Synthesis, characterization and evaluation of computationally designed nanoparticles of molecular imprinted polymers as drug delivery systems

Article abstract of DOI:10.1016/j.ijpharm.2011.12.054

The aim of the present study was to prepare nanoparticles of molecular imprinted polymers (MIPs) with high loading capacity for naltrexone as template drug. To achieve this goal, a computational protocol was employed to select the most appropriate monomer for MIP preparation. Density functional theory (DFT) method at the B3LYP level of theory in conjugate with the 6-31+G(d) basis set was used to evaluate the extent of interaction between naltrexone and a small library of frequently used vinylic monomers. The results revealed that acrylic acid (AA) and methacrylic acid (MAA) can be considered as suitable monomers. To select the best monomer, two MIPs with AA and MAA monomer were synthesized and their loading capacity, selectivity and release profile were evaluated. The experimental results showed that the MIPs synthesized using AA (MIP-AA) exhibited a surprisingly high loading capacity to naltrexone (75 mg of drug/g of MIP) compared to MIP-MAA (34 mg of drug/g of MIP). In vitro release dynamics of the drug from MIPs was also investigated and modeled. It was found that non-Fickian-type diffusion mechanism was responsible for drug release. The results can lead to the conclusion that MIPs designed by computational approach can be considered as promising candidates for drug delivery systems.

Full text of DOI:10.1016/j.ijpharm.2011.12.054

International Journal of Pharmaceutics 424 (2012) 67–75
Contents lists available at SciVerse ScienceDirect
International Journal of Pharmaceutics
journal homepage: www.elsevier.com/locate/ijpharm
Pharmaceutical Nanotechnology
Synthesis, characterization and evaluation of computationally designed
nanoparticles of molecular imprinted polymers as drug delivery systems
K. Rostamizadeha,∗, M. Vahedpourb, S. Bozorgib
a Department of Medicinal Chemistry, School of Pharmacy, Zanjan University of Medical Sciences, Zanjan, Iran
b Department of Chemistry, Zanjan University, 45195-313, Zanjan, Iran
a r t i c l e i n f o
a b s t r a c t
Article history:
The aim of the present study was to prepare nanoparticles of molecular imprinted polymers (MIPs) with
high loading capacity for naltrexone as template drug. To achieve this goal, a computational protocol
was employed to select the most appropriate monomer for MIP preparation. Density functional theory
(DFT) method at the B3LYP level of theory in conjugate with the 6-31+G(d) basis set was used to evaluate
the extent of interaction between naltrexone and a small library of frequently used vinylic monomers.
The results revealed that acrylic acid (AA) and methacrylic acid (MAA) can be considered as suitable
monomers. To select the best monomer, two MIPs with AA and MAA monomer were synthesized and
their loading capacity, selectivity and release profile were evaluated. The experimental results showed
that the MIPs synthesized using AA (MIP–AA) exhibited a surprisingly high loading capacity to naltrexone
(75 mg of drug/g of MIP) compared to MIP–MAA (34 mg of drug/g of MIP). In vitro release dynamics of
the drug from MIPs was also investigated and modeled. It was found that non-Fickian-type diffusion
mechanism was responsible for drug release. The results can lead to the conclusion that MIPs designed
by computational approach can be considered as promising candidates for drug delivery systems.
© 2011 Elsevier B.V. All rights reserved.
Received 10 September 2011
Received in revised form
20 November 2011
Accepted 27 December 2011
Available online 2 January 2012
Keywords:
Molecularly imprinted polymers (MIP)
Density functional theory (DFT)
Drug delivery systems
Naltrexone
1. Introduction
polymerized in the presence of a template molecule. The template
is then extracted, leaving sites which are complementary in both
Naltrexone is an efficient narcotic antagonist used mainly in the
treatment of narcotic addiction. It is currently given orally as tablets
or capsules in a daily dose of 50 mg. There are a number of issues
associated with administration of naltrexone hydrochloride includ-
ing bioavailability and patient compliance. On the other hand, for
naltrexone treatment to be effective, it is necessary to provide a
sufficient level of the drug concentration for a long period of time.
To overcome these problems, there is a need for development of
novel drug delivery systems capable of delivering such drugs more
effectively and efficiently. Therefore, substitution of common drug
delivery systems has attracted significant attention of the scientists
working in the pharmaceutical fields (Asadi et al., 2011a,b; Dong
and Feng, 2004). Molecular imprinting polymers (MIPs) which pos-
sess great potential in a number of applications in the life sciences
represent a promising approach as drug carriers (Yin et al., 2006).
MIPs are known as smart materials with a pre-determined selec-
tivity. To prepare MIPs, a functional monomer and a crosslinker are
chemical functionality and shape to those of the template.
The application of MIPs in different drug delivery systems
has dramatically expanded in the past decade (Piletsky et al.,
2006; Zhao et al., 2009; Singh and Chauhan, 2008; Cirillo et al.,
2009; Chianella et al., 2006). Alvarez-Lorenzo et al. (2006) have
imprinted norfloxacin into soft contact lenses prepared from
poly(hydroxyethylmethacrylate)-based hydrogels. The hydrogels
showed controlled release of norfloxacin for more than 24 h.
The effectiveness of MIPs can be improved by increasing
their loading capacity and selectivity. Since an optimal template–
functional monomer interaction has a determining impact on
successful imprinting, the choice of appropriate monomers play
a critical rule in successful imprinting process (Hiratania and
Alvarez-Lorenzo, 2004). The best functional monomers generally
are selected by trial and error method which is expensive and
time consuming. Nowadays, with the appreciation of the advent of
computational chemistry, computer-aided study of MIPs has been
suggested as a rational and fast technique to search for optimal
imprinting conditions (Gholivand et al., 2010; Nicholls et al., 2009).
