International Journal of Pharmaceutics 378 (2009) 136–139

Contents lists available at ScienceDirect

International Journal of Pharmaceutics

journal homepage: www.elsevier.com/locate/ijpharm

Note

Characterization of thermal behavior of deep eutectic solvents and

their potential as drug solubilization vehicles
Henry G. Morrison a,∗ , Changquan C. Sun b,∗∗ , Sesha Neervannan c
a
Amgen Inc, One Amgen Center Drive, Thousand Oaks, CA 91320, USA
b
Department of Pharmaceutics, University of Minnesota, 308 Harvard St. SE, WDH 9-127B, Minneapolis„ MN 55455, USA
c
Allergan Inc, 2525 Dupont Drive, Irvine, CA 92612, USA

a r t i c l e i n f o a b s t r a c t

Article history: Deep eutectic solvent (DES) is a new class of solvents typically formed by mixing choline chloride with
Received 11 February 2009 hydrogen bond donors such as amines, acids, and alcohols. Most DES’s are non-reactive with water,
Received in revised form 13 May 2009 biodegradable, and have acceptable toxicity profiles. Urea–choline chloride and malonic acid–choline
Accepted 19 May 2009
chloride eutectic systems were characterized using differential scanning calorimetry (DSC) and thermal
Available online 27 May 2009
microscopy. A potential new 2:1 urea–choline chloride cocrystal with a melting point of 25 ◦ C was char-
acterized at the eutectic composition. The formation of this cocrystal suggests that DES should not be
Keywords:
universally explained by simple eutectic melting, and may be useful in guiding the search for new DES
Deep eutectic solvent (DES)
Solubilization
systems. The lack of nucleation of the malonic acid–choline chloride system prohibited the construction
Eutectic melting of a phase diagram for this system using DSC. We also investigated possible uses of DES in solubilizing
Urea–choline chloride poorly soluble compounds for enhanced bioavailability in early drug development such as toxicology
Malonic acid–choline chloride studies. For five poorly soluble model compounds, solubility in DES is 5 to 22,000 folds more than that
in water. Thus, DES can be a promising vehicle for increasing exposure of poorly soluble compounds in
preclinical studies.
© 2009 Elsevier B.V. All rights reserved.

1. Introduction erties. (Abbott et al., 2003, 2004a,b) This class of solvents, termed
deep eutectic solvent (DES), have increased solubilities for inor-
Mixtures of some high melting compounds can exhibit sub- ganic salts, aromatic acids and amino acids. DES’s are advantageous
ambient eutectic melting temperatures and can be used as solvents because they can be easily prepared in high purity at low cost, and
at ambient temperature. Eutectic mixtures have been used to their components are biodegradable with low toxicity.
generate liquid molten salts at ambient temperatures through In this investigation, we characterized the thermal behavior of
interactions between quaternary ammonium salts and metal salts. two known DES systems and then explored whether DES can be
(Welton, 1999; Wasserscheid and Keim, 2000; Abbott et al., 2001, used to enhance solubility of poorly water soluble compounds.
2003, 2004a,b; Wasserscheid and Welton, 2007) Ionic liquids con-
taining aluminum chloride are highly reactive with water, while 2. Materials and methods
other metal containing salts such as ZnCl2 , SnCl2 and FeCl3 form
ionic solutions that do not react with water. (Abbott et al., 2001; Choline chloride (Alfa Aesar, Ward Hill, MA) was dried and stored
Wasserscheid and Welton, 2007) While useful in the fine chemical in a sealed chamber containing P2 O5 (0–3% RH). Urea (J.T. Baker,
industry, applications of ionic liquids in the pharmaceutical indus- Phillipsburg, NJ), malonic acid (Alfa Aesar, Ward Hill, MA), ben-
try have been very limited due to issues with toxicity, purity, and zoic acid (J.T. Baker, Phillipsburg, NJ), danazol (MP Biomedicals Inc.,
high costs. Solon, OH), itraconazole (Sigma, St Louis, MO), and AMG517 free
Recently, it was shown that substituted quaternary ammonium base (Amgen, Thousand Oaks, CA) were stored in sealed vials at
salts with urea or carboxylic acids can also form eutectics that are room temperature, while griseofulvin (MP Biomedicals Inc., Solon,
liquid at ambient temperature and exhibit interesting solvent prop- OH) was stored in a sealed vial in the freezer.
The eutectic mixtures were formed by heating and stirring two
components between 100 and 150 ◦ C until a colorless liquid was
∗ Corresponding author. Tel.: +1 805 313 5502.
formed. A portion of each warm eutectic mixture liquid was loaded
∗∗ Corresponding author. Tel.: +1 612 624 7322. into a hermetically sealed aluminum pan. The sample was then
E-mail addresses: [email protected] (H.G. Morrison), [email protected] cooled to −60 ◦ C and then heated to 125 ◦ C at 1 ◦ C/min on a differen-
(C.C. Sun). tial scanning calorimeter (DSC, Q1000, TA Instruments, New Castle,

