Growth, shrinking, and breaking of pluronic micelles in the presence of drugs and/or β-cyclodextrin, a study by small-angle neutron scattering and fluorescence spectroscopy

Article abstract of DOI:10.1021/la100596q

The associative structures between F127 Pluronic micelles and four drugs, namely, lidocaine (LD), pentobarbital sodium salt (PB), sodium naproxen (NP), and sodium salicylate (SAL), were studied by small-angle neutron scattering (SANS). Different outcomes for the micellar aggregates are observed, which are dependent on the chemical nature of the drug and the presence of charge or otherwise: the micelles grow with LD, are hardly modified with PB, and decrease in size with both NP and SAL. The partition coefficient, determined by fluorescence spectroscopy, is directly correlated to the amount of charge, following NP ≈ SAL < PB < LD. All drugs are found to lie at the interfacial layer, with a slightly deeper localization of LD and more superficial for PB. All drugs can form inclusion complexes with heptakis(2,6-di-O-methyl) β-cyclodextrin (hep2,6 β-CD). Hep2,6 β-CD, as shown in previous studies (Joseph, J.; Dreiss, C. A.; Cosgrove, T. Langmuir, 2008, 24, 10005-10010; Dreiss, C. A.; Nwabunwanne, E.; Liu, R.; Brooks, N. J. Soft Matter, 2009, 5, 1888-1896), is also able to form a complex with F127, resulting in micellar breakup. In the ternary mixtures, a fine balance of forces is involved, which results in drastic micellar changes, as observed from the SANS patterns. Depending on the ratio of drug, polymer, and hep2,6 β-CD and the nature of the interactions (which is directly linked to the drug chemical structure), the presence of drug either hinders micellar breakup by β-CD (at high enough concentration of LD or PB) or leads to micellar growth (NP). These effects are mainly attributed to a preferential drug/β-CD interaction (except for PB), which, at least in the conditions studied here, explains the higher β-CD concentration needed for micellar breakup to occur.

Full text of DOI:10.1021/la100596q

pubs.acs.org/Langmuir
© 2010 American Chemical Society
Growth, Shrinking, and Breaking of Pluronic Micelles in the Presence
of Drugs and/or β-Cyclodextrin, a Study by Small-Angle
Neutron Scattering and Fluorescence Spectroscopy
†
,‡
ꢀ
Margarita Valero and Cecile A. Dreiss*
†
´
´
Dpto. Quımica Fısica, Facultad de Farmacia, Universidad de Salamanca, Campus Miguel de Unamuno,
s/n, 37007 Salamanca, Spain, and Pharmaceutical Science Division, King’s College London,
Franklin-Wilkins Building, 150 Stamford Street, SE1 9NH London, United Kingdom
‡
Received February 9, 2010. Revised Manuscript Received April 11, 2010
The associative structures between F127 Pluronic micelles and four drugs, namely, lidocaine (LD), pentobarbital
sodium salt (PB), sodium naproxen (NP), and sodium salicylate (SAL), were studied by small-angle neutron scattering
(SANS). Different outcomes for the micellar aggregates are observed, which are dependent on the chemical nature of the
drug and the presence of charge or otherwise: the micelles grow with LD, are hardly modified with PB, and decrease in
size with both NP and SAL. The partition coefficient, determined by fluorescence spectroscopy, is directly correlated to
the amount of charge, following NP ≈ SAL<PB<LD. All drugs are found to lie at the interfacial layer, with a slightly
deeper localization of LD and more superficial for PB. All drugs can form inclusion complexes with heptakis(2,6-di-
O-methyl) β-cyclodextrin (hep2,6 β-CD). Hep2,6 β-CD, as shown in previous studies (Joseph, J.; Dreiss, C. A.; Cosgrove,
T. Langmuir, 2008, 24, 10005-10010; Dreiss, C. A.; Nwabunwanne, E.; Liu, R.; Brooks, N. J. Soft Matter, 2009, 5,
1888-1896), is also able to form a complex with F127, resulting in micellar breakup. In the ternary mixtures, a fine
balance of forces is involved, which results in drastic micellar changes, as observed from the SANS patterns. Depending
on the ratio of drug, polymer, and hep2,6 β-CD and the nature of the interactions (which is directly linked to the drug
chemical structure), the presence of drug either hinders micellar breakup by β-CD (at high enough concentration of LD
or PB) or leads to micellar growth (NP). These effects are mainly attributed to a preferential drug/β-CD interaction
(except for PB), which, at least in the conditions studied here, explains the higher β-CD concentration needed for
micellar breakup to occur.
Introduction
Pluronics or Polaxamers, in particular, meet a number of essential
requirements for drug delivery;9-12 they are available commer-
cially in large quantities and at low cost with a range of polymer
architectures (block length, hydrophobic/hydrophilic ratio, mole-
cular weight, etc.), thus providing systematic variability of key
parameters and, quite critically, avoiding lengthy and costly
synthesis, with formulations being readily prepared most often
by simple dissolution of the drugs into the micellar core. Poly-
(ethylene) blocks provide protection against opsonization and
uptake by the macrophages of the reticuloendothelial system
(RES).13 In addition, they have been shown to improve the treat-
ment of tumors with the multidrug-resistant (MDR) phenotype
by affecting several mechanisms of drug resistance in cancer
cells,10 which points to their selectivity. Although nondegradable,
The poor solubility in biological fluids of nearly 50% of newly
approved drugs remains the main limitation in oral administra-
tion, resulting in low gastrointestinal absorption and bioavailabi-
lity.3 Among the existing strategies, polymeric micelles are ex-
tremely attractive candidates for solubilizing hydrophobic drugs,
increasing their stability against chemical degradation and meta-
bolism by biological agents, extending circulation times in the
blood,4 and improving their bioavailability.5-8 They present
similar features to natural carriers in terms of size, structure, and
transport properties and can selectively deliver drugs to specific
sites of action. Their similarity to biological carriers confers them
a fundamental property for pharmaceutical formulations, that is,
low toxicity. Selective targeting is another key feature, allowing
reduced dose, decreasing side effects and economic cost, an aspect
which is particularly important in the treatment of chronic disea-
ses. Triblockcopolymers ofpoly(ethylene oxide)-poly(propylene
oxide)-poly(ethylene oxide) (PEO-PPO-PEO), referred to as
molecules in the range 10-15000 kg mol-1 are filtered by the
3
kidney and cleared in urine14 and have been approved by regu-
latory institutions (FDA) for use in pharmaceutical formula-
tions.6 In addition, Pluronics solutions display a sol-gel transi-
tion which can be tuned to body temperature (e.g., by addition of
salts or chemical modification), making them attractive for the
design of injectable implants for minimally invasive applications.
*To whom correspondence should be addressed. E-mail: cecile.dreiss@
kcl.ac.uk.
(1) Joseph, J.; Dreiss, C. A.; Cosgrove, T. Langmuir 2008, 24, 10005–10010.
(2) Dreiss, C. A.; Nwabunwanne, E.; Liu, R.; Brooks, N. J. Soft Matter 2009, 5,
1888–1896.
(3) Chiappetta, D. A; Sosnik, A. Eur. J. Pharm. Biopharm. 2007, 66, 303–317.
(4) Kwon, G. S. Crit. Rev. Ther. Drug Carrier Syst. 2003, 20, 357–403.
(5) Lavasanifar, A.; Samuel, J.; Kwon, G. S. Adv. Drug Delivery Rev. 2002, 54,
(9) Alexandridis, P.; Hatton, T. A. Colloids Surf., A 1995, 96, 1–46.
(10) Kabanov, A. V.; Batrakova, E. V.; Alakhov, V. Y. J. Controlled Release
2002, 82, 189.
169.
(11) Kabanov, A. V.; Batrakova, E. V.; Alakhov, V. Y. Adv. Drug Delivery Rev.
2002, 54, 759–779.
(6) Adams, M. L.; Lavasanifar, A.; Kwon, G. S. J. Pharm. Sci. 2003, 92, 1343–
€
1355.
(12) Sezguin, A.; Yuksel, N.; Baykara, T. Eur. J. Pharm. Biopharm. 2006, 64,
261–268.
(7) Francis, M. F.; Cristea, M.; Winnick, F. M. Pure Appl. Chem. 2004, 76,
1321–1335.
(13) Barratt, G. Cell. Mol. Life Sci. 2003, 60, 21–37.
(14) Grindel, J. M.; Jaworski, T.; Piraner, O.; Emanuele, R. M.; Balasubramanian,
M. J. Pharm. Sci. 2002, 91, 1936–1947.
(8) Kataoka, K.; Harada, A.; Nagasaki, Y. Adv. Drug Delivery Rev. 2001, 47,
113–131.
