
Continuous
CrystallizationTableting/
Encapsulation
API
Solution
Excipient
Continuous
Purification
API-Excipient
Composite Particles
Manufacturing process API-excipient composite particles
Widely tunable chemical, physical, and mechanical properties
Potential to enhance API bioavailability
Potential to control drug release profile
Applicable to multiple API forms
Streamlined downstream processing
Control over crystallization kinetics through heterogeneous nucleation
API properties masked by the excipient to simplify downstream process development
Advantages:
29
Controlling API nucleation by tuning the nanopore shape
No pore 15nm 40nm
120nm 300nm The scale bar is 200nm
Polymer surfaces with
nanopores of various shapes
and sizes were fabricated by
Nanoparticle Imprint
Lithography (NpIL), as well as
Nanoimprint lithography (NIL)
30
Growth
direction
100nm
Pores with crystals
(011
)
(100)
Empty pores
Aspirin: Crystal orientation suggests nucleation of (011) & (100) from the ledge
Scale bar is 100nm Why not (002) & (100)?
Crystal orientation verified by XRD
31
Hetero-
nucleant Type
Induction
time
(τ, h)
Error
(h)
Linearity
(h)
Type A
(60° and 120°)
9.6 0.1 0.99
Type B
(80° and 100°)
13.6 0.3 0.98
Type C
(90°)
8.8 0.1 0.99
no polymer 42.7 0.8 0.99
no pattern 15.8 0.3 0.96
spheres 36.5 0.7 0.95
Nucleation of mefenamic acid in anisole on HPMC surfaces
Type A
60° and 120°
Type B
80° and 100° Type C
90°
absence of
nano-pores
spherical
nano-pores
-1.5
-1.0
-0.5
0.0
0.0 5.0 10.0 15.0 20.0
Ln (
P)
Time (h)
Type AType BType Cno polymerno patternspheres
32
Publication List
1. Zhang, H.; Quon, J.; Alvarez, A. J.; Evans, J.; Myerson, A. S.; Trout, B.; Development of Continuous Anti-Solvent/Cooling
Crystallization Process using Cascaded Mixed Suspension, Mixed Product Removal Crystallizers, Organic Process
Research & Development, 2012, 16, 915–924.
2. Quon, J. L.; Zhang, H.; Alvarez, A.; Evans, J.; Myerson, A. S.; Trout, B. L.; Continuous Crystallization of Aliskiren
Hemifumarate; Crystal Growth & Design, 2012, 12, 3036–3044.
3. Wong, S. Y.; Tatusko, A. P.; Trout, B. L.; Myerson, A. S.; Development of Continuous Crystallization Processes Using a
Single-Stage Mixed-Suspension, Mixed-Product Removal Crystallizer with Recycle, Crystal Growth & Design, 2012, 12,
5701–5707.
4. Alvarez, A, Singh, A., and Myerson, A.S. (2011). Crystallization of Cyclosporine in a Continuous Multistage MSMPR
Crystallizer. Crystal Growth and Design 11, 4392-4400.
5. Wong, S.Y., Cui, Y., and Myerson, A.S., (2013). Contact Secondary Nucleation as a Means of Creating Seeds for
Continuous Tubular Crystallizers. Crystal Growth and Design 13, 2514-2521.
6. Haitao Zhang, Richard Lakerveld, Patrick L. Heider, Mengying Tao, Min Su, Christopher J. Testa, Alyssa N D'Antonio,
Paul I Barton, Richard D Braatz, Bernhardt L. Trout, Allan S. Myerson, Klavs F Jansen, James M. B. Evans* (2013).
Application of continuous crystallization in an integrated continuous pharmaceutical pilot plant. To be submitted.
7. Ferguson S., Ortner, F., Quon, J., Peeva, L., Livingston, A., Myerson, A.S., (2013). Use of Continuous MSMPR
Crystallization with Integrated Nanofiltration Membrane Recycle For Enhanced Yield and Purity in API Crystallization.
Crystal Growth and Design (in review).
8. Mascia, S., Heider, P.L., Zhang, H., Lakerveld, R., Benyahia, B., Barton, P.I., Braatz, R.D., Cooney, C.L., Evans, J.M.B.,
Jamison, T.,Jensen, K.F., Myerson, A.S., and Trout, B.L., (2013). End-to-End Continuous Manufacturing of
Pharmaceuticals: Integrated Synthesis, Purification, and Final Dosage Formation. Angewandte Chemie International
Edition (in press).
33
Acknowledgements
Allan Myerson, Richard Braatz, Paul Barton
Students, Postdocs, Staff Scientists
• Examples 1 & 2: Haitao Zhang, Justin Quon, Alejandro Alvarez, Richard Lakerveld, Sal Mascia, James Evans
• Example 3: Chris Lai Tsai-ta
• Example 4: Ying Diao, Vilamli Lopez-Mejias, Li Tan
Novartis Pharma
34
35
Thank you! Questions?
Back-up
Single Stage MSMPR: 25°C, τ=2hr
Seeding condition: 100% MTeq β seed
Polymorph transform: β α
| Business Use Only 38
Polymorph transformation in MSMPR
Pure β form α crystal found State transition
Case Study objectives:
• Study the effect of seeding
• Effect of process parameters (residence time, temperature) on steady state polymorphism and stability
Simulation methodology:
• Population balance equations (partial integro-differential equations)
• Mass balance between liquid and solids
• Method of characteristics solves PDE at steady state
• Obtain steady state yield and polymorphism
| Business Use Only 39 |Confidential
T=25°C
Cfeed
300 rpm
Polymorph dynamic simulation
Yield?
Polymorph ratio?
Single stage MSMPR, T= 25˚C and τ = 60min
Regardless of the seeding conditions, steady state
polymorphism remains unchanged (stable S.S.)
40
Simulation case 1: seeding efficacy
Dynamic Simulation Results
Different seed
concentration
Polymorph dominance at different 𝝉 was studied (25˚C):
Negligible β stable form at S.S. for 𝜏<400 mins
To reach > 50wt% stable form at S.S., 𝜏>800 mins (13 hrs)
41
Pure β
Pure α
Simulation case 2: residence time
Steady state polymorphism at different residence time
Dynamic simulation can be used to determine the S.S.
polymorphism and the S.S. stability
• Further investigate effect of temperature and residence time
• May be difficult to obtain desired form at short residence time (>13hrs
in the case of the commercial β L-glutamic acid)
Design MSMPR system to achieve desired polymorph (β
form) while reducing total residence time:
• Continuous seeding from MSMPR cascade
Metastable Form Stable Form
Low Temp High Temp
Short RT Long RT 13hrs
Stable form
dominated 42
Implications from polymorph simulations
T1 T2