Rapid self-assembly of core-shell organosilicon microcapsules within a microfluidic device

Article abstract of DOI:10.1021/ja0612403

The preparation of hierarchically structured organosilicon microcapsules from commercially available starting materials is described. Using a microfluidic device, an emulsion of dichlorodiphenylsilane is formed in a continuous phase of aqueous glycerol. The silane droplets undergo hydrolysis, condensation, and crystallization within minutes to form self-assembled, core-shell microcapsules. The microparticles have been characterized with light and electron microscopy, nuclear magnetic resonance spectroscopy (NMR), diffusion-ordered NMR spectroscopy (DOSY), Fourier transform infrared spectroscopy (FTIR), differential scanning calorimetry (DSC), and powder X-ray diffraction (XRD). The characterization data show that the microcapsule walls consist of amorphous, oligomeric poly(diphenylsiloxane) surrounded by a spiny layer of crystalline diphenylsilanediol. Glycerol is occluded within the wall material but is not covalently bound to the silicon components. Glycerol is a crucial element for producing low-dispersity microcapsules with well-ordered surface spines, as the use of methyl cellulose as viscomodifier yields amorphous surfaces.

Full text of DOI:10.1021/ja0612403

Published on Web 07/01/2006
Rapid Self-Assembly of Core-Shell Organosilicon
Microcapsules within a Microfluidic Device
Jeremy L. Steinbacher,† Rebecca W. Y. Moy,† Kristin E. Price,†
Meredith A. Cummings,§ Chandrani Roychowdhury,† Jarrod J. Buffy,|
William L. Olbricht,‡ Michael Haaf,§ and D. Tyler McQuade*,†
Contribution from the Department of Chemistry and Chemical Biology, Cornell UniVersity,
Ithaca, New York 14853-1301, Department of Chemical and Biomolecular Engineering,
Cornell UniVersity, Ithaca, New York 14853-5201, Department of Chemistry, Ithaca College,
361 Center for Natural Sciences, Ithaca, New York 14850, and Department of Chemistry,
UniVersity of Minnesota, Minneapolis, Minnesota 55455
Received February 21, 2006; Revised Manuscript Received May 24, 2006; E-mail: dtm25@cornell.edu
Abstract: The preparation of hierarchically structured organosilicon microcapsules from commercially
available starting materials is described. Using a microfluidic device, an emulsion of dichlorodiphenylsilane
is formed in a continuous phase of aqueous glycerol. The silane droplets undergo hydrolysis, condensation,
and crystallization within minutes to form self-assembled, core-shell microcapsules. The microparticles
have been characterized with light and electron microscopy, nuclear magnetic resonance spectroscopy
(NMR), diffusion-ordered NMR spectroscopy (DOSY), Fourier transform infrared spectroscopy (FTIR),
differential scanning calorimetry (DSC), and powder X-ray diffraction (XRD). The characterization data show
that the microcapsule walls consist of amorphous, oligomeric poly(diphenylsiloxane) surrounded by a spiny
layer of crystalline diphenylsilanediol. Glycerol is occluded within the wall material but is not covalently
bound to the silicon components. Glycerol is a crucial element for producing low-dispersity microcapsules
with well-ordered surface spines, as the use of methyl cellulose as viscomodifier yields amorphous surfaces.
Introduction
assembly has been employed to produce microcapsules and
colloidosomes of varied packing and surface characteristics11-13
Self-assembly1-3 is an important tool for organizing matter
at small length scales. Nature self-assembles complex systems
ranging from proteins4 to nanobiocomposites.5 Researchers have
expanded the scope of molecular self-assembly in recent years
to include nanotubes,6 supramolecular electronic materials,7
ordered polypeptide hydrogels,8 bilayered pyramidal nano-
columns,9 and metal nanostripes.10 At larger length scales, self-
and even macroscopic objects.14,15 A complication inherent
in many self-assembled systems is the extensive synthesis
required to generate precursors. Here, we present an alternative
method involving the rapid self-assembly of structured organo-
silicon microcapsules using commercially available starting
materials.