Langmuir 2010, 26(13), 10561–10571
Published on Web 05/14/2010
DOI: 10.1021/la100596q 10561

Article
Despite the interest in polymeric micelles for delivery purposes
Valero and Dreiss
Scheme 1. Structure of Lidocaine, Pentobarbital, Naproxen, and
Salicylic Acid
and their already extensive use in existing formulations, there is
generally little rationale behind the formulations used due to a lack
of physical characterization of the nanostructures. An abundance
of studies has been published on drug solubilization in Pluronics
and their release characteristics.3,6 Drug loading efficiency appears
to be highly dependent on drug type and polymer structure,12 but
structure-property relationships to guide formulations have still
to be established. Indeed, very few studies have focused on
characterizing micellar shape and size and describing structural
changes occurring upon drug encapsulation, apart from a few
exceptions.2,15-17 Small organic molecules,18 and hence drugs,16,19
have been shown to modify the cmc, cmt, aggregate size and shape,
phase behavior (gelation boundaries), and stability of Pluronic
micelles. Rationalizing these changes is fundamental, since micellar
size, shape, and aggregation number are directly correlated to the
locus of drug solubilization, hence drug loading capacity, release
profile in vivo, circulation time, and therefore drug bioavailability.
In this contribution, we have selected four drugs, namely, the
anesthetics lidocaine and pentobarbital, the NSAID naproxen
(the latter two as sodium salts), and the analgesic sodium salicy-
late, having all related structures (Scheme 1). All these drugs are
well-known and with wide clinical applications. Sodium salicylate
(SAL) is the metabolite of aspirin, and both together are probably
the most widely used drugs in the world, presenting a wide range
of pharmacological actions, with the most known being analgesic,
anti-inflammatory, and antipyretic properties, besides stroke
prevention. However, the high incidence of gastrointestinal side
effects is also well-known and the main cause for avoiding its
prescription. Salicylic acid is used in cosmetic formulations as a
denaturant, hair-conditioning agent, and skin-conditioning agent
in a wide range of cosmetic products at concentrations ranging
from 0.0008% to 3%. More recently, β-hydroxy acids (including
sodium salicylate) have been included in cosmetics to reduce signs
of skin aging. Therefore, although a very well-know drug, it is still
the subject of intense research, for instance, regarding its role on
increasing skin’s sensitivity to ultraviolet radiation,20 and appro-
priate formulations and new drug carriers are still actively sought.21
Naproxen (NP) is another NSAID widely used in the treatment
of chronic inflammatory diseases and as a topical anti-inflammatory.
As most NSAIDs, naproxen also presents a high incidence of
gastrointestinal side effects. The correct formulation of NP for
avoiding gastrointestinal side effects, in addition to good dissolu-
tion rate, is a wide subject of interest.22 In addition, the photo-
reactivity of naproxen has been described,23 and how it could be
controlled by using appropriate drug carriers.24
with other drugs, among others anti-inflammatory drugs and
anesthetics,25 is currently being carried out.
In general, transdermal drug delivery has been shown in recent
years to be a very useful route of administration, as it is non-
invasive and avoids gastrointestinal absorption. This results in the
avoidance of first-pass metabolism in the liver, reduced gastro-
intestinal side effects, and no interaction with food. Therefore, it is
a very interesting alternative to oral or parenteral administration.
Transdermal administration strongly depends on high drug penet-
ration through the skin, which can be improved by the use of
electric fields, iontophoretic transdermal delivery. This system is a
novel drug delivery system that depends on formulation para-
meters such as drug content and pH,26 and opens a wide field of
study on the effect of drug carriers and additives. Overall, the
recentinterestintransdermal route emphasizesagainthe necessity
of better formulations of drugs for efficient and safe delivery; in
order to achieve this, the understanding of drug/carrier interac-
tions and microstructures formed is paramount.
In this contribution, we study the structural modifications
resulting from the addition of increasing amounts of each of these
four drugs to F127 Pluronic micelles (5 wt %) at room tempera-
ture and under natural conditions of pH. We have used a combi-
nation of small-angle neutron scattering (SANS) and UV-vis
fluorescence emission spectroscopy to examine the partitioning
of the drugs inside the micelles and rationalize it in terms of the
chemical structure and charge of the drug. While SANS provides
direct information on the modifications of the micelles upon
addition of the drug, fluorescence provides binding constants,
partition coefficients, and information on changes in the drug
microenvironment occurring in the presence of the polymer.
In addition to the need of a better characterization of
polymer-drug complexes, a substantial improvement to the
development of drug formulations is the design of triggers
to control the release and delivery of a payload at selected sites
of action. We have previously reported that a derivative of
β-cyclodextrin, namely, heptakis(2,6-di-O-methyl) β-cyclodextrin
(hep2,6 β-CD), is able to fully break up Pluronic micelles.1,2 This
interesting property is attributed to the specific complexation of
the central block PPO with β-CD, which, by imparting hydro-
philicity to the otherwise water-insoluble polymer, decreases the
critical micelle concentration (cmc). We propose that such a
mechanism could be developed and exploited to achieve the
controlled release of an active compound encapsulated in the
micellar core. Cyclodextrins, cyclic oligosaccharides with a hol-
low toroidal shape and a hydrophobic cavity, are well-known to
form inclusion complexes through noncovalent interactions with
Lidocaine (LD) is a local anesthetic used in the relief of pain,
which can be used in topical or parenteral administration. Lidocaine
and fentanyl are the first drugs approved by regulatory agencies
for iontophoretic transdermal application (active transdermal libe-
ration) and research to exploit this delivery system in combination
(15) Sharma, P. K.; Bhatia, S. R. Int. J. Pharm. 2004, 278, 361–377.
(16) Sharma, P. K.; Reilly, M. J.; Jones, D. N.; Robinson, P. M.; Bhatia, S. R.
Colloids Surf., B 2008, 61, 53–60.
(17) Foster, B.; Cosgrove, T.; Hammouda, B. Langmuir 2009, 25, 6760–6766.
(18) Nagarajan, R. Colloids Surf., B 1999, 6, 55–72.
(19) Scherlund, M.; Brodin, A.; Malmsten, M. Int. J. Pharm. 2000, 211, 37–49.
(20) Kornhauser, A.; Wei, R. R.; Yamaguchi, Y.; Coelho, S. G.; Kaidbey, K.;
Barton, C.; Takahashi, K.; Beer, J. Z.; Miller, S. A.; Hearing, V. J. J. Dermatol. Sci.
2009, 55, 10–17.
(21) Jung, A. Int. J. Toxicol. 2003, 22, 1–108.
(22) Alleso, M.; Chieng, N.; Rehder, S.; Rantanen, J.; Rades, T.; Aaltonen, J.
J. Controlled Release 2009, 136, 45–53.
ꢀ
(23) Bosca, F.; Miranda, M. A. J. Photochem. Photobiol., B 1998, 43, 1–26.
(24) Valero, M.; Carrillo, C. J. Photochem. Photobiol., B 2004, 74, 151–160.
(25) Delgado-Charro, M. B. J. Drug Delivery Sci. Technol. 2009, 19, 75–88.
(26) Valjakka-Koskela, R.; Hirvonen, J; Monkkonen, J.; Kiesvaara, J.; Antila,
S.; Lehtonen, L.; Urtti, A. Eur. J. Pharm. Sci. 2000, 11, 343–350.
10562 DOI: 10.1021/la100596q
Langmuir 2010, 26(13), 10561–10571

Valero and Dreiss
Article
a variety of molecular guests that can fit into their cavity.27-29 By
threading onto polymers, stable structures known as pseudopo-
lyrotaxanes have been reported, following the pioneering work
of Harada and co-workers.30-33 Formation of a Pluronic-
polyrotaxane would involve sliding of the β-CD along the
polymer chain to preferentially localize on the PPO central block,
for which there is a better geometrical fit than with PEO.31,34
However, cyclodextrins can also form 1:1 (or 1:2) host-guest
complexes with drugs; in fact, they have been widely used in the
pharmaceutical industry for enhancing drug bioavailability, as
they are biocompatible and well-tolerated by the body.35,36 The
driving force for the complexation is attributed to the exclusion of
high energy water from the cyclodextrin cavity, hydrogen and
hydrophobic binding. The complexation of drug with cyclodex-
trin is described by an association constant (Kassoc) given by eq 1:
Materials and Methods
Materials. Pluronic copolymer F127 comprising a central
block of 65 PPO units and two side blocks of PEO (100 units each)
was obtained from Sigma-Aldrich UK (Mw =12600 kg mol-1).
