http://informahealthcare.

com/phd
ISSN: 1083-7450 (print), 1097-9867 (electronic)

Pharm Dev Technol, 2014; 19(8): 976–986

! 2014 Informa Healthcare USA, Inc. DOI: 10.3109/10837450.2013.846374

RESEARCH ARTICLE

Curcumin amorphous solid dispersions: the influence of intra and

intermolecular bonding on physical stability
Lindsay A. Wegiel1, Yuhong Zhao2,3, Lisa J. Mauer2, Kevin J. Edgar4, and Lynne S. Taylor1
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Maastricht on 06/09/14

1
Department of Industrial and Physical Pharmacy, College of Pharmacy, Purdue University, West Lafayette, IN, USA, 2Department of Food Sciences,
Purdue University, West Lafayette, IN, USA, 3Department of Food Sciences, Forestry College, Northeast Forestry University, Harbin, China, and
4
Department of Sustainable Biomaterials, Virginia Polytechnic Institute and State University, Blacksburg, VA, USA

Abstract Keywords
We have investigated the physical stability of amorphous curcumin dispersions and the role of Amorphous, curcumin, FTIR, hydrogen
curcumin–polymer intermolecular interactions in delaying crystallization. Curcumin is an bonding, physical stability, polymers, solid
interesting model compound as it forms both intra and intermolecular hydrogen bonds in the dispersion
crystal. A structurally diverse set of amorphous dispersion polymers was investigated;
poly(vinylpyrrolidone), Eudragit E100, carboxymethyl cellulose acetate butyrate, hydroxypropyl History
methyl cellulose (HPMC) and HPMC-acetate succinate. Mid-infrared spectroscopy was used to
determine and quantify the extent of curcumin–polymer interactions. Physical stability under Received 5 July 2013
different environmental conditions was monitored by powder X-ray diffraction. Curcumin Revised 12 September 2013
chemical stability was monitored by UV-Vis spectroscopy. Isolation of stable amorphous Accepted 13 September 2013
Published online 4 November 2013
For personal use only.

curcumin was difficult in the absence of polymers. Polymers proved to be effective curcumin
crystallization inhibitors enabling the production of amorphous solid dispersions; however, the
polymers showed very different abilities to inhibit crystallization during long-term storage.
Curcumin intramolecular hydrogen bonding reduced the extent of its hydrogen bonding with
polymers; hence most polymers were not highly effective crystallization inhibitors. Overall,
polymers proved to be crystallization inhibitors, but inhibition was limited due to the
intramolecular hydrogen bonding in curcumin, which leads to a decrease in the ability of the
polymers to interact at a molecular level.

Introduction can be expected upon disrupting its crystalline structure.

