pubs.acs.

org/Langmuir
© 2009 American Chemical Society

Surface-Active Monomer as a Stabilizer for Polyurea Nanocapsules

Synthesized via Interfacial Polyaddition in Inverse Miniemulsion
Eva-Maria Rosenbauer,†,‡ Katharina Landfester,†,‡ and Anna Musyanovych*,†,‡
†
Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany, and ‡Institute of
Organic Chemistry III - Macromolecular Chemistry and Organic Materials University of Ulm, Albert-Einstein-
Allee 11, 89081 Ulm, Germany

Received May 13, 2009. Revised Manuscript Received June 25, 2009

A surface-active monomer, polyisobutylene-succinimide pentamine (Lubrizol U), was used as a stabilizer for
synthesizing polyurea nanocapsules with aqueous core via polyaddition at inverse miniemulsion droplet interface.
Because of the presence of amine groups in the Lubrizol molecule, it is covalently incorporated into the polymeric
interfacial layer after reaction, resulting in more compact (less permeable) capsule shell. The influence of the stabilizer
and the monomer concentration on the shell thickness, colloidal stability, average capsule size, and capsule size
polydispersity were examined in detail. Different materials, such as a water-soluble fluorescent dye and aqueous
dispersion of magnetite nanoparticles with 10 nm in size, were used as inner phase of the polyurea capsules. The
encapsulation efficiency was studied using fluorescein as a marker. As an example for biomedical application, the
fluorescein-containing capsules were utilized in cell uptake experiments and visualized using fluorescence microscopy.

1. Introduction entrapped FITC molecules remained within the capsules cavity.

The preparation of polymeric nanocapsules has attracted a The formation of a polymeric shell by adsorption on the template
widespread interest in the past decades due to their potential core can be also achieved using the layer-by-layer technique,
applications in numerous fields, including sensors, vesicles for which is based on an alternate adsorption of oppositely charged
enzymes and chemicals, controlled drug delivery systems, etc.1-3 polyelectrolytes (or nanoparticles) onto the fluid droplet surface6
Particularly stable polymeric nanocapsules with an aqueous core or onto the surface of a sacrificial core. The thickness of the
are of high importance for the protection, storage, and delivery of capsule shell can be adjusted by the number of layer-by-layer
hydrophilic organic/inorganic compounds. Generally, there a two deposition cycles. The size of the capsule depends on the colloidal
major approaches to synthesize polymeric nanocapsules, which template size. Capsules with diameter varied from 70 nm to 10 μm
can be differentiated by the presence or absence of a sacrificial can be obtained.6-12 Direct shell polymerization around the
core template. In the first method the core particles, usually of particle core was described recently involving dispersion13-15 or
mineral or organic silica origin, are used as templates. The living polymerization mechanisms (e.g., atom transfer radical
particles can then be modified with a polymeric shell or different polymerization or anionic polymerization).16,17 Later, the core
polymeric layers by adsorption of a preformed polymer or by the was eliminated by chemical etching with HF or THF.18-24
direct polymerization onto the core surface. Afterward, the
template cores are removed under physical or chemical conditions
(6) Li, J. B.; Mohwald, H.; An, Z. H.; Lu, G. Soft Matter 2005, 1(4), 259–264.
by dissolution or calcination process resulting in the hollow (7) Caruso, F.; Spasova, M.; Salgueirino-Maceira, V.; Liz-Marzan, L. M. Adv.
nanospheres, which can be subsequently loaded with a preferred Mater. 2001, 13(14), 1090–1094.
(8) Caruso, F. Adv. Mater. 2001, 13(1), 11–22.
target. Feldheim et al. have used gold nanoparticles as templates (9) Caruso, F.; Caruso, R. A.; M€ohwald, H. Chem. Mater. 1999, 11(11), 3309–
for the synthesis of polypyrrole and poly(N-methylpyrrole) 3314.
shells.4 After etching the gold, the hollow polymer capsules with (10) Caruso, F.; Caruso, R. A.; M€ohwald, H. Science 1998, 282(5391), 1111–
1114.
a shell thickness governed by the polymerization time were (11) Park, M. K.; Xia, C.; Advincula, R. C.; Sch€utz, P.; Caruso, F. Langmuir
obtained. The synthesis of fluorescent thermosensitive poly(N- 2001, 17(24), 7670–7674.
isopropylacrylamide) (PNIPAM) nanocapsules with tempera- (12) Wang, K. W.; He, Q.; Yan, X. H.; Cui, Y.; Qi, W.; Duan, L.; Li, J. B.
J. Mater. Chem. 2007, 17, 4018–4021.
ture-tunable size and shell permeability was reported by Gao (13) Xu, X.; Asher, S. A. J. Am. Chem. Soc. 2004, 126(25), 7940–7945.
et al.5 First, the cross-linked PNIPAM shell was prepared by (14) Bourgeat-Lami, E.; Lang, J. J. Colloid Interface Sci. 1998, 197(2), 293–308.
(15) Bourgeat-Lami, E.; Lang, J. J. Colloid Interface Sci. 1999, 210(2), 281–289.
precipitation polymerization in the presence of the isothiocyanate (16) Ali, M. M.; St€over, H. D. H. Macromolecules 2003, 36(6), 1793–1801.
fluorescein (FITC)-trapped silica particles serving as seeds. Then, (17) Ali, M. M.; St€over, H. D. H. J. Polym. Sci., Part A: Polym. Chem. 2006, 44
the SiO2 core was etched by hydrofluoric acid (HF), and the (1), 156–171.
(18) Perruchot, C.; Khan, M. A.; Kamitsi, A.; Armes, S. P.; Von Werne, T.;
Patten, T. E. Langmuir 2001, 17(15), 4479–4481.
*Corresponding author. E-mail: [email protected]. (19) Von Werne, T.; Patten, T. E. J. Am. Chem. Soc. 2001, 123(31), 7497–7505.
(1) Sukhorukov, G.; Fery, A.; M€ohwald, H. Prog. Polym. Sci. 2005, 30(8-9), (20) Fleming, M. S.; Mandal, T. K.; Walt, D. R. Chem. Mater. 2001, 13(6),
885–897. 2210–2216.
(2) Yang, S.; Zhang, Y.; Yuan, G.; Zhang, X.; Xu, J. Macromolecules 2004, 37 (21) Mori, H.; Seng, D. C.; Zhang, M.; M€uller, A. H. E. Langmuir 2002, 18(9),
(26), 10059–10062. 3682–3693.
(3) Chu, L. Y.; Yamaguchi, T.; Nakao, S. Adv. Mater. 2002, 14(5), 386–389. (22) Kamata, K.; Lu, Y.; Xia, Y. J. Am. Chem. Soc. 2003, 125(9), 2384–2385.
(4) Marinakos, S. M.; Novak, J. P.; Brousseau, L. C.III; House, A. B.; Edeki, E. (23) Mandal, T. K.; Fleming, M. S.; Walt, D. R. Chem. Mater. 2000, 12(11),
M.; Feldhaus, J. C.; Feldheim, D. L. J. Am. Chem. Soc. 1999, 121(37), 8518–8522. 3481–3487.
(5) Gao, H.; Yang, W.; Min, K.; Zha, L.; Wang, C.; Fu, S. Polymer 2005, 46 (24) Zhou, Q.; Wang, S.; Fan, X.; Advincula, R.; Mays, J. Langmuir 2002, 18(8),
(4 spec. iss.), 1087-1093. 3324–3331.

