Solid Form Screening using High Throughput Crystallisation of Pharmaceuticals Rajni Miglani, Iain Oswald, Alastair Florence Strathclyde Institute of Pharmacy and Biomedical Sciences, University of Strathclyde, Glasgow, G4 0RE, UK
Introduction Pharmaceutical development (fig 1) is frequently confronted with the phenomenon of multiple solid forms (polymorphs, solvate, co-crystals, salts and amorphous) of the same chemical entity. The properties of two solid forms of a same drug can vary substantially with potentially significant impact on pharmacokinetics, ease of manufacturing, stability and
Material and Method
bioavailability. Hence it is critical to have in depth knowledge of solid forms of any candidate molecule so that appropriate form can be taken forward as a clinical candidate.
Material and Method
Best Solid Form
Traditional experimental screening to identify appropriate solid form of a compound involves thousands of crystallisations
Drug Discovery
Regulatory Approval and Market
Clinical Development
Preclinical Development
involving material, lots of time and money.
DXR Raman (Thermo Scientific) instrument was used for collecting high quality Raman spectra, which is equipped with
high precision X, Y,Z motorized stage with a holder for well Our Focus
The main aim of this work is to develop an efficient methodology (in terms of material, time and money requirement)of physical / solid form screening using a High Throughput Crystallisation(HTC) in a multi well plate.
A layout of antisolvent screening experiment is shown in fig 2.
The 532nm laser power was used for this purpose.
3a
Figure 1: Schematic flow chart indicating process of drug discovery and drug development
In this method, solid form screening was carried out using custom made quartz 96 / 48 multi well plate formats (Hellma, Germany) formats with an automated system for collecting Raman spectra (Thermo DXR system).
plate (fig 3).
3b
Results :
Salt Screening of Amoxapine Figure 3: a) DXR Raman (Thermo Scientific) instrument b) Motorized stage with well plate reader
The array automation software, which provides various sampling options to cover different eventualities was used to allocate
Figure 2: Layout of antisolvent screen using 48 well plate
points in the well.
Various chemometric tools (group analysis, principal component analysis, hierarchical clustering) were then used to classify the samples.
This method has been applied to three drug
Salt screening of amoxapine was carried out using 8 acids in two ratio (1:1, 2:1) using 12 solvents using
(pharmaceutical) molecules;
96 well plate.
Olanzapine, Clozapine
20 novel salts have been observed during this screening. Counter ions / acids used for screening and their results (molar ratio
Novel observed forms
and
of 1:1 for amoxapine vs counter acid) are shown in table 1. Differences in the Raman spectra of representative member from each group in 1530-1650 cm-1 region are highlighted in fig 5.
were scaled up for
Amoxapine
characterization Conter ion
Methanol
Ethanol
S1
S1
Sucinnic acid Maleic acid
Results :
M2
M1
M2
S1
S1
S1
S1
M2
M2
M2
M2
M1
Toluene
Diisoprop ylether
THF
S2
S1
S3
S4
M2
M2
M1
M2
M1
M2
E1
E1
E1
E1
E1
E1
E2
E1
E1
E1
E1
A1
A1
A1
A1
---
A1
A1
A1
A1
A1
A1
C1
C1
C1
C1
F1
F2
F2
LM2
---
LM1
L-Malic acid
Polymorph screening of
M3
Ethyl Isobutyl Nitro Cyclo acetate acetate methane hexane
S1
Adipic acid
Fumaric acid
Figure 4: Snapshot of the real time Raman data collection
M2
S1
PTSA Citric acid
Solid Form Screening of Olanzapine
M3
Acetonitrile Acetone
C1 F1
C1 F2
LM1
F1
C1 F2
F1
LM1
C1
C1
F1
F1 F2
LM1
LM2
F2 LM2
F1
F2 ---
C1 F1
F2 ---
water
E1 A1
C1
C1
F1 F2
F2
LM1
LM1
Table 1: Novel forms obtained from salt screening of amoxapine in molar ratio of 1:1 (amoxapine and counter ion). The groups are colour coded according to crystalline form.
Results : Solid Form / Physical Screening of Clozapine
Polymorph screening of clozapine was
olanzapine was carried out using 48
carried out using 96 solvents using 48 solvents
solvents and two concentrations of solute
in one concentration of solute(2.4mg). In total, 96
using 96 well plate. In total, 96 crystallisations
crystallisations utilizing 230mg of solute with ~24 ml of 48
utilizing 144mg of solute with ~20 ml of 48 solvents yielded 21 novel forms
solvents yielded 5 novel forms.
6c S
6a
H N
H N N
Cl
N
7c
N
Figure 5: Raman spectra of representative member from each group of amoxapine salts
N
N
6b Figure 6: a) Twenty one groups (color coded) of novel forms obtained from HTC b) Raman spectra of representative member from each group of olanzapine c) Chemical structure of olanzapine
Grouping of these novel forms in 21 groups (according to colour code) on 96 well plate is shown in fig 6a. Differences in the Raman spectra of representative member from each group in 1400-1650 cm-1 region are
Summary
N
7a
The high throughput crystallisation method employed
7b Figure 7: a) Five Groups (colour coded) of novel forms obtained from HTC b) Raman spectra of representative member from each group of clozapine c) Chemical structure of clozapine
here has been shown to be an effective way of exploring the crystallisation
Grouping of novel forms (according to colour code) in one
space of the molecules under study. A large number of solid forms have been identified for
olanzapine, amoxapine and clozapine using HTC.
experiment on 48 well plate are shown in figure 7a. Differences in the Raman spectra of each
group (representative member) in 1400 This methodology has various advantages in the context of solid form screening when only small amounts of compound 1700cm-1 are highlighted in fig 7b. Out of 21 novel forms, 13 forms are available, produces results in a matter of days by relying on automated data collection procedures and chemo-metric analysis. The results have been characterised highlighted in fig 6b.
using single X-ray diffraction.
of multi well plate approaches are complementary to the larger scale automated methods, identifying those conditions that merit further, more detailed
investigation.
Further automation of steps would be advantageous (e.g. solvent dispensing), however as a relatively low-cost initial screen, this approach adds considerable value to the existing technology base in SIPBS.
Acknowledgement: RM thanks Commonwealth Scholarship Commission for studentship funding and CPOSS