The value of amorphous solid dispersions may be limited by the
Curcumin is a yellow–orange pigment derived from the rhizome
potential physical instability of these systems over pharmaceut-
of Curcuma longa. Curcumin exhibits keto–enol tautomerism: the
ically relevant time scales, in particular, drug crystallization,
di-keto form is shown in Figure 1a, and the keto–enol form is
which will negate the solubility advantage offered by using an
shown in Figure 1b. This compound has been of great interest
amorphous formulation.
recently due to its promising and diverse pharmacological and
The tendency of a drug to crystallize depends on both
biological properties, including anti-oxidant1,2, anti-inflamma-
environmental and formulation parameters. Factors such as
tory2, anti-fungal3, anti-tumor4, anti-cancer5, anti-viral6, anti-
temperature and relative humidity (RH) can have a pronounced
13
obesity7 and cardioprotective effects8. Curcumin is moderately
20
effect on the physical stability of solid dispersions16,17. Polymers
hydrophobic (logP 2.5) and has very low bioavailability due, in
are typically added to inhibit drug crystallization during processing
part, to its low aqueous solubility (11 ng/mL at pH 5.0)9.
and storage. The physical stability of a solid dispersion will not
In addition, curcumin degrades very quickly at neutral or alkaline
only depend on the inherent crystallization tendency of the drug but
pH, which can further decrease its bioavailability10,11. Numerous
also on the ability of the polymer to inhibit crystallization18.
attempts to increase curcumin solubility have been reported in the
Curcumin is an intermediate crystallizer according to the classi-
literature, including complexation with cyclodextrins9, formation
fication system of Van Eerdenbrugh et al.19. It is therefore crucial
of nanoparticles12 and formulation into solid dispersions13. One
to understand polymer properties important for crystallization
of the most promising strategies for curcumin is formulation
inhibition of curcumin if stable amorphous dispersions are to be
as a solid dispersion, as the amorphous state of the drug has a
formed.
significant solubility advantage over its crystalline counter-
Recent studies from our group have suggested correlations
part9,14,15. This is due to the high melting point of curcumin
between crystallization rates and the extent to which different
(183  C), which reflects strong crystal lattice energy; substantial
polymers form drug–polymer intermolecular interactions20–23.
improvements in solution concentration and biological exposure
Furthermore, using techniques such as infrared (IR) spectroscopy
to monitor drug polymer interactions appears to be a reliable
method for predicting amorphous solid dispersion physical
Address for correspondence: Lynne S. Taylor. Tel: þ1-755-496-6614. stability23. Curcumin is an interesting model compound as it
Fax: þ1-755-494-6545. E-mail: lstaylor@purdue.edu has considerable intramolecular hydrogen bonding; therefore,
 DOI: 10.3109/10837450.2013.846374 Curcumin solid dispersions – role of interactions on physical stability
there is less opportunity for this compound to form intermolecular
hydrogen bonds with polymers (Figure 1). In this study, disper-
sions of curcumin with select polymers were formed. IR
spectroscopy was used to evaluate the extent of curcumin–
polymer intermolecular interactions, and the impact of these
interactions upon the rate of crystallization was evaluated. Model
polymers were chosen for diversity in hydrogen bonding potential,
including poly(vinylpyrrolidone) (PVP), Eudragit E100 (E100),
carboxymethyl cellulose acetate butyrate (CMCAB), hydroxypro-
pyl methyl cellulose (HPMC) and HPMC-acetate succinate
(HPMCAS) (Figure 2).
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Maastricht on 06/09/14
For personal use only.
Figure 1. Chemical structure of (a) curcumin di-keto form, (b) curcumin
keto–enol form and (c) curcumin keto–enol form showing resonance
assisted intra-molecular hydrogen bond. *Denotes atoms involved in
intermolecular hydrogen bonding in the crystalline form I polymorph.
Figure 2. Chemical structures of (a) Eudragit E100, (b) PVP, (c) CMCAB, (d) HPMC and (e) HPMCAS.
977
Materials
Dichloromethane (ChromAR grade) and acetone (ChromAR
grade) were obtained from Mallinckrodt Baker, Inc., Paris, KY.
Absolute ethanol was obtained from Pharmaco Products, Inc.,
Brookfield, CT and Aaper, Shelbyville, KY. Curcumin was
obtained from Spectrum Chemical Manufacturing Corp., New
Brunswick, NJ. PVP (K29/32, Mw 58 000) was purchased from
Sigma-Aldrich (St. Louis, MO). HPMC (grade 606, hypromellose
United States Pharmacopeia substitution type 2910, Mw 35 600)
and HPMCAS (grade AS-MF, Mw 17 000) were provided by Shin-
Etsu Chemical Co., Ltd., Tokyo, Japan. Eudragit E100 (Mw
47 000) was obtained from Rohm GmbH, Darmstadt, Germany.
CMCAB (641-0.5, Mw 35 000) was obtained from Eastman
Chemical Co., Kingsport, TN. The polymers were kept in
desiccators with P2O5 for at least 1 week prior to use to remove
any residual moisture.
Methods
Formation of solid dispersions
Attempts to make curcumin amorphous by rapid solvent evapor-
ation were only successful by using the small scale spincoating
technique described below. Attempts to obtain larger quantities of
neat amorphous curcumin by rotary evaporation (technique
described below) were not successful. However, a solvent-free
process (cryomilling in a liquid nitrogen bath) permitted prepar-
ation of a larger quantity of amorphous curcumin powder.
Spincoating
Solutions were prepared by dissolving curcumin and polymer at
different dry weight ratios in a 1:1 (by weight) mixture of
dichloromethane and ethanol. All mixtures were visually
inspected to confirm that both curcumin and polymer were fully
dissolved, forming a uniform one-phase solution. Two or three
drops of each solution were placed on a thallium bromoiodide
(KRS-5) optical crystal, which was immediately rotated on a
 978 L. A. Wegiel et al.
KW-4A spin coater at 500/2500 rpm for 18 and 30 s, respectively,
to produce a thin film.
Rotary evaporation
Solutions were prepared by weighing out curcumin and polymer
at a dry weight ratio of 3:1, then dissolving the mixture in a 1:1
(by weight) mixture of dichloromethane and ethanol. All mixtures
were visually inspected to confirm that the curcumin and the
polymers were fully dissolved, forming uniform one-phase
solutions. The solvent was removed using a rotary evaporator
(Brinkman Instruments, Westbury, NY) with a water bath