12084 DOI: 10.1021/la9017097 Published on Web 07/20/2009 Langmuir 2009, 25(20), 12084–12091
Rosenbauer et al. Article

Preparing polymeric nanocapsules without any sacrificial core

is more beneficial, since those capsules can be synthesized in fewer
steps and the target material can be placed inside the “core”
already in the beginning of the capsule process formation. There
are several techniques that are commonly used to obtain capsules
directly. For example, by the “ouzo effect”,25,26 an induced phase
separation within emulsion or miniemulsion droplets,27,28 the self-
assembly of block copolymers,29 cross-linking of polymerizable
liposomes,30 templating of vesicles,31,32 via interfacial polymer Figure 1. Lubrizol U (polyisobutylene-succinimide pentamine).
deposition under solvent displacement,33-35 or by interfacial Length of the polyisobutylene chain, n = 1-10.
cross-linking reactions.36,37
A highly promising way to create polymeric nanocapsules in a contrast agents, like Magnevist and Gadovist for applications in
controlled manner is via the miniemulsion process where mono- magnetic resonance imaging (MRI).46
disperse and stable droplets in the size range from 50 to 500 nm can For an efficient stabilization of the aqueous droplets in a
be formed. The size of the droplets mainly depends on the type and hydrophobic continuous phase, only a limited number of surfac-
effective amount of the surfactant used. The choices of monomers/ tants can be applied. Usually different types of sorbitan esters,
polymers and chemical reactions that can be utilized to form the known also as Span, or specially synthesized block copolymers
nanocapsules with a hydrophilic or hydrophobic core are almost are used. With the aim to increase the capsules stability and
unlimited. Therefore, this technique gives the opportunity to encapsulation efficiency of low molecular weight compounds, in
produce capsules with desired properties for a broad range of the current paper we used polyisobutylene-succinimide penta-
applications. As an example, poly(methyl methacrylate) nano- mine (Lubrizol U, for structure see Figure 1). Because of the
capsules with an aqueous core containing antiseptic agent were presence of amine groups, this molecule acts in polyaddition
obtained by controlled nanoprecipitation of a preformed polymer reaction as a monomer as well as a stabilizing agent at the same
onto the droplet’s surface.38 Van Zyl et al.39 reported the synthesis time and can therefore be considered as “surfmer”. After polym-
of nanocapsules with a liquid core and molar mass-controlled erization the surfmer is expected to be covalently bounded into the
polystyrene shell by an in situ miniemulsion polymerization polymeric interfacial layer, leading to the improved properties of
the capsules, e.g., more compact and impermeable shell. The
reaction performed in the presence of a RAFT (reversible addi-
influences of the surfactant and monomer concentration on the
tion-fragmentation chain transfer) agent. By carrying out the
shell thickness, colloidal stability, average capsule size, and
anionic polymerization at the interface of stable water-in-oil
capsule size polydispersity were examined in detail. Furthermore,
droplets, poly(butyl cyanoacrylate) nanocapsules containing
the conversion of monomer into polymer was studied by FTIR,
DNA molecules were prepared.40,41 Polyurethane nanocapsules
and the incorporation efficiency of Lubrizol was estimated from
with a hydrophobic core were obtained via the interfacial poly-
the GPC results.
addition in the direct miniemulsions.42,43 Previously, polyurethane
To show the versatility in application of the obtained capsules,
and polyurea nanocapsules with an aqueous core were successfully different materials, such as a water-soluble fluorescent dye and
formulated performing the interfacial polyaddition/polyconden- aqueous dispersion of magnetite nanoparticles with 10 nm in size,
sation reactions in the inverse miniemulsion system.44,45 Nano- were used as inner phase of the polyurea capsules. The encapsula-
capsules produced in such a way could be easily loaded with tion efficiency was studied using fluorescein as a marker.

