University of California
Santa Barbara
Polyelectrolytes in the Synthesis of
Bio-Inspired Composite Materials
A dissertation submitted in partial satisfaction of the
requirements for the degree of
Doctor of Philosophy
in
Chemistry
by
Brandon J. McKenna
Committee in charge:
Professor Galen D. Stucky, Chair
Professor Steven K. Buratto
Professor Alison Butler
Professor J. Herbert Waite
September 2007
The dissertation of Brandon J. McKenna is approved.
____________________________________________ Steven K. Buratto
____________________________________________ Alison Butler
____________________________________________ J. Herbert Waite
____________________________________________ Galen D. Stucky, Committee Chair
September 2007
ACKNOWLEDGEMENTS
I would like to thank my advisor Professor Galen D. Stucky for the
opportunities he has given me while working in his laboratory. The
scientific atmosphere supported by Galen promotes curiosity driven
research through intellectual freedom and has fostered my
independence and personal development. He has been an exemplary
role model of integrity, and his sheer joy in the process of discovering
and learning new science, across a wide range of subfields, has inspired
us all and kept us grounded in our pursuits.
I would also like to Professor J. Herbert Waite, who has effectively
served as my unofficial co-advisor. His expansive knowledge,
principled philosophies, precise and artistic prose, and passion for
discovery have guided me to enjoy the purity in science in an age of
increasing hype. Thanks also to my entire advising committee,
including Prof. Alison Butler and Prof. Steve K. Buratto, who provided
invaluable direction at a critical time in my studies.
The entire Stucky lab, in each of its incarnations that I have
experienced, has been critical to my growth as a Ph.D. student. Each
member has provided individual insight, and the list of support is long.
At large, the group has provided a positive and supportive atmosphere
v
of dedication, hard work, and of course diversified learning. Several
members have particularly contributed to my experience. I would like
to thank Prof. Henrik Birkedal for laying the groundwork for turning a
physicist into a materials chemist, imparting a basis for conducting
research and writing scientifically, and earnestly upholding the wonder
of discovery. I thank Dr. Peter Stoimenov for guidance, meaningful
conversations, and his appreciation for beauty, either in science or his
photographs; he is truly the best. I thank Dr. Muhammet Toprak for
insightful and extended discussions, for his friendship, and for our
fruitful collaborations from which I learned important complementary
research practices. Thanks to Dr. Todd A. Ostomel for leading the way
and rocking in the lab late into the nights; philosophical discussions of
resolving these practices with his initials were priceless. And thanks to
my “Taiwanese research brother,” Frank (Chia-Kuang) Tsung. During
our parallel pursuits of “real” science, I have appreciated his friendship
and look forward to continuing camaraderie in Berkeley.
I would also like to thank Dr. Celia Wrathall, who for all her
humility has no power to delete her acknowledgement this time. Celia
has provided critical support for my time here and beyond.
Many many thanks to Dr. Burçin Temel for her motivation,
direction, and support. Her companionship has made the last few years
vi
among my best. Thanks also to all the friends who have supplied
needed comic relief throughout my years here.
Finally, I would like to give thanks to my parents for their
emotional support and life guidance. Thanks to my mother for
fostering a childhood of wonder and for her patience with my
inquisitive nature—whether it regarded trains, windmills, or pockets
full of rocks! Thanks to my father for instilling the edicts of hard work,
the beauty in simplicity, and adaptability.
This work was partially supported by the MRSEC program of
the National Science Foundation under award No. DMR00-80034, and
made use of MRL Central Facilities granted by NSF under Award No.
DMR05-20415. This work was also supported in part by the U.S.
Army Research Laboratory and the U.S. Army Research Office under
contract number DAAD19-03-D-0004. This work is also supported in
part by the NASA University Research, Engineering and Technology
Institute on Bio Inspired Materials (BIMat) under award No. NCC-1-
02037. This work was supported in part by the Public Health Service
from NIH grant R01 DE 014572, NSF 0233728.
vii
VITA OF BRANDON J. MCKENNA
September 2007
EDUCATION
• Doctor of Philosophy, Chemistry University of California Santa Barbara, CA, September 2007 (anticipated)
Thesis title: “Polyelectrolytes in the Synthesis of Bio-Inspired Composite Materials”
• Artium Baccalaureus, Chemistry & Physics Harvard University, Cambridge, MA June 2002 AWARDS & SCHOLARSHIPS
• Dow Materials Use Prize 8th Annual New Venture Competition Technology Management Program University of California, Santa Barbara, 2007
• Certificate in Technology Management
Graduate Program in Management Practice University of California, Santa Barbara, 2007
• Chemistry and Biochemistry Department Fellowship University of California, Santa Barbara, 2002-2003, 2003-2004
• MRL Distinguished Graduate Fellowship
University of California, Santa Barbara, 2002-2003 • Advanced Standing Award
Harvard University, 1997-2001
RESEARCH EXPERIENCE
University of California, Santa Barbara Graduate Researcher, Chemistry and Biochemistry Department, 2002-present with Prof. Galen Stucky
viii
• Currently studying the ability of polyelectrolytes and
nanoparticles to assemble ordered inorganic structures, in two distinct projects:
1) Developing a device for targeted and magnetic drug delivery.
2) Biomimetic mineralization with acidic macromolecules.
Harvard University, Cambridge, MA Undergraduate Researcher, Physics Department and DEAS, Jun. 2000-Apr. 2002 with Prof. Charles Marcus
• Studied the lithographical and electrochemical fabrication of nano-electrodes, to produce quantized conductance, e.g. in single atoms.
Harvard University, Cambridge, MA Undergraduate Researcher, Harvard-Smithsonian Center for Astrophysics, Apr.-Jun. 2000 with Dr. Jim Phillips
• Arranged and tested a picometer-sensitive laser optics system, for testing Einstein’s Equivalence Principle.
University of New Hampshire, Durham, NH Undergraduate Researcher, Constraint Computation Center, Jun.-Aug. 1999 with Prof. Eugene Freuder
• Modeled and programmed solutions to constraint satisfaction problems.
PUBLICATIONS & PRESENTATIONS
1. McKenna, B. J., Waite, J. H., Stucky, G. D. “Complex Coacervates as Intermediates in Non-classical Crystallization,” in preparation
2. McKenna, B. J., Waite, J. H., Stucky, G. D. “Biomimetic Control
of Calcite Morphology with Homopolyanions,” submitted
ix
3. Toprak, M. S., McKenna, B. J., Waite, J. H., Stucky, G. D. “Control of Size and Permeability of Nanocomposite Microspheres,” Chem. Mat. 2007, 19, 4263-4269
4. Huang, X., Bronstein, L. M., Retrum, J., Dufort, C., Stein, B.,
Stucky, G., McKenna, B. J., Dragnea, B. “Self-Assembled Virus-like Particles with Magnetic Cores,” Nano Lett. 2007, 7, 2407-2416
5. McKenna, B. J., Waite, J. H., Stucky, G. D. “Morphological
Control of CaCO3 with Anionic Homopolymers,” poster presented at 2006 Symposium on Recent Advances in Nanoscale Materials Research, and 2007 Materials Research Outreach Program, UCSB.
6. Toprak, M., McKenna, B. J., Waite, J. H., Stucky, G. D.
“Tailoring Magnetic Microspheres with Controlled Porosity,” MRS Symp. Proc. 2007, 969, W03-11
7. Toprak, M., McKenna, B. J., Mikhaylova, M., Waite, J. H.,
Stucky, G. D. “Spontaneous Assembly of Magnetic Microspheres,” Adv. Mat. 2007, 19, 1362-1368
8. McKenna, B. J., Waite, J. H., Stucky, G. D. “Biomimetic
Materials,” August 2006 talk at US Gypsum Corporation, invited by Creative Realities, Inc. as Techmax Thought Leader
9. Toprak, M., McKenna, B. J., Stoimenov, P., Waite, J. H., Stucky,
G. D. “Controlled Assembly of Magnetic Microspheres,” poster presented at Spring 2006 MRS Meeting
10. McKenna, B. J., Waite, J. H., Stucky, G. D. “Complex
Coacervate Mineralization,” oral talk at Fall 2005 MRS Meeting 11. McKenna, B. J., Birkedal, H., Bartl, M. H., Deming, T. J.,
Stucky, G. D. “Micrometer-sized Spherical Assemblies of Polypeptides and Small Molecules by Acid-Base Chemistry,” Angew. Chem. Int. Ed. 2004, 43, 5652-5622
12. McKenna, B. J., Birkedal,H., Bartl, M. H., Deming, T. J.,
Stucky,G. D., “Self-Assembling Microspheres from Charged Functional Polyelectrolytes and Multivalent Ions,” poster presentation at UC Systemwide Bioengineering Symposium 2004
13. McKenna, B. J., Birkedal, H., Bartl, M. H., Deming, T. J.,
Stucky,G. D., “Self-Assembling Microspheres from Charged
x
Functional Polyelectrolytes and Small-Molecule Counterions,” MRS Symp. Proc. 2004, 823, 23; also presented poster at Spring Meeting.
PROFESSIONAL AFFILIATIONS
• Materials Research Society student membership • American Physical Society student membership
TEACHING EXPERIENCE
University of California, Santa Barbara Chemistry and Biochemistry Department
• Teaching Lab. Assistant, Fall 2003, Winter 2004, and Spring 2004 Inorganic Synthesis
Taught upper level laboratory techniques, including analysis, organometallic synthesis, and various advanced characterization techniques (HPLC, transient absorption, bulk electrolysis, etc.)
• Teaching Lab. Assistant, Fall 2002, Winter 2003, and Spring 2003 General chemistry
Taught sections, guided student experiments, prepared quizzes, graded laboratory reports.
SKILLS
• Materials synthesis (sol-gel, nanoparticles, organic); X-ray diffraction (XRD), surface area and pore size analysis (BET), mass spectrometry (MS), X-ray photoelectron spectroscopy (XPS), atomic force microscopy (AFM), UV-vis and IR spectroscopy, Raman spectroscopy, fluorimetry, thermogravimetric analysis and differential scanning calorimetry (TGA/DSC), scanning electron microscopy (SEM), transmission electron microscopy (TEM), nuclear magnetic resonance (NMR), zeta potential measurements, dynamic light scattering (DLS), light optical microscopy (LOM), confocal laser scanning microscopy (CLSM) including FRAP, circular dichroism (CD).
xi
• Special Courses: Transmission Electron Microscopy– Principles
and Practice
• Programming Languages: C++, JAVA, LISP, Mathematica
xii
ABSTRACT
Polyelectrolytes in the Synthesis of Bio-Inspired Composite Materials
by
Brandon J. McKenna
This original research dissertation contains studies on complex
coacervation, methods of modifying coacervates to create new
materials as particularly applied to targeted drug delivery, and the use
of coacervating polyanions for the assembly of intricate structures of
calcium carbonate.
Complex coacervation is a liquid-liquid phase separation that
typically produces microspherical droplets from the combination of a
variety of oppositely charged ions, including polymers and
nanoparticles. The chemical space of coacervating components was
found dependent on the number of charged groups and pH.
Coacervates were shown to present chemically active surfaces that
could be solidified by various methods, some of which also induced
hollow interiors. The resulting assemblies were considered for targeted
drug delivery by using superparamagnetic magnetite nanoparticles as
xiii
assembling components. Control over microsphere sizes was obtained
from variation of several parameters, and porosities were examined as a
function of cross-linking extent to determine encapsulation capabilities.
Coacervates were further found to direct mineral growth, first in
the form of shells, and then in the form of complicated structures that
require substrate interaction via a solution-amorphous-crystalline
mechanism. A ternary phase diagram approach revealed a great
diversity of morphologies that could be modulated by the action of
coacervating polyanions. Detailed analysis of one particular stacked
lamellar structure suggested an assembly mechanism that may have
relevance for biomineralization of nacre.
xiv
TABLE OF CONTENTS
Approval page………………………………………………………….ii
Dedication……………………………………………………………..iv
Vita…………………………………………………………………...viii
Abstract………………………………………………………………xiii
I. Complex Coacervation for the Synthesis of New Materials……...….1
II. Assembly and Silica Coating of Polypeptide/Small Molecule
Coacervates…………………………………………………………...35
III. Superparamagnetic Nanoparticle Coacervates for Targed Drug
Delivery…………………………………………………………….....61
IV. Calcium Carbonate Mineralization via Complex Coacervation...110
V. Morphological Ternary Diagram Studies of Non-classical Calcium
Carbonate Mineralization with Homopolyanions…...………………134
VI. Towards a Model for Nacre Formation…..……………………...183
xv
Abstract:
It is presented in this chapter a generalized scheme for the
assembly of microspherical devices from complex coacervates. First, a
review of complex coacervation and related fields is presented, and
several important properties displayed by most coacervates are
highlighted. Polyelectrolyte association is generally enhanced by any
parameter that increases the degree of attraction between oppositely
charge moieties of the coacervating components. Then I present
several methods for the stabilization of colloidal microspheres,
including the formation of various inorganic shells, the use of organic
crosslinking, and physical phase changes. It is anticipated that the
extended variability of assembling components and stabilization
methods will, in turn, extend the variety of potential device
applications.
2
Complex coacervation, in its broadest definition, is a liquid
phase separation that occurs between multiple solution species that
electrostatically but dynamically attract one another and remain highly
solvated, particularly around their binding moieties.1 Most commonly,
there are two components, at least one of which is a charged polymer
and the solvent is water; often, the other component is an oppositely
charged polymer. The most commonly cited example is the mixing of
gelatin with gum Arabic. Above gelation temperatures, and at pH
values (4-5) that render gelatin with a net positive charge and gum
Arabic with a net negative charge, the two polymers mutually attract
and manifest in the appearance of micron-sized colloidal droplets that
make the solution turbid, i.e. with a milky white appearance. Over time
or with force, the droplets coalesce and eventually form a single,
continuous, lower liquid layer that is dense with the biopolymers. This
layer is called the “coacervate phase”, and the other, usually larger
supernatant portion is called the “equilibrium phase”, as it is in
dynamic equilibrium with the coacervate and containing a low
concentration of the biopolymers. The droplets, prior to coalescing into
the continuous coacervate phase, are often referred to as “coacervates”
(Fig. 1.1A). The phenomenon is perhaps better understood in contrast
to flocculation, the solid phase counterpart that usually takes the shape
of fractal-like aggregates due to the rapid, diffusion-limited growth
3
mode.2-5 The main difference, of course, is that flocculations, or
“flocs”, contain less labile (more solid-like) bonds; i.e., they are less
dynamic on the timescale of diffusion (Fig. 1.1B).
Complex coacervation has found use primarily in
microencapsulation, where it is used for taste- masking,6-9 drug
formulations,10-12 and carbonless copy paper.13 The phenomenon is
also receiving newfound attention in the field of biomaterials, having
been found more prevalent in nature than previously suspected. For
instance, the DNA/histone interaction may be a coacervate-like
interaction.14 Waite et al. have suggested that mussel fiber anchoring15
and sandcastle worm gluing16, 17 employ coacervation processes in
combination with high density phase inversion into foams and
subsequent covalent and noncovalent crosslinking. The exquisite
architecture of marine diatoms have been postulated to occur via liquid
Figure 1.1. A) Light optical micrograph of a coacervate solution containing poly-L-lysine and citric acid. B) Micrograph of flocculation in a solution of poly-L-arginine and citric acid.
(A) (B)
100 μm
4
phase separation processes,18-20 which suitable describes coacervation
as it is known to occur for mimetic polymers.21
“Coacervation” has also been used to describe other, similar
interactions aside from association of polyelectrolytes. For instance
charged polymers and smaller anions can form coacervates, as with
polyacrylic acid and Ca2+. There are also examples of small
counterions forming coacervates, such as Cd(NO )3 2 with sodium
succinate and SrCl2 with ammonium molybdate. Such combinations
tend to dehydrate and crystallize, losing their fluidity. Finally, “simple
coacervation” contrasts with complex coacervation in that it occurs
with a single species. AOT (sodium
bis(ethylhexyl)octylsulfosuccinate) phase separation in solution is a
common example of simple coacervation. Although it has been shown
to depend on the concentration of Na+,22 sodium’s role as a noncovalent
crosslinker is dubious.
These descriptions accord with generally accepted definitions of
“coacervation,” and especially with those first espoused by Bungenburg
de Jong, who conducted some of the first in-depth studies that revealed
some of the trends elaborated below.1 Nonetheless, “coacervation” has
also been occasionally relegated to more specific instances; some
biological texts describe it as a phase separation between a
polysaccharide and protein—only a subset of components that may
5
form a coacervate. The confusion probably originates from the most
widespread example of coacervation (gelatin/gum arabic), popularized
further in the field of biology by the research activities of Oparin.23
Oparin’s experiments incorporated starch-polymerizing enzymes into
these coacervates, which were then observed to grow and divide,
leading him to conclude that complex coacervation may have supplied
the most primitive “protocellular” compartments at the origin of life.24
The components were made more familiar still by patents for
carbonless copy paper, for which such coacervates are used to
encapsulate a pH-sensitive dye.13
Aside from occasionally receiving a limited definition,
sometimes altogether different terms are used to describe coacervation.
The most common example is “polyelectrolyte complex,” abbreviated
as PEC or sometimes IPEC for “interpolyelectrolyte complex.” This
term is less specific in describing the resultant phase, however, as it
may also describe flocculation25 or even soluble dimers.26 Some
reports simply refer to “polyelectrolyte association,” either because the
research was not focused on the nature of interaction, because such
nature was not confirmed, or because the term has not completely
penetrated various realms of science. Recently, phenomena related to
coacervation have been given other monikers. Block Ionomer
Complexes (BICs)27-34 occur between polyelecrolytes and oppositely
6
charged micelles, which have also previously been described
straightforwardly.35-37 Although coulombic interactions contribute to
their assembly, the importance of surfactant amphiphilicity legitimizes
this term as a subset of coacervation. Recently, “Polymer Induced
Liquid Precipitates” (PILPs) have been observed as precursors in
mineralizing solutions; it remains to be shown that their emergence is
caused in the same way as coacervation.38 However, the term PILP has
also been applied to describe ordinary coacervate-like polyelectrolyte
association.39 Finally, as will be described, our own research has
referred to coacervates as “acid-base microsphere assemblies.”20, 40
The term “coacervation” has been applied to systems seemingly
dissimilar to those of the classical definition. For instance, phase
separation has been observed between partially condensed silicic acid,
such as polysilicic acid, and various small polar molecules,41
polyethers, and polyvinyl alcohol.42 These so-called coacervates differ
in two significant ways: firstly, the components are not charged but
rather interact purely via hydrogen bonding, and secondly, the ensuing
aggregation and phase separation is due to hydrophobicity of carbon
segments rather than electrostatics. In yet further departures from the
classical definition, “coacervation” has been used to describe liquid
precipitates of single macromolecules in two-component solvents, and
phase separations in organic solvents, such as with ethylcellulose in
7
hexanes.43 In these cases, “coacervation” is taken very generally to
mean “liquid phase separation of partially soluble components.”
Coacervate Properties
All coacervates demonstrate similar dependencies on solution
variables such as pH. It can generally be stated that polyelectrolyte
association is enhanced by any parameter that increases the degree of
attraction between oppositely charge moieties of the coacervating
components. The following trends illustrate this point.
1) Concentration dependence roughly follows a solubility
product. Being soluble, coacervates dissolve upon dilution,
and the component concentrations limiting coacervation
roughly fall along a curve. These curves also describe
concentrations in the equilibrium phase. Approximate
solubility curves determined by dilution and optical
microscopy appear in figure 1.2.
8
0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045
0.000
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0.009
PLD
[CO
O- ] (
mM
)
[Ca2+] (mM)
11kDa, pH7 11kDa, pH 10 33 kDa, pH 7 33 kDa, pH 10
Coacervate Visibility, by Dilution
2+Figure 1.2. Solubility curves for PLD/Ca
coacervates, determined by incremental dilutions and visibility under an optical microscope.
2) Salt disrupts coulombic attraction and inhibits coacervation.
The dynamic bonds are disrupted with significant effect on
the effective solubility products. This effect appears to be
due to a combination of ionic strength and a counterion
substitution effect. PLK/citrate coacervates are eradicated
by ~10mM chloride salts of Na, K, and
tetramethylammonium, for ~0.5 wt% plK and citrate.
Higher valency spectator salts are more effective than their
monovalent counterparts at inhibiting coacervation,
requiring only 6 mM CaCl or less than 1mM MgCl . 2 2
3) There is an optimal ratio of components. At the point of
charge balance, coacervates generally repel less and grow
9
larger. The amount of material in the equilibrium phase is
minimized. Due to the aforementioned salt and substitution
effects, solubility curves appear to ‘bend back’ and reflect
working component ratios. Outside of these ratios, both
ionic strength and counterion substitution overcome the
drive for coacervation.
4) There is an optimal pH range. The charges of amines and
carboxylates, for instance, are controlled by their pKa
values, so that a pH range usually exists that maximizes the
overall number of complementary charges.
5) The optimal component ratio changes predictably with pH.
Higher pH values cause deprotonation and create higher net
negative charge, and hence require more polycation
component to achieve maximal coacervation.
6) Less polar solvents promote polyelectrolyte association by
dehydration and by decreasing the solution dielectric
constant. Adding small amounts of alcohol to PLK/citrate
coacervates causes significant swelling of coacervate
spheres, and eventual aggregation into shapeless gels [Fig.
1.3].
10
100 μm
Figure 1.3. Micrograph depicting the effect of ethanol added to a solution of PLK/citrate coacervates. The coacervates initially swell and slowly aggregate into a solid mass.
7) The degree of coacervation correlates with coacervate phase
volume, charge neutrality, bulk viscosity, and colloidal
turbidity. Maximal polymer-polymer noncovalent
crosslinking, as determined by component concentrations,
pH, etc., in turn maximizes coacervate phase volume, which
may be quantified by centrifugation. This usually also
results in maximal solution turbidity if all other variables are
constant (mixing order, etc.). Charge neutrality reflects
balanced Coulombic interactions and maximal crosslinking,
and can be determined by zeta potential measurements.
Maximal attraction also reduces the number of free polymer
11
in equilibrium mixtures, reducing viscosity of the bulk
colloidal solution.
8) Similarly sized components coacervate more effectively.
The degree of coacervation is lessened when component
molecular weights mismatch. This effect has been observed
with more complicated block copolymer systems, in which
phase separation is completely inhibited between mixtures
containing charge block segments of different sizes.44
The previous section described established coacervation properties.
The following are general observations I have noted from engineering
several coacervate systems.
1) The effect of temperature is variable. Increased
temperatures can dissolve some coacervates, as is the case
with PLK/cit, at 70 ºC. PLR/cit undergoes a solid-to-liquid
transition at 40 ºC; coacervate microspheres can be seen to
‘bud off’ of flocs with temperature increase. In contrast,
PLE/Ca2+ solutions are transparent at room temperature, but
become turbid with heating. The PLE/Ca2+ structures remain
stable and solidified at room temperature, often in the form
of dumbbells or other multiplets (Fig. 1.4).
2)
12
2+Figure 1.4. PLE/Ca multiplet assemblies following heating and cooling steps.
2+The phase separation is attributed to desolvation of Ca with
concomitant change in tertiary structure of PLE from random
coil to beta sheet [Fig. 1.5].
190 200 210 220 230 240 250-0.35
-0.30
-0.25
-0.20
-0.15
-0.10
-0.05
0.00
0.05
0.10
Ellip
ticity
Wavelength (nm)
21C 30C 40C 50C 60C 70C 75C
Circular Dichroism, PLD/Ca2+
Figure 1.5. Circular Dichroism spectra of a PLD/Ca2+ solution at various temperatures.
13
The desolvation mechanism is supported by observations of
PLD/Mg2+ systems, which are soluble at room temperature but
can be observed to form coacervates at 80-90 ºC [Fig. 1.6].
3) In polymer/small-counterion mixtures, the larger polymer
dictates the net charge (electrophoretic mobility) and
coacervation is optimized with the small counterion in
excess. In PLD/Ca2+ coacervates, for example, the CaCl2
component may be required in excess of aspartate
monomers before coacervation is appreciable. The
electrophoretic mobility of these coacervates will always
correspond to a net negative charge, reflecting the dynamic
association of calciums at the surface.
4) The order of component addition affects size distribution.
For a given set of final concentrations, coacervates are
(A) (B)
100 μm 100 μm
Figure 1.6. LOMs of PLD/Mg2+ solutions at (A) 80 ºC and (B) 90 ºC.
14
largest when two concentrated solutions are mixed and then
diluted. If one solution is already dilute, coacervates are
larger when the smaller component is added to the larger
(diluted) one. Therefore, they are smallest when a small
volume of the larger polymer is added to a diluted solution
of the smaller component.
A variety of components can be used to effect coacervation,
including: natural8, 45-50 and synthetic,51 organic polyelectrolytes,
inorganic polyelectrolytes,24 block copolymers,27, 28, 52 dendrimers,53
small multivalent organic ions,20, 40 inorganic ions,54 and nanoparticles
and quantum dots that are appropriately capped.55-57 However, there
is currently no reliable method for predicting when coacervation will
occur rather than flocculation or dissolution. Towards this effort,
presented in the following chapter, we have found predictive trends
with polymer/small counterion combinations of carboxylate/amine-
containing compounds.20 The pH requirements of these systems were
demonstrated to be highly dependent on component pKa. Furthermore,
polyamine coacervation requires at least tri-carboxylated counterions,
and polycarboxylate coacervation usually requires counterions with at
least 5 amine groups. However, exceptions exist, and in general it is
necessary to test specific systems individually.
15
Results of work done in our laboratory suggested a general
mechanism of coacervating components as spherical templates,
followed by a process of solidification/stabilization. The coacervates
alone remain liquid out of solution, and in solution eventually wet the
container as spread-out droplets (Fig. 1.7). However, the spherical
shape could be preserved by condensation of inorganic silica, or with
temperature changes. In order to extend the variety of potential future
applications for various environments, further work was conducted to
engineer other kinds of stabilization processes, including inorganic
shell formation, organic covalent crosslinking, and physical
crosslinking.
(A) (B)
100 μm
Figure 1.7. Examples of normal coacervate sedimentation and wetting on glass slides, if not stabilized.
Materials and Methods.
SiO2 Shells. Standard silica treatments have been described
elsewhere. For mesoporous silica, a polymer stock solution 1g of P123
(BASF, Pluronic) was dissolved in 37.5 mL pH 3 DI H2O. Sol-gel
batches were typically prepared by mixing 10 μL tetramethoxysiloxane
16
(TMOS) in 258 μL of the P123 solution, and hydrolyzing for 20 min.
Aqueous coacervate solutions were prepared by mixing 10 μL
polyallylamine hydrochloride (PAH) (Sigma-Aldrich) at a monomeric
concentration of 25 mM with 10 μL of a 10mM solution of trisodium
citrate solution, and diluting to 100 μL at pH7. A 10 μL aliquot of the
hydrolyzed TMOS/P123 solution was mixed with the full coacervate
solution and allowed to react for 3h.
Au shells: In a typical synthesis, poly-L-lysine (70kDa) and
trisodium citrate solutions, each of ~0.5 wt.%, were mixed in a 1:9 ratio
to form coacervate templates. A gold salt solution was prepared
separately by dissolving 5 0mg HAuCl into 3 mL H4 2O. Ten
microliters of this gold solution was pipetted into 100 μL of
coacervates. 0.1 M sodium citrate (20 μL) was added after 10 min. to
continue gold reduction on coacervate surfaces. Additional 10 μl gold
acid was added, followed by addion of another 20 μl portion of citrate.
The solutions were centrifuged after 1h, the supernatant was removed,
and the gold spheres were redispersed in DI water.
Mineral shells: Calcium carbonate shells were typically
prepared by adding 0.5 mL of 100 mM by monomer PAA (15kDa) to
5.25 mL CaCl , and then diluting with 40mL of DI H2 2O and NaOH
solution, to a pH of 8.5. A fresh solution of 75mM Na CO2 3 (75mM)
was prepared, and three portions of 750 μL of this were successively
17
mixed with the coacervate solution in 5 min. intervals. The colloids
were stable and allowed to sediment for 24-48 h, after which the
solutions were decanted and the spheres were redispersed and stabilized
in pH 10 NaOH to 50mL.
Calcium phosphate shells were prepared from solutions of 5
mg/mL poly-L-aspartic acid (PLD) (33kDa) , 180mM CaCl2, and
10mM KH PO2 4. Coacervates were formed from 20 μL PLD and 7.5 μL
CaCl2, to which 10 μL phosphate solution was added. Products were
collected and washed within 1 h.
Covalent cross-linking: EDC (1-ethyl-3-(3-
dimethylaminopropyl) carbodiimide hydrochloride) crosslinking was
performed on coacervates of poly-L-glutamic acid (PLE) and
pentalysine. EDC solutions (50 mg/mL) were prepared fresh, and
added in aliquots of 20 μL to 100 μL coacervate solutions. Crosslinked
spheres were observed without further purification.
