Effects of crystal form on solubility and pharmacokinetics: A crystal engineering case study of lamotrigine

Article abstract of DOI:10.1021/cg901010v

In this contribution, we describe how the supramolecular synthon approach, can be used for discovery of novel crystal forms and for enhancing the relevant preclinical properties of a low solubility antiepileptic drug, lamotrigine (6-(2,3dichlorophenyl)-l,

Full text of DOI:10.1021/cg901010v

DOI: 10.1021/cg901010v
Effects of Crystal Form on Solubility and Pharmacokinetics: A Crystal
Engineering Case Study of Lamotrigine
2010, Vol. 10
394–405
Miranda L. Cheney,†,‡ Ning Shan,*,† Elisabeth R. Healey,† Mazen Hanna,†
Lukasz Wojtas,‡ Michael J. Zaworotko,*,‡ Vasyl Sava,# Shijie Song,# and
Juan R. Sanchez-Ramos#
†Thar Pharmaceuticals Inc, 3802 Spectrum Boulevard, Suite 120, Tampa, Florida 33612, ‡Department
of Chemistry, University of South Florida, 4202 East Fowler Avenue, CHE205, Tampa, Florida 33620,
and #Departments of Neurology and Neurosurgery, University of South Florida, College of Medicine,
James A. Haley VA Hospital, Tampa, Florida 33612
Received August 22, 2009; Revised Manuscript Received November 4, 2009
ABSTRACT: In this contribution, we describe how the supramolecular synthon approach can be used for discovery of novel
crystal forms and for enhancing the relevant preclinical properties of a low solubility antiepileptic drug, lamotrigine (6-(2,3-
dichlorophenyl)-1,2,4-triazine-3,5-diamine). Ten novel crystal forms are reported: lamotrigine methylparaben cocrystal form I
(1:1) (1), lamotrigine methylparaben cocrystal form II (1:1) (2), lamotrigine nicotinamide cocrystal (1:1) (3), lamotrigine
nicotinamide cocrystal monohydrate (1:1:1) (4), lamotrigine saccharin salt (1:1) (5), lamotrigine adipate salt (2:1) (6),
lamotrigine malate salt (2:1) (7), lamotrigine nicotinate dimethanol solvate (1:1:2) (8), lamotrigine dimethanol solvate (1:2)
(9), and lamotrigine ethanol monohydrate (1:1:1) (10). A selected set of the reported crystal forms were studied to determine their
dissolution rate, solubility, and pharmacokinetic behavior. The solubilities were measured in aqueous media and in acidified
aqueous media (pH = 1). It was observed that 5 and 2 exhibited the highest concentration in the aqueous media and acidified
aqueous media, respectively. In the pharmacokinetic study, the serum concentration of lamotrigine, measured in Spra-
gue-Dawley rats, reached the highest level after a single-dose oral administration of 5.
1. Introduction
Pharmaceutical salts are materials formed by an ionic API
and a suitable, pharmaceutically acceptable counterion.9
They have been a part of crystal form selection for decades
as they offer diversity of composition and can therefore
exhibit a wide range of physicochemical properties.10 Phar-
maceutical salts have been used to enhance the solubility of
poorly soluble APIs which represent approximately 40% of
the drugs on the market11 and as many as 60% of APIs in
development.12 Improving the solubility or dissolution rate of
a BCS Class II API is possible via pharmaceutical salt forma-
tion.10 For example, in the late 1950s Juncher and Raaschou13
developed three novel salt forms of penicillin V that exhibited
superior dissolution profiles in comparison to the original
API. When conducting a pharmacokinetic study, it was
observed that the salt form enabling the highest in vivo
exposure of penicillin V was the same form that possessed
the most rapid dissolution rate.
A more recently applied technique for crystal form deve-
lopment is pharmaceutical cocrystallization. Pharmaceutical
cocrystals can be defined as multiple component crystals in
which at least one component is molecular and a solid at room
temperature (the cocrystal former) and forms a supramole-
cular synthon with a molecular or ionic API.14 Pharmaceu-
tical cocrystals have demonstrated that they can profoundly
modify the physicochemical properties of the parent API
molecule,13-22 and at least 90 APIs have been studied in the
context of cocrystallization. Often APIs that are targeted for
pharmaceutical cocrystallization experience undesirable solu-
bility and/or stability and possess multiple hydrogen bonding
sites.15 A recently published study by Bak and co-workers
highlighted the ability of a series of pharmaceutical cocrystals
to improve the solubility of AMG 517.16,17 In particular, the
AMG 517:sorbic acid cocrystal was studied with respect to its
It is of fundamental importance for the pharmaceutical
industry to understand the nature of crystal packing and its
influence upon the physicochemical properties of active phar-
maceutical ingredient (API) crystal forms. Indeed, crystal form
screening and identification of an optimal crystal form have
become an integral part of the drug development process.1,2
Often a pharmaceutical crystal form tends to afford a high
purity product with reliable reproducibility and scalability,
and it is thermodynamically more stable when compared to
amorphous solids. Stable crystal forms are therefore more
desired by both drug developers and regulatory bodies. In
addition, although crystallization has been widely studied
scientifically since at least the early 19th century, this does
not mean that crystallization is predictable3 or even control-
lable.4 New crystal forms are therefore likely to be patentable
in their own right since they meet the primary criteria for
patentability: novelty, lack of obviousness, and utility.5
Furthermore, it has been known for over 100 years that rate
of dissolution of a solid is at least partly determined by the
thermodynamic solubility of a compound,6 and it is well
recognized that, if absorption is limited by solubility, solubility
critically influences the bioavailability and pharmacokinetics
of an API. Given that the majority of APIs fall into Biophar-
maceutical Classification Scheme7 (BCS) classification II8 (low
solubility, high permeability), the importance of API crystal
form screening and selection is increasing in scope and sig-
nificance. In short, the existence of multiple crystal forms of an
API (e.g., salts, cocrystals, solvates, or hydrates) affords both
challenges and opportunities to the pharmaceutical industry.
*To whom correspondence should be addressed. E-mail: nshan@tharpharma.
com (N.S.); xtal@usf.edu (M.J.Z.).
r
pubs.acs.org/crystal
Published on Web 12/03/2009
2009 American Chemical Society