Hollow and solid microcapsules, such as those reported here,
are particularly interesting because of applications in diverse
fields, ranging from the encapsulation of drugs16 to use as
catalysts.17,18 Though microcapsules have been used for decades,
recent advances such as rapid prototyping19 and soft lithogra-
phy20 have brought the field of microfluidics to the forefront
† Department of Chemistry and Chemical Biology, Cornell University.
‡ Department of Chemical and Biomolecular Engineering, Cornell
University.
§ Ithaca College.
| University of Minnesota.
(1) Lehn, J. M. Angew. Chem., Int. Ed. Engl. 1990, 29, 1304-1319.
(2) Lindsey, J. S. New J. Chem. 1991, 15, 153-180.
(3) Whitesides, G. M.; Mathias, J. P.; Seto, C. T. Science 1991, 254, 1312-
1319.
(11) Noble, P. F.; Cayre, O. J.; Alargova, R. G.; Velev, O. D.; Paunov, V. N.
J. Am. Chem. Soc. 2004, 126, 8092-8093.
(12) Subramaniam, A. B.; Abkarian, M.; Stone, H. A. Nat. Mater. 2005, 4, 553-
556.
(4) Brenner, S. L.; Zlotnick, A.; Griffith, J. D. J. Mol. Biol. 1988, 204, 959-
972.
(5) Braun, P. V. Natural Nanobiocomposites, Biomimetic Nanocomposites, and
Biologically-Inspired Nanocomposites. In Nanocomposite Science and
Technology; Ajayan, P. M., Schadler, L. S., Braun, P. V., Eds.; Wiley-
VCH: Weinheim, Germany, 2003.
(13) Hsu, M. F.; Nikolaides, M. G.; Dinsmore, A. D.; Bausch, A. R.; Gordon,
V. D.; Chen, X.; Hutchinson, J. W.; Weitz, D. A. Langmuir 2005, 21,
2963-2970.
(6) Raez, J.; Manners, I.; Winnik, M. A. J. Am. Chem. Soc. 2002, 124, 10381-
(14) Bowden, N.; Terfort, A.; Carbeck, J.; Whitesides, G. M. Science 1997,
276, 233-235.
10395.
(7) Percec, V.; Glodde, M.; Bera, T. K.; Miura, Y.; Shiyanovskaya, I.; Singer,
K. D.; Balagurusamy, V. S. K.; Heiney, P. A.; Schnell, I.; Rapp, A.; Spiess,
H. W.; Hudson, S. D.; Duan, H. Nature 2002, 419, 384-387.
(8) Wright, E. R.; McMillan, R. A.; Cooper, A.; Apkarian, R. P.; Conticello,
V. P. AdV. Funct. Mater. 2002, 12, 149-154.
(15) Gracias, D. H.; Tien, J.; Breen, T. L.; Hsu, C.; Whitesides, G. M. Science
2000, 289, 1170-1172.
(16) Thies, C. In Drugs and the Pharmaceutical Sciences; Benita, S., Ed.; Marcel
Dekker: New York, 1996; Vol. 73, pp 1-19.
(17) Ley, S. V.; Ramarao, C.; Gordon, R. S.; Holmes, A. B.; Morrison, A. J.;
McConvey, I. F.; Shirley, I. M.; Smith, S. C.; Smith, M. D. Chem. Commun.
2002, 1134-1135.
(9) Percec, V.; Imam, M. R.; Bera, T. K.; Balagurusamy, V. S. K.; Peterca,
M.; Heiney, P. A. Angew. Chem., Int. Ed. 2005, 44, 4739-4745.
(10) Huang, J. X.; Kim, F.; Tao, A. R.; Connor, S.; Yang, P. D. Nat. Mater.
2005, 4, 896-900.
