Improving the poor aqueous solubility of nutraceutical compound pterostilbene through cocrystal formation

Article abstract of DOI:10.1021/cg1016092

Pterostilbene is a nutraceutical compound being investigated for its ability to form cocrystals with pharmaceutically acceptable coformers to improve its poor aqueous solubility. Cocrystals of a 2:1 and 1:1 stoichiometric molar ratio of pterostilbene with piperazine or glutaric acid were synthesized on a multigram scale and fully characterized including single-crystal X-ray diffraction. Furthermore, both cocrystals were evaluated for physical stability with respect to humidity and temperature as well as kinetic solubility. Each cocrystal displayed remarkable stability, showing no evidence of dissociation across a range of durations, temperatures, and relative humidities. The aqueous concentration of pterostilbene measured over five hours from dissolution of the pterostilbene-piperazine cocrystal was six times higher than the solubility of the single-component pterostilbene. Although X-ray powder diffraction results indicate partial transformation of the pterostilbene-piperazine cocrystal to pterostilbene after 5 h, this increase in concentration was sustained throughout the dissolution experiment. Within 5 min of slurrying the pterostilbene- glutaric acid cocrystal in water, precipitation of pterostilbene was observed; thus no dissolution data under these conditions were obtained.

Full text of DOI:10.1021/cg1016092

ARTICLE
pubs.acs.org/crystal
Improving the Poor Aqueous Solubility of Nutraceutical Compound
Pterostilbene through Cocrystal Formation
Sarah J. Bethune,* Nate Schultheiss,† and Jan-Olav Henck
SSCI, a division of Aptuit, 3065 Kent Avenue, West Lafayette, Indiana, United States
S Supporting Information
b
ABSTRACT: Pterostilbene is a nutraceutical compound being investigated for its ability
to form cocrystals with pharmaceutically acceptable coformers to improve its poor
aqueous solubility. Cocrystals of a 2:1 and 1:1 stoichiometric molar ratio of pterostilbene
with piperazine or glutaric acid were synthesized on a multigram scale and fully
characterized including single-crystal X-ray diffraction. Furthermore, both cocrystals
were evaluated for physical stability with respect to humidity and temperature as well as
kinetic solubility. Each cocrystal displayed remarkable stability, showing no evidence of
dissociation across a range of durations, temperatures, and relative humidities. The
aqueous concentration of pterostilbene measured over five hours from dissolution of the
pterostilbene-piperazine cocrystal was six times higher than the solubility of the single-
component pterostilbene. Although X-ray powder diffraction results indicate partial
transformation of the pterostilbene-piperazine cocrystal to pterostilbene after 5 h, this
increase in concentration was sustained throughout the dissolution experiment. Within
5 min of slurrying the pterostilbene-glutaric acid cocrystal in water, precipitation of
pterostilbene was observed; thus no dissolution data under these conditions were obtained.
’ INTRODUCTION
individual components. Cocrystal dissolution that results in
undersaturated concentrations of individual components leads
to solution stability. Cocrystal dissolution that leads to super-
saturated concentrations of one or both components may result
in precipitation of the component(s), and therefore the cocrystal
is not stable in solution. Recent studies on a homologous series of
carbamazepine cocrystals have been published relating aqueous
stability and kinetic solubility.8 Coformers with high aqueous
solubility, such as glutaric, malonic, and glycolic acids, resulted in
cocrystals that were unstable in water; as the cocrystal dissolved,
the concentration of carbamazepine in solution increased beyond
the solubility of carbamazepine dihydrate and thus precipitated.
Coformers with lower aqueous solubility, such as fumaric and
1-hydroxy-2-naphthoic acid, often resulted in stable cocrystals in
water, and no precipitation was observed.
In a recent study, we demonstrated the ability of pterostilbene
to form cocrystals with the APIs carbamazepine and caffeine and
the aqueous solubility of each cocrystal was measured. The
caffeine cocrystal resulted in approximately 27-fold increase
in concentration (kinetic, sustained for at least 5 h) in compar-
ison to pterostilbene.9 Within this study, we opted to screen
pterostilbene with pharmaceutically acceptable coformers that
possess relatively high aqueous solubilities10 with the intentions
of using cocrystals to increase the poor solubility profile of
pterostilbene.
