Ternary and quaternary phase diagrams: Key tools for chiral resolution through solution cocrystallization

Article abstract of DOI:10.1039/c3ce41141a

The goal of this contribution is to guide the reader through the construction and the understanding of quaternary phase diagrams in the pursuit of optimal conditions for a chiral resolution through cocrystallization in solution. The overall description will be illustrated by experimental results on a system involving RS-2-(2-oxopyrrolidin-1-yl) butanamide as chiral API to be separated, and S-mandelic acid as chiral coformer.

Full text of DOI:10.1039/c3ce41141a

View Article Online
View Journal
CrystEngComm
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: G. Springuel, L. Collard and T. Leyssens,
CrystEngComm, 2013, DOI: 10.1039/C3CE41141A.
This is an Accepted Manuscript, which has been through the RSC Publishing peer
review process and has been accepted for publication.
CrystEngComm
Accepted Manuscripts are published online shortly after acceptance, which is prior
www.rsc.org/crystengcomm
Volume 12 | Number 5 | May 2010 | Pages 1317–1658
to technical editing, formatting and proof reading. This free service from RSC
Publishing allows authors to make their results available to the community, in
citable form, before publication of the edited article. This Accepted Manuscript will
be replaced by the edited and formatted Advance Article as soon as this is available.
To cite this manuscript please use its permanent Digital Object Identifier (DOI®),
which is identical for all formats of publication.
More information about Accepted Manuscripts can be found in the
Information for Authors.
Please note that technical editing may introduce minor changes to the text and/or
graphics contained in the manuscript submitted by the author(s) which may alter
content, and that the standard Terms & Conditions and the ethical guidelines
that apply to the journal are still applicable. In no event shall the RSC be held
responsible for any errors or omissions in these Accepted Manuscript manuscripts or
any consequences arising from the use of any information contained in them.
HOT ARTICLE
COMMUNICATION
Hall et al.
HIGHLIGHT
Wang, Su et al.
Alginate-mediated routes to the
selective synthesis of complex metal
oxide nanostructures
Long
Ionothermal synthesis of a new
open-framework zinc phosphate
NIS-3 with low framework density
pH effect on the assembly of
metal–organic architectures
www.rsc.org/crystengcomm
Registered Charity Number 207890

CrystEngComm
Page 1 of 9Journal Name
Dynamic Article Links ►
View Article Online
DOI: 10.1039/C3CE41141A
Cite this: DOI: 10.1039/c0xx00000x
www.rsc.org/xxxxxx
ARTICLE TYPE
Ternary and Quaternary Phase Diagrams: Key Tools for Chiral
Resolution through Solution Cocrystallization.
Géraldine Springuel, Laurent Collard and Tom Leyssens*
5
Received (in XXX, XXX) Xth XXXXXXXXX 20XX, Accepted Xth XXXXXXXXX 20XX
DOI: 10.1039/b000000x
The goal of this contribution is to guide the reader through the construction and the understanding of
quaternary phase diagrams in the pursuit of optimal conditions for a chiral resolution through
cocrystallization in solution. The overall description will be illustrated by experimental results on a
10 system involving RS-2-(2-oxopyrrolidin-1-yl) butanamide as chiral API to be separated, and S-mandelic
acid as chiral coformer.
enantiospecific system, we were able to develop a new resolution
tool based on enantiospecific cocrystallization in solution.21
The goal of this article is to describe the thermodynamics behind
50 this system, introducing quaternary phase diagrams.
Developing any robust and optimal crystallization process in
solution ideally requires construction of appropriate phase
diagrams. The overall aspect of these diagrams strongly depends
on the nature of the system under study.22-25 This is also the case
55 for our resolution technique based on enantiospecific
Introduction
In recent years, cocrystals have been considered as attractive
alternate solid forms,1,2 especially for pharmaceutical compounds
15 for which salt formation is no valid option. Cocrystallization of
an active pharmaceutical ingredient (API) with a biologically
acceptable coformer can improve solubility, dissolution rate,
bioavailability, chemical stability, and physical properties of an
API.3,4 The list of new pharmaceutical cocrystals continues to
20 increase, and for some APIs a multitude of cocrystals have been
identified.
Even if the primary focus of cocrystal research remains oriented
toward the ability to control pharmaceutical properties, recent
contributions also investigate novel applications based on the
25 specific properties of these systems.5,6 Amongst these, the
interesting behavior of cocrystal systems comprising two chiral
partners has been highlighted by several research groups,7,8
reviving the idea of using cocrystallization as a chiral resolution
tool.