(18) Kobayashi, S.; Akiyama, R. Chem. Commun. 2003, 449-460.
(19) Qin, D.; Xia, Y. N.; Whitesides, G. M. AdV. Mater. 1996, 8, 917-919.
9
9442
J. AM. CHEM. SOC. 2006, 128, 9442-9447
10.1021/ja0612403 CCC: $33.50 © 2006 American Chemical Society

Self-Assembly of Core−Shell Organosilicon Microcapsules
A R T I C L E S
of their preparation. Due to the small dimensions involved,
microfluidic fluid dynamics are predominantly characterized by
laminar flow, leading to unique fluid morphologies21-23 and the
production of monodisperse emulsions.24,25 To create micro-
particles, the emulsions are subsequently captured through
polymerization or coascervation. For example, flow-through
channel arrays have been used to create monodisperse micro-
emulsions26,27 that were subsequently polymerized via UV
initiation to form solid microspheres.28 Straightforward geom-
etries such as T-intersections have yielded monodisperse single
and double emulsions29 that have been polymerized to form solid
acrylic microspheres and biphasic microspheres30 as well as
nonspherical plugs and disks.31 Microfluidic flow focusing
devices32 have also been used recently to create single and
double emulsions that yielded solid acrylic microspheres, disks,
and rods33,34 in addition to multicored acrylic microspheres35,36
and dye-labeled microspheres.33,34,36 The above two-dimensional
flow focusing devices have evolved into three-dimensional,
coaxial systems such as those employed to produce hollow nylon
capsules,37 hollow cured adhesive capsules,38 solid enzyme-
containing acrylic microspheres,39 and continuously produced
hollow and solid microfibers.40
Figure 1. Schematic of the microfluidic device and the formation of the
self-assembled microcapsules.
Figure 2. Light microscope images of (A) a population of microcapsules
and (B) a single capsule.
Results and Discussion
The core-shell organosilicon microcapsules were prepared
in a simple microfluidic device41 made from flexible PVC
tubing. A continuous phase of aqueous glycerol flowed through
the tube and a disperse phase of dichlorodiphenylsilane (Cl2Ph2Si)
was introduced through a small-gauge needle inserted orthogonal
to the continuous phase flow, forming a T-junction in the center
of the channel (Figure 1, Figure S1). Disposable syringes and
syringe pumps supplied the fluids, allowing independent control
of each phase. The enhanced viscosity of the glycerol solutions
(η ) 11-60 mPa‚s) provided viscous confinement of the
disperse phase. Otherwise, the Cl2Ph2Si wet and spread along
the PVC channel walls after roughly a minute of flow in a water
continuous phase, forming polydisperse microcapsules. When
glycerol or 1% methyl cellulose was employed, no channel
wetting occurred and conversion was quantitative based on mass
yield.
The morphology of the microparticles was characterized by
light microscopy and scanning electron microscopy (SEM).
Light microscope images reveal spheroidal microparticles with
diameters of 100-600 µm (Figure 2A) with serrated edges
visible at higher magnifications (Figure 2B). Microcapsule size
is governed by the radius r of the drop that forms in the
apparatus in Figure 1. Predicting this size is difficult, but it is
reasonable to expect that it would depend on the Reynolds and
capillary numbers (Re and Ca, respectively), interfacial surface
tension (γ), needle radius (rn), the tube radius (R), and the ratios
of disperse to continuous phase viscosities (η and ηd), densities
(F and Fd), and velocities (V and Vd). Then, dimensional analysis
yields eq 1.
Though these efforts are exciting developments in micro-
particle preparation and microfluidics, all of the investigations
to date have created particles whose morphologies closely
mirrored those of the templating fluids. Here, we present the
microfluidic preparation of microparticles possessing structured
morphologies that differ from the fluids from which they were
prepared. The initially homogeneous emulsion droplets self-
assemble within minutes of the emulsification event, forming a
hollow, core-shell morphology.