Pterostilbene, a naturally occurring analog of resveratrol, is
classified asa nutraceutical compound1 withits greatest abundance
found in a variety of different berries.2 In a number of animal
studies, pterostilbene was found to lower cholesterol, regulate
blood sugar levels, and be as effective as resveratrol as an anti-
carcinogenic agent.3 Furthermore, initial studies on comparative
pharmacokinetics of pterostilbene and resveratrol resulted in the
former showing enhanced lipophilicity and membrane perme-
ability and thus greater bioavailability over the latter.4
Although pterostilbene has shown some promising initial
pharmacokinetic studies, a potential drawback is its extremely
low aqueous solubility, approximately 21 μg/mL.5 A common
approach to increasing the solubility of a poorly soluble com-
pound, especially in the pharmaceutical and agrochemical in-
dustry, is through the formation of a crystalline salt.
Unfortunately pterostilbene is not chemically amenable for such
a modification. Thus, another potential option is the formation of
a cocrystal (or multicomponent crystal) where the individual
components are connected by noncovalent interactions, typically
hydrogen bonds.6 One benefit in forming cocrystals is that it
allows for changes to be made to the crystal structure of the
parent molecule without introducing chemical modifications,
which could change the bioactivity profile of the active pharma-
ceutical ingredient (API). Cocrystals have been shown to be
effective in altering the physicochemical properties of APIs,
particularly aqueous solubility and dissolution rates.7
Received: December 3, 2010
As a cocrystal dissolves, the concentration in solution may be
supersaturated, saturated, or undersaturated with respect to the
Revised:
Published: May 17, 2011
April 15, 2011
r
2011 American Chemical Society
2817
dx.doi.org/10.1021/cg1016092 Cryst. Growth Des. 2011, 11, 2817–2823
|

Crystal Growth & Design
ARTICLE
Scheme 1. Molecular Structures of Pterostilbene, Piperazine,
and Glutaric Acid (Left to Right)
In this paper, we describe the synthesis and characterization of
two pterostilbene cocrystals with coformers piperazine and
glutaric acid, Scheme 1. Piperazine is a pharmaceutically accepted
basic salt former but also has pharmacological activity as an
anthelmintic.11 Although glutaric acid is not listed as pharma-
ceutically acceptable according to Stahl and Wermuth,11 the
toxicity profile of glutaric acid falls within the range of several
pharmaceutically acceptable compounds that are included. Both
cocrystals were generated on a multigram scale and characterized
by X-ray powder and single-crystal diffraction, differential scan-
ning calorimetry (DSC), thermogravimetric analysis (TGA), and
1H NMR. To evaluate the physical stability, accelerated tests
were conducted over a number of temperatures, humidities, and
durations. Additionally, powder dissolution was studied for both
cocrystals to evalute their ability to increase the poor aqueous
solubility of the nutraceutical compound pterostilbene.
Figure 1. X-ray powder patterns of pterostilbene form I, piperazine (as
received), glutaric acid (as received), 1, and 2 (top to bottom).15
General Methods. High-Throughput Screening (HTS). Specified
amounts of filtered 0.1 M pterostilbene solutions were dispensed into
the wells of a 96-well microplate using the Symyx CORE(X). Stoichio-
metric amounts of several pharmaceutically acceptable or GRAS-
listed carboxylic acid coformer stock solutions were added to the wells
(∼1À2 mg per well). Four different experiment types were conducted:
slow evaporation, fast evaporation, sonication followed by slow evapora-
tion, and sonication followed by fast evaporation. Upon completion,
each well was analyzed by XRPD.
Physical Stability. Physical stability was evaluated at approximately
40 °C in ambient RH, 60 °C in ambient RH, ambient temperature in
75% RH, ambient temperature in 98% RH, and 40 °C in 75% RH. XRPD
was used to detect dissociation. Vials of each cocrystal 1 and 2 were
subjected to each condition for durations of 2 weeks, 1 month, and 2
months. Upon completion of the duration allowed, the samples were
immediately analyzed by XRPD.
’ EXPERIMENTAL METHODS
Reagents. Pterostilbene was acquired from Aptuit Laurus Ltd.,
India. Piperazine and glutaric acid were purchased from Sigma-Aldrich
and used as received. All other chemicals were purchased from various
suppliers and used without further purification.
Pterostilbene-Piperazine, 1. Cocrystal 1 was prepared by liquid-
assisted grinding (LAG), solvent evaporation, or slow cooling. Under
LAG conditions, a 2:1 mixture of pterostilbene (∼65 mg, ∼0.25 mmol)
and piperazine (∼11 mg, ∼0.13 mmol) was added to an agate mill.