30 Chiral resolution is commonly used in the pharmaceutical
industry given the large number of chiral drug candidates.9,10 In
many instances, enantiomers possess different pharmacological,
and toxicological properties.11-13 Marketing of enantiopure drugs
is preferred over the racemate formulation in order to ensure a
35 higher activity for a reduced drug dosage or to avoid unwanted
side effects. As chiral synthesis is not always feasible or
economically viable, most chiral compounds are synthesized as
racemic mixtures, with a chiral resolution step somewhere along
the production process. For chiral compounds that do not easily
40 form salts chiral column chromatography often remained the only
viable alternative.1
cocrystallization
in
solution.
Although,
typically,
cocrystallization processes in solution are developed using
ternary phase diagrams,26,27-31 these no longer suffice, as both
enantiomers of the racemic mixture need to be taken into account,
60 yielding a total of 4 variables (amount of solvent, coformer, and
both enantiomers of target compound).32 We therefore develop
the concept of a quaternary cocrystal phase diagram, and
emphasize on the necessity of these phase diagrams in the pursuit
of optimal conditions for the development of a chiral resolution.
65 Although some isothermal quaternary phase diagrams have been
described for the resolution of diastereomeric salts in solution, the
thermodynamics behind an enantiospecific cocrystal system have
not yet been investigated.33 We will start by giving an overall
description of the quaternary phase diagram. In a second stage,
70 we will consider some well-chosen cross-sections of this
quaternary diagram, which according to us are very helpful in
developing an effective resolution. Finally, we will investigate
the effect of temperature, by studying how these cross-sections
vary with respect to this variable.
Materials and Methods
Starting Materials.
Recently, however, we showed how cocrystallization can be used
as an effective resolution tool for those compounds that do not
easily form salts. Opposite to salts, cocrystals frequently show
45 enantiospecific behavior, with only one of two enantiomers
forming a cocrystal with a chiral coformer.7,14-20 Using such an
75 S-2-(2-oxopyrrolidin-1-yl)butanamide (Levetiracetam) (S-1) was
purchased from Xiamen Top Health Biochem. Tech. Co., Ltd. S-
Mandelic acid (S-2) (Figure 1, right), R-mandelic acid (R-2) were
purchased from Acros Organics. (RS)-2-(2-oxopyrrolidin-1-
This journal is © The Royal Society of Chemistry [year]
[journal], [year], [vol], 00–00 | 1

CrystEngComm
Page 2 of 9
View Article Online
DOI: 10.1039/C3CE41141A
Figure 1. Chemical structures of 2-(2-oxopyrrolidin-1-yl)butanamide (1)
(left), and S-mandelic acid (S-2) (right).
yl)butanamide (RS-1) was prepared from S-2-(2-oxopyrrolidin-1-
yl)butanamide (Figure 1, left). 10g of S-2-(2-oxopyrrolidin-1-
yl)butanamide together with catalytic amount (0.05 eq) of
MeONa were added to 10 ml of MeOH. The solution was kept at
reflux under continuous stirring for 24h, and then cooled to room
temperature. The compound crystallizes spontaneously. After
5
10 filtration, the compound was washed twice with MeOH and used
as such.
Figure 2. Schematic quaternary solubility phase diagram of a four
50 component system in which two compounds are related enantiomerically
(R-1 and S-1) and crystallize out as a racemic compound (RS-1), and in
which a selective cocrystal formation can be achieved (S-1:S-2).
X-ray Powder Diffraction (XRPD).
X-ray diffraction measurements were performed with a Siemens
D5000 diffractometer equipped with a Cu X-ray source operating
15 at 40kV and 40 mA and a secondary monochromater allowing to
select the Kα radiation of Cu (λ = 1.5418 Å). A scanning range of
2θ values from 2° to 72° at a scan rate of 0.6° min-1 was applied.
quaternary phase diagrams related to cocrystal systems involving
various chiral compounds, after which we will specify two cross-
55 sections, which according to us are helpful for the development
of a chiral resolution through cocrystallization.
High-Performance Liquid Chromatography (HPLC).
General Description of Quaternary Phase Diagrams
A reverse HPLC system Waters Alliance 2695 equipped with a
20 PDA detector (Waters 2998) turned to 210 nm was used to
monitor the Waters Atlantis T3 analytical column (50 mm x 4.6
mm x 3.5 µm). The injection volume was 10 µl and column
temperature was keep at 35°C. A chiral HPLC system Waters 600
equipped with a PDA detector (Waters 996) set to 205 nm for
25 detection. A Daicel Chiralpak IA analytical column (250 mm x
4.6 mm x 5 µm) has been used. An isocratic flow rate of 1 ml/min
(80% Isohexane, 20% Ethanol) was used for elution. The
injection volume was 25 µl and column temperature was kept at
20°C. In both systems, the columns are equilibrated in the initial
30 conditions (for 10 min for reverse HPLC and for 30 min for chiral
HPLC).
As mentioned, phase diagrams are powerful and sensitive tools
for the development of crystallization processes in solution. For a
60 chiral resolution through crystallization, the optimal conditions,
combine maximum yield with a required level of enantiopurity.