(20) Xia, Y. N.; Whitesides, G. M. Angew. Chem., Int. Ed. 1998, 37,
551-575.
(21) Thorsen, T.; Roberts, R. W.; Arnold, F. H.; Quake, S. R. Phys. ReV. Lett.
2001, 86, 4163-4166.
(22) Dreyfus, R.; Tabeling, P.; Willaime, H. Phys. ReV. Lett. 2003, 90.
(23) Zheng, B.; Tice, J. D.; Ismagilov, R. F. Anal. Chem. 2004, 76, 4977-
4982.
(24) Garstecki, P.; Stone, H. A.; Whitesides, G. M. Phys. ReV. Lett. 2005, 94.
(25) Cristini, V.; Tan, Y. C. Lab Chip 2004, 4, 257-264.
(26) Sugiura, S.; Nakajima, M.; Iwamoto, S.; Seki, M. Langmuir 2001, 17,
5562-5566.
(27) Kobayashi, I.; Mukataka, S.; Nakajima, M. Langmuir 2005, 21, 7629-
7632.
(28) Sugiura, S.; Nakajima, M.; Itou, H.; Seki, M. Macromol. Rapid Commun.
2001, 22, 773-778.
(29) Okushima, S.; Nisisako, T.; Torii, T.; Higuchi, T. Langmuir 2004, 20,
9905-9908.
(30) Nisisako, T.; Torii, T.; Higuchi, T. Chem. Eng. J. 2004, 101, 23-29.
(31) Dendukuri, D.; Tsoi, K.; Hatton, T. A.; Doyle, P. S. Langmuir 2005, 21,
2113-2116.
(32) Anna, S. L.; Bontoux, N.; Stone, H. A. Appl. Phys. Lett. 2003, 82, 364-
366.
(33) Seo, M.; Nie, Z. H.; Xu, S. Q.; Mok, M.; Lewis, P. C.; Graham, R.;
Kumacheva, E. Langmuir 2005, 21, 11614-11622.
(34) Xu, S. Q.; Nie, Z. H.; Seo, M.; Lewis, P.; Kumacheva, E.; Stone, H. A.;
Garstecki, P.; Weibel, D. B.; Gitlin, I.; Whitesides, G. M. Angew. Chem.,
Int. Ed. 2005, 44, 724-728.
FVR
η
ηV
γ
V η F R
Ca Vd ηd Fd rn
r ) f
,
,
,
,
,
(1)
(35) Nie, Z. H.; Xu, S. Q.; Seo, M.; Lewis, P. C.; Kumacheva, E. J. Am. Chem.
Soc. 2005, 127, 8058-8063.
(
(
)
(
)
)
Re
(36) Martin-Banderas, L.; Flores-Mosquera, M.; Riesco-Chueca, P.; Rodriguez-
Gil, A.; Cebolla, A.; Chavez, S.; Ganan-Calvo, A. M. Small 2005, 1, 688-
692.
Once the apparatus geometry and fluid compositions are fixed,
the drop size can vary with the Reynolds and capillary numbers
and the ratio of flow rates of continuous and disperse phase. If
the disperse phase flow rate also is fixed, then the drop size
(37) Takeuchi, S.; Garstecki, P.; Weibel, D. B.; Whitesides, G. M. AdV. Mater.
2005, 17, 1067-1072.
(38) Utada, A. S.; Lorenceau, E.; Link, D. R.; Kaplan, P. D.; Weitz, D. A. Science
2005, 308, 537-541.
(39) Jeong, W. J.; Kim, J. Y.; Choo, J.; Lee, E. K.; Han, C. S.; Beebe, D. J.;
Seong, G. H.; Lee, S. H. Langmuir 2005, 21, 3738-3741.
(40) Jeong, W.; Kim, J.; Kim, S.; Lee, S.; Mensing, G.; Beebe, D. J. Lab Chip
2004, 4, 576-580.