Approximately 10 μL of solvent (ethanol or nitromethane) was added,
and the material was ground for 20 min at a rate of 30 Hz. Single crystals
were grown by slow evaporation of a 1:1 mixture of pterostilbene
(135.2 mg, 0.53 mmol) and piperazine (45.5 mg, 0.53 mmol) in ethanol
(2 mL). After 1 day, rod-shaped crystals were harvested. The material
was scaled up by dissolution of pterostilbene (5.12 g, 20.0 mmol) and
piperazine (862 mg, 10.0 mmol) in ethanol (∼70 mL) with heat. The
homogeneous solution was stirred in an oil bath for approximately 1 h, after
which time the heat was removed and the solids were precipitated. The
white crystalline solid was filtered and dried, yielding 10.54 g, 88%, of 1.
Pterostilbene-Glutaric Acid, 2. Cocrystal 2 was also prepared
by LAG, solvent evaporation, or slow cooling. Under LAG conditions,
a 1:1 mixture of pterostilbene (∼36 mg, ∼0.14 mmol) and glutaric acid
(∼19 mg, ∼0.14 mmol) was added to an agate mill. Approximately
10 μL of solvent (toluene or 2-propanol) was added, and the material
was ground for 20 min at a rate of 30 Hz. Single crystals were grown by
slow evaporation of a 1:1 mixture of pterostilbene (25.0 mg, 0.10 mmol)
and glutaric acid (13.1 mg, 0.10 mmol) in toluene (3 mL). After 1 day,
plate-shaped crystals were harvested. The material was scaled up
by dissolution of pterostilbene (3.03 g, 11.8 mmol) and glutaric acid
(1.55 g, 11.7 mmol) in toluene (∼40 mL) with heat. The homo-
geneous solution was stirred, and upon cooling the solids precipitated.
The whitecrystalline solid was filteredand dried, yielding 3.79 g, 83%, of2.
Powder Dissolution Experiments. Concentration measurements
were performed using ultraviolet (UV) spectroscopy on a Spectramax
Microplate Reader. For pterostilbene, a standard curve was produced by
serial dilutions; absorbance readings at 315 nm for pterostilbene were
used to establish a linear regression. The small amount of methanol used
to prepare pterostilbene standards did not cause shifting in the absor-
bance spectrum. The UV spectra for the cocrystal formers, glutaric acid
and piperazine, do not overlap at 315 nm. Powder dissolution of
cocrystal 1 was carried out such that an excess of solid material was
slurried in water at ambient conditions, and aliquots were taken at
specific time points to derive a concentration versus time profile to
estimate the maximum concentration before transformation to pteros-
tilbene occurred. Aliquots were centrifuged, supernatant was extracted,
and appropriate dilutions were made to maintain absorbance readings
within the standard curve. Absorbance measurements were taken at
315 nm for pterostilbene, and concentrations were calculated from the
standard curve. All experiments were repeated three times to evaluate
the standard deviation. Particle size was not controlled for the powder
dissolution experiments because the maximum concentration was sus-
tained for several hours, and an initial dissolution rate, a property that
would be directly affected by particle size, was not calculated from the
data collected.