In this paragraph, we give a general description of a quaternary
phase diagram for a cocrystal system involving two chiral
partners. We illustrate the theoretical diagram, using
65 experimental results on a model compound, 2-(2-oxopyrrolidin-1-
yl)butanamide (RS-1). This latter is a pharmaceutical compound
that does not easily form salts and which hitherto required chiral
chromatography for effective resolution. This molecule, shown in
Figure 1, is a racemic nootropic drug and its biologically activity
70 is essentially associated with the S-enantiomer (S-1). S-1 is
marketed under the name Levetiracetam, an active anticonvulsant
used to treat epilepsy. The R-enantiomer, R-1, is
pharmacologically inactive. S-mandelic acid (S-2) has been
chosen as cocrystal former since S-mandelic acid effectively
75 cocrystallizes with S-1 but not with R-1.21,38 Acetonitrile was
selected as a solvent, as both 1 and S-2 show moderate solubility
with respect to this solvent.
Phase Diagram Construction
Each phase diagram was obtained by preparing a set of vials with
specific compositions of RS-1, S-1, and S-2, in acetonitrile
35 according to the section of the quaternary diagram considered.
All vials were hermetically sealed and heated up to 60°C until
complete dissolution. They were stored at constant temperature
and seeded with all of the solid forms possibly occurring in the
corresponding ternary phase diagram (for each diagram the
40 appropriate selection was made from RS-1 form I,34-37 S-1, S-2,
and S-1:S-2 cocrystal). After two weeks, the system was assumed
to have reached equilibrium. For each vial, the final liquid phase
was analysed using both achiral and chiral HPLC to determine
the exact solution composition (quantities of each component).
45 The crystallized solid phases were analysed by XRPD.
The phase diagram associated with this system comprises
following six variables:
80
1
Amount of S-enantiomer of the target compound
(corresponding in this work to the enantiomer that effectively
cocrystallizes with the coformer and will be recovered in the
solid phase after resolution)
2
3
4
5
6
Amount of R-enantiomer of the target compound
Amount of coformer
Amount of solvent
Temperature
Pressure
85
Results and Discussion
In what follows, we will discuss the overall aspect of the
2
| Journal Name, [year], [vol], 00–00
This journal is © The Royal Society of Chemistry [year]

Page 3 of 9
CrystEngComm
View Article Online
DOI: 10.1039/C3CE41141A
Figure 3. Down: Detail of the upper part of ternary phase diagrams (in molar percentage) involving both enantiomers of 1 in acetonitrile at 9°C (left) and -
10°C (right). Solid lines are solubility curves, dotted lines are geometrical constructions to visualize the different stability zones. Up: These diagrams
correspond to the front side of the quaternary phase diagrams. The schematic ternary phase diagram allows an overview of the different zones of solid
state stability in suspension. Experimental data used for the construction of these diagrams is given in the supporting information.
5
Describing and representing a system with six variables is not
Table 1. Summary of Solubilities and Eutectic Compositions for the
always very convenient. To facilitate graphical representation and
interpretation, it is recommended to reduce the number of
Different Ternary Diagrams Involving S-1, R-1, S-2, and Acetonitrile at
9°C and -10°C, as Shown in Figures 3 to 5.
10 variables. As most crystallization processes are performed under
constant pressure, this variable will be held constant to
atmospheric pressure. Furthermore, diagrams will be constructed
at given temperatures, hereby allowing reducing the number of
variables to 4. Each isothermal system can then be represented in
15 the form of a 3D tetrahedron with the apexes representing the
pure components as shown in figure 2. The three sides
culminating in the solvent apex correspond to three ternary phase
diagrams typically encountered in literature. The front side,
highlighted in figure 3, represents the symmetric ternary phase
20 diagram of a racemic compound (RS-1). The right side of the
tetrahedron, highlighted in figure 4, corresponds to the
asymmetric ternary phase diagram of a cocrystal system,
involving the S-enantiomer (S-1) and the chiral coformer (S-2).
The reverse side of the tetrahedron corresponds to the ternary
25 phase diagram involving R-1 and S-2 (Figure 5), and shows a
single non-symmetric eutectic point. This latter ternary diagram
is in strong contrast with that of the S-1:S-2 system (Figure 4) as
no cocrystal is formed between the R-enantiomer (R-1) and the
chiral coformer (S-2).
9°C
-10°C
Solubility (mmole/ml Acetonitrile)
S-1/R-1
RS-1
S-2
0.1395
0.1175
0.4778
0.0881
0.0582
0.3654
Eutectic composition (mmole/ml Acetonitrile)
A/A’ S-1 : 0.1658 R-1 : 0.0312
B/B’ S-1 : 0.3214 S-2 : 0.7522
C/C’ S-1 : 0.3569 S-2 : 1.0164
S-1 : 0.0921 R-1 : 0.0110
S-1 : 0.1606 S-2 : 0.1814
S-1 : 0.1297 S-2 : 0.6960
D/D’ R-1 1.74191 S-2 : 3.25671 R-1 : 0.3733 S-2 : 0.8263
40 1 Obtained by extrapolation.
dotted line in Figure 3.