(41) Quevedo, E.; Steinbacher, J.; McQuade, D. T. J. Am. Chem. Soc. 2005,
127, 10498-10499.
9
J. AM. CHEM. SOC. VOL. 128, NO. 29, 2006 9443

A R T I C L E S
Steinbacher et al.
reveal an exterior coated with ∼10-µm long crystals (Figure
5B). Images of microcapsules produced at various continuous
phase flow rates (Figure S2) show that the spiny structure is
little affected by flow rate. The hierarchical structure was
revealed by examining manually broken microcapsules. Frag-
ments viewed parallel to the shell wall show that the micro-
spheres are hollow (Figure 5C). The walls comprise an
amorphous layer surrounded by an outer layer of crystals. Also,
SEM images reveal spines with needle-shaped and prismatic
crystals, which correspond to the pinacoid and prism forms,
respectively, of the triclinic system reported for pure Ph2Si-
(OH)2.44 This observation led us to hypothesize that the
crystalline, exterior spines are composed of hydrolyzed mono-
mer and the interior, amorphous layer is oligomeric or polymeric
diphenylsiloxane.
Figure 3. Plots of mean particle size in µm, r, versus continuous phase
flow rate for 70% (gray diamonds, thick line) and 80% (black squares, thin
line) aqueous glycerol (v/v) continuous phases. Error bars correspond to 1
standard deviation from the mean diameters.
To confirm the composition of the microcapsules, we
performed attenuated total reflectance Fourier transform infrared
spectroscopy. The absorbances corresponding to aromatic C-H
and C-C, Si-C, and Si-O bonds in addition to terminal
hydroxyl groups, consistent with Ph2Si(OH)2 and hydroxyl-
terminated oligomers, are present (Figure S3).
For more detailed characterization, solid state and solution
NMR spectroscopy were performed. Solid state 29Si NMR
spectra (Figure S4) indicated three unique silicon nuclei (δ )
62, 71, 80 ppm), most likely corresponding to hydrolyzed
monomer, Ph2Si(OH)2, and the terminal and internal Si nuclei
of diphenylsiloxane oligomers. To better clarify structure,
microcapsules dissolved in d6-DMSO were analyzed using
solution NMR spectroscopy. The 29Si NMR spectrum (Figure
S5) indicated the presence of Ph2Si(OH)2 (δ ) -34.54 ppm)
with decreasing amounts of the dimeric 1,1,3,3-tetraphenyl-
disiloxane (δ ) -40.25 ppm),45 the trimer (δ ) -40.36, -40.71
ppm),45 and the siloxane repeat unit (δ ) -46.77 ppm).46 1
H
Figure 4. Histograms of particle size produced at four flow rates, as
indicated, using a continuous phase of 70% aqueous glycerol.
NMR spectroscopy (Figure S6) confirms the presence of
Ph2Si(OH)2 (δ ) 7.61, m; 7.34, m), associated oligomers (δ )
7.20-7.60, m; 7.09-7.59, m), and a significant amount of
glycerol (δ ) 3.29, quartet; 3.37, quartet; 3.43, quintet).
To determine whether the glycerol was physically or co-
valently incorporated into the siloxane oligomers, further NMR
can depend only on the continuous phase flow rate. When tested
experimentally, we found that the microparticle size could be
controlled by adjusting the continuous phase flow rate in both
70% and 80% aqueous glycerol systems (Figure 3). The
measured particle size distributions have low diameter CVs of
6-9%, nearly qualifying as monodisperse (Figure 4).42
The droplet radius, r, may be quantitatively estimated by the
following equation:
1
studies were performed. H-29Si heteronuclear multiple bond
correlation47 experiments showed no correlation between silicon
atoms of the monomeric or oligomeric species and the protons
of the glycerol (Figure S7). The lack of three-bond correlation
points to chemically distinct compounds. However, because a
lack of signal is inconclusive, diffusion-ordered NMR spec-
troscopy48,49 was used to determine whether a covalent linkage
existed. Since this technique relies on translation of the
molecules in solution, the silicon oligomers should move with
the glycerol if chemically linked. If, on the other hand, no
covalent bond exists, the two will diffuse separately. The DOSY
pseudo 2-D spectrum shows that, on average, the glycerol
diffuses faster than all of the silicon-containing compounds
γ
η ꢀ˘
r ≈
(2)
where γ and η are as above and ꢀ˘ is the shear rate.43 Because
the shear rate is approximately proportional to the continuous
phase velocity (V),21 capsule size should be proportional to V-1
.