Glutaraldehyde crosslinking was performed on poly-L-lysine
(PLK) and citrate coacervates, as described elsewhere (See Chapter 3).
Typically, 20 μL of 2 mg/mL PLK solution was mixed vigorously with
120 μL TSC (1.6 mg/ml) for 15 seconds, using a vortex-mixer. To this,
120 μL of 2.5 wt% glutaraldehyde was added, and excess aldehydes
were quenched by the addition of glycine after reacting for between 1
and 5000 min. Samples were purified by successive centrifugations.
18
Flocculation stabilization: Cationic coacervates were prepared
from 25 μL FITC-labeled PLK at a monomeric concentration of 2.5
mM and 50 μL TSC at a concentration of 5mM. To this, 10 μL PLE
(0.5 wt.%) was added and quickly mixed. Anionic coacervates were
prepared from PAA and CaCl2 solutions to final concentrations of
10mM (monomer) and 10.5 mM, respectively, to a total volume of 100
uL. To this, 10 μL diaminoethylene (50 mM) was added.
Physical stabilization: Mixtures either of poly-L-Arginine
(PLR) or poly-L-Histidine (PLH) were made with TSC, such that both
components comprised ~0.5wt% of solution. PLR/TSC solutions were
heated to 60ºC, either in an oven, water bath, or microscope heat stage.
PLH solutions were adjusted to pH ~5 with diluted HCl, and pH was
increased with aqueous NaOH. PLD/Mg coacervates were made from
final concentrations of 13mM PLD (33kDa) and 125 mM MgCl . 2
Results
SiO2: As described in the following chapter, prehydrolyzed
TMOS induced a silica layer on the surface of primary amine-
containing polymers.20, 40 Silica nanoparticles could form shells around
the coacervates and induce hollow cores in the coacervate interior; the
hollow core effect was more evident with large coacervates (3-5 um).20,
40, 55, 56, 58 Polymer surfactants were found to introduce wormhole
19
mesopores using a surfactant/sol-gel strategy (Fig. 1.8); more work is
required to determine if pore ordering is possible with these systems.
This is an attractive route for manufacturing aqueous reaction vesicles
that incorporate enzymes or other components too instable for porous
vesicles that require thermal treatment.
100 nm
Figure 1.8. (A) LOM of PAH/citrate coacervates stabilized by shells of mesoporous silica at pH 7. (B) TEM image of wormhole structured shells.
Au and other metals: using
the reduction properties of citrate
in solution, Au salt precursor could
be deposited onto coacervate
surfaces (Fig. 1.9). The scheme
works with Ag as well, but is not
generalized to work with Cu, for
instance, which deposits in
metallic rings on microsphere
Figure 1.9. SEM image of PLK/citrate coacervates coated with Au shell and subject to focused ion beam. Thanks to Dr. Peter Stoimenov for permission to use this micrograph
20
surfaces.
Salts of divalent metals: Using CaCO and Ca/PO3 4 as model
systems, coacervates of Ca2+ with PAA or PLD were used as
microspherical templates and Ca2+ sources for the deposition of low Ksp
salts. The phosphate spheres (Fig. 1.10) formed more rapidly and could
produce well defined spheres with little side product.
(A) (B)
5 μm 25 μm
Figure 1.10. (A) LOM and (B) SEM of PLD/Ca2+ coacervates stabilized by shells of phosphate mineral.
EDC cross-linking: Taking advantage of the innate proximity
of carboxylate and primary amine groups, EDC was used to form amide
bonds in certain coacervates (Fig. 1.11). The primary obstacle for this
procedure is that EDC works optimally at pH values in the range of 4.5-
5, whereas the coacervates are most stable at higher pH. HOBt can be
used to improve the amidation yield.
21
50 μm
Figure 1.11. LOM of PLE/K5 coacervates stabilized by EDC cross-linking.
Glutaraldehyde cross-linking: Glutaraldehyde is commonly
used in the fixation of biological samples by the cross-linking of amine
groups into Schiff bases. The advantages of this technique over EDC
cross-linking are the range of permissible pH values and the stability of
glutaraldehyde in solution (EDC decomposes slowly in water).
However, glutaraldehyde cross-links are unstable for extended periods
and free glutaraldehyde is unsafe for certain biological applications.
Furthermore, while the Schiff base bonds can be made irreversible by
NaBH reduction, the resulting bonds are not easily degradable. 4
Flocculating counter-polyelectrolytes: Some coacervates could
be solidified by the addition of oppositely-charged polyelectrolyte. In
order to induce stable microspheres, the polymer must have a ‘solid-
like’ or flocculation-inducing interaction with one of the components of
the original coacervates, as opposed a ‘liquid-like,’ coacervation-
22
inducing interaction. This distinction is important for studies of LbL
assembly, for which polyelectrolyte interactions should be essentially
irreversible. For instance, at room temperature and neutral pH, PLD
forms a coacervate with PLK, whereas PLE forms flocs with PLK.
Therefore, coacervates of PLK/citrate could be stabilized by the
addition of PLE. These assemblies are also able to induce hollow cores
in the coacervates, as was done with silica NPs (Fig. 1.12).
Figure 1.12. Coacervates of FITC-PLK/citrate stabilized by PLE. (A) LOM, demonstrating structural integrity, and (B) demonstrating partial component redistribution.
(A) (B)
50 μm 5 μm
The flocculation interaction can also be used to stabilize
coacervates by using smaller counterions. For instance, oxalate added
to PLK/citrate spheres can induce solidified plK/oxalate shells, and
diaminoethylene can cap PAA/Ca2+ coacervates.
pH change: Some coacervates undergo a liquid-solid phase
transition upon changes in pH. For instance, PLH/citrate coacervates
are liquid between pH [4-5.5], but above pH 6, the imidazole moiety is
23
neutralized and the spheres solidify and aggregate. Similarly, chitosan
is insoluble above pH 6, and will solidify at higher pH values in
microspherical combinations with gum Arabic.
Temperature change: As mentioned above, coacervate phase
transitions with temperature are highly dependent on the exact system
being used. Liquefied PLR/citrate mixtures reversibly solidify at
room temperature (See following chapter). Such behavior is also
known for gelatin/gum arabic coacervates, but in that instance the
solidification is due to the tertiary structure of the gelatin component
rather than changes in the stability of carboxyl/amine electrostatics.
This procedure constitutes a facile microencapsulation method for
applications where the physical interactions are stable.
From these various interactions and coacervate sources, a
generalized scheme is presented in Table 1.1. Several different types of
polyanion and polycation sources can be used to form coacervates,
although it is not currently possible to predict the nature of interaction
between any given two: soluble, liquid, or solid. And liquid
coacervates can be stabilized by various mechanisms, which depend in
part on the nature of the selected polyelectrolytes.
24
Polyanions Polycations StabilizationPolypeptides (polyaspartate) & polysaccharides (gum arabic)
Polypeptides (polylysine, gelatin)
Inorganics (silica, gold, calcium carbonate)
Synthetic polymers (polyacrylate)
Synthetic polymers (polyallylamine)
Silica nanoparticles (also creates hollow core)
Small organics (citrate, EDTA) Small organics (amino alkanes)
Solidification (cooling, gelation)
Inorganics (dichromate) Metal ions (Ca2+, Fe2+, Zn
Counter-polyelectrolyte (polyglutamate) 2+)
Citrate-capped Nanoparticles (Au, CdSe, Fe
Amine-terminated Dendrimers
Covalent crosslinking (glutaraldehyde, EDC) 3O4…)
Table 1.1. Generalized scheme outlining the variety of nano-sized coacervating components and various methods for their stabilization.
Conclusion
Complex coacervation is a liquid-liquid phase separation that
occurs, somewhat commonly, between oppositely charged species in
solution. For polymer-based coacervates, some trends exist to describe
the degree of polyelectrolyte association as it depends on pH,
component concentrations, and added salts. Complex coacervation is a
long-known phenomenon that nevertheless has remained obscure to
most sciences, has continued to be rediscovered in various instances
and with different names, and yet has been little-developed in the
advancement of materials science. There remains much progress to be
made to control coacervation, including an understanding of kinetics, a
model to predict the nature of phase separation, a method for reducing
their typical large size dispersion, a better understanding of
encapsulation capabilities of both hydrophobic and hydrophilic agents,
25
and an expanded chemical toolbox for tailoring coacervate microsphere
functionalities, such as porosity or environmental robustness. We have
sought to expand the types of chemistry and the range of nano-
components that can be integrated with the process, in order to produce
organized core-shell devices that may find use as: drug delivery
vehicles and artificial cells, industrial fillers, chemical microreactors,
absorbents of toxins for environmental remediation, or colloidal
sensors.
REFERENCES
1. Bungenberg de Jong, H.G., in Crystallisation- coacervation-
flocculation in colloid science, H.R. Ed. Kruyt, Editor. 1949,
Elsevier: Amsterdam. p. 232-258.
2. Biggs, S., et al., Aggregate structures formed via a bridging
flocculation mechanism. Chemical Engineering Journal, 2000.
80(1-3): p. 13-22.
3. Larsson, A., C. Walldal, and S. Wall, Flocculation of cationic
polymers and nanosized particles. Colloids and Surfaces a-
Physicochemical and Engineering Aspects, 1999. 159(1): p. 65-
76.
26
4. Thomas, D.N., S.J. Judd, and N. Fawcett, Flocculation
modelling: A review. Water Research, 1999. 33(7): p. 1579-
1592.
5. Berlin, A.A., I.M. Solomentseva, and V.N. Kislenko,
Suspension flocculation by polyelectrolytes: Experimental
verification of a developed mathematical model. Journal of
Colloid and Interface Science, 1997. 191(2): p. 273-276.
6. Gouin, S., Microencapsulation: industrial appraisal of existing
technologies and trends. Trends in Food Science & Technology,
2004. 15(7-8): p. 330-347.
7. Gibbs, B.F., et al., Encapsulation in the food industry: a review.
International Journal of Food Sciences and Nutrition, 1999.
50(3): p. 213-224.
8. Schmitt, C., et al., Structure and technofunctional properties of
protein-polysaccharide complexes: A review. Critical Reviews
in Food Science and Nutrition, 1998. 38(8): p. 689-753.
9. Shahidi, F. and X.Q. Han, Encapsulation of Food Ingredients.
Critical Reviews in Food Science and Nutrition, 1993. 33(6): p.
501-547.
10. Barry, B.W., Novel mechanisms and devices to enable
successful transdermal drug delivery. European Journal of
Pharmaceutical Sciences, 2001. 14(2): p. 101-114.
27
11. Lim, S.T., et al., Preparation and evaluation of the in vitro drug
release properties and mucoadhesion of novel microspheres of
hyaluronic acid and chitosan. Journal of Controlled Release,
2000. 66(2-3): p. 281-292.
12. Madan, P.L., Microencapsulation .1. Phase Separation or
Coacervation. Drug Development and Industrial Pharmacy,
1978. 4(1): p. 95-116.
13. Green, B.K., and Schleicher, Lowell, Oil-containing
Microscopic Capsules and Method of Making Them, U.S.P.
Office, Editor. 1957: U.S.A.
14. Takahagi, M. and K. Tatsumi, Aggregative organization
enhances the DNA end-joining process that is mediated by
DNA-dependent protein kinase. Febs Journal, 2006. 273(13): p.
3063-3075.
15. Waite, J.H., et al., Mussel adhesion: Finding the tricks worth
mimicking. Journal of Adhesion, 2005. 81(3-4): p. 297-317.
16. Stewart, R.J., et al., The tube cement of Phragmatopoma
californica: a solid foam. Journal of Experimental Biology,
2004. 207(26): p. 4727-4734.
17. Zhao, H., et al., Cement proteins of the tube-building polychaete
Phragmatopoma californica. Journal of Biological Chemistry,
2005. 280(52): p. 42938-42944.
28
18. Sumper, M., A phase separation model for the nanopatterning
of diatom biosilica. Science, 2002. 295(5564): p. 2430-2433.
19. Sumper, M. and E. Brunner, Learning from diatoms: Nature's
tools for the production of nanostructured silica. Advanced
Functional Materials, 2006. 16(1): p. 17-26.
20. McKenna, B.J., et al., Micrometer-sized spherical assemblies of
polypeptides and small molecules by acid-base chemistry.
Angewandte Chemie-International Edition, 2004. 43(42): p.
5652-5655.
21. Brunner, E., K. Lutz, and M. Sumper, Biomimetic synthesis of
silica nanospheres depends on the aggregation and phase
separation of polyamines in aqueous solution. Physical
Chemistry Chemical Physics, 2004. 6(4): p. 854-857.
22. Menger, F.M. and B.M. Sykes, Anatomy of a coacervate.
Langmuir, 1998. 14(15): p. 4131-4137.
23. Oparin, A.I., et al., Synthesis and Dissociation of Starch in
Coacervate Drops. Doklady Akademii Nauk Sssr, 1962. 143(4):
p. 980-&.
24. Oparin, A.I., et al., Coacervate Drops with Participation of
Polyphosphates and Model of Probiont Division. Doklady
Akademii Nauk Sssr, 1977. 232(2): p. 485-488.
29
25. Dautzenberg, H., et al., Stoichiometry and structure of
polyelectrolyte complex particles in diluted solutions. Berichte
Der Bunsen-Gesellschaft-Physical Chemistry Chemical Physics,
1996. 100(6): p. 1024-1032.
26. Li, Y.J., et al., Complex-Formation between Polyelectrolyte and
Oppositely Charged Mixed Micelles - Soluble Complexes Vs
Coacervation. Langmuir, 1995. 11(7): p. 2486-2492.
27. Lysenko, E.A., et al., Block ionomer complexes with polystyrene
core-forming block in selective solvents of various polarities. 2.
Solution behavior and self-assembly in nonpolar solvents.
Macromolecules, 2002. 35(16): p. 6344-6350.
28. Lysenko, E.A., et al., Block ionomer complexes from
polystyrene-block-polyacrylate anions and N-cetylpyridinium
cations. Macromolecules, 1998. 31(14): p. 4511-4515.
29. Bronich, T.K., et al., Effects of block length and structure of
surfactant on self-assembly and solution behavior of block
ionomer complexes. Langmuir, 2000. 16(2): p. 481-489.
30. Kabanov, A.V. and V.A. Kabanov, Interpolyelectrolyte and
block ionomer complexes for gene delivery: Physicochemical
aspects. Advanced Drug Delivery Reviews, 1998. 30(1-3): p.
49-60.
30
31. Lysenko, E.A., et al., Block ionomer complexes with polystyrene
core-forming block in selective solvents of various polarities. 1.
Solution behavior and self-assembly in aqueous media.
Macromolecules, 2002. 35(16): p. 6351-6361.
32. Oh, K.T., et al., Block ionomer complexes as prospective
nanocontainers for drug delivery. Journal of Controlled
Release, 2006. 115(1): p. 9-17.
33. Solomatin, S.V., et al., Environmentally responsive
nanoparticles from block ionomer complexes: Effects of pH and
ionic strength. Langmuir, 2003. 19(19): p. 8069-8076.
34. Solomatin, S.V., et al., Colloidal stability of aqueous
dispersions of block ionomer complexes: Effects of temperature
and salt. Langmuir, 2004. 20(6): p. 2066-2068.
35. Wang, Y.L., et al., Polyelectrolyte-micelle coacervation: Effects
of micelle surface charge density, polymer molecular weight,
and polymer/surfactant ratio. Macromolecules, 2000. 33(9): p.
3324-3331.
36. Wang, Y.L., et al., Effects of salt on polyelectrolyte-micelle
coacervation. Macromolecules, 1999. 32(21): p. 7128-7134.
37. Zhang, G.Z., et al., Interpolymer complexes comprising block
copolymers due to specific interactions. Materials Science &
31
Engineering C-Biomimetic and Supramolecular Systems, 1999.
10(1-2): p. 155-158.
38. Gower, L.B. and D.J. Odom, Deposition of calcium carbonate
films by a polymer-induced liquid-precursor (PILP) process.
Journal of Crystal Growth, 2000. 210(4): p. 719-734.
39. Wohlrab, S., H. Colfen, and M. Antonietti, Crystalline, porous
microspheres made from amino acids by using polymer-induced
liquid precursor phases. Angewandte Chemie-International
Edition, 2005. 44(26): p. 4087-4092.
40. McKenna, B.J., Birkedal, H., Bartl, M. H., Deming, T. J.,
Stucky, G. D. Self-Assembling Microspheres from Charged
Functional Polyelectrolytes and Small-Molecule Counterions.
in Mater. Res. Soc. Symp. Proc. 2004.
41. Iler, R.K., Association between Polysilicic Acid and Polar
Organic Compounds. Journal of Physical Chemistry, 1952.
56(6): p. 673-677.
42. Iler, R.K., Coacervates of Polyvinyl-Alcohol and Colloidal
Silica. Journal of Colloid and Interface Science, 1975. 51(3): p.
388-393.
43. Arshady, R., Microspheres and Microcapsules, a Survey of
Manufacturing Techniques .2. Coacervation. Polymer
Engineering and Science, 1990. 30(15): p. 905-914.
32
44. Harada, A. and K. Kataoka, Chain length recognition: Core-
shell supramolecular assembly from oppositely charged block
copolymers. Science, 1999. 283(5398): p. 65-67.
45. de Kruif, C.G., F. Weinbreck, and R. de Vries, Complex
coacervation of proteins and anionic polysaccharides. Current
Opinion in Colloid & Interface Science, 2004. 9(5): p. 340-349.
46. Doublier, J.L., et al., Protein-polysaccharide interactions.
Current Opinion in Colloid & Interface Science, 2000. 5(3-4): p.
202-214.
47. Veis, A. and C. Aranyi, Phase Separation in Polyelectrolyte
Systems .1. Complex Coacervates of Gelatin. Journal of
Physical Chemistry, 1960. 64(9): p. 1203-1210.
48. Weinbreck, F., et al., Complex coacervation of whey proteins
and gum arabic. Biomacromolecules, 2003. 4(2): p. 293-303.
49. Weinbreck, F., et al., Diffusivity of whey protein and gum arabic
in their coacervates. Langmuir, 2004. 20(15): p. 6389-6395.
50. Weinbreck, F., R.H. Tromp, and C.G. de Kruif, Composition
and structure of whey protein/gum arabic coacervates.
Biomacromolecules, 2004. 5(4): p. 1437-1445.
51. Prokop, A., et al., Water soluble polymers for immunoisolation
I: Complex coacervation and cytotoxicity, in
Microencapsulation - Microgels - Iniferters. 1998. p. 1-51.
33
52. Stuart, M.A.C., et al., Assembly of polyelectrolyte-containing
block copolymers in aqueous media. Current Opinion in Colloid
& Interface Science, 2005. 10(1-2): p. 30-36.
53. Leisner, D. and T. Imae, Interpolyelectrolyte complex and
coacervate formation of poly(glutamic acid) with a dendrimer
studied by light scattering and SAXS. Journal of Physical
Chemistry B, 2003. 107(32): p. 8078-8087.
54. Delacruz, M.O., et al., Precipitation of Highly-Charged
Polyelectrolyte Solutions in the Presence of Multivalent Salts.
Journal of Chemical Physics, 1995. 103(13): p. 5781-5791.
55. Cha, J.N., et al., Microcavity lasing from block peptide
hierarchically assembled quantum dot spherical resonators.
Nano Letters, 2003. 3(7): p. 907-911.
56. Cha, J.N., et al., Spontaneous formation of nanoparticle vesicles
from homopolymer polyelectrolytes. Journal of the American
Chemical Society, 2003. 125(27): p. 8285-8289.
57. Toprak, M.S., et al., Spontaneous assembly of magnetic
microspheres. Advanced Materials, 2007. 19(10): p. 1362-+.
58. Wong, M.S., et al., Assembly of nanoparticles into hollow
spheres using block copolypeptides. Nano Letters, 2002. 2(6): p.
583-587.
34
Chapter II
Assembly and Silica Coating of
Polypeptide/Small Molecule Coacervates
*Related versions of this chapter have been published as:
Brandon J. McKenna, Henrik Birkedal, Michael H. Bartl, Timothy J. Deming, and Galen D. Stucky, “Micrometer-sized Spherical Assemblies of Polypeptides and Small Molecules by Acid-Base Chemistry,” Angew. Chem. Int. Ed. 2004, 43, 5652-5622
Brandon J. McKenna, Henrik Birkedal, Michael H. Bartl, Timothy J. Deming, and Galen D. Stucky, “Self-Assembling Microspheres from Charged Functional Polyelectrolytes and Small-Molecule Counterions,” MRS Symp. Proc. 2004, 823, 23
35
Abstract:
The charged polyamino acids were found to assemble into
microspheres in combination with certain small, multivalent
counterions. Assembly was found to be highly dependent on the total
number of ionizable groups and their pKa values, relative to the
polyelectrolyte pKa values and solution pH. Silica shells were induced
on the surfaces of the microspherical assemblies, using either silicic
acid condensation or silica nanoparticle precursors. The use of
nanoparticles created a hollow interior whereas silicic acid did not,
suggesting that strong coulombic forces drive the redistribution of
assembly components.
36
Self-assembled microspheres are important for their potential to
contain and protect one material, while displaying the properties of a
different one on the exterior. Such systems could find applications in
chemical storage and transport, and in particular, biocompatible
microspheres are desirable for applications in drug delivery. Aside
from polymeric micelles and liposomes,1 other microsphere systems
explored for this application require sacrificial templates and/or
surfactants for their self-assembly, or otherwise use organic solvents.2 It
was previously shown that microspheres can be obtained directly by
self-assembly of Cys Lysn m block copolypeptides with either citrate-
coated silver and gold nanoparticles3 or CdSe/CdS nanocrystal quantum
dots,4 or of poly-L-lysine (PLK) with citrate-coated CdSe quantum
dots.5 These assemblies were mechanically stabilized by adding an
outer layer of negatively charged colloidal silica, which also yield
“hollow spheres”—that is, microspherical assemblies with internal
(core) voids of solution and shells of the assembling components,
covered with silica.
The goal of this work was to determine the nature of the
interaction between citrate-capped inorganic nanoparticles and charged
block co-polyamino acids. It was first presumed that the two polymer
segments had specific interactions with the two different nanoparticles,
37
resulting in macromolecular, composite lyotropic liquid crystals similar
in structure to liposomes. In other words, the cysteine blocks of the
polymer would bind to Au NPs, and the lysine blocks would bind to
silica, and these building blocks would self-assemble into larger
structures. However, this conjectured assembly mechanism was not
proven, and it will be demonstrated that it is inconsistent with further
scientific observations.
Herein we report that large spherical assemblies can be obtained
without nanoparticles but simply by reaction of one of several
polyelectrolytes and certain small, functionalized molecular
counterions.6 These assemblies can then be further functionalized, and
we show how those based on polyamines can be protected by silica
either in the form of colloidal silica or by condensation of silicic acid.
Many earlier efforts have concentrated on the ability of multivalent ions
to aggregate oppositely charged polymers, and such systems have been
described by theory7, 8 and studied experimentally,9, 10 particularly in
the case of DNA.11 However, in none of these studies was sphere
formation observed. In other studies, spherical assembly using
polyelectrolytes was observed, but these approaches have required
either amphiphilic block copolymers,12-14 proteins,15 two different
polyelectrolytes,16 hydrophobic molecules,17 or the presence of both
components of an insoluble salt (calcium carbonate).18 The work
38
reported herein demonstrates spherical assembly using only a single
polyelectrolyte with one small counterion without stabilization by an
inorganic species.
The directed formation of hierarchically arranged silica seen in
diatoms and sponges provides a promising framework for designing
synthetic patterned nanoscale materials. Directed biomineralization can
provide novel methods for the assembly of highly ordered structural
materials, as demonstrated by the recent in vitro utilization of some
biological or biomimetic peptides.19-25 Using a biomimetic approach to
silica condensation has many benefits, including: room-temperature
synthesis, neutral or moderate pH’s, the opportunity for hierarchical
ordering, and the ability to vary the resulting structure by tailoring the
active components. Stucky and Morse have shown how silicateins
from marine sponges act as catalysts for silica condensation and as
scaffolds for the directed growth of polysiloxanes.26, 27 Synthetic
poly(amino acid)s have been shown to mimic silicateins and direct the
formation of silica structures, such as spheres.19 Kröger, Sumper and
co-workers have reported that silaffins and polyamines from diatoms
can template organized silica condensation into 0.3-1- m spheres in the
presence of inorganic phosphates.21, 24, 25, 28 In an extension of this
work, Brunner, Lutz, and Sumper have very recently shown that sulfate
and phosphate induce microscopic phase separation of polyallylamine
39
(PAA) and that this phase-separated state in turn has high silica-
precipitation activity to yield silica spheres.29 Clarson and co-workers
have shown that silica microspheres can be obtained from PLK using
silicic acid in phosphate or citrate buffers.22, 23 It is shown herein that
preformed assemblies also condense silicic acid, and propose that the
formation of microspheres in the work of the groups of Sumper,
Brunner, and Clarson may be understood by initial formation of
spherical templates, like those presented herein, prior to silica
condensation. This model is similar to the microscopic phase-separation
picture put forth by Brunner, Lutz, and Sumper.29 The silica spheres
from these systems are in direct contrast to the disordered precipitates
that result when multivalent anions are not used. 20
Materials. Poly-L-lysine hydrochloride (30 kDa), FITC-labeled
poly-L-lysine (70 kDa), poly-L-histidine (10 kDa), poly-L-arginine (30
kDa), poly-L-ornithine (50 kDa), poly-L-glutamate (15 kDa) and poly-
L-aspartate (35 kDa) were obtained from Sigma and used as received.
Snowtex 0 colloidal silica was obtained from Nissan Chemicals.
Organic amines were purchased from Sigma-Aldrich.
In extension of our work based on citrate-stabilized
nanoparticles,3-5 we obtained assemblies by the reaction of citrate (final
concentration 0.5 wt. %) and PLK (final concentration 0.6 wt. %) at pH
40
7. Therefore, nanoparticles themselves are not necessary entities;
molecular citrate itself suffices, as shown in figure 2.1. After mixing
the two components, the solution immediately turned from clear to
cloudy; the resulting colloidal assemblies did not sediment. Within a
few minutes of mixing, spheres were observed by light microscopy, as
illustrated in Figure 2.1a. Thus, a route to assembly has been found that
eliminates the need for nanoparticle reactants. After drying, these
assemblies cling to the glass slide and lose their shape. However, they
can be stabilized by a protective silica shell through the addition of
colloidal silica that condenses on the preformed assemblies (Fig. 2.1b).
It was shown that positively charged amine-containing groups of the
polymers would be able to attract negatively charged colloidal silica.
Adding just a small amount of colloidal silica solution, Snowtex 0, 5%
by volume, is enough to add a visible silica shell to the preformed
spheres. These spheres can be purified by centrifugation and they
maintain their shape out of solution, and can thus be imaged by SEM,
etc.
41
Figure 2.1. Images of PLK/citrate microspheres. (a) Optical micrograph of assemblies prior to colloidal silica condensation, with dust particles digitally removed. (b) SEM image of spheres coated with colloidal silica.
Figure 2.2. Chemical structures of the reactants tested for microsphere formation. The polypeptides in the second row are shown with protonated side chains.
To investigate how generalizable the sphere formation process
is, we explored a host of multivalent organic acids other than citrate
(Table 2.1, Fig. 2.2). PLK-containing spheres were obtained in the
presence of two other triacids (isocitrate and trimesate), but with neither
diacids nor ethylenediaminetetracacetic acid (EDTA). Assuming that
the role of the counterion is to bridge polycations, the diacids may not
42
offer the required kinetic or thermodynamic cooperativity. The failure
of EDTA to form spheres with PLK at any pH value is puzzling, but
may be related to the better hydration of its carboxylate groups.
Table 2.1. Microsphere synthesis at room temperature.[*]
Acid nCOOH pKa(n) PLK PLO PLH PLR
citric 3 6.43 5.5-9.0 5.5-9.5 4.5-6.0 Precipitate
isocitric 3 6.40 5.5-9.0 5.0-9.5 5.0-6.0 Precipitate
trimesic 3 4.7 4.5-8.0 4.5-9.0 4.0-6.0 Precipitate
EDTA 4 10.26 (6.16) NO NO NO 6-10
carbonate 2 10.33 NO NO NO NO
alkanedicarboxylic acids, n(CH2)=0-6
2 3.85-5.69 NO NO NO NO (precipitate with oxalate)
tartaric 2 4.34 NO NO NO NO
malic 2 5.2 NO NO NO NO
fumaric 2 4.54 NO NO NO NO [*] Entries indicate which combination of small organic acid and polycation yields assemblies or precipitates, and the approximate pH ranges over which assemblies are visible in the optical microscope. NO=no assembly or precipitate.
