Solid phase behavior in the chiral systems of various 2-hydroxy-2-phenylacetic acid (mandelic acid) derivatives

Article abstract of DOI:10.1021/je500845d

The solid phase behavior of a series of monosubstituted F-, Cl-, Br-, I-, and CH₃- and two 2,4-halogen-disubstituted 2-hydroxy-2-phenylacetic acid (mandelic acid) derivatives was investigated. The study includes detailed information about melting temperature, melting enthalpy, X-ray diffraction data, as well as selected binary phase diagrams of the respective chiral systems. Aside from the known metastable conglomerate 2-chloromandelic acid, evidence for two more metastable conglomerates was found.

Full text of DOI:10.1021/je500845d

Article
pubs.acs.org/jced
Solid Phase Behavior in the Chiral Systems of Various 2‑Hydroxy-2-
phenylacetic Acid (Mandelic Acid) Derivatives
Jan von Langermann,*,†,‡ Erik Temmel,‡ Andreas Seidel-Morgenstern,‡,§ and Heike Lorenz*,‡
†Institute of Chemistry, University of Rostock, Albert-Einstein-Strasse 3A, 18059 Rostock, Germany
‡Max Planck Institute for Dynamics of Complex Technical Systems, Sandtorstrasse 1, 39106 Magdeburg, Germany
§Chemical Process Engineering, Otto von Guericke University Magdeburg, Universitatsplatz 2, 39106 Magdeburg, Germany
̈
S
* Supporting Information
ABSTRACT: The solid phase behavior of a series of
monosubstituted F-, Cl-, Br-, I-, and CH3- and two 2,4-
halogen-disubstituted 2-hydroxy-2-phenylacetic acid (mandelic
acid) derivatives was investigated. The study includes detailed
information about melting temperature, melting enthalpy, X-
ray diffraction data, as well as selected binary phase diagrams of
the respective chiral systems. Aside from the known metastable conglomerate 2-chloromandelic acid, evidence for two more
metastable conglomerates was found.
enantiomers in equal and ordered quantities in its crystal
structure. Within a binary phase diagram two symmetrical
eutectics are found, shown in Figure 1A. Enantioselective
crystallization is only feasible from enriched solutions, e.g.,
obtained via chromatography or partially selective synthesis.9,10
(b) Simple eutectics (conglomerates) are basically equi-
molecular mixtures of two crystalline enantiomers that are, in
theory, separable via mechanical techniques. The corresponding
phase diagram shows a single eutectic exactly at a mole fraction
of 0.5, Figure 1B. Most importantly, pure enantiomers can be
obtained directly from a racemic mixture, which represents a
significant advantage from a crystallization separation point of
view. Thus, conglomerates are typically preferred over racemic
compounds.
(c) Solid solutions (pseudoracemates), which are relatively
rare for chiral substances, show an extraordinary phase
behavior. For this type any mixture of enantiomers forms a
unique crystalline state with a stochastic distribution of the
elementary building blocks, which also explains its designation
as solid solutions. Phase diagrams show at the racemic
composition a melting temperature maximum (I), a melting
temperature minimum (III), or an identical melting temper-
ature (II) as found for the enantiomers, Figure 1C. In theory
such pseudoracemates are basically inseparable by classical
crystallization processes. However, processes with a series of
crystallization steps (fractional crystallization) may be used for
enantioenrichment.11,12
1. INTRODUCTION
Enantiomers are pairs of chiral molecules that are fully identical
except its mirror-image symmetry. These closely related
molecules possess identical physical-chemical properties, e.g.,
melting and boiling point, with the only exception that plane-
polarized light is rotated to different directions. Aside from
these fundamental physical-chemical properties enantiomers
also show strong differences in biological activity.1 While one
enantiomer may result in the desired positive effect, its
counterpart molecule could be harmful. A well-known example
is the drug Contergan, which contained thalidomide as a
racemic mixture and was used as a mild sedative for pregnant
women.2 While the target enantiomer, (+)-(R)-thalidomide,
results in the desired effect, the other enantiomer, (−)-(S)-
thalidomide, causes unfortunately strong teratogenic effect at
the unborn child. As a result up to 20000 children were born
worldwide with body deformities, which eventually forced
pharmaceutical companies to remove the drug from the market
in 1961 (Europe) and 1962 (Canada).