3
Heptakis(2,6-di-O-methyl)-β-cyclodextrin was obtained from
Sigma-Aldrich UK (H0513, Mw=1331.4 g mol-1) and is equally
3
referred to in the text as β-CD or hep2,6 β-CD.
The drugs naproxen sodium salt (NP, M1275), pentobarbital
sodium salt (PB, P3761), sodium salicylate (SAL, 71945), and lido-
caine (LD, L7757) were purchased from Sigma. The solvents used
were hexane for HPLC, dichloromethane, chloroform, methanol,
acetone, and acetonitrile, all of them being analytical reagent
grade, from Fisher Scientific; ethanol (EtOH, for HPLC) from
Fluka; propanol (PrOH) and butanol 99% from Aldrich; and
octanol 99% from Lancaster Synthesis. All materials were used as
received.
Preparation of the Solutions. Drugs/Pluronics solutions
were made by simple mixing. In the case of lidocaine (the only
one not in salt form), the solutions were heated above the drug
melting point (68 ꢀC) to facilitate solubilization inside the micellar
core. The solutions were then cooled back to room temperature.
Ternary solutions containing polymer, drug, and β-CD were
made by mixing the appropriate amounts of F127 and β-CD stock
solutions with the drug in solid form and mixing. All solutions
were prepared by weight, and “%” always refers to weight % (wt %).
SANS Measurements. Most SANS measurements were
carried out on LOQ at the ISIS facility (Rutherford Appleton
Laboratory, Didcot, U.K.). SANS patterns of β-CD on its own
were measured on D11 at the Institut Laue Langevin (ILL)
(Grenoble, France). At ISIS, the instrument uses incident wave-
lengths from 2.2 to 10.0 A, sorted by time-of-flight, and a fixed
sample-detector distance of 4.1 m. This provides access to scat-
tering vectors Q from 0.009 to 0.287 A-1. On D11, three config-
Kassoc
½drug - CDꢀ
½drugꢀ þ ½CDꢀ T
ð1Þ
½drugꢀ½CDꢀ
where [drug] and [CD] represent the free concentration of drug
and cyclodextrin and [drug - CD] is the concentration of the
complex. Complexation is a dynamic process, and the drug can be
displaced if a more favorable interaction with another host is
introduced in the system. Cyclodextrins have been used in combi-
nation with polymers, through either covalent links or physical
association37,38 to modify the drug release properties (e.g., by com-
plexing the drug and either increasing its solubility or decreas-
ing its diffusivity, acting as channelling agents, or cross-linking
agents, etc); however, the mechanisms invoked are very different
from the one reported in this study.
In view of exploiting β-CD-induced micellar breakup for the
controlled release of drugs, we report here investigations on
systems comprising polymeric micelles, drug, and β-CD. From
the above, it is clear that these ternary mixtures lead to three
possible types of interaction, which are in competition with each
other: polymer-drug (micellar encapsulation); cyclodextrin-
drug (inclusion complex); and polymer-cyclodextrin (most prob-
ably inclusion complex). Understanding the mechanisms of inter-
action and the nature of the complexation processes is not only
necessary to improve drug loading and release profile but also
fascinating from a fundamental viewpoint. In this Article, we
examine the structures formed by the interaction of four drugs
with Pluronics in the presence and absence of β-CD. We observe
how the presence of drug affects the disruptive action of β-CD on
the micelles, depending on the structure of the drug. SANS
enables us to monitor structural changes in the micelles upon
drug or β-CD addition, while fluorescence spectroscopy provides
information on the drug microenvironment, enabling us to infer
where the drug is localized and how strongly the different species
interact, in binary or ternary mixtures; a combination of these two
techniques leads us to propose a partial description of the inter-
action mechanisms observed in these complex mixtures.
urations were used to cover a Q range from 0.007 to 0.185 A-1
,
where Q is the modulus of the scattering vector. The scattering
intensity was converted to the differential scattering cross section
in absolute units using the standard procedures at each facility.
Solutions of F127 were 5 wt % concentrated with 0.5, 1, and
2 wt % drug salts or 0.67 wt % lidocaine. β-CD concentrations
ranged from 5 to 13 wt %, corresponding to PO/βCD ratios of 6.9
to 2.6. All samples were measured in D2O, in order to optimize
the contrast and minimize the incoherent background for SANS
experiments. The solutions were equilibrated for a period of at
least 24 h to ensure that the complexation process was complete
before performing the measurements.
Fluorescence Measurements. Measurements were performed
on a Cary Eclipse fluorescence spectrophotometer (Varian, Oxford,
U.K.). Two regimes of drug concentration were used: dilute
(10-3 wt %) and concentrated (2 wt % for NP, PB, and SAL,
and 0.3% for LD, because of its low aqueous solubility:39 5.06
mg mL-1). The aqueous solutions were prepared using Milli
3
Q water.
The following setup conditions were used for the drugs. Lido-
caine: λexc= 262 nm, slits=5-10 nm, PMT=800 V. Pentobarbi-
tal: λexc= 240 nm, slits = 5-10 nm, PMT = 800 V. Naproxen
sodiumsalt:λexc= 317nm, slits=5-5 nm, PMT=400 V. Sodium
salicylate: λexc= 296 nm, slits=5-5 nm, PMT=500 V.
The partition coefficients were determined using the method
proposed by de la Guardia et al.40
(27) Wenz, G. Angew. Chem., Int. Ed. Engl. 1994, 33, 803–822.
(28) Szejtli, J. Chem. Rev. 1998, 98, 1743–1753.
(29) Nepogodiev, S. A.; Stoddart, J. F. Chem. Rev. 1998, 98, 1959–1976.
(30) Harada, A.; Kamachi, M. Macromolecules 1990, 23, 2821–2823.
(31) Harada, A.; Kamachi, M. J. Chem. Soc., Chem. Commun. 1990, 19, 1322–
1323.
(32) Harada, A.; Li, J.; Kamachi, M. Nature 1992, 356, 325–327.
(33) Harada, A. Coord. Chem. Rev. 1996, 148, 115–133.
(34) Harada, A.; Li, J.; Kamachi, M. Macromolecules 1993, 28, 5698–5703.
(35) Vyas, A.; Saraf, S.; Saraf, S. J. Inclusion Phenom. Macrocyclic Chem. 2008,
62, 23–42.
ꢀ
ꢁ
ꢀ
ꢁ
ꢀ
ꢁ
- 1
- 1
I
IM
I0
1
- 1
¼
- 1
1 þ
ð2Þ
I0
γKCM
where CM is the micellar concentration with CM=(CS - cmc), CS
being the total surfactant concentration. K is the binding constant
(36) Davis, M. E.; Brewster, M. E. Nat. Rev. 2004, 3, 1023–1035.
(37) Bibby, D. C.; Davies, N. M.; Tucker, I. G. Int. J. Pharm. 2000, 197, 1–11.
(38) van de Manakker, F.; Vermonden, T.; van Nostrum, C. F.; Hennink, W. E.
Biomacromolecules 2009, 10, 3157–3175.
(39) Kang, L.; Jun, H. W.; McCall, J. W. Int. J. Pharm. 2000, 206, 35–42.
(40) de la Guardia, M.; Peris-Cardells, E.; Sancedon, J.; Carrion, J. L. Micro-
chem. J. 1991, 44, 193–200.
ꢀ
ꢀ
Langmuir 2010, 26(13), 10561–10571
DOI: 10.1021/la100596q 10563

Article
Valero and Dreiss
of the drugto the micelle and γ is the ratio ofmolar absortivities of
the drug in both media (it is 1 in all cases). The cmc value used is
0.26 ( 0.03 wt %.16
The partition coefficients can be calculated from the binding
constant following eq 3.41
K ¼ ðPMW - 1ÞV
ð3Þ
where V is the partial molar volume of the polymeric surfactant
(a value of 1.24 was used).
The binding constant of each drug to cyclodextrin, KB, was
determined using the following expression:
ðF0 þ F¥KB½CDꢀÞ
F ¼
ð4Þ
1 þ KB½CDꢀ
where F is the measured fluorescence intensity, and F0 and F¥ are
the fluorescence intensity when all the drug is free and complexed,
respectively (both are experimental parameters). KB is the binding
constant, obtained by fitting the experimental data. [CD] is the
concentration of free β-cyclodextrin, which, in dilute systems,
corresponds to the analytical concentration, since [CD] . [drug].
In concentrated solutions instead, the free β-CD concentration
was determined by means of a Taylor’s series expansion of the
equation that relates the total concentration to the binding con-
stant and free cyclodextrin concentration as described in the lite-
rature.42 A nonlinear least-squares method was used to fit the
experimental results to eq 4.
