Preparation and characterization of novel amphophilic hydrogels with covalently attached drugs and fluorescent markers

Article abstract of DOI:10.1021/ma102044n

This paper describes the synthesis of styrene-based macromonomers with covalently attached model drugs (ibuprofen and naproxen) or fluorescent markers (pyrene) and their incorporation into linear or hyperbranched p-(chloromethyl)styrene copolymers. Altern

Full text of DOI:10.1021/ma102044n

Macromolecules 2010, 43, 10017–10030 10017
DOI: 10.1021/ma102044n
Preparation and Characterization of Novel Amphiphilic Hydrogels with
Covalently Attached Drugs and Fluorescent Markers
Caiping Lin† and Ivan Gitsov*,†,‡,§
†Department of Chemistry, and ‡The Michael M. Szwarc Polymer Research Institute , College of
Environmental Science and Forestry, State University of New York, Syracuse, New York 13210, United States,
and §Syracuse Biomaterials Institute, Syracuse, New York 13210, United States
Received September 2, 2010; Revised Manuscript Received October 22, 2010
ABSTRACT: This paper describes the synthesis of styrene-based macromonomers with covalently attached
model drugs (ibuprofen and naproxen) or fluorescent markers (pyrene) and their incorporation into linear or
hyperbranched p-(chloromethyl)styrene copolymers. Alternatively the copolymers were produced by post-
polymerization modification of linear or hyperbranched poly[p-(chloromethyl)styrene], PPCMSt, with the
same compounds. The incorporation of these copolymers into amphiphilic conetworks was achieved by two
methods: Williamson ether synthesis between PPCMSt and poly(ethylene glycol), PEG, with hydroxyl end
groups or by nucleophilic substitution between the chloromethyl moieties in PPCMSt and the amine end
groups in poly(oxyalkylenediamine), Jeffamine. The dynamic and equilibrium swelling properties were
studied on representative Jeffamine hydrogels. The swelling studies showed that the conetworks absorb water
quickly and reach equilibrium in 1-2 h, the equilibrium swelling ratio of gels based on linear or
hyperbranched copolymer being 181-358% and 244-480%, respectively. Preliminary drug release studies
in different aqueous media showed that the release kinetics and the amount of drugs released from hydrogels
depend on the physical properties of drugs, the microstructure of polymer network, and the drug-polymer
interaction and more particularly on the hydrolysis dynamics of ester linkage between the drug and the
polymer matrix.
Introduction
solvents. In aqueous media these networks were also able to
encapsulate pH sensitive stains or model compounds and release
The development of novel techniques for drug delivery and
controlled release is part of the continuing efforts toward more
efficient therapeutic strategies.1-5 In recent years, amphiphilic
hydrogels and polymer conetworks have attracted significant
attention as promising materials for drug delivery and wide range
of other applications due to their unique structure and proper-
ties.6-14 Since the amphiphilic cross-linked materials often contain
hydrophilic and hydrophobic chains, they are able to swell in and
interact with both aqueous and organic media and under certain
conditions they can absorb increased amounts of hydrophobic
molecules, which then can be controllably released by application
of external stimulus. That is why the creation of new hydrogels of
improved binding and release capabilities continues to be extensively
pursued. Often the research has been focused on conetworks,
consisting of polymers with block structures, which are prepared
by copolymerization of comonomers with different hydrophilic/
hydrophobic balance.15 Alternatively, amphiphilic hydrogels, formed
by dendrimers16 as the cross-linking agents, have been developed
by us and other research groups.17-19 The dendritic conetworks
represent interesting substitutes of the traditional systems with
their well-defined highly branched and globular domains. In our
previous studies we have explored the synthesis of hydrogels
based on dendritic poly(benzyl ether)s and linear poly(ethylene
glycol)s, PEG.17,18 The synthetic strategy was based on the reaction
of PEG with isocyanate or epoxy end groups as the hydrophilic
component and hydrophobic dendritic poly(benzyl ethers) with
amino groups at the periphery. The hydrogels obtained had high
degree of cross-linking and swelled in both water and organic
them over a period of time,18b and thus have intriguing potential
for biomedical and biotechnological applications.
The hyperbranched polymers are imperfect analogues of
dendrimers sharing many of their unique characteristics, including
the well-defined globular shape and the presence of multiple and
modifiable surface and interior functionalities.20 These proper-
ties, combined with the much lower cost, make these macro-
molecules attractive dendrimer substitutes for use in biological
and pharmaceutical applications.
Despite their promising potential the hyperbranched structures
have been rarely used as cross-linking agents. In an interesting
series, Wooley and co-workers have prepared amphiphilic networks
by in situ cross-linking of mixtures, containing hyperbranched
fluoropolymer (HBFP) and diamino-terminated PEG.21-24 The
changes in the PEG/HBFP ratio influenced not only the compo-
sition, but also the degree of cross-linking and the topology of
resulting network. Recently we have shown that another factor,
which strongly affects the characteristic properties of the hyper-
branched conetworks, is the macromolecular architecture of the
hydrophilic and hydrophobic constituents.25
In this paper, we describe the synthesis and characterization of
novel model drug-containing p-chloromethylstyrene derivatives
and the copolymerization of these derivatives with p-chloromethyl-
styrene to afford new copolymers with covalently attached drugs.
They are further combined with PEG or poly(oxyalkylenedia-
mine), Jeffamine, to prepare amphiphilic hydrogels via Williamson
ether synthesis or nucleophilic substitution. Conetworks of dif-
ferent cross-linking densities and hydrophilicity are obtained by
varying the amount of the two polymeric components, the archi-
tecture, and the molecular weight of the poly(styrene) derivatives.
*Corresponding author.
r 2010 American Chemical Society
Published on Web 11/15/2010
pubs.acs.org/Macromolecules