Li et al. (2009) used molecular dynamics simulations and computa-
tional screening to identify functional monomers capable of strong
interaction with sulfadimidine as template. The combination of
molecular dynamics simulations and quantum mechanics calcu-
lations were also used to select the optimum monomer and progen
∗
Corresponding author at: Department of Medicinal Chemistry, School of Phar-
macy, Zanjan University of Medical Sciences, Postal Code 45139-56184, Zanjan, Iran.
Tel.: +98 241 4248876; fax: +98 241 4248876; mobile: +98 9144199104.
E-mail addresses: rostamizadeh@zums.ac.ir, Rostamizadeh@gmail.com
(K. Rostamizadeh).
0378-5173/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpharm.2011.12.054

68
K. Rostamizadeh et al. / International Journal of Pharmaceutics 424 (2012) 67–75
solvent (Dong et al., 2009; Chianella, 2006). Chianella et al. (2006)
have demonstrated that with the aid of computational design, it is
possible to prepare MIPs with high binding capacity for Abacavir,
which is a HIV-1 reverse transcriptase inhibitor. Leapfrogk algo-
rithm was used to rational design of MIPs to be employed as carriers
for controlled release of simazine into water (Piletska et al., 2005).
It was found that the speed of herbicide release was correlated with
the calculated binding characteristics.
The present research was focused on the development of a fast
and reliable screening method to aid the design and synthesis of
nanoparticles of MIP with high loading capacity and selectivity for
naltrexone. Nanoparticles properties as drug delivery vehicles were
also evaluated.
As a reference, non-imprinted particles (NIPs) of each monomer
were also prepared without naltrexone by using the same synthetic
protocol.
2.4. MIPs characterization
The structure of MIPs was characterized using FT-IR spec-
troscopy (Bruker, Tensor 27, Germany). The hydrodynamic
diameter and the zeta potential (ꢀ) of nanoparticles were deter-
mined in phosphate-buffer saline (PBS, pH 7.4) by dynamic
light scattering using a NanoZS apparatus (Malvern Instruments,
Worcestershire, UK) equipped with a 633 nm laser at a fixed scat-
tering angle of 173◦. The temperature of the cell was kept constant
at 25 ◦C.
2. Experimental
2.5. Evaluation of loading capacity of MIP–AA and MIP–MAA
2.1. Materials and methods
The loading capacity of MIP–AA and MIP–MAA were assayed
through analysis of naltrexone concentration in the soxhelt solu-
tion by UV spectrophotometer at ꢁ = 280 nm. The loading capacity
of MIPs (Q, mg/gMIP) was calculated based on the concentration of
naltrexone in the soxhelt solution (c, mg ml−1), the volume of sox-
helt solution (v, ml) and the weight of the MIPs (w, g) according to
the following Eq. (1):
Azobisisobutyronitrile (AIBN) was supplied by Kemikalieimport
and all other chemicals were purchased from Merck (Germany) and
used as received. Naltrexone hydrochloride was supplied by Sun
Pharmaceuticals Ltd. (India).
Due to solubility problems in most organic solvents, naltrexone
hydrochloride was converted into the free base by diethyl ether
extraction of an aqueous solution after addition of Na2CO3. The
resulting naltrexone free base was then used as template for the
synthesis of MIPs.
CV
Q =
(1)
W
The computer used to simulate monomer–template interactions
was a Pentium IV 1.2 GHz, 256 MB of RAM, running a Windows 2007
operative system.
In order to avoid the effect of presence of unreacted monomer
and byproducts of the polymer synthesis in the absorbance change,
the amount of naltrexone was calculated from the absorbance of
the solution after subtracting the absorbance of the supernatant
produced of NIPs in the same conditions.
2.2. Computational modeling
The aim of calculations was to compare some monomers
interaction with naltrexone in order to search for the optimal
naltrexone–monomer complex for MIP synthesis. To achieve this
goal, three methods were considered: the first was geometrical
parameters of complex and its shift in comparison with the isolated
naltrexone and monomers. The second was the relative stabil-
ity of complex and the third was the atom in molecules analysis
(AIM).
To study the kinetic of template extraction, naltrexone was
removed using acetic acid/methanol by batch method.
2.6. Selectivity studies
To estimate the selectivity of the naltrexone imprinted particles,
the loading capacity of MIPs toward methadone was also studied.
Methadone was chosen due to their molecular structures simi-
lar to naltrexone to some extent and its ability to form hydrogen
bond with MIPs. The procedure followed for binding experiment of
methadone was similar to that of naltrexone.
2.3. Synthesis of molecular imprinted PAA (MIP–AA) and PMAA
(MIP–MAA) nanoparticles
Particles of MIPs were prepared using precipitation polymer-
ization. Briefly, to prepare MIP–AA, naltrexone (0.0341 g), AA
monomer (0.041 ml), and ethylene glycol dimethacrylate (EGDMA)
as cross linker (molar ratio of functional monomers to cross-linker
was 1:5.4) were dissolved in 3 ml of THF in three neck flask. The
solution was stirred slightly for 4 h at 2 rpm to allow monomers
and template molecules to form complex, and then AIBN (0.041 g)
as initiator was added. The solution finally syringed to 25 ml water.
The mixture was saturated with dry nitrogen and polymerization
was carried out by placing the flask in an oil bath for 24 h at 90 ◦C.