(25) Ganachaud, F.; Katz, J. L. ChemPhysChem 2005, 6(2), 209–216. 2. Experimental Section
(26) Vitale, S. A.; Katz, J. L. Langmuir 2003, 19(10), 4105–4110.
(27) Dowding, P. J.; Atkin, R.; Vincent, B.; Bouillot, P. Langmuir 2004, 20(26), 2.1. Materials and Methods. 1,6-Diaminohexane (HMDA,
11374–11379. Fluka), toluene 2,4-diisocyanate (TDI, 98%, Aldrich), and cyclo-
(28) Tiarks, F.; Landfester, K.; Antonietti, M. Langmuir 2001, 17(3), 908–918. hexane (HPLC-grade) were used without further purification.
(29) Wong, M. S.; Cha, J. N.; Choi, K. S.; Deming, T. J.; Stucky, G. D. Nano Sodium dodecyl sulfate (SDS, 99% Merck) and Lubrizol U (poly-
Lett. 2002, 2(6), 583–587.
(30) Lestage, D. J.; Urban, M. W. Langmuir 2005, 21(10), 4266–4267. isobutylene-succinimide pentamine, Mw = 384-875 g mol-1,
(31) Hotz, J.; Meier, W. Adv. Mater. 1998, 10(16), 1387–1390. determined from GPC, HLB < 7, containing 50:50 w/w %
(32) Hotz, J.; Meier, W. Langmuir 1998, 14(5), 1031–1036. mineral oil as a diluent, Lubrizol, France) were used as surfac-
(33) Fessi, H.; Puisieux, F.; Devissaguet, J. P.; Ammoury, N.; Benita, S. Int. J. tants. Fluorescein (free acid, Riedel-de H€aen) was utilized as a
Pharm. 1989, 55(R1-R4), 1590.
(34) Crespy, D.; Landfester, K. Macromol. Chem. Phys. 2007, 208(5), 457–466. hydrophilic fluorescent marker. Demineralized water and phos-
(35) Alvarez-Roman, R.; Barre, G.; Guy, R. H.; Fessi, H., Eur. J. Pharm. phate saline buffer (PBS, 0.15 M, pH 7.4) were used as an aqueous
Biopharm. 2001, 52(2), 191–195. phase throughout the experiments.
(36) Soto-Portas, M. L.; Argillier, J. F.; Mechin, F.; Zydowicz, N. Polym. Int.
2003, 52(4), 522–527.
2.2. Preparation of Polyurea Capsules. For the synthesis
(37) Sun, Q.; Deng, Y. J. Am. Chem. Soc. 2005, 127(23), 8274–8275. of polyurea capsules, a certain amount of HMDA (for details see
(38) Paiphansiri, U.; Tangboriboonrat, P.; Landfester, K. Macromol. Biosci. Table 1) and 0.750 g of PBS buffer, which form the hydrophilic
2006, 6(1), 33–40. liquid cores, were homogenized for 3-5 min at room temperature.
(39) Van Zyl, A. J. P.; Bosch, R. F. P.; McLeary, J. B.; Sanderson, R. D.;
Klumperman, B. Polymer 2005, 46(11), 3607–3615.
This mixture was added to 6 g of cyclohexane containing a defined
(40) Musyanovych, A.; Landfester, K. Colloid Polym. Sci. 2008, in press. amount of Lubrizol U. After pre-emulsification at room tem-
(41) Lambert, G.; Fattal, E.; Pinto-Alphandary, H.; Gulik, A.; Couvreur, P. perature for 1 h with a magnetic stirrer (in order to obtain an
Pharm. Res. 2000, 17(6), 707–714. effective macroemulsion), the miniemulsion was prepared by
(42) Torini, L.; Argillier, J. F.; Zydowicz, N. Macromolecules 2005, 38(8), 3225–
3236.
ultrasonicating the mixture for 180 s at 70% amplitude
(43) Johnsen, H.; Schmid, R. B. J. Microencapsulation 2007, 24(8), 731–742. (Branson sonifier W450 Digital, tip size 6.5 mm) under ice cooling
(44) Landfester, K.; Tiarks, F.; Hentze, H. P.; Antonietti, M. Macromol. Chem. for preventing evaporation of cyclohexane. Then a defined
Phys. 2000, 201(1), 1–5. amount of TDI (for details see Table 1) dissolved in 4 g of
(45) Crespy, D.; Stark, M.; Hoffmann-Richter, C.; Ziener, U.; Landfester, K.
Macromolecules 2007, 40(9), 3122–3135.
cyclohexane was dropwisely added to the miniemulsion. The
(46) Jagielski, N.; Sharma, S.; Hombach, V.; Mail€ander, V.; Rasche, V.; Landfester, reaction was carried out for 3 h at 25 °C. The redispersion of
K. Macromol. Chem. Phys. 2007, 208(19-20), 2229–2241. the nanocapsules in the aqueous phase was performed by mixing

Langmuir 2009, 25(20), 12084–12091 DOI: 10.1021/la9017097 12085

Article Rosenbauer et al.

Table 1. Reaction Composition and Characterization of Synthesized Polyurea Capsulesa

average diameter, nm
b
HMDA/TDI, molar ratio HMDA, mol (mg) TDI, mol (mg) DLS (in cyclohexane) DLS (in aqueous phase) TEM (in cyclohexane)
-4 -4
1:0.4 6.5
10 2.6
10 - c
- c
-c
(1:0.39) (75) (28)
1:1 6.5
10-4 6.5
10-4 255 246 235
(1:0.97) (75) (112)
-4 -4
1:1.5 6.5
10 9.8
10 280 242 224
(1:1.46) (75) (170)
1:2 6.5
10-4 1.3
10-3 440 350 268
(1:1.94) (75) (225)
-4 -4
1.5:1 9.8
10 6.5
10 365 - c
-c
(1.5:0.98) (112) (112)
2:1 1.3
10-3 6.5
10-4 -c -c -c
(2:0.98) (150) (112)
a
Each sample contains 0.75 g of PBS buffer as an aqueous phase and 10 g of cyclohexane with 25 mg of dissolved Lubrizol U as continuous phase. b The
values in parentheses correspond to the HMDA þ Lubrizol U/TDI molar ratio. c The measurements were not possible due to the presence of broken
capsules or their strong aggregation.