10018 Macromolecules, Vol. 43, No. 23, 2010
Lin and Gitsov
The substance-release properties of the amphiphilic hydrogels are
evaluated and correlation of the observed phenomena with net-
work characteristics is made, as well.
solution was vigorously stirred under nitrogen and refluxed. The
reaction process was monitored by thin layer chromatography
(TLC) and was stopped after the spots of the starting reagents
disappeared on the TLC plate. K2CO3 was filtered off and after
solvent removal about 90% of crude mixture was obtained. The
reaction products were purified by flash chromatography with
hexane/diethyl ether mixture with increasing ether content as
follows: 32:1-16:1-8:1 and 4:1. These products were character-
ized by 1H NMR, 13C NMR, FT-IR, and elemental analysis.
Ibuprofen-Modified p-(Chloromethyl)styrene. This was pre-
pared from ibuprofen and p-chloromethylstyrene, yield 46%. Rf =
0.46 in hexane/diethyl ether (4:1). The product was a colorless
Experimental Section
Materials. Linear poly(ethylene glycol), PEG, (L-PEG 5k: Mn =
5000; Mw = 5400, Mw/Mn = 1.08) was purchased from Polymer
Source, Inc. Poly(oxyalkylenediamine), Jeffamine ED-6000 (JEFF,
nominal molecular weight 6,000) was obtained from Texaco
Chemical Company. Sodium hydride, NaH, (95%, dry), ethanol
(95%), tetrahydrofuran, THF, (99.8%), 2,20-Azobis(isobutyro-
nitrile), AIBN (98%), (s)-(þ)-4-isobutyl-R-methylphenylacetic
acid (ibuprofen) (99%), (s)-(þ)-6-methyoxy-R-methyl-2-naphtha-
leneacetic acid, (naproxen) (98%), 1-pyrenecarboxylic acid
(98%), N,N-diisopropylethylamine, DIPEA (99.5%); copper(I)
chloride (g99%), 2,20-bipyridil (g99%), and 18-crown-6 (g99%)
were all purchased from Aldrich. Chlorobenzene (99%) and
p-chloromethylstyrene (90%) were obtained from Acros. Toluene
(99.9%) from Fisher Scientific, K2CO3 (g99%, anhydrous)
from Spectrum, dichloromethane (99.9%) from J. T. Baker
and chloroform (99.8%) from Mallinckrodt were used as received.
p-Chloromethylstyrene was distilled under low pressure (0.4 mmHg,
45 °C); AIBN was purified by recrystallization in ethanol. THF
was dried over benzophenone-sodium and distilled immediately
before use. Acetone, toluene, and chlorobenzene were dried
1
liquid. H NMR (CDCl3): δ 0.93 (6H, d, 2CH3), 1.54 (3H, q,
CH3), 1.87 (1H, sept. CH), 2.47 (2H, d, CCH2Ph), 3.78 (1H, m,
PhCH2COO), 5.13 (2H, q, COOCH2Ph), 5.26 (1H, d, H(Ph)Cd
CH2), 5.74 (1H, q, H(Ph)CdCH2), 6.69 (1H, m, PhCHdCH2),
7.08-7.38 (8H, m, ArH). 13C NMR (CDCl3): δ 18.40 (CH3),
22.37, 22.39 (2CH3), 30.18 (CH), 45.02 (CHCH3), 45.13 (CCH2Ph),
66.02, 66.09 (COOCH2Ph), 114.13 and 114.22 (PhCHdCH2),
125.52, 125.79, 126.23, 127.08, 127.22, 128.02, 128.61, 129.31,
129.33, 135.59, 136.35, 136.38, 136.46 (ArC), 174.8 (ArCOO).
FT-IR, ν: 2958, 1740, 1514, 1466, 1379, 1332, 1168, 1093, 990,
911, 850, 799, 714 cm-1. Elemental analysis: C: 81.95% (theor),
81.40% (found); H: 8.13% (theor), 8.45% (found) and O: 9.92%
(theor), 10.04% (found).
Naproxen-Modified p-(Chloromethyl)styrene. This was pre-
pared from naproxen and p-chloromethylstyrene, yield 55%.
Rf = 0.33 in hexane/diethyl ether (4:1). The product was a white
crystalline substance with melting point 40 °C. 1H NMR
(CDCl3): δ 1.63 (3H, m, CHCH3), 3.92 (1H and 3H, m, CH3CH
and CH3O), 5.13 (2H, q, COOCH2Ph), 5.23 (1H, m, H(Ph)Cd
CH2), 5.70 (1H, q, H(Ph)CdCH2), 6.65 (1H, m, H(Ph)CdCH2),
7.10-7.45 (7H, m, ArH and NaH), 7.63-7.74 (3H, m, NaH).
13C NMR (CDCl3): δ 18.44, 18.51 (CH3), 45.44 (CHCH3), 55.28
(CH3O), 66.22 (COOCH2Ph), 105.54, 114.20, 118.95, 125.57,
125.83, 125.97, 126.00, 126.26, 127.12, 128.21, 128.89, 129.28,
133.68, 135.46, 135.51, 136.33, 136.37, 137.40, 137.77 (PhCH =
CH2, ArC and NaC) 157.66 (CH3O-NaC) 174.8 (NaCOO).
FT-IR, ν: 2980, 1734, 1606, 1506, 1485, 1393, 1267, 1233, 1177,
1034, 855 cm-1. Anal. Calcd: C, 79.74; H, 6.40. Found: C, 79.44;
H, 6.36.
˚
using molecular sieve (mesh size 5 A) prior to use. Deionized
(DI) water (18.3 MΩ) was purified by a Barnstead NANOPure
ultrapure water system. NaOH (99%, Mallinckrodt), was used
to prepare solutions with pH = 10.1 for the release the drugs
from hydrogels.
Instrumentation. Size-exclusion chromatography (SEC) ana-
lyses were conducted in THF using a SEC line consisting of a
Waters 510 pump, Waters U6K universal injector, an Applied
Biosystems 785A programmable UV-vis detector, and a Viscotek
Model 250 instrument for dual refractive index and viscometry
detection. The separations were achieved at 40 °C across a set of
˚
˚
three 5 μ PLgel columns (50 A, 500 A and a mixed C) from
Polymer Laboratories at eluent flow rate of 1 mL/min. The
apparent molecular weights and polydispersities of the polymers
were determined by calibration with 20 monodisperse polysty-
rene standards and software OmniSEC Ver. 3.1 from Viscotek
Corporation.
1-Pyrenecarboxylic Acid Modified p-(Chloromethyl)styrene.
This was prepared from 1-pyrenecarboxylic acid and p-chloro-
methylstyrene, yield 54%. Rf = 0.37 in hexane/diethyl ether
(4:1). The product was isolated as light yellow needle crystals
The chemical compositions of polymers, copolymers and
conetworks were determined by FT-IR and NMR. The samples
were analyzed in a powder/particle form by Magna 750 FT-IR
spectrometer (Nicolet) using a MTEC 300 photoacoustic module
in the spectral range from 500 to 4000 cm-1. 1H and 13C NMR
spectra were recorded in CDCl3 at room temperature with a
Bruker Avance 300 instrument (1H: 300 MHz).
Elemental analyses were conducted by Galbraith Labora-
tories (Knoxville, TN).
The UV spectroscopic measurements for the release of the drugs
from the hydrogels were conducted on a DU 640B UV-vis spec-
trometer (Beckman) from 200 to 800 nm with a scanning speed
600 nm/min at room temperature.
The surface morphology of the conetworks formed was investi-
gated by scanning electron microscopy (SEM) using JEOL 5800
LV SEM instrument with acceleration voltage 5-10 kV. The
swollen hydrogel samples were immersed in water for 48 h and
then quickly frozen with liquid nitrogen. The frozen hydrogels
were subjected to cryofracturing, followed by lyophilization.
The dry samples were sputter coated with palladium and gold
for SEM examination.
Synthesis. General Synthesis of p-(Chloromethyl)styrene
Derivatives with Ibuprofen, Naproxen, and Pyrene. In a 25 mL
round-bottom flask, K2CO3 (2.5 equiv), 18-crown-6 (0.2 equiv),
and a drug or a model compound (1 equiv) were added and
dissolved in acetone. Then this mixture was purged with nitrogen
and p-(chloromethyl)styrene (1.05 equiv) was added. The reaction
1
with melting point 101 °C. H NMR (CDCl3): δ 5.30 (1H, m,
H(Ph)CdCH2), 5.55 (2H, s, COOCH2Py) 5.80, (1H, q, H(Ph)Cd
CH2), 6.77 (1H, m, H(Ph)CdCH2), 7.36-7.62 (4H, m, ArH),
8.13-8.30 (7H, m, PyH). 8.68 (1H, d, PyH) and 9.31 (1H, d,
PyH). 13C NMR (CDCl3): δ 66.68, 66.82 (COOCH2Ph), 114.33,
114.45, 123.19, 124.09, 124.14, 124.83, 126.07, 126.20, 126.30,
126.47, 127.13, 127.72, 128.49, 128.58, 128.88, 129.50, 129.67,
130.33, 130.96, 131.27, 134.40, 136.35, 136.48, 138.00 (PhCH =
CH2, ArC and PyC)167.69 (PyCOO). FT-IR, ν: 3050, 1710,
1596, 1507, 1387, 1252, 1230, 1195, 1134, 1088, 1043, 849, 709 cm-1
.
Anal. Calcd: C, 86.17; H, 5.00. Found: C, 85.67; H, 4.98.
Synthesis of Linear Poly[p-(chloromethyl)styrene], LPPCMSt
by Classic Free Radical Polymerization. p-(Chloromethyl)styrene
(5 mmol) was mixed with AIBN at monomer to initiator ratios,
[M]/[I] = 6.25-100, then chlorobenzene or toluene (1.8 mL) was
added. The resulting solution was stirred in a nitrogen atmosphere
in an oil-bath at 75-80 °C with chlorobenzene as the solvent or
60 °C with toluene as the solvent. The viscosity of the solution
increased as polymerization proceeded and the color of solution
changed to a light yellow. After 24 h the polymer was precipitated
into methanol as a white powder and thenfiltered. The precipitate
was dissolved in THF and reprecipitated into methanol. The
process was repeated twice and the final product was then dried
under vacuum at room temperature (yield 72-95%). The polymers
had molecular weights averaging between 2.7 ꢀ 103 Da and