Figure 1. Small-angle neutron scattering patterns from solutions
of 5% F127 in D2O (4) and in the presence of 0.67% lidocaine (O),
2% pentobarbital (b), 2% naproxen ( ꢁ ), or 2% sodium salicylate
(0), showing different effects onthe micellaraggregates, depending
on the drug added.
Results and Discussion
1. Micellar Structural Changes in Binary and Ternary
Systems Monitored by SANS. 1. 1. Effect of Drugs on F127
micelles. Lidocaine. Previous measurements2 on concentrated
F127 micellar solutions (15 wt %) have shown an increase in the
aggregation number and micellar size when incorporating lido-
caine up to 2 wt %. In addition, the presence of drug drove F127
micelle organization into a macrolattice, leading to gelation at
a lower polymer concentration than in the absence of drug.2
Measurements performed in more dilute conditions, 5 wt %
micellar solutions in the presence of 0.67 wt % lidocaine (near
saturation), are shown in Figure 1. The scattering of the drug
dissolved as a molecular species is negligible, and hence, changes
in the SANS patterns arise solely from changes in micellar struc-
ture (this applies to all the drugs studied here). The incorporation
of lidocaine leads to a clear shift of the micellar scattering to lower
Q values, accompanied by an overall increase in the intensity; the
combination of these two features reflects an increase in micellar
size. This is in agreement with the results of Sharma et al.16 who
showed an increase in core size from 43 to 58 A and corona from
66 to 71 A upon incorporation of lidocaine to saturation. Our data
also indicate that lidocaine is encapsulated into the micelles, lead-
ing to a net increase in their size. This is in agreement with the high
selectivity of Pluronics for aromatics (over aliphatics)43 and pre-
viously reported formulations of lidocaine with Pluronics.15,16,19,44,45
Pentobarbital Sodium Salt. In comparison to lidocaine, the
effect of pentobarbital on F127 micellar structure is minimal
(Figure 1). Addition of 2 wt % leads to a slight increase in inten-
sity, but no shift to lower Q values (hence no change in micel-
lar size), accompanied by a slight decrease of the signal at lower Q,
Figure 2. SANS patterns from 5% F127 micellar solutions alone
(b) and in the presence of 7% (]), 9% (4), and 11% ( ꢁ ) hep2,6
β-CD, where full demicellization occurs.
reflecting an enhancement of intermicellar interactions, probably
arising from increased charge repulsions. This effect is gradual
when adding 0.5-2 wt % PB (only the highest concentration is
shown here; 1 wt % PB is shown in Figure 3). It is possible that
most PB stays solubilized in water, therefore not disturbing very
much the structure of the micelles. Some interaction however with
F127 micelles could occur with the outer layer, explaining the
stronger micellar repulsions observed and the slight increase in
intensity. No published data have been found on the interaction
of pentobarbital with Pluronic micelles.
Naproxen Sodium Salt. Formulations in F127 micelles have
been shown to significantly improve naproxen solubility and half-
life.46,47 Figure 1 shows the effect of adding 2 wt % naproxen salt
to 5 wt % micellar solutions of F127. The intensity of the scatter-
ing pattern decreases significantly, together with a shift of the main
scattering peak (0.025 A-1) to a higher Q value, while micellar
features are all smoothed out. This reflects a shrinking of the
(41) Song, G. P.; Guo, R.; Yang, Y. F. J. Yangzhou Technol.Coll. (Nat.Sci.)
1992, 12, 42–47.
(42) Connors, K. A. Binding Constants, the Measurement of Molecular Complex
Stability; John Wiley and Sons: New York , 1987; p 42.
(43) Nagarajan, R.; Barry, M; Ruckenstein, E. Langmuir 1986, 2, 210.
(44) Scherlund, M.; Malmsten, M.; Brodin, A. Int. J. Pharm. 1998, 173, 103–116.
(45) Ricci, E. J.; Lunardi, L. O.; Nanclares, D. M. A.; Marchetti, J. M. Int. J.
Pharm. 2005, 288, 235–244.
(46) Suh, H.; Jun, W. Int. J. Pharm. 1996, 129, 13–20.
(47) Suh, H.; Jun, W.; Dziminaskl, M. T. Biopharm. Drug Dispos. 1997, 18,
623–633.
10564 DOI: 10.1021/la100596q
Langmuir 2010, 26(13), 10561–10571

Valero and Dreiss
Article
interact with PPO and hence reduces micellar breakup. These two
hypothese are later discussed in the text for each drug and referred
to by hypothesis (1) and (2).
Pentobarbital Sodium Salt (NaPB). The addition of
9-13 wt % β-CD (equivalent to PO units/β-CD ratios of 3.8 to
2.6) to 5 wt % F127 micellar solutions in the presence of 1% PB
(Figure 3) leads to a complete disruption of the micelles. The
patterns obtained are almost identical to the ones obtained for
pure micelles of F127 in the absence of drug (shown in Figure 2)
and to the scattering pattern of β-CD alone in solution (shown in
the inset of Figure 3). This confirms the absence of any residual
micellar scattering, and hence, the polymer has been totally
reduced to its unimeric form. The breakup is achieved with the
lowest amount of β-CD studied (9%) and does not change with
further addition of β-CD, showing that even if further β-CD/
polymer complexation occurs, it does not alter the structures
formed (however, it is possible that the onset of micellar breakup
occurs below 9% β-CD, not studied here). In contrast, the same
complexes at a higher PB concentration of 2 wt % (also shown in
Figure 3) lead to a completely different picture: in this case, the
presence of PB clearly weakens the disruptive action of β-CD on
the micelles, as shown by the very gradual smoothing out of
micellar scattering, instead of its collapse. At the highest β-CD
concentration studied (in this case 9%), the scattering pattern is
still very structured and originates from micelles, instead of
unimers. Comparing 9% β-CD addition to the micellar solutions
containing either 1 or 2 wt % PB, it is very clear that the drug is
able to reduce the disruptive action of β-CD on the micelles,
provided it is above a critical concentration, which may be linked
to either a critical PB/F127 or PB/ β-CD ratio, or a combination,
depending on the mechanism of action. Interestingly, the impact
of PB on F127 micelles was found to be moderate (Figure 1),
suggesting weak drug/polymer interactions, and therefore point-
ing to hypothesis (2), namely, a favorable interaction between
β-CD and PB (hypothesis (2)), which would reduce the avail-
ability of β-CD to complex the polymer. The fluorescence data
presented below give additional evidence to discuss this hypoth-
esis in more quantitative terms.
Figure 3. SANS patterns from 5% F127 micellar solutions with
the drug pentobarbital at two concentrations: 1% (O) and 2% (b),
and varying amounts of hep2,6 β-CD. F127 solutions with 1% PB
and 9% (2), 11% (ꢁ), and 13% (3) hep2,6 β-CD show full break-
up of the micelles. Instead, F127 with a higher amount of drug
(2% PB) and 5% (0), 7% (]), and 9% (4) hep2,6 β-CD shows
only a slight decrease in the size of the micelles. The scattering from
solutions of 11% hep2,6 β-CD alone (9) is also shown for com-
parison (this last set was measured at ILL, hence on a different
Q range, and scaled down from a solution containing 13% hep2,6
β-CD).
micelles (which is gradual upon addition of 0.5 to 2 wt % NP) and
a weakening of intermicellar interactions. This suggests a mecha-
nism of interaction whereby the charged drug interacts strongly
with the micelles, inserting itself in the aggregate (possibly the
palisade layer, as the drug is largely soluble in water), introducing
electrostatic repulsions and therefore reducing the packing of the
chains; as a consequence, the micellar aggregation number de-
creases. This result is in agreement with the data published by
Sharma and Bhatia15 who reported a decrease in the aggregation
number in 2 wt % F127 micelles from 88.6 to 51.9 upon addition
of naproxen.