Consequently, nowadays almost all pharmaceuticals are
provided as enantiopure preparations to prevent such negative
effects. Unfortunately most chemical syntheses of chiral
substances on industrial scale result in racemic mixtures,
which require subsequent enantioseparation steps. Well-
established techniques include chiral chromatography, enantio-
selective membrane separation, and crystallization.3−5 From
these methods crystallization has repeatedly proven its
applicability and cost effectiveness, especially on an industrial
scale.6 However, detailed information about the physical−
chemical properties of these substances, including their phase
diagrams, is essential to plan and optimize crystallization
processes.7,8 Herein enantiomeric substances show three basic
types of phase behavior (Figure 1).1
In this study 18 substituted 2-hydroxy-2-phenylacetic acid
(mandelic acid) derivatives were synthesized, in racemic and
enantiomerically pure form, and their physical−chemical
properties investigated and discussed in detail (Table 1). The
Received: September 16, 2014
Accepted: January 9, 2015
(a) Binary (Racemic) compounds, which represent the vast
majority, show a crystalline racemate that consists of both
© XXXX American Chemical Society
A
DOI: 10.1021/je500845d
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
Figure 1. Schematic binary phase diagrams of chiral substances that form at racemic composition (A) a binary compound, (B) a simple eutectic, and
(C) solid solutions; Rs − (R)-enantiomer, solid, Ss - (S)-enantiomer, solid, RSs − racemic compound, solid, l. - liquid.
Table 1. Specifications of the Investigated Mandelic Acid Derivatives
final mole fraction
entry
R1
R2
R3
name
mandelic acid
CAS no.
origin
purity
analysis method
1
2
3
4
5
H
H
H
included as literature reference
included as literature reference
included as literature reference
included as literature reference
racemic: 10421-85-9
F
H
F
H
H
F
2-fluoromandelic acid
3-fluoromandelic acid
4-fluoromandelic acid
2-chloromandelic acid
H
H
Cl
H
H
H
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
synthesis
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
> 0.995
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
HPLC, DSC, XRPD
(S)-enantiomer: 52950-19-3
racemic: 16273-37-3
6
H
H
Br
H
H
Cl
H
H
Br
H
H
Cl
H
H
Br
3-chloromandelic acid
4-chloromandelic acid
2-bromomandelic acid
3-bromomandelic acid
4-bromomandelic acid
(S)-enantiomer: 32222-43-8
racemic: 492-86-4
7
(S)-enantiomer: 76496-63-4
racemic: 7157-15-5
8
(S)-enantiomer: 52923-26-9
racemic: 49839-81-8
9
(S)-enantiomer: 52950-20-6
racemic: 6940-50-7
10
(S)-enantiomer: 123484-90-2
racemic: 89942-35-8
11
12
13
14
15
I
H
H
H
I
2-iodomandelic acid
3-iodomandelic acid
4-iodomandelic acid
2-methylmandelic acid
3-methylmandelic acid
H
I
racemic: 873963-87-2
H
H
racemic: 70466-89-6
CH3
H
H
H
H
racemic: 85589-35-1
CH3
racemic: 65148-70-1
(S)-enantiomer: 132748-42-6
racemic: 18584-20-8
16
17
18
H
Cl
F
H
H
H
CH3 4-methylmandelic acid
(S)-enantiomer: 75172-62-2
F
2-chloro-4-fluoromandelic acid racemic: 365525-72-0
(S)-enantiomer: 1630325-00-6 synthesis
Cl
4-chloro-2-fluoromandelic acid racemic: 1214345-56-8
synthesis
(S)-enantiomer: 1630519-50-4 synthesis
selected derivatives comprise o-, m-, and p-substituted
compounds with halogen atoms Cl, Br, and I and a methyl
group as potential substitutents (Table 1). This unique class of
substances is considered as highly valuable intermediates for the
direct synthesis of various pharmaceuticals.13,14 Furthermore,
mandelic acid and also its derivatives are used as chiral resolving
agents. Despite its synthetic values detailed crystallization
relevant physical-chemical data is only known for a few main
mandelic acid derivatives. For example, 2-chloromandelic acid
is an alternative intermediate for the synthesis of (S)-
clopidogrel hydrogensulfate, a major pharmaceutical drug
(commercially produced by Bristol-Myers Squibb and Sanofi,
trade name: Plavix). The solid phase behavior of 2-
chloromandelic acid was recently intensively studied in detail
and its unusual behavior reported.10,15−19 It was characterized
as a racemic compound with a eutectic at a mole fraction of x =
0.56 and a corresponding eutectic temperature of 360.7 K. In
addition, a metastable conglomerate was found with a eutectic
temperature of ca. 356.2 K.19 This detailed knowledge allows
the use of other potential crystallization pathways via the
conglomerate by avoiding a crystallization of the racemic
compound. Another major example is 3-chloromandelic acid
that exhibits different polymorphic forms for the racemic
mixture and enantiomer.20−22 These polymorphs can be
transformed into each other by a careful selection of
crystallization procedures.