However, the power law regressions to the data above yield a
dependence roughly proportional to V-0.5. The deviation from
the expected dependence could arise from confinement effects
due to the channel walls or changes in the interfacial tension
associated with the rapidly polymerizing capsule walls.
SEM images show that the microcapsules are spheroidal with
rough surfaces (Figure 5A). Higher levels of magnification
(44) Kantz, M. R.; Desando, R. J. Microscope 1968, 16, 65-69.
(45) Behbehani, H.; Brisdon, B. J.; Mahon, M. F.; Molloy, K. C.; Mazhar, M.
J. Organomet. Chem. 1993, 463, 41-45.
(46) Williams, E. A. NMR Spectroscopy of Organosilicon Compounds. In The
Chemistry of Organosilicon Compounds; Patai, S., Rappaport, Z., Eds.;
Wiley: New York, 1989; Vol. 1, pp 511-554.
(47) Claridge, T. D. W. High-Resolution NMR Techniques in Organic Chemistry;
Elsevier: Amsterdam, 1999.
(42) Jillavenkatesa, A.; Dapkunas, S. J.; Lum, L. H. Particle Size Characteriza-
tion. In National Institute of Standards and Technology Special Publication
960-1; U.S. Government Printing Office: Washington, DC, 2001; p 149.
(43) Taylor, G. I. Proc. R. Soc. London, Ser. A 1934, 146, 0501-0523.
(48) Johnson, C. S. Prog. Nucl. Magn. Reson. Spectrosc. 1999, 34, 203-256.
(49) Stchedroff, M. J.; Kenwright, A. M.; Morris, G. A.; Nilsson, M.; Harris,
R. K. Phys. Chem. Chem. Phys. 2004, 6, 3221-3227.
9
9444 J. AM. CHEM. SOC. VOL. 128, NO. 29, 2006

Self-Assembly of Core−Shell Organosilicon Microcapsules
A R T I C L E S
Figure 5. SEM images of organosilane microcapsules. (A) Population of typical, spheroidal microcapsules produced in the microfluidic device. (b) Higher
magnification image of the spinulose surface. (C) A microcapsule fragment showing the inner amorphous shell and the outer spiny layer.
Figure 7. Crystalline Ph2Si(OH)2 constitutes the surface spines and
Figure 6. Pseudo 2-D DOSY spectrum of the structured microcapsules
dissolved in d6-DMSO. The aromatic protons of the organosilicon compo-
nents clearly diffuse more slowly than the protons corresponding to the
glycerol.
oligomeric diphenylsiloxane, the amorphous shell.
occluded within the crystals rather than chemically bonded to
silicon species (Figure 6).
We suggest that the hierarchical structures of the spiny
microcapsules begin assembling immediately after the Cl2Ph2Si
droplets snap-off at the T-junction (Scheme 1). The outermost
Cl2Ph2Si first hydrolyzes to Ph2Si(OH)2, creating an interfacial
layer of mixed composition between the Cl2Ph2Si droplet and
water/glycerol phase. We expect that this hydrolysis creates a
water-depleted, glycerol-rich corona surrounding the droplet.