maintained at 60  C. The samples were then placed under
vacuum for 24 h at room temperature to remove any residual
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Maastricht on 06/09/14
solvents. The obtained material was subsequently cryomilled in a
liquid nitrogen bath as described below to obtain a uniform
powder sample.
Cryomilling
The rotary evaporated samples were added to the cryomill in a
liquid nitrogen bath and milled for 2 min at a rate of two strikes
per second then cooled for 2 min. This process was repeated once.
For the dry powdered curcumin, for which the goal was to obtain
an amorphous material, the sample was added to the cryomill in a
liquid nitrogen bath. The sample was milled for 2 min at a rate of
10 strikes per second then cooled for 2 min; this was repeated for
30 cycles (1 h of total milling time).
IR spectroscopy
For personal use only.
IR spectra of the spin-coated thin films were obtained in
absorbance mode using a Bio-Rad FTS 6000 spectrophotometer
(Bio-Rad Laboratories, Hercules, CA) equipped with globar IR
source, KBr beamsplitter, and DTGS detector. The scan range was
set from 4000 to 500 cm–1 with 4 cm–1 resolution, and 128 scans
were co-added. The absorbance intensity of the spectral region of
interest was between 0.6 and 1.2. During measurements, the spin-
coated samples, the sample and optical compartments of the
spectrophotometer were flushed with dry air23.
Storage treatments
Amorphous curcumin and dispersions (25% polymer and 75%
curcumin) were stored at six different environmental conditions
for up to 350 d and sampled periodically. The storage conditions
were: room temperature (22–25  C) in a desiccator with P2O5 (0%
RH), room temperature (22–25  C) in a desiccator with potassium
chloride (84% RH), 40  C in a desiccator with sodium chloride
(75% RH), 60  C in a desiccator with potassium sulfate (98% RH)
as well as 50  C and 90  C in desiccators with DrieriteÔ (W.A.
Hammond Drierite Company, Ltd., Xenia, OH) (anhydrous
calcium sulfate, 0% RH).
Powder X-ray diffraction
Immediately after cryomilling, the samples were analyzed using a
Shimadzu XRD-6000 diffractometer (Shimadzu Corporation,
Kyoto, Japan) equipped with a Cu-Ka source and set in Bragg-
Brentano geometry, scanning between 5 and 35 2 at 8 /min with
a 0.04 step size. Samples were also analyzed as a function of
time following the storage treatments listed above. Before each
day of measurement, the accuracy of the 2 angle was checked by
verifying that the [111] peak of a Si-standard sample was between
28.423 and 28.463 . The measured photon intensity of the [111]
peak was used to normalize the data collected from samples on
that day. Samples where sharp peaks were observed were deemed
crystalline, and those samples that lacked sharp peaks were
Pharm Dev Technol, 2014; 19(8): 976–986
labeled X-ray amorphous. The detection limit was found to be
about 5–10 wt% crystalline material by adding small amounts
crystalline material to a solid dispersion until the powder X-ray
diffraction (PXRD) showed visible peaks in the diffractogram.
Differential scanning calorimetry
Thermal analysis was performed on the samples prepared by
rotary evaporation using a TA Q2000 differential scanning
calorimeter (DSC) equipped with a refrigerated cooling accessory
(TA Instruments, New Castle, DE). Nitrogen, 50 mL/min, served
as the purge gas. Samples (3–7 mg) were weighed into aluminum
T-zero sample pans with pin holes (TA instruments) and sealed.
Samples were heated from 0  C to 120  C at 10  C/min, quickly
cooled to 0  C at the maximum instrument cooling rate (15  C/
min), then heated from 0  C to 200  C at 10  C/min; transitions
are reported from this second heating scan. For the PVP solid
dispersions, the Tg could not be resolved without modulation.
These samples were heated from 25  C to 120  C at 10  C/min
then quickly cooled to 0  C at the maximum cooling rate of the
instrument, held isothermally at 0  C for 5 min, then heated at
2  C/min to 200  C with a modulation of  1  C every 60 s. DSC
heating curves were analyzed using Universal Analysis 2000
software (TA Instruments).
Ultraviolet-visible spectra
Effects of formulation and storage treatments on curcumin content
of solid dispersions prepared by the rotary evaporation technique
were determined using an ultraviolet-visible spectrometer
(Beckman DU800 Spectrophotometer, Fullerton, CA) at max
424 nm using a method adapted from Jasim and Ali24. Curcumin
concentration in the sample was calculated using a standard curve
constructed over the concentration range of interest. The results
were treated statistically using SPSS software version 17 (SPSS
Inc., Chicago, IL), and data were expressed as means  standard
deviations. One-way analysis of variance with post hoc Tukey
(HSD) test was employed to identify significant differences
(p50.05) between data sets.
Results
Curcumin is an intermediate crystallizer according to the
classification system of Van Eerdenbrugh et al.19. Therefore, it
was possible to make the compound amorphous by spin coating,
but a larger amorphous sample could not be prepared via rotary
evaporation. However, cryomilling enabled preparation of a larger
mass of amorphous powder.
IR spectroscopic investigation of curcumin–polymer
interactions
Drug–polymer interactions in solid dispersions can be explored
using IR spectroscopy23, whereby the presence of specific
interactions is often used to infer miscibility25,26. The IR spectrum
of a miscible solid dispersion typically contains peaks that are
shifted relative to those found in the spectrum of the physical
mixture of amorphous drug and polymer. This is due to specific
interactions between the chemical entities of the drug and
polymer, usually hydrogen bonding. If these interactions differ
in strength or extent relative to self-associations in these systems,
then there are changes in IR peak positions and shape for the
functional groups involved.
In order to understand the origin of any spectroscopic changes,
it is first important to consider the inter- and intra-molecular
interactions formed in pure curcumin. Curcumin exhibits keto–
enol tautomerism: the di-keto form is shown in Figure 1a, and the
keto–enol form is shown in Figure 1b. Single crystal studies have
 DOI: 10.3109/10837450.2013.846374 Curcumin solid dispersions – role of interactions on physical stability 979
Figure 3. IR spectra of amorphous and
crystalline curcumin.
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Maastricht on 06/09/14