Article
Crystal Growth & Design, Vol. 10, No. 1, 2010 395
stability in vitro as well as its ability to modify the plasma
concentration of AMG 517 in Sprague-Dawley rats. It was
foundthat after oral administrationof a 500 mg/kgdose ofthe
cocrystal, the Cmax (maximum plasma concentration) and
AUC0-inf (area under curve) were 8 and 9 times greater,
respectively, compared to oral administration of the same
dose of pure API. Childs et al.18 highlighted a cocrystal that
exhibited approximately 4-fold increase in plasma concentra-
tion over the pure API after a single oral dose to dogs. A
review of the literature reveals that there have been eight17-19
pharmaceutical cocrystal case studies with pharmacokinetic
details reported to date, all of which support that pharma-
ceutical cocrystals are a viable option to enhance the clinical
performance of a poorly soluble API. The importance of
hydrates and solvates in pharmaceutical development has
also been recognized.20,21 Various examples have demon-
strated that the formation of hydrates and solvates can
significantly alter the physicochemical properties of APIs,
such as chemical stability, solubility, and dissolution rate.22
With this in mind, we report a study of lamotrigine (6-(2,3-
dichlorophenyl)-1,2,4-triazine-3,5-diamine), a triazine drug
amenable to crystal engineering design strategies that exhibits
poorsolubilityanddissolutionratein itspure crystallineform.
Lamotrigine is marketed as Lamictal by GlaxoSmithKline for
oral administration as a compressed or chewable tablet. It is
primarily used as an anticonvulsant drug for the treatment of
epilepsy as well as in the treatment of psychiatric disorders
such as bipolar disorder.23 In particular, lamotrigine is used
for the treatment of generalized seizures associated with
Lennox-Gastaud syndrome,24 and it can be used in conjunc-
tion with other antiepileptic drugs such as carbamazepine,
phenytoin, phenobarbital, primidone, or valproate.25 An
additional and perhaps less common use for lamotrigine is
for the treatment of neuropathic pain, cluster headaches, and
migraines.26 Lamotrigine, a white to pale cream-colored
powder, exhibits a pKa of 5.7 and a melting point of 218 °C.
It is very slightly soluble in water (0.17 mg/mL at 25 °C) and
slightly soluble in 0.1 M HCl (4.1 mg/mL at 25 °C).27
Lamotrigine tablets are supplied for oral administration as
25 mg, 100 mg, 150 mg, and 200 mg tablets. Various attempts
have been made to solve the use limitations of lamotrigine,
such as poor solubility at the higher pH conditions. Briefly,
these approaches have involved the exploration of a plethora
of crystal forms,28 including salt forms, and reduction of the
particle size.29 Recently, eight novel crystal forms of lamo-
trigine were reported by Galcera et al.,30 with only two salt
forms (saccharinate and DL-hemitartrate dimethylsulfoxide
solvate) reaching a greater maximum aqueous solubility than
pure lamotrigine. A benzoate dimethylformamide solvate,31
hydrogen phthalate dimethylformamide solvate,32 methanol
solvate,33 isoethoinate,34 dimethylformamide solvate,35
methanesulfonate,36 and monohydrate37 have also been re-
ported; however, only limited solubility data are available
concerning these crystal forms. To date, no polymorphic form
of lamotrigine has been identified in-house or, to the best of
our knowledge, reported in the literature. Some of the afore-
mentioned crystal forms, particularly those involving certain
salt forms, are undesirable for certain routes of administra-
tion, such as parenteral, due to their acidity. Other formula-
tions contain ingredients that are not safe for human
consumption such as dimethylformamide. Clearly, there is
a strong scientific and clinical need to develop novel forms
of lamotrigine that have significantly improved physicochem-
ical properties, including aqueous solubility, which can be
Scheme 1. Molecular Structures of Cocrystal and Salt Formers
formulated for use in various delivery routes, such as oral
administration.
To generate novel crystal forms of lamotrigine, an analysis
based upon crystal engineering,2 molecular recognition, and
supramolecular synthon formation38,39 was conducted to
determine complementarities between a number of pharma-
ceutically acceptable and/or approved materials containing
carboxylic acid, alcohol, and primary amide moieties and
lamotrigine. The ultimate goal of this approach, namely, the
supramolecular synthon approach, was to effectively prior-
itize all possible guest molecules for crystal form screening of
drugs, and to avoid the “tactless” high-throughput screening
based on trial-and-error. In practice, the supramolecular
synthon approach was performed by careful analysis of the
Cambridge Structural Database (CSD).40,41 It showed that
lamotrigine can form complexes with two dominant supra-
molecular synthon motifs, with or without the aminopyridine
dimer. Among all pharmaceutically acceptable and/or ap-
proved compounds with hydrogen-bonding functionality, a
variety of guest molecules that were likely to form either of
these two motifs with lamotrigine were selected for this study.
All selected guest molecules except butylated hydroxyanisole
successfully formed complexes with lamotrigine, resulting in
10 novel lamotrigine crystal forms. Details of the supramole-
cular synthon approach and the development of 10 crystal
forms of lamotrigine from established cocrystallization tech-
niques39,42 are presented herein. Solubility and pharmacoki-
netic studies were conducted upon a subset of these crystal
forms.
2. Materials and Methods
2.1. Materials. Lamotrigine was supplied by Jai Radhe Sales,
India, with a purity of 99.79% and was used without further
purification. All other chemicals were supplied by Sigma-Aldrich
and used without further purification.
2.2. Crystal Form Synthesis. Lamotrigine was reacted with seven
compounds shown in Scheme 1, namely, methylparaben, nicotina-
mide, saccharin, adipic acid, L-malic acid, nicotinic acid, and
butylated hydroxyanisole resulting in the formation of 10 crystalline
cocrystals, salts, or solvates of lamotrigine. Multiple functional
groups that would facilitate the complex formation with lamotrigine
have been preidentified based on the supramolecular synthon
approach. Pharmaceutically acceptable and/or approved materials
possessing those functional groups were awarded high priority in
the selection of materials for form screening. The above seven
compounds covered a variety of those high-priority materials, while
each of them represented a distinct class. To date, all compounds
except butylated hydroxyanisole formed complexes with lamotri-
gine successfully. Interestingly, attempts of complexing butylated
hydroxyanisole with lamotrigine resulted in the formation of lamo-
trigine dimethanol solvate.
Synthesis of Lamotrigine Methylparaben Cocrystal Form I (1:1),
1. 0.0117 g (0.046 mmol) of lamotrigine and 0.0750 g (0.490 mmol)

396 Crystal Growth & Design, Vol. 10, No. 1, 2010
Cheney et al.
of methylparaben were dissolved in ca. 2 mL of tetrahydrofuran
(THF) and the resulting solution was left to evaporate at room
temperature. Colorless crystals were afforded within seven days. 1
crystallized concomitantly with methylparaben and lamotrigine
THF solvate. Single crystals of 1 were isolated from this mixture.
Synthesis of Lamotrigine Methylparaben Cocrystal Form II (1:1),
2. This cocrystal was made via multiple methods: (i) solvent-drop
grinding - 0.0722 g (0.282 mmol) lamotrigine was ground with
0.0458 g (0.301 mmol) of methylparaben with 40 μL of THF for 30
min in a mechanical ball-mill with ca. 100% conversion; (ii) slurry -
0.0486 g (0.190 mmol) of lamotrigine and 0.0294 g (0.193 mmol) of
methylparaben were slurried with ca. 3 mL of water at room
temperature for 24 h. 2 was isolated via filtration in 70% yield;
(iii) melt - 0.0751 g (0.293 mmol) of lamotrigine and 0.0485 g (0.319
mmol) of methylparaben were placed in an oven at 115 °C for 2 h. 2
was obtained via slow cooling to room temperature in 97% yield.
Single crystals of X-ray diffraction quality were obtained from slow
cooling of the melt.
Synthesis of Lamotrigine Nicotinamide Cocrystal (1:1), 3. This
cocrystal was prepared via multiple methods: (i) solvent-drop
grinding - 0.2081 g (0.813 mmol) of lamotrigine and 0.2056 g
(1.68 mmol) of nicotinamide were ground with 40 μL of methanol
for 30 min in a mechanical ball-mill with ca. 100% conversion; (ii)
melt - 0.7105 g (2.77 mmol) of lamotrigine and 0.3496 g (2.86
mmol) of nicotinamide were heated at 125 °C for 2.5 h resulting in
98% yield; (iii) lamotrigine nicotinamide cocrystal hydrate, 4, can
be dehydrated to isolate 3 after heating at 160 °C for 6 h.
2.3. Crystal Form Characterization. Single-Crystal X-ray Dif-
fraction. Single crystals were obtained for nine compounds. At-
tempts to crystallize 3 did not afford crystals suitable for single
crystal X-ray crystallographic analysis. Single crystal analysis for 1,
2, and 5-10 was performed on a Bruker-AXS SMART APEX CCD
diffractometer with monochromatized Mo KR radiation (λ =
0.71073 A) connected to a KRYO-FLEX low-temperature device
while 4 was collected using Cu KR radiation (λ = 1.54178 A). Data
for 1, 2, and 5-10 were collected at 100 K. Data for 4 was collected
at 293 K. Lattice parameters were determined using the difference
vector method, and reflection data were integrated using SAINT.43
Structures were solved by direct methods and refined by full matrix
least-squares based on F2 using the SHELXTL package.44 All non-
hydrogen atoms were refined with anisotropic displacement para-
meters. All hydrogen atoms bonded to carbon, nitrogen, and
oxygen atoms were placed geometrically and refined with an
isotropic displacement parameter fixed at 1.2 times Uq of the atoms
to which they were attached. Hydrogen atoms bonded to methyl
groups were placed geometrically and refined with an isotropic
displacement parameter fixed at 1.5 times Uq of the carbon atoms.
Powder X-ray Diffraction (PXRD). 2-8 were characterized by a
D-8 Bruker X-ray powder diffractometer using a Cu KR radiation
(λ = 1.54178 A), 40 kV, 40 mA. Data were collected over an angular
range of 3° to 40° 2θ value in continuous scan mode using a step size
of 0.05° 2θ value and a scan speed of 1.0°/min. Crystal forms 1, 9,
and 10 were not stable to loss of solvent or could not be made in
sufficient quantities to collect an experimental PXRD.
Synthesis of Lamotrigine Nicotinamide Cocrystal Monohydrate
(1:1:1), 4. This cocrystal was made from two methods: (i) slurry -
0.0641 g (0.250 mmol) of lamotrigine and 0.0614 g (0.503 mmol) of
nicotinamide (1:2 molar ratio) were slurried with ca. 1 mL of ethyl
acetate for 24 h. The resulting solid was isolated and filtered for
further use with 92% yield; (ii) solution - 0.1021 g (0.399 mmol) of
lamotrigine and 0.0515 g (0.422 mmol) nicotinamide dissolved in
600 μL of n-butanol. The resulting solution was left to slowly
evaporate at room temperature. Colorless crystals of 4 were formed
within two weeks in 95% yield. The crystals were dehydrated in an
oven at 160 °C for 6 h to form anhydrous cocrystal 3.
Synthesis of Lamotrigine Saccharinate Salt (1:1), 5. This salt was
made via multiple methods: (i) slurry - 50.20 g (196 mmol)
lamotrigine and 35.50 g (194 mmol) of saccharin were slurried in
500 mL water overnight under ambient conditions. The solid was
isolated via filtration in 83% yield. (ii) solution - 0.0102 g (0.040
mmol) of lamotrigine and 0.0103 g (0.056 mmol) of saccharin were
dissolved in ca. 2 mL of methanol. The solution was left to slowly
evaporate at room temperature. Colorless crystals of 5 were af-
forded within seven days in 94% yield.
Synthesis of Lamotrigine Adipate Salt (2:1), 6. 0.0158 g (0.062
mmol) of lamotrigine and 0.0108 g (0.074 mmol) of adipic acid were
dissolved in ca. 2 mL of methanol and the resulting solution was left
to slowly evaporate at room temperature. Colorless crystals of 6
appeared within seven days in 91% yield.
Synthesis of Lamotrigine Malate Salt (2:1), 7. 0.0199 g (0.078
mmol) of lamotrigine and 0.0120 g (0.089 mmol) of L-malic acid
were dissolved in ca. 2 mL of methanol and the resulting solution
was left to slowly evaporate at room temperature. Colorless crystals
of 7 appeared within seven days in 92% yield.
Synthesis of Lamotrigine Nicotinate Dimethanol Solvate (1:1:2), 8.
A solution of ca. 2 mL of methanol, 0.0148 g (0.058 mmol) of
lamotrigine, and 0.0075 g (0.061 mmol) of nicotinic acid was left at
room temperature to slowly evaporate. Colorless crystals of 8
formed after two days in 78% yield.
Synthesis of Lamotrigine Dimethanol Solvate (1:2), 9. 0.0213 g
(0.083 mmol) of lamotrigine and 0.0148 g (0.082 mmol) of butylated
hydroxyanisole were dissolved in ca. 2 mL of methanol, and the
resulting solution was left to slowly evaporate at room temperature.
Colorless crystals of 9 were afforded within five days in 93% yield.
Synthesis of Lamotrigine Ethanol Monohydrate (1:1:1), 10. 0.0819
g (0.320 mmol) of lamotrigine and 0.0415 g (0.340 mmol) of nicoti-
namide were dissolved in ca. 3 mL of a 1:1 ethanol/water solution
mixture while heating followed by rapid cooling. The sample was left
at room temperature and allowed to slowly evaporate. Colorless
crystals of 10 were afforded within two weeks in 74% yield.
Calculated PXRD. Calculated PXRD diffractograms were gen-
erated from the single crystal structures using Mercury 1.5
(Cambridge Crystallographic Data Centre, UK) for the following
complexes: 1-2, 4-10.
Differential Scanning Calorimetry (DSC). DSC was performed
on a Perkin-Elmer Diamond DSC with a typical scan range of
25-280 °C, scan rate of 10 °C/min, and nitrogen purge of ca. 30 psi.
Thermal Gravimetric Analysis (TGA). TGA analysis was per-
formed on a Perkin-Elmer STA 6000 with a typical scan range of
30-300 °C, scan rate of 10 °C/min, and nitrogen purge of ca. 20 psi.
The resulting thermograms were processed using Pyris version 9.
Fourier Transform Infrared Spectroscopy (FT-IR). FT-IR analy-
sis was performed on a Perkin-Elmer Spectrum 100 FT-IR spectro-
meter equipped with a solid-state ATR accessory.
Ultraviolet/Visible Spectroscopy (UV/vis). UV/vis analysis was
performed on a Perkin-Elmer Lambda 25 UV/vis spectrophoto-
meter.
High Performance Liquid Chromatography (HPLC). Analysis
was performed on an HPLC system (Perkin-Elmer Instruments
LLC) comprising the following units: a series 200 Gradient Pump; a
785A UV/vis Detector; a series 200 Autosampler; an NCI 900
Network Chromatography Interface and a 600 Series Link. The
system was operated by a Total Chrome Workstation. The sample
holder temperature was kept at 4 °C with a flow rate of 1 mL/min.
The column was a Microsorb-MV 300-5 C-18 (250 ꢀ 4.6 mm ꢀ
1/400). The mobile phase consisted of a mixture of phosphate buffer
(pH 3.0) with methanol (1/1, v/v). The phosphate buffer was
prepared from 50 mmol/L Na2HPO3 water solution with pH-
controlled HCl titration.
2.4. Dissolution Study. Dissolution studies were performed on 2,
3, 4, and 5 allowing for representative crystal forms from different
crystal form categories (i.e., salt, cocrystal, and solvate) to be
compared against the original API. Both deionized water (25 °C)
and pH 1 aqueous solution (0.1 M HCl, 37 °C) were used. The pH
and temperature conditions were selected to facilitate a direct
comparison to the literature and subsequent animal pharmacoki-
netic study. The crystal forms were sieved to achieve a particle size
between 53 and 75 μm. The dissolution study was conducted using
an excess of free-flowing solid in solution; that is, 100 mg of each solid
was used in ca. 100 mL of water and 500 mg of each solid was used in
ca. 50 mL of 0.1 M HCl. The slurries were stirred with a magnetic stir
bar at a rate of ca. 200-300 rpm. Aliquots were filtered with 0.45 μm
filters after 5, 10, 20, 30, 40, 50, 60, 75, 90, 105, 120, 150, 180, and 240
min. The resulting solution was processed and the concentration of
lamotrigine was measured using a UV/vis spectrophotometer. The
pH values of the resulting solutions and crystal forms of the solid in