Each of the ternary phase diagrams discussed above were
determined experimentally at 9°C and at -10°C for the system
under study. Saturated liquid curves were obtained by HPLC
45 analysis of the mother liquor and the nature of the solid forms
was determined using XRPD analysis.
Some important numbers (solubility of pure components and
eutectic points) are summarized in Table 1, with the remaining
data used for the creation of figures 3 to 5, given in the
50 supporting information. Table 1 shows the pure enantiomer S-1,
and a fortiori R-1, to have higher solubility compared to the
racemic compound at 9°C and -10°C. At 9°C, S-1 is about 1.8
times more soluble than RS-1, while their relative solubility
decreases slightly to 1.51 at -10°C. For the data concerning the
55 cocrystal ternary phase diagram, one observes the compound of
interest S-1 to show a solubility that is about 3.42 times lower
30 As R-1 was not at our disposal, the reciprocity of the system was
used to construct certain zones of the diagrams involving this
enantiomer. The ternary diagram involving R-1 and S-2 is
identical to that constructed using S-1 and R-2. Furthermore, the
ternary diagram of the racemic RS-1 compound is symmetric and
35 thus the part of the diagram that requires excess amounts of R-1
can be obtained by a mere symmetry operation, represented by a
This journal is © The Royal Society of Chemistry [year]
Journal Name, [year], [vol], 00–00 | 3

CrystEngComm
Page 4 of 9
View Article Online
DOI: 10.1039/C3CE41141A
Figure 4. Down: Detail of the upper part of ternary phase diagrams (in molar percentage) involving S-1 and S-2 in acetonitrile at 9°C (left) and -10°C
(right). An equimolar S-1:S-2 cocrystal is formed within a certain composition range. Solid lines are solubility curves, dotted lines are geometrical
constructions to visualize the different stability zones. Up: These diagrams correspond to the right side of the quaternary phase diagrams. The schematic
ternary phase diagram permits an overview of the different zones of solid state stability in suspension. Experimental data used for the construction of these
diagrams is given in the supporting information
5
Figure 5. Down: Detail of the upper part of ternary phase diagrams (in molar percentage) involving R-1 and S-2 in acetonitrile at 9°C (left) and -10°C
(right). No cocrystal is formed between R-1 and S-2. Solid lines are solubility curves, dotted lines are geometrical constructions to visualize the different
stability zones. Up: These diagrams correspond to the backside of the quaternary phase diagrams. The schematic ternary phase diagram permits an
overview of the different zones of solid state stability in suspension. Experimental data used for the construction of these diagrams is given in the
supporting information
10
compared to the coformer S-2 at 9°C (+/- 4.14 times at -10°C).
This relatively important difference in solubility between
15 cocrystal components could explain the observed asymmetry in
the ternary diagram shown in Figure 4.39,40 Unsurprisingly the
zone in which the S-1:S-2 cocrystal is the only stable solid phase
in suspension skews toward the S-2 axis due to the higher
4
| Journal Name, [year], [vol], 00–00
This journal is © The Royal Society of Chemistry [year]

Page 5 of 9
CrystEngComm
View Article Online
DOI: 10.1039/C3CE41141A
Figure 6. Cross-sections of the tetrahedral diagram, with a fixed solvent molar percentage (here, 89mol%) at 9°C (right) and -10°C (left). Each top of the
triangle corresponds to a suspension of pure components in the appointed molar percentage of solvent.
solubility of this latter.41 Figure 4, furthermore, shows
incongruent diagrams at both temperatures; implying that a
suspension of a pure cocrystal S-1:S-2 phase is unstable and will
spontaneously transform either to a mixture of S-1:S:2 cocrystal
diagram. Technically, HPLC analyses can be realized for an
ensemble of different compositions inside the tetrahedron, with
the measured liquid composition referring to a point on the
saturated liquid surface. However, this would be a lengthily and
5
and S-1, or to a pure S-1 suspension, depending on the amount of 40 not easily performable task especially when working at different
solvent.
temperatures. We therefore present two cross sections which
according to us are easily obtainable and useful for the
development of an effective resolution. The general idea of these
surfaces is to reduce a three dimensional representation to more
10 As for the final ternary diagram (Figure 5), the solubility of R-1
is seen to strongly depend on the concentration of S-2 in solution,
and vice versa. At respectively 9°C and -10°C, the solubility of
R-1 at the eutectic point is 12.48 times and 4.23 times more 45 easily readable two dimensional tools.