Powder dissolution of cocrystal 2 was attempted under the same
experimental conditions as cocrystal 1. However, XRPD results indicate
the presenceofcrystallinepterostilbene after 5 min whenthe cocrystal was
slurried in water at ambient conditions. Intrinsic dissolution in 900 mL of
2818
dx.doi.org/10.1021/cg1016092 |Cryst. Growth Des. 2011, 11, 2817–2823

Crystal Growth & Design
ARTICLE
Table 1. Crystal Data for Cocrystals 1 and 2
Table 2. Hydrogen-Bond Geometries for Cocrystals 1 and 2
d(H A), d(D A), θ(DHA),
cocrystal
empirical formula
1
2
3 3 3
3 3 3
C36H42N2O6
C21H24O7
cocrystal
DÀH
A
Å
Å
deg
3 3 3
MW
598.72
orthorhombic
P212121, 4
5.2586(7)
11.7922(14)
51.155(7)
90
388.40
monoclinic
P21/c, 4
7.2644(3)
32.8801(16)
7.9819(4)
90
1
O211ÀH211 N11
O212ÀH212 N14
1.87(4)
1.82(5)
2.711(5)
2.746(5)
3.224(5)
3.338(5)
160(4)
166(4)
168(4)
173(5)
3 3 3
crystal syst
space group, Z
a, Å
3 3 3
N11ÀH11 O211#1 2.35(4)
N14ÀH14 O212#2 2.60(5)
3 3 3
3 3 3
b, Å
2
O11ÀH11 O21
1.821(17)
1.792(17)
1.850(19)
2.6798(12) 170.9(16)
2.6463(12) 173.5(16)
2.7001(12) 173.7(16)
3 3 3
c, Å
O15ÀH15 O16
3 3 3
R, deg
O21ÀH21 O12
3 3 3
β, deg
90
96.000(2)
90
γ, deg
90
vol, Å3
density, g/cm3
3172.2(7)
1.254
1896.07(15)
1.361
temp, K
296
120(2)
0.71073
0.102
X-ray wavelength
μ, mmÀ1
F(000)
0.71073
0.085
1280
724
Θmin, deg
Θmax, deg
reflns
1.77
1.24
27.57
32.58
collected
independent
observed
threshold expression
R1 (observed)
wR2 (all)
37364
4250
30177
6677
1843
5136
>2σ(I)
0.0545
0.1255
>2σ(I)
0.0472
0.1396
water under ambient conditions was also attempted. However, XRPD
results indicate precipitation of crystalline pterostilbene on the surface of
the pellet after 30 min. Therefore, a concentration vs time profile and
intrinsic dissolution rate value were unobtainable under these conditions.
Analytical Characterization. X-ray Powder Diffraction (XRPD).
Patterns were collected using a PANalytical X’Pert Pro or Inel XRG-3000
diffractometer. An incident beam of Cu KR radiation was produced using
an Optix long, fine-focus source. PANalytical data were collected and
analyzed using X’Pert Pro Data Collector software (v. 2.2b). Prior to the
analysis a silicon specimen (NIST standard reference material 640c) was
analyzed to verify the accuracy of the diffractometer optics using the
known silicon 111 peak position at 28.441 °2θ to within (0.01°.
PANalytical diffraction patterns were collected using a scanning posi-
tion-sensitive detector (X’Celerator) located 240 mm from the specimen.
Single-Crystal X-ray Diffraction (SCXRD). Data sets were collected on
a Bruker SMART APEX II diffractometer using Mo KR radiation. Data
were collected using APEXII software.12 Initial cell constants were found
by small widely separated “matrix” runs. Data collection strategies were
determined using COSMO. Scan speed and scan width were chosen
based on scattering power and peak rocking curves. Temperature
control for cocrystal 2 was provided with an Oxford Cryostream low-
temperature device.
Figure 2. Labeled thermal ellipsoids plot (displayed at a 50% prob-
ability level) of the 2:1 pterostilbene-piperazine cocrystal 1.
Cocrystal 1. The two crystallographically nonequivalent pterostil-
benes were distinguished with use of the SHELXL “RESI” command.
The compound crystallizes in the noncentrosymmetric space group
P212121. Due to the absence of heavy atom anomalous scatterers,
determination of crystal handedness was not pursued, and Friedel
opposites were merged. Coordinates for the amine protons H11 and
H14 and phenol protons H21 (on each pterostilbene) were allowed to
refine. An extinction correction was applied; the EXTI parameter refined
to a small but nonzero number.
Cocrystal 2. Coordinates for the carboxylic acid protons H11 and
H15 and the phenol proton H21 were allowed to refine.
’ RESULTS AND DISCUSSION
Cocrystals 1 and 2 were synthesized through a variety of
different reaction conditions: liquid-assisted grinding (LAG),
slow evaporation, and slow cooling. Evidence of cocrystal 2 was
first observed through HTS experiments where a unique XRPD
pattern was detected in wells containing glutaric acid. Initial
characterization by XRPD of cocrystals 1 and 2 displayed unique
patterns in comparison to pterostilbene, piperazine, or glutaric
acid, Figure 1. Proton NMR revealed the 2:1 and 1:1 pterostil-
bene/coformer stoichiometries for cocrystals 1 and 2, respec-
tively, while the TGA thermograms showed negligible weight
loss until decomposition; thus the cocrystals were characterized
as nonsolvated/hydrated.
Unit cell constants and orientation matrix were improved by least-
squares refinement of reflections thresholded from the entire data set.
Integration was performed with SAINT,13 using this improved unit cell
as a starting point. Precise unit cell constants were calculated in SAINT
from the final merged data set. Lorenz and polarization corrections were
applied. Where indicated, absorption corrections were made using the
multiscan procedure in SADABS.