Article
Macromolecules, Vol. 43, No. 23, 2010 10019
2.04 ꢀ 104 Da as estimated by SEC. They were further character-
precipitated into methanol. The off-white powder precipitate
was dried under vacuum. SEC analyses showed that the polymers
were obtained with weight-average molecular weights between
2.4 ꢀ 103 Da and 4.5 ꢀ 104 Da. The ratio of secondary to
primary benzyl chloride groups as determined by 1H NMR was
between 0.18 and 0.33 at monomer to catalyst ratios from 5 to
20. 1H NMR (CDCl3): δ 1.55, 1.90, 2.30, 2.75, 3.0 4.50 (CH2Cl),
4.75 (CHCl) and 6.50-7.41 (ArH). FT-IR: 3023.5, 2930.6,
2856.1, 1699.3, 1605.5, 1511.6, 1487.8, 1445.1, 1423.7, 1266.4,
1
ized by FT-IR and H NMR. FT-IR: 3023.9, 2927.5, 2850.3,
1700.9, 1608.4, 1511.9, 1484.9, 1446.4, 1423.2, 1265.1, 836.9,
798.4, 709.7, 674.9, 609.4 cm-1 1H NMR (CDCl3): δ 0.94,
.
1.16, 1.46, 1.73, 4.51 (CH2Cl), 6.50 and 7.06 (ArH).
General Procedure for the Radical Homopolymerization of
p-(Chloromethyl)styrene Derivatives, Containing Ibuprofen,
Naproxen, and Pyrene. Modified p-(chloromethyl)styrene deri-
vatives and AIBN ([M]/[I] = 12.5) were added to the solvent.
The mixture solution was stirred in an oil bath at 60 °C under
nitrogen atmosphere. The viscosity of solution increased and the
color of solution became darker as polymerization proceeded.
After 24 h, the polymer formed was precipitated into methanol
as white or light yellow powder and then filtered. The precipitate
was then dissolved in THF and reprecipitated into methanol.
The isolated polymer was dried under vacuum at room tem-
perature. Finally the products were characterized by 1H NMR
and SEC.
1165.0, 1017.6, 899.6, 828.6, 795.9, 706.5, 677.0, 631.9 cm-1
.
Modification of Linear or Hyperbranched PPCMSt with
Ibuprofen, Naproxen, and 1-Pyrenecarboxylic Acid. In a 25 mL
round-bottom flask, K2CO3 (2.5 equiv), 18-crown-6 (0.2 equiv),
poly(4-vinyl benzyl chloride) (1 equiv) and drugs (ibuprofen or
naproxen) or model compounds (1-pyrenecarboxylic acid) (1.05
equiv) were added and dissolved in THF. The reaction solution
was vigorously stirred under nitrogen and refluxed at 70 °C for
24 h. Then K2CO3 was removed by filtration. The clear polymer
solution was precipitated into methanol to form a white or light
yellow powder. The precipitation was repeated twice and the
final precipitate was dried under vacuum. Yields of 90-97%
were obtained. UV-vis spectra and NMR were used to confirm
the attachment and to calculate the modification degree.
Synthesis of Hydrogels from Jeffamine and Linear or Hyper-
branched PPCMSt, with or without Attached Drugs. To a test
tube was added linear or hyperbranched PPCMSt with or without
drugs attached, Jeffamine (400, 200, or 100 wt %, relative to
PPCMSt) and dry THF (300 mg polymer/1 mL THF). The
mixture was allowed to equilibrate with a magnetic stirring at
45 °C for 0.5 h before the addition of DIEA (3 molar equivalency
with Jeffamine), followed by curing at 50 °C. After 24 h THF
was added to swell the hydrogel and the solid part was washed
with chloroform to remove the unreacted residual Jeffamine and
PPCMSt fragments. Then water was used to wash out any side
products. After that the hydrogel samples were quickly frozen in
liquid nitrogen and dried by lyophilization for 24 h. Yields of
50-85% were calculated using the total amounts of the two
original building blocks. FT-IR and swelling measurements
were conducted on representative hydrogel samples.
Synthesis of Hydrogels from PEG and Linear or Hyperbranched
Ibuprofen-, Naproxen- and Pyrene-Modified PPCMSt. Linear or
hyperbranched modified PPCMSt and PEG were mixed at
different ratios in a pear-shaped flask with two openings. Dried
THF (100 mg polymer/1 mL THF) was used as the solvent to
dissolve both reagents. The solution was flushed with nitrogen
and NaH was added to catalyze the reaction. The mixture was
stirred at room temperature under nitrogen atmosphere for 24 h,
and then ethanol was added to stop the reaction and quench the
excess NaH. The solution was centrifuged and the solid part was
washed with ethanol and chloroform to remove the unreacted
residual PEG and PPCMSt fragments. Water was used to wash
out the side product-NaCl. Hydrogels were dried using lyophi-
lization for 24-48 h. Yield (10%-90%) was calculated from the
total amounts of the two original building blocks. FT-IR and
swelling measurements were conducted on representative hydrogel
samples.
Homopolymer of Ibuprofen-Modified p-(Chloromethyl)styrene.
This was prepared from ibuprofen-modified p-(chloromethyl)-
styrene, AIBN as initiator, and toluene as the solvent (yield
1
67.5%). H NMR (CDCl3): δ 0.88 (2CH3), 1.45 (CH3), 1.86
(CH), 2.42 (CCH2Ph), 3.75 (COOCH2Ph), 4.32, 4.89, and 6.69
(-CH2-CH-), 6.87, 7.04, and 7.17 (ArH).
Homopolymer of Naproxen-Modified p-(Chloromethyl)styrene.
This was prepared from naproxen-modified p-(chloromethyl)-
styrene (yield 66.4%). 1H NMR (CDCl3): δ 1.46 (CHCH3), 3.78
(1H and 3H, m, CH3CH and CH3O), 4.83 and 4.97 (COOCH2Ph),
6.23 and 6.78 (-CH2-CH-), 7.02, 7.31, and 7.53 (ArH and NaH).
Homopolymer of 1-Pyrenecarboxylic Acid Modified p-(Chloro-
methyl)styrene. This was prepared from 1-pyrenecarboxylic acid
modified p-(chloromethyl)styrene (yield 15%). 1H NMR (CDCl3):
δ 5.46 (COOCH2Py) 7.25-8.30 (ArH and PyH), 8.50, 8.93, and
9.19 (PyH).
General Procedure for the Radical Copolymerization of p-(Chloro-
methyl)styrene and Its Derivatives. p-(Chloromethyl)styrene was
mixed with modified p-(chloromethyl)styrene derivatives at
different ratios [p-(chloromethyl)styrene:p-(chloromethyl)styrene
derivative = 9:1, 5:1, 3:1, 1:1], and then AIBN ([M]/[I] = 12.5)
and chlorobenzene or toluene were added. The reaction mixtures
were processed in a similar way to the radical homopolymeri-
zation of p-(chloromethyl)styrene derivatives described above.
The polymerization products were isolated as white powders.
Copolymer of p-(Chloromethyl)styrene and Ibuprofen-Modified
p-(Chloromethyl)styrene. 1H NMR (CDCl3): δ 0.90, 1.15, 1.50,
1.85, 4.50(CH2Cl), 4.97(CH2Ib), 6.53 and 7.08 (ArH).
Copolymer of p-(Chloromethyl)styrene and Naproxen-Modi-
1
fied p-(Chloromethyl)styrene. H NMR (CDCl3): δ 0.90, 1.12,
1.40, 1.58, 3.89 (CH3CH and CH3O), 4.46(CH2Cl), 5.0(CH2Na),
6.48 (ArH), 7.0, 7.40, and 7.64(NaH).
Copolymer of p-(chloromethyl)styrene and Pyrene-Modified
p-(Chloromethyl)styrene. 1H NMR (CDCl3): δ 0.90, 1.17, 1.50,
4.36 (PhCH2Cl), 5.43(PhCH2Py), 6.48 and 6.97 (ArH), 8.10,
8.60, and 9.30 ((PyH).
Synthesis of Hyperbranched Poly[p-(chloromethyl)styrene],
HPPCMSt, by Metal-Mediated Controlled Radical Polymerization.
Hyperbranched PPCMSt was synthesized according to pre-
viously published procedure.26 The preparation was conducted
as follows: 2,20-bipyridyl (8.4 mmol) and CuCl (4.2 mmol) were
added to a cylindrical reaction vessel, followed by addition of
chlorobenzene (8.0 mL) and p-(chloromethyl)styrene (21 mmol).
Then the reaction mixture was flushed with nitrogen and placed
in an oil-bath (115 °C). After ∼40 min, a green solid precipi-
tated. Polymerization went on for 4 h. Then the reaction vessel
was moved out of the oil-bath and allowed to cool down for 30 min.
THF (50 mL) was added to the reaction mixture, which was then
stirred at room temperature overnight to complete the dissolu-
tion of the polymer and to allow for catalyst oxidization. The
solution was filtered through alumina to remove the insoluble
salts. The resulting clear brown liquid was concentrated and
Swelling Experiments. Before the swelling measurements were
performed, all hydrogels (irregular particles) were extracted with
ethanol, chloroform, and multiple DI water portions. Hydrogels
were then dried in the lyophilizer. The swelling behavior of networks
with different Jeffamine and PPCMSt content was investigated
as a function of time. All measurements were done in triplicates.
Kinetics of Swelling. Dynamic swelling (weight) studies were
conducted to elucidate the mechanism of water diffusion into
the polymer networks based on PPCMSt and Jeffamine as deter-
mined by the dynamic portion of the gravimetric curve. The
kinetics of swelling was investigated as follows: ∼15 mg of
hydrogels were placed in a vial containing 20 mL of DI water. The
hydrogels were weighted periodically throughout the experiment
at room temperature. The water uptake, Mt, was followed as a