0378-5173/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpharm.2009.05.039
 H.G. Morrison et al. / International Journal of Pharmaceutics 378 (2009) 136–139 137

Fig. 1. Phase diagram urea–choline chloride mixtures.

DE). X-ray powder diffraction (XRPD) patterns were collected using

a Phillips X-ray automated powder diffractometer (X’Pert, Phillips
Analytical, Netherlands) equipped with a sub-ambient chamber
accessory. Diffraction patterns were collected between 3–40◦ 2
at a scan rate of 0.5◦ 2 per minute with 45 kV voltage and
40 mA current. Microphotographs were taken using a microscope
(Nikon Eclipse E600 POL, Melville, NY) with a hot-stage accessory
(Linkam LTS 350, United Kingdom). For hot-stage microscopy, the
urea–choline chloride eutectic was heated from −100 ◦ C to 30 ◦ C at
a rate of 5 ◦ C/min, while the malonic acid–choline chloride eutectic
was heated from −150 ◦ C to −50 ◦ C at a rate of 5 ◦ C/min.
Solubility was determined by saturating a given solvent with a
test compound at room temperature for 24 h while being agitated.
The suspension was then filtered using a syringe membrane filter
(0.25 ␮m). Concentration in each saturated solution was deter-
mined using a HPLC (HP1100, Agilent, Santa Clara, CA) equipped
with a Luna 3u C18(2) 100A column (30 × 4.60 mm, 3 ␮m). Analy-
ses were run with a gradient method using mixtures of 98% (wt%)
water + 2% acetonitrile and 98% acetonitrile (wt%) + 2% water as the
mobile phases. Phase identity of equilibrium solid was analyzed by
XRPD.

3. Results and discussion

Fig. 1 shows the phase diagram for urea and choline chloride.
Similar to a previous report, the eutectic composition is 67.7%
(mol%) urea and 32.3% choline chloride. (Abbott et al., 2003) Upon
heating the eutectic mixture from −60 ◦ C, an exotherm with an
onset temperature of −27 ◦ C was observed (Fig. 2). This was fol-

Fig. 3. Hot-stage microscope images for the urea–choline chloride eutectic.

lowed by an endotherm with an onset temperature of 17 ◦ C. This

type of thermal behavior is consistent with crystallization of a
glass that is followed by melting. On sub-ambient hot-stage, the
liquid solidified into a glass, which did not exhibit birefringence,
upon cooling to −90 ◦ C (Fig. 3). Upon heating, crystallization of
the amorphous glass commenced at −40 ◦ C and completed at
−24 ◦ C. Melting of the crystalline phase began at approximately
17 ◦ C and completed at 26 ◦ C. The unique XRPD pattern of the eutec-
tic mixture from −40 ◦ C to −20 ◦ C confirmed the presence of a
new crystalline phase with a composition of two urea for each
choline chloride, hereafter called 2:1 cocrystal (Fig. 4). Cocrystals
are crystals that contain two different molecules that do not rely on
ionization of the API and the counter-ion to make a solid. They have
Fig. 2. DSC curve for the urea–choline chloride eutectic. recently been recognized as a way to enhance solubility/dissolution,
138 H.G. Morrison et al. / International Journal of Pharmaceutics 378 (2009) 136–139

Fig. 4. XRPD patterns for the (A) urea–choline chloride eutectic point mixture (B)
choline chloride form B (C) choline chloride form A and (D) urea.