HCl generated via hydrolysis catalyzes diol-diol and diol-
silyl chloride condensation, producing oligomeric diphenyl-
siloxane near the surface of the droplet. As the interfacial layer
becomes further depleted in water, the relative rates of hydrolysis
and oligomerization become significant factors of microcapsule
morphology. Eventually, crystals of Ph2Si(OH)2 nucleate on top
of the oligomeric shell, and the rate of crystallization becomes
an important influence as well.52-54 Diffusion of glycerol into
the interior and Cl2Ph2Si out of the droplet interior to the
interface causes thickening of the oligomeric shell as well as
outward growth of the spiny crystals. The rate of outward crystal
growth is faster than shell thickening, leading to evacuation of
silane from the droplet core, which is presumably replaced by
(Figure 6). The difference in diffusion rate indicates that the
glycerol is not covalently attached to the oligomers.
Differential scanning calorimetry on microcapsule samples
indicates an endothermic transition between 137 and 152 °C
on heating, followed immediately by an exothermic transition
between 160 and 172 °C (Figure S8). No further transitions
were observed on the cooling ramp or the second heating ramp.
We propose that the endothermic transition corresponds to the
Tm of Ph2Si(OH)2, which melts between 132 and 174 °C,
depending on purity.44,50 During the subsequent exothermic
transition, the hydrolyzed monomer condenses completely to
oligomeric or polymeric poly(diphenylsiloxane). On the cooling
and second heating ramps, no further transitions were observed
due to this heat-induced condensation. Finally, the X-ray powder
diffraction pattern (Figure S9), though exhibiting broad peaks
due to the composite nature of the microcapsules, is consistent
44
with a combination of patterns reported for pure Ph2Si(OH)2
and poly(diphenylsiloxane).51
Based on the characterization experiments, we assert that the
amorphous core of the capsule is comprised of oligomeric
diphenylsiloxane and the outer spines are pure, crystalline
Ph2Si(OH)2 (Figure 7). Also, glycerol is only physically
(52) Zheng, B.; Roach, L. S.; Ismagilov, R. F. J. Am. Chem. Soc. 2003, 125,
11170-11171.
(53) Chen, D. L.; Gerdts, C. J.; Ismagilov, R. F. J. Am. Chem. Soc. 2005, 127,
(50) Takiguchi, T. Bull. Chem. Soc. Jpn. 1959, 32, 556-557.
(51) Grigoras, S.; Qian, C.; Crowder, C.; Harkness, B.; Mita, I. Macromolecules
1995, 28, 7370-7375.
9672-9673.
(54) Zheng, B.; Gerdts, C. J.; Ismagilov, R. F. Curr. Opin. Struct. Biol. 2005,
15, 548-555.
9
J. AM. CHEM. SOC. VOL. 128, NO. 29, 2006 9445

A R T I C L E S
Steinbacher et al.
Scheme 1. Formation Mechanism of the Hierarchically Structured Microcapsules
1
continuous phase material. Thus, hollow, hierarchically struc-
tured capsules assemble from originally homogeneous droplets
within minutes.
Given a channel diameter of /16” (1.59 mm) and length of
30 cm, residence times ranged between 4.0 and 60 s. Though
the reaction begins immediately, residence times are too short
for complete reaction in the channel. Visual inspection suggests
that reaction is complete within 5-10 min. Moreover, a rapid
isolation experiment was performed whereby the microcapsule
formation was quenched via isolation and drying within 5 min
of droplet emulsification. These microcapsules showed no
discernible differences between those allowed to mature for
2 h.
A degree of control over the crystal morphology is achievable
by varying the composition of the continuous phase. For
reference, a standard continuous phase of 70% glycerol yields
the spines shown in Figure 8A. If a pure water continuous phase
is used, misshapen microcapsules with less dense, higher-
dispersity spines are produced briefly before the microfluidic
device clogs (clogging is not observed when glycerol is
used) (Figure 8B). On the other hand, the use of 0.66% (w/v)
aqueous methyl cellulose results in spineless microcapsules
(Figure 8C). This concentration of methyl cellulose produces a
viscosity approximately equal to 70% (v/v) aqueous glycerol
(22 mPa‚s), which should provide similar diffusion character-
istics. The polymeric methyl cellulose apparently significantly
slows the crystallization rate of Ph2Si(OH)2 on the microcapsule
surface. Finally, we varied the pH of the continuous phase.