10020 Macromolecules, Vol. 43, No. 23, 2010
Scheme 1. Synthesis of Drug/Marker-Substituted Monomers from p-(Chloromethyl)styrene
Lin and Gitsov
function of time until equilibrium was attained. Mt of the
hydrogels was calculated from the average of the final weights
using the formula:
Mt ¼ ðWwet - WdryÞ=Wdry ꢀ 100%
Swelling Ratio. The weight swelling ratio was calculated as the
ratio of the weight of the swollen hydrogels to the weight of the
original dry gel according to the following procedure. The weight
of the dry samples was recorded before they were immersed in a
large excess of the corresponding solvent (DI water, toluene,
THF, or CH2Cl2) at room temperature. After at least 24 h the
specimens were weighted again after the solvent droplets on the
surface of the samples were carefully blotted with lint-free paper.
The weight swelling ratio (q) was calculated as follows:
q ¼ Wwet=Wdry
Wwet and Wdry are the weights of swollen and dry sample,
respectively.
Figure 1. FT-IR spectra of p-(chloromethyl)styrene (PCMSt) and its
ibuprofen (Ib-PCMSt), naproxen (Na-PCMSt), and pyrene (Py-PCMSt)
derivatives.
Ibuprofen, Naproxen, and Pyrene Release Studies from the
Hydrogels. Prior to the release studies, calibration curves for
ibuprofen, naproxen, and 1-pyrenecarboxylic acid were created
in NaOH solutions at pH=10 to determine the corresponding
molar extinction coefficients (ε) in this medium. All three sub-
stances showed good linear absorbance/concentration relation-
ships (R2 >0.99). The absorbance intensity of ibuprofen was
between 0.2 and 2.3 AU at concentrations 2.31 ꢀ 10-5 to 2.89 ꢀ
10-4 M, the absorbance intensity of naproxen was between
0.1 and 2.18 AU at 1.67 ꢀ 10-5 to 4.18 ꢀ 10-4 M, and the absor-
bance intensity was between 0.09 and 3.5 AU for 1-pyrenecar-
boxylic acid at concentrations ranging from 4.38 ꢀ 10-6 to 8.75 ꢀ
10-5 M. The ε values in NaOH solution were 8156 L mol-1
cm-1at λ=223 nm for ibuprofen, 4954 L mol-1 cm-1at λ =
270 nm for naproxen and 38 851 L mol-1 cm-1at λ = 242 nm for
1-pyrenecarboxylic acid. For all release measurements, a known
amount (about 30 mg) of the dry hydrogel was preswollen over-
night in DI water, and the solvent droplets on the surface of the
specimens were carefully blotted with lint-free paper before their
immersion in a vial with NaOH solution (15 mL). The release of
the drugs from the hydrogels was studied at room temperature. At
various time intervals, 3 mL of the basic solution were taken to
measure the drug concentration by UV-vis. The spectra of the
aliquots from the vial were recorded until there were no measur-
able changes in the absorbance intensity. The amount of the
release percentage was calculated from the following equation:
or via passive binding. In turn the inclusion could be performed
before or after the cross-linking operation. When a controlled
release of the encapsulated substance is desired, the passive binding
has certain limitations mostly associated with the partition co-
efficient of the substrates.18b,27 Therefore, in this study, the incorpo-
ration and release of the model drugs and the fluorescent marker
will be explored through the chemical bond formation and scission
pathway. Initial pilot experiments show that the efficiency of the
covalent attachment of the investigated substances into preformed
and preswollen hydrogels is very low (5-10%). This method also
tends to create networkswithunevensubstratedistributionwhere
the surface of the hydrogel is “saturated” while the interior of the
cross-linked matrix remains mostly “empty”. Therefore, the adopted
strategy for all subsequent experiments involves the incorporation
of the desired substance before the polymer synthesis (macro-
monomer approach) or before the cross-linking reaction (post-
polymerization modification).
Macromonomer Formation. Synthesis and Characterization
of p-(Chloromethyl)styrene Derivatives. Three new monomers,
based on p-(chloromethyl)styrene, PCMSt, are synthesized by
reaction with drugs (ibuprofen, naproxen) or a model fluo-
rescent compound (1-pyrenecarboxylic acid) using K2CO3/
18-crown-6 as a catalyst, Scheme 1. After stirring the reaction
mixture at reflux for 24 h no remaining drugs can be detected
by TLC manifesting the completion of the reaction. The FT-IR
spectra of the purified new monomers are shown in Figure 1.
It is seen, that the strong peak at 1710-1740 cm-1 due to the
carboxyl groups is observed in all of the produced monomers.
Figures 2 and 3 show the 1H and 13C NMR spectra of the
original PCMSt and the substituted PCMSt derivatives. It is
clearly seen that the -CH2- signals at 4.7 ppm (benzyl
% release ¼ ðWt=WtotalÞ ꢀ 100
Here Wt is the weight of released drugs at time t and Wtotal is the
total weight of drugs being attached in the gel structure.
Results and Discussion
Incorporation of substances in cross-linked materials could be
achieved by two general strategies—through covalent link formation

Article
Macromolecules, Vol. 43, No. 23, 2010 10021
Figure 2. 1HNMR of p-(chloromethyl) styrene, PCMSt, and its derivatives with ibuprofen (Ib-PCMSt), naproxen (Na-PCMSt), and pyrene (Py-PCMSt).
chloride moiety) shift to lower field in all three new mono-
mers (Ib-PCMSt, Na-PCMSt, and Py-PCMSt). The inte-
grals of all hydrogen peaks correspond to the theoretical
values calculated for the chemical structures shown in
Scheme 1. The attachment is also confirmed by the appearance
of carboxyl group signals at δ = 168-175 ppm in the 13C NMR
spectra, Figure 3. The chemical structure of the PCMSt
derivatives and their purity are also confirmed by the ele-
mental analysis.
Homopolymerization and Copolymerization of p-(Chloro-
methyl)styrene Derivatives. The radical homopolymerization
and copolymerization of the p-(chloromethyl)styrene and its
derivatives are initially examined using AIBN as initiator
and toluene or chlorobenzene as solvents. The polymeriza-
tion in chlorobenzene yields polymers with lower molecular
weights and higher polydispersities. The results from the
polymerizations in toluene are summarized in Table 1 and
Figure 4. It should be noted that at identical reaction con-
ditions p-chloromethylstyrene (PCMSt) produces polymers
with the highest yield. The products also have monomodal
molecular weight distribution and the lowest polydispersity.
The homopolymers from Ib-PCMSt or Na-PCMSt are
formed with lower yields, but have markedly higher molec-
ular weights and polydispersities. They also have multimodal
molecular weight distributions indicating the occurrence of
chain transfer reactions to monomer probably through the
hydrogens in the methyl groups of the drug fragments. The
polymer of Py-PCMSt is formed with the lowest yield and
molecular weight. This result is somewhat unexpected since
under comparable reaction conditions the radical (co)poly-
merizations of monomers with condensed aromatic rings
tend to afford polymers in good yields and with high molec-
ular weights.28 The low reactivity of this monomer is probably
caused by the bulkiness of the attached pyrene, which disrupts
favorable monomer alignment and chain propagation. SEC
with UV detection tuned at the specific wavelength of the
ibuprofen (223 nm), naproxen (270 nm) and pyrenecar-
boxylic acid (242 nm) shows that all eluting fractions contain
the corresponding attached moiety, which is not cleaved
during the polymerization. Therefore, the products formed
are not fractionated despite their multimodal molecular
weight distributions and are used directly in the network
synthesis.
The copolymerization of PCMSt and derivatives proceeds
at the same reaction conditions as their homopolymeriza-
tion, Scheme 2.
The results obtained show that under the same reaction
conditions the copolymers form with lower yields and molecular
weights than the PCMSt homopolymer, Table 2. However,
these parameters are notably higher for the Py-PCMSt copoly-
mers compared with the corresponding values of the Py-PCMSt
homopolymer and could serve as an indirect proof of the
previously discussed causes for the low reactivity of this mono-
mer. The copolymer composition is calculated from the integral
ratio of the peak for the CH2 group, linked to the drug (marker),
and the -CH2Cl peak in 1H NMR spectra. An example with
Na-PCMSt is shown in Figure 5. Since the copolymerizations
are driven to high conversions the actual composition of the
copolymers formed is similar to the feed ratio of two monomers
in the initial polymerization mixtures, Table 2.
It is also found that the reaction time increases the final
yield and molecular weight of the copolymers, Table 3.