Article
Crystal Growth & Design, Vol. 10, No. 1, 2010 397
Table 1. pKa Values and the Resulting ΔpKa Values for the Lamotrigine
Salts
acid
pKa
ΔpKa
adipic acid
L-malic acid
nicotinic acid
saccharin
4.43
3.44
4.92
1.60
1.27
2.26
0.78
4.10
those solutions were also determined. The experiment was repeated
twice to allow for statistical analysis.45
2.5. Animal Pharmacokinetic (PK) Study. Twenty-four hour
animal PK studies were conducted using a single-dose oral admin-
istration of lamotrigine as well as 3, 4, and 5. Five male Spra-
gue-Dawley rats (225-250 g) with preimplanted indwelling jugular
vein catheters were used for each crystal form. The animals were
allowed water ad libitum and fasted overnight before drug admin-
istration. The crystal forms were administered via oral gavage with a
dosage of 10 mg/kg lamotrigine or its equivalent, in the 1 mL
suspension vehicle of a 5% PEG 400 and 95% methyl cellulose
aqueous solution. All crystal forms were observed insoluble in the
vehicle. After dosing, 0.2 mL of blood was withdrawn at 0, 30, 60,
120, 180, 240, 480, 720, and 1440 min. The blood samples were
immediately clotted and centrifuged at room temperature to obtain
the rat serum samples, which were subsequently processed and
analyzed by HPLC according to the literature.46 The PK data was
processed by the Microsoft Office Excel 2003 software package.
Figure 1. Motif 1 involves breaking the aminopyridine dimer and
motif 2 retains the aminopyridine dimer but breaks the exterior
bifurcated interaction.
either motif is feasible given a complementary secondary
component. Disruption of motif 1 occurs in 26 out of 81
aminopyridine entries (32%) and disruption of motif 2
occurs in 39 out of those 81 entries, or 48% of the time. In
order to understand the hydrogen bonding of the primary
amine moiety of the diaminopyridine moiety, the 39 entries
that break motif 2 were analyzed. Among those entries 95%
(37/39) show the exterior amine moiety hydrogen bonding to
the second molecule while only two entries (5%) show the
exterior amine moiety hydrogen bonding to a molecule of the
same kind, that is, another diaminopyridine (Refcodes:
AMCQUN, GICWOF). On the basis of these CSD statis-
tics,41 breaking the supramolecular synthons of motif 1 and
motif 2 is feasible; however, there remains a tendency toward
persistence of the aminopyridine dimer (listing of the re-
fcodes for these searches can be found in the Supporting
Information).
Identification of complementary cocrystal formers for
lamotrigine by statistical examination of the percentage of
occurrence of supramolecular homosynthons versus supra-
molecular heterosynthons was also addressed via a CSD
analysis.41 Interactions between an aminopyridine moiety
and carboxylic acid, primary amide, and alcohol moieties
were examined in order to determine if the supramolecular
heterosynthon or the supramolecular homosynthon would
be more prominent (Table 2, aminopyridine was chosen
instead of diaminopyridine to provide a larger data set for
the statistical analysis).
3. Results and Discussion
3.1. Salt vs Cocrystal. Lamotrigine has the ability to form
both salts and cocrystals due to its relatively basic nature
(pKa = 5.7). By selecting cocrystal formers with a range of
varying acidities, the formation of lamotrigine pharmaceu-
tical salts or cocrystals would be expected. The pKa difference
between lamotrigine and the adduct, ΔpKa (i.e., ΔpKa
=
pKa base - pKa acid), is widely accepted as the key to
predicting whether a salt or cocrystal forms.9,47 It is gene-
rally considered that if ΔpKa > 3 the resulting compound
will be a salt (exemplified in this study by saccharin), whereas
the result is typically a cocrystal if ΔpKa < 0. For ΔpKa
between 0 < ΔpKa < 3 the outcome can be either a salt or
cocrystal or a complex with partial proton transfer.48 The
pKa and ΔpKa values49 involved in this study are summarized
in Table 1. It is noted that the ΔpKa values of three cocrystal
formers (i.e., adipic acid, L-malic acid, and nicotinic acid) fall
in the variable region. Interestingly, all of these cocrystal
formers produce lamotrigine salts, as evidenced by the
proton location and bond length analysis from the single
crystal X-ray diffraction data. Analysis of the carbonyl
region of the solid-state FT-IR spectrum also supports the
formation of lamotrigine salts.
3.2. CSD Analysis. In order to prepare novel crystal forms
of lamotrigine, a crystal engineering2 study incorporating
supramolecular design and the molecular functionality of
lamotrigine, that is, the supramolecular synthon approach,
was conducted. The crystal structure of pure lamotrigine
exhibits two dominant supramolecular synthon motifs, the
aminopyridine dimer (motif 1) and the amine-aromatic
nitrogen synthon (motif 2). Motif 1 and motif 2 are depicted
in Figure 1. The introduction of an additional complemen-
tary component to the crystal lattice of lamotrigine could
lead to an interruption of motifs 1 or 2 by either: breaking the
aminopyridine dimer (breaking motif 1) or breaking the
exterior bifurcated interactions between the aromatic nitro-
gen moieties of one lamotrigine to the primary amine moi-
eties of two additional lamotrigine molecules (breaking
motif 2). An analysis of the CSD indicates that breaking
To conduct the supramolecular homosynthon and hetero-
synthon analysis, a broad distance range was initially se-
lected and then reduced by visual inspection to determine the
appropriate ranges for defining hydrogen bond contact
limits. The values given in Table 2 are a refined data
set which includes only aminopyridine and one additional
moiety sustained by a specified supramolecular hetero- or
homosynthon interaction within a defined distance range.
Our analysis concluded that, in general, the supramolecular
heterosynthons were more dominant than the homo-
synthons. The alcohol moiety was the most statistically
favored to interact with the aminopyridine moiety as there
was 33% (100/307) versus 21% (66/307) occurrence for
the supramolecular heterosynthon and homosynthon,
respectively. There was also a preference for the aminopyr-
idine-acid supramolecular heterosynthon, with a higher