2. EXPERIMENTAL SECTION
2.1. Chemicals. All monosubstituted derivatives of
benzaldehyde, sodium cyanide, and sodium sulfate were
obtained from Sigma-Aldrich, USA. 2-Chloro-4-fluoro and 4-
chloro-2-fluorobenzaldehyde and the hydroxynitrile lyase from
B
DOI: 10.1021/je500845d
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
Manihot esculenta were a gift from Julich Fine Chemicals (now
Codexis), Germany/USA. Diisopropylether and poly(vinyl
alcohol) (as Mowiol 18-88) was purchased from Fluka,
Germany. Concentrated hydrochloric acid and concentrated
acetic acid were obtained from Merck, Germany. Methanol was
provided by Fisher Scientific, USA. Triethylammonium acetate
was used as 1 mol·kg−1 solution from Calbiochem, (Merck
Millipore, Germany). The immobilization reagent Sylgard 184
was a product sample from Dow Corning, USA. Deionized
water was used throughout the study.
2.2. Synthesis of Mandelic Acid Derivatives. All
required mandelic acid derivatives were synthesized from the
respective benzaldehyde derivatives and HCN yielding the
benzaldehyde cyanohydrin, which was hydrolyzed to form the
final product (Figure 2). All reactions including or forming
hydrogen cyanide were executed within a well ventilated fume
hood with an electrochemical HCN-detector for self-
protection.
within all shown investigations (see also Table 1). Exemplary
analysis results are assembled in the Supporting Information
(SI) file available with this article.
2.3. DSC Analysis. DSC experiments were performed with
a DSC131 differential scanning calorimeter from Setaram
(France). Pure samples were crushed into a fine powder, (0.010
to 0.015) g thereof given into an aluminum crucible, and then
used for the determination of melting point and enthalpy. For
the investigation of binary phase diagrams calculated amounts
of racemate and enantiomer were cogrounded and dissolved in
acetone. After evaporation of the solvent, the solid was crushed
again into a fine powder and ca. 15 mg thereof used for DSC
analysis.
DSC measurements were performed in closed aluminum
crucibles at a heating rate of 2 K·min−1 in a temperature range
from 30 to 140 (160 °C, if higher melting temperatures were
found) and highly pure helium with a mole fraction purity of >
0.99999 % (5.0 purity) as purge gas. The DSC apparatus was
regulary calibrated against temperature and enthalpy of fusion
of water and highly pure standard materials (indium, tin, lead)
also with respect to the heating rates applied. The
determination of melting/liquidus and solidus temperatures
from the DSC curves was executed as described in previous
work.24,25
2.4. XRPD Analysis. All solid samples were analyzed
throughout this study by X-ray powder diffraction (XRPD) to
identify the present solid phases. For this purpose a X’Pert Pro
diffractometer (PANalytical GmbH, Kassel, Germany) with Cu
Kα radiation was used. Measurements covered a 2Theta-range
from 3 to 40° with a step size of 0.0167° and a counting time of
50 s per step. Mostly samples were studied on (background-
free) Si single crystal discs.
2.5. Induction of Metastable Conglomerates. 2-
Bromomandelic acid or 2-chloro-4-fluoromandelic acid was
grounded to a fine powder and then completely dissolved in
acetone. A few very small crystals (typically 2−4 particles) of
the metastable conglomerate of 2-chloromandelic acid were
added and the resulting mixture covered with a punctured
Parafilm, which facilitated a very slow evaporation process.
After total evaporation of the solvent the resulting solid was
carefully ground and then subjected to analysis with DSC and
XRPD.