The particles were collected using a centrifuge at 15,000 rpm for
20 min and the process was followed by a final rinsing step with
water and acetone, respectively. The nanoparticles were then dried
and loaded naltrexone was extracted from MIPs through soxhelt
method using a mixture of acetic acid and methanol (1:1) as sol-
vent for 24 h to ensure that the drug is nearly completely removed
from the polymer matrix. Synthesis of MAA based MIPs (MIP–MAA)
was carried out using the same procedure stated above with MAA
monomer (0.049 ml).
2.7. Drug reloading to the MIPs by soaking procedure
In order to investigate the reloading capacity of MIPs, naltrex-
one was reloaded to the soxhelted MIPs by soaking method. As
brief, after removal of the naltrexone from the MIPs by soxhelt
method, they were dried at room temperature. 0.0350 g of MIPs
were immersed in 25 ml of phosphate buffer with a known con-
centration of naltrexone (0.00239 M) and soaked for 1 day at room
temperature while stirring. To study kinetic of naltrexone reload-
ing on MIPs, sampling was carried out with 1 h interval. Samples
were centrifuged for 15 min (15,000 rpm) and the naltrexone con-
centration in the supernatant was monitored. Any particles present
in the samples withdrawn were returned to the medium after cen-
trifugation and analysis of the drug concentration. NIPs were also
subjected to the reloading study with equal initial concentration
and condition. The amount of naltrexone reloaded to the poly-
mer was calculated by subtracting the amount of free drug from
its initial value.

K. Rostamizadeh et al. / International Journal of Pharmaceutics 424 (2012) 67–75
69
2.8. In vitro release study
3. Results and discussion
Release studies were carried out for both drugs entrapped dur-
ing MIPs synthesis and for drugs reloaded to the soxhelted MIPs
in water media (phosphate buffer). The nanoparticles were dis-
persed in flasks containing phosphate buffer (pH 7.2) at 37 ◦C for
24 h with shaking. These conditions were maintained throughout
the experiment. Samples were withdrawn from solution medium
at appropriate time intervals to determine the amounts of drug
released.
3.1. Computational selection of the functional monomer
One of the most important advantages of MIP as drug deliv-
ery systems is the possibility to enhance loading capacity and
selectivity by selection of a monomer with high affinity toward
template. The computational design is a promising approach for
selection of adequate monomers for MIP. In this work, five func-
tional monomers Acrylamide (AAM), methylacrylamide (MAAM),
MAA, allylamine, and AA were theoretically selected as possible
functional monomers. Monomers selection was based on their
functionalities able to form non-covalent interactions especially H–
bonding with naltrexone.
2.9. Mechanism and mathematical modeling of drug release from
MIPs particles
Geometries of monomers, naltrexone and their correspond-
ing complexes were fully optimized using the density functional
theory (DFT) method at the B3LYP level of theory in conjugate
with the 6-31+G(d) basis set (Becke, 1993; Lee et al., 1988). All
of the calculations were performed with the Gaussian 03 pro-
gram (Frisch et al., 2003). First, semiemprical and Hartree–Fock
(HF) methods were employed to optimize the complexes as
an initial guesses. The obtained results were improved using
the B3LYP calculation method and used to calculate the rel-
ative energy of collisional complexes. Using such level of the
method gives the opportunity to generate a wave function in
a suitable form to execute the topological analysis atom in the
molecule by means of the AIM2000 series program (Bader, 1990;
Biegler-Kônìng et al., 2000). Finally, the counter poise (CP) pro-
cedure was used to correct the interaction energy for basis set
superposition error (BSSE). The optimized molecular structure of
naltrexone, the five functional monomers, and the most stable com-
plexes of naltrexone–AA, naltrexone–MAAM, naltrexone–MAA,
naltrexone–AAM, and naltrexone–Allylamide is presented in Fig. 1
and Fig. 1S in supplementary data.
There is a number of reports deal with the mathematical model-
ing of drug release from polymeric systems. The following equation
is one of the generalized empirical equations which is widely used
to describe drug release (Ritger and Peppas, 1987).
Mt
= ktn
(2)
M
∞
where Mt/M is the fractional release of drug at time t (Mt) and the
∞
drug released at equilibrium time (M ), k is the constant character-
∞
istic of the drug–polymer system, and n is the diffusion exponent
characteristic of the release mechanism. The slope and the inter-
cept of the plot drawn between ln Mt/M and ln t give the value
∞
of n and k, respectively. Normal Fickian diffusion is characterized
by n = 0.5 and Case II diffusion by n = 1.0. A value of n between 0.5
and 1.0 indicates a mixture of Fickian and Case II diffusion, which
is usually called non-Fickian, or anomalous, diffusion.
The binding characteristics of MIPs were also studied using
different adsorption isotherm models including Hguchi model,
Ritger–Peppas model, Peppas–Sahlin model, Ritger–Peppas model
(m = 0.5), and Zero-order.
Table 1
Bond length, bond angles and topological analysis of AA, naltrexone, and their complexes.