While acidic to weakly basic continuous phases (pH ) 0.38,
4.70, 7.0, and 9.26) had little effect on the capsule morphology,
strongly basic continuous phases (pH ≈ 14) led to rapid
decomposition when capsules remained suspended in the
continuous phase effluent.
Microcapsules with vastly different characteristics may be
produced by using silanes other than Cl2Ph2Si. For example,
phenyltrichlorosilane (PhCl3Si) results in collapsed micro-
capsules with smooth, folded surfaces (Figure 9 A). These
collapsed microcapsules do not reinflate in any solvents, owing
to the stiff, inflexible phenylsilsesquioxane structure. On the
other hand, the use of silicon tetrachloride (SiCl4) as the disperse
phase forms spherical, hollow silica microspheres (Figure 9B).
We hypothesize that they are cracked due to the evolution of
HCl gas during the hydrolysis reaction. The rapidly expanding
gas cannot diffuse through the nonporous silica walls. Thus,
the pressure inside the microcapsules increases until the fracture
stress of the walls is reached, rupturing the walls and allowing
the gas to escape. These two examples illustrate that the
dichlorodiphenylsilane has unique properties relative to other
compounds with similar structures.
Conclusion
In this report, we have described the microfluidic preparation
of organosilicon-based microcapsules that possess high-surface-
area spiny exteriors, representing structure added beyond the
templating fluids. Using a variety of characterization techniques,
we have shown that the capsules are composed of an inner layer
of amorphous, oligomeric diphenylsiloxane and an outer layer
of spiny, crystalline Ph2Si(OH)2. The use of glycerol as a
viscomodifier is an important component for producing low-
dispersity, structured microcapsules. The high concentration of
glycerol results in a water-depleted corona around the as-
sembling capsules, creating an environment where crystallization
is accelerated relative to oligomerization. On the other hand, a
low concentration polymeric viscomodifier, methyl cellulose,
results in amorphous microcapsules without spines. Finally,
unique microcapsule structures may be formed by using other
silane starting materials.
Figure 8. Comparison of microcapsule surfaces prepared with continuous
phase compositions of (A) 70% (v/v) aqueous glycerol, (B) pure water,
and (C) 0.66% (w/v) aqueous methyl cellulose. The scale bar in (A) applies
to all three frames.
We conjecture that few other materials can be prepared in
one step while still achieving the complexity observed here.
However, it may be possible to generalize the mechanism to
include other chemical systems. This preparation proceeds via
Figure 9. SEM images of microcapsules produced from (A) phenyltri-
chlorosilane and (B) silicon tetrachloride. The scale bar in (A) applies to
both frames.
9
9446 J. AM. CHEM. SOC. VOL. 128, NO. 29, 2006

Self-Assembly of Core−Shell Organosilicon Microcapsules
A R T I C L E S
a solvolysis followed by simultaneous and competing condensa-
tion and crystallization of the solvolysis products. The relative
rates of these three processes for a different system could be
tuned to have similar ratios, perhaps yielding analogous,
hierarchical structures.
Keck SEM administered by the NSF-MRSEC at Cornell,
Malcolm Thomas, facility manager.
Supporting Information Available: Materials and methods,
complete tabulated NMR data and spectra, FTIR spectrum, DSC
trace, XRD pattern, and photographs of the microfluidic device.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Acknowledgment. We thank NSF Sensors (CTS-0329899),
ARO MAP-MURI, Cornell Center for Materials Research, the
Beckman Foundation, NYSTAR, 3M, Dreyfus, Ivan Keresztes
for assistance with the HMBC and DOSY experiments, and the
JA0612403
9
J. AM. CHEM. SOC. VOL. 128, NO. 29, 2006 9447