10022 Macromolecules, Vol. 43, No. 23, 2010
Lin and Gitsov
Figure 3. 13C NMR of p-(chloromethyl)styrene, PCMSt, and its derivatives with ibuprofen (Ib-PCMSt), naproxen (Na-PCMSt), and pyrene
(Py-PCMSt).
Figure 4. SEC traces of homopolymers from p-(chloromethyl)styrene, PCMSt and its derivatives with ibuprofen (Ib-PCMSt), naproxen (Na-PCMSt),
and pyrene (Py-PCMSt). See the Experimental Section for analysis conditions.
Table 1. Yields and Molecular Weight Characteristics of Polymers
formed from p-(Chloromethyl)styrene (PCMSt) and its Ibuprofen
(Ib-PCMSt), Naproxen (Na-PCMSt), and Pyrene (Py-PCMSt)
Derivatives
The attempted formation of functional hyperbranched
macromolecules by self-condensing vinyl copolymerization26
of PCMSt and its derivatives produces copolymers in very
low yields (5-10%) and broad multimodal molecular weight
distributions. That is why the incorporation of ibuprofen,
naproxen, and pyrene in this macromolecular architecture is
achieved by postpolymerization modification of preformed
hyperbranched poly(PCMSt) as described in the section below.
Postpolymerization Approach. Modification of Hyperbranched
Poly(PCMSt). Modified hyperbranched poly(PCMSt),
a
Mp
polymer
[M]/[I] yield (%)
Mn
Mw/Mn
poly(PCMSt)
12.5
12.5
12.5
12.5
89.9
67.5
66.4
15.1
5800
10 400 13 700
8700 38 000
8900
1.50
2.16
2.57
2.18
poly(Ib-PCMSt)
poly(Na-PCMSt)
poly(Py-PCMSt)
a Mp is the apparent molecular weight at the peak maximum.
1200
800

Article
Macromolecules, Vol. 43, No. 23, 2010 10023
Scheme 2. Co polymerization of p-(Chloromethyl)styrene and Its Derivatives
Table 2. Molecular Weights of Copolymers of p-(Chloromethyl)styrene (PCMSt) and Naproxen- or Pyrene-Modified PCMSt Prepared at 80 °C
in Chlorobenzene at [M]/[I] = 12.5
feed ratio of
modified PCMSt (%)
content of modified PCMSt
in copolymer (%)
a
polymer or copolymer
poly(PCMSt)
yield (%)
Mn
Mp
Mw/Mn
0
0
84.0
37.1
36.6
47.5
72.0
71.9
64.9
79.3
74.4
3500
2300
2300
2300
2300
2400
2200
2100
2000
4500
2700
2800
2700
2800
3000
2800
2900
2900
1.33
1.21
1.23
1.23
1.30
1.26
1.26
1.31
1.34
poly(PCMSt -stat-Na-PCMSt)
poly(PCMSt -stat-Na-PCMSt)
poly(PCMSt -stat-Na-PCMSt)
poly(PCMSt -stat-Na-PCMSt)
poly(PCMSt -stat-Py-PCMSt)
poly(PCMSt -stat-Py-PCMSt)
poly(PCMSt -stat-Py-PCMSt)
poly(PCMSt -stat-Py-PCMSt)
10.0
16.7
25.0
50.0
10.0
16.7
25.0
50.0
10.0
16.1
25.0
44.5
5.0
16.0
23.4
41.3
a Mp is the apparent molecular weight at the peak maximum.
Figure 5. 1H NMR spectra of homopolymers - poly(PCMSt) and poly(Na-PCMSt) and the copolymer of the same comonomers.
H-PPCMSt, is prepared by esterification under mild basic
conditions as shown in Scheme 3. The degree of modifica-
tion is controlled by changing the molar composition of
reaction mixture. The incorporation of the corresponding
moiety is confirmed by the appearance of strong UV
absorbance in the SEC traces of the modification pro-
ducts at 312 nm (naproxen) and 332 nm (pyrene), respec-
tively. Pure PPCMSt has no UV absorbance at these two
wavelengths.
NMR is also used to verify the chemical transformations
of the benzyl chloride groups and the apparent degree of
substitution (DS). The values are determined by 1H NMR as
the ratio of the peak areas for the initial and substituted
methylene protons in the 4-benzyl chloride moiety. Figure 6

10024 Macromolecules, Vol. 43, No. 23, 2010
Lin and Gitsov
Table 3. Molecular Weight Characteristics of PCMSt Copolymers Prepared at Different Reaction Times with [M]/[I] = 12.5 in Toluene at 60 °C
feed ratio of modified content of modified PCMSt
PCMSt (%)
a
polymer or copolymer
poly(PCMSt)
poly(PCMSt-stat-Ib-PCMSt)
poly(PCMSt-stat-Na-PCMSt)
poly(PCMSt-stat-Py-PCMSt)
reaction time (h)
in copolymer (5)
yield (%)
Mn
Mp
Mw/Mn
24
24
72
24
72
24
72
0
0
84.0
31.1
79.9
37.1
92.4
71.9
93.8
3500
3500
5600
2300
5200
2400
4500
4500
7700
2700
7700
3000
1.33
1.40
1.58
1.21
1.50
1.26
1.75
10.0
10.0
10.0
10.0
10.0
10.0
7.2
6.9
10.0
8.5
5.0
8.1
6300 11 100
a Mp is the apparent molecular weight at the peak maximum.
Scheme 3. Attachment of Model Compounds to Hyperbranched Poly(PCMSt)
shows for example the 1H NMR spectrum of hyperbranched
PPCMSt and the pyrene modified product. The decrease in
the relative integral intensities of the signals between 4 and
5 ppm indicate that the covalent attachment occurs both
through the primary and secondary methyl chloride groups
with slight preference for the primary methyl chloride moiety
(1:3.63 before to 1:3.06 after Py attachment). Ibuprofen and
naproxen modification of PPCMSt proceeds in a similar
fashion. It should be emphasized that multimodifications of
PPCMSt can also be achieved with more than one drug and
the resulting macromolecules can be used directly to prepare
templates and formulations, where the amount and type of
drugs could be tailored to desired levels and specific applica-
tions.
Network Formation. Preparation of Amphiphilic Hydrogels
from PPCMSt and Reactive Aliphatic Polyethers. Amphiphilic
hydrogels are synthesized from linear or hyperbranched PPCMSt
(modified or not modified) and reactive aliphatic polyethers
(Jeffamine and PEG) by two reactions: Williamson ether
synthesis and nucleophilic substitution as shown in Scheme 4.
The reactions lead to the formation of two types of networks—
type I (containing hyperbranched PPCMSt) and type II (with
linear PPCMSt). The hydrogels produced by Williamson ether
reaction are well characterized and extensively discussed in our
previous paper.25 That is why in following sections, only the
hydrogels prepared by the nucleophilic substitution reaction
will be described in detail.
Both linear and hyperbranched PPCMSt react with Jeffamine
and form cross-linked structures (JEFF-L-PPCMSt and JEFF-
H-PPCMSt, respectively) at different feeding ratios. At 45-
50 °C with small amounts of THF, the reaction mixtures turn
milky and the viscosity increases rapidly in the presence of
triethylamine as catalyst. The yields range from 50 to 85% as
calculated from the weight of dried gels. The decrease of the
absorption bands, characteristic for primary amino groups
at 3370-3400 cm-1, and the appearance of the N-H bending
(scissoring) absorbance at 1508 cm-1 in the FT-IR spectra of
isolated products confirms the network formation. The coupling
efficiency is evaluated at three different weight feeding ratios
of the two building blocks (Jeffamine/PPCMSt = 4, 2, and 1).
The yield of formed networks increases by ∼12% upon
decrease in the Jeffamine amount from 80 to 50 wt %. This
phenomenon is probably caused by the decreased ammo-
nium salt formation at the Jeffamine chain ends, which
would enable their more efficient participation in the cross-
linking reaction. At the same reaction conditions, the yields
of hydrogels synthesized from linear PPCMSt are higher
than those from the hyperbranched analogue probably due
to differences in reactivity and steric factors.
All gels have a uniform appearance and most of the
swollen gels are transparent. An example of modified and
unmodified JEFF-L-PPCMSt networks is shown in Figure 7.
It should be emphasized, that the incorporation of sub-
stances of biological importance via hydrolyzable linkage
prior to the final formation of the hydrogels offers several
notable advantages over the passive or chemical binding of
the same compounds into preformed networks, the most
important ones being the increased load and the homoge-
neous distribution across the entire gel matrix.
The evaluation of the accessibility and release of the
compounds bonded in the hydrogels under different condi-
tions will be discussed in the following section.
Swelling Behavior. A. Dynamic Swelling Properties. The
ability to control swelling and solute diffusion through
hydrogels determines their usefulness as tissue engineering
scaffolds, drug delivery platforms, and enzyme immobilization
templates. The polymer composition, the structure and size
of the gel pores, the water content, and the nature and size of
the incorporated substances jointly affect the processes of
swelling and solutes diffusion.
The time dependence of the swelling ratio is shown in
Figure 8 for hydrogels with different chemical composition
composed of Jeffamine and linear or hyperbranched PPCMSt.
Here, the water uptake at time zero corresponds to the start
of swelling process. All measurements are performed at room
temperature and the water uptake is monitored gravimetrically.
The figure illustrates that the swelling increases with time up
to a certain level, and then levels off to reach equilibrium