Species
ꢂ(r) (eÅ−3
)
ꢀꢂ(r)(eÅ−3
)
Bond lengths(Å)
Ellipticity
Acrylic acid and naltroxone complex
H56–O52
H56–O18
C17–O18
C17–C14
C14–O15
O53–O15
C8–O15
0.2959
0.04014
0.3589
0.2372
0.2077
0.006104
0.2394
0.2952
0.2733
0.3036
0.04426
0.36203
0.2580
0.2863
0.004988
0.005274
−1.3728
0.1343
1.0021 (0.0212)
1.7540
1.2465 (0.0086)
1.5315 (−0.0083)
1.4933 (−0.0094)
3.1665
1.4233 (0.0139)
1.3957 (0.0067)
1.3656 (−0.0215)
0.9966 (0.0178)
1.6794
1.2497 (0.0116)
1.4756 (0.0056)
1.3475 (−0.0367)
0.01193
0.01215
0.07287
0.05550
0.02998
0.3733
−0.2805
−0.4257
−0.2305
0.02358
−0.3846
−0.6835
−0.5153
−1.4179
0.16815
−0.4720
−0.5300
−0.5416
0.02492
0.02332
0.03566
0.2104
C8–C9
C9–O3
O3–H30
0.01549
0.01708
0.01342
0.00775
0.06689
0.05053
−1.7484
−1.6471
H30–O53
O53–C50
C50–C49
C50–O52
6-RCP
8-RCP
Naltrexone
H30–O3
0.3267
0.2647
0.3081
0.2482
0.2133
0.2378
0.3807
−1.5554
−0.3579
−0.7545
−0.4236
−0.2627
−0.4249
−0.1561
0.9788
0.01908
0.001896
0.2210
0.02904
0.03683
0.05395
0.03485
C9–O3
C9–C8
C8–O15
1.3871 [1.36]
1.3891 [1.38]
1.4094 [1.36]
1.5027 [1.44]
1.5398 [1.53]
1.2379 [1.21]
C14–O15
C14–C17
C17–O18
Acrylic acid
H56–O52
C50–O52
C50–C49
C50–O53
0.3127
0.2954
0.2549
0.3829
−1.45332
−0.4331
−0.5135
−0.4768
0.9809 [0.952]
1.3842 [1.355]
1.4700 [1.476]
1.2381 [1.220]
0.017843
0.076494
0.06662
0.02584
The extent of bond lengths in pretenses refers to deviation bond length of complex from corresponding bond in monomer or naltrexone and the value in bracket refers to
experimental value.

70
K. Rostamizadeh et al. / International Journal of Pharmaceutics 424 (2012) 67–75
Fig. 1. Optimized molecular structure of acrylic acid, naltrexone and their complex: bond lengths presented in Table 1.
3.1.1. Molecular structure calculations
exhibited 20.694, 21.562, 20.507, 20.106, and 17.184 kcal/mol less
energies than that of the original reagents (Table 2).
The most relevant calculated bond lengths of the monomers,
naltrexone and their complexes are listed in Table 1. The bond
lengths of some species have been compared with the experi-
mental values. The comparison shows that the bond lengths of all
molecules calculated at B3LYP level are relatively in good agree-
ment with the experimental data.
The calculation showed variation in the AA–naltrexone com-
plex bond lengths compared to the free AA and naltrexone which
is indicative of the complex formation. For instance, variation in
H56–O52 and C50–O53 bond lengths of acrylic acid and its com-
3.1.3. Atom in molecules theory (AIM)
Theory of AIM developed by Bader is based on the CP of the
molecular electronic charge density, ꢂ(r). It has been proven that
AIM provides valuable information about many different chemical
systems by analysis of the molecular electron density distribution.
The application of the atoms and molecules theory to understand
the nature of the bonds in deeper detail is an interesting approach.
Atoms in molecules theory is based on the CP of the molecular elec-
tronic charge density, ꢂ(r). These are points where the electronic
˚
plex with naltrexone were calculated to be 0.0212 and 0.1094 A,
respectively (Table 1). This difference can be explained by the new
O18–H56, O53–O15, and H30–O3 bonds which form two 6 and 8
member ring structure. Since the lengths of newly formed bonds
were longer than the covalent bonds, they seem to be van der Waals
bonds. Similarly, calculations demonstrated the complex formation
between other monomers with naltrexone.
Table 2
The total energy and relative energy of species obtained at B3LYP level.
Total energy
(Hartree)
Relative energy
(kcal/mol)
Naltroxene–AA complex
Naltroxene–MAA complex
Naltroxene–allylamine complex
Naltrexone–AAM complex
Naltrexone–MAAM complex
Naltrexone
AA
MAA
Allylamine
AAM
−1398.31748
−1304.44298
−1437.62819
−1378.494385
−1417.781117
−1131.24041
−267.07707
−20.694
−21.562
−20.507
−20.106
−17.184
3.1.2. Relative stability of the complexes
Table 2 presents the total energies and relative energies of com-
plexes with respect to the monomers and template at the B3LYP
level of computation. To perform calculation, only one stable van
der Waals complex formed through the interaction of two or three
atoms of each monomer with appropriate atoms of naltrexone was
considered. The results revealed that the obtained collisional com-
plexes of naltrexone with AA, MAA, allylamine, AAM, and MAAM
−306.38778
−173.20257
−247.2219297
−286.513319
MAAM

K. Rostamizadeh et al. / International Journal of Pharmaceutics 424 (2012) 67–75
71
density gradient (ꢀ2ꢂ(r)) vanishes and which are characterized by
the three eigenvalues (ꢁi, i = 1, 2, 3) of the Hessian matrix of ꢂ(r). The
first two eigenvalues correspond to the perpendicular curvatures
and the latter provides curvatures along the internuclear axis. The
CP’s are labeled (r, s) according to their rank, r, (number of non-zero
eigenvalues) and signature s (the algebraic sum of the signs of the
eigenvalues).
The ε is a measure of the ratio of the rate of density decrease
in the two directions perpendicular to the bond path at the BCP,
the general shape of the bond and the degree of ␲-character. A
value of zero indicates a symmetrical distribution of density about
the bond path, such as found in standard single and triple bonds,
while large values indicate a preference for density build up in a
particular orientation. The values of the ε for AA–naltrexone are
listed in Table 1.
