Developing catalytic screening in Chem. Dev.

Report
Introducing Milligram
Screening Reactions into the
Kilogram World of Chemical
Development
Simon Yates, AstraZeneca.
FreeSlate European User Meeting
24th September 2013
Our Journey so far…
What is Chemical Development?
• We produce a scalable process, not just a compound
• From a few grams in the lab to multi-kilo plant campaigns
• Emphasis on SELECT criteria
• Safety
• Economics
• Legal
• Environment
• Control
• Throughput
Scavenging Challenge?
Transition Metals are toxic!
In number of metal catalysed reactions in Med Chem
The time to develop process for early compounds
Conventional work up takes time to develop
Development time by using scavengers
Of scavengers and carbons to choose from
Time, effort and missed opportunity by automated screening
Reproducibility by screening under inert conditions
What are scavengers?
But what kit?
Integrity 10
• 10 reactions at a time
• Minimum volume ~2mL
But what kit?
• Inerted Glovebox
• 20mL reaction tubes
• 100mg reactions
• 24 wells
• Very manual
• Lots of programming
MT Flexiweigh
2x MT Mini Mapper
Wanted to….
In number of reactions we could run (up from 24)
Reaction scale
Manual intervention, to make this as routine as possible
Ease of programming and data analysis
‘Future proof’ our investment
Purchased a Symyx CM2
Solid Dispensing
10G needle
(5mL syringe)
9 heat-stir bays
Bespoke filter equipment
- Symyx filter too small
Vial gripper
16G to 20G needle (with N2 pressure)
- 1mL & 500uL syringes attached to 10
off deck solvents
Scavenging Workflow
96 x 1mL vials
~40 different
scavengers
Stock of reaction
solution
Seal, heat, stir
16 hours
Scavenging Workflow
Backing Solvent
Air Gap
Overshoot
ICP sample
HPLC sample
RAS
Centrifuge for
5 min
ICP = Inductively Coupled Plasma
Scavenging today
• We have run 20+ screens and saved projects time and money
• Numerous examples of where scavengers used in all scales of manufacture
• Pd is most common metal scavenged, now have a generic plate of 22 scavengers.
• Improvements in whole workflow reduced time from 5 working days to 2 days.
Cross Coupling
R1-M
R2-X
R1-R2
Background
Original plan was to run X-Coupling
Timing forced by closure of our existing facility
In number of reactions we could run
cover more experimental space, quicker.
Reaction scale
Apply our previous learning
What we developed
Pre-screening
~10 reactions
Discovery
96-well
96-well plate
plate
screening
screening
Development
Optimisation
of
Optimisation of
continuous
continuous
parameters
parameters
(24-plate)
DoE (24-plate)
Life Cycle
Management
Generic Reaction Plates
• Developed ‘Generic plates’ for 5 different reaction types
• Based on literature and in house expertise
18
Generic plate formation
Stock
solutions of
- Ligands
- Metals
- Internal std
19
Manual
Pipette
-or Automated
dispensing
Evaporate off
carrier solvent
and store
Running a project…
Define:
Bases and
Solvents
Add in solid(s)
Reactants
Add in
reaction
solvent
Running a project…
Quickly add in
Aq. Base
‘Start of
reaction’
Seal up
Heat and Stir
2 x Manual
Sample Prep
~2hrs and
20hrs
Future improvements
Reactions run on 200-400uL scale at 10C
below bp
Solvent loss / corrosion of sealing
material
Fully closed plate
Can’t sample by CM2
1st sample will always be manual
2nd sample could be automated – middle of night.
Need a pierceable, but re-sealable,
membrane
Any ideas welcome!!
24 well - DoE type
Expanding on 96 well hit(s)
A
B
Discrete variables
Continuous variables
Material consumption
Profile reactions
Statistical Analysis
24 well – DoE type
Catalyst Plate
24 x 4mL vials
Weigh in
Ligands
Metal
Add reaction
solvent
Cap, heat and
stir, 60min.
Reaction Plate
24 x 4mL vials
Weigh in
Reactants, Bases
Internal Standard
Add reaction
solvent/liquid
reagents
24 well – DoE type
Transfer from
Catalyst plate to
reaction plate
Reaction Plate
Heat and Stir
Automatically
sample
4 times over
16 hours
Run LC-Mass Spec
on samples
Future improvements
1 Temperature (plate) per run
Lose a key factor
Multiple Plates / run
 sampled into 1 HPLC plate
Future improvements
Sampling always run at the end, not
good for fast reactions
Allow sampling as part of ‘dispense’
And/or move vials to heat zones
to ‘start’ reactions
(Josh Denette and Kristin Price at Pfizer)
Data Handling
MDB
RAS
29
Asymmetric Hydrogenation
Successfully built capability in Sodertalje, Sweden
site end of 2012
Opportunity to purchase new equipment
CM3 (in glovebox) and SPR
Off deck hydrogenation – make use
of CM3 during long reactions
Knowledge of LEA / CM2
CM2 / CM3 used as 1 resource for
X-Coupling and Asymm. Hydrog.
Commissioning - NOW
Conclusion
We have come along way since 2007
Scavenger was the warm up act
• Scavenging and X-Coupling saved millions of dollars already
• Success led the way to keep Asymm. Hydrog. in house
• Asymm. Hydrog. will have similar if not bigger impact
Continue to develop workflows
Keep control of the data
96 x 2 x 5 x 50 = 48K data points / year
Conclusion
User community carry on sharing and learning
What’s good for me, could be good for you too.
Acknowledgments
AstraZeneca
John Leonard, Gair Ford, Barney Squires
Phil Hogan, Keith Mulholland, Andy Campbell
Rachel Munday, Kevin Leslie
Sarah Thompson, Andy Poulton
Ex AstraZeneca
Paul Murray and co. (catsci.com)
Per Ryberg and co. (SP Technical Research Institute of Sweden)
Symyx/FreeSlate
Steve Yemm, Zack Hogan, Colin Masui, Anny Tangkilisan, Rob Rosen, Eric Carlson
Peter Huefner, Grant Gavranovic, Justin Fisher, Jonathan Harris
Peter Gravil, Jos De Keijzer, Rick Sidler, Tony Mani, Guillaume Magan, Ludovic Edvard
Confidentiality Notice
This file is private and may contain confidential and proprietary information. If you have received this file in error, please notify us and
remove it from your system and note that you must not copy, distribute or take any action in reliance on it. Any unauthorized use or
disclosure of the contents of this file is not permitted and may be unlawful. AstraZeneca PLC, 2 Kingdom Street, London, W2 6BD, UK,
T: +44(0)20 7604 8000, F: +44 (0)20 7604 8151, www.astrazeneca.com

similar documents