1.0 g of the synthesized nanocapsule cyclohexane dispersion with

an aqueous solution of SDS (0.02 mol L-1) for 5 min. The
resulting mixture was subjected to the ultrasonication for 60 s at
70% amplitude, 10 s pulse, 5 s pause (Branson sonifier W450
Digital, tip size 6.5 mm) under ice cooling, and the cyclohexane
was evaporated afterward by heating the mixture up to 70 °C and
stirring (400 rpm) for 30 min until ∼1.0 g of the suspension was
removed.
Polyurea capsules containing the fluorescent marker were Figure 2. Schematic presentation of spinning drop experimental
prepared by adding 7.5
10-5 mol L-1 of fluorescein dye setup.
dissolved in PBS (0.15 M, pH 7.4) into the aqueous phase prior
to pre-emulsification of both phases. The presence of ions in PBS 20 °C for several minutes under rotation at 6000-8000 rpm until
converts free acid form of fluorescein into the salt form, therefore “macroscopic” droplets at the axis of rotation were observed. At
improving the water solubility of fluorescein. equilibrium, the shape of the droplet is a balance between inter-
The encapsulation of magnetite was performed as followed: facial (∼γ/a) and centrifugal (∼ΔFω2a2) stresses, where ΔF is the
75 mg of HMDA was mixed with 750 mg of an aqueous 4 wt % density difference between the phases.
magnetite dispersion prepared according to ref 47. The homo- The interfacial tension of the fluid can be obtained using the
genized solution was added dropwise to an organic phase of 6 g of Vonnegut formula:
cyclohexane and 50 mg of Lubrizol U. The resulting mixture was ðF1 -F2 Þω2 a3
miniemulsified for 180 s at 70% amplitude (Branson sonifier γ ¼
W450 Digital, tip size 6.5 mm) under ice cooling after pre- 4
emulsification for 1 h with a mechanical stirrer. Directly after In our experiment the L/a ratio of the droplet exceeds 4, stating
ultrasonication, a solution containing 170 mg of TDI and 4 g of that the droplet form is nearly cylindrical.48-50
cyclohexane was added to the reaction mixture, and the reaction FT-IR measurements were performed to follow the kinetics of
was carried out for 4 h at 25 °C under gentle stirring. the polyaddition. The samples were prepared using KBr pellets,
2.3. Characterization of Samples. The average size and size and the measurements were carried out on a FT-IR 113v Bruker
distribution of the final polymer capsules were determined by spectrophotometer equipped with a DTGS detector using 100
dynamic light scattering (DLS) using a Zeta Nanosizer (Malvern signal-averaged scans at a resolution of 2 cm-1. The solid used for
Instruments), equipped with a detector to measure the intensity of the IR measurements was obtained by freeze-drying the samples
the scattered light at 173° to the incident beam. Transmission from the cyclohexane or aqueous phases.
electron microscopy (TEM) (Philips EM400) was used to study The residual TDI concentration and the amount of nonreacted
the capsule morphology, the mean diameter, and the shell thick- surfactant Lubrizol U were evaluated from GPC data. The
ness of polymer capsules. An average of five TEM images of each samples were dissolved in cyclohexane, and the measurements
miniemulsion sample, which contain not less than 100 nanocap- were performed using an apparatus consisting of the Spectra
sules, were analyzed, and an average capsule size and shell System P2000 pump, an autosampler Agilent 1100, and two
thickness were determined. The interfacial tension between demi- detectors, the Shodex differential refractometer RI-71 and the
neralized water and the miniemulsion was determined using a Knauer UV variable wavelength monitor detector. Separation
commercial spinning drop tensiometer SVT 20N (DataPhysics, was performed at a flow rate of 1 mL min-1 using two mixed-bed
Germany). The obtained value was compared with the values of linear M columns (PSS, Mainz, Germany) with a particle size of
interfacial tension of cyclohexane/water and cyclohexane plus 5 μm. The amount of TDI and Lubrizol was calculated by
Lubrizol U/water in order to get the information about the integrating the area under the GPC curve. The calibration curve
coverage of the droplets with Lubrizol U. The working principle obtained with a known amount of TDI (or Lubrizol U) was used
of the spinning drop tensiometer is based on the theory derived by to calculate the residual amounts of TDI and Lubrizol U in the
Vonnegut.48 A glass capillary is filled with two fluids of interest, experimental samples.
whereas the fluid with the lower density being in a minority, e.g., The uptake studies into HeLa cells were studied after 24 h. For
water (∼1.2 g) and miniemulsion (∼0.012 g) (see Figure 2). Then, the incubation, HeLa cells were seeded at a density of 50 000 cells
the equipment tube was placed horizontally and equilibrated at
(49) Joseph, D. D.; Arney, M. S.; Gillberg, G.; Hu, H.; Hultman, D.; Verdier,
(47) Ramı́rez, L. P.; Antonietti, M.; Landfester, K. Macromol. Chem. Phys. C.; Vinagre, T. M. J. Rheol. 1992, 36(4), 621–662.
2006, 207(2), 160–165. (50) Princen, H. M.; Zia, I. Y. Z.; Mason, S. G. J. Colloid Interface Sci. 1967, 23
(48) Vonnegut, B. Rev. Sci. Instrum. 1942, 13(1), 6–9. (1), 99–107.

12086 DOI: 10.1021/la9017097 Langmuir 2009, 25(20), 12084–12091

Rosenbauer et al. Article

per cm2 on the first day. On the second day, fluorescein-contain-

ing capsules were added at a concentration of 75 μg mL-1 to the
media. The intracellular localization of the capsules was con-
firmed by confocal laser scanning microscopy (CLSM, Fluoview
on an IX71 equipped with two lasers, 488 and 543 nm, and a 60