398 Crystal Growth & Design, Vol. 10, No. 1, 2010
Cheney et al.
Table 2. Comparison of Supramolecular Homosynthon versus Supramolecular Heterosynthon with Aminopyridines and Complementary Moieties
no. of entries
w/ both
groups
% homosynthon occurrence:
aminopyridine refined data set
(distance range)
%homosynthon occurrence: %heterosynthon heterosynthon distance
moiety refined data set (distance occurrence refined range (A) refined data
complementary
moiety
range)
data set
set
carboxylic acid
91
40/91 (44%) (N N 2.92-3.17)
3 3 3
0
42/91-acid (46%)
28/91-carboxylate
(31%)
N(py) O 2.50-2.80
3 3 3
N(am)
(acid and carboxylate)
O
2.71-3.10
3 3 3
primary amide
alcohol
15
2/15 (13%) (N N 3.04-3.077)
3 3 3
1/15 (6%) (N O 2.98)
3 3 3
1/15 (6%)
N(py) N 2.97
3 3 3
N(am) O 3.06
3 3 3
307
66/307 (21%) (N N 2.92-3.17)
3 3 3
78/307 (25%) (O O2.61-
100/307 (33%)
N(py) O 2.67-2.90
3 3 3
3 3 3
2.92)
N(am) O 2.78-3.19
3 3 3
percentage of occurrence attributed to the carboxylic acid
group (43/91, 46%) than the carboxylate group (28/91,
31%). Interestingly, the carboxylic acid homosynthon does
not occur in the presence of the aminopyridine functional
group. There are 15 structures that contain aminopyridine
and primary amide moieties, 2 of which form the aminopyr-
idine supramolecular homosynthon, 1 forms the amide
dimer, and 1 forms the aminopyridine-amide supramole-
cular heterosynthon. Unfortunately, this paucity of data
precludes determination of the reliability of the aminopyr-
idine-amide supramolecular heterosynthon. However, the
CSD analysis indicates that aminopyridines are likely to
form supramolecular heterosynthons with molecules con-
taining alcohols and carboxylic acids.
3.2.1. Motif 1 versus Motif 2. The crystal forms presented
herein break either motif 1 or motif 2 present in pure
lamotrigine by incorporating a complementary cocrystal
former. Motif 1 is broken in 4 out of 9 structures, while
the remaining crystal forms contain motif 1 but break
motif 2. Specifically, crystal forms 2, 4, 7, 9, and 10 break
motif 2, while forms 1, 5, 6, and 8 break motif 1. The
individual motifs and how they impact the physicochemical
properties of the crystal form are discussed in the following
segments.
Figure 2. 1 breaks motif 1 as shown in the hydrogen bonding
pattern.
3.2.2. Lamotrigine Methylparaben Cocrystal Form I, 1.
Cocrystals of lamotrigine methylparaben form I (1) crystal-
lize in the space group P21/n (Figure 2). The asymmetric unit
contains one lamotrigine and one methylparaben molecule.
The lamotrigine aminopyridine dimer is not observed in 1.
Instead, the structure is a corrugated tape sustained by two
hydrogen bonds linking lamotrigine and methylparaben.
Specifically, the aromatic nitrogen N2 of the triazine ring is
hydrogen bonded to the hydroxyl moiety of the methylpar-
aben [O1-H10O N2: O N 2.651(4) A, H N 1.831 A,
3 3 3
3 3 3
3 3 3
O-H N 164.8°] and the amine in the 5-position of lamo-
3 3 3
Figure 3. Breaking motif 2 shown in the hydrogen bonding of 2;
disordered molecular moieties have been omitted for clarity.
trigine is hydrogen bonded to the carbonyl group of the ester
moiety of methylparaben [N5-H3N O2: N O 2.823(4)
3 3 3 3 3 3
A, H O 1.971 A, N-H O 162.7°]. The tapes are held
172.3°]. The lamotrigine dimers connect to neighbor-
ing dimers via methylparaben molecules, thereby forming
supramolecular ribbons that extend parallel to the
3 3 3
3 3 3
together by various weak interactions including C-H
N
3 3 3
and Cl Cl interactions.
3 3 3
3.2.3. Lamotrigine Methylparaben Cocrystal Form II, 2.
Lamotrigine methylparaben form II (2) can be obtained via
grinding, slurry, or melt. Crystals of 2 exist in space group P1
with one lamotrigine and one methylparaben in the asym-
metric unit (Figure 3). The chlorinated phenyl ring backbone
of lamotrigine is disordered over two distinct positions with
40% and 60% occupancies, respectively. The attached chlo-
rine atoms are refined over three positions with occupancies
of 40%, 40%, and 20%, respectively. A comparison of the
crystal structures of 1 and 2 reveals that they exhibit different
molecular packing arrangements. Unlike 1, 2 exhibits
lamotrigine centrosymmetric aminopyridine dimers [N3-
a-axis [N5-H5A O1:
3 3 3
N
O
3.036(4) A,
H
O
N
3 3 3
3 3 3
3 3 3
2.205 A, N-H O 157.3°; O1-H1C N2: O
3 3 3 3 3 3
2.711(4) A, H N 1.876 A, O-H N 173.3°]. The methyl-
3 3 3
3 3 3
paraben molecule also serves as a bridge to join these
supramolecular ribbons via N-H O interactions [N5-
3 3 3
H5B O2: N O 2.931(4) A, H O 2.129 A, N-H
O
3 3 3
3 3 3
3 3 3
3 3 3
151.2°].
3.2.4. Lamotrigine Nicotinamide Cocrystal, 3. The cocrys-
tal of lamotrigine and nicotinamide (3) was prepared from a
melt of a 1:1 ratio of lamotrigine and nicotinamide, as
evidenced by PXRD characterization. Because of their basic
nature, both lamotrigine and nicotinamide should remain
neutral in the crystal lattice of 3. Efforts to prepare quality
H3B N4: N N 3.121(4) A, H N 2.246 A, N-H
N
3 3 3 3 3 3 3 3 3
3 3 3