Article
Macromolecules, Vol. 43, No. 23, 2010 10025
Figure 6. 1H NMR (CDCl3) of hyperbranched PPCMSt and its modification derivatives with 1-pyrenecarboxylic acid, poly(PCMSt-stat-Py-PCMSt).
Scheme 4. Syntheses of Hydrogels from Poly(PCMSt) and Polyethers

10026 Macromolecules, Vol. 43, No. 23, 2010
Lin and Gitsov
after approximately 2 h. This equilibrium swelling ratio
depends naturally on the amount of hydrophilic component
(Jeffamine) used in the preparation of the conetwork. The
results also show that the equilibrium process strongly
depends on the polymers architecture used in the hydrogel
construction. Thus, cross-linked materials, built of hyper-
branched PPCMSt, swell ∼50% more than those from linear
PPCMSt at the same feeding ratios. In that aspect the
Jeffamine hydrogels show swelling behavior similar to the
previously explored PPCMSt hydrogel series containing
different PEG segments.25
Naturally, the equilibrium swelling ratio of any hydrogel
reflects the affinity of the components toward the selected
media. For the networks in this study the highest swelling
ratio is observed in CH2Cl2, a good solvent for both con-
stituent blocks. Since Jeffamine is the only component that
can dissolve in water, the swelling ratio in water for both
types of hydrogels decreases with the weight percentage
decrease of the hydrophilic building block. For all systems,
the hydrogel swelling ratio depends also on the cross-linking
density, chemical balance of two components and the poros-
ity parameters.
The exact diffusion mechanism in these hydrogels and the
influence of the specimens shape and size need further investi-
gation.
There is marked difference in the swelling ratio of gels with
linear and hyperbranched PPCMSt with the same JEFF
content. In distinction to the values of the hyperbranched
networks, the weight swelling ratios of linear PPCMSt
hydrogels are significantly lower in all four solvents. The
reason is not only the cross-linking density difference, caused
by the different polymer architecture as described in the pre-
vious paper,25 but also the different hydrogel morphology as
revealed by scanning electron microscopy (SEM), Figure 9.
All hydrogels formed from hyperbranched PPCMSt have
macroporous organization, which is independent of the weight
percentage of Jeffamine therein. The pore diameter is about
2-10 μm with a relatively wide pore size distribution. The
morphology of hydrogels formed from linear PPCMSt does
not show any macropore structure. The lack of macropores,
which serve as effective solvent reservoirs, affects signifi-
cantly the ability of these hydrogels to swell. It is noteworthy
that both hydrogel lines, formed by the two methods, are
stable and show no sign of degradation even after repeated
swelling in different solvents.
B. Equilibrium Swelling of Hydrogels. The amount of
solvent absorbed by a network is quantitatively represented
by the equilibrium water weight swelling ratio (q). It is mea-
sured at room temperature in four solvents for the hydrogel
family of Jeffamine and hyperbranched or linear PPCMSt.
The results of swelling experiments depending on networks
composition are summarized in Table 4. The values included
are the average results from three separate determinations.
The dissimilarity in the swelling behavior of PEG and
Jeffamine gels even at close interjunction lengths (the mo-
lecular weights of both polyethers are between 5 and 6 kDa)
is very obvious. The weight swelling ratio of the Jeffamine
hydrogels is more than two times lower than that of the
PEG conetworks discussed in our previous paper25 (Tables 4
and 5). One of the contributing factors is the lower affinity of
Jeffamine to these four solvents, the main probable reason
being the presence of poly(propylene oxide) in this copolymer.
This different composition would ultimately lead to differ-
ences in the hydrogels morphology. The hydrogels, formed
by the Williamson ether reaction contain also numerous
Figure 7. Photographs of swollen hydrogels from (a) Jeffamine and
linear PPCMSt, Jeffamine/PPCMSt = 2, and (b) Jeffamineand pyrene-
modified linear PPCMSt.
Figure 8. Time dependence of the swelling process, measured at room temperature for amphiphilic hydrogels from Jeffamine (JEFF) and
hyperbranched (HPPCMSt) or linear (LPPCMST) poly[p-(chloromethyl)styrene]. The numbers signify the Jeffamine to PPCMSt ratio.

Article
Macromolecules, Vol. 43, No. 23, 2010 10027
Table 4. Yields and Weight Swelling Ratios of Hydrogels Formed by Reaction of Jeffamine (JEFF, Mn = 6000) and HPPCMSt (Hyperbranched
Polymer: Mn = 5000, Mw = 16 500) or LPPCMSt (Linear Polymer: Mn = 10 100, Mw = 16 800)a
weight swelling ratios
hydrogel
JEFF wt %
yield (%)
DI water
toluene
THF
CH2Cl2
JEFF-HPPCMSt4
JEFF-HPPCMSt2
JEFF-HPPCMSt1
JEFF-LPPCMSt4
JEFF-LPPCMSt2
JEFF-LPPCMSt1
80
67
50
80
67
50
52.6
65.5
75.5
70.1
74.3
82.6
4.80 ( 0.05
3.04 ( 0.04
2.44 ( 0.08
3.58 ( 0.02
2.27 ( 0.02
1.81 ( 0.01
8.39 ( 0.24
7.15 ( 0.06
8.83 ( 0.15
3.32 ( 0.05
3.90 ( 0.04
3.15 ( 0.05
10.03 ( 0.32
9.38 ( 0.29
10.98 ( 0.11
3.81 ( 0.02
4.28 ( 0.10
3.88 ( 0.06
17.48 ( 0.16
15.86 ( 0.11
18.89 ( 0.32
7.32 ( 0.06
8.08 ( 0.07
6.98 ( 0.04
a The numbers signify the Jeffamine to PPCMSt ratio.
Figure 9. SEM micrographs of Jeffamine based hydrogels: (a) JEFF-HHPCMSt4; (b) JEFF-HPPCMSt2, (c) JEFF-HPPCMSt1, (d) JEFF-LPPCMSt4,
(e) JEFF-LPPCMSt2, (f) JEFF-LPPCMSt1.
Table 5. Yields and Weight Swelling Ratios of Hydrogels Series,25 Formed by Reaction of Linear PEG (L PEG: Mw = 5400) and Hyperbranched
PPCMSt (HPPCMSt: Mn = 5000; Mw = 16 500) or Linear PPCMSt (LPPCMSt: Mn = 10 080; Mw = 16 800)a
weight swelling ratios
hydrogel
PEG wt %
yield (%)
DI water
toluene
THF
CH2Cl2
L PEG5k-HPPCMSt4
L PEG5k-HPPCMSt2
L PEG5k-HPPCMSt1
L PEG5k-LPPCMSt4
L PEG5k-LPPCMSt2
L PEG5k-LPPCMSt1
80
67
50
80
67
50
87.7
90.7
90.6
93.8
97.9
98.9
19.2 ( 0.2
16.6 ( 0.4
16.0 ( 0.4
14.7 ( 0.1
10.0 ( 0.1
9.6 ( 0.1
16.3 ( 0.1
17.0 ( 0.1
19.3 ( 0.7
14.8 ( 0.1
14.6 ( 0.1
19.3 ( 0.1
17.4 ( 0.1
17.5 ( 0.1
18.4 ( 0.1
15.6 ( 0.1
14.0 ( 0.8
18.5 ( 0.4
39.0 ( 0.3
37.2 ( 0.7
36.6 ( 1.1
37.7 ( 0.1
34.1 ( 0.9
41.0 ( 0.3
a The numbers signify the PEG to PPCMSt ratio.
Table 6. Yields and Weight Swelling Ratios of Hydrogels Formed by Reaction of Jeffamine (JEFF, Mn = 6000) and Modified PPCMSt
weight swelling ratios
hydrogel
JEFF wt %
yield (%)
DI water
toluene
THF
CH2Cl2
JEFF-HPPCMSt2a
JEFF-HPPCMSt-Ib
JEFF-HPPCMSt-Na
JEFF-HPPCMSt-Py
JEFF-LPPCMSt2a
JEFF-LPPCMSt-Ib
JEFF-LPPCMSt-Na
JEFF-LPPCMSt-Py
67
67
67
67
67
67
67
67
65.5
69.2
69.5
72.4
74.3
73.1
74.3
83.6
3.04 ( 0.04
2.88 ( 0.01
2.85 ( 0.01
2.93 ( 0.01
2.27 ( 0.02
3.94 ( 0.01
3.34 ( 0.03
3.44 ( 0.01
7.15 ( 0.06
3.77 ( 0.03
4.58 ( 0.02
4.17 ( 0.03
3.90 ( 0.04
3.10 ( 0.06
2.84 ( 0.01
2.81 ( 0.01
9.38 ( 0.29
5.51 ( 0.02
5.70 ( 0.04
5.29 ( 0.12
4.28 ( 0.10
3.21 ( 0.03
3.14 ( 0.01
3.26 ( 0.02
15.86 ( 0.11
11.06 ( 0.07
11.35 ( 0.04
10.78 ( 0.08
8.08 ( 0.07
6.06 ( 0.01
5.68 ( 0.05
6.08 ( 0.05
a The numbers signify the Jeffamine to PPCMSt ratio.
micropores, while their analogues derived from Jeffamine
have few pores, Figure 9. It is not clear at this point, however,
whether the cross-linking chemistry or the fabrication technol-
ogy are the contributing factors for the different morpholo-
gies observed.
The effect of drug attachment on the equilibrium water
content in hydrogel at the similar polymer composition is
investigated as well. Comparing the data in Tables 4 and 6, it
is evident that the yields are higher for networks of types
I and II, when the PPCMSt used is modified with Ib, Na, or
Py. The equilibrium weight swelling ratio for the hydrogels
with modified hydrophobic segments is generally lower in all
solvents. The results in aqueous medium could be probably
explained by the increased overall hydrophobicity of the
resulting networks. The SEM micrographs also reveal the
absence of interconnected porous structure for both types of