oil lens, Olympus). The detailed procedure of the sample prepara-
tion is given in ref 51.
The concentration of encapsulated fluorescein was measured in
the supernatant obtained after centrifugation of the capsules
redispersed in water. The fluorescent spectra were performed on
a FluoroMax-3 spectrofluorometer (HORIBA Jobin Yvon, Inc.).
The excitation wavelength was 490 nm, and the emission was
measured at a wavelength of 520 nm. The calibration curve was
obtained from the fluorescent intensities of defined fluorescein
concentrations dissolved in SDS aqueous solution (0.02 mol L-1,
pH 7.0). The total content of the fluorescein inside the capsules
was calculated from the difference in fluorescent intensities
Figure 3. Synthesis of polyurea nanocapsules in the inverse
between the calibration curve and the measured samples.
(water-in-oil) miniemulsion system.
The thermal properties of the capsule shell and the amount of
encapsulated magnetite were determined from the thermogravi- results of our previous cell uptake experiments,51,52 we did not
metric analysis (TGA) performed on a Perkin-Elmer TGA appa- observe any toxic effects of SDS on the cell vitality (within the
ratus under a nitrogen atmosphere and heating rate of 10 °C min-1. experimental concentrations, i.e., 0.02 mol L-1).
Freeze-dried samples were used for the measurements. In the case
3.1. Variation of HMDA to TDI Ratio. In order to study
of magnetite-containing capsules, the nonencapsulated magnetite
was separated by gradient centrifugation, and the residual pellet of the influence of HMDA and TDI amounts on the average size and
capsules was used for TGA measurements. shell thickness of the capsules, the formulation process was
carried out with different molar ratios of these reagents keeping
3. Results and Discussion the other parameters constant. The reaction composition and
characterization of the final polyurea capsules are summarized in
A series of nanocapsules consisting of aqueous core and Table 1.
polyurea shell were prepared by an interfacial polyaddition The morphological studies of the obtained nanoparticles were
reaction at the water-in-oil droplets interface,45 as shown in performed by TEM. Some selected TEM images are presented in
Figure 3. In the beginning of the formulation process, HMDA Figure 4. On the basis of the DLS and TEM results, it can be
and fluorescent marker were mixed together in PBS buffer concluded that stable and well-defined polymeric capsules within
(0.15 M, pH 7.4), then added into the solution containing a size range between 255 and 440 nm were obtained when the
cyclohexane with different amounts of the Lubrizol U, and amount of HMDA was equal or less in molar ratio compared to
miniemulsified by ultrasonication, resulting in the narrow sized the TDI concentration. In the case of using an excess of TDI, after
droplets in a size range between 100 and 500 nm. The salts the consumption of the existing amine groups (of HMDA and
presented in buffer serve as a costabilizer and prevents the also of the surfactant, see below), the isocyanate groups can react
Ostwald ripening of the droplets. Because of the hydrophilic with a water leading to the primary amine groups formation.53
character of HMDA, it is dissolved inside the droplets which are These amine groups could then react with the isocyanate group
stabilized by the surfactant Lubrizol U. Afterward, a solution of again (if any still present) and be built into the polymeric chain.
the hydrophobic TDI in cyclohexane is dropwisely added to the For the same systems, the polydispersity index (PDI) was in the
miniemulsion. range of 0.1-0.2, indicating a narrow size distribution. It is
Because of the high reactivity of the amino groups in HMDA interesting to note that, according to the principle of miniemul-
compared to the hydroxyl groups in water, the hydrophilic groups sion technique, the capsule size should not be affected by the
of the diisocyanate react more rapidly with the HMDA than with amount of added TDI. In our case the increase in the average
water, especially at room temperature.42 Moreover, the presence diameter with the decrease in the HMDA/TDI molar ratio was
of the alkyl chain between two amino groups in HMDA deter- clearly observed. This could be assigned to the reaction of TDI
mines that this molecule is amphiphilic and preferably stays at the with Lubrizol U molecules dissolved in the continuous phase. In
droplet interface, and therefore, it can be assumed that the order to maintain the equilibrium between the concentration of
isocyanate groups can reach amino groups faster compared to surfactant molecules present at the droplets interface and in the
the water hydroxyls concentrated inside the droplet. The polymer continuous phase, some amount of Lubrizol U desorbs from the
is formed simultaneously via polyaddition reaction and precipi- droplets, resulting in their coalescence and subsequent bigger final
tates at the interfacial layer of the droplets. After synthesis in capsule size. This effect is not significant at low concentrations of
cyclohexane, the polyurea capsules were transferred into an used TDI. However, the final particles were almost 2 times larger
aqueous solution of SDS. Lubrizol U is an efficient surfactant than the size of the initial droplets when the HMDA to TDI molar
for a water-in-oil system, whereas SDS is required for the ratio was 1:2.
stabilization of nanocapsules in the aqueous medium in order From the TEM images it could be also seen that the polyurea
to be suitable for biomedical applications. Although SDS is particles have a low contrast interior surrounded by a high
known to be not biocompatible, in the present work it is used contrast (dark) polymeric shell (see Figure 4a,b). The capsule
as a model hydrophilic surfactant. Moreover, on the basis of the shell thickness usually varies in the range between 15 and 30 nm
and shows the tendency to increase with increasing amount of
(51) Musyanovych, A.; Schmitz-Wienke, J.; Mail€ander, V.; Walther, P.; Landfester, TDI introduced into the system. In the case of using high amounts
K., Macromol. Biosci. 2008, 8(2), 127–139.
(52) Lorenz, M. R.; Holzapfel, V.; Musyanovych, A.; Nothelfer, K.; Walther,
P.; Frank, H.; Landfester, K.; Schrezenmeier, H.; Mail€ander, V. Biomaterials 2006, (53) Borsus, J. M.; Jerome, R.; Teyssie, P. J. Appl. Polym. Sci. 1981, 26(9), 3027–
27(14), 2820–2828. 3043.

Langmuir 2009, 25(20), 12084–12091 DOI: 10.1021/la9017097 12087

Article Rosenbauer et al.