2.6. HPLC Analysis. The enantiomeric composition and
purity of the investigated phases of mandelic acid and its
derivatives was verified before DSC and XRPD measurements
via High Performance Liquid Chromatography (HPLC). All
chromatographic analyses were performed with an Agilent HP
1200 using the same chiral stationary phase column (Chiro-
biotic T 250·4.6 mm, particle size: 5 μm) and constant
parameters to ensure comparability between the results. The
eluent consisted of 80 % 1 % (m/v) triethylammonium acetate
in H2O and 20 % methanol and was adjusted to a pH of 4.02
with concentrated acetic acid. A temperature of 25 °C and a
flow of 0.5 mL·min−1 were applied to achieve reasonable
retention times. The samples were dissolved in the mobile
phase and adjusted to a concentration of one weight percent
(cinj = 1 % (m/v)). A total of 5 μL of all solutions were injected
into the system and investigated with a diode array detector
(DAD) at 254 nm.
Figure 2. Synthesis of mandelic acid derivatives (enantiomerically pure
and racemic mixture).
The required amount of sodium cyanide (5-fold excess) was
dissolved in water and chilled to 5 °C. Afterward concentrated
hydrochloric acid was slowly added to the cyanide mixture until
pH < 2 was obtained. Throughout the whole procedure the
solution temperature was kept below 10 °C to prevent
evaporation of toxic hydrogen cyanide. The resulting aqueous
solution was extracted 3 times with diisopropylether and the
resulting diisopropylether-solutions were combined. (A)
Enantiopure mandelic acid derivative: The respective benzalde-
hyde derivative and immobilized enzyme were added to the
hydrogen cyanide solution and stirred overnight at room
temperature. Immobilized enzyme was obtained as recently
reported.23 Afterward the immobilized enzyme was filtered off,
the resulting solution dried with sodium sulfate and the solvent
evaporated under reduced pressure. The resulting raw
cyanohydrin was then refluxed with an excess of concentrated
hydrochloric acid for 2 h. The resulting solution was allowed to
cool to room temperature and evaporated to dryness. The
resulting raw material derivative was then purified by
recrystallization from toluene. If required multiple recrystalliza-
tion steps were executed. (B) Racemic mandelic acid derivatives
were obtained from a nonselective chemical reaction of the
corresponding benzaldehyde derivative with hydrogen cyanide.
For this purpose a biphasic system (50% (v/v)) consisting of
0.05 mol·kg−1 citrate-phosphate buffer pH 6 and the above-
mentioned hydrogen cyanide solution was allowed to react at
room temperature overnight. The resulting mixture was
subjected to the identical purification protocol as described
for the synthesis of enantiopure mandelic acid derivatives.
Chemical identity as well as purity of the materials was
determined via XRPD, NMR, DSC, and (chiral) HPLC
analyses. A final mole fraction purity of > 0.995 was found
An exemplary chromatogram, retention times, and separation
factors for all studied mandelic acid derivatives are reported in
the SI, Figure SI-18 and Table SI-1. Regarding the shape of the
chromatographic elution profiles of the racemic mixtures (see
C
DOI: 10.1021/je500845d
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
D
DOI: 10.1021/je500845d
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
the SI, Figure SI-18), there is evidence that the adsorption
behavior of the enantiomers is based on different adsorption
isotherms, which is remarkable for chiral compounds. It appears
that the (R)-enantiomers exhibit a favorable Langmuir-isotherm
for all derivatives tested. In contrary, a linear or even an
unfavorable anti-Langmuir-isotherm can be observed for the
(S)-enantiomers.
2.7. Reproducibility Analysis. The reproducibility of the
chosen DSC technique was determined on the basis of its
standard deviation by a series of measurements of different
samples of racemic 2-chloromandelic acid (racemic com-
pound). The standard deviation, s.d., was calculated as shown
in eq 1.
derivatives are not comparable, 408.9 (4), 393.2 (7), 391.2
(10), 415.3 (13), and 416.9 (16) K with the corresponding
melting enthalpies of 30 (4), 16.2 (7), 17.8 (10), 30.4 (13),
and 25.6 (16) kJ·mol−1. Noteworthy, all 4-substituted
derivatives always exhibit the highest melting temperature
both for the enantiomer and racemic mixture.