For the O53–O15 and H52–H30 bonds, high values of ε ascribe
to the symmetrical van der Waals bonds of the coordination cites
(Table 1 and Fig. 1S in supplementary data) and also, high value of
ε for C8–C9 bonds in various species show high ␲ bonding char-
acter which are in consistent with the experimental predictions.
Whereas low values of ε belong to the other bonds especially hetro-
atome containing newly bonds with the asymmetrical distribution
of electron charge density or low ␲ bonding character.
In essence, computational results indicate that the acid form
monomers interact with the naltrexone stronger than the amine
form monomers. So it can be concluded that the AA and MAA
monomers are more favorable monomers for MIP preparation than
the others. On the other words, the MIP synthesized with AA and
MAA are expected to possess the highest loading capacity and
selectivity to naltrexone compared to the MIP prepared with other
monomers. To identify the most appropriate monomer among AA
and MAA to prepare MIP, experimental approach was considered.
Four types of CP are of interest in molecules: (3, −3), (3, −1),
(3, +1) and (3, +3). A (3, −3) point corresponds to a maximum in
ꢂ(r) characterized by ꢀ2ꢂ(r) < 0 and occurs generally at the nuclear
positions. A (3, +3) point indicates electronic charge depletion and
it is characterized by ꢀ2ꢂ(r) > 0. It is known as bond critical point
(BCP). The (3, +1) points, or ring critical points (RCP), are saddle
points. Finally, a (3, −1) point or bond critical point, BCP, is gener-
ally found between two neighboring nuclei indicating the existence
of a bond between them. BCPs and RCPs were analyzed in this work.
The trajectories of ꢂ(r) that start at BCP and terminate at such nuclei
define a ‘bond path’. Then, a ‘molecular graph’ is the network of
bond paths and it is interesting to see that the molecular graph for
a molecule at equilibrium geometry is identified with the corre-
sponding network of chemical bonds. The charge density along a
bond path attains its minimum value at the BCP and the associated
curvature or eigenvalue of the Hessian of ꢂ at r, ꢁ3, is thus positive.
On the other hand, the charge density in an interatomic surface
attains its maximum value at the BCP and two directed long axes
perpendicular to the bond path are thus negative.
The Laplacian ꢀ2ꢂ(r) is associated with the charge density and
it is the sum of the curvatures in the electron density along any
orthogonal coordinate axes at the point (r). The sign of ꢀ2ꢂ(r)
indicates whether the charge density is locally depleted (2ꢂ(r) > 0)
or locally concentrated (ꢀ2ꢂ(r) < 0). Thus, when the negative cur-
vatures dominate at the BCP, the electronic charge is locally
concentrated within the region inter atoms leading to an interac-
tion named as covalent or polarized bonds and being characterized
by large ꢂ values, ꢀ2ꢂ(r) < 0 and |ꢁ1|/ꢁ3 > 1. On the other hand,
if the positive curvature is dominant, the electronic density is
locally concentrated in each of the atomic basins. The interac-
tion is now referred to as a closed-shell and it is characteristic
of highly ionic bonds, hydrogen bonds and van der Waals inter-
actions. It is characterized by relatively low ꢂ values, ꢀ2ꢂ(r) > 0
and |ꢁ1|/ꢁ3 < 1.
3.2. MIPs synthesis and characterization
On the basis of theoretical calculations, two monomers (AA,
MAA) strongly interacting with the template were identified and
used for the synthesis of nanoparticles of MIPs. As a matter of
fact, two sets of molecular imprinted and non imprinted poly-
mers (AA–MIP, AA–NIP and MAA–MIP, MAA–NIP) were prepared
under similar condition. MIPs were characterized by FTIR. For
instance, Fig. 2 illustrates the FTIR spectrum of MIP–AA, naltrexone,
NIP–AA and physical mixture of NIP–AA and drug. FTIR spectrum
of MIPs–AA shows a broad absorption band at 3445.1 cm−1 which
is due to –OH stretching and indicating the acid functional group
in this polymer. The infrared absorption bands at 1734.4 cm−1 due
to C O stretching of the acid and at 1161.4 cm−1 due to the C–O
stretching of acid and the CH2 asymmetric stretching vibration
at 2982.2 cm−1 were also observed in the spectra. One surprising
observation was that there was no significant difference between
MIP and NIP spectrum and the absorption bands corresponding to
the naltrexone were not observed in the MIP spectra. A possible
explanation for this phenomenon could be the low amount of the
drug compared to the polymer. In order to confirm this assumption,
physical mixtures of drug and NIP were also prepared by mixing the
powders together. The drug/NIP ratio in the physical mixtures was
the same as in the MIP. As it can be seen from Fig. 2, the absorp-
tion bands corresponding to naltrexone in the physical mixture
spectrum were also not detectable.
Other interesting parameter is the ellipticity (ε) defined as
(ꢁ1/ꢁ2) − 1. It is indicative of the similarity between the perpen-
dicular curvatures (ꢁ1 and ꢁ2) at the BCP. In terms of the orbital
model of electronic structure, the ε provides a quantitative mea-
sure of the ␲-bond character and of the delocalization electronic
charge.
The topological analysis of the electronic charge density was
performed for all bonds in the complexes, monomers and nal-
trexone. Calculations exhibited that three cites corresponding
with H30–O3, O53–O15 and H56–O18 were responsible for
AA–naltrexone and MAA–naltrexone complexes formation. Table 1
and Fig. 1S in supplementary data represent the topological charac-
teristic in the BCPs and RCPs of all species involved in the study. The
data shows that the newly formed bonds strength for monomers
containing acid functional group was stronger than the monomers
with amine functional moiety. This is presumably coming from
the atoms electronegativity and complexation ability of acid and
amine groups. Among two acidic monomers, it was shown that the
newly bonds in MAA–naltrexone complex are stronger than that of
AA–naltrexone complex.