Article
Crystal Growth & Design, Vol. 10, No. 1, 2010 399
Figure 5. Hydrogen bonding of 5 interrupting motif 1.
Figure 4. Hydrogen bonding of 4 interrupting motif 2.
single crystals of 3 for single crystal XRD analysis are
unsuccessful to date.
3.2.5. Lamotrigine Nicotinamide Cocrystal Monohydrate,
4. 4 crystallizes in the space group P1, with the asymmetric
unit consisting of one molecule of lamotrigine, one molecule
of nicotinamide, and one water molecule (Figure 4). Lamo-
trigine molecules form aminopyridine dimers [N3-
H2N N4: N N 3.039(2) A, H N 2.243 A, N-H
154.2°]. Supramolecular ribbon motifs are formed along the
b-axis as the lamotrigine dimers are linked by water mole-
N
3 3 3
3 3 3
3 3 3
3 3 3
cules via N-H O and O-H N interactions [N5-
3 3 3 3 3 3
H3N O1-H5O N2: N O 2.930(2) A, H O 2.089
3 3 3 3 3 3 3 3 3 3 3 3
A, N-H O 165.3°, O N 2.822(2) A, H N 1.967 A,
3 3 3
3 3 3
3 3 3
O-H N 171.7°]. In addition, nicotinamide molecules
3 3 3
form centrosymmetric amide dimers that pack in the layers
between the ribbons of lamotrigine supramolecular units
[N7-H8N O2: N O 2.915(3) A, H O 2.056 A,
3 3 3
3 3 3
3 3 3
N-H O 176.7°]. The water molecules within the ribbons
Figure 6. Hydrogen bonding in compound 6, highlighting motif 1,
breaking the aminopyridine dimer.
3 3 3
hydrogen bond to nicotinamide dimers via additional
O-H N interactions [O1-H6 N6: O N 3.039(3) A,
3 3 3 3 3 3 3 3 3
H
for 6 h, it desolvated and converted to 3.
N 2.162 A, O-H N 166.5°]. By heating 4 at 160 °C
3 3 3
2.818(2) A, H O 2.054 A, N-H O 144.7°]. Each tetra-
3 3 3
3 3 3
3 3 3
mer is hydrogen bonded to four additional tetramers via
either sulfonyl-amine or sulfonyl-chlorine interactions
3.2.6. Lamotrigine Saccharinate Salt, 5. The salt of lamo-
trigine and saccharin (5) crystallizes in space group P21/c.
The asymmetric unit contains one lamotrigine cation and
one saccharin anion. The structure of 5 does not contain the
aminopyridine dimer; however, formation of the dimer is
possible with protonation of the most basic nitrogen (N2) as
it is not incorporated in the lamotrigine dimer. The basic
supramolecular unit in 5 is a tetramer formed between two
lamotrigine and two saccharin ions where N2 is protonated
(Figure 5). Lamotrigine and saccharin are associated via two
2-point recognition aminopyridine-sulfonamide supramole-
cular heterosynthons [N2þ-H1N N6: N N 2.819(2) A,
[N5-H4N O2: N O 2.889(2) A, H O 2.194 A,
3 3 3
3 3 3
3 3 3
N-H O 135.5°; Cl1 O3 3.068(3) A].
3 3 3 3 3 3
3.2.7. Lamotrigine Adipate Salt, 6. 6 crystallizes in space
group of P21/c with two lamotrigine cations and one adipate
dianion in the asymmetric unit. The molecular packing of 6
(Figure 6) is based upon 2-point supramolecular hetero-
synthons between the aminopyridinium and carboxylate
moieties, involving an N-H
O
hydrogen bond
3 3 3
[N3-H3N O1: N O 2.942(2) A, H O 2.110 A,
3 3 3 3 3 3 3 3 3
N-H O 157.5°] and an Nþ-H O- charge-assisted
3 3 3 3 3 3
hydrogen bond [N2þ-H1N O2-: N O 2.627(2) A,
3 3 3
3 3 3
3 3 3 3 3 3
H
N
N 1.940 A, Nþ-H N 177.5°; N3-H2N O1:
3 3 3 3 3 3
H
O 1.790 A, N-H O 158.1°]. Proton transfer is
3 3 3
3 3 3
3 3 3
3 3 3
O 2.830(2) A, H O 1.954 A, N-H O 173.5°].
3 3 3 3 3 3
evidenced by the C-O bond distances of the carboxylate
group (1.248(2) and 1.272(2) A) and the geometry of the
lamotrigine triazine ring. The C3-N2-N1 angle of the
triazine ring in the crystal structure of 6 is 122.4(2)°. Each
discrete unit, comprised of two lamotrigine cations and one
adipate dianion, is hydrogen bonded to eight nearby lamo-
trigine-adipate units through the lamotrigine NH2 moieties
The C3-N2-N1 angle of the triazine ring in the crystal
structure of 5 is 123.4° which correlates to the previously
reported values for protonated lamotrigine50 and the ex-
pected trend for protonated aminopyridines, that is, higher
angles than those of a neutral aminopyridine.50,51 The tetra-
mer is formed from two adjacent supramolecular heterosyn-
thon dimers that are further connected by primary
and neighboring carboxylate [N3-H2N O1: N
O
3 3 3
3 3 3
151.5°;
amine-carbonyl interactions [N3-H3N O1:
3 3 3
N
O
2.861(2) A,
H
O
2.057 A, N-H
O
3 3 3
3 3 3
3 3 3

400 Crystal Growth & Design, Vol. 10, No. 1, 2010
Cheney et al.
Figure 8. Breaking motif 1 shown in the hydrogen bonded assem-
bly of 8.
Figure 7. Hydrogen bonding of 7 breaks motif 2.
N-H
O
hydrogen bonds [N3-H2N O3:
3 3 3
N
159.4°,
O
3 3 3
2.764(3) A,
3 3 3
N5-H4N O2: N O 2.760(2) A, H O 1.931 A,
3 3 3 3 3 3 3 3 3
H
O
1.9223 A, N-H
O
3 3 3
3 3 3
N-H O 156.5°]. Each adipate dianion is hydrogen
N3-H3N O3: N O 2.872(3) A H O 2.061 A
3 3 3
3 3 3
3 3 3
3 3 3
bonded to four additional discrete units via C-O H-N
N-H O 152.8°]. In addition, four methanol molecules
3 3 3
3 3 3
interactions. The overall packing can be viewed as staggered
supramolecular units of lamotrigine-adipate running para-
llel to either (010) or (001) with Cl pi interactions.
attach to each tetramer by hydrogen bonding to lamotrigine
cations and nicotinate anions. Two methanol molecules
interact with carboxylate groups via O-H O hydrogen
3 3 3
3 3 3
3.2.8. Lamotrigine Malate Salt, 7. The salt formed by
lamotrigine and L-malic acid (7) crystallizes in P21/c. The
asymmetric unit is comprised of two lamotrigine cations and
one L-malate dianion. Lamotrigine dimers are formed via a
bonds [O1S-H6S O2: O O 2.777(2) A, H O 1.974
3 3 3
3 3 3
3 3 3
A, O-H O 159.7°], while the other two methanol mole-
3 3 3
cules are inserted between N5 and the aromatic nitrogen of
nicotinate [N5-H4N O4S: N O 2.755(3) A, H
3 3 3 3 3 3
O
N
3 3 3
3 3 3
noncentrosymmetric dimer sustained by N-H N hydro-
1.881 A, N-H O 171.8°, O4S-H7S N6: O-
3 3 3 3 3 3
3 3 3
gen bonds [N5-HN5A N14: N N 2.956(6) A, H
N
N
2.699(3) A, H N 1.863 A, O N 172.7°].
3 3 3 3 3 3
3 3 3
3 3 3
3 3 3
3 3 3
2.078 A, N-H N 174.6°; N15-H15A N4: N
3 3 3 3 3 3
3.2.10. Lamotrigine Dimethanol Solvate, 9. The crystal
structure of lamotrigine monomethanol solvate has pre-
viously been reported.33 In the structure of the monometha-
nol solvate, the basic supramolecular unit is comprised of a
lamotrigine dimer with the adjacent methanol molecule
hydrogen bonded to the exterior. A lamotrigine dimethanol
solvate (9) was obtained from an attempted cocrystalliza-
tion of lamotrigine and butylated hydroxyanisole from
methanol. 9 crystallizes in C2/c with the asymmetric unit
comprised of one lamotrigine and two methanol molecules
(Figure 9). 9 retains the lamotrigine dimer motif and the
supramolecular unit consists of one lamotrigine dimer and
two separately hydrogen bonded methanol molecules. The
lamotrigine supramolecular homosynthon dimer in 9 is
3.082(6) A, H N 2.215 A, N-H N 168.6°]. The
3 3 3
3 3 3
L-malate dianion hydrogen bonds to the lamotrigine dimer
such that a supramolecular chain is formed (Figure 7).
Proton transfer occurs between both carboxylate groups
(C20-O1: 1.275(5) A; C20-O2: 1.235(6) A; C23-O4:
1.257(6) A; C23-O5: 1.275(6) A); of L-malate and aromatic
nitrogen atoms of lamotrigine [N12þ-HN12 O1-:
3 3 3
N
O 2.632(5) A, H O 1.766 A, N-H O 167.8°;
3 3 3
3 3 3
3 3 3
C19-N12-N11 angle 123.2°]. The lengths of the C-O
bonds are typical of a carboxylate moiety and the
C19-N12-N11 angle of 123.2° is consistent with that of a
protonated aromatic nitrogen. The supramolecular chains
that are generated from two lamotrigine cations alterna-
ting with one L-malate dianion extend perpendicular to the
centrosymmetric [N3-H1N N4: N N 3.059(2) A,
3 3 3 3 3 3
bc-plane. The chains interact via NH O hydrogen bonds
3 3 3
H
N5 amines form hydrogen bonds with methanol molecules
N 2.189 A, N-H N 170.0°]. In addition, N3 and
3 3 3 3 3 3
[N3-HN3A O1: N O 2.790(5) A, H O 2.086 A,
3 3 3
3 3 3
3 3 3
N-H O 136.3°; N13-H13A O5: N O 2.928(5) A,
[N3-H2N O2: N O 2.824(2) A, H O 2.210 A,
3 3 3 3 3 3 3 3 3
3 3 3 3 3 3 3 3 3
H
O 2.135 A, N-H O 149.6°] to chains that run
N-H O 126.6°; N5-H3N O2: N O 2.862(2) A,
3 3 3 3 3 3 3 3 3
3 3 3
3 3 3
through the ac-plane, thereby generating a 3D structure.
3.2.9. Lamotrigine Nicotinate Dimethanol Solvate, 8. The
methanol solvate of the salt formed by lamotrigine and
nicotinic acid (8) crystallizes in space group P21/c (Figure
8) with proton transfer observed from the carboxylic acid
H
also observed between methanol molecules and aromatic
O 2.109 A, N-H O 143.1°]. Hydrogen bonds are
3 3 3
3 3 3
nitrogen atoms N1 and N2 [O1-H6O N1: N
2.987(2) A,
O2-H5O N2: N O 2.654(2) A, H N 1.824 A,
N-H O 169.9°]. The lamotrigine dimers and methanol
molecules hydrogen bond to form a ribbon that extends
along the c-axis. Two inversion center related ribbons
interact via CH-N and Cl-pi interactions thus forming a
ribbon bilayer. The bilayers stack along the b-axis in an
abab motif.
O
3 3 3
3 3 3
H
N
2.159 A, N-H O
168.6°;
3 3 3
3 3 3
3 3 3
3 3 3
3 3 3
group to N2 on the triazine ring [N2þ-H1N O2-: N
2.729(3) A, H O 1.880 A, N-H O 161.7°; C3-N2-N1
O
3 3 3
3 3 3
3 3 3
3 3 3
3 3 3
angle 122.8°]. The structure of 8 reveals that the nicotinate
anion breaks the lamotrigine dimer. Similarly to 5, two pairs
of lamotrigine nicotinate adducts interact to form tetrameric
motifs sustained by charge assisted Nþ-H O- and
3 3 3