Also all fluoro-substituted mandelic acid derivatives tend to
be excluded from similarities, since probably other strong
hydrogen bonds are formed in solid state in comparison to
analogous chloro-substituted mandelic acid derivatives.27,28
Unfortunately enantiopure iodine-substituted mandelic acid
derivatives were not obtainable by synthesis, mostly due to size
limitations by the iodine-substituent within the active site of the
biocatalyst. Methyl-substituted mandelic acid derivatives (14,
15, and 16) were included in this study to compare the
substitution-effect without the introduction of polar halogen-
substituents.
3.2. XRPD-Phase Analysis Results. Analysis of the
corresponding XRPD measurements reveals that at least a
few similarities within the investigated mandelic acid derivatives
library were found. The most prominent example is the
occurrence of at least three very similar metastable conglom-
erates, as mentioned above. The metastable conglomerates of
2-chloromandelic acid 5, 2-bromomandelic acid 8, and 2-
chloro-4-fluoromandelic acid 17 exhibit very similar X-ray
diffraction results (Figure 3).
n
i=1
∑
(xi − x)2
n − 1
̅
s. d. =
(1)
xi is the observed value; x is the mean value; n is the number
̅
of samples
With a total number of 15 measurements (n = 15) for
melting temperatures a standard deviation of
0.45 K was
found. The melting enthalpies show a standard deviation of
0.6 kJ·mol−1 (n = 13). These values need to be considered
throughout this study as a measure for accuracy of all reported
values.
3. RESULTS AND DISCUSSION
3.1. Melting Temperatures and Melting Enthalpies of
Mandelic Acid Derivatives. Table 2 summarizes the main
results of DSC measurements of mandelic acid and its
corresponding mandelic acid derivatives with a standard
uncertainty u of u(Tfus) = 0.45 K and u(ΔHfus) = 0.6 kJ·
mol−1. The reported values represent measurements of
enantiopure and racemic mandelic acid derivatives and include
information about the presence of metastable conglomerates
and further observed polymorphism. Noteworthy, a recent
study indicates also the presence of other currently unknown
polymorphs of racemic mandelic acid derivatives, which is in
accordance with the already reported polymorphism behavior
of 2-chloro- and 3-chloromandelic acid 5, 6.26 Results for
entries 7−18 were obtained in this study. The other entries
comprehend literature data as comparison and completeness
for mandelic acid 1, all three monofluoro derivatives 2−4 and
2- and 3-chloromandelic acid 5, 6.
As shown in Table 2, all thermodynamic stable and a few
metastable phases of mandelic acid and all its derivatives were
characterized as racemic compounds. Another interesting
behavior is the occurrence of metastable conglomerates, e.g.,
2-chloromandelic acid 5, which was recently characterized in
detail.17,19 In this study evidence for at least two other
metastable conglomerates was found for 2-bromomandelic acid
8 and 2-chloro-4-fluoromandelic acid 17 (see also below). Such
a behavior was expected since strong structural and chemical
similarities to 2-chloromandelic acid are present. For these
substances, also similar melting temperatures and enthalpies
Figure 3. XRPD patterns of metastable conglomerates of 2-substituted
mandelic acid derivatives; (S)-5, (S)-2-chloromandelic acid; rac-5,
racemic 2-chloromandelic acid; (S)-8, (S)-2-bromomandelic acid; rac-
8, racemic 2-bromomandelic acid; (S)-17, (S)-2-chloro-4-fluoroman-
delic acid; rac-17, racemic 2-chloro-4-fluoromandelic acid; I, intensity
(arbitrary units); 2Θ, diffraction angle.
Slight deviations occur mainly only at the dominant
reflections at 6.7, 13.4, 19.9, and 26.6° (values for derivative
5), which also explain the very similar values of melting
temperature and enthalpy for 5 and 8. Both, an increase of the
size of the substituent (2-chloro to 2-bromo) and the
introduction of a second substituent (2-chloro to 2-chloro-4-
fluoro) leads to shifts to smaller 2Theta values due to widening
of the crystal lattice. The shown metastable conglomerates of 8
and 17 are unfortunately not directly obtainable from solution.