The hydrodynamic diameter of nanoparticles was 358 nm
(Fig. 3). Polydispersity index and ꢀ potential of MIP particles were
determined to be 0.86 and 6.28 mV, respectively.
3.3. Evaluation of the MIPs loading capacity (Q)
The evaluation of the MIPs loading capacity was carried out by
measuring the template extracted from MIPs. The results showed
that QMIP–AA was 75 mg naltrexone/g MIP, which was more than
that of MIP–MAA (QMIP–MAA = 34 mg naltrexone/g MIP). The kinetic
of naltrexone extraction was also monitored during extraction pro-
cess by batch method (Fig. 4). As it is clear the extraction of template
using acetic acid/methanol was almost complete in 6 h, as there
is no change in concentration of naltrexone. As it can be seen
According to the theory of AIM, the positive value of ꢀ2ꢂ(r) for
the newly formed bonds describe the intermolecular interaction
character. All complexes showed RCPs. In the acid form monomers,
two RCPs were calculated which indicate that two ring structures
are formed.

72
K. Rostamizadeh et al. / International Journal of Pharmaceutics 424 (2012) 67–75
2982.1990
3442.0942
1161.4660
3185.7592
2929.4241
2823.8743
1734.4503
1730.6806
939.0576
1236.8586
3438.3246
1455.4974
2963.3508
2961.0601
799.5812
2362.0795
1157.6963
1457.4924
805.5046
3432.8236
1738.2199
1264.0163
1158.0020
1735.7798
4000
3500
3000
2500
2000
1500
1000
500
Wavenumber cm-1
Fig. 2. FTIR spectra of MIP–AA (purple), naltrexone (red), NIP–AA (blue), and physical mixture of NIP–AA and drug (green). (For interpretation of the references to color in
this figure legend, the reader is referred to the web version of the article.)
the efficiency of MIPs–AA is quite higher than that of MIPs–MAA.
Apparently, this can be associated with the optimum association
of AA with the template. It is obvious that although the intensive
interaction of monomer with template would lead to the high incor-
poration of drugs into the MIP, but on the other hand it would
make it difficult for template to be extracted from MIP, so it can
be concluded that the best monomer is the one capable of mod-
erate interaction with template. The result was well consistent
with the computational results, which confirmed the validity of
the model and method proposed to benefit screening molecularly
Fig. 3. Hydrodynamic diameter of MIP–AA particles.
imprinted systems rapidly in an experiment-free way instead of
trial and error approach. In addition, these values are acceptable
for practical application of MIP as drug delivery device.
3.4. Evaluation of the MIPs reloading performance
The time course of reloading using the soxhelted MIPs synthe-
sized with the two monomers, AA and MAA, is shown in Fig. 5. The
reloading capacity for MIP–AA and MIP–MAA were obtained to be
250 and 120 mg naltrexone/g MIPs. A possible explanation for this
behavior could be the intensive interaction of MIPs with naltrexone
which subsequently leads to significant surface adsorption of drug
on nanoparticles. The curve of MIPs indicated that the adsorption of
naltrexone increased quickly with time and then reached equilib-
rium. In the first step, the adsorption rate was fast, and the contact
time to nearly reach equilibrium was 10, and 5 min for AA–MIP and
MAA–MIP, respectively. The observation can be explained in terms
of small size of particles which facilitates loading process.
3.5. Selectivity studies
Probably, the most important feature of MIPs is their selectiv-
ity. In order to evaluate the selectivity of the synthesized MIPs,
methadone, which is an analogue of naltrexone, to a certain extent,
and possesses capability to form hydrogen bonds with MIPs was
considered in this study (Fig. 6). As it can be seen, the binding
observed for naltrexone in MIP–AA and MIP–MAA were 2.5 and
1.3 times as high as that of methadone, respectively. This can be
Fig. 4. Naltrexone extraction kinetics from MIPs particles.
attributed, in part, to the unique shape of the template cavity

K. Rostamizadeh et al. / International Journal of Pharmaceutics 424 (2012) 67–75
73
Fig. 5. Reloading kinetics of naltrexone into MIPs and NIPs particles.
created on the MIP particles which play an important role in the
template selective binding to the MIPs. On the other words, the
difference in adsorption of naltrexone to MIP–AA and MIP–MAA
devices indicated that in case of MIPs–AA specific binding occurred
more than MIP–MAA. Fig. 5 reports that naltrexone binds to
naltrexone imprinted MIP–AA to a greater extent than that of
MIP–MAA but this difference was not observed for their corre-
sponding NIPs. This trend may be explained by the fact that the NIPs
adsorption arises from nonspecific adsorption of particles. Over-
all, these findings can be considered to be further evidence that
AA appear to be the most useful monomer to form stable complex
with naltrexone. This is in agreement with computational studies
demonstrating that the observed effects are a consequence of the
optimal association of drug with monomer.
Fig. 7. Release profile of naltrexone from MIPs and reloaded MIPs.
incubation in the drug solution (reloaded MIPs). The release behav-
ior of both particles was evaluated.
The profile of the release of naltrexone per gram of the MIPs
and reloaded MIPs is presented in Fig. 7. The overall amount of
drug released from MIP–AA, and MIP–MAA was 37, and 19 mg/gMIP
in about 85 h, respectively which corresponds to 49% and 56% of
total encapsulated drug of MIP–AA and MIP–MAA, respectively
(Fig. 8). The total amounts of naltrexone released from reloaded
AA–MIP, and MAA–MIP were 72.8, and 36.1 mg/gMIP in 22 h, respec-
tively. It can be seen that the amount of drug released from the
reloaded MIPs–AA was higher and faster than that of the MIP–MAA.