Naproxen Sodium Salt. The effect of β-CD (9-13 wt %) on
F127 micelles in the presence of 1 wt % naproxen is very similar
to that of PB: the micelles are fully broken up, as shown by the
collapse of the scattering pattern (Figure 4), and all curves
superimpose with the ones shown for 1% PB data in the presence
of β-CD. This is somehow surprising as the interaction of F127
micelles with NP seemed stronger than with PB (Figure 1). How-
ever, the resulting structures may be identical (namely, unimers),
but the mechanism of interaction different. At a higher concen-
tration of NP (2 wt %), a very interesting and unexpected effect is
observed (Figure 4): the addition of 7 and 9 wt % β-CD leads to
an increase in micellar size, therefore offsetting the disruptive
effect that NP had on the micelles (Figure 1). At higher concen-
trations of β-CD, however (11 and 13 wt %), this effect is no
longer seen and Pluronics are reduced to unimers, as in the case of
the lower drug concentration (1%). These intriguing results
suggest that at low concentration β-CD could be complexing
preferentially with NP, therefore decreasing drug/polymer inter-
action. As a result of this preferential β-CD/drug interaction, the
micelles would grow (since charge repulsions induced by the drug
are reduced). At higher concentrations, however, possibly after
full β-CD/drug complexation, β-CD could interact with the
polymer by breaking up the micelles, resulting in a pattern similar
to the one observed in the absence of drug and at a lower drug
concentration (1%). Further studies on a range of β-CD and drug
ratios would need to be carried out to pinpoint the occurrence
Sodium Salicylate. The effect of 2 wt % SAL on the Pluronic
micelles is very similar to that of naproxen (Figure 1). With the
addition of 2 wt %, a significant reduction in micellar size occurs,
shown by the collapse of the intensity and disappearance of micellar
structural features. This suggests again a significant interaction of
the salt with the micellar aggregates. The similarity with NP is not
surprising, as NP and SAL structures are fairly related, with NP
bearing an additional aromatic ring. The effect of sodium sali-
cylate on Pluronic micelles, to our knowledge, has not been repor-
ted elsewhere.
1.2. Effect of β-CD on Drug-Containing Micelles. We have
recently shown that the gradual addition of hep2,6 β-CD to F127
Pluronic micelles leads to their gradual disruption;1,2 the effect
is highly dependent on temperature. At 25 ꢀC, the addition of
7-11 wt % hep β-CD to 5 wt % F127 micellar solutions leads
to complete disruption of the micelles to their unimers as shown
in Figure 2 (the micellar disruption observed at 37 ꢀC is more
gradual2).
Lidocaine. We have also previously shown that the disruptive
effect of β-CD on Pluronic micelles is weakened in the presence of
the drug lidocaine.2 This can be explained by two competitive
processes of interaction, or a combination of both: (1) the hydro-
phobic drive of the drug stabilizes the micelles against the disrup-
tive action of β-CD; (2) the formation of a drug/β-CD inclu-
sion complex reduces the number of β-CD molecules available to
Langmuir 2010, 26(13), 10561–10571
DOI: 10.1021/la100596q 10565

Article
Valero and Dreiss
Figure 4. SANS patterns from 5% F127 micellar solutions with
the drug naproxen at two concentrations: 1% (ꢁ) and 2% (b), and
varying amounts of hep2,6 β-CD. F127 solutions with 1% napro-
xen and 9% (2), 11% (ꢁ), and 13% (3) hep2,6 β-CD show full
breakup of the micelles. Instead, solutions of F127 with a higher
amount of drug (2% NP) and 7% (]) and 9% (4) hep2,6 β-CD
show a growth of the micelles (as compared to F127 with 2% NP
(b)), while full breakup is observed with 11% (ꢁ) and 13% (1)
hep2,6 β-CD.
Figure 5. Comparison of the SANS patterns obtained from solu-
tions of 13% hep2,6 β-CD alone (9) and the complexes formed by
(i) 5% F127, 2% naproxen and 13% hep2,6 β-CD (1); (ii) 2%
naproxen and 13% hep2,6 β-CD (3).
out in three stages: (i) determination of the partition coefficient of
the drugs into F127; (ii) determination of the binding constant of
the drugs to hep2,6 β-CD; (iii) determination of the binding of the
drugs to hep2,6 β-CD in the presence of F127.
2.1. Interaction of the Drugs with F217: Partition Coefficient.
Lidocaine. In water, the drug displays three bands centered
around290, 350, and 410 nm(Figure6A(1)), whichare not always
present, depending on factors which are not well understood. It is
clear however that several species exist in equilibrium, whose
proportion varies with properties of the environment. Addition of
F127 produces a hyperchromic effect and a change in the shape of
the spectrum at both drug concentrations (Figure 6A(2) and (3)).
In addition, a red shift in the lower wavelength band (from 290 to
304 nm) is also observed at high LD concentrations. Together, all
these changes show that LD interacts with the polymer and pro-
motes structural changes in the aggregate.
The binding constant of LD to F127 and its partition coeffi-
cient determined by fitting the fluorescence intensity at the maxima
of the spectrum (cf. Materials and Methods) are presented in
Table 1. They show that the presence of increasing amounts of
drug promotes its partition into the micelles, as PMW increases
from 2.21 in dilute conditions to 28.8 in concentrated solutions.
This latter value is in better agreement with the value of 20 repor-
ted for this system at saturation obtained from UV spectroscopy.16
In order to bring additional insight into the microenvironment
of the drug inside the micelles, we next attempted to examine the
bands of the fluorescence spectra in more detail. We performed
pH variation studies of aqueous solutions of LD, which show that
the emission band centered around 290 nm is present at all pH,
while the species emitting at 410 appears at high pH (cf. Suppor-
ting Information). This suggests that inclusion of LD inside F127
micelles at high drug concentration promotes the appearance of
species that are formed in basic media (since emission at 410 nm is
enhanced). The enhanced dissociation of weak acid compounds
by polymer addition has been reported before.48 This effect depends
on the solvation of the acid in the presence of polymer.48 In order
to establish whether the drug inside the micelle presents a change
in solvation, the effect of solvent on LD emissive behavior was
studied. As expected, an important solvatochromism is observed
(cf. Supporting Information). In these homogeneous media, LD
of these phase transitions. However, these results highlight the
delicate balance of forces involved and show how this balance can
be reversed by small changes in species ratio. The results also
provide strong evidence for a preferential interaction of β-CD
with the drug and open up an interesting perspective for delivery,
where β-CD can cause a displacement of the drug from the inte-
rior of the micelles into β-CD inclusion complexes. These hypo-
theses are examined in more detail with fluorescence spectroscopy.
In order to better characterize the nature of the complexes, the
scattering of naproxen (2%) in the presence of β-CD (13%) is
shown in Figure 5, together with the scattering from β-CD alone at
the same concentration and the three-component mixture (with
5% F127). The scattering patterns obtained from both the binary
and ternary mixtures are lower than the scattering from β-CD on
its own, confirming therefore that total breakup of the micelles (in
the ternary mixture) and full complexation (in both binary and
ternary mixtures) have occurred. As the scattering from the drug
alone is negligible, the modified scattering pattern obtained from
the drug/β-CD mixture compared to β-CD alone confirms the
formation of a complex. The decrease in intensity at low Q in both
mixtures reflects the occurrence of strong interactions, attributed to
charge repulsion coming from the drug included in the complexes.
2. Complexation Behavior from Fluorescence Measure-
ments. Fluorescence is a very sensitive technique to study changes
in the microenvironment of a given compound. Fluorescence
requires low absorbance at the excitation wavelength; for this
reason, it is usually carried out in dilute systems and the results are
extrapolatedto concentrated conditions. However, the purpose of
drug formulations is usually to encapsulate high loadings of drug,
and therefore, rather concentrated systems are the norm. In addi-
tion, simple extrapolation of the results obtained in dilute solu-
tions is not always a true representation of the systems at high
concentration. Finally, SANS measurements require relatively
highconcentrationsto beable todetect structural changes, and we
therefore needed to reproduce the same conditions for the fluore-
scence measurements. For these reasons, the fluorescence mea-
surements were performed on two sets of samples: dilute (10-3%)
and concentrated (2%, except 0.3% for LD). The study was carried
(48) Sekhon, S. S.; Arora, N.; Singh, H. P. Solid State Ionics 2003, 160, 301–307.