Having found some variability in one assembly component, the
cationic poly(amino acid) was then varied. The triacids created spheres
with poly-L-ornithine (PLO) and with poly-L-histidine (PLH). PLO
shows roughly the same behavior as PLK, which is mostly expected
since their sidechains differ structurally by only one methylene unit. A
striking feature of all the spherical assemblies is that they only existed
within certain pH ranges; it appears that all the components must carry
43
a minimum charge5. PLH provides a particularly interesting case in
that spheres were not obtained at pH 7 but only below the pKa value of
the imidazole side chain of free histidine, 6.0. This result reflects an
assembly requirement for charged groups. Indeed, in all cases the
lower pH boundary for assembly was defined by the highest pKa value
of the acid; in other words, all the carboxyl groups must be
deprotonated, and the spheres disassemble at pH values roughly one
unit below the acid pKa. This behavior is reversible: the spheres
reassemble upon increasing the pH value. Similarly, spheres will not
form if the polymer chain is under-protonated, and assembly is only
seen up to a pH value similar to the formal pKa of the polypeptide side
chains. This situation suggests that the spheres form primarily by
electrostatic attraction, which is most likely accompanied by COO-
+/NH acid-base hydrogen bonding that provides further stabilization. 3
The PLH spheres show remarkable behavior when the pH is
elevated, however. Around pH 7, PLH becomes insoluble, so the
spheres do not dissolve. Instead, spheres of PLH brought to basic
conditions aggregate but maintain their individual shapes for a few
days, before losing their shapes (Fig. 2.3a). This is important because
PLH shows promise as an agent for gene delivery30, and for such
applications the assemblies must be able to survive physiological
conditions.
44
100 μm
50 μm (b)(a)
Figure 2.3. a) Poly-L-histidine/citrate spheres following pH increase. b) Poly-L-arginine/EDTA spheres at pH 12.
Poly-L-arginine (PLR), with its pKa 12.5 side chain, also
showed remarkably different behavior with the triacids, by forming
precipitates even at the highest pHs. However, the PLR-triacid systems
were coaxed into assembling spheres by heating the solution, which
indicates there is an important entropic contribution to both the stability
of the spheres and their kinetics of formation. After cooling, these
spheres resolidify and retain their shapes so that stabilization by silica
was unnecessary (Fig. 2.4). Also, unlike the other cationic poly(amino
acid)s, PLR formed spheres with EDTA, at room temperature. This
may be because the highest pKa of EDTA is below that of the arginine
side chain, so a proton transfers from EDTA to the side chain and
creates an additional acid-base bridge. As with PLH microspheres, the
PLR/EDTA microspheres solidified at high pHs (10-12) (Fig. 2.3b).
45
(b) (a)
Figure 2.4. SEM images of PLR/citrate microspheres following heating and cooling stages. Wrinkling in (b) reflects dehydration that is more pronounced in larger spheres.
The stability of PLK/citrate spheres were also tested at
concentrations of three monovalent salts: NaCl, KCl, and
tetramethylammonium chloride. There is little variability between the
three cases; the spheres shrink and finally cease to be visible after
increasing the salt concentration by about 10 mM, or roughly double
the final value of [COO- +] and [NH3 ]. The spheres may dissolve as a
result of electrostatic screening or competition between Cl- and citrate
binding; it is likely a combination of these two effects, since adding
more citrate can reform spheres, but adding too much citrate (around 15
mM, depending on the PLK concentration) prohibits sphere formation.
The size of the assemblies depends on the ratio of [COO- +] and [NH3 ].
In agreement with the observations of Brunner, Lutz, and Sumper in the
PAA/phosphate system,29 we found that the size of the PLK/citrate
assemblies reach a maximum as the [COO- +]:[NH ] ratio is increased. 3
46
Figure 2.5. Chemical components tested in the synthesis of polyanion/small counterion microspheres.
The microsphere assembly was further extended by combining
anionic poly(amino acid)s with cationic molecules containing multiple
amine groups (Fig. 2.5). Poly-L-aspartate (PLD) and poly-L-glutamate
(PLE) formed spheres with pentalysine and tetraethylenepentamine.
Tetravalent tris(ethyleneamine)amine also gave spheres with PLD;
however, with PLE, assembly required cooling and produced a
combination of spheres and precipitate. Spherical assembly for both
polymers did not occur with the explored divalent or trivalent cations:
1,4-bis(3-aminopropyl)piperazine, 3,3 -diamino-N-
methyldipropylamine, melamine, diethylenetriamine, 2,6-
diaminopyridine, N,N,N ,N ,N -pentamethyldiethylenetriamine, 1-(2-
aminoethyl)piperazine, 1,3-diaminopropane, 1,6-diaminohexane, 1,8-
diaminooctane, and 1,12-diaminododecane.
47
The surfaces of the assemblies are chemically active, and a shell
of colloidal silica can be deposited. The addition of colloidal silica to a
solution of these assemblies gives them a protective shell (Fig. 2.1b).
Confocal microscopy was performed on spheres made from FITC-
labeled PLK (FITC=fluorescein isothiocyanate) to determine polymer
distribution. Before the addition of silica, the uncoated assemblies
were mobile in solution before adhering to the glass slide and losing
shape, and were thus difficult to observe; however, all independent
cross-sectional snapshots clearly showed that they are full of polymer,
with no noticeable uneven distribution. This is in contrast to the
organization found in spheres coated with colloidal silica; figure 2.6
shows a series of cross-sectional scans, from which it is obvious that
the polymer has rearranged and left a cavity in the center.
Figure 2.6. Cross-sectional images taken with confocal microscopy at equal vertical intervals. The sphere was made from FITC-labeled PLK and citrate, and was coated with colloidal silica.
It appears that electrostatic attraction to the negatively charged
silica nanoparticles draws most of the polymer towards the shell,
48
leaving a void in the center. This same behavior has recently been
observed in systems made with citrate-stabilized gold nanoparticles.31
Figure 2.7 shows a superposition of fluorescent and transmission
images, demonstrating that the polymer becomes located inside rim of
the silica coat.
(b) (c)(a)
Figure 2.7. Cross-sectional images of a colloidal-silica-coated sphere made from FITC-labeled PLK and trimesate, a) the polymer fluorescence in a plane through the center of the sphere, b) the optical transmission image at that focal plane, c) an overlay of (a) and (b).
Inspired earlier work,22, 23, 29 we also obtained silica spheres by
adding prehydrolyzed tetramethylorthosilicate (TMOS) to a solution of
preformed polycation/citrate spheres. TMOS was prehydrolyzed at a
final concentration of 0.1 M, in HCl (pH 2-3), for 5 min, before adding
a small amount of this precursor ( 5 % v/v) to a solution of
polycation/citrate spheres. The prehydrolyzed TMOS could also be
quickly neutralized before addition to the spheres. Condensation of the
49
silicic acid was evident from the dark, thin shells seen under the optical
microscope, and from the stability of the resulting spheres after drying
(Fig. 2.8).
Figure 2.8. PLO/isocitrate spheres functionalized by condensed silicic acid: a) an optical image in which the spheres have a thick, dark outline, in contrast to those in figure 2.1a, b) SEM image of a centrifuged sample.
Confocal microscopy was performed on these spheres, as had
been done with those coated with colloidal silica, to verify the
templating action of the assemblies. Cross-sectional images of spheres
after silicic acid condensation also show that their interiors are indeed
full of polymer. Moreover, there is no central cavity as was observed
for the colloidal silica-coated spheres (Figs. 2.9 & 2.10).
(a)
50
20 μm
Figure 2.9. Confocal micrograph of FITC-labeled PLK/citrate spheres stabilized by silicic acid condensation.
(a) (c)(b)
Figure 2.10. Cross-sectional images of FITC-labeled PLK/isocitrate spheres functionalized by condensed silicic acid. Galleries of the fluorescence (a), and transmission (b) responses at different sections through a sphere, c) an overlay of the two images at the central focal plane. The scale bars represent 5 m.
Observations presented here contradict the hollow sphere
assembly model in which block copolymers exhibit specific
interactions with different nanoparticles. The results suggest a more
51
general assembly process, best described by the existing model of
complex coacervation. The substitution of various nanoparticles (Ag,
CdSe) for Au and the identical assembly by homopolylysine suggest
that cysteine-NP specificity is not fundamentally required. It has been
shown herein that citrate itself (the nanoparticle ligand) can be used in
place of the full inorganic nanoparticle, and hollow spheres are still
produced; furthermore, the assembly is observed with the combination
of a range of polyelectolyte/counterion components. In the context of
complex coacervation, which was unknown to the researchers at the
time, the fact of spherical assembly is hardly surprising, but
nevertheless requires verification such as the evidence from confocal
microscopy presented here.
Without asymmetric components, the mechanism of hollow
center formation has to occur by some other mechanism, in order to
break the spatial symmetry, besides surfactant-like self-assembly. Two
alternative mechanisms for hollow center formation were first
considered: 1) the chirality of peptides coupled with condensing
counterions may induce asymmetric tertiary structure that induces
spherical assembly. For instance, extended planar beta sheets could
curve and reduce their free energy by curving into spheres. Rotello et
al have demonstrated examples of hydrogen bonding that could beget
hollow microcapsules.32 2) The assembly could be template by
52
microscopic air bubbles. Air bubbled through commercial fish tanks,
for example, is used to remove organic contaminants and preserve the
purity of the water.
However, we have shown that the hollow space is formed by a
third mechanism: previously full (non-hollow) coacervates expand and
condense to their surfaces by the action of strongly oppositely charged
colloids. The results were corroborated by independent research, which
as well found a variety of assembling chemical components. The work
of Wong, et al has demonstrated other surface nanoparticles that can
yield hollow structures, for instance using ZnO, SnO2, CdSe
carboxylated polystyrene, polyacrylic acid, and polystyrene sulfonate.33
Here, we have demonstrated that the production of a hollow cavity is
optional, as silicic acid precursors can be used as an alternative to
produce spherical shells with interiors that remain “full”.
Conclusion
Complex coacervates can self-assemble from low
concentrations of a charged polyamino acid and an oppositely charged,
multivalent ion. The coacervates have the same properties as earlier
described nanoparticle systems, but they do not require the
complexities of large nanoparticles in order to achieve highly ordered
systems. We have shown that all naturally occurring charged
53
poly(amino acid)s can assemble form coacervates in combination with
small molecules bearing the proper number of charged groups and pKa
values. The pH/pKa requirements and the negative influence of salt
concentration confirm the key role of electrostatics in coacervate
formation. However, there do appear to be some additional
thermodynamic requirements for specific polycation/counteranion
pairs. That is, in some cases the components remain dissolved, and in
other cases a precipitate may form. Importantly, the coacervate
surfaces are chemically active, as shown by the silica condensation
reactions. The resulting shells of silica are uniformly spherical, but the
distribution of the internal components depends on the silica source
used. In the case of silica nanoparticles, the resulting central void may
be useful if the full assembly is to be used as a vesicle for
transportation, storage, or isolating chemical reactions. Finally, the
silica surfaces can be further functionalized for applications in delivery
or detection.
REFERENCES
1. Euliss, L. E.; DuPont, J. A.; Gratton, S.; DeSimone, J.,
Imparting size, shape, and composition control of materials for
nanomedicine. Chemical Society Reviews 2006, 35, (11), 1095-1104.
54
2. Shchukin, D. G.; Sukhorukov, G. B.; Mohwald, H., Smart
inorganic/organic nanocomposite hollow microcapsules. Angewandte
Chemie-International Edition 2003, 42, (37), 4472-4475.
3. Wong, M. S.; Cha, J. N.; Choi, K. S.; Deming, T. J.; Stucky, G.
D., Assembly of nanoparticles into hollow spheres using block
copolypeptides. Nano Letters 2002, 2, (6), 583-587.
4. Cha, J. N.; Bartl, M. H.; Wong, M. S.; Popitsch, A.; Deming, T.
J.; Stucky, G. D., Microcavity lasing from block peptide hierarchically
assembled quantum dot spherical resonators. Nano Letters 2003, 3, (7),
907-911.
5. Cha, J. N.; Birkedal, H.; Euliss, L. E.; Bartl, M. H.; Wong, M.
S.; Deming, T. J.; Stucky, G. D., Spontaneous formation of
nanoparticle vesicles from homopolymer polyelectrolytes. Journal of
the American Chemical Society 2003, 125, (27), 8285-8289.
6. McKenna, B. J., Birkedal, H., Bartl, M. H., Deming, T. J.,
Stucky, G. D. In Self-Assembling Microspheres from Charged
Functional Polyelectrolytes and Small-Molecule Counterions, Mater.
Res. Soc. Symp. Proc., 2004; 2004; pp 4.12.1-6.
7. Arenzon, J. J.; Stilck, J. F.; Levin, Y., Simple model for
attraction between like-charged polyions. European Physical Journal B
1999, 12, (1), 79-82.
55
8. Ha, B. Y.; Liu, A. J., Counterion-mediated attraction between
two like-charged rods. Physical Review Letters 1997, 79, (7), 1289-
1292.
9. Sabbagh, I.; Delsanti, M., Solubility of highly charged anionic
polyelectrolytes in presence of multivalent cations: Specific interaction
effect. European Physical Journal E 2000, 1, (1), 75-86.
10. Delacruz, M. O.; Belloni, L.; Delsanti, M.; Dalbiez, J. P.;
Spalla, O.; Drifford, M., Precipitation of Highly-Charged
Polyelectrolyte Solutions in the Presence of Multivalent Salts. Journal
of Chemical Physics 1995, 103, (13), 5781-5791.
11. Raspaud, E.; de la Cruz, M. O.; Sikorav, J. L.; Livolant, F.,
Precipitation of DNA by polyamines: A polyelectrolyte behavior.
Biophysical Journal 1998, 74, (1), 381-393.
12. Buchhammer, H. M.; Mende, M.; Oelmann, M., Formation of
mono-sized polyelectrolyte complex dispersions: effects of polymer
structure, concentration and mixing conditions. Colloids and Surfaces
a-Physicochemical and Engineering Aspects 2003, 218, (1-3), 151-159.
13. Otsuka, H.; Nagasaki, Y.; Kataoka, K., PEGylated nanoparticles
for biological and pharmaceutical applications. Advanced Drug
Delivery Reviews 2003, 55, (3), 403-419.
14. Solomatin, S. V.; Bronich, T. K.; Bargar, T. W.; Eisenberg, A.;
Kabanov, V. A.; Kabanov, A. V., Environmentally responsive
56
nanoparticles from block ionomer complexes: Effects of pH and ionic
strength. Langmuir 2003, 19, (19), 8069-8076.
15. Weinbreck, F.; Tromp, R. H.; de Kruif, C. G., Composition and
structure of whey protein/gum arabic coacervates. Biomacromolecules
2004, 5, (4), 1437-1445.
16. Biesheuvel, P. M.; Stuart, M. A. C., Electrostatic free energy of
weakly charged macromolecules in solution and intermacromolecular
complexes consisting of oppositely charged polymers. Langmuir 2004,
20, (7), 2785-2791.
17. van Bommel, K. J. C.; Jung, J. H.; Shinkai, S., Poly(L-lysine)
aggregates as templates for the formation of hollow silica spheres.
Advanced Materials 2001, 13, (19), 1472-+.
18. Gower, L. B.; Odom, D. J., Deposition of calcium carbonate
films by a polymer-induced liquid-precursor (PILP) process. Journal of
Crystal Growth 2000, 210, (4), 719-734.
19. Cha, J. N.; Stucky, G. D.; Morse, D. E.; Deming, T. J.,
Biomimetic synthesis of ordered silica structures mediated by block
copolypeptides. Nature 2000, 403, (6767), 289-292.
20. Coradin, T.; Durupthy, O.; Livage, J., Interactions of amino-
containing peptides with sodium silicate and colloidal silica: A
biomimetic approach of silicification. Langmuir 2002, 18, (6), 2331-
2336.
57
21. Kroger, N.; Lorenz, S.; Brunner, E.; Sumper, M., Self-assembly
of highly phosphorylated silaffins and their function in biosilica
morphogenesis. Science 2002, 298, (5593), 584-586.
22. Patwardhan, S. V.; Clarson, S. J., Silicification and
biosilicification - Part 1. Formation of silica structures utilizing a
cationically charged synthetic polymer at neutral pH and under ambient
conditions. Polymer Bulletin 2002, 48, (4-5), 367-371.
23. Patwardhan, S. V.; Mukherjee, N.; Steinitz-Kannan, M.;
Clarson, S. J., Bioinspired synthesis of new silica structures. Chemical
Communications 2003, (10), 1122-1123.
24. Poulsen, N.; Sumper, M.; Kroger, N., Biosilica formation in
diatoms: Characterization of native silaffin-2 and its role in silica
morphogenesis. Proceedings of the National Academy of Sciences of
the United States of America 2003, 100, (21), 12075-12080.
25. Sumper, M.; Lorenz, S.; Brunner, E., Biomimetic control of size
in the polyamine-directed formation of silica nanospheres. Angewandte
Chemie-International Edition 2003, 42, (42), 5192-5195.
26. Shimizu, K.; Cha, J.; Stucky, G. D.; Morse, D. E., Silicatein
alpha: Cathepsin L-like protein in sponge biosilica. Proceedings of the
National Academy of Sciences of the United States of America 1998,
95, (11), 6234-6238.
58
27. Zhou, Y.; Shimizu, K.; Cha, J. N.; Stucky, G. D.; Morse, D. E.,
Efficient catalysis of polysiloxane synthesis by silicatein alpha requires
specific hydroxy and imidazole functionalities. Angewandte Chemie-
International Edition 1999, 38, (6), 780-782.
28. Sumper, M., A phase separation model for the nanopatterning of
diatom biosilica. Science 2002, 295, (5564), 2430-2433.
29. Brunner, E.; Lutz, K.; Sumper, M., Biomimetic synthesis of
silica nanospheres depends on the aggregation and phase separation of
polyamines in aqueous solution. Physical Chemistry Chemical Physics
2004, 6, (4), 854-857.
30. Putnam, D.; Gentry, C. A.; Pack, D. W.; Langer, R., Polymer-
based gene delivery with low cytotoxicity by a unique balance of side-
chain termini. Proceedings of the National Academy of Sciences of the
United States of America 2001, 98, (3), 1200-+.
31. Murthy, V. S.; Cha, J. N.; Stucky, G. D.; Wong, M. S., Charge-
driven flocculation of poly(L-lysine)-gold nanoparticle assemblies
leading to hollow microspheres. Journal of the American Chemical
Society 2004, 126, (16), 5292-5299.
32. Thibault, R. J.; Hotchkiss, P. J.; Gray, M.; Rotello, V. M.,
Thermally reversible formation of microspheres through non-covalent
polymer cross-linking. Journal of the American Chemical Society 2003,
125, (37), 11249-11252.
59
33. Rana, R. K.; Murthy, V. S.; Yu, J.; Wong, M. S., Nanoparticle
self-assembly of hierarchically ordered microcapsule structures.
Advanced Materials 2005, 17, (9), 1145-+.
60
Chapter III
Superparamagnetic Nanoparticle
Coacervates
for Targeted Drug Delivery
*Related versions of this chapter have been published as:
Muhammet S. Toprak, Brandon J. McKenna, J. Herbert Waite, and Galen D. Stucky, “Spontaneous Assembly of Magnetic Microspheres,” Adv. Mat. 2007, 19, 1362-1368 Muhammet S. Toprak, Brandon J. McKenna, J. Herbert Waite, and Galen D. Stucky, “Tailoring Magnetic Microspheres with Controlled Porosity,” MRS Symp. Proc. 2007, 969, W03-11 Muhammet S. Toprak, Brandon J. McKenna, J. Herbert Waite, and Galen D. Stucky, “Control of Size and Permeability of Nanocomposite Microspheres,” Chem. Mat. 2007, 19, 4263-4269
61
Abstract:
We present here a facile bio-inspired method for the synthesis
of inorganic/organic hybrid drug delivery devices based on complex
coacervation. Microspheres were spontaneously formed from the
interaction between cationic polyamines and citrate-coated magnetite
nanoparticles, without the use of a template or surfactant. Control over
the microsphere size distributions were achieved by varying
amine/carboxylate ratios, aging times, temperature, polymer molecular
weight, salt content, and dilution factors. The hybrid spheres were
found to be stabilized by the addition of glutaraldehyde, and the effect
of this organic cross-linking was found to affect the microsphere
porosities and the diffusion coefficients of dextran molecules of various
molecular weights, within the hybrid magnetic microspheres.
62
The ability to assemble materials that are organized over
different length scales has a recognized importance for the development
of new functional materials. In particular, the potential for application
of emerging nanoscale objects can often only be realized by arranging
such components into larger scale assemblies, such that the product
exhibits the functionalities of each of its constituents. One embodiment
of this concept is in the fabrication of microspheres. Microspheres have
been widely considered for isolation of chemical reactions, making
them an active research topic in recent decades. In particular,
biocompatible microspheres are desirable for applications in drug
delivery or other bio-related applications.
Targeted drug delivery has been sought for at least three
decades.1, 2 In the general scheme, drug-encapsulating vesicles
circulate in the bloodstream until they encounter a chosen cell type to
which they then attach and release their drug or toxin into the cellular
environment. The advantages of this process over systemic delivery are
reduced side effects and reduced drug costs, because the drug is not
exposed to the rest of the body and a much smaller amount is required
for the desired effect. Prohibitively expensive but highly effective (or
highly toxic) drugs could find commercial application.3-8
63
The designs of such systems have at least 3 requirements: an
encapsulating material, targeting moieties, and a mechanism for drug
release. Practically, however, more functions are required of such
devices: circulation time in the body must be increased because
particulates are rapidly removed by the liver, so it has become standard
practice to integrate biologically inert surface materials such as PEG.6, 9
The devices must also be biocompatible/biodegradable, they must have
proper shape (usually spherical) to interact with cells,10-12 and they must
be small enough to circulate through capillary pores and reticular
endothelial system (RES), and preferably to endocytose.13
Liposomes are the most common material for encapsulation, as
phospholipids are biocompatible and their assembly is well
understood.2, 14 The greatest difficulty has been modulating their
stability to not release in the bloodstream and yet selectively release
inside of cells. For this reason, a number of new ‘smart’ materials have
been investigated that destabilize under altered solution conditions,
most commonly pH.15-17 This, in effect, combines the ‘encapsulating’
and ‘trigger’ considerations into one. Cancer cells have slightly lower
pHs than other cells,18 so materials with proper pKa’s or which
incorporate chemical bonds whose destruction is catalyzed by H+ have
been researched. I.e., the first category comprises materials with
noncovalent interactions such as polymeric micelles,15, 16, 19, 20 and layer
64
by layer (LbL) polyelectrolyte capsules deposited onto sacrificial
cores,21, 22 and the second category comprises molecules such as lipids
with reversible covalent linkages, such as ketals, vinyl ethers, or ortho
esters.23
Complex coacervation presents new opportunities for single-
step syntheses of ordered micron-scale objects that are composed from
predefined nano-scale objects. Coacervation is a spontaneous aqueous
phase separation, in which liquid-like microspheres are produced from
oppositely-charged chemical entities.24 This imparts some advantages
over other systems which require sacrificial templates, surfactants, or
else the use of organic solvents. A new advantage of coacervation has
emerged with the discovery of nanoparticle-incorporating coacervation
in our laboratory.25, 26 Nanoparticles have established importance in
materials science for their useful physical properties (fluorescent,
magnetic, electrochemical) that emerge at the nanoscale. However, for
commercial application, nano-entities now require appropriate
organization in order to achieve bulk material properties that take
advantage of their functions. Coacervation is now one method to
obtaining microspherical packages containing dense collections of
nanoparticles.
We have previously studied such assemblies made from organic
ions, metallic nanoparticles, or quantum dots; self assembly of
65
Cys Lysn m block copolypeptides with citrate coated gold or silver
nanoparticles, CdSe/CdS quantum dots, or of poly-L-lysine (PLK) with
citrate coated CdSe quantum dots.25-27 We have sought to extend the
variability of assembly components in order to expand the variety of
potential uses. We have utilized the coacervates’ chemical properties to
form inorganic shell structures. Maghemite nanoparticles and block
copolypeptides containing polyaspartic acid were used to form~100 nm
clusters.28 Some of these assemblies were mechanically stabilized by
the addition of an outer layer of negatively charged colloidal silica. A
report by Wong et al. demonstrated the formation of supramolecular
aggregates between cationic polyamines and multivalent counteranions
via ionic cross-linking in a two-step process; and negatively charged
nanoparticles deposit around these aggregates to form a multilayer-
thick shell.29
We have considered coacervation as a potential ‘smart
encapsulating material’ for use in targeted drug delivery. As
mentioned, assembly is facile and fast, and the distinct coacervate
entities require no grinding and present smooth spherical surfaces that
are available for chemical (covalent) modification, such as the
attachment of targeting antbodies or PEG. The range of coacervating
components includes several inexpensive biocompatible polymers of
varying pKa’s, and the expanded number of stabilization methods
66
provides a platform for tailorable release mechanisms. Finally,
coacervates readily incorporate various compounds24 including
hydrophobic liquids,30-32 which simplifies constraints for drug
encapsulation.
In particular, there are many benefits to be derived from
integrating magnetic components into a coacervate microsphere
formulation. Colloidal particles with magnetic properties have gained
increasing attention both technologically and for fundamental studies
due to the tunable anisotropic interaction they exhibit.33, 34 They find
widespread and diverse use in many fields, such as environmental
remediation (removal of toxic and radioactive waste), therapy35
(controlled drug targeting,36 hyperthermia37) and diagnostic biomedical
applications (ELISA, NMR imaging, sensing).38-40 Magnetic
nanoparticles have been used extensively in the field of biomagnetics.
This field currently consists of a broad range of applications, including
drug delivery,36 cell separation,41 biosensing,42 separation of
biochemical products,43 and cell labeling and sorting.44 In the case of
drug delivery, magnetic fields are utilized to direct the particles, and
hence the encapsulated drug, to specific locations within the body—
thus expediting delivery of drugs and minimizing side effects.
We have foremost considered magnetic nanoparticles (MNPs)
for targeted drug delivery. For this application, superparamagnetic
67
nanoparticles, which are smaller than the bulk material’s characteristic
domain size, are desirable, since upon removal of the magnetic field
they lose their magnetic moment and do not aggregate.40, 45 For these
studies, superparamagnetic magnetite (Fe O3 4) nanoparticles were used,
because of their suspected biodegradability and known chemical
synthesis. We consider them for targeted drug delivery for three
functions:
1) Their location in the body can be imaged with magnetic
resonance imaging (MRI). Application as MRI ‘negative’
contrast agents is, itself, important, but for targeted drug
delivery, this would provide a method for visualizing and
confirming proper device delivery prior to release.
2) They may impart directed delivery capability. While
ordinary magnetic fields could not completely counteract
blood flow and permit magnetically guided delivery, a
magnetic field applied near the targeted site (especially near
the body surface) could increase the residence time of the
device, to orient appropriately and hence the likelihood of
attaching to the targeted epithelial cells.
3) They may provide a mechanism for ‘triggered release’.
Superparamagnetic nanoparticles have been studied for
hyperthermia applications, in which an external oscillating
68
magnetic field (in the form of RF) induces NP oscillation
and localized heating.45, 46 This ordinarily confers the
advantage of heating and liberating gaseous O2 within
cancer cells, which weakens the cell and makes it more
susceptible to toxins (systemic or not). In a coacervate-
based design, the MNPs could also destabilize or ‘self-
destruct’ the coacervate this way, and release the
encapsulated drug on command.
In one example a recent synthetic method, Caruso et al. reported
the LbL production of composite magnetic core-shell particles; the
shells consisted of magnetite nanoparticle/polyelectrolyte multilayers
and the colloidal cores were polystyrene latex microspheres.22 For the
various applications of such systems, critical parameters to control
include shell permeability, biocompatibility, mechanical stability, and
size control. Möhwald et al. improved and well tailored the properties
of these polyelectrolyte microcapsules, exemplified with PAH/PSS
microcapsules templated on MnCO3 and melamine formaldehyde (MF)
particles, by glutaraldehyde cross-linking.47
In this study we present a spontaneous, single-step synthesis of
MNP spheres using a combination of positively charged homopolymers
and negatively charged magnetic nanoparticles. Polyamino acids have
attracted considerable interest in recent years due to their non-toxicity,
69
biocompatibility, nutritional function and pharmacological efficacy as
drug carriers.48, 49 Polyamines provide one promising means of
stabilizing drug carriers, since they can self-assemble into micelles via
specific interactions with polyacids50-52 that are capable of trapping
drugs within a core, while their exteriors can be designed to be stable
within a wide range of physiological environments. As with our earlier
investigations, hybrid coacervates could also be formed using other
cationic homopolyamino acids. Herein, poly-L-lysine (PLK) was
chosen as the polycation to interact with citrate-capped magnetite NPs,
because PLK a biodegradable and its coacervation properties were
understood from previous studies.50, 51
The application of coacervate devices depends on the control of
several other important parameters, including their size, stabilization,
and loading capability. We present studies for each of these
parameters: We demonstrate several solution methods for adjusting
microsphere size. We employ glutaraldehyde crosslinking for
coacervate stabilization; this method was selected because the reaction
is well-understood, easily conducted, and easily tracked by the
absorbance of newly formed imide bonds, which also render a positive
surface charge. We show loading capability by demonstrating the
tailorability of critical pore sizes and permeabilities, via cross-link
density. Cross-linked hybrid microspheres were furthermore screened
70
for their toxicity and tested for their ability to incorporate/encapsulate
functional molecules as potential drug delivery system.