For these derivatives seeding with a few crystals of metastable
conglomerate 5 was required to induce the formation of the
respective metastable conglomerate. In Figure 3 traces of the
respective seeds of metastable 5 are still found in the XRPD
were found, e.g., enantiopure 2-chloromandelic acid 5 (Tfus
=
392.4 K and ΔHfus = 24.6 kJ·mol−1) and enantiopure 2-
bromomandelic acid 8 (Tfus = 392.5 K and ΔHfus = 22.4 kJ·
mol−1). However, a clear tendency of the melting points or
melting enthalpy between all similarly substituted mandelic acid
derivatives was not observed, as shown for all racemic 3-
substituted derivatives (3, 6, 9, and 12). In addition, the
melting points of all five racemic 4-substituted mandelic acid
E
DOI: 10.1021/je500845d
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
pattern of metastable 17, as highlighted herein. The metastable
conglomerate of 8 is relatively stable at ambient conditions,
which allowed its investigation of the respective physical-
chemical properties, but metastable racemic 17 was only
sporadically obtained hereby. This indicates a significant lower
stability of this solid phase. Presumable the presence of the
additional fluoro-substituent at 4-position destabilizes the
metastable conglomerate in favor of its stable racemic
compound counterpart. In addition, the stable racemic
compound of 17 shows the highest melting point (370.7 K;
Table 2) compared to 5 and 8. Furthermore, the presence of
various other metastable conglomerates within this group of 2-
substituted mandelic acid derivatives can be assumed.
In comparison, three similar 4-substituted derivatives (4-
chloromandelic acid 7, 4-bromomandelic acid 10 and 4-
iodomandelic acid 13) show hardly any similarities within
XRPD-experiments (Figure 4). This indicates clearly that even
for such highly similar substances various different crystal
structures may occur.
Figure 4. XRPD-results of racemic 4-chlorosubstituted mandelic acid
derivatives; rac-13, racemic 4-iodomandelic acid; rac-10, racemic 4-
bromomandelic acid; rac-7, 4-chloromandelic acid; I, intensity
(arbitrary units); 2Θ, diffraction angle.
A similar behavior was found for the XRPD patterns of
racemic 2,3 and 4-iodomandelic acid (11, 12, 13) derivatives
(shown in the SI, Figure SI-7). Also considerably different
melting temperatures ((382.1/370.1/415.3) K) and enthalpies
of melting ((25/24.6/30.4) kJ·mol−1) were observed. Herein
differences in the crystal lattices, very likely due to diverse
hydrogen-bonding patterns, are present, as also reported within
the 2-, 3- and 4-fluoro-substituted mandelic acids (2, 3, and
4).27,28 Further results of additional XRPD-measurements, both
for the racemic and enantiopure composition of all investigated
derivatives, are given in the SI.
3.3. Selected Binary Phase Diagrams. Strong differences
within the investigated derivative library were also visualized by
the determination of the respective binary melting phase
diagram (as derived from DSC-analysis). As shown in Figure
5A and B, even small differences in halogen-substitution result
in a significant different solid/liquid phase behavior. 2-
Bromomandelic acid 8 exhibits a eutectic composition close
to the racemic mixture at a mole fraction of about x = 0.54,
while 3-bromomandelic acid 9 has a considerable higher
eutectic composition of about x = 0.9. Noteworthy, the binary
Figure 5. Exemplary binary melting point phase diagrams of (A) 2-
bromomandelic acid 8 (racemic compound with a metastable
conglomerate, triangle), (B) 3-bromomandelic acid 9 (racemic
compound), and (C) 3-methylmandelic acid 15 (racemic compound).
Detailed experimental data is available in the SI (Tables SI-2, -3, and
-4). Panels A−C only show half of the full binary phase diagram, with
the racemic composition on the left and the pure enantiomer (all (S)
here) on the right side. The shown liquidus curves (dashed lines) are
based on theoretical calculations using eqs 2 and 3 (see text). The
eutectic temperatures are shown as straight lines.
phase diagrams of 8 and 9 show significant similarities to the
binary phase diagrams of 2- and 3-chloromandelic acid 5, 6,
which were recently investigated in detail.19,20
F
DOI: 10.1021/je500845d
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
Funding
This comparison also includes the presence of an additional
metastable conglomerate of 8, indicated in Figure 5A by the
triangle at the racemic composition below the circle
representing the melting temperature of the racemic com-
pound. Furthermore, an exchange of the substitution itself may
result in various other solid phases. For example, strong
differences were found between the structural very similar
derivatives 3-bromo- and 3-methylmandelic acid 9 and 15,
Figure 5B and C. The binary phase diagram of 15 shows, in
comparison to 9, a relatively low eutectic composition around
0.62 (versus 0.9). In accordance to the results of other 3-
halogen-substituted mandelic acid derivatives, a metastable
conglomerate was not observed for 15 in this study. In addition,
the shown phase diagrams partly exhibit a depression of the
measured solidus temperatures, which may indicate a partial
solid solution behavior (solvus line), e.g., wider deviations of
the solidus temperatures close to the pure enantiomer side
(Figure 5A and C) and the racemate side (Figure 5B). The
eutectic compositions were calculated from the intersection of
the liquidus lines of the enantiomer and the racemate according
Financial support by the European Union (EU-Project
“INTENANT, Integrated Synthesis and Purification of
Enantiomers”; NMP2-SL-2008-214129) and the German
Federal Ministry of Education and Research (BMBF,
Bundesministerium fur Bildung and Forschung, Project No.