10566 DOI: 10.1021/la100596q
Langmuir 2010, 26(13), 10561–10571

Valero and Dreiss
Article
Figure 6. Fluorescence spectra of the drugs in different media. (A) (1) LD in H2O, presenting three bands centered at λmax = 290, 350, and
410 nm; (2) LD 10-3% and F127 5%, presenting a new band at λmax = 330 nm (corresponding to the emission of LD in a very high HBD
environment) and a shoulder around 410 nm (corresponding to the emission of LD in a low HBD environment); (3) LD 0.3% and F127 5%,
presenting two bands, one at 304 nm (corresponding to species emitting at 290 nm in water, but red-shifted by F127 addition due to the
partition of LD into the apolar micelle environment) and a second band at 410 nm (corresponding to the emission of LD in a low HBD
environment). (B) (1) PB in H2O. The spectrum presents three bands centered at λmax = 260, 350, and 410 nm; (2) PB 10-3% and F127 5%,
λmax = 260 nm, 335 nm (corresponding to the emission of PB in a very high HBD environment) and λmax = 410 nm (corresponding to the
emission of PB in a low HBD environment); (3) PB 2% and F127 5%, λmax = 298 nm, 410 nm, and a shoulder around 350 nm (corresponding
tothe emissionofPBinahighHBDenvironment);the changesobservedarerelated tothe inclusionofPBinthemicelleandshowthatonlythe
region of the micelle with acidic character is affected. (C) NP 2% in H2O and in the presence of 0-5% Pluronic F127. Inset: Variation of the
fluorescence intensity with F127 concentration at the emission maximum of 2% NP. The increase in fluorescence which follows premicellar
and micellar aggregate formation shows that NP is in contact with water molecules at the micelle surface. (D) SAL 2% in H2O and in the
presence of 0-5% of Pluronic F127. Inset: Variation of the fluorescence intensity with F127 concentration at the emission maximum of 2%
SAL. Quenching occurring in the region of the cmc indicates hydrogen bond formation with the water shell of the micelle.
presents one or two bands centered at different wavelengths, which
vary with the properties of the solvent. The main observations can
be summarized as follows: (i) the band emitting at low wavelength
is red-shifted when the polarity decreases (290 nm in water and
300 nm in hexane or DCM); (ii) the emission of the drug in apolar
solvents strongly increases; (iii) the band appearing at high wave-
length, when present, is linearly red-shifted as the hydrogen bond
donor (HDB) ability decreases, in good agreement with the pH
data. On this basis, the increase in intensity (at both drug concen-
trations) and the red shift of the 290 nm band (at high drug con-
centration) observed with F127 addition can be interpreted as LD
being displaced into the lower polarity micellar environment. The
position of the higher wavelength bands, related to the HBD abi-
lity of the micelle, shows that at low concentration LD is in con-
tact with two different environments, one with high HBD ability
(higher even than that of water, with a band at 335 nm) and another
one with a very low HBD environment (shoulder centered at
410 nm). Therefore, LD is inside a micellar region with this dual
nature, which could be the micelle interface. Considering the struc-
ture of lidocaine (Scheme 1), the LD benzene group is most pro-
bably inside the apolar region of the micelle with the substituents
protruding into the water shell. Previous studies using UV spectro-
scopy have indeed reported limited partitioning of LD inside the
F127 micellar core,16 and FTIR studies have shown that unchar-
ged local anesthetics are mainly located in the membrane/water
interface, competing with the polar heads of phospholipids in
binding with water molecules.49,50
At high drug concentration instead, the enhancement of the
410 nm band points to a change in the microenvironment around
the drug. LD is less in contact with the protonating zone of the
micelle, that is, the OH terminal groups of PEO or the water mole-
cules. On this basis, it is possible to speculate that as the drug
concentration increases, the micelles become tighter, changing the
solvation of the micelle. Overall, the good partitioning of LD
inside F127 micelles concurs with the net increase in aggregate size
detected by SANS (Figure 1 and ref 2).
Sodium Pentobarbital. The fluorescence spectrum of PB
in water presents three main bands: 260, 350, and 410 nm
(49) Ueda, I.; Chiou, J. S.; Krishna, P. R.; Kamaya, H. Biochim. Biophys. Acta
1994, 1190, 421–429.
(50) Fraceto, L. F.; de Paula, E. Rev. Cie^nc. Farm. Basic Apl. 2006, 27, 27–35.
Langmuir 2010, 26(13), 10561–10571
DOI: 10.1021/la100596q 10567

Article
Table 1. Partition Coefficients of the Drugs in 5% Pluronic
F127 (in Water)
Valero and Dreiss
However, and in contrast to LD, the basic character of the
aggregate is not modified by PB (as the 410 nm band does not
change). This behavior seems to indicate that PB only modifies
one of the two regions of the micelle which it is in contact with. By
considering the different volumes of the hydrophobic (large in
LD) and hydrophilic (large in PB) regions of LD and PB, together
with the fluorescence data, it is possible to surmise that PB is
“anchored” to the surface of the micelles through the lateral
chains, whereas LD is inside the micelle at the core/shell interface.
Localization of most of the PB molecules at the surface of the
micelles, together with the presence of charge, concurs with the
results from the SANS analysis, which showed that PB may affect
intermicellar interactions but only marginally the micellar struc-
ture, whereas LD, partitioned also at the interface but with most
of the molecules deeper inside the micelles, significantly affects the
structures.
drug
K/(g/g)
PMW
lidocaine
dilute
dilute
2.21 ( 0.15(F330)
1.50 ( 0.13 (F330
2.20 ( 0.09 (F290
concentrated
)
)
2.77 ( 0.12 (F290
)
concentrated
28.8 (F293)
34.5 ( 0.9 (F293
)
pentobarbital
dilute
dilute
8.83 ( 0.99 (F335
)
6.59 ( 1.00 (F335
concentrated
)
concentrated
7.79 ( 2.0 (F298
)
7.28 ( 2.32 (F298
7.60 ( 1.90 (F350
)
)
8.18 ( 1.69 (F350
)
salicylate
naproxen
concentrated
4.55 ( 0.32
concentrated
4.67 ( 0.37
Sodium Salicylate and Naproxen Sodium Salt. Addition of
F127 to a dilute drug solution, in both cases, produces no changes
in the fluorescence intensity, indicating no partition. At high drug
concentration, F127 addition changes the emission intensity of
NP and SAL, showing the occurrence of interaction with F127
(Figure 6C and D).
concentrated
3.18 ( 0.90
concentrated
3.56 ( 0.95
(Figure 6B(1)). The addition of F127 to aqueous solutions of PB
(at 10-3% and 2%) produces a strong hyperchromic effect, ref-
lecting drug/micelle interaction (Figure 6B(2) and (3)). The shape
of the spectrum is similar at both drug concentrations, but, as
observed for LD, the maxima positions are different.
Further analysis of the fluorescence behavior highlights some
differences between the two drugs. In the case of salicylate, a strong
quenching by F127 addition is produced (Figure 6D, inset).
Salicylate has been reported to be quenched by intra- and inter-
molecular hydrogen bond formation in aqueous solutions,51
whereas it does not interact with alcohols.52 Therefore, this sharp
quenching occurring in the region of the cmc clearly points to
the inclusion of the drug inside the micelles and the formation
of hydrogen bonds with the water shell of the micelle. Naproxen,
in contrast, is quenched by intermolecular hydrogen bonds
formed with alcohols, but not in water.53 On this basis, the in-
crease in fluorescence (Figure 6C, inset) which follows premicellar
and micellar aggregate formation shows that in these aggregates,
NP, like SAL, is in contact with water molecules at the micelle
surface.
The partition coefficients (Table 1) are of the same order of
magnitude for both (3.56 and 4.67 for NP and SAL, respectively),
and lower than those for the other two drugs (LD and PB). The
similarity in partition coefficient for both compounds concurs
with the results from SANS, which showed an almost identical
effect of both drugs on the micelles (Figure 1). It is known that the
partition in Pluronics is favored for aromatic compounds com-
pared to nonaromatics.43 Salicylate, naproxen, and lidocaine are
aromatic compounds, LD and SAL have quite a similar aromatic
structure, and the main difference between these two drugs is the
presence of charge in the former. Moreover, NP presents a more
hydrophobic aromatic ring than SAL, but the same charge. The
presence of charge clearly decreases the binding constant, follow-
ing the trend: NP ≈ SAL<PB<LD. Therefore, it is possible to
speculate that the main cause of poor partitioning inside the
micelles arises from charge repulsion between the drug molecules;
for comparison, the partition coefficient reported for molecular
naproxen is 355 ( 64.16
The partition coefficients determined at both drug concentra-
tions are the same (Table 1), and they are also identical at both
maxima (298 and 350 nm) appearing in the concentrated system.
This suggests that the presence of PB does not change the pro-
perties of the micelle, which is in very good agreement with the
SANS data, where very little change was detected in the binary
systems (Figure 1). It is much lower than LD (at high concen-
tration), in good agreement with the better ability of aromatic
compounds to be solubilized inside Pluronic micelles.43 In addi-
tion, the presence of heteroatoms and electron pairs in the PB
structure, added to the presence of charge, is likely to contribute
to the lower partition of PB inside the micelles, compared to LD.
These results certainly corroborate the substantial micellar
growth seen with lidocaine, compared to the very weak effect of
PB on the micellar structure (Figure 1).