Article
Crystal Growth & Design, Vol. 10, No. 1, 2010 401
Figure 9. Crystal form 9 breaking motif 2.
Figure 10. Crystal form 10 interrupting motif 2.
3.2.11. Lamotrigine Ethanol Monohydrate, 10. Crystals of
10 form in space group P21/c with an asymmetric unit that is
comprised of one lamotrigine, one ethanol, and one water
molecule. 10 exhibits the lamotrigine dimer as shown in
Figure 10. In the crystal structure of 10, the basic supramo-
lecular unit is the lamotrigine aminopyridine dimer sustained
by two symmetrically related hydrogen bonds [N3-
H2N N4: N N 3.006(2) A, H N 2.191 A, N-H
N
3 3 3 3 3 3 3 3 3
3 3 3
153.7°]. These lamotrigine dimers are further hydrogen
bonded via water molecules [N3-H1N O1: N
O
3 3 3
3 3 3
2.928(2) A, H O 2.295 A, N-H O 128.8°; N5-
3 3 3
3 3 3
H3N O1: N O 2.848(2) A, H O 1.973 A, N-H
O
O
3 3 3 3 3 3 3 3 3
3 3 3
3 3 3
173.2°] and ethanol molecules [N5-H4N O2: N
3 3 3
3.042(2) A, H O 2.346 A, N-H O 136.2°] to form a
3 3 3
3 3 3
supramolecular ribbon parallel to the b-axis. Adjacent su-
pramolecular ribbons are connected through hydrogen
bonds that occur between water and ethanol molecules
[O1-H9O O2: O O 2.829(2) A, H O 1.942 A,
3 3 3
3 3 3
3 3 3
O-H O 178.8°], thereby generating a stacked motif. The
3 3 3
supramolecular synthons involved in the structure of 10 are
reminiscent of pure lamotrigine; however, the water and
ethanol molecules force neighboring lamotrigine dimers
further apart.
3.3. Solubility and Dissolution Study. As shown in Figures
11 and 12, solubility and dissolution studies were conducted
for 2, 3, 4, 5 and pure lamotrigine. The solubility of 5 has
been reported elsewhere30 and is re-examined in this study.
The maximum concentration of pure lamotrigine in water
and pH 1 HCl solution differs by approximately 10% with
the higher solubility of lamotrigine exhibited under more
acidic conditions. The dissolution study in water revealed

402 Crystal Growth & Design, Vol. 10, No. 1, 2010
Cheney et al.
that 5 reached a maximum concentration of ca. 0.45 mg/mL.
This observation is in agreement with the reported literature
aqueous solubility.52 However, given that the solubility of
lamotrigine increased under more acidic conditions, the
improved solubility of 5 in water might be rationalized by
the inadvertent decrease in pH from 5.5 to 5.1. Interestingly,
5 was the only crystal form to decrease the pH value of the
solution during the water dissolution study. The maximum
concentrations of the other crystal forms in water were ca.
0.21 mg/mL, 0.30 mg/mL, 0.23 mg/mL, and 0.28 mg/mL for
2, 3, 4, and pure lamotrigine, respectively. It was also found
that 4 exhibited the lowest concentration in water after 4 h,
which is not surprising as hydrates are typically considered to
be less soluble than the corresponding anhydrate.21,53 The
remaining solid phase after the dissolution study was char-
acterized by PXRD analysis. It was observed that 2 and 5
retained their respective crystal forms after the water dis-
solution study, while 3 and 4 converted to a hydrated form of
lamotrigine.
concentration of ca. 3.8 mg/mL. 4, however, also reaches
a maximum concentration of ca. 3.8 mg/mL after only
5 min, but it then proceeds to decline over the remainder
of the study. This particular type of profile is a product
of the “spring and parachute effect” and has been exhibi-
ted by a number of pharmaceutical cocrystals reported
recently.16,18,54 This profile is significant because it shows
that a greater concentration of API can be achieved at a
much faster rate depending upon the crystal form. A similar
trend is also exhibited by 4 under aqueous conditions;
however, the maximum concentration of 4 was less than that
of pure lamotrigine. A slurry of 3, stirred for 5 min in acidic
media, achieved a concentration ca. 36% greater than pure
lamotrigine. Interestingly, 5, which exhibits the highest con-
centration in aqueous solution, achieves the lowest concen-
tration in acidic media (1.3 mg/mL). The pH values of the
resulting solutions after the 0.1 M HCl slurry for 2, 3, 4, 5,
and pure lamotrigine were also measured, but no significant
pH change was observed. Therefore, the dissolution profiles
of the acidic slurry were not affected by the pH change, even
though pH 1 buffer was not used. All remaining solids con-
verted to the lamotrigine hydrochloride salt after slurry, as
evidenced by PXRD analysis. It is noted that such conver-
sions are concomitant with the crystallization of methylpar-
aben and saccharin during the acidic slurry of 2 and 5,
respectively. More detailed results of the dissolution study
are available in the Supporting Information.
An examination of the dissolution profiles generated
at pH 1 indicates that 2 sustains the highest concentra-
tion throughout the 4-h study achieving a maximum
A recently published article suggests that the solubility of
the cocrystal is directly proportional to the solubility of its
components;55 more specifically, the solubility ratio plotted
against the solubility of the cocrystal former divided by the
solubility of the API should result in a linear relationship. A
similar analysis of crystal forms 2-5 reported herein, how-
ever, does not generate a linear plot. In fact, within this set of
case studies, no clear correlation exists with respect to the
solubility of the salt/cocrystal former and the solubility of the
resulting crystal form. The aqueous solubility of nicotina-
mide is ca. 1 g/mL, the highest of all cocrystal formers studied
(methylparaben = 1 g/400 mL and saccharin = 1 g/290 mL),
and it is also a hydrotrope that is frequently used
for solubility improvement;56 however, the crystal forms
Figure 11. Dissolution profiles in water for lamotrigine and crystal
forms 2-5.
Figure 12. Dissolution profiles at pH = 1 for lamotrigine and crystal forms 2-5.

Article
Crystal Growth & Design, Vol. 10, No. 1, 2010 403
Figure 13. Rat serum concentrations of lamotrigine, 3, 4, and 5.
containing nicotinamide (3 and 4) are not the most soluble in
aqueous solutions. Instead, 5 is the most soluble crystal form
most likely due to the acidic saccharin molecule altering the
pH of the solution to favor the dissolution of lamotrigine.
The low correlations may be due to the small data set or the
inclusion of both protonated and unprotonated species in the
data set.
Correlations between solubility and crystal packing were
also investigated. 5, which breaks motif 1, achieved a greater
concentration in water than under acidic conditions (pH =
1). 2 and 4, which break motif 2, show markedly improved
concentrations in the acidic solution (pH = 1) but exhibited
much lower concentrations in water. On the basis of this data
set, it can be concluded that, for this particular case study,
lamotrigine cocrystals that break motif 2 are more soluble
than pure lamotrigine under acidic conditions (pH = 1),
while lamotrigine salts that break motif 1 are more soluble
than pure lamotrigine in aqueous solutions. However, the
enhanced aqueous solubility of 5 may be attributed to the
inherent acidity of saccharin.
required to reach the therapeutic concentration in humans
could be decreased. It is possible that 5 could enable faster
onset and a better clinical performance. The result of this
preliminary PK study showed that 5 would be the desired
crystal form for further pharmaceutical development.
An analysis of the serum concentrations for 3, 4, and 5 in
the rat in terms of crystal packing concluded that 5, which
broke motif 1, experienced an initial boost in serum concen-
tration of lamotrigine. Meanwhile, serum levels for cocrys-
tals 3 and 4, which retained the aminopyridine dimer and
broke motif 2, were an average of ca. 54% less than that of
pure lamotrigine, thus suggesting that crystal forms that
break motif 1 could exhibit higher serum concentrations
than forms that break motif 2. Interestingly, the greatest
improvements in the PK study and the dissolution study
occur when the aminopyridine dimer is broken. This case
study also illustrates that pharmaceutical cocrystals and salts
can have a significant impact upon drug development as they
can greatly alter the PK profile of the parent drugs.
3.4. Animal Pharmacokinetic (PK) Study. The serum con-
centration of lamotrigine that resulted from single-dose oral
gavage of 3, 4, 5, and pure lamotrigine was measured in
Sprague-Dawley rats over a 24-h time period (Figure 13). 2
was not studied due to its instability after 3 months of aging
at 40 °C in variable humidity. An examination of the serum
concentrations after dosing with 3, 4, and 5 shows that the
PK profile can be substantially altered via cocrystal or salt
formation. Two hours after dosing the average serum con-
centration of 5 was 3.5 μg/mL, which is ca. 1.5 times the level
shown for pure lamotrigine (2.3 μg/mL). 3 and 4 showed a
decrease in the serum concentration by ca. 40% and 68%,
respectively, compared to the pure lamotrigine. The area
under the curve (AUC0-24h) for 3, 4, 5 and pure lamotrigine
was calculated to be 37, 26, 66, and 60 μg/mL, respectively. A
comparison of the average serum concentrations showed
that 5 exhibited the highest initial serum concentration, the
lowest Tmax (time to reach maximum serum concentration)
and the greatest AUC0-24h. Thus, 5 allowed the faster and
higher absorption of lamotrigine in the gastrointestinal tract
of rats. Given the increase of absorption in rats, the time
4. Conclusions
Insummary, theworkpresented herein exemplifieshowsalt
or cocrystal formers can generate novel crystalline forms of
preexisting APIs with different physicochemical properties.
Lamotrigine was targeted for crystal form development with
the goal of improving its solubility and clinical performance.
Ten crystal forms of lamotrigine were developed via the
supramolecular synthon approach. The new crystal forms
differed from pure lamotrigine in that either supramolecular
synthon motif 1 (the aminopyridine dimer) or motif 2 (the
amine-aromatic nitrogen hydrogen bond) were broken. Motif
2 was broken in 5 out of 10 structures (50%), while 4 out of 10
structures (40%) broke motif 1. Interestingly, the majority of
the crystal forms that did not contain motif 1, with the
exception of 1, were all lamotrigine salts.
Several crystal forms were tested to determine solubility,
dissolution rate, and rat PK profiles. Out of 10 crystal forms,
four (2, 3, 4, and 5) were selected for dissolution studies and
three (3, 4, and 5) were selected for PK studies. The solubility/
dissolution study was conducted under aqueous conditions