shown that the keto–enol form is favored in all crystal forms27,28. Table 1. OH peak position in the IR spectra of amorphous curcumin and
The keto–enol tautomer is further stabilized in the most stable curcumin–polymer dispersions at 80% w/w curcumin.
polymorphic form by a resonance-assisted intramolecular hydro-
For personal use only.

gen bond (Figure 1c) where the hydrogen atom is symmetrically Sample OH peak position (cm–1)
positioned between the two oxygen atoms27. Formation of an Amorphous curcumin 3400
intramolecular hydrogen bond leads to increased delocalization of HPMC–curcumin 3386
the -conjugate electrons as depicted in Figure 1c29. Figure 3 HPMCAS–curcumin 3394
shows the IR spectrum of crystalline curcumin (stable polymorph CMCAB–curcumin 3399
and monoclinic form) whose structure is known to be similar to E100–curcumin 3396
PVP–curcumin 3187
that depicted in Figure 1c27. As expected, the spectrum is missing
the strong carbonyl stretching band that would be anticipated for
the di-keto form at around 1700 cm–1. Rather, there is a prominent
stretching band at 1626 cm–1 consistent with formation of a keto– Table 2. C¼O stretching vibration peak positions in IR spectra of the
enol tautomer stabilized by a strong intramolecular hydrogen model polymers, and observed shifts in curcumin–polymer dispersions at
bond. The spectrum of amorphous curcumin (Figure 3) is very 80% w/w curcumin.
similar to that of crystalline curcumin with regard to the pattern of
peaks in this spectral region, indicating that the structure of C¼O peak C¼O peak
position in position with 80% Shift
curcumin is predominantly present as the intramolecular hydrogen
Sample polymer (cm–1) curcumin (cm–1) observed (cm–1)
bonded keto–enol tautomer in the amorphous form. This is to be
expected since ab initio calculations indicated that the keto–enol PVP 1682 1655 27
tautomers are lower energy than the corresponding di-keto form30. HPMC N/A N/A N/A
HPMCAS 1743 1735 8
Strong intramolecular hydrogen bonding makes the keto–enol
CMCAB 1749 1745 4
group less available for intermolecular hydrogen bonding. E100 1730 1728 2
Curcumin also has two phenol groups that have the potential for
intermolecular hydrogen bonding. However, in crystalline curcu-
min, intramolecular hydrogen bonds are formed between the wavenumbers would be expected upon formation of these
phenolic groups and methoxy substituents, oriented ortho to one interactions31. The C¼O stretching vibrations assigned to the
another on the phenyl ring27. Therefore, it might be anticipated carbonyls for the relevant polymers are listed in Table 2, along
that curcumin hydrogen bonding with a polymer will be weak due with the solid dispersion shifts observed in the curcumin–polymer
to limited possibilities for intermolecular hydrogen bonding. dispersion at 80% w/w curcumin (larger shift indicates stronger/
Changes in the OH region can be used to provide information more extensive hydrogen bonding).
about curcumin–polymer intermolecular hydrogen-bonding inter- For all of the solid dispersions of curcumin with cellulosic
actions. The OH peak positions of amorphous curcumin and polymers (HPMC, HPMCAS or CMCAB), curcumin–polymer
curcumin–polymer dispersions at 80% w/w curcumin (lower interactions were observed but they appeared to be minor
wavenumber indicates stronger hydrogen bonding) are listed in (Figures 4–6). For example, in the presence of 20% polymer,
Table 1. In addition, all of the polymers, except HPMC, have the OH peak occurred at 3386 cm–1, 3394 cm–1 and 3399 cm–1 for
C¼O groups. The carbonyl group of the polymer can potentially HPMC, HPMCAS and CMCAB, respectively, compared to
hydrogen bond with the phenolic OH groups in curcumin, and 3400 cm–1 for the pure amorphous material (Table 1). Based on
resultant shifts of the carbonyl stretching frequency to lower the relative extent of the OH peak shift, it can be inferred that
 980 L. A. Wegiel et al.
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Maastricht on 06/09/14
Figure 4. IR spectra of curcumin–CMCAB solid dispersions showing (a)
the OH stretching region (3600–2800 cm–1) and (b) the carbonyl
stretching region (1775–1580 cm–1). Percentage indicates the weight
For personal use only.
percent of curcumin in the solid dispersion.
Figure 5. IR spectra of curcumin–HPMCAS solid dispersions showing
(a) the OH stretching region (3700 cm–1–2700 cm–1) and (b) the carbonyl
stretching region (1800–1580 cm–1). Percentage indicates the weight
percent of curcumin in the solid dispersion.
curcumin formed the weakest hydrogen bonds with CMCAB and
the strongest with HPMC with regard to the cellulosic polymers.
In the carbonyl region, where the polymer acceptor group
provides additional information about hydrogen bonding, it was
Pharm Dev Technol, 2014; 19(8): 976–986
Figure 6. IR spectra of curcumin–HPMC solid dispersions showing the
OH stretching region (3600–2700 cm–1). Percentage indicates the weight
percent of curcumin in the solid dispersion.
Figure 7. IR spectra of curcumin–PVP solid dispersions showing (a) the
OH stretching region (3600–2650 cm–1) and (b) the carbonyl stretching
region (1720–1480 cm–1). Percentage indicates weight % curcumin in
solid dispersion.
confirmed that CMCAB has a weaker hydrogen bonding inter-
action with curcumin based on the observation that the polymer
carbonyl peak shifts only by about 4 cm–1 in the presence of
curcumin, versus an 8 cm–1 wavenumber shift with HPMCAS
(HPMC does not have a carbonyl group). Overall the IR
spectroscopy results indicate weak hydrogen bonding between
the cellulosic polymers and curcumin.
PVP contains an amide carbonyl on the pyrrolidone monomer,
which is a strong hydrogen bond acceptor, thus stronger
interactions would be anticipated relative to the cellulosic
polymers, which have with their weaker acceptor groups22.
Stronger interactions were indeed observed for the curcumin–
PVP system, based on the IR data. In the hydroxyl region of the
IR spectrum (Figure 7a), at 80% curcumin, the curcumin OH peak
had a maximum at around 3187 cm–1. This is considerably lower
than the peak maximum in pure curcumin (3400 cm–1) and is the
lowest value seen for this peak out of all of the curcumin–polymer
dispersions studied. This suggests that the hydrogen bonds formed
in this system were the strongest/most extensive. In the carbonyl
spectral region (Figure 7b), a new shoulder at 1655 cm–1 emerges
due to the presence of a hydrogen bonded PVP carbonyl. This is a
 DOI: 10.3109/10837450.2013.846374 Curcumin solid dispersions – role of interactions on physical stability
wavenumber shift of 27 cm–1 relative to the non-hydrogen bonded
carbonyl in pure PVP (1682 cm–1) and confirms curcumin–
polymer hydrogen bonding. As expected, the ratio of the free
carbonyl peak to the hydrogen bonded carbonyl peak increased as
the concentration of the polymer increased. The large shift in the
carbonyl peak relative to that observed for the cellulosic polymers
provides further evidence that the strongest hydrogen bonds are
formed in the PVP–curcumin dispersion.
In the case of the E100-curcumin system, the curcumin OH
peak occurred at 3396 cm–1 in the presence of 20% polymer
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Maastricht on 06/09/14
For personal use only.
Figure 8. IR spectra of curcumin-E100 solid dispersions showing: (a) OH
stretching region (3600–2700 cm–1) and (b) carbonyl stretching region
(1770–1450 cm–1). Percentage indicates weight % curcumin in solid
dispersion.
Figure 9. Images of curcumin solid dispersions with 75% curcumin and 25% polymer. (a) Curcumin, (b) curcumin–E100, (c) curcumin–PVP, (d)
curcumin–CMCAB, (e) curcumin–HPMCAS and (f) curcumin–HPMC.
981
(Figure 8a). Thus, there is only about a 4 cm–1 shift relative to
amorphous curcumin, indicating that interspecies hydrogen
bonding is limited or weak. Consistent with this interpretation,
little to no evidence of hydrogen bonding can be inferred from the
carbonyl peak position where only a 2 cm–1 wavenumber shift is
observed. This is the smallest shift of all the polymers studied.
However, E100 contains dimethylamino groups, which can also
potentially interact with curcumin through ionic interactions
(Figure 8b). In the E100 spectrum, there are two peaks at
2820 cm–1 and 2770 cm–1 that are reported to correspond to the
non-ionized dimethylamino groups32. As the concentration of
curcumin was increased, these peaks quickly decreased in
intensity and were fully eradicated when 60% curcumin was
present in the dispersion. The disappearance of these peaks
provides evidence for an acid–base interaction32 between the
acidic phenol hydroxyls of curcumin and the amino groups of
E100, as it indicates the presence of ionized dimethylamino
groups. This acid–base interaction potentially can lead to an
increase in physical stability33–35 and will be discussed later.
Crystallization tendency of curcumin solid dispersions
Storage stability was tested at 0% RH (25  C, 50  C and 90  C),
84% RH (25  C), 75% RH (40  C) and 98% RH (60  C).
Interestingly, the nature of the dispersion polymer had an
impact on color (Figure 9). The E100 solid dispersions were
rich dark red in color, while the other solid dispersions all had an
orange color similar to that of pure curcumin. Chemical
degradation did not appear to be the cause of the color change,
as the curcumin-E100 dispersions did not show a decreased level
of curcumin content compared to the other polymer dispersions.
In fact, all of the dispersions were found to be chemically stable
following storage for up to 200 d for all stability conditions apart
from the most extreme storage condition of 98% RH (60  C)
(Figure 10). Since the latter system was physically unstable and
crystallized after 1 d, its chemical instability is not of relevance to
this study. Pure crystalline curcumin did not show evidence of
degradation at any environmental condition and proved to be
chemically stable over the time frame of this study. The color
 982 L. A. Wegiel et al. Pharm Dev Technol, 2014; 19(8): 976–986
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Maastricht on 06/09/14

Figure 11. PXRD of amorphous curcumin versus storage time at 50  C,

0% RH.
Figure 10. Chemical stability of curcumin-E100 solid dispersion (75%
curcumin w/w) versus storage temperature, relative humidity.
between the polymers could be made. At all other storage
Table 3. Long-term physical stability of amorphous curcumin-polymer conditions, the physical stability of the solid dispersions was
dispersion at 75% w/w curcumin by PXRD (days before crystalline highly dependent on the nature of the matrix polymer.
material is observed).
The poorest crystallization inhibitor was PVP. At 50  C, the
PVP solid dispersion started to crystallize after 6 d, as compared
Solid Tg 0% RH 0% RH 0% RH 84% RH 75 % RH 98% RH
dispersions ( C) 25  C 50  C 90  C 25  C 40  C 60  C
to amorphous curcumin, which exhibited crystallization at day 3.
In addition, at high RH conditions (84% RH) and room
Cur amorphous 43.6 4100 3 1 14 1 1 temperature, crystallization occurred by 22 d for the PVP
E100 62.4 4350 4350 1 4350 4200 1 dispersion versus 10 d for amorphous curcumin alone. At 40  C
PVP 62.8 4350 6 1 22 16 1
HPMC 68.3 4350 4350 1 75a 100a 1b
(75% RH), crystallization of the PVP solid dispersion was
HPMCAS 65.3 4350 75a 1b 75a 30a 1b inhibited for 16 d versus amorphous curcumin, which crystallized
CMCAB 61.7 4350 75a 1 75a 30a 1b after only 1 d. Overall, PVP inhibited crystallization to some
extent; however, the PVP system crystallized faster than disper-
For personal use only.