The effect of pH and solvent was studied next to obtain more
insight on the nature of the interactions. The bands centered at
298 and 350 nm appear at acid pH (below pH = 3), whereas a
strong band around 410 nm appears at pH >12 (cf. Supporting
Information). The fluorescent behavior of PB in homogeneous
solvents shows a strong solvatochromic effect (cf. Supporting
Information). Overall, we observed that (i) the band centered at
298 nm clearly appears in apolar solvents (but it is also present at
low pH); (ii) the band centered at 355 nm appears in solvents with
high HBD, in good agreement with the bands present at low pH;
(iii) the band centered at 410 nm appears exclusively in acetone
(very low HBD) as a sole band; (iv) PB is strongly fluorescent in
propanol, butanol, and octanol, and all these solvents have in
common a high viscosity. Quite remarkably, in propanol, all the
bands appear clearly (300, 350, and 410 nm). Taken together,
these results point to the inclusion of PB inside the micelles in an
apolar (ε ∼ 20) environment similar to propanol and with double
acid-basic character, therefore at the interface, as observed for
LD, most probably in the PEO headgroups and in contact with
the water shell. Therefore, PB must be included in the micelle
through its hydrophobic substituents, with the polar ring in
contact with the water shell, as observed for LD. However, there
are clear differences between the spectra of PB and LD in F127.
The low wavelength band maxima positions show that, as for LD,
the acid character of the micelle decreases upon PB addition.
Stability of the micelles, and therefore cmc and aggregation
number (Nagg), is known to be the result of a repulsive-attractive
balance of forces between the hydrophobic chains and the hydro-
philic headgroups of the surfactant, respectively, acting mainly on
(51) Bisht, P. B.; Tripathi, H. B.; Pant, D. D. J. Photochem. Photobiol., A 1995,
90, 103–108.
(52) Escandar, G. M. Spectrochim. Acta, Part A 1999, 55, 1743–1752.
ꢀ
ꢀ
(53) Velazquez, M. M.; Valero, M.; Rodrı
´
guez, L. J.; Costa, S. M. B.; Santos,
M. A. J. Photochem. Photobiol., B 1995, 29, 23–31.
10568 DOI: 10.1021/la100596q
Langmuir 2010, 26(13), 10561–10571

Valero and Dreiss
Article
the interfacial region of the aggregates.54 Our results show how
the presence of additives contributes to modify this balance of
forces in the aggregate, resulting in micellar size changes that limit
the partition, owing to the stability of the loaded micelle. In
summary, and combining the fluorescence data with the SANS
results, an increase in the attractive forces between the hydro-
phobic chains increases micellar size and consequently Nagg (LD,
high partition); incontrast, an increase in repulsive forcesbetween
the hydrophilic headgroups decreases micellar size (NP and SAL,
low partition). Inthe case of PB, SANS data show that additionof
the drug only affects very marginally the micellar structure, with a
slight increase in intermicellar repulsions, while the fluorescence
data show that all the drugs are localized at the micelle interface.
Hence, for PB, the large polar head allows the ionizable group to
stay in water far from the hydrocarbon lateral chains, which are
likely to be included in the micelle; therefore, the presence of PB
does not contribute to the overall repulsive forces in the core-
shell region of the aggregate, which effectively leads to minor micellar
structural changes, but to a slight increase in surface repulsions.
2.2. Binding Constant of the Drugs to β-CD, in the Absence
and Presence of Pluronics. In the following section, we examine
the interactions in the ternary systems composed of drugs, Pluro-
nics and heptakis(2,6-di-O-methyl)-β-cyclodextrin. Prior to study-
ing the ternary mixtures, the complexation of the drugs with
hep2,6 β-CD is studied first for each drug, and a binding constant
determined, both at low and high drug concentration. The effect
of the presence of polymer (5%) on the complexation behavior of
the drug is then studied.
Lidocaine. The presence of hep2,6 β-CD produces an increase
in LD emission intensity and a red shift of the 290 nm band,
showing the inclusion of LD inside the β-CD cavity (cf. section
2.1). At both drug concentrations, the intensity variation fits
well to a 1:1 complex model (Figure 7). The binding constant
determined at low LD concentration (Table 2) is very low 3.80 (
0.1 (g/g), and it is increased to 13.01 ( 0.84 (g/g) (Table 2) at high
drug concentration.
Figure 7. Experimental fluorescence data (dots) and fitting (solid
line) for the determination of drug/β-CD binding constant. (A) Dilute
drug systems (10-3%). (B) Concentrated drug systems (0.3% for
LD and 2% for the other drugs).
The same studies were then performed in the presence of F127.
At low drug concentration in 5% F127, β-CD addition increases
the emission intensity, and the maximum is blue-shifted. These
changes show that LD/β-CD inclusion also occurs in the presence
of Pluronic micelles, despite the low binding constant measured
in the absence of polymer. In the concentrated LD systems in
the presence of 5% F127, addition of β-CD leads to an irregular
pattern in the emission intensity of the drug. A break in the
fluorescence trend is observed around 3% of β-CD, reflecting a
modification in the behavior of the system, probably related to
micellar breakup. These changes show that the competition bet-
ween LD and Pluronic is still present at these high concentrations.
These results suggest that drug/CD complexation lead to a weaken-
ing of the breaking up of the micelles by β-CD, as reported
earlier,2 because of the reduced availability of β-CD molecules,
therefore bringing support to hypothesis (2) (we cannot, however,
discard hypothesis (1)).
Sodium Pentobarbital. The addition of β-CD to dilute aqu-
eous solutions of the drug produces a change in fluorescence
intensity and a strongblue shift (from 358 to 338 nm), showing the
inclusion of the drug inside the hydrophobic cavity of hep2,6
β-CD. The fluorescence intensity data fit well to a 1:1 stoichiom-
etry (Figure 7A). The binding constant obtained (Table 2) is
higher than that for LD. Addition of hep2,6 β-CD to aqueous
solutions of PB and F127 produces the same changes as in the
Table 2. Binding Constants (g/g) of 1:1 Drug/hep2,6 β-CD
at 25ꢀC (in Water)
drug
without F127
with F127
lidocaine
dilute
dilute
3.80 ( 0.1 (F290
)
-
concentrated
-
concentrated
13.01 ( 0.84 (F290
)
pentobarbital
dilute
dilute
-
48.92 ( 1.9 (F352
51.21 ( 8.8 (F262
concentrated
)
)
concentrated
-
72.2 (F299
197.9 (F355
)
)
salicylate
naproxen
dilute
dilute
440.9 ( 30.1
concentrated
0.025 ( 0.001
277.5 ( 47.5
concentrated
4.50 ( 0.46
dilute
625 ( 25
concentrated
56.55 ( 20.6
dilute
211.6 ( 58.4
concentrated
123.4 ( 25.89
absence of polymer, showing that hep 2,6 β-CD/PB complexation
occurs all the same. This confirms again a competition between
drug and Pluronic for complexing with the cyclodextrin.
The addition of increasing amounts of hep2,6 β-CD to 2% PB
aqueous solutions slightly increases the fluorescence intensity up
(54) Israelachvilli, J. N. Intermolecular and Surface Forces, 2nd ed.; Academic
Press: San Diego, 1992; p 367.
Langmuir 2010, 26(13), 10561–10571
DOI: 10.1021/la100596q 10569

Article
Valero and Dreiss
to 2% β-CD (Figure.7B), in good agreement with a 1:1 complex.
Further addition of cyclodextrin produces a decrease in the emis-
sion intensity. These changes point out to a PB/hep2,6 β-CD
complex formation. However, the interaction seems to be diffe-
rent at low and high β-CD concentration, probably reflecting the
formation oftwo types of complexes, asdescribed for barbiturates
with hydroxypropyl β-CD.55 From these data, it is not possible to
obtain a good fit; however, for indicative purposes only, we deter-
mined a binding constant using the fluorescence variation of the
different bands appearing in the spectra: its value varies between
72.2 and 198 (g/g).
The addition of hep2,6 β-CD to aqueous PB solutions in the
presence of 5% F127 decreases the fluorescence intensity. All
these changes clearly show that the drug is being displaced from
F127 micelles into the β-CD cavity. The spectrum is similar to the
one of PB at low β-CD concentration in the absence of F127,
indicating that only one type of complex is now formed, in good
agreement with a competition between the micelle and the cyclo-
dextrin for binding the drug. It is also clear that the presence of PB
decreases the free avalaible CD for Pluronic complexation, but as
observed for LD the existence of other contributions cannot be
ruled out.