Figure 4. TEM images (from cyclohexane phase) of polyurea capsules or frazzles prepared at different HMDA/TDI molar ratios: (a) 1:2, (b)
1:1.5, (c) 2:1.

of HMDA compared to the TDI (e.g., 1:0.4; 1.5:1, or 2:1 molar

ratio), the DLS measurement data reveal different and not
reproducible size fractions with broad polydispersity values
(PDI > 0.5). This can be explained by the main production of
low-molecular-weight polymeric chains due to unreacted amino
groups, which does not lead to the formation of compact capsule
structure. The absence of the intact capsules and strong polymer
aggregation was also confirmed by the TEM studies (see
Figure 4c).
From Table 1, the DLS values are presenting the hydrody-
namic diameter, whereas the size of dry capsules was determined
using TEM. The DLS values measured in cyclohexane phase are
slightly higher compared to those measured in an aqueous phase
after redispersion step. The reason for this observation could be
assigned to the high mobility of the hydrophobic chains origi-
nated from the Lubrizol U molecules that are attached onto the Figure 5. Influence of the Lubrizol U concentration on the capsule
surface. This assumption was confirmed by TEM studies of the size; the HMDA/TDI molar ratio is 1:1.5. DLS values of polyurea
polyurea capsules prepared from the cyclohexane phase. The capsules dispersed in cyclohexane.
mean diameter of the capsules is slightly smaller due to the drying
effect but comparable with the DLS values measured in the information about the coverage of miniemulsion droplets with
aqueous phase, when the hydrophobic chains are strongly Lubrizol U, the interfacial tension between the miniemulsion and
adsorbed/collapsed onto the capsule surface. water was determined via spinning drop. For pure cyclohexane/
Thermogravimetric analysis was performed on polymeric cap- water system the interfacial tension was determined to be 48 mN/
sules obtained at a HMDA/TDI molar ratio of 1:1, 1:1.5, and 1:2 m (in the literature 50 mN/m54). In the case of saturated Lubrizol
in order to examine the thermal properties of the polymeric wall. U/cyclohexane solution the tension decreased up to 5.6 mN/m,
The weight loss at ∼287 °C was observed for all samples, indicating which was taken as a saturation value for system containing
high thermal stability of the produced polyurea capsules. inverse micelles. The interfacial tension of miniemulsion/water
3.2. Variation of the Surfactant (Lubrizol U) Amount. was measured to be 35 mN/m, therefore indicating the absence of
The surfactant is responsible for the stability of the miniemulsion micelles and incomplete coverage of the droplets with Lubrizol
droplets as well as the resulting nanocapsules. It decreases the U molecules.
interfacial tension of the droplets and prevents their coagulation. In the concentration range between 1.0 and 2.0 wt % the size
The size of the capsules can be adjusted by performing the changes negligibly, and only PDI value increases. However, as the
reaction in the presence of different surfactant amounts. The amount of Lubrizol U increases further, a strong aggregation of
influence of Lubrizol U concentration on the final capsule size the capsules was observed and dispersion shows instability. We
and shell thickness was studied by varying the surfactant weight assume that the main reason for this might be due to the fact that
ratio in cyclohexane. The amount of monomers was kept constant the critical micelle concentration (cmc) of Lubrizol U is reached.
at a HMDA/TDI molar ratio of 1:1.5; this means that a Although in the presence of 1 wt % Lubrizol U the mean diameter
considerable amount of isocyanate groups can be consumed for of the capsules was smaller, the amount of Lubrizol U corre-
the reaction with amine groups of the surfactant. The average size sponded to 0.25 and 0.5 wt % was found to be optimal in order to
of polyurea nanocapsules was determined by DLS, and the prepare narrow sized (PDI e 0.15) polyurea nanocapsules. In this
obtained results are plotted in Figure 5. Selected TEM images case, the capsule size was around 250 nm (after redispersion in
of polyurea nanocapsules manufactured with different amounts water) and the shell thickness between 15 and 20 nm.
of Lubrizol U are presented in Figure 6. 3.3. Polymer Formation Characterized by FT-IR Spec-
The obtained results show that with increasing surfactant con- troscopy. To estimate the amount of TDI that was converted into
centration from 0.075 to 1.0 wt % the capsule size rapidly decreases a polymer, FT-IR measurements were carried out on freeze-dried
from 1240 to 190 nm. This is in agreement with the theoretical samples of nanocapsules. The FT-IR spectra taken at different
approximations. About 15 mg of Lubrizol U (0.15 wt % on reaction times are shown in Figure 7a.
continuous phase) is required for complete coverage of the droplets,
although stable miniemulsion (for at least 1.5 h) was produced with (54) D’Ans, J.; Lax, E. Taschenbuch f€
ur Chemiker und Physiker, 4th ed.; Springer-
25 mg (0.25 wt % on continuous phase). Furthermore, to get the Verlag: Berlin, 1992; Vol. I.

12088 DOI: 10.1021/la9017097 Langmuir 2009, 25(20), 12084–12091

Rosenbauer et al. Article

Figure 6. TEM images of polyurea nanocapsules or frazzles obtained with different amounts of Lubrizol U (corresponding to the continuous
phase): (a) 0.075, (b) 0.5, and (c) 2.5 wt %. HMDA/TDI molar ratio is 1:1.5.

Figure 7. FTIR spectra of freeze-dried polyurea capsules prepared from cyclohexane phase: (a) conversion kinetics of toluene 2,4-
diisocyanate during 24 h (HMDA/TDI molar ratio is 1:1.5); (b) after transferring into aqueous phase.

Figure 8. Hydrolysis of isocyanate group in the aqueous phase to form amine groups.