Cross Section: Fixed Molar Percentage of Solvent
important than the solubility of pure R-1, clearly implying a
15 strong intermolecular interaction between these compounds in
solution, even if they do not form a thermodynamically stable
cocrystal in the solid phase.
Although in the current contribution we selected an
enantiospecific cocrystal system, resolution remains possible
20 even if a diasteriomeric pair of cocrystals is formed. For these
cases, the diagram represented in Figure 5 will be that of a
ternary phase diagram of a cocrystal system. (A schematic
representation of this case is shown in the supporting
information).
25 It must also be noted that, beside diastereomeric cocrystals, other
multicomponent cocrystal systems can also occur eg. comprising
both enantiomers as well as the coformer.42 Such a phase would
be stable in the centre of the tetrahedral diagram and would bring
increased complexity to the diagram. We here decided to work on
30 a system for which no such phase is observed.
The first of these diagrams is obtained by making a cross section
parallel to the basis of the tetrahedron. This procedure is
equivalent to fixing the molar percentage of solvent. Ideally, this
50 section has to be located below solubility points of pure
components. Although cases can appear where the cocrystal is
much less soluble than its components, one should not expect this
to be the general rule.43
The screening procedure consisted in preparing a set of vials with
55 given compositions of RS-1, S-1, and S-2, allowing for
construction of the part of the diagram enriched in S-1. The part
of the diagram enriched in R-1, could not be obtained
straightforwardly due to lack of R-1, and was hence constructed
using RS-1, S-1, and R-2. The full diagram based on R-1, S-1,
60 and S-2 is the exact mirror image of the one obtained for R-1, S-
1, and R-2. The left part of the cross section considered here was
thus obtained using symmetry properties of the system.
A variable amount of acetonitrile was added to achieve a fixed
solvent molar percentage (89 mol%) in each vial.
Developing a Chiral Resolution
65 In the thus created isothermal plane (Figure 6), each top of the
triangle corresponds to a suspension of the pure component in the
appointed molar percentage of solvent. Relative compositions of
S-1, R-1, and S-2 can be read in a similar manner as for common
To develop a chiral resolution, ideally one would have at his
disposal the entire quaternary phase diagram, with the interior of
the diagram characterized by an ensemble of saturated liquid
35 surfaces, similar to the liquidus lines appearing in a ternary phase
This journal is © The Royal Society of Chemistry [year]
Journal Name, [year], [vol], 00–00 | 5

CrystEngComm
Page 6 of 9
View Article Online
DOI: 10.1039/C3CE41141A
Figure 7. Racemic Cross-section of the tetrahedral diagram (overall amount of R-1=S-1) at -10°C.Detail of the upper part, is shown below.
ternary phase diagrams. Screening this surface allows identifying
different zones such as presented in figure 6, showing the cross
sections for 89 mol% solvent at 9°C and -10°C, clearly
liquid curve, which implies that only one component needs to be
defined to define the entire system.
The main utility of this diagram is to show whether or not an
5
highlighting the temperature effect. As expected, lowering the 40 effective resolution is feasible at a given temperature. For this to
crystallization temperature leads to a reduction of the size of the
undersaturated zone. A more detailed description of these
diagrams, as well as the analysis of the different stability zones,
occur following conditions have to be satisfied:
1. Formation of a cocrystal phase involving only one of the API
enantiomers
10 can be found elsewhere.21 These kind of sections yield insight
into the different phases appearing inside the quaternary phase
diagram: a liquid monophasic field (grey lines in Figure 6), four
biphasic fields (R-1 + Liquid, RS-1 + Liquid, S-1:S-2 + Liquid ,
and S-2 + Liquid) surrounding three triphasic fields (R-1 + RS-1
15 + liquid, RS-1 + S-1:S-2 + Liquid, and S-2 + S-1:S-2 + Liquid).
2. Formation of a zone in the cross section where this phase
appears alone (pink triangles in Figure 6), or in combination
with the pure coformer (blue circles in Figure 6)
3. Crossing of this zone (pink triangles in Figure 6) with the so-
called "racemic line" (red-dotted line in Figure 6).