4X indicates monitoring was stopped after this time point.
a
System only partially crystallized.
b
Crystallized into a metastable polymorph.
change observed in the presence of E100 can be attributed to a
change in the ionization state of curcumin. E100 is a basic
polymer and therefore can cause ionization of curcumin if an
acid–base reaction occurs. It is well known that curcumin changes
color as a function of pH and has a darker color at high pH36. This
observation is in good agreement with the inferences made from
the IR analysis.
Curcumin-polymer solid dispersion physical stability as a
function of time was found to depend greatly on the polymer used
to form the dispersion, as well as on the storage conditions
(Table 3). The difference in crystallization tendency could not be
correlated with the glass transition temperature (Tg) as all of the
dry solid dispersions had Tg values close to 65  C (Table 3), while
pure amorphous curcumin has a Tg of 44  C. Thus, although the
Tg in the solid dispersions is increased relative to curcumin alone,
the significant differences in physical stability of the solid
dispersions containing different polymers could not be explained
by small differences in Tg values. In addition, there was no trend
with pure polymer Tg values as E100 has a Tg of 41  C, and
CMCAB has a Tg of 137  C.
Neat amorphous curcumin was used as a negative control and
was found to crystallize at all storage conditions (Figure 11), apart
from 0% RH and 25  C, where it was physically stable for more
than 100 d. The crystal form was the form I polymorph, listed in
the Cambridge Crystallographic Data Center under reference code
BINMEQ, with characteristic reflections at 8.9, 17.4, 23.5 and
25.1 2. It is thus unsurprising that all of the solid dispersions
were also physically stable at this storage condition, with no
crystallinity detected for storage periods approaching 1 year. At
the harshest storage conditions of 60  C (98% RH) and 90  C (0%
RH), all of the samples crystallized within a day. Thus for the
mildest and harshest crystallization conditions, no differentiation
sions containing any of the other polymers at comparable storage
conditions.
The dispersions containing E100 were the most stable against
crystallization at the intermediate temperature and RH storage
conditions. The E100 dispersion was stable at 0% RH (25  C and
50  C) and was the only dispersion stable at both 84% RH (25  C)
and 75% RH (40  C). For the latter conditions, the dispersions
were stable to crystallization for up to 350 d and 200 d,
respectively. E100 was unable to prevent crystallization at the
harsh conditions of 60  C (98% RH) and 90  C (0% RH) where
complete crystallization was seen after 1 d. However, this system
was X-ray amorphous for considerably longer time than any of the
other dispersions stored at the same condition, showing much
better physical stability.
The cellulosic polymers provided some enhancement of
physical stability, with HPMC proving to be the most effective
cellulosic crystallization inhibitor. However, the curcumin–
HPMC dispersion only had long-term stability at 0% RH (25  C
and 50  C). Crystallization occurred after 75 d for the sample at
84% RH (25  C) and after 100 d when stored at 75% RH (40  C).
At all other storage conditions, the sample showed evidence of
crystalline material after 1 d. Interestingly, complete crystalliza-
tion was never reached at some of the storage conditions. The
system instead underwent partial crystallization, and then no
additional crystal growth was observed over the time scale of the
study (Figure 12). The two other cellulosic polymers (CMCAB
and HPMCAS) were similar in their ability to inhibit crystalliza-
tion at a given storage condition. Both systems were stable at 0%
RH (25  C) for up to 350 d; however, after 75 d at 0% RH (50  C)
crystalline material was observed. In addition, it only took 30 d
and 75 d to observe crystallization at 75% RH (40  C) and 84%
RH (25  C), respectively. Again, complete crystallization was not
observed at these intermediate storage conditions. In addition, for
dispersions prepared with the cellulosic polymers, a different
polymorphic form of curcumin was found to crystallize from the
polymeric dispersion following storage at some conditions
 DOI: 10.3109/10837450.2013.846374 Curcumin solid dispersions – role of interactions on physical stability
(Figure 13). At 98% RH (60  C), curcumin crystallized to form II
from all of the cellulose dispersions. This polymorph (CCDC
reference code of BINMEQ), could be identified from the
characteristic reflections at 13.9, 26.0 and 27.1 2 (the 26.0
and 27.1 2 peaks are not resolved by our instrument). In
addition, form II was seen in the HPMCAS dispersion following
storage at 0% RH (90  C). Furthermore, form II did not convert to
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Maastricht on 06/09/14
Figure 12. PXRD of curcumin–CMCAB solid dispersion (75% curcumin
w/w) after 250 d at different conditions (only partial crystallization
observed).
For personal use only.
Figure 13. PXRD of curcumin–polymer solid dispersions (98% RH,
60  C and 1 d). The form II polymorph could be identified by the presence
of the characteristic peaks between 26.0 and 27.1 2 and the absence of
the form I peak at 17.4 2.
Figure 14. IR spectra of the carbonyl stretching region (1710 cm–1–1630 cm–1) of (a) curcumin–PVP solid dispersion and (b) resveratrol–PVP solid
dispersion23. Percentage indicates the weight percent of polyphenol in the solid dispersion.
983
the more stable polymorphic form over the time frame of the
study.
Discussion
Curcumin has a relatively high inherent crystallization tendency,
so it is difficult to prepare pure, amorphous curcumin using
methods such as rapid solvent evaporation. When made amorph-
ous using cryomilling, it remains amorphous only when stored at
room temperature and extremely low RH, but crystallizes rapidly
when exposed to higher humidity or temperature. The rapid
crystallization behavior and the extensive intramolecular hydro-
gen bonds present in both crystalline and amorphous curcumin
make this an interesting model compound with which to evaluate
the crystallization inhibition ability of different polymers. This is
because the presence of extensive intramolecular hydrogen
bonding in curcumin might be expected to hinder curcumin–
polymer interactions in the solid dispersions that are thought to be
important for inhibition of crystallization20–23.
Prior to 2011, only one polymorph of curcumin had been
reported. However, recently, Sanphui et al.28 reported two new