The spectra reveal a strong broad band ranging from about molecule can react with the primary amine group of HMDA. In
3450 to 3200 cm-1 that usually corresponds to the combination of the final capsule, Lubrizol U is then covalently bounded in the
the stretching vibration bands of N-H groups in amine or imine outer polymeric layer, stabilizing the capsule with its hydrophobic
bonds of the polyurea as well as the surfactant Lubrizol U. The isobutylene chain oriented into the organic phase. Although the
adsorption band between 2950 and 2750 cm-1 corresponds to the reaction of primary amine groups is known to be faster in
C-H stretching vibrations in CH, CH2, and CH3 groups. From comparison to the secondary ones (∼100 times), it could be
the features of the NdCdO peak, which appears at 2275 cm-1, it possible that several amine groups from the Lubrizol U molecule
can be seen that the conversion of TDI is almost completed after participate in the polyaddition reaction with TDI. As a result, the
2 h of the reaction. The presence of a small residual peak of the formation of nanocapsules might be performed without the
isocyanate group which is visible after the reaction can be presence of HMDA. To prove this hypothesis, formation of
attributed to the groups, which are incorporated into the capsule nanocapsules at the droplet’s interface was carried out with
shell only by one-side reacting with HMDA. After transferring different amounts of Lubrizol U and without addition of HMDA.
the capsules into an aqueous phase, the isocyanate peak totally It was found that the capsules are formed when the concentration
disappears (see Figure 7b), as a fact of the hydrolysis reaction of surfactant exceeds 200 mg (2 wt % on continuous phase).
occurring between isocyanate groups and water to form amine The amount of Lubrizol U that was reacted with TDI during
groups via formation of an unstable carbamic acid intermediate53 the polymerization and the residual monomer concentration of
(Figure 8). This reaction occurs also during the capsule formation TDI after the polyaddition reaction were estimated from the
however to a low extent due to the presence of amine groups results of GPC analysis. The following molar ratios were studied:
which are more reactive compared to the water hydroxyls. Lubrizol U/TDI 1:3.4; 1:1.7; 1:1; 1:0.8; 1:0.4 (or Lubrizol U þ
3.4. Determination of the Lubrizol U Amount Bounded HMDA/TDI 1:1.42; 1:1.35; 1:1.27; 1:1.21; 1:1.01), keeping the
Covalently in the Polymeric Shell. The presence of amine other parameters constant. The obtained results demonstrate that
groups in the molecule of Lubrizol U allows incorporation of the independently from the initial concentration of Lubrizol U, the
surfactant into the capsule’s shell under the polyaddition reaction. total amount of the surfactant was covalently bounded into the
For example, one side of TDI molecule can react with the amine polyurea shell. Additionally, the FTIR spectra of the polyurea
group of Lubrizol U, whereas the other isocyanate group of the capsules synthesized with different Lubrizol U amounts (25 and

Langmuir 2009, 25(20), 12084–12091 DOI: 10.1021/la9017097 12089

Article Rosenbauer et al.

Table 2. Composition and Encapsulation Efficiency of Polyurea Nanocapsulesa

HMDA/TDI molar ratiob fluorescein solution,c mg PBS, mg fluorescein,d mol encapsulation efficiency,d %

1:1 (1:0.97) 375 375 2.8
10-8 95

1:1.5 (1:1.46) 250 500 1.9
10-8 95
-8
375 375 2.8
10 97
500 250 3.8
10-8 98
750 0 5.6
10-8 96
-8
1:2 (1:1.94) 375 375 2.8
10 97
a
Each sample contains 10 g of cyclohexane with 0.25 wt % of dissolved Lubrizol U as a continuous phase. The amount of dispersed phase was kept
constant at 750 mg. b The values in parentheses correspond to the HMDA þ Lubrizol U/TDI molar ratio. c Corresponds to a stock solution with a
concentration of 7.5
10-5 mol L-1. d Corresponds to the samples of polyurea capsule after redispersion in the SDS aqueous solution (0.02 mol L-1, pH 7.0).

400 mg) reveal that the NdCdO peak (at 2275 cm-1) is lower in calculated, knowing the exact concentration of dye at certain
the case of using higher amounts of Lubrizol U (data not shown). intensity from the calibration curve. The calibration curve was
However, due to the presence of -NH and -NH2 groups in evaluated from the defined concentrations of the fluorescein dye
Lubrizol molecule as well as in the reaction product, it is not in the SDS aqueous solution (0.02 mol L-1 at pH 7.0). The
possible to differentiate precisely how many and which groups influence of different monomer ratio on the encapsulation effi-
participated in the reaction. Therefore, we compared the shell ciency was studied at 0.25 wt % of Lubrizol U. The fluorescence
permeability of the capsules synthesized in the presence of measurements performed on freeze-dried samples after substitu-
Lubrizol U and with block copolymer surfactant, i.e., tion with SDS aqueous solution showed very low intensities,
(butylene-co-ethylene)-b-(ethylene oxide) (see Encapsulation Ef- confirming that the amount of nonencapsulated fluorescein in the
ficiency section). cyclohexane phase is negligible (less than 1 wt %). However, it is
The results of GPC analysis also provide the evidence that TDI worth to mention that due to the turbidity of the supernatant, the
is almost completely (more than 97 wt %) consumed during the experimental error of the measurements was relatively high. This
reaction. Therefore, final capsules can be immediately utilized was not the case for supernatants obtained after centrifugation of
after the synthesis without performing any deactivation or redispersed polyurea capsules. Furthermore, from the control
purification step. With the molar excess of TDI (HMDA/TDI = experiment it was found that after transferring the nanocapsules
1:1.5 or 1:2) the functional diisocyanate groups react covalently into the aqueous phase, the surface adsorbed fluorescein mole-
either with one, alternatively with two HMDA functional groups cules (if any) desorbs into the continuous (aqueous) phase. The
as well as with a amine group from the Lubrizol U. Both ratios results of fluorescence measurements are summarized in Table 2.
result in the formation of stable nanocapsules, and the only one As it can be seen from the Table 2, an encapsulation efficiency of
difference is the average size (see Table 1). Larger capsules more than 95% was achieved, indicating that the obtained
440 nm vs 280 nm (measured in cyclohexane) are produced at nanocapsules are composed of a stable and impermeable poly-
HMDA/TDI = 1:2. meric shell. For comparison, the encapsulation efficiency experi-
3.5. Encapsulation Efficiency. The encapsulation efficiency ments were performed with polyurea capsules stabilized with block
was studied by using the fluorescent dye as a hydrophilic marker. copolymer surfactant (butylene-co-ethylene)-b-(ethylene oxide)
Fluorescein was dissolved in the disperse phase, and the interfacial described previously.45 It was found that after several washing
polyaddition reaction was carried out resulting in fluorescent steps more than 50 wt % of the encapsulated material leaked out.
polyurea capsules. The determination of nonencapsulated fluor- In contrast, the leakage was less than 10% in the case of using
escein amount was performed on nanocapsule samples dispersed Lubrizol U, confirming that degree of the shell cross-linkage is
in cyclohexane (directly after polymerization) and in aqueous higher in contrast to (butylene-co-ethylene)-b-(ethylene oxide).
phase (after redispersion in SDS solution). Although from the 3.6. Uptake of Polyurea Capsules into the HeLa Cells. In
hydrophilic nature of the fluorescein it is clear that its molecules order to evidence one of the potential applications of the
will preferably stay inside the miniemulsion droplets before nanocapsules, for example, in a biomedical field, the fluores-
polymerization, the possibility of solubilization by the surfactant cein-containing polyurea capsules were used as markers in the
aggregates should not be excluded as well. cellular uptake experiments. The nanocapsules redispersed in the
Because of the low solubility of fluorescein in the cyclohexane, SDS aqueous solution were added to the HeLa cells, and after
it was not possible to measure the amount of nonencapsulated 24 h incubation the uptake efficiency was detected using fluor-
fluorescein directly in the supernatant obtained after centrifuga- escent confocal microscopy (see Figure 9). The experiments
tion of the capsules. Therefore, the capsule dispersion was freeze- confirmed that the capsules are not causing the cell death (more
dried first in order to remove the cyclohexane, substituted with a than 96% of the cells stay vital) and were efficiently internalized
known amount of SDS aqueous solution (0.02 mol L-1 at into the cell and not simply attached to the cell membrane.
pH 7.0), and then stirred for several hours in order to recover 3.7. Encapsulation of Magnetite. As was shown above and
all free fluorescein molecules. Afterward, the capsules were in the previous publications of our group, the miniemulsion
centrifuged down, and the supernatant was examined with a technique allows efficient entrapment of small organic molecules
fluorescence spectrometer on the presence of fluorescein. (e.g., fluorescein, as shown here), water-soluble complexes,46 and
The fluorescence of supernatant obtained after centrifugation shear-sensitive DNA molecules40 into the aqueous core nanocap-
of polyurea capsules redispersed in aqueous phase was measured sules. The presence of magnetite nanoparticles inside a capsule
as well. One could expect that during the redispersion step the offers promising drug delivery applications. In this regard,
usage of sonication might result in the breakage some of the magnetite nanoparticles of 10 nm dispersed in aqueous phase
capsules and therefore affect the final fluorescein concentration in and stabilized with SDS were used as a core material in order to
the continuous phase. synthesize the magnetic polyurea nanocapsules. The reaction was
By measuring the intensity of the fluorescein dye in the super- performed as described above, via interfacial polyaddition at
natant, the effectively encapsulated molar amount of dye can be the molar ratio of HMDA/TDI 1:1.5 and 50 mg of Lubrizol U.