Starting from a racemic mixture of a chiral API, Figure 6, allows
45
The Gibbs phase rule applies for this diagram. Describing the 50 determining the quantities of coformer required (for a given
intensive degrees of freedom (DoF) in a system at equilibrium, in
function of the number of components and the number of phases
in the system,44 the rule can be written as F=C-P+2, with F the
20 intensive degrees of freedom, C the number of components, and P
amount of solvent) to reach the zone in which the cocrystal is the
thermodynamically stable phase in suspension. Crystallization of
the enantiospecific cocrystal will lead to enantiomeric excess in
solution in favour of the R-enantiomer. As shown in Figure 6,
the number of phases. In a system in which temperature and 55 starting from a racemic mixture, such a resolution procedure is
pressure are constant, as is the case here, the rule can be reduced
to F’’=C-P. Our model system counts four components (C=4): R-
1, S-1, S-2, and the solvent. In this system, the liquid monophasic
25 field (P=1) has three degrees of freedom (DoF), implying that the
knowledge of at least three intensive variables is required to
identify the composition of each phase; in this case, the
composition of the liquid. If the amount of three components is
known, the fourth can be deduced from the first three. In a similar
30 manner, a biphasic fields (P=2) has two DoF. The composition of
the solid phase is 100% of pure component, cocrystal or racemic
compound, depending on the biphasic field considered. At this
stage, the liquid composition is given by a fixed point on the
saturated liquid surface, and can be deduced knowing the amount
35 of only two components. The triphasic fields (P=3) have only one
DoF because the liquid composition evolves along a saturated
barely possible at 9°C, but becomes easily feasible at -10°C,
clearly emphasizing the importance of the temperature for these
resolution processes.
Cross Section: Racemic Composition Plane
60 Although the diagram mentioned above is convenient to illustrate
the feasibility of a chiral resolution the most useful diagram is the
so-called "racemic composition plane". This plane, illustrated on
the top left of Figure 7, is the cross section in the tetrahedral
diagram that contains all points for which the overall composition
65 of the molecule to be resolved (liquid plus solid phases) is
racemic (R-1=S-1; or ee =0). This plane is crucial for
determining suitable conditions under which resolution can occur,
and furthermore orients the trials to allow studying the
effectiveness of the chiral resolution. To obtain this plane, the
70 screening procedure consisted in preparing a set of vials of
6
| Journal Name, [year], [vol], 00–00
This journal is © The Royal Society of Chemistry [year]

Page 7 of 9
CrystEngComm
View Article Online
DOI: 10.1039/C3CE41141A
Table 2. Efficiency Results of Cocrystallization in Solution from an Initial
different conditions all over the diagram. Figure 7, shows this
section for the model system at -10°C, the temperature which, as
illustrated above, is suitable for chiral resolution. Although
frontiers (black lines) are added for a better visual effect, these
cannot be directly deduced from HPLC analysis. The
composition of the liquid phase remaining after crystallization
cannot be deduced directly from this section. Indeed, for a
biphasic field, the equilibrium composition of the liquid phase is
identified by the conode or tie line connecting the composition of
Racemic Mixture of RS-1 for a Temperature of -10°C.
Initial composition (mmole/ml Acetonitrile)
Efficiency
5
RS-1
S-2 (R-2)a
ACN 85 mol %
10 the pure phase with the overall composition. This point is outside
the two-dimensional space considered here, and requires
knowledge of the entire quaternary phase diagram.
However, this diagram is extremely useful as it immediately
shows all conditions that lead to an effective resolution (S-1:S-2
15 + liquid zone (pink triangles in Figure 7); and S-2 + S1:S-2 +
liquid zone (bleu circles in Figure 7)). Knowing the initial
amount of RS-1, one can deduce the amounts of solvent and S-2
required for an effective resolution.
1
2
3
1.6349
1.3698
1.4818
1.8041
2.0691
1.9571
0.83
0.81
0.79
ACN 89 mol %
As mentioned, this diagram can also help to orient trials that
20 allow studying the effectiveness of resolution. Resolution
efficiency is given by the ratio between the masse of S-1 in the
recovered solid and the initial mass of S-1 in the racemate.
As mentioned above, lowering the temperature will lead to an
increased zone in which resolution can occur. In addition, for a
25 given temperature, decreasing the amount of solvent is expected
to increase the efficiency, as less S-1 will be lost in the liquid
phase. As stated above, the liquid phase composition cannot be
determined directly from this diagram. To check efficiency, we
therefore decided to screen within the zone where the S-1:S-2
30 cocrystal is the only stable phase (experimental points refer to
Figure 7). Results in Table 2 confirm the efficiency to depend on
the amount of solvent, with the efficiency increasing by 30%
when changing the amount of solvent from 93mol% to 85mol%.45
The efficiency furthermore varies according to the amount of S-2,
35 with respect to RS-1.
From the data in Table 2, we show that the resolution technique
through cocrystallization can be extremely effective, as
conditions were identified for which more than 80% of the initial
amount of S-1 (points 1 and 2 in Table 2 are assessed to a
40 efficiency of 83% and 81% respectively) is recovered from
solution in a single crystallization step. HPLC analysis was
performed on the recovered co-crystal solid phase (after an
appropriate filtration and washing procedure to remove remaining
mother liquor), and showed 100% enantiopurity with respect to
45 the desired enantiomer, here S-1.