Therefore, three equilibria are present: PB/CD, PB/micelle,
and probably F127/CD, as reflected by the binding, partition, and
breaking of the Pluronic micelle, respectively. If we suppose that
the presence of F127 does not modify the PB/β-CD binding
constant (K = 51.21 (g/g)) and if the drug partitioning is not
competing with its complexation with β-CD, at 1% drug, the
available (free) β-CD concentration would be 6.2 and 8.2% for
7 and 9% total β-CD, respectively. On the opposite, if we consider
that the partitioning is 100% competitive for the complexation
(in other words, that all the druginsidethe micelles is not available
for β-CD complexation), at 1% PB and 7-9% β-CD, the free
β-CD would be 6.7-8.7%. In real conditions, the effective free
β-CD must be between these two extreme scenarios, that is,
6.2-6.7% for 7% total and 8.2-8.7% for 9% total. At 9% total
β-CD (8.2-8.7% effective), the micelles break (Figure 3), as it
would be expected since they break with 7% β-CD in the absence
of drug(Figure 2). With 2% PB instead, the freeβ-CD determined
for the same scenarios is between 5.5 and 6.3% and 7.4-8.3% for
7 and 9% total β-CD, respectively. For 9% total β-CD, how-
ever (i.e., above 7% effective), the micelles are not fully broken
(Figure 3). Therefore, in this case, the drug competition for β-CD,
even including the partitioning of the drug inside the micelles,
cannot be the only factor involved in weakening the micellar
breaking process, and thus, hypothesis (1) must be considered in
addition to hypothesis (2).
indicate that both drugs complex with β-CD at both concentra-
tions. The emission intensity variation shows the formation of a
1:1 complex for both NP and SAL (Figure 7), as for the other
drugs.
For these two drugs however, there is a major difference bet-
ween the binding constant in dilute and concentrated systems
(Table 2). This could arise from drug aggregation (naproxen
excimer formation is indeed observed) but also from inner filter
effects, which cannot be avoided in these conditions, hence, these
results should be treated with care.
The binding constant of dilute salicylate systems (440.9 (
30.1 g/g) (or 587.1 ( 43.3 M-1) (Table 2) is much higher than
those reported previously for hep2,6 β-CD (140 ( 15 M-1).56
Comparing with the other drugs used in this study, the binding
constant of naproxen is the highest. The high binding constant of
NP to hep2,6 β-CD, even in the presence of F127, gives further
support to the hypothesis that the micellar growth observed at
7 and 9% CD with 5% F127 (Figure 4) is due to the decrease of
NP concentration inside the micelle, as a direct result of its com-
plexation with hep2,6 β-CD.
In dilute systems, in the presence of F127, the addition of
hep2,6 β-CD changes the fluorescence of these drugs, giving
evidence of drug complexation. The binding constant of SAL
and NP decreases in the presence of polymeric micelles. There-
fore, given that in dilute solutions of NP and SAL no partition
into F127 is observed (cf. section 2.1), only two binary systems are
present: drug/CD and probably F127/CD. Hence, the decrease
must be due exclusively to the binding of F127 to hep2,6 β-CD. In
these conditions (625 ( 25 (g/g) and no partition), the concentra-
tions of free β-CD determined at 2% NP are 5.1% and 7.0% for
total β-CD concentrations of 7 and 9%, respectively, and 9.0%-
11.0% free β-CD for total β-CD of 11-13%. These values could
justify the effect observed from SANS (Figure 4), since at 2% NP
the micelles are not broken up with 7-9% β-CD (i.e., below 7%
effective) and micellar break up is only observed from 11 to 13%
(above 7% effective). At 1% NP instead, for 9% total β-CD, there
is 8.0% free β-CD, thus above 7%, which could therefore explain
why the micelles are broken up (Figure 4). In this perspective, and
with this yet limited set of data, this suggests that hypothesis
(2) alone could explain the weakened disruption of the micelles by
β-CD in the presence of drug.
Therefore, although the high affinity of both drugs, NP and
SAL, to hep2,6 β-CD may decrease the disruptive action of β-CD
on the micelles at low β-CD, with further addition of β-CD
instead, the micelles may be more readily broken up by β-CD,
since the β-CD/F127 binding increases. This again demonstrates
that a fine balance of forces is involved, which controls micelle
aggregation and break up, but this also opens up very interesting
perspectives in terms of finely tuning the release of the drug at
specific sites in these ternary mixtures.
Sodium Salicylate and Sodium Naproxen. The addition of
hep2,6 β-CD to both dilute and concentrated solutions of sali-
cylate produces an increase in fluorescence intensity (Figure 7).
However, neither the emission maxima position nor the shape of
the spectrum is modified.
Conclusion
Addition of hep2,6 β-CD to solutions of NP also produces
changes in the emission spectrum, but the effect is dependent on
drug concentration. In the dilute system, an increase in fluores-
cence is observed, besides the appearance of an iso-emissive point
around 335 nm and a blue shift from 355 to 350 nm. A blue shift
has been reported when the drug is in apolar solvents.53 In the
concentrated drug systems, addition of hep2,6 β-CD leads to an
increase in the monomer emission (355 nm), with a simultaneous
decrease in the excimer emission band (440 nm). All these changes
In this contribution, we have used a combination of SANS and
fluorescence spectroscopy to gain more insight into the complexa-
tion behavior of the Pluronic F127, heptakis(2,6-di-O-methyl)
β-cyclodextrin, and four selected drugs (namely, lidocaine (LD),
pentobarbital (PB), sodium naproxen(NP), and sodium salicylate
(SAL)) in binary and ternary mixtures. In particular, we have
attempted to explain the structural micellar changes observed in
the SANS patterns by an analysis of the interactions involved
(reflected by binding constants and partitioning coefficients), the
(55) Aki, H.; Niiya, T.; Iwase, Y.; Yamamoto, M. J. Pharm. Sci. 2001, 90, 1186–
1197.
~
(56) Junquera, E.; Pena, L.; Aicart, E. J. Pharm. Sci. 1998, 87, 86–90.
10570 DOI: 10.1021/la100596q
Langmuir 2010, 26(13), 10561–10571

Valero and Dreiss
Article
consideration of the chemical structure of the drugs, and the
modifications occurring in their microenvironment upon addition
of polymer, β-CD, or both. The main findings are summarized
below.
above); however, at a higher drug concentration (2% NP), gra-
dual addition of β-CD led to micellar growth, up to 11% β-CD
where micellar breakup occurred. These effects are attributed to a
preferential β-CD/drug complexation over β-CD/polymer inter-
action, as deduced from the binding constants. However, it is clear
that a fine-tuning of the component ratio controls the balance of
forces and the mechanism of interaction. In the case of PB, the
results suggest that the effect is not exclusively related to the
partitioning of the drug inside the micelles or its binding to β-CD,
but the presence of the drug in the aggregate also modifies the
polymer/β-CD binding, contributing to the overall effect. This
study is a first attempt to shed light on the nature of these complex
interactions, but further work should enable us to understand
better the balance of forces, in order to be able to control and
adjust the behavior for targetted drug delivery.
The binding constants of the drugs to the Pluronic micelles,
determined by fluorescence spectroscopy, were found to be dir-
ectly correlated to the amount of charge, following NP ≈ SAL<
PB <LD. All drugs were found to lie at the micellar interface,
with a slightly deeper localization of lidocaine (LD) inside the
micelles. The hydrophobic character (and absence of very hydro-
philic groups) of LD gave rise to a substantial growth of the
micelles, while charge repulsions in NP and SAL led to the shrin-
king of the micelles. PB was found to adopt a more superficial
position, with the hydrophobic substituents inside the apolar
region of the micelle, the polar ring protruding into the palisade
zone and probably into the solution; taking into account the size
of both parts of the drug, its interaction with the micelle can be
defined as “anchored” to the micelle through its lateral chains,
therefore affecting only marginally the overall structure of the
aggregates.
Acknowledgment. M.V.J. wishes to thank Junta de Castilla
ꢀ
y Leon, Spain, for supporting her stay in London. ISIS and ILL
are acknowledged for provision of beam time.
All drugs were found to form complexes with β-CD; the strength
of the binding constant (at low drug concentration) followed the
order: LD < PB < SAL < NP. While hep2,6 β-CD has been
shown to gradually break up F127 micelles through β-CD/
polymer complexation, this effect is weakened in the presence
of drug, provided the drug is above a critical concentration (∼2%
PB with 5% F127). In the case of NP, a low drug concentration
(1%) did not alter micellar breakup by β-CD (from 9% β-CD and
Supporting Information Available: Graphs showing the
fluorescence spectra of lidocaine and pentobarbital at vary-
ing pH and in a range of solvents (hexane, dichloromethane,
chloroform, methanol, acetone, acetonitrile, ethanol, propa-
nol, butanol, octanol, and water) and a table showing the
position of emission maxima of both drugs in these solvents.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Langmuir 2010, 26(13), 10561–10571
DOI: 10.1021/la100596q 10571