12090 DOI: 10.1021/la9017097 Langmuir 2009, 25(20), 12084–12091

Rosenbauer et al. Article

The images reveal irregularly shaped capsules and the presence

of magnetite nanoparticles within a polymeric matrix. From the
scanning electron microscopy images (not shown) no magnetite
was found to be adsorbed onto the capsules surface. It could also
be seen that not all capsules are filled with magnetite and that the
distribution of magnetite within a capsule is not homogeneous.
These could be improved by variation of the sonication para-
meters or using higher initial concentrations of surface modified
(in order to avoid aggregation of magnetic nanoparticles) mag-
netite dispersion. Related experiments are now in progress, and
the results will be presented in a separate paper.

4. Conclusion
Polyurea nanocapsules consisting of an aqueous core were
synthesized by a polyaddition reaction at the inverse miniemul-
sion droplet interface. The use of amino-functionalized surfactant
Figure 9. Confocal fluorescent microscopy of HeLa cells after the (Lubrizol U) can improve the stability and impermeability of the
uptake of polyurea capsules (green). The capsules were added to final capsules through the formation of the covalent bonds
the HeLa cells in a concentration of 75 μg mL-1. between surfactant and polymeric shell. The FT-IR data obtained
from the kinetic experiments indicate that the conversion of TDI
is almost completed after 2 h of the reaction. The obtained
capsules were characterized for their average size, shell thickness,
morphology, and loading efficiency when produced with different
monomer molar ratio and surfactant concentration. It was found
that stable and well-defined polymeric capsules within a size range
between 255 and 440 nm were obtained when the amount of
HMDA (including Lubrizol U) was equal or less in molar ratio
compared to the TDI concentration. The capsule shell thickness
varies in the range between 15 and 30 nm, showing a tendency to
increase with increasing amount of TDI introduced into the
system. The encapsulation efficiency about 95% and 85% was
Figure 10. TEM images of magnetic polyurea nanocapsules (from achieved with water-soluble fluorescent dye molecules and mag-
cyclohexane phase). netite aqueous dispersion, respectively, thus indicating the effec-
tiveness and flexibility of the proposed miniemulsion approach.
The hydrodynamic diameter and the PDI value of the obtained Looking for a potential application, as an example, the fluor-
capsules were within the size range of the polyurea capsules escein-containing polyurea capsules redispersed in the SDS aqu-
prepared without magnetite (Dz = 273 nm, PDI = 0.14), eous solution were used as markers in the cell uptake studies. The
indicating that there is no influence of the magnetite on the obtained results confirmed that the capsules are not toxic and
interfacial polymerization process. The amount of encapsulated were efficiently internalized into the cell.
magnetite was estimated from the TGA curve performed with the
freeze-dried capsules. The obtained results demonstrate that Acknowledgment. This research was supported by German
about 85% of the introduced magnetite was encapsulated, that Research Foundation (DFG). The authors thank Julia Dausend
corresponds to 0.11 gmagnetite per gpolymer. The morphology of and Dr. Volker Mail€ander for help in cell uptake experiments and
nanocapsules was studied by TEM (see Figure 10). discussions.

Langmuir 2009, 25(20), 12084–12091 DOI: 10.1021/la9017097 12091