Experimental Methods
Chemicals. FeCl .4H O, FeCl .6H O, NH2 2 3 2 4OH, HCl, Poly-L-
Lysine, (PLK 14 kDa, 26 kDa, 46 kDa, 67 kDa), trisodium citrate
(TSC), and FITC-Dextran with different MW (4, 10, 20, 40, 70, and
250 kDa), NaCl, CaCl , MgCl2 2, Phosphate Buffered Saline (PBS),
ammonium acetate, hydroxylamine HCl, o-phenanthroline, and
methanol were obtained from Sigma-Aldrich. Glutaraldehyde and
glycine were obtained from Ted Pella Inc., CA, USA. All chemicals
were used as received and aqueous solutions were prepared by
dissolving the corresponding chemicals in DI water, 18 MΩ.
Surface modified magnetic nanoparticles. Superparamagnetic
nanoparticles were prepared by using a previously described
procedure.53 In a typical process, a mixture of Fe2+ 3+and Fe were
hydrolyzed with NH3 solution at pH >10 in an oxygen-free atmosphere.
Afterwards, the reaction mixture was heated to 80oC under Ar flow,
followed by the addition of TSC.54 Subsequently, the reaction mixture
was cooled, and the magnetic nanoparticles (MNPs) were collected
with the help of a strong magnet and washed 4 times with hot water to
71
remove the excessive reagents. Collected particles were then dispersed
in TSC and stored in a refrigerator for future use.
Coacervate formation and cross-linking. Poly-l-lysines,
regardless of molecular weight, were dissolved in deionized water at
concentrations of 2 mg/mL A solution of magnetic nanoparticles
(MNP) having a net negative surface charge was mixed with a chosen
amount of PLK solution, upon which the solution turned cloudy. A
typical sample was fabricated by mixing 20 μL of 2 mg/mL PLK
solution with 120 μL MNP (1.6 mg/ml TSC). The reaction was mixed
vigorously for 15 seconds using a vortex-mixer.
Glutaraldehyde, GA, has been used frequently as a cross-linking
reagent, because of its lower cost, nontoxicity, and high solubility in
aqueous solution.55 All samples in the following section were cross-
linked by the addition of 120 μL of 2.5 wt% glutaraldehyde, excess of
which was quenched by the addition of glycine. Cross-linked
coacervates are referred to as spheres/microspheres in the following
sections.
Size Effects.
Particle Size Analysis Particle size analysis was performed on
several SEM micrographs counting a minimum of 150 cross-linked
72
microspheres. Size distribution plots are presented using the average
nanocomposite sphere size with one standard deviation, dave ± SD.
- +[COO ]/[NH3 ] ratio, R: The effect of R on microsphere size
was investigated in the range R=1-20. A set of MNP solutions were
prepared by suspending MNPs in TSC solutions with concentrations in
the range 0.77 mM – 15.3 mM while keeping the MNP concentration
constant at 3x1013 MNP/mL. The volumes of PLK (20 μL) and MNP
(120 μL) were preserved.
Temperature: The effect of temperature on the size of hybrid
spheres was investigated in the range 30-70oC. Solutions of PLK4 and
MNP with R=7 were equilibrated in a water bath for 5 min at the
designated temperature. Solutions were then mixed and kept at the
target temperature for another 3 min, after which GA was added to
initiate cross-linking.
Aging: The effect on size by aging coacervate solutions was
studied in the range of 2 min to 3 days. PLK4 and MNP solutions were
mixed to yield an R factor of 7 and aged for different durations prior to
cross-linking.
Polymer Molecular Weight: The influence of polymer molecular
weight, MW, on size was examined using commercially available grade
PLKs, as detailed in Table 3.1. PLK1-PLK6 (2 mg/mL) and MNP
solutions were mixed at R=7 and aged for 6 min prior to cross-linking.
73
Table 3.1. Poly-L-Lysines Used for Experiments
a Dispersity indexPoly-L-Lysine MW, kDaa Mw/Mn PLK1 14 1.4 PLK2 28 1.3 PLK3 46 1.4 PLK4 67 1.1 PLK5 130 1.6 PLK6-FITC b 59 NA a Data provided by the supplier; b 0.008 moles FITC/mole lysine unit, provided by the supplier.
Ionic Strength. The effect of ionic strength on size was
investigated by using chloride salts with different cations; namely
NaCl, CaCl , and MgCl2 2. Stock solutions of these salts at 50 mM
concentration were prepared and added to PLK solutions prior to
mixing with TSC and glutaraldehyde crosslinking.
Dilution. The effect of dilution on size was investigated by
adding different volumes of DI water to PLK solution, prior to mixing
with TSC and glutaraldehyde crosslinking.
Scanning Electron Microscopy, SEM. Scanning Electron
Microscopy analysis was performed using FEI XL40 Sirion FEG
Digital Scanning Microscope. Following cross-linking and quenching
steps, samples were centrifuged and the solution decanted, and then the
microspheres were re-dispersed in DI water. This process was repeated
three times per sample and the resultant suspension was dropped on a
74
Si wafers and dried. Gold coating was performed at 10 mA for 2 min,
prior to SEM analysis. Several SEM micrographs were taken and
particle size analysis was performed by counting a minimum of 150
coacervates. Size distribution plots are presented as the average
coacervate size with one standard deviation, Dave ± SD.
Transmission Electron Microscopy, TEM. Transmission
Electron Microscopy analysis was performed using a FEI Tecnai G2
Sphera Microscope. A drop of sample, from the procedure used in SEM
preparation, was diluted in alcohol and a drop was placed onto the TEM
grid. Following drying, the sample was loaded into a TEM column for
analysis. Particle size distribution was obtained from three micrographs
counting a minimum of 150 particles. Electron diffraction (ED) was
also performed to identify the crystalline phase.
UV-Vis Spectroscopy, UV-Vis. The concentration of magnetite
nanoparticles was determined by a colorimetric method, using UV-Vis
spectrometry by using phenanthroline complexation.56 A known
amount of stable magnetite stock solution was leached in HNO3,
reduced by hydroxylamine. Additions were made of ammonium acetate
buffer followed by o-phenantroline. The absorption of the resulting
iron-phenantroline complexes was observed at 510 nm, as a function of
concentration. An estimate value of 5×1014 MNP/mL was calculated
using the concentration obtained from UV-vis analysis and the average
75
particle size obtained from TEM analysis. MNP solutions in TSC were
prepared from this stock solution to have a final MNP concentration of
3×1013 MNP/mL.
Confocal Microscopy. A Leica microscope equipped with an
ArKr laser was used for performing laser scanning confocal
microscopy. Samples for confocal microscopy were prepared by using
fluorescein isothiocyanate (FITC) labeled PLK and FITC-dextran. A
drop of sample was sandwiched between a glass slide and a coverslip,
which was sealed to avoid evaporation.
Infrared Spectroscopy, IR. Infrared spectroscopy was performed
using a Nicolet Magna 850 IR/Raman spectrophotometer. Infrared
spectra of pure TSC, uncoated and citrate and coated magnetite
particles were obtained using KBr pellets.
Incubation tests. After spontaneously assembled composite
microspheres were cross-linked for different durations, they were
collected by centrifugation and then re-dispersed in DI water. FITC-
dextran solutions with different MWs were then added in 20 µL (~2
mg/mL) portions to 100 µL of the microspheres. After 5 min of
incubation/equilibration, the samples were transferred to glass-slides
and analyzed by confocal laser scanning microscopy (CLSM).
Quantification of Diffusion Coefficient/Permeability. The
microsphere permeability was quantified by means of fluorescence
76
recovery after photobleaching (FRAP) using FITC-dextran as a
molecular probe. To follow the diffusion of FITC-dextran into the
microsphere, the microsphere’s interior was photochemically bleached
with the CLSM ArKr laser (488 nm), at 100% intensity, for sufficient
durations. Imaging was typically performed at about 4% of the
maximal laser intensity. The interval between image scans varied,
depending on recovery rates established in preliminary experiments.
Recovery was considered complete when the intensity of the
photobleached region plateaued. For quantitative analysis, the
fluorescence intensity signals within closed circular areas were
averaged to yield intensity values at each interval.
Results and Discussion.
As determined by TEM (Fig. 3.1a), the synthesized magnetic
iron oxide nanoparticles particle size distribution (Fig. 3.1b), yielded an
average size of 10.5±0.3 nm. The electron diffraction pattern (Fig. 3.1a,
inset), indicated a polycrystalline sample and was indexed to magnetite.
The room temperature magnetization curve, as measured by vibrating
sample magnetometry, is shown in figure 3.1c. The absence of
remanence magnetization and coercivity confirms that the nanoparticles
are superparamagnetic. The mean size of the magnetic core was also
77
calculated to be 10.5±0.5 nm, by assuming a lognormal distribution of
particle volumes for the Langevin function.57
[440]
[220]
[400]
[511]
[422]
[311]
[440]
[220]
[400]
[511]
[422]
[311]
Figure 3.1. (a) TEM micrograph, and SAED of magnetic nanoparticles; (b) size distribution histogram; (c) Room temperature magnetization curve of iron oxide nanoparticles.
The MNP surfaces were modified with citrate in order to effect
the desired self-assembly with polycations, and changes in the MNPs
with different processing were assessed by IR analysis. Uncoated
MNPs only had significant IR absorption bands at 580 cm-1 and 3400
cm-1. In the range 1000-100cm-1, the IR bands of solids usually relate to
crystal lattice ion vibrations.58 The band observed around 600-550 cm-1
corresponds to intrinsic stretching vibrations of the tetrahedral metal
site.59 The 3400 cm-1 band is due to physically adsorbed water. Surface
modification with citrate yielded extra bands at 1395 and 1620 cm-1,
which were assigned to asymmetric and symmetric COO- stretching
-20000 -10000 0 10000 20000
-50
-40
-30
-20
-10
0
10
20
30
40
50
60
-60
Mag
netiz
atio
n [e
mu/
g]
Applied Field [Oe]
(a)
(c)
8 9 10 11 12 130
10
20
30
40
Dh [nm]
% c
ount
s
(b)
78
vibrations, from citrate. Surface modification also affected zeta
potential (Fig. 3.2b). The uncoated MNPs displayed an isoelectric point
around pH 7, whereas the citrate coated MNPs remained negatively
charged, due to the COO- groups of trisodium citrate.
4000 3500 3000 2500 2000 1500 1000 50
--- uncoated magnetite... trisodium citrate___ magnetite w/ citrate
% T
rans
mitt
ance
Wavenumber [cm-1]
1 2 3 4 5 6 7 8 9 10 11 12
-40
-30
-20
-10
0
10
20
30
40
ζ po
tent
ial [
mV
]
pH
uncoated magnetite magnetite w/ citrate
(a) (b)
Figure 3.2. (a) IR spectra of uncoated magnetite, pure trisodium citrate, and citrate coated magnetite nanoparticles; (b) Zeta potential measurements for uncoated and citrate coated magnetite nanoparticles.
Hybrid microspherical coacervates of MNP and PLK were
assembled by Coulombic interactions between the positively charged
amine groups and negatively charged carboxylates, an interaction that
has been hypothesized as charge-stabilized hydrogen bonding.25 Light
microscopy imaging revealed oil droplet like formations diameters of
~1 µm (Fig. 3.3a). Confocal fluorescence microscopy images and
cross-section analysis of the coacervates, prepared using PLK6, show
dome-like features (Fig. 3.3b) because the liquid coacervates wet the
substrate surface. There is synthetic flexibility for tailoring and the
amount of MNP loading within a desired range. Changing R, while
79
keeping [MNP] constant, hybrid spheres with 3-13 wt% MNP loading
have been produced.
(a) (b)
Figure 3.3. (a) Light microscopy and (b) Confocal microscopy images of un-cross linked PLK-MNP spheres.
Mechanical stability of the assemblies is essential for their
successful application as devices. However, coacervation is dynamic,
and the interactions are easily disturbed by various factors that alter the
degree of ionization, such as as ionic strength, pH, and temperature. For
example, a deviation of ±1 unit from the pI of the polyelectrolyte leads
to dissolution. The addition of even a small amount (10 mM) of
indifferent salt disturbs the system by electrostatic screening. Therefore
PLK was cross-linked using glutaraldehyde (GA), a known effective
cross-linker between amine groups. Cross-linking introduces new -
C=N- imide bonds that give rise to absorbance in the UV region.47
Therefore, the cross-linking was followed and verified by UV-Vis
spectroscopy (Fig 3.4b). The broad absorption in the UV region of 250-
80
300 nm increased with the treatment time, indicating more imide bonds.
The absorbance at 266 nm as a function of time is shown in the inset,
which indicates that the reaction initially proceeds very quickly and
then gradually slows. Cross-linking for extended periods resulted in
aggregation; sedimentation can be monitored by decreasing absorption.
Hence, the degree of cross-linking can be conveniently tuned by
reaction time prior to quenching with glycine.
250 300 350 4000
2
0 10 20 30
0.6
0.7
0.8
0.9
1.0
1.1
1.2
Abs
orpt
ion
@ 2
66 n
m [a
rb. u
.]Time [min]
Abs
orpt
ion
[arb
. u.]
Wavelength [nm]
(a) (b)
(c)
2 μm
Figure 3.4. (a) Optical micrograph of as-prepared magnetic coacervates; (b) UV-Vis spectra of PLK/MNP coacervates cross-linked with 2.5 wt% GA for different durations: 0, 1, 4, 6, 9, 13, 25 min., from bottom to top. The inset shows the absorption due to newly formed imide bonds at 266 nm as a function of time; (c) SEM micrograph of cross-linked magnetic microspheres.
Size Study Results. Cross-linked hybrid PLK-MNP microspheres
were examined with SEM for their size and morphology (Fig. 3.5).
Individual MNPs as well as aggregates of MNPs can easily be seen in
figure 3.5b. The formation and size of hybrid coacervates was found to
81
- +be dependent on the [COO ]/[NH3 ] ratio R, PLK MW, aging time,
temperature, ionic strength, and dilution.
Figure 3.5. SEM micrographs of cross-linked PLK-MNP spheres: (a) General view; (b) Close-up view showing individual nanoparticles and aggregates
The effect of R on size was investigated in the range of 1-20
(Fig. 3.6a). Below R=3, no coacervation occurred. This contrasts with
an earlier report where excess positive charge was required for vesicle
formation.25 Because R=1 indicates a net charge balance of negatively
and positively charged participating ions, this result indicates that the
MNP/citrate bridges are too mobile and labile to effectively form
noncovalent crosslinks. Coacervates appeared at R=3, with an average
size of 1.2 μm, which decreases with increasing R, down to 560 nm at
R=20. This trend is likely the result of increasing ionic strength,
although it may also reflect faster nucleation kinetics or a
thermodynamic drive to decrease surface tension (by reducing the
number of MNPs per total surface area). The microspheres are
(a) (b)
200 nm 1 μm
82
positively charged at all R values, revealing the dominant role of PLK
in these formations, in a “cloud” of smaller anions.
R R1 R2 R3 R5 R6 R7 R8 R10 R15 R20 ---0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
Dav
e [μm
]
R [CH3COO-]/[NH3+]
T30 T40 T50 T60 T70
500
600
700
800
900
1000
1100
1200
1300
Par
ticle
Siz
e [n
m]
Temperature [oC]
(a) (b)
Figure 3.6. Depende +nce of sphere size on: (a) [COO-]/[NH3 ] ratio, R; (b) Temperature.
The effect of temperature on the size of the hybrid spheres in
the temperature range 30-70oC is shown in Figure 3.6b. The average
size of the hybrid spheres is around 700 nm at ambient temperature.
This value increases to ~1 μm at 70oC, which is expected, since
aggregation of like-charged bodies requires thermal energy to
overcome Coulombic repulsion.
Aging in the range of 2 min–3d also affected the coacervate size
(Fig. 3.7a). A distinct size distribution difference was observed, and
larger MNP/PLK hybrid spheres were found with longer aging times.
The sphere size ranged from an average diameter of 700 nm at 2 min to
about 1.7 μm at 24 hrs. Prolonged aging resulted in the disappearance
83
of the coacervates because they adhered to container walls, as was
visualized using PLK6.
0 10 20 30 40 50 60 70 80 90 2000 3000 4000 5000
0.0
0.4
0.8
1.2
1.6
2.0
2.4
Dav
e [μm
]
Aging time [min]13K 28K 46K 67K 130K
0.0
0.4
0.8
1.2
1.6
2.0
2.4
Dav
e[μm
]MW PLL [kDa]
(a) (b)
Figure 3.7. (a) Aging time; and (b) Poly-L-Lysine molecular weight dependence of the size of PLK-MNP spheres, R=7.
The microsphere size shows a sub-linear square root
dependence to the aging time. This is in agreement with earlier findings
on polymer-colloidal particles, and indicates a diffusion limited process
in which polymer and/or MNP adsorbs onto coacervate surfaces faster
than bulk diffusion within the coacervates.60, 61 Other mechanisms
could also contribute to growth, such as entropy-minimizing
coalescence and Ostwald ripening, in which larger coacervates grow at
the expense of smaller ones.
MW dependence of the hybrid spheres size (Fig 3.7b) is
summarized in table 3.2 with the average sphere size, Dave, standard
deviation, SD, and minimum and maximum sizes observed for each set.
No coacervates were observed for the lowest two MWs (13 kDa and 26
84
kDa). Hybrid coacervate size thereafter increased with PLK MW, from
600 nm (at 46 kDa) to 1.9 μm (at 130 kDa). The threshold MW of PLK
for hybrid coacervate formation is suggested to be between 26 kDa and
46 kDa, for this particular system.
Table 3.2. Size of hybrid spheres as a function of PLK MW at R=7.
Poly-L-Lysine
MW, kDa
D SD Min Max ave, μm
PLK1 14 -- -- -- -- PLK2 26 -- -- -- -- PLK3 46 0.61 0.13 0.34 0.88 PLK4 67 0.95 0.29 0.31 1.61 PLK5 130 1.64 0.5 0.66 2.95
The effect of ionic strength on the coacervates size was
investigated by using salts with cation valencies of +1 and +2 (Figs. 3.8
and 3.9). The average size was observed to decrease with increasing
ionic strength. In the case of NaCl addition, the average microsphere
size was reduced from 700 nm to 300 nm with 25 mM of added ionic
strength (Fig. 3.8). No formation was observed above 25 mM ionic
strength for NaCl. In the case of CaCl2, the average size decreased to
350 nm with an added ionic strength of 6 mM, above which coacervates
dissolved (Fig. 3.9). The decrease in average size was attributed to
electrostatic screening of noncovalent bridges. When Mg2+ was used,
no spherical assemblies were observed at any ionic strength conditions.
85
The significant difference in responses to these two different salts is
attributed to complexation of Ca2+ with citrate ions. This greater
activity of Mg2+ is attributed to its much larger radius of hydration
(~4X larger than hydrated Ca2+), which causes significant shielding,
spatially, of citrate ions and thus even more interference with
assembly.62
100 1000-5
0
5
10
15
20
25
30
35
40
45
50
55
60
% c
ount
dave [nm]
50 mM NaCl 0 μL 6 μL 16 μL 35 μL 60 μL 140 μL
0 50 100 150 200 250-0.1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
d ave [μm
]
VNaCl [μL]
dave
0
4
8
12
16
20
24
28
32
Ionic strength
Ioni
c S
treng
th [m
M](a) (b)
Figure 3.8. The effect of adding aliquots of 50 mM NaCl on the average hybrid microsphere diameter and solution ionic strength. Error bars indicate one standard deviation.
100 1000
0
5
10
15
20
25
30
35
40
% c
ount
dave [nm]
50 mM CaCl2 0 μL 3 μL 6 μL
0 5 10 15-0.1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
d ave [μm
]
VCaCl2 [μL]
dave
-2
0
2
4
6
8
10
12
14
16
18
20
Ioni
c St
reng
th [m
M]
Ionic Strength
(b) (a)
Figure 3.9. The effect of adding aliquots of 50 mM CaCl2 on the average hybrid microsphere diameter and solution ionic strength. Error bars indicate one standard deviation.
86
The effect of dilution on the average size of hybrid
microspheres was similar to that observed for increasing ionic strength
(Fig. 3.10). Dilution is expected to have two effects on size.
Kinetically, rapidly-formed charged coacervate ‘nuclei’ are too
spatially separated to overcome repulsion and then coalesce.
Thermodynamically, dilution increases the total amount of material in
the equilibrium phase.
100 1000
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
% c
ount
dave [nm]
(DI H2O)
0 μL 25 μL 50 μL 100 μL 200 μL 300 μL 500 μL 1000 μL
0 100 200 300 400 500 600 700 800 900 10000.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
d ave [μm
]
VH2O [μL]
(a) (b)
Figure 3.10. Diameter distributions of hybrid microsphere samples at different dilutions, as calculated from SEM micrographs. Error bars indicate one standard deviation.
Based on the data obtained, there is synthetic flexibility for tailoring
the size of coacervates. Size control can be achieved by changing the
ratio R, the temperature of synthesis, MW of the polyamine or by
controlling the aging time of the coacervates.
Permeability/Porosity Study Results.
Prior to further detailed confocal microscopy studies, the
fluorescence emission data were collected for microsphere components
87
and microspheres at different processing stages, using 488 nm
excitation (Fig. 3.8).
500 550 600 650 700
MNP w TSC coa w/ MNP coa no MNP cross-linked PLK
Fluo
resc
ence
inte
nsity
λ [nm]
500 550 600 650 700
FITC-dextran MS + 40 μL FITC-dextran MS + 20 μL FITC-dextran MS w/ MNP
Fluo
resc
ence
inte
nsity
λ [nm]
(a) (b)
Figure 3.8. Fluorescence emission of microspheres at various stages.
As control experiments, both hybrid coacervates and cross-
linked PLK were analyzed (Fig 3.8a). Coacervate solutions with MNPs
alone showed an emission band centered at 520 nm, which was also
observed for a suspension of MNPs. Cross-linked magnetic
microspheres showed an emission band around 550 nm, as did cross-
linked PLK; therefore this band was assigned to emission from imide
cross-links. Emission of magnetite entrapped within these cross-linked
microspheres could not be resolved, perhaps due to exciton transfer to
PLK. The emission of FITC-labeled dextran, centered at 520 nm (Fig.
3.8b), slightly blue-shifted to 510 nm at higher concentrations, because
of interaction with the reaction medium. Importantly, this analysis
indicates the proper wavelengths for differentiating signals of imide
bonds and FITC: FITC detection was performed in the range 510-530
88
nm, while the cross-linked microspheres were detected in the range of
540-570 nm.63
The hydrodynamic size distribution of the various FITC-dextran
molecules is presented in figure 3.9a; the average hydrodynamic size
increased with MW, as expected (Fig. 3.9b). Thirteen samples of
microspheres were prepared with cross-linking durations of 1, 2, 3, 5, 8,
12, 16, 24, 33 min, 1 hr, 3 hr, 24 hr, and 48 hr. The results are
summarized in Table I under three categories: (i) permeable to FITC-
dextran, as observed by complete microsphere filling (Fig. 3.11c); (ii)
impermeable to dextran, such that negatively charged FITC-dextran
molecules are adsorbed only on the microsphere surfaces (Fig. 3.11d);
and (iii) critically permeable, for sample sets in which both (i) and (ii)
are observed. These observations show that porosity of these composite
microspheres could be tailored by an adjustable and measurable cross-
linking treatment.
89
1 10 100-5
0
5
10
15
20
25
30
35
40
4 kD a 10 kD a 20 kD a 40 kD a 70 kD a 150 kD a 250 kD a
Vo
lum
e [
%]
S ize [d .nm ] (a) (c)
0 50 100 150 200 250
2
4
6
8
10
12
14
16
18
20
d [n
m]
M w (FITC-dextran) [kDa] (b) (d)
Figure 3.11. (a) Hydrodynamic size distribution of FITC-dextran molecules with different MWs; (b) Average size of FITC-dextran molecules with different MWs; Confocal microscopy images of hybrid microspheres that are (c) permeable and (d) impermeable to FITC-dextran.
90
FITC-dextran with different MWs CL Duration 4 kDa 10 kDa 20 kDa 40 kDa 70 kDa 150 kDa 250 kDa t=1 min t=2 min t=3 min t=5 min t=8 min t=12 min t=16 min t=24 min t=33 min
t=1 hr t=3 hr t=24 hr t=48 hr
Permeable
Critical Perm-eability
Imperm-eable
Table 3.3. Permeability of hybrid microspheres, cross-linked (CL) for different durations, to FITC-dextran molecules with different MWs.
The two shortest durations of cross-linking, 1 and 2 min.,
resulted in the largest pores and permitted entry of FITC-dextran
molecules with an average hydrodynamic radius of 18 nm (250 kDa).
Prolonging cross-linking for 3 and 5 minutes reduced the critical pore
size and allowed the diffusion of 12 nm molecules. Eight minute cross-
linking reduced the critical pore size to under 12 nm, and 12 min cross-
linking further reduced it to around 8 nm. Extended periods of cross-
linking (between 16 min-48 hrs) resulted in microspheres with similar
porosities—all were permeable to 5 nm dextran. As expected, the
longest duration of cross-linking in this series resulted in the smallest
pore size. Importantly, pore sizes did not change significantly for
samples cross-linked beyond 16 min., despite an increasing UV
91
absorption peak. This suggests that some GA molecules coupled to
PLK did not function as intermolecular cross-linkers, despite Schiff
base formation.
Diffusion coefficients were measured by adding excess
fluorescein-labeled dextran to cross-linked microspheres and
performing fluorescence recovery after photobleaching (FRAP)
according to established procedures.47, 64, 65 Diffusion of FITC-dextran
molecules into the microspheres was monitored with time to generate a
fluorescence recovery curve.
The fluorescence recovery as a function of time, in a central
planar cross-section of solid microspheres, is given exactly by (Fig.
3.12a):
2
0
(2 1)
20 0 2
1
8( ) 1(2 1)
ntDR
exactn
eC t R Cn
π
ππ
⎛ ⎞−− ⎜ ⎟
∞ ⎝ ⎠
=
⎛ ⎞⎜ ⎟
= −⎜−⎜⎜ ⎟
⎝ ⎠
∑ 2 ⎟⎟ (1)
where Cexact(t) and Co denote the fluorescence/concentration at times t
and t = 0, respectively. D is the diffusion constant, and Ro is the radius
of the microsphere.
92
Figure 3.12. (a) Characteristic recovery vs. time curve, for diffusion into a solid sphere cross-section. (b) Radial concentration profiles at equal time increments, using 20 summation terms.
The equation was derived from the diffusion equation in spherical
coordinates, using separation of time and space variables. The r-
dependence was matched to a Bessel equation or order ½, and boundary
conditions and initial conditions were used to describe an initially
empty sphere within a medium that maintains constant concentration at
the sphere surface (Fig. 3.12b).66
2
000
1 0
2 ( 1)( , ) 1 sinnn tDR
n
R n rC r t C er n R
ππ
π
⎛ ⎞∞ − ⎜ ⎟⎝ ⎠
=
⎛ ⎞⎛ ⎞−⎜ ⎟= + ⎜ ⎟⎜ ⎟⎝ ⎠⎝ ⎠∑ (2)
This full space-time (Eq. 2) equation was integrated to obtain (1). For
times after 40% recovery, the expression can be approximated by using
only the first term of the infinite summation (Fig. 3.13a):
2
0240% 0 0 2
8( ) 1tD
RC t R C eπ
ππ
⎛ ⎞− ⎜ ⎟
⎝ ⎠⎛ ⎞⎜ ⎟= −⎜⎝ ⎠
⎟ (3)
1
5 10 15 20 25 30
0.2
0.4
0.6
0.8
1
-1.5 -1 -0.5 0.5 1 1.5
0.2
0.4
0.6
0.8(b) (a)
C t r
93
Finally, the effects of partially unbleached dye at t=0 and instrumental
time delay between bleaching and measurements can be simultaneously
corrected by fitting with an “unbleached” parameter u:
2
020 0 2
8( ) 1 (1 )tD
RC t R C u eπ
ππ
⎛ ⎞− ⎜ ⎟
⎝ ⎠⎛ ⎞⎜= − −⎜⎝ ⎠
⎟⎟ (4)
The fluorescence recovery curves for different samples of known cross-
sectional areas were fit to this equation, using least-squared
minimization (Fig. 3.13b). This yielded the diffusion coefficients of
FITC-dextran molecules into composite microspheres of different
cross-linking durations.