404 Crystal Growth & Design, Vol. 10, No. 1, 2010
Cheney et al.
and under acidified (pH = 1) aqueous conditions. The dis-
solution profiles for 2 and 4 achieved concentration levels
similar to lamotrigine in aqueous media, while 3 maintained a
concentration equivalent to the maximum solubility of lamo-
trigine. The average concentrations achieved from 2, 3, 4, and
5 during dissolution measurements from the acidic media
surpassed the levels of pure lamotrigine by ca. 48%, 19%,
18%, and 58%, respectively. In the rat PK study, the serum
concentrations for 3 and 4 were less than pure lamotrigine by
37% and 26%, respectively. 5, however, exhibited an initial
increase in the serum concentration of ca. 66%. After ap-
proximately 3 h, the serum concentration of 5 reduced to a
level similar to that of pure lamotrigine.
The influence of a particular cocrystal former upon solubi-
lity and rat PK was examined. The analysis compared the
solubility of the cocrystal former with the solubility and serum
concentration of the subsequent crystal form. For this data
set, the most soluble cocrystal former did not lead to the most
soluble crystal form. In addition, a comparison of the rat PK
data to the solubility of the crystal forms revealed that the
crystal forms which achieved the greatest aqueous solubility
also reached the highest concentrations in the rat PK data. In
contrast, thesolubilityof the crystalforminthe acidic solution
did not correspond to the PK data. A further analysis showed
that, except for 5, the Tmax for lamotrigine, 3 and 4 (10-12 h)
were significantly longer than the time required for the
equilibrium of solutions in the dissolution studies. This
discrepancy could be rationalized by the fact that serum
concentrations of all crystal forms reached ca. 90% Cmax
within 60 min, which correlated to the dissolution profiles.
The slow increase of serum concentration after 60 min could
be the results of the prolonged absorption of the drug in the
gastrointestinal tract as well as slow metabolism and excre-
tion.
information is available free of charge via the Internet at http://
pubs.acs.org/.
References
.
(1) (a) Gardner, C. R.; Walsh, C. T.; Almarsson, O. Nat. Rev. Drug
Discovery 2004, 3, 926. (b) Vishweshwar, Peddy; McMahon, J. A.; Bis,
J. A.; Zaworotko, M. J. J. Pharm. Sci. 2006, 95, 499. (c) Allen, L. V.;
Popovich, N. G.; Ansel, H. C. Ansel’s Pharmaceutical Dosage Forms
and Drug Delivery Systems, 8th ed.; Lippincott Williams and Wilkins:
New York, 2005; (d) Blagden, N.; de Matas, M.; Gavan, P. T.; York, P.
Adv. Drug Delivery Rev. 2007, 59, 617.
(2) (a) Desiraju, G. R. Crystal Engineering: The Design of Organic
Solids; Elsevier: New York, 1989; (b) Pepinsky, R. Phys. Rev. 1955,
100, 971. (c) Schmidt, G. M. J. Pure Appl. Chem. 1971, 27, 647.
(3) Maddox, J. Nature 1988, 335, 201. Ball, P. Nature 1996, 381, 648.
(4) Dunitz, J. D.; Bernstein, J. Acc. Chem. Res. 1995, 28, 193.
(5) Trask, A. V. Mol. Pharmaceutics 2007, 4, 301.
(6) Noyes, A. A.; Whitney, W. R. J. Am. Chem. Soc. 1897, 19, 930.
(7) Takagi, T.; Ramachandran, C.; Bermejo, M.; Yamashita, S.; Yu,
L. X.; Amidon, G. L. Mol. Pharmaceutics 2006, 3, 631.
(8) Hiten, J. S.; Gunta, S.; Dasharath, M. P.; Chhagan, N. P. Bio-
pharm. Drug Dispos. 2009, 30, 524-531.
(9) Stahl, P. H.; Wermuth, C. G. Handbook of Pharmaceutical Salts:
Properties, Selection, and Use; WILEY-VCH: Zurich, 2002.
(10) Berge, S. M.; Bighley, L. D.; Monkhouse, D. C. J. Pharm. Sci.
1977, 66, 1.
(11) Lipinski, C. A. Am. Pharm. Rev. 2002, 5, 82.
(12) Dublin, C. H. Drug Delivery Technol. 2006, 6, 34.
(13) Juncher, H.; Raaschou, F. Antibiot. Med. Clin. Ther. 1957, 4, 497.
.
(14) (a) Almarsson, O.; Zaworotko, M. J. Chem. Commun. 2004, 1889.
(b) Shan, N.; Zaworotko, M. J. Drug Disc. Today 2008, 13, 440. (c)
Zaworotko, M.; Arora, K. Pharmaceutical co-crystals: A new opportu-
nity in pharmaceutical science for a long-known but little studied class
of compounds. In Polymorphism in Pharmaceutical Solids, 2nd ed.;
Brittain, H. G., Ed.; Informa Healthcare: London, 2009.
(15) Lipinski, C. A. J. Pharmacol. Toxicol. 2000, 44, 235.
(16) Stanton, M. K.; Bak, A. Cryst. Growth Des. 2008, 8, 3856.
(17) Bak, A.; Gore, A.; Yanez, E.; Stanton, M.; Tufekcic, S.; Syed, R.;
Akrami, A.; Rose, M.; Surapaneni, S.; Bostick, T.; King, A.;
Neervannan, S.; Ostovic, D.; Koparkar, A. J. Pharm. Sci. 2008,
97, 3942.
The influence of supramolecular synthon motif upon solu-
bility and PK was also examined. Of the crystal forms where
the solubility and rat PK were measured (3, 4, 5) the only
crystal form that broke the aminopyridine dimer (5) also
achieved the highest concentration in aqueous solution and
rat serum, suggesting that breaking the lamotrigine amino-
pyridine dimer can lead to crystal forms with desirable
physicochemical properties. Therefore, when considering
both aqueous dissolution and animal PK data of the crystal
forms presented herein collectively, 5 exhibited the targeted
physicochemical properties with substantial improvements,
and would be an appropriate candidate for further develop-
ment. Given the promising results of 5 shown in this study, it
must be noted that the successful development of a novel
crystal form of lamotrigine encompasses many factors above
and beyond solubility study and rat PK study. For example,
the crystal form must be proven physically stable during the
accelerated stress testing and downstream processing. In addi-
tion, the crystal form has to be scaled up and further evaluated
by studies including additional PK study using larger animals,
animal toxicity study, and animal efficacy study. All of the
above considerations will be taken into account in the plan-
ning and implementation of the future development of 5.
(18) McNamara, D. P.; Childs, S. L.; Giordano, J.; Iarriccio, A.;
Cassidy, J.; Shet, M. S.; Mannion, R.; O’Donnell, E.; Park, A.
Pharm. Res. 2006, 23, 1888.
(19) (a) Chen, A. M.; Ellison, M. E.; Peresypkin, A.; Wenslow, R. M.;
Variankaval, N.; Savarin, C. G.; Natishan, T. K.; Mathre, D. J.;
Dormer, P. G.; Euler, D. H.; Ball, R. G.; Ye, Z. X.; Wang, Y. L.;
Santos, I. Chem. Commun. 2007, 419; E. (b) Dova; E. Mazurek, J. M.;
Anker, J. Tenofovir disoproxil hemi-fumaric acid co-crystal. WO/2008/
143500, 2008; (c) Hickey, M. B.; Peterson, M. L.; Scoppettuolo, L. A.;
Morrisette, S. L.; Vetter, A.; Guzman, H.; Remenar, J. F.; Zhang, Z.;
.
Tawa, M. D.; Haley, S.; Zaworotko, M. J.; Almarsson, O. Eur. J.
Pharm. Biopharm. 2007, 67, 112. (d) Imamura, M. Cocrystal of
C-glycoside derivative and L-proline. WO 2007/114475 A1, 2007. (e)
.
Peterson, M.; Bourghol Hickey, M.; Oliveira, M.; Almarsson, O.;
Remenar, J. Modafinil compositions. US 2005267209 A1, 2005; (f)
Remenar, J.; MacPhee, M.; Peterson, M.; Morissette, S. L.; Almarsson,
.
O. Novel crystalline forms of conazoles and methods of making and
using the same, WO/2005/092884, 2005.
(20) Reutzel-Edens, S. M.; Bush, J. K.; Magee, P. A.; Stephenson,
G. A.; Byrn, S. R. Cryst. Growth Des. 2003, 3, 897.
(21) Khankari, R. K.; Grant, D. J. W. Thermochim. Acta 1995, 248, 61.
(22) (a) Hickey, M. B.; Peterson, M. L.; Manas, E. S.; Alvarez, J.;
.
Haeffner, F.; Almarsson, O. J. Pharm. Sci. 2007, 96, 1090. (b)
Kobayashi, Y.; Ito, S.; Itai, S.; Yamamoto, K. Int. J. Pharm. 2000, 193,
137. (c) Salole, E. G.; Alsarraj, F. A. Drug Dev. Ind. Pharm. 1985, 11, 855.
(d) Pudipeddi, M.; Serajuddin, A. T. M. J. Pharm. Sci. 2005, 94, 929.
(23) (a) Bowden, C. L. Expert Opin. Pharmacol. 2002, 3, 1513. (b) Goa,
K. L.; Ross, S. R.; Chrisp, P. Drugs 1993, 46, 152.
(24) Motte, J.; Trevathan, E.; Arvidsson, J. F. V.; Barrera, M. N.;
Mullens, E. L.; Manasco, P.; Dulac, O.; Mai, J.; Billard, C.; Vallee,
L.; Oberarztin, F.; Ried, S.; Christe, W.; Rating, D.; Vigevano, F.;
Tassinari, C.; Bernardina, B. D.; Majoie, H.; Weber, A.; Laegreid,
L.; EegOlofsson, O.; Bermejo, M.; Martinez, A.; Castello, J. C.;
Serra, M. T. I.; Kirkham, F.; Noronha, M.; Wallace, S.; Black, A.;
Berkovic, S.; Pedespan, J.; Echenne, B.; Pinsard, N.; Parain, D.;
Acknowledgment. We thank Mr. Raymond Houck for
stimulating discussions and valuable input to the project.
Supporting Information Available: Characterization results of
lamotrigine crystal forms, refcodes from the CSD statistical analy-
sis, and more detailed experimental results of dissolution study. This