031A123) is highly acknowledged.
̈
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors want to thank Jacqueline Kaufmann, Luise
Borchert, and Nora Doering for their help in the experimental
work and Dr. Liane Hilfert for the help with NMR-analysis.
ABBREVIATIONS
■
HPLC, high performance liquid chromatography; DSC,
differential scanning calorimetry; Tfus, melting temperature;
ΔHfus, melting enthalpy; XRPD, X-ray powder diffraction; x,
mole fraction
to simplified equations of Schroder−van Laar and Prigogine−
̈
Defay (eqs 2 and 3) with x, mole fraction; R, gas constant;
REFERENCES
Tfus,E, Tfus,R, melting temperature of enantiomer and racemic
■
compound; ΔHfus,E, ΔHfus,R, enthalpy of fusion of enantiomer
(1) Jacques, J.; Collet, A.; Wilen, S. H. Enantiomers, Racemates and
Resolutions; Krieger Publishing Company: Malabar, FL, 1981.
(2) Franks, M. E.; Macpherson, G. R.; Figg, W. D. Lancet 2004, 363,
1802−1811.
and racemic compound.1
⎛
⎝
⎞
⎠
ΔHfus,E
1
1
⎜
⎜
⎟
⎟
ln x =
−
(3) Collins, A. N.; Sheldrake, G. N.; Crosby, J. Chirality in Industry:
The Commercial Manufacture and Applications of Optically Active
Compounds; John Wiley & Sons Ltd.: Hoboken, NJ, 1992.
(4) Subramanian, G. N.; Collet, A.; N, C. B. Chiral Separation
Techniques, 2nd ed.; Wiley VCH: Weinheim, Germany, 2001.
(5) Toda, F. Enantiomer Separation, 1st ed.; Springer: Heidelberg,
Germany, 2004.
(6) Lorenz, H.; Perlberg, A.; Sapoundjiev, D.; Elsner, M. P.; Seidel-
Morgenstern, A. Chem. Eng. Process.: Process Intensificat. 2006, 45,
863−873.
(7) Coquerel, G. Chem. Soc. Rev. 2014, 43, 2286−2300.
(8) Lorenz, H.; Seidel-Morgenstern, A. Angew. Chem. - Int. Ed. 2014,
53, 1218−1250.
R
T
T
fus,E
fus
(2)
(3)
⎛
⎝
⎞
⎠
2ΔHfus,R
1
1
T
fus
⎜
⎜
⎟
⎟
ln 4x(1 − x) =
−
R
T
fus,R
The shown calculated liquidus lines describe the exper-
imentally obtained liquidus temperatures sufficiently although
the used equations principally only apply to ideal systems. We
like to emphasize that liquidus and solidus lines are mainly
presented here to qualitatively evaluate the solid/liquid phase
behavior in the investigated chiral system.
(9) Gedicke, K.; Kaspereit, M.; Beckmann, W.; Budde, U.; Lorenz,
H.; Seidel-Morgenstern, A. Chem. Eng. Res. Des. 2007, 85, 928−936.
(10) von Langermann, J.; Tam, L. M.; Lorenz, H.; Seidel-
Morgenstern, A. Chem. Ing. Tech. 2010, 82, 93−100.
4. CONCLUSIONS
The solid phase behavior of halogen- and methyl-substituted
derivatives of mandelic acid was investigated. The results show
clearly that the entire group of substances exhibit racemic
compound behavior, while only a limited number of similarities
between comparably substituted mandelic acid derivatives were
found and described in this study. A noteworthy exclusion
hereof is the presence of metastable conglomerates, which were
found merely for 2-halogen-substituted derivatives.