10028 Macromolecules, Vol. 43, No. 23, 2010
Lin and Gitsov
Figure 10. SEM micrographs of Jeffamine-based hydrogel with incorporated naproxen moieties, JEFF-LPPCMSt-Na (see also Table 6). Magnification
1000ꢀ.
Figure 11. Size and shape of ibuprofen, naproxen, and 1-pyrenecarboxylic acid.
modified networks. An example with naproxen is shown in
Figure 10. Thus, it is reasonable to assume that the decrease
in the q values for the modified hydrogels in the four solvents
is also probably due to the morphological changes in the
network matrix.
Release examples with modified JEFF-HPPCMSt hydrogels
are shown in Figure 12. The data show that the molecules of
the model compounds are able to slowly diffuse through the
swollen networks and the release is completed in about
3 days. The release mechanism and the total amount released
in one portion are different for the compounds investigated.
At similar attachment levels to PPCMSt, the hydrogels with
ibuprofen and naproxen release ∼7-8 mg drug/g dry gel,
while the pyrene gel releases only 0.32 mg Py/g dry gel. This
result is somewhat unexpected since the mesh (pore) sizes in
these networks are much larger than hydrodynamic radius of
the released substances and therefore their diffusion should
not experience significant interference from the swollen
hydrogel matrices. In this particular case the difference is
most probably caused by the (in)solubility of the pyrene
carboxylate in water. Thus, the more hydrophobic pyrene
moiety strongly interacts with the hydrophobic PPCMSt
domains in the conetwork and shows very low tendency to
migrate into the aqueous medium surrounding the swollen
gel particles.
Release Behavior of JEFF-PPCMSt Hydrogels. The po-
tential use of hydrogels based on Jeffamine and PPCMSt in
biomedical and biotechnological applications is evaluated with
ibuprofen, naproxen and 1-pyrenecarboxylic acid as model
compounds with increasing hydrophobicity. The solvent acces-
sible surfaces of these compounds, modeled by CambridgeSoft
Chem3D software (map property: hydrophobicity) are shown
in Figure 11. It is seen that Ip and Na have similar size and
elongated shape, while Py is ovoid and more compact.
To understand how the diffusion process of the drugs is
influenced by the structure of the polymer network, the
cumulative release profile of these substances is investigated
in basic medium (pH = 10.1). It should be noted that the
conditions are chosen for accelerated bond scission and by
no means are intended to mimic physiological conditions.

Article
Macromolecules, Vol. 43, No. 23, 2010 10029
Figure 12. Cumulative release of ibuprofen (Ib), naproxen (Na), and pyrene carboxylate (Py) from type I hydrogels (67 wt % of Jeffamine),
investigated in basic medium (pH = 10.1).
Figure 13. Cumulative release from networks of type I and II with 67 wt % of Jeffamine, containing covalently attached naproxen (Na) and pyrene (Py).
The influence of the macromolecular architecture of the
hydrophobic PPCMSt components (type I and type II
Jeffamine-based networks) on the release profile of naproxen
and pyrene is shown in Figure 13. It can be seen that the Na
amounts released from hyperbranched PPCMSt containing
hydrogels are almost four times larger than those from the
hydrogel of the linear polymer. It is also noteworthy that the
drug release from the network JEFF-LPPCMSt levels of
after 1 day while with the gel JEFF-HPPCMSt the process
continues for more than 3 days, Figure 13. These results
indicate that the drug diffusion process is not only affected
by the molecular size of the drugs, but is also influenced by
the structure of polymer network. In that aspect one should
take into account that at equivalent molecular weight the
globular hyperbranched polymers are more compact and
occupy less volume in the expanded framework of the
swollen hydrogels. Consequently the surface contact and
the interactions with the penetrating medium will differ signif-
icantly from the same processes in the hydrogels containing
the modified hydrophobic linear segments. Since the rate of
release depends also on the rate of bond scission (chemical or
enzymatic29) the amount of the released covalently linked
substance will be significantly different even at similar initial
attachment levels.
The total substance load in hydrogels with different
macromolecular architectures and the amount of the sub-
stance released from these hydrogels are shown in Table 7.
The cooperative influence of all previously discussed factors
is obvious. The released amounts from JEFF-HPPCMSt
hydrogels are higher than those released from JEFF-
LPPCMSt hydrogels after the interaction with two consecu-
tive fresh portions of basic aqueous medium. Despite the fact
that the gels are immersed in fresh basic solution after the
release profile in the starting medium levels off, the amount
of the released substances is consistently lower than the initial
burst (Table 7, second portion release). Thus, the lower
release is not caused by decreased concentration gradient.
It is most probably due to the lack of interconnected pores