0.5 1 1.5 2
0.1
0.2
0.3
0.4
20 40 60 80 100 120 140
80
90
100
110
(a) (b)
t t
Figure 3.13. (a) Convergence of recovery curves, using 1, 2, and 50 summation terms. Use of 1 term is accurate by ~40% recovery; 2 terms give accuracy at around ~20% recovery. (b) Representative fits to FRAP data, demonstrating experimental and theoretical agreement.
The diffusion coefficients of FITC-dextran molecules of
different MWs into cross-linked microspheres were on the order of 10-
15 m2/s (Fig. 3.14a), which were about four orders of magnitude smaller
than those of the free molecules in water. This decrease may be partly
94
attributed to altered diffusion in the coacervate phase, but it is mainly
due to the cross-linking process. Furthermore, a series of FRAP
experiments were performed on microspheres cross-linked for various
durations, using 10 kDa FITC-dextran molecules. Diffusion coefficients
calculated for FITC-dextran molecules within these microspheres (Fig.
3.14b) showed a trend of decreasing D with greater cross-linking.
Hence, cross-links reduced pore-sizes and restricted free movement of
FITC-dextran molecules. These results suggest that diffusion through,
and hence release from, the nanocomposite network can be controlled
by macromolecule size relative to microsphere pore size.
0 5 10 15 20 25 30 35
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
D x
10-1
5 [m2 /s
]
tcross-link [min]
10 kDa FITC-Dextran
2 3 4 5 6 7 8 9 10 11
1E-15
1E-10
D [m
2 /s]
ddextran [nm]
in 8min cross-linked μspheres in H2O
(a) (b)
Figure 3.14. Diffusion coefficients calculated from FRAP experiments for (a) 10 kDa FITC-dextran diffusing into microspheres cross-linked for different durations, and (b) Dextran molecules with different MWs diffusing into 8 min cross-linked microspheres.
95
Conclusions
This work demonstrates the first example of coacervation that
uses magnetic nanoparticles as assembling components, exemplified
here as PLK/TSC microspheres. One-step formation of these MNP-
based microspheres can be considered as a nanoparticle self-assembly
process induced by the presence of polyelectrolytes. Solid
microspheres can be rapidly generated at ambient temperature, in water,
and without the need of a solid template. The described procedure is
straightforwardly applicable to most nanoparticles that bear the proper
surface functionality. The resulting objects can be tailored by
modifying the conditions during and/or after the assembly step, they
retain their magnetic properties. Glutaraldehyde cross-linking proved
effective in mechanical stabilization of liquid coacervates. The hybrid
sphere size was controlled by adjusting the ratio R, temperature, aging
time, the molecular weight of the polymer, ionic strength, and the
amount of dilution in the synthesis medium.
In addition to providing mechanical stability, cross-linking
further permitted the tailoring of porosity, the extent of which was
monitored by UV-Vis spectroscopy. Cross-linking decreased
microsphere permeabilities and resulted in smaller critical pore sizes.
96
FRAP experiments, performed using FITC-dextran molecules of
various MW, revealed that cross-linking could control microsphere
permeabilities, and that the dextran diffusion coefficients decreased by
four orders of magnitude. These results are of critical importance for
the design of functional microspheres that can control diffusion based
on size selection, or for entrapment of molecules of a chosen size while
allowing free transfer of smaller molecules.
Based on the microsphere porosities as controlled by
glutaraldehyde cross-linking, it is concluded that this design is
appropriate for encapsulating agents larger than 5nm (such as some
protein formulations or genes). For smaller drugs, additional
modifications are necessary; for instance the drugs may be covalently
bound to the coacervate interior, or else the outermost pores must be
blocked by ‘plugging’ with larger molecules or by coating with
additional, probably inorganic, material. Alternatively, other cross-
linking methods might be sought, such as amide bond formation
catalyzed by EDC. This method would be expected to further restrict
pore size because 1) two coacervating components rather than one
contribute to cross-linking and 2) all cross-links would necessarily be
intermolecular.
Although PLK and MNPs with citrate were featured here as a
case study, the design principles and synthetic methods are applicable
97
to a variety of polyamines and nanoparticles that are appropriately
surface-functionalized. The incorporation of magnetic nanoparticles
into this drug delivery device design provides opportunities for MRI
visualization, in vivo magnetic steering, and RF-triggered hyperthermia.
The effectiveness of each of these will require further studies, as will
the evaluation of a hyperthermic, triggered release mechanism in vitro.
In vitro cell separation studies on antibody-coated devices are needed to
prove targeting capabilities. In vitro cell viability studies have already
been performed, and suggest that the hybrid microspheres have low
toxicity.67
Finally, in vivo studies will be required to identify any
additional shortcomings of such devices in the complex and
unpredictable settings of animal and human bodies. Particulate-based
drug delivery systems usually face the problem of short circulation time
due to removal by the liver. Even with modern technology’s best
developed antibody formulations and delivery methods, all particulate-
based devices currently suffer from their inability to infiltrate more than
one-third of the targeted cells. Since cancer requires eradication of
100% of cancer cells in order to avoid progression to metathesis, new
physical and biological breakthroughs in delivery methods are required
to make targeted drug delivery commercially realistic. Meanwhile, a
prospective new material has now been developed and lies in wait of
98
improved targeting strategies. The widespread applicability of these
nanocomposites as magnetically functionalized drug delivery devices or
Magnetic Resonance Imaging contrast agents is supported by their
facile assembly, tunability of properties, encapsulation of dextran, and
low toxicity.
REFERENCES
1. Suit, H. D.; Shwayder, M., Hyperthermia - Potential as an
Antitumor Agent. Cancer 1974, 34, (1), 122-129.
2. Yatvin, M. B.; Weinstein, J. N.; Dennis, W. H.; Blumenthal, R.,
Design of Liposomes for Enhanced Local Release of Drugs by
Hyperthermia. Science 1978, 202, (4374), 1290-1293.
3. Leduc, P. R.; Wong, M. S.; Ferreira, P. M.; Groff, R. E.;
Haslinger, K.; Koonce, M. P.; Lee, W. Y.; Love, J. C.; McCammon, J.
A.; Monteiro-Riviere, N. A.; Rotello, V. M.; Rubloff, G. W.;
Westervelt, R.; Yoda, M., Towards an in vivo biologically inspired
nanofactory. Nature Nanotechnology 2007, 2, (1), 3-7.
4. Moghimi, S. M.; Hunter, A. C.; Murray, J. C., Nanomedicine:
current status and future prospects. Faseb Journal 2005, 19, (3), 311-
330.
5. Rezler, E. M.; Khan, D. R.; Lauer-Fields, J.; Cudic, M.;
Baronas-Lowell, D.; Fields, G. B., Targeted drug delivery utilizing
99
protein-like molecular architecture. Journal of the American Chemical
Society 2007, 129, (16), 4961-4972.
6. Torchilin, V. P., Multifunctional nanocarriers. Advanced Drug
Delivery Reviews 2006, 58, (14), 1532-1555.
7. Vasir, J. K.; Reddy, M. K.; Labhasetwar, V. D., Nanosystems in
drug targeting: Opportunities and challenges. Current Nanoscience
2005, 1, (1), 47-64.
8. Wang, M. D.; Shin, D. M.; Simons, J. W.; Nie, S. M.,
Nanotechnology for targeted cancer therapy. Expert Review of
Anticancer Therapy 2007, 7, (6), 833-837.
9. Stolnik, S.; Illum, L.; Davis, S. S., Long Circulating
Microparticulate Drug Carriers. Advanced Drug Delivery Reviews
1995, 16, (2-3), 195-214.
10. Champion, J. A.; Mitragotri, S., Role of target geometry in
phagocytosis. Proceedings of the National Academy of Sciences of the
United States of America 2006, 103, (13), 4930-4934.
11. Euliss, L. E.; DuPont, J. A.; Gratton, S.; DeSimone, J.,
Imparting size, shape, and composition control of materials for
nanomedicine. Chemical Society Reviews 2006, 35, (11), 1095-1104.
12. Rolland, J. P.; Maynor, B. W.; Euliss, L. E.; Exner, A. E.;
Denison, G. M.; DeSimone, J. M., Direct fabrication and harvesting of
100
monodisperse, shape-specific nanobiomaterials. Journal of the
American Chemical Society 2005, 127, (28), 10096-10100.
13. Vonarbourg, A.; Passirani, C.; Saulnier, P.; Benoit, J. P.,
Parameters influencing the stealthiness of colloidal drug delivery
systems. Biomaterials 2006, 27, (24), 4356-4373.
14. Heath, T. D.; Fraley, R. T.; Papahadjopoulos, D., Antibody
Targeting of Liposomes - Cell Specificity Obtained by Conjugation of
F(Ab')2 to Vesicle Surface. Science 1980, 210, (4469), 539-541.
15. Bae, Y.; Nishiyama, N.; Fukushima, S.; Koyama, H.; Yasuhiro,
M.; Kataoka, K., Preparation and biological characterization of
polymeric micelle drug carriers with intracellular pH-triggered drug
release property: Tumor permeability, controlled subcellular drug
distribution, and enhanced in vivo antitumor efficacy. Bioconjugate
Chemistry 2005, 16, (1), 122-130.
16. Bae, Y.; Fukushima, S.; Harada, A.; Kataoka, K., Design of
environment-sensitive supramolecular assemblies for intracellular drug
delivery: Polymeric micelles that are responsive to intracellular pH
change. Angewandte Chemie-International Edition 2003, 42, (38),
4640-4643.
17. Guo, X.; Szoka, F. C., Chemical approaches to triggerable lipid
vesicles for drug and gene delivery. Accounts of Chemical Research
2003, 36, (5), 335-341.
101
18. Gatenby, R. A.; Gillies, R. J., Why do cancers have high aerobic
glycolysis? Nature Reviews Cancer 2004, 4, (11), 891-899.
19. Nishiyama, N.; Kataoka, K., Current state, achievements, and
future prospects of polymeric micelles as nanocarriers for drug and
gene delivery. Pharmacology & Therapeutics 2006, 112, (3), 630-648.
20. Nishiyama, N.; Kataoka, K., Nanostructured devices based on
block copolymer assemblies for drug delivery: Designing structures for
enhanced drug function. In Polymer Therapeutics Ii: Polymers as
Drugs, Conjugates and Gene Delivery Systems, 2006; Vol. 193, pp 67-
101.
21. Shchukin, D. G.; Sukhorukov, G. B.; Mohwald, H., Smart
inorganic/organic nanocomposite hollow microcapsules. Angewandte
Chemie-International Edition 2003, 42, (37), 4472-4475.
22. Caruso, F.; Caruso, R. A.; Mohwald, H., Production of hollow
microspheres from nanostructured composite particles. Chemistry of
Materials 1999, 11, (11), 3309-3314.
23. Huang, Z.; Szoka, F. C., Bioresponsive Liposomes and Their
Use for Macromolecular Delivery. In Liposome Technology, Third ed.;
Gregoriadis, G., Ed. Informa Healthcare: 2006; pp 165-196.
24. Bungenberg de Jong, H. G., In Crystallisation- coacervation-
flocculation in colloid science, Ed. Kruyt, H. R., Ed. Elsevier:
Amsterdam, 1949; Vol. II, pp 232-258.
102
25. Cha, J. N.; Birkedal, H.; Euliss, L. E.; Bartl, M. H.; Wong, M.
S.; Deming, T. J.; Stucky, G. D., Spontaneous formation of
nanoparticle vesicles from homopolymer polyelectrolytes. Journal of
the American Chemical Society 2003, 125, (27), 8285-8289.
26. Wong, M. S.; Cha, J. N.; Choi, K. S.; Deming, T. J.; Stucky, G.
D., Assembly of nanoparticles into hollow spheres using block
copolypeptides. Nano Letters 2002, 2, (6), 583-587.
27. Cha, J. N.; Bartl, M. H.; Wong, M. S.; Popitsch, A.; Deming, T.
J.; Stucky, G. D., Microcavity lasing from block peptide hierarchically
assembled quantum dot spherical resonators. Nano Letters 2003, 3, (7),
907-911.
28. Euliss, L. E.; Grancharov, S. G.; O'Brien, S.; Deming, T. J.;
Stucky, G. D.; Murray, C. B.; Held, G. A., Cooperative assembly of
magnetic nanoparticles and block copolypeptides in aqueous media.
Nano Letters 2003, 3, (11), 1489-1493.
29. Rana, R. K.; Murthy, V. S.; Yu, J.; Wong, M. S., Nanoparticle
self-assembly of hierarchically ordered microcapsule structures.
Advanced Materials 2005, 17, (9), 1145-+.
30. Luzzi, L. A.; Gerraughty, R. J., Effects of Selected Variables on
Extractability of Oils from Coacervate Capsules. Journal of
Pharmaceutical Sciences 1964, 53, (4), 429-&.
103
31. Weinbreck, F.; Minor, M.; De Kruif, C. G., Microencapsulation
of oils using whey protein/gum arabic coacervates. Journal of
Microencapsulation 2004, 21, (6), 667-679.
32. Green, B. K., and Schleicher, Lowell Oil-containing
Microscopic Capsules and Method of Making Them. 2,800,457, 1957.
33. Promislow, J. H. E.; Gast, A. P., Low-energy suspension
structure of a magnetorheological fluid. Physical Review E 1997, 56,
(1), 642-651.
34. Wirtz, D.; Fermigier, M., One-Dimensional Patterns and
Wavelength Selection in Magnetic Fluids. Physical Review Letters
1994, 72, (14), 2294-2297.
35. Tartaj, P.; Morales, M. D.; Veintemillas-Verdaguer, S.;
Gonzalez-Carreno, T.; Serna, C. J., The preparation of magnetic
nanoparticles for applications in biomedicine. Journal of Physics D-
Applied Physics 2003, 36, (13), R182-R197.
36. Hafeli, U.; Pauer, G.; Failing, S.; Tapolsky, G., Radiolabeling of
magnetic particles with rhenium-188 for cancer therapy. Journal of
Magnetism and Magnetic Materials 2001, 225, (1-2), 73-78.
37. Johannsen, M.; Gneveckow, U.; Eckelt, L.; Feussner, A.;
Waldofner, N.; Scholz, R.; Deger, S.; Wust, P.; Loening, S. A.; Jordan,
A., Clinical hyperthermia of prostate cancer using magnetic
104
nanoparticles: Presentation of a new interstitial technique. International
Journal of Hyperthermia 2005, 21, (7), 637-647.
38. Hogemann, D.; Josephson, L.; Weissleder, R.; Basilion, J. P.,
Improvement of MRI probes to allow efficient detection of gene
expression. Bioconjugate Chemistry 2000, 11, (6), 941-946.
39. Mornet, S.; Vasseur, S.; Grasset, F.; Duguet, E., Magnetic
nanoparticle design for medical diagnosis and therapy. Journal of
Materials Chemistry 2004, 14, (14), 2161-2175.
40. Neuberger, T.; Schopf, B.; Hofmann, H.; Hofmann, M.; von
Rechenberg, B., Superparamagnetic nanoparticles for biomedical
applications: Possibilities and limitations of a new drug delivery
system. Journal of Magnetism and Magnetic Materials 2005, 293, (1),
483-496.
41. Bergemann, C.; Muller-Schulte, D.; Oster, J.; a Brassard, L.;
Lubbe, A. S., Magnetic ion-exchange nano- and microparticles for
medical, biochemical and molecular biological applications. Journal of
Magnetism and Magnetic Materials 1999, 194, (1-3), 45-52.
42. Baselt, D. R.; Lee, G. U.; Natesan, M.; Metzger, S. W.;
Sheehan, P. E.; Colton, R. J., A biosensor based on magnetoresistance
technology. Biosensors & Bioelectronics 1998, 13, (7-8), 731-739.
43. Ugelstad, J.; Berge, A.; Ellingsen, T.; Schmid, R.; Nilsen, T. N.;
Mork, P. C.; Stenstad, P.; Hornes, E.; Olsvik, O., Preparation and
105
Application of New Monosized Polymer Particles. Progress in Polymer
Science 1992, 17, (1), 87-161.
44. Chemla, Y. R.; Crossman, H. L.; Poon, Y.; McDermott, R.;
Stevens, R.; Alper, M. D.; Clarke, J., Ultrasensitive magnetic biosensor
for homogeneous immunoassay. Proceedings of the National Academy
of Sciences of the United States of America 2000, 97, (26), 14268-
14272.
45. Jordan, A.; Scholz, R.; Wust, P.; Fahling, H.; Felix, R.,
Magnetic fluid hyperthermia (MFH): Cancer treatment with AC
magnetic field induced excitation of biocompatible superparamagnetic
nanoparticles. Journal of Magnetism and Magnetic Materials 1999,
201, 413-419.
46. Pankhurst, Q. A.; Connolly, J.; Jones, S. K.; Dobson, J.,
Applications of magnetic nanoparticles in biomedicine. Journal of
Physics D-Applied Physics 2003, 36, (13), R167-R181.
47. Tong, W. J.; Gao, C. Y.; Mohwald, H., Manipulating the
properties of polyelectrolyte microcapsules by glutaraldehyde cross-
linking. Chemistry of Materials 2005, 17, (18), 4610-4616.
48. Lee, D.; Rubner, M. F.; Cohen, R. E., Formation of
nanoparticle-loaded microcapsules based on hydrogen-bonded
multilayers. Chemistry of Materials 2005, 17, (5), 1099-1105.
106
49. Malachowski, L.; Holcombe, J. A., Immobilized poly-L-
histidine for chelation of metal cations and metal oxyanions. Analytica
Chimica Acta 2003, 495, (1-2), 151-163.
50. McKenna, B. J., Birkedal, H., Bartl, M. H., Deming, T. J.,
Stucky, G. D. In Self-Assembling Microspheres from Charged
Functional Polyelectrolytes and Small-Molecule Counterions, Mater.
Res. Soc. Symp. Proc., 2004; 2004; pp 4.12.1-6.
51. McKenna, B. J.; Birkedal, H.; Bartl, M. H.; Deming, T. J.;
Stucky, G. D., Micrometer-sized spherical assemblies of polypeptides
and small molecules by acid-base chemistry. Angewandte Chemie-
International Edition 2004, 43, (42), 5652-5655.
52. Wang, S. B.; Xu, F. H.; He, H. S.; Weng, L. J., Novel alginate-
poly(L-histidine) microcapsules as drug carriers: In vitro protein release
and short term stability. Macromolecular Bioscience 2005, 5, (5), 408-
414.
53. Tourinho, F. A.; Franck, R.; Massart, R., Aqueous Ferrofluids
Based on Manganese and Cobalt Ferrites. Journal of Materials Science
1990, 25, (7), 3249-3254.
54. Deng, Y. H.; Yang, W. L.; Wang, C. C.; Fu, S. K., A novel
approach for preparation of thermoresponsive polymer magnetic
microspheres with core-shell structure. Advanced Materials 2003, 15,
(20), 1729-+.
107
55. Walt, D. R.; Agayn, V. I., The Chemistry of Enzyme and
Protein Immobilization with Glutaraldehyde. Trac-Trends in Analytical
Chemistry 1994, 13, (10), 425-430.
56. Atkins, R. C., Colorimetric Determination of Iron in Vitamin
Supplement Tablets - General Chemistry Experiment. Journal of
Chemical Education 1975, 52, (8), 550-550.
57. Chantrell, R. W.; Popplewell, J.; Charles, S. W., Measurements
of Particle-Size Distribution Parameters in Ferrofluids. Ieee
Transactions on Magnetics 1978, 14, (5), 975-977.
58. Brabers, V. A. M., Infrared Spectra of Cubic and Tetragonal
Manganese Ferrites. Physica Status Solidi 1969, 33, (2), 563-&.
59. Ahn, Y.; Choi, E. J.; Kim, S.; Ok, H. N., Magnetization and
Mossbauer study of cobalt ferrite particles from nanophase cobalt iron
carbonate. Materials Letters 2001, 50, (1), 47-52.
60. Murthy, V. S.; Cha, J. N.; Stucky, G. D.; Wong, M. S., Charge-
driven flocculation of poly(L-lysine)-gold nanoparticle assemblies
leading to hollow microspheres. Journal of the American Chemical
Society 2004, 126, (16), 5292-5299.
61. Spicer, P. T.; Pratsinis, S. E., Shear-induced flocculation: The
evolution of floc structure and the shape of the size distribution at
steady state. Water Research 1996, 30, (5), 1049-1056.
108
62. Maguire, M. E.; Cowan, J. A., Magnesium chemistry and
biochemistry. Biometals 2002, 15, (3), 203-210.
63. Toprak, M. S., McKenna, B. J., Waite, J. H., Stucky, G. D. In
Tailoring Magnetic Microspheres with Controlled Porosity, Mater.
Res. Soc. Symp. Proc., 2007; Materials Research Society: 2007; pp
0969-W03-11.
64. Ibarz, G.; Dahne, L.; Donath, E.; Mohwald, H., Controlled
permeability of polyelectrolyte capsules via defined annealing.
Chemistry of Materials 2002, 14, (10), 4059-4062.
65. Ibarz, G.; Dahne, L.; Donath, E.; Mohwald, H., Resealing of
polyelectrolyte capsules after core removal. Macromolecular Rapid
Communications 2002, 23, (8), 474-478.
66. Crank, J., Diffusion in a Sphere. In The Mathemetics of
Diffusion
Second ed.; Oxford University Press Inc.: New York, 1999; pp 89-103.
67. Toprak, M. S.; McKenna, B. J.; Mikhaylova, M.; Waite, J. H.;
Stucky, G. D., Spontaneous assembly of magnetic microspheres.
Advanced Materials 2007, 19, (10), 1362-+.
109
Abstract:
This chapter marks a transition in research focus from
coacervate-based, microspherical, “core-shell” devices to the
employment of coacervates for directing the higher-order growth of
inorganic structures. Using calcium carbonate as a model inorganic
component, coacervates were first used as templates to control the
mineralization into spheres. The formation of mineral shells further
expanded the framework for device design, and the formation of solid
spheres presented a facile route to an industrially important product.
However, coupled with interaction at substrate surfaces, the
mineralizing coacervates were discovered to guide rod-like growths or
cones. Growth of rods occurred using pre-existing coacervate funnels
with carbonate vapor infusion, whereas cones were produced by first
creating a condition of mutual inhibition by direct addition of carbonate
salt.
111
Thus far, two primary kinds of interaction have been
demonstrated between coacervates and inorganic material:1-6
1) Inorganic material can comprise one of the assembling components of coacervation (various nanoparticles).7, 8
2) Coacervates can behave as templates for the surface deposition of inorganic material (silica shells).
In further research, a third type of interaction was considered:
3) The use of the entire coacervate volume as a template for mineralization.
Using calcium carbonate mineralization as a model inorganic
component, the coacervates used this way will be shown to result in
three kinds of phenomena:
a) Solidification of mineral preferentially at the coacervate surface (as before)
b) Petrification of the entire internal coacervate volume into solid microspheres
c) Coalescence of solidifying spheres into larger structures, as controlled by forces in solution
d) Mineralization at the coacervate/substrate interface, whereby solidifying coacervates behave as conduits for the mineral components.
The interaction was found to depend primarily on three factors:
the solution pH (which controls both mineralization and coacervation),
the component concentrations, and the order of addition of the various
components.
Control over mineral morphology has both industrial and
fundamental importance. Industrially, calcium carbonate microspheres
112
are employed as fillers for paint and as basic whitening elements in
paper manufacturing. The properties of these materials are partly
determined by the qualities of the CaCO3 particles, most notably size
and shape. Therefore, there have been efforts to understand and control
the formation of spherical CaCO particles,9-183 or generally, to control
mineral morphology into a variety of complex structures.19 As will be
elaborated, this also has great importance for understanding
biomineralization, as morphological control of mineral is inextricably
linked to molecular, crystal controlling processes and emergent bulk,
structural organization.
Templating Results
The formation of CaCO3 shells was sought first, and this has
previously been achieved in different ways. For instance, Gower et al
report the use of their PILP process,20 combined with the action of an
O/W emulsion to produce oil-core microcapsules of CaCO .153 Xu et al
have reported the use of phytic acid in a similar vapor-induced process
to form shells of CaCO .123 However, these previous reports did not
consider the possibility of coacervation when in fact, both types of
precursor components can couple with Ca2+ at the proper pH to form
complex coacervates (Fig. 4.1). Phytic acid has a particularly sharp
113
transition region from dissolution to coacervation to flocculation,
occurring between pH 4 and 5.
Conditions for the production of CaCO3 shells were primarily
sought with PAA (MW 15k) and PLD (MW 33k), which have
sufficient molecular weights to associate efficiently with Ca2+. From
solutions of 10mM (monomer), optimal coacervation conditions were
sought as functions of pH and [Ca2+]. Values of pH at and above 7
were sufficient for coacervation with these molecular weights, and
between pH 9-10 the coacervates partially solidified, aggregated, and
sedimented. This may be due to enhanced charge interaction with
complete ionization of the polyanion, but is probably also due to partial
solidification and neutralization due to the formation of calcium
hydroxides in solution (Ksp = 5 x 10-6). A pH above the
bicarbonate/carbonate pKa of 8.3 is desirable, however, so that
(A) (B)
2+Figure 4.1. Optical micrographs of pH/Ca -induced coacervates of (A) PAA2kDa at pH 10 and (B) phytic acid at pH 4.5.
114
-mineralization occurs faster than HCO3 diffusion to the coacervate
interior. Therefore, pHs of 8.5-9.0 were chosen for initial coacervate
template solutions.
2+Selection of the proper [Ca ] was important for localizing
mineralization. At the lowest levels, coacervation is too weak to form
templates. Slightly elevating these conditions (as to charge neutrality,
[COO- 2+] = 2 [Ca ]) may yield some coacervates but the relatively low
[Ca2+] still impedes mineralization and especially hampers its
occurrence at the template surface. In contrast, the highest levels of
[Ca2+] unfavorably raise levels in the equilibrium solution, such that
mineral occurs non-specifically, away from coacervate surfaces.
Careful optimization was therefore required to mineralize specifically
and produce high yields of CaCO shells. 3
2+A better understanding of [Ca ] conditions that select for
CaCO3 shells was sought for more general application to different
polymer concentrations and different polymer types. The extent of
coacervation was investigated for this purpose by using turbidimetry.
Solutions of constant [polymer] and varying [Ca2+] were measured by
UV-vis spectrometry by adding incremental volumes of each
component to initial solutions containing only polymer, and the each
sample’s absorbance was recorded, using averaged values between 400-
500nm (Fig. 4.2). The resulting plots demonstrate several expected
115
trends. In the first leg, the low amount of scattering indicates that little
or no coacervation occurs, because too few Ca2+ cross-links are
available for polymers association. Next, there is a rapid “S-curve”-
like increase in turbidity, indicating a critical range for coacervation.
This reaches a maximum, indicating optimal coacervation for the given
polymer concentration. After this point, turbidity decreases because
extra CaCl2 salt increases the ionic strength of solution
0 20 40 60 80 100 1200.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
Abso
rban
ce
[Ca] (mM)
.8mMD 4mMD 9mMD
0 20 40 60 80 1000.0
0.5
1.0
1.5
2.0
2.5Ab
sorb
ance
[Ca2+] (mM)
(A) (B)
Figure 4.2. Turbidity curves of (A) PLD and (B) PAA with increasing Ca2+ concentrations.
For solutions of 5mM and 10mM monomer of both PAA and
PLD, it was found that the critical zone for the structure directing
capacity of coacervates occurred between the midway point of the S-
curve turbidity increase and the maximum turbidity point. In the case
of PLD, there was little variation in the optimal [Ca2+] for different
polymer concentration; maximum turbidity occurred with ~45mM Ca2+
at [D]’s of 0.8 mM, 3.6 mM, and [10]mM. For 10mM PAA, the
maximum of 10.5 mM Ca2+ was used, and the final procedure was
described in Chapter 1.
116
Finally, the carbonate solution was considered. NaHCO3 at pH
9-9.5 was used in concentrations and pHs that minimized the salt
inhibition effect. The maintenance of proper pH was also important
because CaCO3 mineralization releases H+, which inhibits coacervation
and mineralization. Hence, acidification was prevented by multiple
carbonate and NaOH additions. The final procedure was described in
Chapter 1. Micrographs are shown in figure 4.3.
The resulting shell material was characterized by XRD (Fig. 4a)
and TGA/DSC (Fig. 4.4b). Interestingly, the shells appear completely
(A) (B) (C)
2+Figure 4.3. PAA/Ca coacervates mineralized by carbonate addition. (A) LOM, (B) SEM, and (C) TEM indicating a less dense interior.
0 100 200 300 400 500
50
60
70
80
90
100
Temperature (deg. C)
%M
ass
-10
0
10
20
30
40
50
mW
/mg
10 20 30 40 50 60 70
50
100
150
200
250
300
350
Inte
nsity
2 Theta
(A) (B)
Figure 4.4. (A) XRD of hollow spheres indicating amorphous mineral. (B) TGA/DSC scan during calcination of hollow spheres.
117
amorphous. The organic component was determined to be 48%.
Calcined shells demonstrated hollow interiors (Fig. 4.5) and diffracted
as calcite, as expected.