(11) Temmel, E.; Wloch, S.; Muller, U.; Grawe, D.; Eilers, R.;
̈
Lorenz, H.; Seidel-Morgenstern, A. Chem. Eng. Sci. 2013, 104, 662−
673.
(12) Binev, D.; Wloch, S.; Lorenz, H.; Seidel-Morgenstern, A. 1st Int.
Workshop Ind. Pharm. Crystallisat. (HIW 2012) 2012, 127−135.
(13) Coppola, G. M.; Schuster, H. F. a-Hydroxy Acids in
Enantioselective Synthesis; John Wiley & Sons: Hoboken, NJ, 1997.
(14) Groger, H. Adv. Synth. Catal. 2001, 343, 547−558.
̈
(15) Davey, R. J.; Sadiq, G.; Back, K.; Wilkinson, L.; Seaton, C. C.
Chem. Commun. 2012, 48, 1976−1978.
ASSOCIATED CONTENT
* Supporting Information
■
S
(16) Gilks, S. E.; Davey, R. J.; Mughal, R. K.; Sadiq, G.; Black, L.
Cryst. Growth Des. 2013, 13, 4323−4329.
A summary of XRPD-results for all investigated mandelic acid
derivatives, exemplary DSC curves, NMR and HPLC results are
provided. This material is available free of charge via the
Internet at http://pubs.acs.org.
(17) He, Q.; Rohani, S.; Zhu, J.; Gomaa, H. Cryst. Growth Des. 2010,
10, 5136−5145.
(18) He, Q.; Zhu, J.; Gomaa, H.; Jennings, M.; Rohani, S. J. Pharm.
Sci. 2009, 98, 1835−1844.
(19) Lorenz, H.; Von Langermann, J.; Sadiq, G.; Seaton, C. C.;
Davey, R. J.; Seidel-Morgenstern, A. Cryst. Growth Des. 2011, 11,
1549−1556.
AUTHOR INFORMATION
Corresponding Authors
■
*(Synthesis related) Phone: +49-3814986456. Fax: +49-
3814986452. E-mail: jan.langermann@uni-rostock.de.
*(Solid phase behavior related) Phone: +49-3916110293. Fax:
+49-3916110524. E-mail: lorenz@mpi-magdeburg.mpg.de.
(20) Le Minh, T.; Von Langermann, J.; Lorenz, H.; Seidel-
Morgenstern, A. J. Pharm. Sci. 2010, 99, 4084−4095.
(21) Zhang, Y.; Ray, A.; Rohani, S. Chem. Eng. Sci. 2009, 64, 192−
197.
G
DOI: 10.1021/je500845d
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
(22) Coles, S. J.; Threlfall, T. L.; Tizzard, G. J. Cryst. Growth Des.
2014, 14, 1623−1628.
(23) von Langermann, J.; Wapenhensch, S. Adv. Synth. Catal. 2014,
356 (14-15), 2989−2997.
(24) Lorenz, H.; Sapoundjiev, D.; Seidel-Morgenstern, A. J. Chem.
Eng. Data 2002, 47, 1280−1284.
(25) Lorenz, H. In Crystallization: Basic Concepts and Industrial
Applications; Beckmann, W., Ed.; Wiley-VCH: Weinheim, Germany,
2013; pp 33−74.
(26) Coles, S. J.; Ellis, A. L.; Leung, K.; Sarson, J.; Threlfall, T. L.;
Tizzard, G. J. CrystEngComm 2014, DOI: 10.1039/c4ce01832j.
(27) Larsen, S.; Marthi, K. Acta Crystallogr. Sect. B 1994, 50, 373−
381.
(28) Larsen, S.; Marthi, K. Acta Crystallogr. Sect. B 1997, 53, 280−
292.
(29) Lorenz, H.; Seidel-Morgenstern, A. Thermochim. Acta 2004,
415, 55−61.
(30) Kuhnert-Brandstatter, M.; Ulmer, R. Mikrochim. Acta 1974, 62,
̈
927−935.
(31) Rose, H. A. Anal. Chem. 1952, 24, 1680−1681.
(32) Collet, A.; Jacques, J. N. Bull. Soc. Chim. France 1973, 12, 3330−
3334.
H
DOI: 10.1021/je500845d
J. Chem. Eng. Data XXXX, XXX, XXX−XXX