hygroscopicity, stability and the IP position with respect to the

development of active pharmaceutical ingredients. This 2:1 cocrys-
tal suggests that the formation of a new compound with a very low
melting point is responsible for the observed deep eutectic behavior
of the urea–choline chloride system. The formation mechanism of
the DES therefore should not be restricted to simple eutectic melt-
ing. We expect the 2:1 cocrystal to form eutectic systems with urea
and choline choride. However, we did not detect distinctive eutec-
tic melting temperatures in mixtures containing either 60 mol% or
70 mol% urea (Fig. 1). This can happen if the eutectic composition
between the cocrystal and either urea or choline chloride is very
close to that of the cocrystal. Heating mixtures of compositions
different from the 2:1 cocrystal showed XRPD patterns of excess
component, urea or choline chloride, but not that of the 2:1 cocrys-
tal. This suggests that a small excess of urea or choline chloride can
substantially slow down crystallization of the 2:1 cocrystal.
Similar work was performed to construct the phase diagram
for the malonic acid–choline chloride system. The eutectic com-
position was suggested to be 50 mol % malonic acid. (Abbott et
al., 2003) When the mixture was heated on the DSC from −80 ◦ C Fig. 6. Hot-stage microscope images for the malonic acid–choline chloride eutectic.
to 100 ◦ C, no thermal events were observed except for the shift in
baseline at above −50 ◦ C (Fig. 5). This behavior is consistent with a
readily nucleation and crystal growth during heating. The lack of
glassy material that undergoes glass transition but does not crystal-
nucleation prohibited the derivation of a phase diagram for this
lize upon heating. This interpretation is supported by sub-ambient
system using DSC. The previous work on this system was based on
microscopy. Upon cooling the liquid to −125 ◦ C, the sample solid-
freezing point measurement of the liquid upon cooling. (Abbott et
ifies as a glass (Fig. 6). Upon heating, the glassy material liquefied
al., 2004a,b) Our attempts to determine the phase diagram for the
between −84 and −60 ◦ C. The liquid is very viscous when handled
malonic acid–choline chloride system by preparing powder mix-
at room temperature and may be a reason why we do not observe
tures in a liquid nitrogen bath (instead of starting with the liquid
eutectic) were unsuccessful because of the rapid eutectic melting
upon mixing.
The solubilization potential of the urea–choline chloride and
malonic acid–choline chloride eutectic systems were studied using
several poorly soluble compounds including benzoic acid, griseo-
fulvin, danazol, itraconazole and a recently reported experimental
drug AMG517 (Stanton and Bak, 2008). Solubility was measured in
the pure DES, mixtures of the DES and water (75:25 and 50:50 by
weight), and pure water. The data is summarized in Tables 1 and 2.
The results indicate that the solubility of the test compounds
increased by 5 to 22,000 folds when compared with their solubility
in water.
We also determined the solubility of the various compounds
in aqueous solutions of the individual components of the eutec-
tic mixtures (Table 3). Compositions for solvents were based on
the molar ratios required to mimic the 75:25 DES–water mixtures
with the absence of one component from the eutectic. Solubiliza-
Fig. 5. DSC curve for the malonic acid–choline chloride eutectic. tion power of these solvents is significantly lower than that of DES
 H.G. Morrison et al. / International Journal of Pharmaceutics 378 (2009) 136–139 139

Table 1
Summary of solubility data in urea–choline chloride eutectic/water.

Solvent Drug solubility (mg/mL)a

Benzoic acid Danazol Griseofulvin AMG517 Itraconazole

Pure eutectic 229 0.048 0.034 0.010 <0.001

75:25 Eutectic:water 23 (4.4) 0.0061 (9.4) 0.016 (9.2) 0.00022 (9.4) <0.001 (8.9)
50:50 Eutectic:water 14 (4.1) 0.002 (9.4) 0.015 (8.7) <0.0001 (9.2) <0.001 (9.0)
Pure water 3 (3.8) <0.0005 (8.9) 0.007 (8.9) <0.0001 (9.5) <0.001 (8.5)
a
pH in brackets

Table 2
Summary of solubility data in malonic acid–choline chloride eutectic/water.