4
5
6
7
8
1.1000 (1.0997)
1.0180 (1.0243)
0.9596 (0.9548)
0.9981 (0.9992)
0.8854
1.3085 (1.3088)
0.74 (0.77)
0.74 (0.74)
0.73 (0.73)
0.69 (0.75)
0.63
1.3905 (1.3843)
1.4489 (1.4538)
1.4103 (1.4094)
1.5231
ACN 93 mol %
0.60
0.49
9
0.5754
0.5120
0.8914
0.9547
10
60 a Between brackets, repeating experiments using R-2 as coformer.
model system for which all of the different phase diagrams were
constructed. The potency of the resolution technique through
cocrystallization is illustrated, as conditions are identified for
which more than 80% of the initial amount of one enantiomer can
65 be recovered in the cocrystal solid phase in
a single
crystallization step. Cocrystals formed during the resolution are
100% enantiopure. The presented system is not an isolated case,
as at least two other systems have been tested by the authors.
Obviously, the R-1 can be recovered within the cocrystal phase,
if the coformer of opposite chirality is used. In this case, a R-1:R-
2 cocrystal will be formed, leaving the S-1 in the mother liquor.
Experiments 4 to 7 were repeated using the R coformer, and
50 similar yields and enantiopurity were obtained (see Table 2).
70 Notes and references
Institute of Condensed Matter and Nanosciences, Université Catholique
de Louvain, Louvain-La-Neuve, Belgique. Fax: +32 (0)10 472707; Tel:
+32 (0)10 472811; E-mail: tom.leyssens@uclouvain.be
† Electronic Supplementary Information (ESI) available: data used to
75 create the diagrams, schematic representation of a quaternary phase
diagram wherein a diastereomeric pair of cocrystals is formed. See
DOI: 10.1039/b000000x/
Conclusions
In this study, we introduced quaternary phase diagrams for the
development and optimisation of robust chiral resolution
processes through solution cocrystallization. Besides an overall
55 description of the diagram, some strategically well-chosen cross
sections were detailed and conditions to achieve effective
resolution identified. Theoretical concepts are illustrated using a
1
2
B. S. Sekhon, Int. J. PharmTech Res., 2010, 2, 1584.
P. Vishweshwar, J. A. McMahon, J. A. Bis, M. J. Zaworotko,
J. Pharm. Sci., 2006, 95, 499.
80
3
N. Shan, M. J. Zaworotko, Drug Discovery Today, 2008, 13,
440.
This journal is © The Royal Society of Chemistry [year]
Journal Name, [year], [vol], 00–00 | 7

CrystEngComm
Page 8 of 9
View Article Online
DOI: 10.1039/C3CE41141A
4
5
6
W. Jones, W. D. S. Moetherwell, A. V. Trask, MRS Bull.,
2006, 31, 875.
J. Urbanus, C. P. M. Roelands, J. Mazurek, D. Verdoes, J. H.
ter Horst, CrystEngComm, 2011, 13, 2817.
K. H.-Y. His, K. Chadwick, A. Fried, M. Kenny, A. S.
Myerson, CrystEngComm, 2012, 14, 2386.
H. G. Brittain, Cryst. Growth Des., 2011, 11, 2500.
B. S. Sekon, Ars Pharm., 2009, 50, 99.
O. McConnell, A. Bach II, C. Balibar, N. Byrne, Y. Cai, G.
Carter, M. Chlenov, L. Di, K. Fan, I. Goljer, Y. He, D. Herold,
M. Kagan, E. Kerns, F. Koehn, C. Kraml, V. Marathias, B.
Marquez, L. McDonald, L. Nogle, C. Petucci, G.
Schlingmann, G. Tawa, M. Tischler, R.T. Williamson, A.
Sutherland, W. Watts, M. Young, M.-Y. Zhang, Y. Zhang, D.
Zhou, D. Ho, Chirality, 2007, 19, 658.
37 C. Herman, T. Leyssens, F. Debaste, B. Haut, J. Cryst.
Growth, 2012, 342, 57.
38 G. Springuel, B. Norberg, K. Robeyns, J. Wouters, T.
Leyssens, Cryst. Growth Des., 2012, 12, 475.
39 T. Leyssens, G. Springuel, R. Montis, N. Candoni, S. Veesler,
Cryst. Growth Des., 2012, 12, 1520.
40 R. A. Chiarella, R. J. Davey, M. L. Peterson, Cryst. Growth
Des., 2007, 7, 1223.
41 A. Ainouz, J-R. Authelin, P. Billot, H. Lieberman, Int. J.
Pharm., 2009, 374, 82.
42 S. Müller, M. C. Afraz, R. de Gelder, G. J. A. Ariaans, B.
Kaptein, Q. B. Broxterman, A. Bruggink, Eur. J. Org. Chem.,
2005, 1082.
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
7
8
9
43 R. Thakuria, A. Delori, W. Jones, M. P. Lipert, L. Roy, N.
Rodriguez-Hornedo,
Int.
J.