The increased release of naltrexone from reloaded MIPs may arise
due to higher loading of drug in this polymer, whereas in case of
reloaded nanoparticles the release was fast. It is of great importance
to note that the drug release in the MIPs was in a controlled manner.
This observations can be explained by the fact that the most loading
of reloaded MIPs comes from surface adsorption which shows burst
3.6. Release dynamics of the drug from MIPs
Two methods were followed in order to achieve the loaded MIPs:
drug incorporation during the process of generation of the nanopar-
ticles and hence trapping of the drug in the polymeric matrix
(MIPs), or adsorption of naltrexone into soxhelted MIP particles by
Fig. 6. Comparison loading capacity of MIPs toward naltrexone and its analogue,
Fig. 8. comparison of the percent of naltrexone release from MIPs and reloaded
methadone.
MIPs.

74
K. Rostamizadeh et al. / International Journal of Pharmaceutics 424 (2012) 67–75
Fig. 9. Plot for the evaluation of the diffusion exponent n and the MIPs characteristic
constant k.
release, but in case of MIPs, naltrexone incorporated in the MIP
cavity were released which is controlled by different phenomenon.
In order to investigate of release mechanism of drugs from MIPs,
the values of the diffusion exponent n and the MIPs characteristic
constant k for polymers in phosphate buffer were evaluated from
the slope and intercept of the plot ln Mt/M vs. ln t (Fig. 9). It is
∞
clear that a value of n = 0.7 were obtained for both release case,
which indicates that both MIPs were quite similar in their drug
release behavior which may result from non-Fickian type diffusion
mechanism. The MIPs characteristic constant k for both MIPs was
calculated to be 5.6.
3.7. Modeling release kinetics
The binding characteristics of MIPs were also studied using
different adsorption isotherm models: including Hguchi model,
Ritger–Peppas model, Peppas–Sahlin model, Ritger–Peppas model
(m = 0.5), and Zero-order. These mathematical models are valid only
for the first 60% of the drug release. To distinguish the models that
described the data properly from those that did not fit the data
correctly, the sum of the squared residuals (SSR) was obtained.
The model that best explains the experimental data is the one
that shows the minimal value for the SSR. However, since a larger
number of model parameters could lead to a higher probability of
obtaining a smaller SSR value, it was necessary to use a discrimina-
tory criterion that was independent of the number of parameters
that each model had. For this reason, the Akaike Information Crite-
rion (AIC) was applied. The AIC can be defined as
AIC = N(ln SSR) + 2p
where N is number of experimental data points, SSR is the sum
of the squared residuals and p is the number of parameters. The
model that provides the best fit for the experimental data is selected
based on the smallest value for the AIC. Additionally, the fit of the
predicted curve to the experimental data and the validity of the cal-
culated parameters were also examined. The findings showed that
MIPs prepared gives rise to a binding isotherm that is accurately
modeled by the Ritger–Peppas model (Table 3).

K. Rostamizadeh et al. / International Journal of Pharmaceutics 424 (2012) 67–75
75
4. Conclusion
Chianella, I., Karim, K., Piletska, E.V., Preston, C., Piletsky, S.A., 2006. Computational
design and synthesis of molecularly imprinted polymers with high binding
capacity for pharmaceutical applications-model case: adsorbent for abacavir.
Anal. Chim. Acta 559, 73–78.
Cirillo, G., Iemma, F., Puoci, F., Parisi, O.I., Curcio, M., Spizzirri, U.G., Picci, N., 2009.
Imprinted hydrophilic nanospheres as drug delivery systems for 5-fluorouracil
sustained release. J. Drug Target. 17 (1), 72–77.
Dong, Y., Feng, S.S., 2004. Methoxy poly(ethylene glycol)-poly(lactide) (MPEG-
PLA) nanoparticles for controlled delivery of anticancer drugs. Biomaterials 25,
2843–2849.
Dong, C., Li, X., Guo, Z., Qi, J., 2009. Development of a model for the rational design of
molecular imprinted polymer: computational approach for combined molecular
dynamics/quantum mechanics calculations. Anal. Chim. Acta 647, 117–124.
Frisch, MJ., Trucks, GW., Schlegel, HB., Scuseria, GE., Robb, MA., Cheeseman, JR.,
Montgomery Jr., JA., Vreven, T., Kudin, KN., Burant, JC., Millam, JM., Iyengar,
SS., Tomasi, J., Barone, V., Mennucci, B., Cossi, M., Scalmani, G., Rega, N., Peters-
son, G.A., Nakatsuji, H., Hada, M., Ehara, M., Toyota, K., Fukuda, R., Hasegawa,
J., Ishida, M., Nakajima, T., Honda, Y., Kitao, O., Nakai, H., Klene, M., Li, X., Knox,
JE., Hratchian, HP., Cross, JB., Adamo, C., Jaramillo, J., Gomperts, R., Stratmann,
RE., Yazyev, O., Austin, AJ., Cammi, R., Pomelli, C., Ochterski, JW., Ayala, PY.,
Morokuma, GA., Voth, K., Salvador, P., Dannenberg, JJ., Zakrzewski, VG., Dapprich,
S., Daniels, AD., Strain, MC., Farkas, O., Malick, DK., Rabuck, AD., Raghavachari,
K., Foresman, JB., Ortiz, JV., Cui, Q., Baboul, AG., Clifford, S., Cioslowski, J., Ste-
fanov, BB., Liu, G., Liashenko, A., Piskorz, P., Komaromi, I., Martin, RL., Fox, DJ.,
Keith, T., Al-Laham, MA., Peng, CY., Nanayakkara, A., Challacombe, M., Gill, PMW.,
Johnson, B., Chen, W., Wong, MW., Gonzalez, C., Pople, JA., 2003. Gaussian 03:
Revision B. 03, Gaussian, Inc., Pittsburgh PA.