(A) (B)
(C)
Figure 4.5. SEM images of calcium carbonate shells following calcination.
Internally mineralized coacervates could be prepared in several
ways. Regarding the above procedure for shell formation, it was
apparent from optical microscopy that the “shells” could extend to the
spheres’ centers when mineralization was too slow (data not shown).
In other words, petrified spheres could be produced with lower [Ca2+],
at somewhat lower pHs, or with less carbonate. Alternatively, solid
spheres could be obtained by adding carbonate first and calcium last
(Fig. 4.6). In a typical preparation, 20 uL of PLD (5 mg/mL) was
mixed with 3 μL NaHCO (20mM), and to this was added 7.5 μL CaCl3 2
(180 mM). Small volumes were used for rapid mixing. The primary
further obstacle in this procedure is the aggregation of microspheres as
they become charge neutralized and sediment. Preventing
118
sedimentation by intermittent redispersion can sufficiently maintain
segregated colloids.
25 μm
Figure 4.6. Internally mineralized spheres prepared by adding Ca2+ last.
Vapor Infusion Results
A peculiar demonstration of slow intra-coacervate
mineralization occurred when the ammonium carbonate vapor
technique was applied to pre-mixed coacervate systems.21 As with
shell formation, the resulting products varied as a function of [Ca2+],
but the morphological variation was more complex. Using constant
10mM PLD (33k MW), [Ca2+] of 30+mM resulted in microspheres of
varying sizes and aggregation degrees. Above 50mM [Ca2+], spheres
were large (2-5μm) and highly aggregated (Fig. 4.7A). Many of these
larger formations appear highly clustered at the solution/air interface
119
where pH is highest, as a side product of the mineralization that is often
unreported. Closer to 40mM, the spheres are smaller (0.5-2 μm) and
more distinct (Fig. 4.7B). Again, sedimentation increases aggregation,
but dispersion is prolonged if they are disturbed at least every hour.
(A) (B) (C)
10 μm 10 μm 30 μm
Figure 4.7. SEM images of spherical CaCO3 aggregates from during vapor-treated PLD coacervates using (A) 50mM Ca2+ and (B) 40mM Ca2+. (C) Worm-like shapes produced from coalescence of spheres using 20mM Ca2+.
2+However, at [Ca ] under 30mM, the mineralization rate was
slow enough for more complex morphologies to emerge (Fig. 4.6C). In
addition to the usual microspheres, segmented worm-like growths
appeared. Slightly lower [Ca2+] induced similar and presumably
mechanistically related, tapered spicule-like rods. Between 10-20mM
Ca2+, another, distinct “showerhead” morphology grew (Fig. 4.8). At
early growth stages, these appeared as broad “sunflowers”. All three of
these morphologies appeared to be nucleated by a larger sphere, which
in the case of the showerheads was a broader hemisphere.
120
(A) (B)
Figure 4.8. Sunflower (A & B) and showerhead (C & D) morphologies produced by vapor infusion of PLD coacervates using 15-20 mM Ca2+. Scale bars represent 25 μm.
Efforts were made to observe the growth mechanism of these
new structures. A crystallization cell was specially constructed so that
the mineralization process could be observed in the optical microscope.
These cells were placed vertically in a 50mL centrifuge tube containing
NH4HCO3 as a vapor source. The cell was removed at various times,
carefully turned horizontally, and observed under an optical microscope
to determine crystal growth orientation. Synthetic adjustments were
made to accommodate the new container, and it was confirmed that
showerhead structures grow in the expected direction, with the flat end
of the domes facing the substrate (Fig. 4.9). Observations of spicules
(C) (D)
121
were less conclusive, but they appeared to grow preferentially on the
glass, and roughly parallel to the surface.
Figure 4.9. Optical micrograph of vapor-induced growths viewed directly on their growth substrates.
Previous reports by Gower et al have demonstrated that
markedly similar systems produced rods by a Solution-Liquid-Solid
(SLS) mechanism akin to the VLS mechanism used in inorganic
syntheses such as fibers SiC. This is thought to occur by the action of
“Polymer-Induced Liquid Precipitates” (PILPs) that facilitate
mineralization at a growing surface.20 Because of the similarity of the
mineralization systems and the similarity of structures, it is suggested
that the spicule and showerhead morphologies form by a similar
mechanism. Preferential growth of showerheads at the dome periphery
is assumed to be due to increased surface tension at those locations;
alternatively, free carbonate from solution may be responsible for the
extensions, also explaining restriction to peripheral growth. The
occurrence of these structures has significance in determining the
122
nature of the PILPs, as these experiments demonstrate that the “liquid
precursors” may in fact be coacervates, but which contain levels of
carbonate/bicarbonate above the supersaturation of CaCO . 3
The sensitivity of the vapor infusion method was tested with
respect to several variables, including initial concentrations (Ca2+ and
polymer as described), ammonium carbonate powder mass, solution
height, and the extent of solution capping. All were found to have
significant effect on the final morphology, which reflects the
complexity of the system process (Scheme 4.1). To generalize, high
concentrations, large powder content, and less capping (greater vapor
accessibility) promoted conditions that favored more rapid,
homogeneous mineralization. And at their other extremes, these
variables could inhibit mineralization, so each required careful
engineering to achieve heterogeneous, coacervate-directed mineral
growth mode.
123
Temperature
pH
Ca
Salt
Mineral supersaturation
Coacervation
Gas solubility
Coacervate size
Surface Interaction
Scheme 4.1. Complexity of the vapor infusion crystallization process. Conflicting temperature effects are partially ameliorated by direct addition.
The vapor infusion process was
particularly sensitive to solution height,
which when either too low or too high
could cause homogeneous nucleation.
This suggests an even more complicated
process, which may be best explained by
the emergence of a pH gradient (Fig.
4.10, green arrow in Scheme 4.1), as
dictated by vapor infusion and solution
height. The higher pHs near the
air/water surface promote partial mineralization of spheres. As they
Figure 4.10. An ammonium bicarbonate solution containing thymol blue dye, demonstrating a pH gradient after 24h.
Location/ Morphology
Vapor
Solution height
Mixing order
CO32-
mass
Polymer
Substrate
124
sediment, they may solidify, dissolve, or remain liquid-like.
Appropriate pH gradients, then, initiate coacervate/substrate-mediated
growth. Too-high pHs (low solution height, rapid infusion), induce
rapid solidification. Too uniform pHs, or shallow gradients (large
solution height, slow diffusion) promote homogeneous mineralization
before substrate contact. In this model, appropriate conditions for
complex shape formation are defined by pH gradient that 1) exists over
a short enough distance for sedimentation to occur and 2) is sufficiently
steep for the acidity near the substrate to induce a liquid phase
separation.
A parallel study was conducted using PAA rather than PLD.
However, because the PAA/Ca2+ -7 (K ~10sp ) coacervates are less soluble
than those of PLD/Ca2+ -9 (K ~10sp ), as reflected in the turbidity curves,
lower levels of Ca2+ were required with PAA to decrease the
nonspecific mineralization. The results with PAA were markedly
different, however, resulting in strange “multipod” legs or the
production of umbrella-shaped “cones,” depending on various
experimental conditions (Fig. 4.11).
125
Figure 4.11. (A) Conical and (B) multipedal morphologies produced from vapor-infused coacervates using PAA15kDa rather than PLD.
The PAA cones were studied further with two modifications to
procedure: direct carbonate salt solution addition, and addition of
calcium last. The vapor infusion method was abandoned because
previously turbid coacervate solutions were found to be transparent
after several hours (12h) of infusion, and the effect could be replicated
by direct carbonate addition, if the carbonate concentration was
>150mM. In great contrast, lower levels of carbonate were found to
actually increase turbidity and induce flocculation. Furthermore, the
same conical shapes could be induced by this method with sufficiently
high carbonate concentration, of around 350-1000mM, which implies
that the vapor method also introduces these high concentrations. Ca2+
was added last to remove the possibility that either coacervates or
CaCO3 seeds would form first and critically affect crystallization
behavior. Both of these modifications do much to simplify the
dynamics of the crystallization system. Analysis of the number of
126
independent and dependent variables in the first situation reveals a
complicated framework with much uncertainty (Scheme 4.1). For
instance, temperature affects carbonate sublimation, carbon dioxide
solubility, and supersaturation. Using direct addition rather than
carbonate vapor provided much greater control of the process and
experimental reproducibility.
The loss of turbidity in mineralization solutions may at first be
surprising, because ordinarily solutions of coacervates are
(immediately) turbid and solutions of CaCO3 also become quite turbid
within ~1min. Therefore both mineralization and coacervation must be
inhibited under these conditions. The inhibitory power of
polyelectrolytes in preventing mineralization or “scaling” is well
known, however. And the eradication of coacervation is best described
as a result of the salt effect: NH HCO4 3 initially behaves as an inactive
salt of monovalent ions, which are known to strongly inhibit
coacervation. Only with CO2 escape and pH increase do bicarbonate
ions populate significantly enough to drive mineralization. Indeed, the
mineralization does not generally occur in solutions before a couple
hours, or after the pH appreciably increases with respect to the 8.3 pKa
of bicarbonate/carbonate. Direct mixing is a less commonly reported
method in the literature, except in cases where stopped flow is used to
quickly introduce higher supersaturation.22 This is probably because of
127
the assumption that such highly supersaturated solutions are too
unstable to achieve well controlled crystallization from an equilibrium
solution. The Kitano method is another way of direct mixing with
lower pHs; however, the procedure is more complex and is not usually
employed after the addition of additives.23 Therefore, the resulting
supersaturations and ionic strengths are significantly lower from those
using the standard direct addition employed here.
The results of direct addition are presented in the following
Chapter. Briefly, it was found that both PAA and PLD can induce the
same conical morphologies using direct addition. A systematic
exploration of chemical conditions, combined with direct addition,
resulted in highly reproducible morphological control. None of the
showerhead, sunflower, or spicule morphologies were observed,
however, which suggests that these uniquely depend on assembly by
pre-existing coacervate entities as in the PILP process. In contrast, the
cones appear to grow via a modular, interfacial growth process that
does not occur through a liquid funnel.
Conclusions
This work demonstrates the capability of complex coacervation
to control mineralization processes. By carefully controlling pH and
concentrations, conditions were found that preferentially mineralize
128
CaCO3 at coacervate surfaces or within coacervate volumes. The
mineral phase of coacervate shells was found to be amorphous and to
comprise about half of the capsule weight. Ammonium bicarbonate
vapor infusion was found to induce new, exotic shapes, which form at
the substrate surfaces at appropriate constituent concentrations. The
process was found to be highly dependent on a range of physical
parameters. The variation in shapes suggests a mechanism of
sedimentation and dynamic mineralization that is controlled by solution
pH gradients. Direct addition could replicate the synthesis of cones
with PAA by way of a controlled assembly process that begins with
mutual inhibition of mineral and coacervation. In contrast to cones
derived via direct carbonate addition, vapor infusion-induced
morphologies proceed via a PILP mechanism.
REFERENCES
1. Cha, J.N., et al., Microcavity lasing from block peptide
hierarchically assembled quantum dot spherical resonators.
Nano Letters, 2003. 3(7): p. 907-911.
2. Cha, J.N., et al., Spontaneous formation of nanoparticle vesicles
from homopolymer polyelectrolytes. Journal of the American
Chemical Society, 2003. 125(27): p. 8285-8289.
129
3. Wong, M.S., et al., Assembly of nanoparticles into hollow
spheres using block copolypeptides. Nano Letters, 2002. 2(6): p.
583-587.
4. McKenna, B.J., Birkedal, H., Bartl, M. H., Deming, T. J.,
Stucky, G. D. Self-Assembling Microspheres from Charged
Functional Polyelectrolytes and Small-Molecule Counterions.
in Mater. Res. Soc. Symp. Proc. 2004.
5. McKenna, B.J., et al., Micrometer-sized spherical assemblies of
polypeptides and small molecules by acid-base chemistry.
Angewandte Chemie-International Edition, 2004. 43(42): p.
5652-5655.
6. Murthy, V.S., et al., Charge-driven flocculation of poly(L-
lysine)-gold nanoparticle assemblies leading to hollow
microspheres. Journal of the American Chemical Society, 2004.
126(16): p. 5292-5299.
7. Toprak, M.S., McKenna, B. J., Waite, J. H., Stucky, G. D.
Tailoring Magnetic Microspheres with Controlled Porosity. in
Mater. Res. Soc. Symp. Proc. 2007: Materials Research Society.
8. Toprak, M.S., et al., Spontaneous assembly of magnetic
microspheres. Advanced Materials, 2007. 19(10): p. 1362-+.
9. Ajikumar, P.K., et al., Synthesis and characterization of
monodispersed spheres of amorphous calcium carbonate and
130
calcite spherules. Crystal Growth & Design, 2005. 5(3): p.
1129-1134.
10. Euliss, L.E., et al., Design of a doubly-hydrophilic block
copolypeptide that directs the formation of calcium carbonate
microspheres. Chemical Communications, 2004(15): p. 1736-
1737.
11. Wang, F., et al., A facile pathway to fabricate microcapsules by
in situ polyelectrolyte coacervation on poly(styrene sulfonate)-
doped CaCO3 particles. Journal of Materials Chemistry, 2007.
17(7): p. 670-676.
12. Xu, A.W., et al., Stable amorphous CaCO3 microparticles with
hollow spherical superstructures stabilized by phytic acid.
Advanced Materials, 2005. 17(18): p. 2217-2221.
13. Yu, J.G., M. Lei, and B. Cheng, Facile preparation of
monodispersed calcium carbonate spherical particles via a
simple precipitation reaction. Materials Chemistry and Physics,
2004. 88(1): p. 1-4.
14. Yu, J.G., et al., Facile fabrication and characterization of
hierarchically porous calcium carbonate microspheres.
Chemical Communications, 2004(21): p. 2414-2415.
131
15. Patel, V.M., et al. Synthesis of Calcium Carbonate-Coated
Emulsion Droplets for Drug Detoxification. in ACS Symposium.
2002. Orlando, FL: American Chemical Society.
16. Naka, K., Y. Tanaka, and Y. Chujo, Effect of anionic starburst
dendrimers on the crystallization of CaCO3 in aqueous
solution: Size control of spherical vaterite particles. Langmuir,
2002. 18(9): p. 3655-3658.
17. Colfen, H. and M. Antonietti, Crystal design of calcium
carbonate microparticles using double-hydrophilic block
copolymers. Langmuir, 1998. 14(3): p. 582-589.
18. Qi, L.M., J. Li, and J.M. Ma, Biomimetic morphogenesis of
calcium carbonate in mixed solutions of surfactants and double-
hydrophilic block copolymers. Advanced Materials, 2002.
14(4): p. 300-+.
19. Colfen, H., Precipitation of carbonates: recent progress in
controlled production of complex shapes. Current Opinion in
Colloid & Interface Science, 2003. 8(1): p. 23-31.
20. Gower, L.B. and D.J. Odom, Deposition of calcium carbonate
films by a polymer-induced liquid-precursor (PILP) process.
Journal of Crystal Growth, 2000. 210(4): p. 719-734.
132
21. Sugawara, A., T. Ishii, and T. Kato, Self-organized calcium
carbonate with regular surface-relief structures. Angewandte
Chemie-International Edition, 2003. 42(43): p. 5299-5303.
22. Chen, P.C., et al., Nucleation and morphology of barium
carbonate crystals in a semi-batch crystallizer. Journal of
Crystal Growth, 2001. 226(4): p. 458-472.
23. Kitano, Y., D.W. Hood, and K. Park, Pure Aragonite Synthesis.
Journal of Geophysical Research, 1962. 67(12): p. 4873-&.
133
Chapter V
Morphological Ternary Diagram Studies of
Non-classical Calcium Carbonate
Mineralization with Homopolyanions
*A related version of this chapter has been submitted as:
Brandon J. McKenna, J. Herbert Waite, Galen D. Stucky, “Biomimetic Control of Calcite Morphology with Homopolyanions,” submitted.
134
Abstract:
Biomineralization is an intricate process that relies on precise
physiological control of solution and interface properties. Despite
much research of the process, mechanistic details of biomineralization
are only beginning to be understood, and studies of additives seldom
investigate a wide space of chemical conditions in mineralizing
solutions. We present a ternary diagram-based method that globally
identifies the changing roles and effects of polymer additives in
mineralization. Simple polyanions were demonstrated to induce a great
variety of morphologies, each of which can be selectively and
reproducibly fabricated. This chemical and physical analysis also aided
in identifying conditions that selectively promote heterogeneous
nucleation and controlled cooperative assembly, manifested here in the
form of highly organized cones. Similar complex shapes of CaCO3
have previously been synthesized using double hydrophilic block
copolymers. We have found the biomimetic mineralization process to
be modular and interfacial, generating large mesocrystals with high
dependence on pH and substrate surface.
135
Introduction
Biomineralization is the controlled process of inorganic
assembly by organisms. The most familiar examples of biominerals
include bones, teeth, and seashells. Mineralized biomolecular
materials, as well as the processive strategies used to create them, are
sources of fascination to materials scientists. Particularly impressive is
the degree of exquisite morphological control with which organisms
manipulate minerals composed of multivalent ions. Despite their high
lattice energies, biology physically tunes them to suit a variety of
specific needs.1 Bone, abalone nacre, and the brittlestar skeleton are
preeminent examples of complex multifunctional materials, whose
properties are derived from precise control of morphology. The
resulting hierarchically ordered composite structures incorporate very
small fractions of organics and yet demonstrate greatly enhanced
strengths and toughnesses.2
Biominerals, and the controlling physiological processes by
which they form, have fundamental importance for general biological
understanding, geological interpretations, both of mineral distributions
in the earth’s crust and accurate analysis of the fossil record. With
regard to these latter motivational aspects, it is interesting that
biologically controlled mineralization appears to have begun relatively
136
late in the course of evolution, which may partly explain why scientists
have so far been able to understand genetic aspects like the Central
Dogma, but are still unable to accurately understand or mimic
biomineralization processes. The interpretations of mineralization
processes here primarily regard a fundamental understanding and the
hope of biomimicry.
The inspiration for biomimetic engineering of materials derives
from the variety of ways that nature is observed to manipulate inorganic
minerals for a variety of uses: optical focusing, navigation, gravity
sensing, and mechanical structure.3, 4 The lens of the extinct trilobite is
composed of calcite, a naturally birefringent material; yet the organism
evolved to overcome the double refraction effect by orienting the c-axis
perpendicular to the surface of the lens. Magnetobacteria assemble
nanoparticles of iron oxide for navigation according the earth’s
magnetic field. The otoliths that occur in the inner ears of vertebrates
act as weights that are coupled with cellular sensors to sense changes
from gravitational force; interestingly, while they are calcite in humans,
other species use alternate polymorphs of calcium carbonate or even
entirely different minerals. Siliceous sponge spicules are an example of
structural materials that are organized into elaborate 3d cage-like arrays
that resist forces of ocean currents and may also enhance light
harvesting. Their structure is also controlled at microscopic levels into
137
layered composites that better resist fracture propagation. Thus, nature
is able to mold a variety of minerals and endow them with otherwise
unnatural, enhanced properties, with control from nanoscopic to
macroscopic levels.
There is thus a drive to learn from various methods of biological
control and apply them to manmade materials. It is of practical concern
to harness such synthetic capability for the development of inexpensive,
strong, lightweight composite “biomimetic” materials (e.g. for
biocompatible replacement materials, as for bone repair with scaffolds
and/or mineral sources).5 The ordered arrays of the brittlestar lens and
butterfly wings could provide routes to photonic crystals. Spider silk
and mussel fibers could provide designs for new, tougher or self-
healing polymers. Mussel thread surface contacts and the glue of the
sandcastle worm could provide new principles for better underwater
adhesives. The structure of organic fibers like those in tree wood could
provide new strategies for reinforcement. Understanding the modes of
mineralization of extended single crystal calcite such as starfish arms
and sea urchin frameworks could impart new fabrication methods for
high purity crystalline materials, as required for high efficiency solar
cells. Finally, an understanding of composite structural materials could
provide methods for the synthesis of tougher and harder impact-
resistant, structural materials. In all cases, nature advantageously
138
directs material structure at various scales in order to effect material
function.
However, there is a prohibitive problem with attempting to
directly apply biological methods to inorganic/composite synthesis:
most biominerals are formed with an extremely high degree of control
by the complex machinery of cells. Without full access to the cellular
synthetic approach, scientists must find other methods of structural
mimicry. Cells interact with growing minerals in several different
ways. The coccolithophores, ancient but very prevalent algaes, exhibits
exquisitely assembled CaCO3 shells (the “coccolithes”) on the cell
surfaces. A high degree of intracellular control is involved: vesicles in
close relation to the Golgi apparatus are constructed with a
predetermined shape that controls both mineralization and transport to
the cell surface where the primary components lock together and are
ejected from the cell. The link between cells and minerals is more
apparent considering that biomineralization is thought to originate from
such species as a way to excrete mineral “waste product”, since Ca2+ is
desirable at only low levels within cytoplasm. The intracellular
mechanism is also effective in the spicules of sponge (silica) and coral
(calcite), for which entire specialized cells called sclerocytes are
responsible for encapsulating and constructing the mineral until near
completion.
139
Cells can also control biominerals extracellularly. One way to
do this is within folds of the cell membrane. Within the microcavities,
membrane pumps can modulate pH and concentrations. In bone, it is
well understood that the mineral component undergoes constant
dissolution/reconstruction by osteoclasts/osteoblasts, respectively. And
during the initial construction of bone, it has been observed that cells
cluster to form pockets of solution spatially isolated from the rest of the
organism. This way, cells can create supersaturated conditions within
the enclosed volume and precisely define new mineral. Because of
such complex control mechanisms, there is interest in examples of
extracellular biomineralization in which cells control diffusion of
inorganic material and biological molecules, but which otherwise do
not directly participate in the fine mineralization process. The
principles of these processes should translate to those of more cellular
mechanisms.
Much work towards understanding these principles has focused
on the action of the soluble polymers to determine how the crystal
growth mechanisms are altered in their presence. In classical
crystallization, supersaturation dictates nucleation, polymorph
selection, crystal growth rates, and surface energetics
(thermodynamics) coupled with kinetic mechanisms (kink-step growth)
dictate crystal growth direction and the prominently exposed crystal
140
faces. In contrast to this situation, various additives, particularly acidic
macromolecules, can exhibit profound effects and control over each of
these parameters in the crystallization mechanism. Such additives
include inorganic ions (e.g. Mg2+),6 organic molecules,7 LB
monolayers,8 SAMs,9 various synthetic polymers, and extracted
biopolymers.10, 11 Anionic macromolecules are of particular interest
biomimetically, because the proteins involved in CaCO3 mineralization
typically have low PIs and contain large fractions of phosphoserine,
aspartate or glutamate residues. Synthetic additives also provide model
systems for understanding non-classical crystal growth. The precise
role of acidic polymers remains controversial however, and several
equally plausible hypotheses have been proposed. As antiscalants, they
prevent mineralization by disturbing nucleation events;12, 13 in contrast,
as growth initiators they induce local supersaturation via ion
sequestration.14, 15 As habit modifiers, usually as part of a matrix, they
selectively initiate crystal growth of specific faces, often by providing
appropriate spatial periodicity.16, 17 They may also stabilize amorphous
calcium carbonate (ACC), sometimes prior to subsequent
mineralization.18
Other more complicated polyanion functions have been
suggested. In one mechanism, demonstrated by Gower et al., polymers
induce liquid precursor precipitate (PILP) microspheres, which
141
subsequently nucleate and direct crystal growth.19-21 In oriented
attachment, crystallite assembly is mediated by a partially capping
polymer.22, 23 Such precise, oriented construction has also been
suggested to yield mesocrystals of CaCO 24-273 as well as structured
crystals of BaSO .28, 294 Double hydrophilic block copolymers (DHBCs)
have been suggested as important for dynamic stabilization of growing
surfaces.30 Lastly, polyanions can also self-associate in the presence of
Ca2+ to yield complex coacervates, which can serve as microspherical
templates.31, 32
Although various polyanions have demonstrated all of the
above-mentioned roles to modulate CaCO3 morphology, in-depth
evaluation of the polymer’s roles under varying physical and chemical
conditions is often lacking. In many cases, it has been presumed that a
specific polymer performs a single important function (and hence
directs only one or few potential morphologies). Herein, we show that
even simple homopolyanions can induce many CaCO3 morphologies,
simply by altering experimental conditions, and in turn effectively
change the polymer’s function. The implementation of ternary phase
diagrams enables rapid screening of morphology over a wide range of
chemical conditions, revealing many possible roles for one polymer.
This facilitated the identification of regions of controlled assembly,
which are more representative of biological processes. The crystallinity
142
and polymer distribution of the various morphologies are also analyzed,
with particular regard to a highly oriented, layered structure.
Experimental Section
Polymer solutions. All polymers were dissolved in DI H2O and
diluted to 10 mM per monomer. Polyacrylic acids, sodium salts (PAA),
were received as follows: PAA 90 kDa (PAA90k) (Aldrich), PAA 15
kDa (PAA15k) (Aldrich), PAA 6 kDa (PAA6k) (Polysciences), and PAA
2 kDa (PAA2k) (Aldrich). Poly-L-aspartate (PLD) (30 kDa) and poly-
L-glutamate (PLE) (14 kDa) were received from Sigma and dissolved
in deionized H2O. Polystyrenesulfonic acid (PSS) (17 kDa, 18 wt.% in
H O) was received from Aldrich and neutralized with NaOH (Fisher). 2
Synthesis of fluorescein-labeled PAA: 6’-Aminofluorescein was
received from Fluka, 1-Hydroxybenzotriazole hydrate (HOBT) was
received from Aldrich, and 1-ethyl-3-(3-
dimethylaminopropyl)carbodiimide hydrochloride (EDC) was received
from Pierce.
To a 100 mM (per monomer) polyacrylic acid solution, 6’-
aminofluorescein was added to saturation (less than 5% per monomer),
and the pH was adjusted to 4.5 using 0.01 M HCl (EMD). A separate
solution, containing 2 equivalents of HOBt and 5 equivalents of EDC
(per 6’-aminofluorescein), was added to the first solution while
143
maintaining a pH of 4.5. Following 12 h of reaction, the product was
thrice dialyzed against deionized water for periods of 1d.
Mineral Syntheses: CaCl2 dihydrate (EMD Chemicals Inc.),
NH HCO , (NH )4 3 4 2CO , and NaHCO3 3 (Fisher) were used as received
and dissolved in DI H2O freshly prior to synthesis. Typically, 10.0 mM
(per monomer) of PAA, 1.00 M of a carbonate salt, and 10.00 mM
CaCl2 were micropipetted and mixed in 0.5 dram shell vials
(Fisherbrand, Type I Glass) to a total of 240 μL. The resulting
mineralization solutions were placed in a humidity chamber that was
temperature controlled at 30±1ºC for up to 24 h. Solution volumes
decreased by less than 1 μL per hour. For analyses requiring dry or
pure product, washing steps were performed by removing the
supernatant, twice rinsing with pH 9 NaOH solution, and finally rinsing
with deionized H O. 2
pH measurements: Measurements of pH were obtained of
various mineralizing solution over the course of 2 h, using an ORION*
Thermo Electron Micro pH Electrode.
Optical Microscopy: Samples were collected with pipet tips,
which were sometimes used to mechanically dislodge precipitates from
the container, and samples were pipetted onto a microscope slide and
observed with a Nikon Eclipse ME600 operating in transmission mode
using a 20x lens.
144
SEM: Washed samples were pipetted and dried on a silicon
wafer, and then sputtered with Pd/Au at 10 mA for 2 minutes. Samples
were loaded into an FEI Sirion Sphera operating at 5 kV.
XRD: Larger syntheses of the various products were screened
for morphological sameness with optical microscopy. Washed and
dried products were collected for analysis, and scanned between 20º
and 70º using a Philips XPERT Powder Diffractometer.
TEM: Washed products were suspended in water or absolute
ethanol, and dry cast onto a lacey carbon TEM grid. Samples were
analyzed with a T20 FEI Technai G2 Sphera Microscope operating at
200 kV, to record bright field images and diffraction patterns on
photographic plates.
Confocal Laser Scanning Microscopy (CLSM): A Leica
microscope equipped with an ArKr laser was used for performing laser
scanning confocal microscopy. A 0.5 mL volume of fluorescein-labeled
sample was loaded into a specially prepared glass slide containing a
well below a coverslip. Three-dimensional rendering was performed
with the VolumeJ package within ImageJ software.