Data were reduced with SHELXTL.14 The structures were solved in
all cases by direct methods without incident. Absorption correction was
not applied (maximum μd ≈ 0.03 in both cases).
2819
dx.doi.org/10.1021/cg1016092 |Cryst. Growth Des. 2011, 11, 2817–2823

Crystal Growth & Design
ARTICLE
Onecarboxylicacidmoietyofglutaric acid forms an OÀH
O
3 3 3
hydrogen bond to the hydroxyl oxygen of pterostilbene, while the
second acid moiety forms an acidÀacid dimer with a neighboring
pterostilbeneÀcarboxylic acid pair, generating a four-component
supramolecular assembly, Figure 5. Extension of the structure
reveals an additional series of OÀH O (OÀH Ocarbonyl
3 3 3
3 3 3
and OÀH Ohydroxyl) hydrogen bonds linking the individual
3 3 3
assemblies together. Overall, two primary hydrogen-bonding
synthons are located within the structure and can easily be
described through graph set notations:20 the conventional
R22(8) acidÀacid dimer and a R44(12) hydroxylÀcarbonyl/
hydroxylÀhydroxyl motif. A search of the CSD revealed that
the R22(8) graph set (acidÀacid dimer) is common place (2991
hits); however, the R44(12) graph set (hydroxylÀcarbonyl/
hydroxylÀhydroxyl motif) occurs less frequently (107 hits).
Two pharmaceutical properties of cocrystals 1 and 2 were
evaluated and measured: physical stability with respect to tem-
perature and relative humidity (RH) and aqueous dissolution.
Stability testing is an integral part of the drug development
process. Early in development, stability testing (normal and
accelerated) on new chemical entities is required to help select
the most satisfactory form with the desired pharmaceutical
profile. Stability studies as a function of time versus humidity
and temperature are particularly useful in assessing the shelf life,
expiry date, storage conditions, and packaging required for an
API and drug product. Thus, both cocrystals 1 and 2 were
subjected to a range of different humidities and temperatures
for 2 weeks, 1 month, and 2 months. At each time point, the
materials were analyzed by XRPD to determine whether the
cocrystals displayed physical instability in the presence of
increased temperature or humidity by comparison to the XRPD
pattern of the unstressed cocrystal. No additional analysis was
conducted to assess the chemical stability of either cocrystal 1
or 2. Both cocrystals displayed remarkable physical stability,
with no dissociation or recrystallization based on no observ-
able differences in the XRPD patterns in comparison to
the unstressed materials after 2 weeks, 1 month, and 2 months
of accelerated stress under the following conditions:
40 °C/ambient RH, 60 °C/ambient RH, ambient tempera-
ture/75% RH, ambient temperature/98% RH, and 40 °C/75%
RH.
Figure 3. Extended architecture displaying the hydrogen bond donor/
acceptor ability of piperazine in cocrystal 1.16 The nitrogen atoms on
piperazine are highlighted as blue spheres while the hydrogen bonds are
yellow dashed lines.
Figure 4. Labeled thermal ellipsoids plot (displayed at a 50% prob-
ability level) of the 1:1 pterostilbene-glutaric acid cocrystal 2.
Single crystals of cocrystals 1 and 2 were generated by slow
evaporation techniques from ethanol and toluene, respectively.
After 1 day, crystals suitable for single-crystal X-ray diffraction
were culled from the mother liquors.
Both piperazine and glutaric acid are freely soluble in aqueous
media with estimates of approximately 150 and 850 mg/mL,
respectively.21 Thus, due to their high water solubility, we
anticipated that the aqueous dissolution of the pterostilbene
cocrystals would increase in comparison to pterostilbene. Slurry-
ing cocrystal 1 in water at ambient temperature for approximately
1 day resulted in material that was characterized by XRPD to
be a physical mixture of cocrystal 1 and Form I of pterostilbene.
Therefore aliquots were taken for approximately 5 hours to
generate a concentration versus time profile (Figure 6). The
concentration of pterostilbene from the dissolution of cocrystal 1
was measured to be approximately 6 times higher than that of
Form I of pterostilbene after approximately 5 hours. Although
XRPD results indicate partial transformation of cocrystal 1 to
Form I of pterostilbene after 5 hours, the concentration of
pterostilbene was maintained as a plateau over the course of the
powder dissolution experiment. This sustained increase in pter-
ostilbene concentration for cocrystal 1 is similar to that observed
for the pterostilbene-caffeine cocrystal.9 Solution complexation
between pterostilbene and piperazine may exist and is likely
aiding in the sustained solution concentrations for cocrystal 1.