polymorphs of curcumin, named forms II and III, both of which
are metastable with respect to the original form I polymorph. In
form I, curcumin is present as a non-planar molecule with the
benzene rings at the end of the heptane chain slightly twisted with
respect to one another, but curcumin is planar in forms II and III
(Figure 14)28. Extensive intramolecular hydrogen bonding persists
in all polymorphic forms of curcumin, suggesting that such H-
bonding is energetically favorable. The keto–enol tautomer forms
an intramolecular hydrogen bond in each of the crystal structures.
In form I, the hydrogen is shared equally between the two oxygen
atoms, while in forms II and III, the hydrogen atom is not centered
and favors one oxygen over the other28. All forms also have strong
intramolecular hydrogen bonds between the ortho methoxy and
hydroxyl benzene substituents27. Form I has intermolecular
hydrogen bonds formed between the phenolic OH and the enol
OH group. There are also a number of much weaker CHO
intermolecular interactions. Forms II and III have more extensive
intermolecular hydrogen bonding but are less stable than Form I.
Considering the strong and extensive intramolecular inter-
actions in curcumin, it is reasonable to expect that polymers that
interact only through hydrogen bonds might be poor curcumin
crystallization inhibitors. This supposition arises from the
 984 L. A. Wegiel et al.
expectation that there will be less opportunity for strong/extensive
hydrogen bonding after intramolecular interactions are accounted
for. Furthermore, polymers that contain only hydrogen bond
acceptors, such as PVP, suffer the additional disadvantage of only
being able to interact with a donor group on curcumin. In contrast,
polymers that contain both donor and acceptor groups, such as the
cellulosic polymers, may gain an advantage by forming hydrogen
bonds with both curcumin donors and acceptor moieties, and may
thus be able to better disrupt curcumin self-associations important
to crystal formation. This may well explain why curcumin–PVP
was the least stable of all the dispersions against crystallization.
This result is in direct contrast to the PVP-resveratrol system,
where PVP was found to be an extremely effective crystallization
inhibitor23. Indeed, comparison of curcumin–PVP and resvera-
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Maastricht on 06/09/14
trol–PVP IR spectra shows that much less hydrogen bonding
occurs between PVP and the polyphenol in the case of curcumin
(Figure 11). Given the similar phenol H-bonding groups in
resveratrol and curcumin, it can be supposed that the intramo-
lecular hydrogen bond formed by curcumin is indeed the critical
factor in diminishing its ability to interact with PVP. The
cellulosic polymers are slightly better crystallization inhibitors for
curcumin, presumably because they can form both donor and
acceptor hydrogen bond interactions with curcumin. However,
these interactions appear to be fairly weak based on the IR
spectra; hence the cellulosic polymers fail to prevent curcumin
crystallization under challenging humidity and temperature con-
ditions. Therefore, it appears that for a compound like curcumin,
which forms intramolecular hydrogen bonds, hydrogen bonding
with polymers is impeded and it is more difficult to inhibit
crystallization. A similar observation was made with nimesulide
For personal use only.
where it was seen that there was little differentiation between
chemically dissimilar polymers in terms of their ability to reduce
the crystal growth rate of this compound because of the inability
of the polymer to compete with intramolecular interactions21.
These observations are also in line with the work of Etter37,38,
whereby it is suggested that intramolecular hydrogen bonds that
complete a six-membered ring structure are typically favored over
intermolecular hydrogen bonds in crystals.
Given the reluctance of curcumin to form intermolecular
hydrogen bonds, it is of interest to explore the origin of the
extremely good resistance of the E100-curcumin dispersions to
crystallization. E100 is a polybase and proton acceptor. The
average pKa of the basic monomer is 8.439. Curcumin has pKa
values of 10.51, 9.88 and 8.3840. Based on the IR data, it is
apparent that E100 was able to form ionic interactions with
curcumin. Such interactions with E100 have been seen with a
number of pharmaceutical compounds, including resveratrol23,
dexamethasone phosphate41, and indomethacin42. There are a
number of studies in the literature that have investigated
miscibility and physical stability of solid dispersion systems,
which have ionic interactions. Weuts et al.34 showed salt
formation between a basic drug (loperamide) and an acidic
polymer (poly(acrylic acid); PAA), and correlated the salt
formation with good miscibility and physical stability. Miyazaki
et al.33 reported the inhibition of drug recrystallization by proton
transfer between the acid functional groups of the polymer PAA
and basic functional groups of the drug (p-aminoacetanilide).
In addition, Yoo et al.35 reported that ionic interactions play
a significant role for both drug–polymer miscibility and drug
physical stability in that systems with ionic interactions were
stable for up to 50 d at 60% RH and 25  C, while other non-ionic
systems tended to recrystallize. These reports demonstrate the
significant influence of acid–base ionic interactions on the
physical stability of amorphous solid dispersions. The results
from this study are consistent with these previous reports in that
ionic drug–polymer interactions appear to be the reason for the
Pharm Dev Technol, 2014; 19(8): 976–986
Table 4. The onset glass transition temperatures of polymers and
polymer–curcumin solid dispersions at 25% drug loading.
Solid Polymer Calculated Observed Deviation
dispersions Tg ( C) Tg ( C) Tg ( C) ( C)
Cur amorphous 44
E100 41 42 62 20
PVP 159 96 63 –33
HPMC 145 92 68 –24
HPMCAS 111 80 65 –15
CMCAB 137 89 62 –27
Calculated Tg values were based on the Fox equation, and deviations are
taken as the difference between the calculated and observed values.
extremely good physical stability of curcumin–E100 dispersions
to crystallization, since little evidence of curcumin–polymer
hydrogen bonding was found in this system.
It also important to consider the Tg values of the dispersions
investigated. It has been argued that the dispersions with the
highest Tg will be the most stable43. However, all of the dispersion
Tg values are around 62–68  C (Table 4). Certainly, Tg does not