Pharmaceut.,
2013,
10 S.C. Stinson, Chem. Eng. News, 2001, 79, 56.
11 H. Lorenz, F. Capla, D. Polenske, M.P. Elsner, A. Seidel-
Morgenstern, J. Univ. of Chem. Tech. and Met., 2007, 42 ,5.
12 K.A. Lyseng-Williamson, Drugs, 2011, 71, 489.
13 L. A. Nguyen, H. He, C. Pham-Huy, Int. J. of Biomedical Sci.,
2006, 2, 85.
14 C. Kassai, Z. Juvancz, J. Bálint, E. Forgassy, D. Kozma,
Tetrahedron, 2000, 56, 8355.
15 S. Hanessian, R. Saladino, R. Margarita, M. Simard, Chem.
Eur. J., 1999, 5, 2169.
16 P. Thorey, P. Bombicz, I. M. Szilágyi, P. Molnár, G.
Bánsághi, E. Székely, B. Simándi, L. Párkányi, G. Pokol, J.
Madarász, Thermochimica Acta, 2010,497, 129.
17 E. Székely, G. Bánsághi, P. Thorey, P. Molnár, J. Madarász,
L. Vida, B. Simándi, Ind. Eng. Chem. Res., 2010, 49, 9349.
18 K. Nemák, M. Ács, Z. M. Jászay, D. Kozma, E. Forgassy,
Terahedron, 1996, 52, 1637.
19 M. Kawashima, A. Hirayama, Chem. Letters, 1991, 763.
20 M. R. Caira, L. R. Nassimbeni, J. L. Scott, A. F. Wildervanck,
J. Chem. Crystallogr. 1996, 26, 117.
21 G. Springuel, T. Leyssens, Cryst. Growth Des., 2012, 12,
3374.
22 J. W. Schroer, C. Wibowo, K. M. Ng, AIChE J., 2001, 47,
369.
23 L. Codan, M. U. Bäbler, M. Mazzotti, Cryst. Growth Des.,
2010, 10, 4005.
24 V. S. Sistla, H. Lorenz, A. Seidel-Morgenstern, Cryst. Growth
Des., 2013, 13, 2548.
25 L. Codan, M. U. Bäbler, M. Mazzotti, Org. Process Res. Dev.,
2012, 16, 294.
26 Typically ternary phase diagrams of cocrystal systems are
constructed at a given temperature, varying the amounts of
solvent, compound of interest and coformer.
27 K. Chardwick, R. Davey, G. Sadiq, W. Cross, R. Pritchard,
CrystEngComm, 2009, 11, 412.
28 S. Boyd, K. Back, K. Chardwick, R. J. Davey, C. C. Seaton, J.
Pharm. Sci., 2010, 99, 3779.
29 H. Lorenz, A. Seidel-Morgenstern, Thermochimica Acta,
2002, 382, 129.
30 J. H. ter Horst, M. A. Deij, P. W. Cains, Cryst. Growth Des.,
2009, 9, 1531.
31 P. Marchand, L. Lefèbvre, F. Querniard, P. Cardinaël, G.
Perez, J-J, Counioux, G. Coquerel, Tetraherdon: Asymmetry,
2004, 15, 2455.
32 As for the ternary phase diagrams, we consider a constant
temperature for each quaternary diagram.
33 F. Querniard, J. Linol, Y. Cartigny, G. Coquerel, J. Therm.
Anal. Cal., 2007, 90, 359.
34 Two polymorphic forms and one hydrate of RS-1 are known.
As Form I is the most stable under the given conditions, it is
the only form considered.
35 C. Herman, T. Leyssens, V. Vermeylen, V. Halloin, B. Haut,
J. Chem. Thermodyn., 2011, 43, 677
36 C. Herman, B. Haut, V. Halloin, V; Vermylen, T. Leyssens,
Org. Process Res. Dev., 2011, 15, 774.
http://dx.doi.org/10.1016/j.ijpharm.2012.10.043.
44 G. M. Barrow, in Physical Chemistry, McGraw-Hill
Kogakusha, Tokyo, 3rd edn., 1973, ch. 20, pp. 539–579.
45 One can only reduce the amount of solvent to a certain extent,
as otherwise one is no longer able to dissolve everything prior
to crystallization.
8
| Journal Name, [year], [vol], 00–00
This journal is © The Royal Society of Chemistry [year]

Page 9 of 9
CrystEngComm
View Article Online
DOI: 10.1039/C3CE41141A
Table of Contents
The goal of this contribution is to guide the reader through the construction and the
understanding of quaternary phase diagrams in the pursuit of optimal conditions for a
chiral resolution using solution cocrystallization. The overall description will be
illustrated by experimental results on a system involving RS-2-(2-oxopyrrolidin-1-yl)
butanamide as chiral API to be separated, and S-mandelic acid as chiral coformer.