Cocrystal 1 crystallizes in the orthorhombic space group
P212121, Table 1. The asymmetric unit of cocrystal 1 contains
two pterostilbene molecules and one piperazine, linked through
OÀH N hydrogen bonds (Table 2) from the hydroxy moiety
3 3 3
of pterostilbene to the nitrogen atom of piperazine, Figure 2. The
three-component supermolecules are cross-linked into one-
dimensional rows through NÀH O hydrogen bonds from
3 3 3
the NÀH of piperazine to the oxygen atom of the pterostilbene
molecule resulting in a R44(14) graph set notation, Figure 3.
Thirteen hydrogen-bonded (and two halogen-bonded) co-
crystals exist in the CSD with piperazine, showing its ability to
form neutral multicomponent crystals with a broad range of
functional groups.17 Additionally, the same R44(14) graph set
observed in cocrystal 1 occurs when piperazine is cocrystallized
with paracetamol18 or 6-bromo-2-naphthol;19 however, each of
the other 11 hydrogen-bonded structures displays a different set
of connectivities and hydrogen bonding motifs.
Cocrystal 2 crystallizes in the monoclinic space group P21/c,
Table 1. The asymmetric unit of cocrystal 2 contains one
pterostilbene and one glutaric acid molecule, Figure 4.
2820
dx.doi.org/10.1021/cg1016092 |Cryst. Growth Des. 2011, 11, 2817–2823

Crystal Growth & Design
ARTICLE
Figure 5. The 1:1 four-component supramolecular assembly between glutaric acid and pterostilbene, cocrystal 2 (top) and extended architecture
displaying the graph-set notations of the two different hydrogen-bonded synthons within the structure (bottom).
slurry sufficiently lowers the activation barrier to cause dissocia-
tion followed by rapid precipitation of pterostilbene likely
because little or no complexation between pterostilbene and
glutaric acid exists in solution.
As was demonstrated from our previous studies,9 pterostilbene
can act as a coformer to form cocrystals with caffeine and carba-
mazepine; while in this work, pterostilbene was selected as an API in
the formation of cocrystals with pharmaceutically acceptable cofor-
mers, glutaric acid and piperazine. These cocrystal examples provide
cases where the classification or label of the components (API or
coformer) is not critical, and the cocrystal structure and properties
are not affected by the nomenclature of choice. Whether pteros-
tilbene was treated as an API or a coformer, the main observation is
that the concentration of pterostilbene in solution, measured from
the different cocrystals, varied greatly.
It is noted that the dissolution and solubility of the caffeine and
carbamazepine cocrystals presented earlier9 were monitored by
measuring the UV spectrum of the respective components rather
than that of pterostilbene. For these 1:1 mole ratio cocrystals, the
concentration of caffeine or carbamazepine is directly proportional
Figure 6. Concentration vs time profile for cocrystal 1 and equilibrium
to the concentration of pterostilbene (i.e., as one mole of cocrystal
solubility of pterostilbene at ambient temperature.
dissolves, one mole of both components are released to the
solution). Therefore, monitoring either component by UVÀvis
spectroscopy can be done so longasthe cocrystalremainsintact, as
was observed for these two cocrystals over the course of the results
presented. For the piperazine cocrystal, dissociation and precipita-
tion of pterostilbene was observed within the course of the
experiment according to XRPD results; thus the absorbance of
pterostilbene was monitored to provide an accurate representation
of the solution concentration of pterostilbene. Monitoring the
absorbance of piperazine may have resulted in misleading results
since the cocrystal was observed to dissociate and pterostilbene
was found to precipitate during the dissolution study.
Precipitation of pterostilbene was observed within 5 minutes
upon slurrying of cocrystal 2 in water at ambient temperature.
Therefore, it can be assumed that the kinetic solubility of
cocrystal 2 is higher than the solubility of pterostilbene, but
due to rapid precipitation of pterostilbene, a quantitative assess-
ment of the concentration increase was not attainable. Interest-
ingly, cocrystal 2 remained in tact (i.e., did not dissociate) upon
exposure to 98% relative humidity at ambient temperature,
suggesting that water vapor does not lower the activation barrier
enough to cause dissociation. Stirring cocrystal 2 in an aqueous
2821
dx.doi.org/10.1021/cg1016092 |Cryst. Growth Des. 2011, 11, 2817–2823

Crystal Growth & Design
ARTICLE
Present Addresses
†Halliburton, 2600 South 2nd Street, Duncan, OK 73536, USA.