explain the higher stability of the E100 and the lower stability of
the PVP dispersions against crystallization relative to the other
polymers, since all the dispersions have similar Tg values. It has
been suggested that deviations in Tg predictions reflect the nature
and extent of interactions formed between drug and the
additive44–46. The Fox equation is one of the simplest models to
predict the Tg. In this model, the Tg of a system comprised of two
amorphous components is calculated from the weight-averaged
contribution of each component47:
1 w1 w2
¼ þ : ð1Þ
Tg, mix Tg1 Tg2
where wi and Tgi are the weight fractions and glass transition of
component i. This model is based on the free-volume theory. In
the presence of specific interactions, such as hydrogen bonding or
ion-dipole or ionic interactions, the mathematical prediction of Tg
for mixed systems are no longer valid, since the assumptions of
free-volume additivity and ideal mixing are likely to have been
violated. For example, in the presence of strong specific
interactions between the different components, the free volume
of the mixed system can be expected to be smaller than the
predicted free volume. This would in turn lead to positive
deviations of Tg. Conversely, the net loss of specific interactions
would lead to a negative deviation of Tg. Table 4 lists the
calculated and observed Tg as calculated by the Fox equation.
Interestingly it can be seen that the E100 dispersion has a higher
Tg than that observed for either the pure polymer or pure
curcumin. This leads to a large deviation from the Fox equation
and is consistent with the formation of ionic interaction observed
between curcumin and E100. However, for all the other disper-
sions, negative deviations are observed, with PVP showing the
largest negative deviation. This may suggest a net loss of
interactions for these systems resulting from hydrogen bonding of
curcumin with the polymer. Similar results have been observed for
other systems showing hydrogen bonding interactions48. Clearly,
Table 4 shows that it is hard to predict the stability of the system
to crystallization based on consideration of Tg values alone.
Interestingly, the cellulosic polymers were able to influence
the crystalline form of curcumin that recrystallizes from the solid
dispersion. The PXRD patterns obtained following crystallization
of curcumin from dispersions containing cellulose derivatives
match that of the polymorph form II reported by Sanphui et al.28.
This metastable polymorph has a higher dissolution rate
and increased solubility relative to the stable polymorph28.
 DOI: 10.3109/10837450.2013.846374 Curcumin solid dispersions – role of interactions on physical stability
Clearly, recrystallization to any form would be undesirable, but it
interesting that this uncommon polymorph of curcumin appears
upon crystallization of some solid dispersion formulations.
Therefore, even if the amorphous phase cannot be stabilized for
a pharmaceutically relevant time scale, there might still be
benefits from formulating with a cellulosic polymer that could
influence and stabilize the polymorphic form II, thus leading to a
possible increase in bioavailability. However, this clearly would
not be ideal from a regulatory perspective.
Conclusions
By studying the crystallization behavior of curcumin, additional
insight into the role of compound–polymer interaction on
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Maastricht on 06/09/14
crystallization tendency has been gained. Intramolecular bonding
in curcumin reduced the extent of hydrogen bonding between the
polyphenol and polymers. Thus, for a compound like curcumin,
hydrogen bonding with polymers is impeded, and it is more
difficult to inhibit crystallization. In contrast, ionic interactions
with E100 were not hindered, and when formed, resulted in
increased physical stability. This suggests that the best polymers
for amorphous solid dispersion of compounds prone to intramo-
lecular hydrogen bonding are those that allow the formation of
interactions other then hydrogen bonds (e.g. ionic interactions).
Acknowledgements
We gratefully acknowledge Alicia Winters for her help in preparing
samples and conducting PXRD and IR measurements.
For personal use only.
Declaration of interest
The authors report no declarations of interest.
We thank the USDA (Grant number 09-35603-05068) and Purdue
Research Foundation for financial support. We thank Eastman Chemical
Company and Shin-Etsu Ltd. for their gracious donations of CMCAB and
HPMCAS, respectively.
References
1. Sharma OP. Antioxidant activity of curcumin and related com-
pounds. Biochem Pharmacol 1976;25:1811–1812.
2. Motterlini R, Foresti R, Bassi R, Green CJ. Curcumin, an
antioxidant and anti-inflammatory agent, induces heme oxygenase-
1 and protects endothelial cells against oxidative stress. Free Radical
Biol Med 2000;28:1303–1312.
3. Apisariyakul A, Vanittanakom N, Buddhasukh D. Antifungal
activity of turmeric oil extracted from Curcuma longa
(Zingiberaceae). J Ethnopharmacol 1995;49:163–169.
4. Sadzuka Y, Nagamine M, Toyooka T, et al. Beneficial effects of
curcumin on antitumor activity and adverse reactions of doxorubi-
cin. Int J Pharm 2012;432:42–49.
5. Aggarwal BB, Kumar A, Bharti AC. Anticancer potential of
curcumin: preclinical and clinical studies. Anticancer Res 2003;
23:363–398.
6. Vlietinck AJ, De Bruyne T, Apers S, Pieters LA. Plant-derived
leading compounds for chemotherapy of human immunodeficiency
virus (HIV) infection. Planta Med 1998;64:97–109.
7. Sergent T, Vanderstraeten J, Winand J, et al. Phenolic compounds
and plant extracts as potential natural anti-obesity substances. Food
Chem 2012;135:68–73.
8. Miriyala S, Panchatcharam M, Rengarajulu P. Cardioprotective
effects of curcumin. Adv Exp Med Biol 2007;595:359–377.
9. Tonnesen HH, Masson M, Loftsson T. Studies of curcumin and
curcuminoids. XXVII. Cyclodextrin complexation: solubility, chem-
ical and photochemical stability. Int J Pharm 2002;244:127–135.
10. Wang Y-J, Pan M-H, Cheng A-L, et al. Stability of curcumin in
buffer solutions and characterization of its degradation products.
J Pharm Biomed Anal 1997;15:1867–1876.
11. Priyadarsini KI. Photophysics, photochemistry and photobiology of
curcumin: studies from organic solutions, bio-mimetics and living
cells. J Photochem Photobiol C-Photochem Rev 2009;10:81–95.
985
12. Anand P, Nair HB, Sung BK, et al. Design of curcumin-loaded
PLGA nanoparticles formulation with enhanced cellular uptake, and
increased bioactivity in vitro and superior bioavailability in vivo.