Dynamic Light Scattering (DLS): Mineralizing solutions were
mixed to 2 mL, with component proportions corresponding to selected
morphologies. Samples were loaded into disposable sizing cuvettes
and placed in a Malvern Zetasizer Nano ZS. Measurement durations
145
were set to be determined automatically, and data were accumulated
over the course of 3h. The correlation function was processed using the
absorption, indices of refraction, and viscosity parameters
corresponding to PMMA and H O. 2
Results and Discussion
In order to analyze the morphological phase space of the major
chemical components, morphological maps were obtained in the form
of ternary diagrams. Solution sets of constant volume were prepared by
varying constituent proportions incrementally from stock solutions of
CaCl2, polymer, and carbonate salt. Proper stock solution
concentrations were determined empirically, such that they yielded a
nearly “optimized” ternary diagram—one which is spanned by many
morphologies and which also limits the variety of morphologies within
individual solutions. E.g., too concentrated solutions are too far beyond
equilibrium to controllably product distinct products, and too-dilute
solutions are too far below equilibrium to obtain product or reveal the
widest range of possible products. This concentration optimization
process also maximizes the sampling information obtained; fewer
points within the ternary diagram need to be queried to observe the
range of morphologies.
146
For the most detailed investigation (Fig. 5.1), PAA15k and
NH HCO4 3 were used. Components were mixed to a total of 240 μL,
using increments of 16 μL, in 0.5 dram flat-bottom glass vials, and the
solutions were incubated at 30ºC for 24h. Products formed as the pH
increased due to CO2 evolution, typically requiring a lag time of at least
20 min., at which point the pHs were 8.2; however, most growth
occurred at pHs of 8.5 and above. Over a dozen distinct morphological
regions were identified, and all products were either calcite or
amorphous. Each inset image represents a prominent or distinct
morphology for its region of phase space, and each composition and
morphology can be synthesized reproducibly and selectively, although
there is overlap across the guideline borders. Hybrid “twinned buds”
also appeared, between the “peanut” and “bud” regions (Fig. 5.2).
Ordinary calcite rhombohedra, which occur along the rightmost edge of
the diagram, are not shown. The diversity of shapes demonstrates that
the multiple functions displayed by this simple polyanion strongly
depend on solution conditions. It is apparent that the polymer is not
acting only as a template, a nucleator, an inhibitor, or a habit modifier,
but rather performs all these functions plus new, emergent behavior.
147
Figure 5.2. Twinned bud morphology.
The ternary diagrams in this study display some general trends.
High p
2+roportions of Ca associated rapidly with both polymer and
carbonate components, resulting in large, poorly defined aggregates.
At high [CO 2-3 ], crystallization was highly favored as evidenced by
increased faceting. High levels of polymer inhibited crystallization and
yielded little precipitate, particularly at low [Ca2+]. The absence of
polymer yielded the expected calcite rhombohedra. Low [CO 2-3 ]
products were entirely microspherical liquid charge neutralized
polymer-Ca2+complexes also known as coacervates. Increasing [Ca2+]
and [CO 2-3 ] from this coacervate region induced flocculation of
submicron precursors into larger sponge-like masses. This likely
resulted from the interconnection of polymer aggregates, induced by
mild electrostatic screening of the added salt, coupled with rapid
solidification of ACC.
149
2-With a further increase of [CO3 ], there was a morphological
transition from rapidly formed, interconnected masses into discrete
microspheres in low yield, reflecting slower nucleation and weakened
aggregation. Over extended periods, some of these exhibited PILP-like
growth into rods.21 2- With modestly higher [CO3 ], such microspheres
nucleated the growth of larger conical morphologies, first with rod-like
character and then with planar sides. At yet higher [CO 2-3 ], there was
an increasing preference for growth of the initial microsphere over cone
formation. This distinguished the fibrous cones from the buds, which
had larger, faceted spheres and shorter, less defined rod-like portions,
and which indeed occupied a distinct region of chemical space. The
microspheres showed an increasing tendency to twin with higher [CO 2-3
]. At the highest carbonate concentrations, cones and microspheres
ceased to form, in favor of smaller, discrete shapes with increasing
faceting. The rice, peanut, and hexagonal shapes decreased in size with
greater polymer concentration, supporting PAA’s role in nucleation.
Many of the morphologies reported here have been described
previously, for either CaCO3 or barium salts, but were formed by
employing more complex polymers, instead of altering solution
conditions. ACC in the form of flocs has been made previously, using
polycations;33 anionic PEO-b-PAA has been shown to stabilize ACC,
although as nanoparticulates rather than flocs.34 Domes have been
150
reported using polycations, although these were amorphous rather than
calcitic.35 Microspheres have been made, e.g., using PLD, although
these were characterized as vaterite instead of calcite.14 Cölfen et al
reported the use of PEG-b-PMAA to form peanuts (“dumbbells”),36
plotting part of a morphological map as a function of pH and the
[polymer]/[CaCO ] ratio.373 PEG-b-PEI-COOH has been used to make
BaSO4 peanuts. 38 Rice-like formations have been made using PAA39
or PEO-b-PMAA.40 PEG-b-PEI polymers with hydrophobic moieties
of different lengths have been used to fabricate microsphere-, peanut-,
quadruple-, hexagon-, and rice-like shapes.41
Morphologies that are similar to the bud and cone structures
have also been observed. “Petunias” were fabricated using
carboxymethyl chitosan,42 and “shuttlecocks” were made using PEG-b-
PGL and PEG(84)-b-PHEE with varying degrees of phosphorylation43;
these have been suggested to form around CO2 microbubble
templates.44 Cones were made with the combination of two different
PAA MWs, and their oriented architectures were attributed to
consecutive controlled growth via oriented attachment. Structures that
appear similar to the fibrous cones reported here have also been made
using PAA, but for BaCrO4 or BaSO4 rather than CaCO3. 24, 27, 45 This
study’s microspheres and the precursors observed at the tips of the
conical morphologies are likely related to the PILPs studied by Gower
151
et al because of the precursor proximity to extended growths. Although
Gower’s multi-domained films19 and rod-like structures21 are not
described in this study’s standard ternary diagram, they are readily
accessible using different conditions, such as different substrates or
different CO diffusion rates. 2
Moreover, direct comparison of our results with those of many
other studies is difficult, because different parameters such as these
(substrate and CO2 diffusion) can so radically alter morphology. For
instance, previous studies have used different or unquantifiable
methods of introducing CO 2-3 , different carbonate salts, and/or different
polymers. Furthermore, the chemical space has been less completely
determined for these other conditions and methods. From the results
obtained here, it would be of considerable interest and benefit to
conduct a more thorough analysis of morphological trends over
chemical space for these systems as well.
Variations.
In order to determine important factors for morphological
control, several experimental parameters were varied. Additional
morphological diagrams were constructed to test variations of:
temperature, ionic strength, carbonate sources, polymer type, and PAA
molecular weight. It was found that morphology was primarily
152
2-controlled by [CO3 ], as defined by pH, and by the extent of
polymer/Ca2+ association, particularly as functions of ionic strength and
polymer type. Increments of 24 μL (10.0%) of the stock solutions were
used to prepare solutions within the arrays to determine relative effects.
Temperature: Calcium carbonate’s decreased solubility at
higher temperatures was reflected in morphological changes in the
ternary diagram. Refrigerated samples (4º C) precipitated very little
product, with the exception of buds. Products at 20º C were similar to
the 30 ºC control products, except that cones formed at high [Ca2+],
where flocs otherwise occur; this is likely because of decreased
polymer association at this temperature. Low-[Ca2+] products (rice,
peanut shapes) were larger, reflecting lower nucleation rates. Samples
at 40 ºC also formed cones rather than flocs at high [Ca2+], and
additionally formed cones with high [PAA]; low [Ca2+] products
appeared smaller, however.
Molecular Weight: Because polymer molecular weight (MW) is
known to affect both coacervation (polymer association increases with
MW)31 and the antiscaling properties of PAA (with an optimum around
5 kDa)12, morphological variation with MW was anticipated. Ternary
diagrams constructed with different molecular weight PAAs, using
constant 10mM monomer, and indeed, despite sharing some overall
trends, the diagrams have notable differences (Fig. 5.3). PAA2k was
153
less potent, failing to inhibit precipitation at higher polymer
concentrations, compared to all the higher MW polymers. High [Ca2+]
did not readily yield flocs, instead promoting cones and rough domes,
or no product at all for low [CO 2-]. PAA3 6k solutions formed cones
under more low-carbonate conditions than with other MWs, suggesting
that this MW may be near optimal for controlled assembly. PAA90k
behaved contrarily, more readily forming flocs, and inducing cones
only at [PAA] 3 mM and below. Many shapes, such as peanuts and
rice, became rough and indistinct.
(A)
(B) (C)
Figure 5.3. Molecular weight variations (a) PAA6k, (b) PAA2k, (c) PAA90k
154
Polymer Type: Polyanions PSS, PLD and PLE were also tested,
and although they all clearly affected crystallization, only PLD yielded
similarly controlled products to those made with PAA (Figs. 5.4a-c).
PLD and PAA were also the only two of this set which self-associated
to form coacervates with Ca2+ at room temperature; this may imply the
importance of a minimal level of dynamic Ca2+ binding.
(B) (C) (A)
Figure 5.4. Representative optical micrographs of morphologies using (a) PLD, (b) PLE, and (c) PSS.
Method of carbonate addition: The present study used aqueous
carbonate salt stock solutions, which were mixed directly into
mineralizing solutions. Compared with other methods of introducing
carbonates, such as the sublimation/infiltration method, pre-dissolving
carbonate offers several advantages: the concentration is more easily
quantified, by a known initial amount; products are produced more
quickly, with a typical lag time of only 3 h instead of 8+ h; and results
are more reproducible because variables affecting the infiltration rate
155
are eliminated. Furthermore, infiltration using crushed ammonium
carbonate salts was found to produce some of the same morphologies,
indicating the methods operate similarly. This was elaborated in the
previous chapter. Because it only offered advantages for these studies,
direct addition was chosen.
Carbonate Source, and pH: All three carbonate sources were
capable of inducing the morphologies reported here, but required
different initial concentrations. The pH values of solutions with
ammonium salts remained buffered near bicarbonate’s pKa, typically
starting near pH 8.0 and ending around pH 9. In contrast, using sodium
salt encumbered controlled precipitation, because it induced more
dramatic pH shifts; however, fine tuning pH and carbonate content was
able to produce appropriate conditions, which verifies that ammonium
is not essential to the process (Fig. 5.5).
156
Figure 5.5. Fibrous cones fabricated using sodium bicarbonate.
Because an aqueous solution of 0.5 M (NH ) CO4 2 3 is equivalent
to a 1.0 M NH HCO4 3 solution with half of its carbonate content
converted to CO2(g), the two salts are directly comparable. In fact,
using such stock solutions yielded roughly equivalent diagrams, again
suggesting that the initial concentration of carbonates is less important
than adequate pH elevation.
Ionic Strength: The imbalance of carbonate compared to the
other solution species (roughly 100-fold higher) is particularly striking
because marine biomineralization occurs with reversed proportions, in
which carbonates total to around 3 mM. Since effects on pH change
little after about 100 mM, another explanation for such high
157
concentrations is needed. Therefore, the effect of increasing the ionic
strength with the carbonate precursors was tested.
It was found that flocs are immediately produced when low
levels of carbonate (<100 mM) are added to turbid solutions of complex
coacervates containing PAA and Ca2+. However, higher carbonate salt
content eliminates the turbidity by increasing the ionic strength to
inhibit coacervation, and in turn liberates solubilized polymer. Hence,
in such solutions, there exists a situation of dually cooperative
inhibition, in which an antiscaling polymer inhibits crystallization and
the salt content inhibits polymer association. Breaking this cycle
requires either heterogeneous nucleation by a high energy surface or a
pH increase, which is prominent in these syntheses.
Figure 5.6. Cones made with 125mM NH HCO and 0.5 M NaCl. 4 3
158
The effect of ionic strength was further tested by adding NaCl to
standard starting solutions, to a concentration of 0.5 M. Only products
with >4mM [Ca2+] were altered, and most strikingly, solutions that
otherwise would yield flocs instead produced a small number of fibrous
cones (Fig. 5.6). This reaffirms the importance of the ionic strength
increasing function of carbonate salt for producing complex
morphologies with anionic polymers.
Container/substrate type. Full diagrams based on different
containers were not obtained, because even the use of different glass
containers caused different morphologies (Fig. 5.7). This is probably
not only due to different qualities of glass, but instead largely due to
different physical effects. For instance, larger containers provide more
surface area for faster CO2 escape and therefore vastly different
spatiotemporal pH conditions. The standard glass vials that were used
also limited evaporation because their height created reflux conditions.
From the conditions explored, complex growth was found to occur on
silica, polystyrene, and polyethylene; only random aggregate growth or
flocs were observed on substrates of mica, aminopropylsiloxane-
modified glass, and the same glass again modified with terephthalic
acid. Perhaps charged surfaces disfavor controlled assembly either by
too strongly repelling precursor material or over-adsorbing solution
polymer. Growth on aragonite, or aragonite seeding, resulted only in
159
the appearance of few spheres on the mineral surfaces. No higher order
growth was observed, probably indicating polymorph specificity as
controlled by the polymer, which may instead dissolve aragonite
crystal.
(A) (B) (C)
Figure 5.7. Cones grown on different surfaces: (a) silicon wafer, (b) Petri dish, (c) polystyrene.
Additive Mg2+. Various levels of MgCl2 were tested in another
examination of morphogenesis across chemical space. The most
prominent effect was enhanced mineral inhibition. When added to
solutions from the standard ternary diagram, the products were most
often coacervates. Higher order crystalline morphologies were
obtained however, using double the amount of Ca2+ and only 10% the
amount of polymer, in the case of 5-20 mM MgCl2. This further
confirmed the mineral-inhibiting role of polymer. While control
samples without polymer yielded expected aragonite crystals, only
calcite products were found with added PAA or PLD, at the
concentrations tested. This affirms the calcite specific role of both of
these polymers.46, 47
160
Calcium phosphate. The ternary phase diagram approach was
applied to mixtures of polymer (PAA, PSS, PLD, PLE, alginate,
polyphosphoserine, and gelatin) at pHs of 7.5 and 5.5 and at 40˚C.
However, at all tested concentrations, the products were all amorphous
flocs, or mixtures of flocs with crystallites (Fig. 5.8). Adding extra
NaCl to disturb flocs only resulted in a sharper hydroxyapatite
diffraction peak (in the case of gelatin), but did not induce any higher
order assemblies as was observed for calcium carbonate. This may
reflect a few important differences between phosphate and carbonate
mineralization: 1) the soluble amorphous-phase inducing proteins for
hydroxyapatite may require more specific conformations and primary
structure, 2) in contrast to nacre-like, flat-surface mineralization, bone
is formed within predefined matrices of collagen, and 3) bone
formation involves more precise cellular involvement than, e.g. nacre.
It may be that calcium phosphates are unable to maintain disordered,
metastable structures for as prolonged times because of much lower
solubility products.
161
20 30 40 50 60 70
500
1000
1500
2000
2500
3000
35
40
00
00
Inte
nsity
2 Theta
Gelatin, High NaCl polyPhosphoserine
(A) (B)
Figure 5.8. (A) SEM images of gelatin/Ca-PO4 aggregates with minimal but poorly dispersed HAP mineral induced by increased ionic strength. (B) XRD patterns indicating predominantly amorphous material.
Characterization: All products were analyzed by LOM, SEM,
CLSM, and various diffraction methods. Conical morphologies
(umbrellas, buds, and fibrous cones) were of particular interest because
of their complexity and order, which are indicative a controlled
assembly process. The umbrella morphology is new and also exhibits
particularly interesting qualities, which are highlighted below.
The mechanism by which the mineral structures grow was
examined with real-time optical microscopy of solutions in Petri dishes
or in specially made reaction slides containing shallow wells. A typical
lag of >20 min preceded precipitation, except with coacervates and
flocs, which form sooner— often upon initial mixing. Peanut shapes
and other high-CO 2-3 morphologies formed while suspended in
solution, before eventual sedimentation and further growth on the
container surface.
162
In contrast to the other shapes, cones depend on directional
growth at the container surface, following nucleation. The first step is
the settling of small precipitates, which initiate growth at the substrate
surface. Conical shapes proceed to grow at their bases, where they
interface with the substrate, such that the cones point upwards, as also
observed via time-resolved optical microscopy. This provides direct
evidence of the growth mechanisms, and is in contrast to previously
proposed mechanisms for other similar conical or “flower-like”
shapes.44, 48 The growth direction was confirmed by SEMs of cones
grown directly on Si wafer substrates (Fig. 5.9). Although fibrous
cones also appeared to begin growth at their bases; most fell on their
sides and continue growing in this position. The preferential growth of
conical morphologies at the crystal/substrate surface suggests that the
growth mode is sensitive to surface energetics. In fact, cones grown on
different substrates were quite different in appearance, even when
different glass surfaces were used (Figs. 5.7a-c). Similar dependence
has been previously observed.48
163
Figure 5.9. SEM image of cones grown directly on a silicon wafer.
SEM images of umbrellas highlight the organization involved
during their growth process (Fig. 5.10a). The structures are evidently
modular, indicating growth by stepwise addition of solution precursors,
which appear nearly monodisperse despite the polydispersity in
polymer starting materials. Their periodicity is reminiscent of patterns
found in self-regulating systems, and could reflect local fluctuations in
pH or other chemical concentrations during formation. Fibrous
structures do not appear faceted and may be formed by similar
precursors which orient and smoothen, in a process that is akin to the
SLS mechanism put forth by Gower et al.21 Dynamic light scattering
measurements confirm that particles on the order of 200-300 nm are
present in mineralizing solutions for at least 5h (Fig. 5.11a).
Depending on [Ca2+], these particles appear at or before the bicarbonate
pKa of 8.3; if prevented from sedimentation by disturbing solutions,
164
they appear to grow with increasing pH and, as expected, acquire
increasingly negative surface charge (Fig. 5.11B). These precursor
“seeds” have been observed previously, and are thought to be
amorphous.18, 34, 49 They have been identified as the cause of granular
superstructures in non-classical crystallization processes.39
(A) (B)
(C) (D)
Figure 5.10. Umbrella fragments. (A) SEM, showing modular alignment. (B) TEM image, its diffraction pattern, and an image at an alternate angle revealing 2 connected planar sections. (C) Averaged projection confocal micrograph, showing striped pattern of polymer (scale bar 10μm). (D) Single plane confocal micrograph, showing a different, PILP-like growth and excluded polymer (scale bar 10μm).
165
7.8 8.0 8.2 8.4 8.6-100
0
100
200
300
400
500
600
700
800
900
7.7 7.8 7.9 8.0 8.1 8.2 8.3 8.4 8.5 8.6-40
-35
-30
-25
-20
-15
-10
10
11
12
00
00
00
Size
(nm
)
pH
Size E
-5
0
Zeta
Pot
entia
l (m
V)
pH
Low Ca High Ca
(B) (A)
Figure 5.11. (A) DLS average particle size and (B) Zeta potential of particles of two otherwise cone-producing mixtures, versus pH increase with time.
The crystallinity of products was determined from powder
XRD, single crystal XRD, IR spectroscopy, or TEM, depending on
product size, yield, and morphological purity. All crystalline products
were indexed to calcite, the thermodynamically most stable form of
calcium carbonate (Fig. 5.12). Of note are the flocs, which indicate one
means of achieving conditions that stabilize ACC. Microspheres
diffracted as calcite and are widespread across the diagram with a
gradual transition from flocs to microspheres appears gradual.
Umbrella fragments analyzed with TEM surprisingly revealed that the
full shards diffracted as single crystals, confirming the precise, oriented
arrangement directed by the growth process (Fig 5.10D). Rice shapes
were characterized by TEM analysis and appear as single crystals, but
are comprised of smaller, ~20nm grains (Fig. 5.13).
166
400 600 800 1000 1200 1400 1600 18000
20
40
60
80
100
Tran
smitt
ance
Wavelength (cm-1)
Umbrella Flocs
20 30 40 50 60 700
2000
4000
6000
8000
10000
Inte
nsity
2 Theta
(B) (A)
Figure 5.12. (A) Representative X-ray diffraction spectrum, showing only calcite diffraction peaks from cones. (B) Representative IR spectra of calcite cones and amorphous flocs.
(B) (A)
Figure 5.13. (A) TEM bright field, SADP, and (B) bright field close-up image of rice morphology.
Confocal Laser Scanning Microscopy (CLSM): In order to
determine the polymer distributions within various morphologies, a
fluorescently labeled polymer was used in the syntheses (a
morphological ternary diagram constructed with the labeled polymer
did not deviate from the control), and confocal microscopy was used to
gather cross sectional images; these images and their three-dimensional
167
reconstructions are available in the supplementary information. The
polymer was integrated throughout the structures despite their
crystalline properties, although some cross-sections revealed internal
regions from which the polymer was partially excluded.
The polymer distributions in the conical products demonstrated
some discontinuity, reflecting the organization seen in SEM images.
The nucleating tips found on many cones usually had the highest
concentration of polymer, followed by the longitudinal “legs.” In
agreement with structures observed with electron microscopy, umbrella
fragments had a periodic, striped distribution of polymer that defines
the border between horizontal layers of what must be crystalline (Fig
5.10B). A similar effect was evident in images of PILP-type growths,
which showed dark (crystalline) rods surrounded by a sheath of
excluded polymer (Fig. 5.10C). This patterning is likely due to micro-
phase separation, in which the crystallizing mineral partially excludes
polymer. A similar mechanism has been hypothesized for hierarchical
biomineral structures,50, 51 and importantly represents a new
methodology of achieving periodic, layered composite materials via
cooperative organization on the photonic length scale.
168
The combined results of the ternary diagrams and the various
characterization procedures suggest mechanisms similar to those in
Scheme 5.1. Peanuts might also form according to previously
described rod-to-dumbbell mechanism,51 but the minor presence of
similarly sized single spheres and the existence of the quadruple
morphology both indicate that twinning may occur. As previously
suggested, mineralization originates from precursor aggregates that
arise from metastable solutions following an appropriate pH increase.
The exact composition of these aggregates remains unknown, but they
are probably distinct from PAA/Ca2+ coacervates, since they occur in a
different region of the ternary diagram, and because centrifugations of
solutions throughout mineralization yielded only solid, not liquid,
precipitates. Following their formation, solution conditions (such as
pH, chemical concentrations and physical parameters) dictate whether
these aggregate, crystallize, or assemble interfacially.
Umbrellas are one example of a morphology that forms
interfacially, such that the precursor modules successively position
themselves into crystallographic alignment at the base. While oriented
attachment cannot be ruled out from these experiments, it is reasonable
that in these aqueous media of high ionic strength, the precursors are
first amorphous and then crystallize once they have fused with the
growing structure. Supporting this hypothesis was the observation that
170
a solution of amorphous flocs slowly transformed into several large
cones (Fig. 5.14). Following fusion to the growing structure,
crystallization induces micro-phase separation, partially excluding the
polymer into layers.
(A) (C)
(B) (D)
Figure 5.14. Photographs of large fluorescent cones, under UV (a) and ambient light (b); and still single crystal x-ray diffraction (c) of a whole 2mm cone, pictured in (d).
Conclusions
Polyanionic additives are capable of inducing a multitude of
morphologies in crystallizing systems, thereby exhibiting various
functions in the process. Polymers behave as antiscalants, ACC
stabilizers, nucleation sites, and inducers of precursor assemblies—
often demonstrating multiple functions at once.52 Exploring the phase
171
space of chemical components, using ternary diagrams, is useful for
exploring such diversity. With such diagrams, one can rationally study
conditions that promote specific shapes, that selectively promote ACC,
or that induce controlled assembly processes. We have thereby
demonstrated fundamental connections among previous studies, having
shown that all such properties can be achieved by the same polymer.
This is of particular benefit to a field that has lacked standard synthetic
methods for analyzing the effects of solution additives. The
methodology may provide a deeper understanding of biomineralization
and indicate new methods for synthesizing high performance composite
materials. Given the potential range of possible growth modes, proper
characterization of both synthetic and biological polymers will require
testing across chemical space.
The various morphologies were shown to be dependent on
several critical parameters, and each can be controlled to reproducibly
select mechanisms of crystal growth. Low temperature inhibits most
assemblies, polymer molecular weight affects the inhibition potency
and the degree of association, and polyanions that form complex
coacervates with Ca2+ appear to have a more dramatic role in directing
crystal assembly. Carbonate salt content affects the pH, both by
buffering and by inducing pH increase; pH, in turn, acts as a synthetic
“switch” to assembly—a common method in biological systems.
172
Importantly, carbonate salts also control ionic strength, which is
especially critical for suppressing polymer association. Finally, the
growth mode is highly sensitive to substrate properties.
This systematic investigation has revealed a controlled,
interfacial, modular assembly process involving organic/inorganic
phase separation, manifested in the form of highly organized umbrella
morphologies with periodic polymer/calcite layers of polymer and
crystallographic alignment. Since the PAA and PLD used in these
studies are simple and polydisperse, the process is quite general, and
such a mechanism should also be accessible with more complex and
well defined polymers, as those used in biomineralization. The
amorphous-crystalline transition method is common to many biological
processes. Moreover, the resulting structures share several properties
with nacre: alternating organic/inorganic layers, high crystallographic
alignment with registry between layers, and, as recently characterized,
composition of smaller interconnected crystallites into mesocrystals.53-
55 Despite obvious differences between the two materials, this
controlled growth mechanism may still come to explain common
features, and further research has yet to reveal the effects of different
substrates for more complex additive systems. Regardless of its
biological relevance, this controlled assembly represents a novel, easily
reproducible method for arranging synthetic materials into periodic
173
structures, which may be engineered to attain desirable mechanical or
optical properties.
REFERENCES
1. Sanchez, C., H. Arribart, and M.M.G. Guille, Biomimetism and
bioinspiration as tools for the design of innovative materials
and systems. Nature Materials, 2005. 4(4): p. 277-288.
2. Smith, B.L., et al., Molecular mechanistic origin of the
toughness of natural adhesives, fibres and composites. Nature,
1999. 399(6738): p. 761-763.
3. Mann, S., Biomineralization: Principles and Concepts in
Bioinorganic Materials Chemistry. 2001: Oxford University
Press, Inc. 198.
4. Simkiss, K. and K.M. Wilbur, Biomineralization: Cell Biology
and Mineral Deposition. 1989, San Diego, California 92101:
Academic Press, Inc. 337.
5. Hartgerink, J.D., E. Beniash, and S.I. Stupp, Self-assembly and
mineralization of peptide-amphiphile nanofibers. Science, 2001.
294(5547): p. 1684-1688.
174
6. Han, Y.J., et al., Template-dependent morphogenesis of oriented
calcite crystals in the presence of magnesium ions. Angewandte
Chemie-International Edition, 2005. 44(16): p. 2386-2390.
7. Fu, G., et al., Acceleration of calcite kinetics by abalone nacre
proteins. Advanced Materials, 2005. 17(22): p. 2678-+.
8. Xu, G.F., et al., Biomimetic synthesis of macroscopic-scale
calcium carbonate thin films. Evidence for a multistep assembly
process. Journal of the American Chemical Society, 1998.
120(46): p. 11977-11985.
9. Han, Y.J. and J. Aizenberg, Face-selective nucleation of calcite
on self-assembled monolayers of alkanethiols: Effect of the
parity of the alkyl chain. Angewandte Chemie-International
Edition, 2003. 42(31): p. 3668-3670.
10. Belcher, A.M., et al., Control of crystal phase switching and
orientation by soluble mollusc-shell proteins. Nature, 1996.
381(6577): p. 56-58.
11. Levi, Y., et al., Control over aragonite crystal nucleation and
growth: An in vitro study of biomineralization. Chemistry-a
European Journal, 1998. 4(3): p. 389-396.
12. Loy, J.E., J.H. Guo, and S.J. Severtson, Role of adsorption
fractionation in determining the CaCO3 scale inhibition
175
performance of polydisperse sodium polyacrylate. Industrial &
Engineering Chemistry Research, 2004. 43(8): p. 1882-1887.
13. Reddy, M.M. and A.R. Hoch, Calcite crystal growth rate
inhibition by polycarboxylic acids. Journal of Colloid and
Interface Science, 2001. 235(2): p. 365-370.
14. Roque, J., et al., Crystal size distributions of induced calcium
carbonate crystals in polyaspartic acid and Mytilus edulis
acidic organic proteins aqueous solutions. Journal of Crystal
Growth, 2004. 262(1-4): p. 543-553.
15. Tsortos, A. and G.H. Nancollas, The role of polycarboxylic
acids in calcium phosphate mineralization. Journal of Colloid
and Interface Science, 2002. 250(1): p. 159-167.
16. Addadi, L., et al., A Chemical-Model for the Cooperation of
Sulfates and Carboxylates in Calcite Crystal Nucleation -
Relevance to Biomineralization. Proceedings of the National
Academy of Sciences of the United States of America, 1987.
84(9): p. 2732-2736.
17. Addadi, L. and S. Weiner, Interactions between Acidic Proteins
and Crystals - Stereochemical Requirements in
Biomineralization. Proceedings of the National Academy of
Sciences of the United States of America, 1985. 82(12): p.
4110-4114.
176
18. Faatz, M., F. Grohn, and G. Wegner, Amorphous calcium
carbonate: Synthesis and potential intermediate in
biomineralization. Advanced Materials, 2004. 16(12): p. 996-+.