’ ACKNOWLEDGMENT
We thank Dr. John Depser for single-crystal data collection
and structure solution of 1 and 2, Mr. Rex Shipplett for well-plate
screening experiments, and SSCI’s analytical services for data
collection. We also thank Mr. Eyal Barash, Dr. David Engers, and
Ms. Melanie Roe for careful review of the manuscript.
’ REFERENCES
(1) A “nutraceutical” compound can be defined as, “a food (or part of
a food) that provides medical or health benefits, including the preven-
tion and/or treatment of a disease and possesses a physiological benefit
or reduces the risk of chronic disease”.Kalra, E. K. AAPS Pharm. Sci.
2003, 3, No. 25.
(2) (a) Rimando, A. M.; Kalt, W.; Magee, J. B.; Dewey, J.; Ballington,
J. R. J. Agric. Food Chem. 2004, 52, 4713–4719. (b) Adrian, M.; Jeandet,
P.; Douillet-Breuil, A. C.; Tesson, L.; Bessis, R. J. Agric. Food Chem. 2002,
48, 6103–6105. (c) Langcake, P. Physiol. Plant Pathol. 1981,
18, 213–226.
(3) (a) Rimando, A. M.; Cuendet, M.; Desmarchelier, C.; Mehta,
R. G.; Pezzuto, J. M.; Duke, S. O. J. Agric. Food Chem. 2002,
50, 3453–3457. (b) Cichocki, M.; Paluszczak, J.; Szaefer, H.; Piechowiak,
A.; Rimando, A. M.; Baer-Dubowska, W. Mol. Nutr. Food Res. 2008,
52, 62–70. (c) Suh, N.; Paul, S.; Hao, X.; Simi, B.; Xiao, H.; Rimando,
A. M.; Reddy, B. S. Clin. Cancer Res. 2007, 13, 350–355.
(4) (a) Lin, H.-S.; Yue, B.-D.; Ho, P. C. Biomed. Chromatogr. 2009,
23, 1308–1315. (b) Das, S.; Lin, H.-S.; Ho, P. C.; Ng, K.-Y. Pharm. Res.
2008, 25, 2593–2600.
Figure 7. Comparison between the aqueous solubility of pterostilbene
and the solution concentration of pterostilbene achieved from three
cocrystals. The x-fold increase or decrease is calculated from equilibrium
or kinetic solubility measured for the respective cocrystal. Pterostilbene
solubility (black) is represented as one.
Figure 7 shows the increase or decrease in pterostilbene
solution concentration that resulted from the three cocrystals.
The cocrystal of pterostilbene with carbamazepine in fact
decreased the overall solution concentration of pterostilbene
nearly 3-fold, while cocrystals with piperazine and caffeine
achieved 6- and 27-fold increases in solution concentration of
pterostilbene, respectively.
(5) Ivanisevic, I.; Andres, M. C.; Stephens, K.; Bethune, S.; Bower,
M.; Wolfe, S.; Smit, J.; Gilkinson, A.; Roe, M. manuscript in preparation,
2011.
’ CONCLUSIONS
(6) (a) Etter, M. C. Acc. Chem. Res. 1990, 23, 120–126. (b) Aaker€oy,
C. B.; Salmon, D. J. CrystEngComm 2005, 7, 439–448. (c) MacGillivray,
L. CrystEngComm 2004, 6, 77–78. (d) Desiraju, G. R. Acc. Chem. Res.
2002, 35, 565–573. (e) Moulton, B.; Zaworotko, M. Chem. Rev. 2001,
101, 1629–1658. (f) Lehn, J.ÀM. Science 2002, 295, 2400–2403.
(f) Grepioni, F.; Braga, D. Making Crystals by DesignÀFrom Molecules
to Molecular Materials, Methods, Techniques, Applications; Wiley-VCH:
Weinheim, Germany, 2007; Chapter 2.5 and references therein.
(7) (a) Crystal Growth and Design virtual issue on Pharmaceutical
Cocrystals, http://pubs.acs.org/page/cgdefu/vi/1, and references
therein. (b) Good, D. J.; Rodríguez-Hornedo, N. Cryst. Growth Des.