Biochem Pharmacol 2010;79:330–338.
13. Paradkar A, Ambike AA, Jadhav BK, Mahadik KR. Characterization
of curcumin-PVP solid dispersion obtained by spray drying. Int J
Pharm 2004;271:281–286.
14. Leuner C, Dressman J. Improving drug solubility for oral delivery
using solid dispersions. Eur J Pharm Biopharm 2000;50:47–60.
15. Serajuddin ATM. Solid dispersion of poorly water-soluble drugs:
early promises, subsequent problems, and recent breakthroughs.
J Pharm Sci 1999;88:1058–1066.
16. Rumondor ACF, Stanford LA, Taylor LS. Effects of polymer type
and storage relative humidity on the kinetics of felodipine crystal-
lization from amorphous solid dispersions. Pharm Res 2009;26:
2599–2606.
17. Rumondor ACF, Marsac PJ, Stanford LA, Taylor LS. Phase behavior
of poly(vinylpyrrolidone) containing amorphous solid dispersions in
the presence of moisture. Mol Pharmaceutics 2009;6:1492–1505.
18. Van Eerdenbrugh B, Taylor LS. Small scale screening to determine
the ability of different polymers to inhibit drug crystallization upon
rapid solvent evaporation. Mol Pharm 2010;7:1328–1337.
19. Van Eerdenbrugh B, Baird JA, Taylor LS. Crystallization tendency
of active pharmaceutical ingredients following rapid solvent evap-
oration – classification and comparison with crystallization tendency
from undercooled melts. J Pharm Sci 2010;99:3826–3838.
20. Kestur US, Taylor LS. Role of polymer chemistry in influencing
crystal growth rates from amorphous felodipine. Cryst Eng Comm
2010;12:2390–2397.
21. Kestur US, Van Eerdenbrugh B, Taylor LS. Influence of polymer
chemistry on crystal growth inhibition of two chemically diverse
organic molecules. Cryst Eng Comm 2011;13:6712–6718.
22. Van Eerdenbrugh B, Taylor LS. An ab initio polymer selection
methodology to prevent crystallization in amorphous solid disper-
sions by application of crystal engineering principles. Cryst Eng
Comm 2011;13:6171–6178.
23. Wegiel LA, Mauer LJ, Edgar KJ, Taylor LS. Crystallization of
amorphous solid dispersions of resveratrol during preparation and
storage – impact of different polymers. J Pharm Sci 2013;102:
171–184.
24. Jasim F, Ali F. A novel and rapid method for the spectrofluorometric
determination of curcumin in curcumin spices and flavors.
Microchem J 1992;46:209–214.
25. Marsac PJ, Shamblin SL, Taylor LS. Theoretical and practical
approaches for prediction of drug-polymer miscibility and solubility.
Pharm Res 2006;23:2417–2426.
26. Taylor LS, Zografi G. Spectroscopic characterization of interactions
between PVP and indomethacin in amorphous molecular disper-
sions. Pharm Res 1997;14:1691–1698.
27. Parimita SP, Ramshankar YV, Suresh S, Row TNG.
Redetermination of curcumin: (1E,4Z,6E)-5-hydroxy-1,7-
bis(4-hydroxy-3-methoxyphenyl)hepta-1,4,6-trien-3-one. Acta
Crystallographica Section E-Structure Reports Online 2007;63:
O860–O862.
28. Sanphui P, Goud NR, Khandavilli UBR, et al. New polymorphs of
curcumin. Chem Commun 2011;47:5013–5015.
29. Gilli G, Bellucci F, Ferretti V, Bertolasi V. Evidence for resonance-
assisted hydrogen-bonding from crystal-structure correlations on the
enol form of the beta-diketone fragment. J Am Chem Soc 1989;111:
1023–1028.
30. Benassi R, Ferrari E, Lazzari S, et al. Theoretical study on
curcumin: a comparison of calculated spectroscopic properties with
NMR, UV-vis and IR experimental data. J Mol Struct 2008;892:
168–176.
31. Tang XC, Pikal MJ, Taylor LS. A spectroscopic investigation of
hydrogen bond patterns in crystalline and amorphous phases in
dihydropyridine calcium channel blockers. Pharm Res 2002;19:
477–483.
32. Moustafine RI, Zaharov IM, Kemenova VA. Physicochemical
characterization and drug release properties of Eudragit E PO/
Eudragit L 100-55 interpolyelectrolyte complexes. Eur J Pharm
Biopharm 2006;63:26–36.
33. Miyazaki T, Yoshioka S, Aso Y. Physical stability of amorphous
acetanilide derivatives improved by polymer excipients. Chem
Pharmaceut Bull 2006;54:1207–1210.
 986 L. A. Wegiel et al.
34. Weuts I, Kempen D, Verreck G, et al. Salt formation in solid
dispersions consisting of polyacrylic acid as a carrier and three basic
model compounds resulting in very high glass transition tempera-
tures and constant dissolution properties upon storage. Eur J Pharm
Sci 2005;25:387–393.
35. Yoo SU, Krill SL, Wang Z, Telang C. Miscibility/stability consid-
erations in binary solid dispersion systems composed of functional
excipients towards the design of multi-component amorphous
systems. J Pharm Sci 2009;98:4711–4723.
36. Kuswandi B, Jayus TS, Larasati A, et al. Real-time monitoring of
shrimp spoilage using on-package sticker sensor based on natural
dye of curcumin. Food Anal Methods 2012;5:881–889.
37. Etter MC. Encoding and decoding hydrogen-bond patterns of
organic compounds. Acc Chem Res 1990;23:120–126.
38. Etter MC. Hydrogen bonds as design elements in organic chemistry.
J Phys Chem 1991;95:4601–4610.
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Maastricht on 06/09/14
39. Menjoge AR, Kulkarni MG. Mechanistic investigations of phase
behavior in Eudragit (R) blends. Int J Pharm 2007;343:106–121.
40. Bernabe-Pineda M, Ramirez-Silva MT, Romero-Romo M, et al.
Determination of acidity constants of curcumin in aqueous solution
and apparent rate constant of its decomposition. Spectrochim Acta A
Mol Biomol Spectrosc 2004;60:1091–1097.
For personal use only.
Pharm Dev Technol, 2014; 19(8): 976–986
41. Guzman ML, Manzo RH, Olivera ME. Eudragit E100 as a drug
carrier: the remarkable affinity of phosphate ester for dimethyla-
mine. Mol Pharm 2012;9:2424–2433.
42. Quinteros DA, Rigo VR, Kairuz AEJ, et al. Interaction between
a cationic polymethacrylate (Eudragit E100) and anionic drugs.
Eur J Pharm Sci 2008;33:72–79.
43. Van den Mooter G, Wuyts M, Blaton N, et al. Physical stabilization
of amorphous ketoconazole in solid dispersions with polyvinylpyr-
rolidone K25. Eur J Pharm Sci 2001;12:261–269.
44. Lu Q, Zografi G. Phase behavior of binary and ternary amorphous
mixtures containing indomethacin, citric acid, and PVP. Pharm Res
1998;15:1202–1206.
45. Tong P, Zografi G. Solid-state characteristics of amorphous
sodium indomethacin relative to its free acid. Pharm Res 1999;16:
1186–1192.
46. Tong P, Zografi G. A study of amorphous molecular dispersions of
indomethacin and its sodium salt. J Pharm Sci 2001;90:1991–2004.
47. Fox TG. Influence of diluent and of copolymer composition on
the glass temperature of a polymer system. Bull Am Phys Soc 1956;
1:123.
48. Shamblin SL, Taylor LS, Zografi G. Mixing behavior of colyophi-
lized binary systems. J Pharm Sci 1998;87:694–701.