Solvent Drug solubility (mg/mL)a

Benzoic acid Danazol Griseofulvin AMG517 Itraconazole

Pure eutectic 35 0.160 1.0 0.4727 22

75:25 Eutectic:water 18 (0.2) 0.0044 (0.3) 0.1007 (0.3) 0.014 (0.4) 6.6 (0.6)
50:50 Eutectic:water 11 (0.6) 0.002 (0.6) 0.043 (0.7) 0.002 (0.7) 1.2 (0.85)
Pure water 3 (3.8) <0.0005 (8.9) 0.007 (8.9) <0.0001 (9.5) <0.001 (8.5)
a
pH in brackets

Table 3
Summary of solubility data in the individual aqueous components.

Solventb Drug solubility (mg/mL)a

Benzoic acid Danazol Griseofulvin AMG517 Itraconazole

Aqueous urea 11 (3.9) 0.01 (8.2) 0.053 (8.0) <0.0001 (8.7) <0.001 (5.2)
Aqueous choline chloride 7 (3.1) <0.0005 (6.2) 0.005 (7.2) <0.0001 (8.0) <0.001 (5.8)
Aqueous malonic acid 9 (0.9) <0.0005 (0.9) 0.084 (0.9) 0.005 (0.9) 12 (0.9)
Pure water 3 (3.8) <0.0005 (8.9) 0.007 (8.9) <0.0001 (9.5) <0.001 (8.5)
a
pH in brackets
b
Compositions for solvents were based on the molar ratios required to mimic the 75:25 eutectic:water mixtures sans one component from the eutectic

for all drugs tested. Thus, the optimum solubilization requires the exposure is limited by solubility/dissolution and the typical ideal
presence of both components in the DES and is not simply due to dosing volume is typically 2.5 mL for rats and 0.25 mL for mice.
the complexation of drugs with either component in the DES.
References
4. Conclusions
Abbott, A.P., Capper, G., Davies, D.L., Munro, H., Rasheed, R.K., Tambyrajah, V., 2001.
Preparation of novel, moisture-stable, Lewis-acidic ionic liquids containing qua-
The previously reported DES of urea and choline chloride sys- ternary ammonium salts with functional side chains. Chem. Commun., 2010.
tem resulted from the formation of a urea–choline chloride (2:1) Abbott, A.P., Capper, G., Davies, D.L., Rasheed, R.K., Tambyrajah, V., 2003. Novel sol-
cocrystal with a low melting point, not from a simple eutectic sys- vent properties of choline chloride/urea mixtures. Chem. Commun., 70–71.
Abbott, A.P., Capper, G., Davies, D.L., Rasheed, R.K., 2004a. Ionic liquids based upon
tem. metal halide/substituted quaternary ammonium salt mixtures. Inorg. Chem. 43,
DES could improve solubility of poorly soluble compounds by 5 3447–3452.
to 22,000 folds when compared with the solubility in water. The Abbott, A.P., Boothby, D., Capper, G., Davies, D.L., Rasheed, R.K., Tambyrajah, V., 2004b.
Deep eutectic solvents formed between choline chloride and carboxylic acids. J.
material safety data sheets indicate that the toxicity profile for Am. Chem. Soc. 126, 9142–9147.
rats (oral LD 50) for urea, choline chloride and malonic acid are Stanton, M., Bak, A., 2008. Physicochemical properties of pharmaceutical co-crystals:
8471 mg/kg, 3400 mg/kg and 1310 mg/kg, respectively, while the a case study of ten AMG 517 co-crystals. Cryst. Growth Des. 8, 3856–3862.
Wasserscheid, P., Keim, W., 2000. Ionic liquids - new solutions for transition metal
toxicity profile for mice (oral LD 50) for urea, choline chloride and
catalysis. Angew Chem., Int. Ed. 39, 3772–3789.
malonic acid are 11 g/kg, 3900 mg/kg and 4000 mg/kg, respectively. Wasserscheid, P., Welton, T., 2007. Ionic Liquids in Synthesis, 2nd edition. Wiley-VCH
Because the components in DES are pharmaceutically acceptable, Verlag, Weinheim, Germany.
they could potentially be used as vehicles for oral dosing of rats Welton, T., 1999. Room-temperature ionic liquids. Solvents for synthesis and catal-
ysis. Chem. Rev. 99, 2071–2083.
during early development pharmacokinetic investigations where