Gholivand, M.B., Khodadadiana, M., Ahmadi, F., 2010. Computer aided-molecular
design and synthesis of a high selective molecularly imprinted polymer for
solid-phase extraction of furosemide from human plasma. Anal. Chim. Acta 658,
225–232.
Hiratania, H., Alvarez-Lorenzo, C., 2004. The nature of backbone monomers deter-
mines the performance of imprinted soft contact lenses as timolol drug delivery
systems. Biomaterials 25, 1105–1113.
The aim of the present study was to prepare nanoparticles of
molecular imprinted polymers (MIPs) with high loading capac-
ity for naltrexone as drug carrier. A computational protocol was
employed to select the best monomer for MIP preparation based
on density functional theory (DFT) method at the B3LYP level of
theory in conjugate with the 6-31+G(d) basis set. Screening the vir-
tual library of some vinilic monomers showed that acrylic acid (AA)
and methacrylic acid (MAA) were the most appropriate monomer.
The validity of computational results was verified by synthesis
of two sets of MIPs using AA, and MAA monomers. The experi-
mental results were in agreement with the computational design
and showed that the MIP synthesized using AA (MIP–AA) exhib-
ited the higher loading capacity to naltrexone (75 mg of drug/g
of MIP) compared to MIP–MAA (34 mg of drug/g of MIP). In vitro
release dynamics of the drug was also investigated and modeled.
The results suggested a non-Fickian type diffusion mechanism for
drug release. The results of the computational modeling as well as
the experimental results on the formation of the MIPs provide a
possible way to enhance the loading capacity of polymeric drug
carriers.
Acknowledgments
We are most grateful for the continuing financial support of
this research project by Zanjan University of Medical Sciences and
Zanjan University.
Lee, C., Yang, R.G., Parr, W., 1988. Development of the Colle-Salvetti conelation
energy formula into a functional of the electron density. Phys. Rev. B 37,
785–789.
Li, Y., Li, X., Li, Y., Dong, C., Jin, P., Qi, J., 2009. Selective recognition of veterinary
drugs residues by artificial antibodies designed using a computational approach.
Biomaterials 30, 3205–3211.
Appendix A. Supplementary data
Nicholls, I.A., Andersson, H.S., Charlton, C., Henschel, H., Karlsson, B.C., Karlsson, J.G.,
Mahony, J., Rosengren, A.M., Rosengren, K.J., Wikman, S., 2009. Theoretical and
computational strategies for rational molecularly imprinted polymer design.
Biosens. Bioelectron. 25, 543–552.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.ijpharm.2011.12.054.
Piletska, E.V., Turner, N.W., Turner, A.P.F., Piletsky, S.A., 2005. Controlled release of
the herbicide simazine from computationally designed molecularly imprinted
polymers. J. Control. Release 108 (1), 132–139.
References
Piletsky, S.A., Turner, N.W., Laitenberger, P., 2006. Molecularly imprinted polymers
in clinical diagnostics—Future potential and existing problems. Med. Eng. Phys.
28, 971–977.
Alvarez-Lorenzo, C., Yan˜ez, F., Barreiro-Iglesias, R., Concheiro, A., 2006. Imprinted
soft contact lenses as norfloxacin delivery systems. J. Control. Release 113,
236–244.
Ritger, P.L., Peppas, N.A., 1987. A simple equation for description of solute release.
II. Fickian and anomalous release from swellable devices. J. Control. Release 5,
37–42.
Singh, B., Chauhan, N., 2008. Preliminary evaluation of molecular imprinting of
5-fluorouracil within hydrogels for use as drug delivery systems. Acta Biomate-
rialia 4, 1244–1254.
Yin, J., Wang, S., Yang, G., Yang, G., Chena, Y., 2006. Molecularly imprinted solid-
phase extraction for rapid screening of mycophenolic acid in human plasma. J.
Chromatogr. B 844, 142–147.
Asadi, H., Rostamizadeh, K., Salari, D., Hamidi, M., 2011a. Preparation of biodegrad-
able nanoparticles of tri-block PLA-PEG-PLA copolymer and determination of
factors controlling the particle size using artificial neural network. J. Microen-
capsul. 28 (5), 406–416.
Asadi, H., Rostamizadeh, K., Salari, D., Hamidi, M., 2011b. Preparation and character-
ization of tri-block poly (lactide)- poly (ethylene glycol)- poly (lactide) nanogels
for controlled release of naltrexone. Int. J. Pharm. 416, 356–364.
Bader, R.F.W., 1990. Atoms in Molecules–A Quantum Theory. Oxford University
Press, Oxford.
Zhao, W.F., Fang, B.H., Li, N., Nie, S.Q., Wei, Q., Zhao, C.S., 2009. Fabrication of pH-
responsive molecularly imprinted polyethersulfone particles for bisphenol-a
uptake. J. Appl. Polym. Sci. 113, 916–921.
Becke, A.D., 1993. A new mixing of Hartree-Fock and local density-functional theo-
ries. J. Chem. Phys. 98, 1372–1377.
Biegler-Kônìng, FJ., Derdau, R., Bayles, D., Bader, RFW., 2000. AIM2000. Version 1.