10030 Macromolecules, Vol. 43, No. 23, 2010
Lin and Gitsov
Table 7. Total Drug Loading and Equilibrium Release from Hydrogels with Different Macromolecular Architecture
first portion release second portion release
amount (mg/g dry gel) amount (mg/g dry gel) % release
hydrogela
drug bonded (mg/g dry gel)
% release
JEFF-HPPCMSt2-Ib
JEFF-HPPCMSt2-Na
JEFF-HPPCMSt2-Py
JEFF-LPPCMSt2-Ib
JEFF-LPPCMSt2-Na
JEFF-LPPCMSt2-Py
33.3
33.3
25.3
33.3
19.2
33.3
8.01
7.63
0.34
3.20
2.10
0.23
24.1
22.9
1.3
9.6
11.0
0.7
4.48
4.41
0.27
2.02
1.10
0.12
13.5
13.2
1.0
6.1
5.8
0.4
a The numbers signify the Jeffamine to PPCMSt ratio.
(Figure 10) in the modified conetworks, which prevents the
efficient access of the swelling medium to all hydrolyzable
attachment sites in the hydrogel matrix.
Broniatowski, S. Macromol. Biosci. 2010, 10, 369–377. (c) Patrickios,
C. S. Macromol. Symp. 2010, 291-292, 1–11.
€
(10) Erdodi, G.; Kennedy, J. P. J. Polym. Sci., Part A: Polym. Chem.
2007, 45, 295–307.
The results obtained show that the release process includes
several interrelated phenomena: (1) penetration of the me-
dium into the network matrix and swelling, affected by the
hydrophilic/hydrophobic balance of the hydrogel, the cross-
linking density and the morphology (pore structure and
organization); (2) hydrolysis of the ester linkages (only the
deattached substances can be released under suitable con-
ditions); (3) dissolution of the cleaved molecules in the
surrounding aqueous media (i.e., the amount of drugs that
can be spectroscopically detected outside the gels); (4) the
residual interaction of the detached compounds with their
original carriers (hydrophobic/hydrophobic or π-π inter-
actions) and (5) diffusion process through the swollen gel as a
consequence of the preceding events (1-4).
€
ꢀ
(11) Erdodi, G.; Ivan, B. Chem. Mater. 2004, 16, 959–962.
ꢀ
€
(12) Ivan, B.; Haraszti, M.; Erdodi, G.; Scherble, J.; Thomann, R.;
Mulhaupt, R. Macromol. Symp. 2005, 227, 264–274.
(13) Scherble, J.; Thomann, R.; Ivan, B.; Mulhaupt, R. J. Polym. Sci.,
€
ꢀ
€
Part B: Polym. Phys. 2001, 39, 1429–1436.
(14) (a) Tiller, J. C.; Sprich, C.; Hartmann, L. J. Controlled Release
2005, 103, 355–367. (b) Bruns, N.; Hanko, M.; Dech, S.; Ladisch, R.;
Tobis, J.; Tiller, J. C. Macromol. Symp. 2010, 291-292, 293–301.
(15) (a) Groenewolt, M.; Brezesinski, T.; Schlaad, H.; Antonietti, M.;
ꢀ
Groh, P. W.; Ivan, B. Adv. Mater. 2005, 17, 1158–1162. (b) Kali, G.;
Georgiou, T. K.; Ivan, B.; Patrickios, C. S.; Loizou, E.; Thomann, Y.;
ꢀ
Tiller, J. C. Langmuir 2007, 23, 10746–10755. (c) Georgiou, T. K.;
Patrickios, C. S. Biomacromolecules 2008, 9, 574–582. (d) Kafouris,
D.; Gradzielski, M.; Patrickios, C. S. Macromolecules 2009, 42, 2972–
2980. (e) Rikkou, M. D.; Loizou, E.; Patrickios, C.; Porcar, L. Eur.
Polym. J. 2010, 46, 441–449.
(16) For some current reviews on dendrimers see: (a) Hourani, R.;
Kakkar, A. Macromol. Rapid Commun. 2010, 31, 947–974. (b) Franc,
G.; Kakkar, A. K. Chem. Soc. Rev. 2010, 39, 1536–1544. (c) Menjoge,
A. R.; Kannan, R. M.; Tomalia, D. A. Drug Disc. Today 2010, 15, 171–
185. (d) Rosen, B. M.; Wilson, C. J.; Wilson, D. A.; Peterca, M.; Imam,
M. R.; Percec, V. Chem. Rev. 2009, 109, 6275–6540. (e) Samad, A.;
Alam, M. I.; Saxena, K. Curr. Pharm. Des. 2009, 15, 2958–2969.
(17) (a) Gitsov, I.; Lys, T.; Zhu, C. Polym. Mater. Sci. Eng. 2000, 82, 328–
329. (b) Gitsov, I.; Zhu, C. Macromolecules 2002,35, 8418–8427. (c) Gitsov,
I.; Lys, T.; Zhu, C. In Polymer Gels. Fundamentals and Applications,
Bohidar, H. B., Dubin, P., Osada, Y., Eds., ACS Symposium Series Vol. 833,
American Chemical Society: Washington, DC, 2002; pp 218-232.
(18) (a) Gitsov, I.; Zhu, C. J. Am. Chem. Soc. 2003, 125, 11228–11234.
(b) Zhu, C.; Hard, C.; Lin, C.; Gitsov, I. J. Polym. Sci., Part A: Polym.
Chem. 2005, 43, 4017–4029.
(19) (a) Tanghe, L. M.; Goethals, E. J.; Du Prez, F. Polym. Int. 2003, 52,
191–197. (b) Altin, H.; Kosif, I.; Sanyal, R. Macromolecules 2010, 43,
3801–3808.
(20) For some illustrative reviews on hyperbranched polymers see:
(a) Voit, B. I.; Lederer, A. Chem. Rev. 2009, 109, 5924–5973.
(b) Hobzova, R.; Peter, J.; Sysel, P. Chem. Listy 2008, 102, 906–913.
(c) Haussler, M.; Tang, B. Z. Adv. Polym. Sci. 2007, 209, 1–58.
(d) Seiler, M. Fluid Phase Equilib. 2006, 241, 155–174.
Conclusions
New monomers with attached model compounds (drugs and
fluorescent marker) are successfully synthesized and incorpo-
rated into copolymers with p-(chloromethyl)styrene. The new
copolymers are formed with controllable molecular weight and
chemical compositions and are used for the preparation of new
amphiphilic conetworks with PEG and Jeffamine. The incor-
poration of macromolecules with different types of architecture
into the hydrogels has a pronounced effect on their degree of
swelling, morphology, and the diffusion processes into and out of
the cross-linked systems. The hydrogels, which include different
amounts of bonded model substances (ibuprofen, naproxen and
pyrenecarboxylic acid), are able to release them in aqueous media
under basic conditions. The multiple-step release process depends
on the hydrolysis of ester linkage, the microstructure of hydro-
gels, the physical properties of the attached compounds, as well as
their interaction with the surrounding conetwork matrix.
Acknowledgment. The authors wish to thank Dr. Y. Yuan
(SUNY ESF) for her assistance with the FT-IR analyses. Partial
funding of this research by The National Science Foundation
(CBET-0853454) and US Department of Agriculture (McIntire-
Stennis Award to I.G.) is gratefully acknowledged.
(21) Gan, D.; Mueller, A.; Wooley, K. L. J. Polym. Sci., Part A: Polym.
Chem. 2003, 41, 3531–3540.
(22) Gudipati, C. S.; Finlay, J. A.; Callow, J. A.; Callow, M. E.; Wooley,
K. L. Langmuir 2005, 21, 3044–3053.
(23) Gudipati, C. S.; Greenlief, M.; Johnson, J. A.; Prayongpan, P.;
Wooley,K.L.J. Polym. Sci., Part A: Polym. Chem. 2004,42, 6193–6208.
(24) (a) Bartels, J. W.; Cheng, C.; Powell, K. T.; Xu, J.; Wooley, K. L.
Macromol. Chem. Phys. 2007, 208, 1676–1687. (b) Ohashi, R.;
Bartels, J. W.; Xu, J. Q.; Wooley, K. L.; Schaefer, J. Adv. Funct.
Mater. 2009, 19, 3404–3410.
References and Notes
ꢀ
(1) Liu, M.; Kono, K.; Frechet, J. M. J. J. Controlled Release 2000, 65,
121–131.
(2) Gupta, P.; Vermani, K.; Grag, S. Drug Disc. Today 2002, 7, 569–579.
(3) Greewald, R. B.; Choe, Y. H.; McGuire, J.; Conover, C. D. Adv.
Drug Delivery Rev. 2003, 55, 217–250.
(25) Lin, C.; Gitsov, I. Macromolecules 2010, 43, 3256–3267.
ꢀ
(26) Weimer, M. W.; Frechet, J. M. J.; Gitsov, I. J. Polym. Sci., Part A:
(4) Allen, T. M.; Cullis, P. R. Science 2004, 303, 1818–1822.
(5) Brigger, I.; Dubernet, C.; Couvreur, P. Adv. Drug Delivery Rev.
2002, 54, 631–651.
Polym. Chem. 1998, 36, 955–970.
(27) (a) Merill, E. W.; Dennison, K. A.; Sung, C. Biomaterials 1993, 14,
1117–1126. (b) Gitsov, I.; Lambrych, K. R.; Remnant, V. A.; Pracitto, R.
J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 2711–2727.
(c) Hoffman, A. S. Adv. Drug Delivery Rev. 2002, 54, 3–12.
(28) (a) Gitsov, I.; Todorova, O. G. J. Appl. Polym. Sci. 1992, 46, 1631–
1638. (b) Todorova, O.; Gitsov, I.; Sinigersky, V. Polym. Bull. 1986, 15,
511–515.
ꢀ
(6) Bruns, N.; Scherble, J.; Hartmann, L.; Thomann, R.; Ivan, B.;
€
Mulhaupt, R.; Tiller, J. C. Macromolecules 2005, 38, 2431–2438.
(7) Bruns, N.; Tiller, J. C. Macromolecules 2006, 39, 4386–4394.
ꢀ
€
€
(8) Domjan, A.; Erdodi, G.; Wilhelm, M.; Neidhofer, M.; Landfester,
ꢀ
K.; Ivan, B.; Spiess, H. W. Macromolecules 2003, 36, 9107–9114.
€
(9) (a) Erdodi, G.; Kennedy, J. P. Prog. Polym. Sci. 2006, 31, 1–18.
(29) Ehrbar, M.; Metters, A.; Zammaretti, P.; Hubbell, J. A.; Zisch,
A. H. J. Contr. Rel. 2005, 101, 93–109.
€
(b) Kang, J.; Erdodi, G.; Kennedy, J. P.; Chou, H.; Lu, L. N.; Grundfest-