Article
Peniello, M.; Carriere, J.; Chauvel, P.; Berquin, P.; Garrel, S.;
Crystal Growth & Design, Vol. 10, No. 1, 2010 405
(40) Allen, F. H. Acta Crystallogr. B 2002, 58, 380.
(41) CSD November 2007 release including August 2008 update. Search
parameters: organics only, no ions, 3D coordinates determined,
R < 7.5%
Schaff, J.; Talvik, T.; Douchowny, M.; Gilman, J.; Pellock, J.;
Cheng, R.; Crumrine, P.; Lynch, B.; Hahn, J.; Griesemer, D.;
Wannamaker, B.; Laxer, K.; Bluestone, D.; Bebin, E. New Engl. J.
Med. 1997, 337, 1807.
ꢀꢀ ꢁ
(42) (a) Friscic, T.; Jones, W. Cryst. Growth Des. 2009, 9, 1621. (b) Etter,
(25) Smith, D.; Baker, G.; Davies, G.; Dewey, M.; Chadwick, D. W.
Epilepsia 1993, 34, 312.
M. C.; Reutzel, S. M.; Choo, C. G. J. Am. Chem. Soc. 1993, 115, 4411.
(c) Shan, N.; Jones, W. Green Chem. 2003, 5, 728.
(26) Sporn, J.; Sachs, G. J. Clin. Psychopharm. 1997, 17, 185.
(27) O’Neil, M. J. Merck Index, 14th ed.; Merck Research Laboratories:
Whitehouse Station, NJ, 2006.
(43) SMART, SAINT-Plus, SADABS, XP and SHELXTL; Bruker:
Madison, Wisconsin, USA, 1997.
(44) Sheldrick, G. M. SHELXTL; University of Gottingen: Germany,
1997.
(45) It is noted that this dissolution study did not use the standard USP
apparatus, due to the limitations of our in-house facility. Never-
theless, it is believed that our dissolution study produced reason-
ably comparable results. As part of the future development, a
dissolution study through a contract research organization using
the standard USP apparatus will be conducted.
(46) Castel-Branco, M. M.; Almeida, A. M.; Falcao, A. C.; Macedo,
T. A.; Caramona, M. M.; Lopez, F. G. J. Chromatogr. B 2001, 755,
119.
(28) (a) Floyd, A. G.; Jain, S. Pharmaceutical composition containing
lamotrigine. U.S. Patent 5,942,510, 1999; (b) Garti, N.; Berkovich, Y.;
Dolitzky, B.-Z.; Aronhime, J.; Singer, C.; Liebermann, A.; Gershon, N.
Crystal forms of lamotrigine and proecesses for their preparations. U.S.
Patent 6,861,426 B2, 2005; (c) Lee, G. R. Process for the preparation of
lamotrigine. U.S. Patent 5,925,755, 1999; (d) Parthasaradhi, R. B.;
Rathnakar, R. K.; Raji, R. R.; Muralidhara, R. D.; Chander, R. K. S.
Novel crystalline forms of lamotrigine. U.S. Patent 2005/0119265 A1,
2005; (e) Winter, R. G.; Sawyer, D. A.; Germain, A. Process for the
preparation of lamotrigine, U.S. Patent 5,912,345, 1999.
(29) Aronhime, J.; Samburski, G. Pharmaceutical composition contain-
ing lamotrigine particles of defined morphology. U.S. Patent 2005/
0238724 A1, 2005.
(47) Delori, A.; Suresh, E.; Pedireddi, V. R. Chem.-Eur. J. 2008, 14,
6967. Bhogala, B. R.; Basavoju, S.; Nangia, A. CrystEngComm 2005,
7, 551.
(30) Galcera, J.; Molins, E. Cryst. Growth Des. 2009, 9, 327.
(31) Sridhar, B.; Ravikumar, K. Acta Crystallogr. E 2005, 61, o3805.
(32) Sridhar, B.; Ravikumar, K. Mol. Cryst. Liq. Cryst. 2007, 461, 131.
(33) Janes, R. W.; Lisgarten, J. N.; Palmer, R. A. Acta Crystallogr. C
1989, 45, 129.
(34) Sawyer, D. A.; Copp, F. C. Triazine salt. U.S. Patent 4,847,249,
1989.
(35) Sridhar, B.; Ravikumar, K. Acta Crystallogr. E 2006, 62, O4752.
(36) Palmer, R.; Potter, B.; Leach, M.; Chowdhry, B. J. Chem. Crystal-
logr. 2007, 37, 771.
(48) Childs, S. L.; Stahly, G. P.; Park, A. Mol. Pharmaceutics 2007, 4,
323.
(49) (a) Cabelli, D. E.; Bielski, B. H. J. Z. Naturforschung B 1985, 40,
1731. (b) Gavrilov, A. V.; Kurmaeva, A. I.; Barabonov, V. P. Russ. J.
Phys. Chem. 1980, 54, 902. (c) Koppel, I.; Koppel, J.; Leito, I.; Pihl, V.;
Grahn, L.; Ragnarsson, U. J. Chem. Res., Miniprint 1994, 6, 1173. (d)
Orekhova, Z.; Ben-Hamo, M.; Manzurola, E.; Apelblat, A. J. Solution
Chem. 2005, 34, 687.
(50) Potter, B.; Palmer, R. A.; Withnall, R.; Leach, M. J.; Chowdhry, B.
Z. J. Chem. Cryst. 1999, 29, 701.
(37) Kubicki, M.; Codding, P. W. J. Mol. Struct. 2001, 570, 53.
(38) (a) Desiraju, G. R. Angew. Chem., Int. Ed. 1995, 34, 2311. (b) Bis,
J. A.; Vishweshwar, P.; Weyna, D.; Zaworotko, M. J. Mol. Pharma-
ceutics 2007, 4, 401. (c) Shattock, T. R.; Arora, K. K.; Vishweshwar, P.;
Zaworotko, M. J. Cryst. Growth Des. 2008, 8, 4533. (d) Shan, N.;
Batchelor, E.; Jones, W. Tetrahedron Lett. 2002, 43, 8721. (e) Shan,
N.; Bond, A. D.; Jones, W. Cryst. Eng. 2002, 5, 9. (f) Shan, N.; Bond,
A. D.; Jones, W. Tetrahedron Lett. 2002, 43, 3101. (g) Shan, N.; Jones,
W. Tetrahedron Lett. 2003, 44, 3687.
(51) Boenigk, D.; Mootz, D. J. Am. Chem. Soc. 1988, 110, 2135.
(52) Banerjee, R.; Bhatt, P. M.; Ravindra, N. V.; Desiraju, G. R. Cryst.
Growth Des. 2005, 5, 2299.
´
(53) Rodrıguez-Hornedo, N.; Lechuga-Ballesteros, D.; Hsiu-Jean, W.
Int. J. Pharm. 1992, 85, 149.
(54) 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.
´
(55) Good, D. J.; Rodrıguez-Hornedo, N. Cryst. Growth Des. 2009, 9,
2252.
(39) Weyna, D. R.; Shattock, T.; Vishweshwar, P.; Zaworotko, M. J.
Cryst. Growth Des. 2009, 9, 1106.
(56) Rasool, A. A.; Hussain, A. A.; Ditter, L. W. J. Pharm. Sci. 1991, 80,
387.