2009, 9 (5), 2252–2264. (c) Childs, S. L.; Chyall, L. J.; Dunlap, J. T.;
Smolenskaya, V. N.; Stahly, B. C.; Stahly, G. P. J. Am. Chem. Soc. 2004,
126, 13335–13342. (d) Shiraki, K.; Takata, N.; Takano, R.; Hayashi, Y.;
Terada, K. Pharm. Res. 2008, 25, 2581–2592.
Within this study, we successfully cocrystallized pterostilbene
with piperazine and glutaric acid, showing the ability of ptero-
stilbene to form cocrystals with pharmaceutically acceptable
coformers. Both cocrystals displayed remarkable physical stabi-
lity, with respect to RH and temperature over specified time
periods of up to 8 weeks. Dissolution of cocrystal 2 resulted in
rapid precipitation of pterostilbene, and thus measurement of the
increase in pterostilbene concentration was unattainable. How-
ever, the aqueous solution concentration was measured for
cocrystal 1 revealing a 6-fold increase in pterostilbene concen-
tration compared with the solubility of pterostilbene. Cocrystals
that result in an increased solution concentration of an API or
nutraceutical compound, such as pterostilbene, could lead to
enhanced pharmacological properties, particularly when the
concentration is sustained for several hours.
(8) Childs, S. L.; R.odríguez-Hornedo, N.; Reddy, L. S.; Jayasankar,
A.; Maheshwari, C.; McCausland, L.; Shipplett, R.; Stahly, B. C.
CrystEngComm 2008, 10, 856–864.
(9) Schultheiss, N.; Bethune, S.; Henck, J. O. CrystEngComm 2010,
12, 2436–2442.
’ ASSOCIATED CONTENT
(10) Coformers were selected for the screen if they possessed
estimated aqueous solubilities of 30 mg/mL or greater.
(11) (a) Handbook of Pharmaceutical Salts: Properties, Selection and
Use; Stahl, P. H., Wermuth, C. G., Eds.; Verlag Helvetica Chimica Acta:
Z€urich, 2002. (b) Generally Regarded as Safe: http://www.cfsan.fda.
gov/∼rdb/opa-gras.html and http://www.cfsan.fda.gov/∼dms/gras-
guid.html (c) Everything Added to Food Stuff in the United States:
Http://vm.cfsan.fda.gov/∼dms/eafus.html.
S
Supporting Information. Simulated and experimental
b
XRPD patterns of cocrystals 1 and 2, XRPD patterns after
temperature or %RH stress experiments, XRPD patterns after
solubility experiments, and crystallographic information files for
1 and 2. This material is available free of charge via the Internet at
http://pubs.acs.org.
(12) APEXII, v1.27; Bruker Analytical X-ray Systems, Madison, WI,
2005.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: sarah.bethune@aptuit.com. Phone: 1-765-463-0112.
(13) SAINT, v6.02; Bruker Analytical X-ray Systems, Madison, WI,
1997À1999.
2822
dx.doi.org/10.1021/cg1016092 |Cryst. Growth Des. 2011, 11, 2817–2823

Crystal Growth & Design
ARTICLE
(14) SHELXTL, v5.10; Bruker Analytical X-ray Systems, Madison,
WI, 1997.
(15) Figure generated using proprietary in-house Pattern Match
software:Ivanisevic, I.; Bugay, D. E.; Bates, S. J. Phys. Chem. B 2005,
109, 7781–7787.
(16) Images created using Mercury 2.1.
(17) CSD refcodes: AYISIK, DIVCUH, DIVDOC, DOSJOK, FEK-
QOC, HIBFED, JESMEA, LAKMOA, LOHNOM, MUPPUI, OGEZIJ,
QAMRAY, RAWFAW, TICFOA, and XOKBOO.
(18) Oswald, I. D. H.; Allan, D. R.; McGregor, P. A.; Motherwell,
W. D. S.; Parsons, S.; Pulham, C. R. Acta Crystallogr. 2002,
B58, 1057–1066.
(19) Tian, Y.; Cui, D. Acta Crystallogr. 2008, E64, 2334–2335.
(20) Etter, M. C.; MacDonald, J. C.; Bernstein, J. Acta Crystallogr.
1990, B46, 256–262.
(21) Approximate aqueous solubilities were calculated based on the
total solvent used to produce a solution; actual solubilities may be
greater due to the slow rate of dissolution.
2823
dx.doi.org/10.1021/cg1016092 |Cryst. Growth Des. 2011, 11, 2817–2823