19. Gower, L.B. and D.J. Odom, Deposition of calcium carbonate
films by a polymer-induced liquid-precursor (PILP) process.
Journal of Crystal Growth, 2000. 210(4): p. 719-734.
20. Olszta, M.J., E.P. Douglas, and L.B. Gower, Scanning electron
microscopic analysis of the mineralization of type I collagen via
a polymer-induced liquid-precursor (PILP) process. Calcified
Tissue International, 2003. 72(5): p. 583-591.
21. Olszta, M.J., et al., Nanofibrous calcite synthesized via a
solution-precursor-solid mechanism. Chemistry of Materials,
2004. 16(12): p. 2355-2362.
22. Navrotsky, A., Energetic clues to pathways to
biomineralization: Precursors, clusters, and nanoparticles.
Proceedings of the National Academy of Sciences of the United
States of America, 2004. 101(33): p. 12096-12101.
23. Niederberger, M. and H. Colfen, Oriented attachment and
mesocrystals: Non-classical crystallization mechanisms based
on nanoparticle assembly. Physical Chemistry Chemical
Physics, 2006. 8(28): p. 3271-3287.
177
24. Kulak, A.N., et al., Continuous structural evolution of calcium
carbonate particles: A unifying model of copolymer-mediated
crystallization. Journal of the American Chemical Society,
2007. 129(12): p. 3729-3736.
25. Miura, T., et al., Emergence of acute morphologies consisting of
iso-oriented calcite nanobricks in a binary poly(acrylic acid)
system. Crystal Growth & Design, 2006. 6(2): p. 612-615.
26. Wang, T.P., M. Antonietti, and H. Colfen, Calcite mesocrystals:
"Morphing" crystals by a polyelectrolyte. Chemistry-a European
Journal, 2006. 12(22): p. 5722-5730.
27. Xu, A.W., et al., Uniform hexagonal plates of vaterite CaCO3
mesocrystals formed by biomimetic mineralization. Advanced
Functional Materials, 2006. 16(7): p. 903-908.
28. Wang, T.X. and H. Colfen, In situ investigation of complex
BaSO4 fiber generation in the presence of sodium polyacrylate.
1. Kinetics and solution analysis. Langmuir, 2006. 22(21): p.
8975-8985.
29. Wang, T.X., A. Reinecke, and H. Colfen, In situ investigation of
complex BaSO4 fiber generation in the presence of sodium
polyacrylate. 2. Crystallization mechanisms. Langmuir, 2006.
22(21): p. 8986-8994.
178
30. Yu, S.H. and H. Colfen, Bio-inspired crystal morphogenesis by
hydrophilic polymers. Journal of Materials Chemistry, 2004.
14(14): p. 2124-2147.
31. Bungenberg de Jong, H.G., in Crystallisation- coacervation-
flocculation in colloid science, H.R. Ed. Kruyt, Editor. 1949,
Elsevier: Amsterdam. p. 232-258.
32. Delacruz, M.O., et al., Precipitation of Highly-Charged
Polyelectrolyte Solutions in the Presence of Multivalent Salts.
Journal of Chemical Physics, 1995. 103(13): p. 5781-5791.
33. Ono, H. and Y.L. Deng, Flocculation and retention of
precipitated calcium carbonate by cationic polymeric
microparticle flocculants. Journal of Colloid and Interface
Science, 1997. 188(1): p. 183-192.
34. Guillemet, B., et al., Nanosized amorphous calcium carbonate
stabilized by poly(ethylene oxide)-b-poly(acrylic acid) block
copolymers. Langmuir, 2006. 22(4): p. 1875-1879.
35. Xu, X.R., J.T. Han, and K.L. Cho, Deposition of amorphous
calcium carbonate hemispheres on substrates. Langmuir, 2005.
21(11): p. 4801-4804.
36. Cölfen, H.Q., L., Progr. Colloid Polym. Sci., 2001. 117: p. 200-
203.
179
37. Colfen, H. and L.M. Qi, A systematic examination of the
morphogenesis of calcium carbonate in the presence of a
double-hydrophilic block copolymer. Chemistry-a European
Journal, 2001. 7(1): p. 106-116.
38. Qi, L.M., H. Colfen, and M. Antonietti, Control of barite
morphology by double-hydrophilic block copolymers.
Chemistry of Materials, 2000. 12(8): p. 2392-2403.
39. Donnet, M., et al., Use of seeds to control precipitation of
calcium carbonate and determination of seed nature. Langmuir,
2005. 21(1): p. 100-108.
40. Marentette, J.M., et al., Crystallization of CaCO3 in the
presence of PEO-block-PMAA copolymers. Advanced
Materials, 1997. 9(8): p. 647-651.
41. Sedlak, M. and H. Colfen, Synthesis of double-hydrophilic block
copolymers with hydrophobic moieties for the controlled
crystallization of minerals. Macromolecular Chemistry and
Physics, 2001. 202(4): p. 587-597.
42. Liang, P., et al., Petunia-shaped superstructures of CaCO3
aggregates modulated by modified chitosan. Langmuir, 2004.
20(24): p. 10444-10448.
43. Rudloff, J., et al., Double-hydrophilic block copolymers with
monophosphate ester moieties as crystal growth modifiers of
180
CaCO3. Macromolecular Chemistry and Physics, 2002. 203(4):
p. 627-635.
44. Rudloff, J. and H. Colfen, Superstructures of temporarily
stabilized nanocrystalline CaCO3 particles: Morphological
control via water surface tension variation. Langmuir, 2004.
20(3): p. 991-996.
45. Yu, S.H., et al., Growth and self-assembly of BaCrO4 and
BaSO4 nanofibers toward hierarchical and repetitive
superstructures by polymer-controlled mineralization reactions.
Nano Letters, 2003. 3(3): p. 379-382.
46. Wierzbicki, A., et al., Atomic-Force Microscopy and Molecular
Modeling of Protein and Peptide Binding to Calcite. Calcified
Tissue International, 1994. 54(2): p. 133-141.
47. Zhang, S.K. and K.E. Gonsalves, Synthesis of Calcium
Carbonate-Chitosan Composites Via Biomimetic Processing.
Journal of Applied Polymer Science, 1995. 56(6): p. 687-695.
48. Yu, S.H., H. Colfen, and M. Antonietti, Control of the
morphogenesis of barium chromate by using double-hydrophilic
block copolymers (DHBCs) as crystal growth modifiers.
Chemistry-a European Journal, 2002. 8(13): p. 2937-2945.
49. Bolze, J., et al., Time-resolved SAXS study of the effect of a
double hydrophilic block-copolymer on the formation of CaCO3
181
from a supersaturated salt solution. Journal of Colloid and
Interface Science, 2004. 277(1): p. 84-94.
50. Sumper, M., A phase separation model for the nanopatterning
of diatom biosilica. Science, 2002. 295(5564): p. 2430-2433.
51. Simon, P., U. Schwarz, and R. Kniep, Hierarchical architecture
and real structure in a biomimetic nano-composite of
fluorapatite with gelatine: a model system for steps in dentino-
and osteogenesis? Journal of Materials Chemistry, 2005.
15(47): p. 4992-4996.
52. Furuichi, K., Y. Oaki, and H. Imai, Preparation of nanotextured
and nanoribrous hydroxyapatite through dicalcium phosphate
with gelatin. Chemistry of Materials, 2006. 18(1): p. 229-234.
53. Oaki, Y., et al., Bridged nanocrystals in biominerals and their
biomimetics: Classical yet modern crystal growth on the
nanoscale. Advanced Functional Materials, 2006. 16(12): p.
1633-1639.
54. Rousseau, M., et al., Multiscale structure of sheet nacre.
Biomaterials, 2005. 26(31): p. 6254-6262.
55. Takahashi, K., et al., Highly oriented aragonite nanocrystal-
biopolymer composites in an aragonite brick of the nacreous
layer of Pinctada fucata. Chemical Communications, 2004(8):
p. 996-997.
182
Abstract:
Nacre is a much studied biomineral composite with a lamellar,
brick-wall architecture. The mechanism of nacre formation remains
elusive, despite increasingly detailed research into nacre’s molecular
and crystallographic architecture. The basic and detailed features of
nacre are reviewed, and the material is compared and contrasted with
the umbrella morphology presented in the previous chapter. An
analogous assembly mechanism is proposed which accords with recent
detailed observations of nacre, and which also explains previously
confusing but basic aspects of nacre formation. Remaining questions
and directions for further research are presented.
184
Of the structural biominerals, perhaps the most intriguing
structure is nacre (or “mother-of-pearl”), known for its pearlescent
luster and brick-wall-like architectures.1, 2 Just one of several other
naturally occurring layers in the shells of mollusks, the nacreous layers
of Gastropoda and Bivalvia have attracted scientific attention for many
reasons. Firstly, the pearlescent lustre, as particularly observed in
abalone shells, has been a phenomenon of fascination and of mimetic
interest for construction of periodic structures on the photonic length
scale. Secondly, its brick-wall-like structure is reminiscent of the man-
made version and appears spatially optimized against forces normal to
the surface.
Thirdly, the detailed organizational characteristics highlight
many of the principles that are involved in biomineralization, at large.
For instance, it is a composite with features of down to several
nanometers, the crystal polymorph and habit are both specifically
controlled by organics, the crystal orientation and morphology in turn
control the overall material structure, and the polymer matrix (between
0.5 and 5 wt. %) that is responsible for the 3000x toughness
enhancement of nacre over ordinary aragonite.3 Fourthly, nacre grows
in films, a structure that at once appears more tenable than other 3d
complex architectures in nature and has immediately recognizable
185
modes of application. That mollusks can control various planar growth
modes is encouraging for mimetic pursuits.
And fifthly, in contrast to a great variety of other biominerals,
nacre is formed extracellularly. The shells, as of abalone, are formed
by excretion of components from the organism’s mantle into the
extrapallial fluid. This fluid and the action of the biomolecules it
contains direct the construction of the shell layers by selecting crystal
polymorph, crystal morphology, and the distribution of
macromolecules—naturally all highly interrelated. Therefore, although
the process remains complex, the mode of nacre synthesis removes the
tremendously daunting tasks of mimicking the cellular machinery such
as the Golgi apparatus and attaining such precision of the dynamic
chemical parameters in highly confined spaces. Instead, what are
needed to understand nacre formation are: 1) an understanding of
diffusional processes in a “bulk” medium, which chemists are capable
of, and 2) knowledge of the biological elements, such as protein
sequence and other precursor macromolecules. However, there is a
third requirement to understand the interactions of biomineral
precursors for given conditions of chemical diffusion. These include
nucleation and growth of inorganic minerals, the supramolecular
association of biomacromoleules, chemical crosslinking, and
interactions of the interaction of biomolecules with inorganic material.
186
Nacre is a layered, periodic material with alternating mostly-
organic and mostly-inorganic layers (lamellae). The inorganic layers
consist mostly of planar discs (tablets, slabs) of aragonite oriented with
their (c-axis) [001] planes expressed along the surfaces, and the discs
grow laterally to achieve full coverage. The crystal orientations of
slabs within the same plane are not correlated, but crystallinity is
conserved vertically between adjacent slabs.4-6 Either side of the discs
is in intimate contact with water-soluble proteins, which form the outer
parts of the organic layers. Between these soluble elements are sheets
of an insoluble chitinous material similar to silk fibroin. The growth of
nacre differs between species, for instance in hexagonality of growing
slabs or slab thickness, but there is a clear distinction between the
growth modes in Bivalvia and Gastropoda. The former prefers lateral
growth, finishing one layer before constructing the next, while the latter
prefers normal/axial/vertical growth, in which layers grow laterally
only subsequent to vertical propagation into conical ‘stack of coins’
structures that have been likened to Christmas trees or cairns.
More detailed observations of nacre’s organization and structure
have been revealed by a vast amount of research.7 The chitin-like
polysaccharide is also arranged periodically in a less common parallel-
strand conformation, and may be in some register with the crystal. The
soluble, silk-like proteins have been found not strictly confined to the
187
edges of the aragonitic layers, but rather incorporated more dispersed
throughout these inorganic layers.8 The slabs have furthermore been
found to be composed of small, crystallographically oriented granules
rather than being a continuous mineral.9 The surfaces of the crystalline
portions have been reported to be coated with ~5nm of amorphous
material.10 Finally, the demineralized organic matrix appears
perforated with small channels or pores through which aragonite
mineral bridges are able to propagate crystallinity with preserved
orientation from layer to layer.4, 11 This latter detail marked an
important paradigm shift from a growth model that previously
described organic components with perfect epitaxial alignment and
transmittance of crystal orientation to successive layers.
Despite the occurrence of mineral bridging, the organic content
of nacre nevertheless has extraordinary control over crystallinity.12-14 In
particular, the soluble matrix proteins can specify the calcium carbonate
polymorph in crystallizing solutions. Belcher et al demonstrated that
proteins derived from the organic matrix may spur nucleation of
aragonite in solution conditions that would otherwise favor calcite,
even in the absence of Mg2+.12 Furthermore, these proteins are able to
modulate polymorph selection, causing aragonite to grow on the
surface of a calcite substrate. Finally, there was some evidence that
188
these proteins were sufficient to induce the formation of small layered
structures, as would otherwise be found in nacre.
Despite the cumulative knowledge of nacre structure, the
precise mechanism of formation is still lacking, so some basic questions
remain (Fig. 6.1). For instance, regardless of whether crystallization
occurs over or within matrix sheets, what determines layer thickness
and causes it to be so uniform across layers? That is, why do matrix
materials not prematurely bind to growing crystalline layers? If
organics control the termination of aragonite growth, what prevents
premature termination or overgrowth and ‘blanketing’ of organic layers
over unfinished aragonite tablets? What causes the horizontally-
preferred growth of bivalves versus the vertically-preferred growth of
gastropods? How can one protein induce a layered structure? Why are
the slabs of aragonite composed of granules, and how is polymer
occluded throughout the mineral? Are matrix sheet perforations
programmed or incidental? If the matrix layers are laid down
successively over inorganic layers, why/how do they not blanket over
entire unfinished mineral layers? And finally, how do the different
classes select for lateral or perpendicular/vertical growth modes?
189
Figure 6.1. Modes of growth that are not observed in nacre formation, and which still lack a mechanistic explanation.
(A) (B) (C)
Herein, I suggest that some of the recent observations of
nonclassical crystallization can unify into a general phase separation
mechanism that begins to explain the broader features of nacre as well
as some of the unanswered questions outlined above.
Progress Towards a New Model
The observed umbrella morphologies, as identified through the
systematic ternary diagram approach in the previous chapter, serve as
an embodiment of the generalized mechanism. It should be
emphasized, however, that the umbrellas are very different from nacre
in several important ways. Most notably, the umbrellas are composed
of calcite rather than aragonite. Secondly, they appear to grow where
the crystal contacts the substrate, conferring a conical rather than
lamellar structure. Thirdly, they lack an insoluble matrix and
reinforcing proteins. Finally, there is no initial habit-selecting organic
layer. Therefore, the umbrella obviously fails, in some aspects, as a
model system for nacre formation.
190
However, the umbrellas have several commonalities with nacre,
and hence the two structures may share mechanistic growth aspects.
Firstly, they are both made from calcium carbonate, and therefore
involve similar constraints in solubility, pH, etc. Secondly, they both
contain alternating layers of organic and inorganic material. Thirdly,
the successive layers of each are crystallographically oriented such that
the structures diffract as single crystals. This transfer of crystallinity is
also visible macroscopically, such that the modules appear periodically
both within and across layers; FFTs of images confirm such patterning
(Fig. 6.2). Fourthly, it is also likely that polymer is occluded in small
portions throughout the inorganic layers, and that the inorganic layers
are composed of smaller granules. Both of these features are more
easily observed in the smaller rice morphologies that were reported, as
the umbrellas are often too thick to observe granules, and polymer in
the inorganic layers can not be resolved between the 250nm spacings,
due to photonic limitations to spatial resolution.
(C) (A) (B)
2 μm
Figure 6.2. (A) Umbrella microstructure displays higher order morphological control by crystal structure, evinced by (B) an enhanced Fourier transform and (C) the inverse Fourier transform.
191
Therefore, the structural features of umbrellas at least
demonstrate the kinds of complexity feasibly achievable by a simple
polymer, and therefore it is possible that the simpler system’s formation
mechanism generally applies more broadly to other nonclassical
crystallization systems involving polyelectrolytes, including nacre.
What then is the mechanism of umbrella formation?
There are two broad classes of mechanisms for periodic
structures of this sort: those in which localized chemical oscillations
alternatingly select assembly modes, and those based on building
blocks of well defined size. In the first kind of mechanism, an increase
in pH would cause crystallization of one layer of calcium carbonate,
then creating a localized concentration of H+, perhaps leading to a halt
in crystallization and temporarily preferred adsorption of polymer to the
crystal. The bulk solution and perhaps some of the polymer would act
as proton sinks, in turn raising the pH and inducing the next layer of
crystal onto an incompletely covered previous layer. However, such
periodic pattern formation has ordinarily only been observed when
diffusion is much slower,15 and the modular appearance of various
SEM images suggests the latter mechanism.
Other evidence indicates that the cones grow by the assembly of
discrete packets, rather than by a continuous but periodic growth
process. Consider the other various micromorphological variations of
192
different umbrellas (Fig. 6.3). Some of the structures appear smooth,
with sharp edges reminiscent of nacreous structure (Fig. 6.3A). But
slight variations in concentrations (currently there is no known trend)
induce features which appear derived from modular growth modes. In
some cases, modules appear disorganized, yet they clearly form sharp
steps (Fig. 6.3B). Other cases, as mentioned in the previous chapter,
appear composed of more highly aligned modules (Fig. 6.2). Finally,
some structures strongly resemble PILP rods, having bulbous heads and
demonstrating polymer/crystal phase separation (Fig. 6.3C). As
described by Gower et al, such rods grow by the action of a liquid
droplet that funnels solution precursor material to the crystal growth
site, preserving crystallographic alignment.
Figure 6.3. Various umbrella microstructures. (A) Sharp and lamellar, (B) Particulate, (C) Intermediate, PILP-like. Scale bars 10μm.
(A) (B) (C)
The modular process is supported by DLS data of
PAA/Ca2+ 2-/CO3 solutions: while lower pH solutions contain 5nm
scattering entities, solutions around the pH where much of the growth
occurs (~8.3-8.5) contain “precursors,” which are roughly the same size
as the observed layer spacings (200-300nm) (Fig. 6.4). Given the
193
structure of the umbrellas and the correlation with precursor size, the
most reasonable mechanism is a modular one. Therefore, it becomes
important to understand the nature of these precursors. What is their
composition? How hydrated are they? What state of matter are they?
If they are liquid, what is the diffusion coefficient inside of them? How
do such properties change with parameters like pH and temperature?
What kinds of polymers are able to induce them, and how do different
polymers change their properties? What emergent properties do they
have, as in their abilities to control crystallization?
7.8 8.0 8.2 8.4 8.6-100
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
Siz
e (n
m)
pH
Size E
0
10
20
30
40
50
1 10 100 1000
Volu
me
(%)
Size (d.nm)
Size Distribution by Volume
(A) (B)
7.8 8.18.2 8.35
8.3 8.4
Figure 6.4. (A) Umbrella precursor sizes versus pH. (B) Size distributions by volume at various pHs. If interfacial growth is prevented, by keeping precursors dispersed as in these experiments, the precursors continue to grow and precipitate in indistinct aggregates.
Precursor mechanisms are not new in the field of calcium
carbonate mineralization. The above-mentioned PILP process is one
example. Another proposed mechanism for certain precursors is
oriented attachment, in which small crystallites contain polymer that is
preferentially adsorbed onto one (or more) specific crystal faces. The
crystallites are free to diffuse and eventually orient themselves to one
194
another, perhaps under the control of their local electrostatic forces, and
then ‘sinter’ once they are in alignment. Such a process certainly
occurs in some systems (for instance TiO nanoparticles162 ); however,
the model does not describe the probability of alignment or the effect of
crystal defects on two adjoining crystals, especially as would occur on
the scale of 100-500 nm. Furthermore, Cölfen has also studied the
crystallinity of precursors and found that they are amorphous for an
extended period.17 Other attempts to study the precursors, as by TEM
or filtration, have revealed that they are inherently unstable and
transient.
In fact, a sensible mechanism is suggested by considering the
umbrella precursors as metastable amorphous modules that proceed to
construct highly oriented structures. Such an idea is also in agreement
with reports that biology controls the formation of amorphous calcium
carbonate, either transiently or in storage, for use as precursor material
to intricate crystalline structures.10, 18-21 It is furthermore in agreement
with the suggestion that, prior to crystallization, polymers exist as
hydrated gels rather than as crystalline substrates.22
In this proposed mechanism, the polymer is responsible for
sequestering amorphous nuclei into discrete packages, increasing local
supersaturation, and locally inhibiting the drive crystallization. The
precursors are free to diffuse until they contact an exposed crystal face,
195
upon which this new substrate induces oriented crystallization within
the precursor, similar to the epitaxial process described by the PILP
SLS mechanism. As crystallization proceeds, the precursor phase
separates, excluding many of the polymers to the exterior (Scheme 1).
However, some of the polymers naturally become occluded and
interrupt the crystallization process such that it occurs in smaller
“granular” components. Meanwhile, both occluded polymers and
surface segregated polymers become oriented according to the crystal
lattice, adopting a structure and orientation to conform within it.
Polymers that are unable to attain a proper conformation for a given
polymorph may inhibit the propagation of the crystal and instead begin
to change the crystal structure, as it accords with lattice matching.
(B) (A) (C) Scheme 6.1. Proposed mechanism of nacre formation, based on precursors and phase-separation.
In this way, the kinetics of the polymer and crystallization
cooperatively control one another. Crystallization proceeds to the other
side of the precursor in this fashion, and the excluded polymer then
either binds thermodynamically to certain surfaces or else is released
back into solution (explaining the excess of polymer observed in
196
umbrella solutions). The process then results in a new exposed crystal
face for the next precursor to bind. Because umbrella layers are
aligned, it can be inferred that adsorbed polymers do not provide
complete surface coverage, and that the incidentally exposed crystal
surface edges can behave as “bridges” to propagate crystal orientation
to the next layer.
As described, this interfacial mechanism can describe several
details of the composite structures: preserved crystal orientation,
periodic/lattice alignment of soluble matrix polymers with crystal
interfaces, occluded polymers, and granular crystallites. It also
explains other important observations (Fig. 6.1). For instance, the
periodic nature and crystal orientation of such structures can be seen as
a natural consequence of the growth, as controlled by polymer
adsorption and precursor attachment. The monodispersity layer
thicknesses is the result of two-fold control: by the size of the
precursors and by the size of the first, initiating crystal face of each
layer. Additionally, of great importance is the emergence of the layered
structure by phase separation. This mechanism describes a method of
step-by-step growth of an organic matrix as occurring in concert with
crystal growth, explaining a method of achieving inorganic/organic
layering without explicit sequential control.
197
This suggested model combines several pertinent observations:
the amorphous nature of previously observed precursors, the use of
transient ACC in nature prior to complex crystallization, the
observation of liquid (and therefore amorphous) PILPs, and the
proximity of the conical phases to amorphous ones in the ternary
diagram. The phase separation/amorphous precursor mechanism also
echoes some basic biochemical principles, for instance the preservation
of a high energy and metastable state that organizes in response to a
bistable switch. Based on such principles, phase separation in
biomineralization has been previously suggested by Sumper in an effort
to explain the patterning achieved by diatoms.23 Similar to the model
proposed here, a mix of polyelectrolyte precursors cause a phase
separation or organic and inorganic material, followed by successive
splittings as surface tension is modulated. However, in contrast to the
mechanism proposed here, the diatom model is based on patterning on
organic, coacervate-like spherical templates, whereas the CaCO3 model
requires no template, per se, and organizes in layers according to crystal
orientation.
198
Remaining Questions
Because precursors are induced by certain ordinary
polyelectrolytes, it is likely that they occur during nacre formation with
at least one of the known soluble proteins. What remains uncertain is
whether the formations participate in the process theorized here, or
whether such a process dominates biomineralization. If nacre does
form according to this suggested mechanism, there are of course further
questions:
1. Can these precursors be observed in the extrapallial fluid? In vitro?
2. How do their sizes and compositions correlate with the final product?
3. How do the insoluble matrix proteins participate in the growth process? Sequentially? Prepackaged as precursor stabilizers?
4. Are nacreous precursors stable indefinitely, or do they solidify (less reversibly) as a function of time/size/pH?
5. What specific polymer attributes (e.g., range of binding energies) are required to initiate such a process?
6. Do multiple polymers affect the process?
7. Do precursors have a measurable or calculable surface tension, and can such properties predict interaction with various biological or synthetic templates?
Of course, these questions extend beyond nacre to other
biomineralization systems, as well. As observed by the morphogenesis
with the ternary phase diagram approach, precursors need not
199
crystallize into lamellar or even period structures. The longitudinal
growth of the fibrous cones, and the rods formed in the similar PILP
process, represent one way for precursors to create extended structures
(such as spicules) in a very different way than would be found in
classical crystallization. The precursors may also be endowed with
properties that create favorable interaction with biological templates in
3d structures that are less restricted to orientational growth but still
preserve crystallographic orientation. Such patterning around 3d
objects has been observed for CaCO3 systems without complex
additives,24 but may be enhanced in biology by the influence of organic
additives.
Conclusion
I have described a conceptual model for a mineralization mode
that explains features of the umbrella structure and possibly of nacre.
The model explains fundamentally unanswered questions about the
layered construction of nacre, such as the deposition of matrix layers
and uniform periodicity. It furthermore implies answers to more
detailed observations, such as epitaxial adsorption, mineral bridging,
and granular crystallites. Because it more simply describes natural
features, accounts for various experimental in vitro observations, and
leaves fewer critical unanswered questions, it is an altogether better
200
model to scientifically investigate nacre formation, according to
Occam’s razor and logical positivism. Finally, this proposal yields well
defined scientific questions for future investigations—an often lacking
attribute of the biomineralization field—and hence favorably lends
itself to testing and scientific progress.
REFERENCES
1. Mann, S., Biomineralization: Principles and Concepts in
Bioinorganic Materials Chemistry. Oxford University Press, Inc.: 2001;
p 198.
2. Simkiss, K.; Wilbur, K. M., Biomineralization: Cell Biology
and Mineral Deposition. Academic Press, Inc.: San Diego, California
92101, 1989; p 337.
3. Currey, J. D., Mechanical-Properties of Mother of Pearl in
Tension. Proceedings of the Royal Society of London Series B-
Biological Sciences 1977, 196, (1125), 443-&.
4. Feng, Q. L.; Li, H. B.; Cui, F. Z.; Li, H. D.; Kim, T. N., Crystal
orientation domains found in the single lamina in nacre of the Mytilus
edulis shell. Journal of Materials Science Letters 1999, 18, (19), 1547-
1549.
201
5. Metzler, R. A.; Abrecht, M.; Olabisi, R. M.; Ariosa, D.;
Johnson, C. J.; Frazer, B. H.; Coppersmith, S. N.; Gilbert, P.,
Architecture of columnar nacre, and implications for its formation
mechanism. Physical Review Letters 2007, 98, (26).
6. Checa, A. G.; Rodriguez-Navarro, A. B., Self-organisation of
nacre in the shells of Pterioida (Bivalvia : Mollusca). Biomaterials
2005, 26, (9), 1071-1079.
7. Addadi, L.; Joester, D.; Nudelman, F.; Weiner, S., Mollusk shell
formation: A source of new concepts for understanding
biomineralization processes. Chemistry-a European Journal 2006, 12,
(4), 981-987.
8. Nudelman, F.; Gotliv, B. A.; Addadi, L.; Weiner, S., Mollusk
shell formation: Mapping the distribution of organic matrix components
underlying a single aragonitic tablet in nacre. Journal of Structural
Biology 2006, 153, (2), 176-187.
9. Li, X. D.; Xu, Z. H.; Wang, R. Z., In situ observation of
nanograin rotation and deformation in nacre. Nano Letters 2006, 6,
(10), 2301-2304.
10. Nassif, N.; Pinna, N.; Gehrke, N.; Antonietti, M.; Jager, C.;
Colfen, H., Amorphous layer around aragonite platelets in nacre.
Proceedings of the National Academy of Sciences of the United States
of America 2005, 102, (36), 12653-12655.
202
11. Schaffer, T. E.; IonescuZanetti, C.; Proksch, R.; Fritz, M.;
Walters, D. A.; Almqvist, N.; Zaremba, C. M.; Belcher, A. M.; Smith,
B. L.; Stucky, G. D.; Morse, D. E.; Hansma, P. K., Does abalone nacre
form by heteroepitaxial nucleation or by growth through mineral
bridges? Chemistry of Materials 1997, 9, (8), 1731-1740.
12. Belcher, A. M.; Wu, X. H.; Christensen, R. J.; Hansma, P. K.;
Stucky, G. D.; Morse, D. E., Control of crystal phase switching and
orientation by soluble mollusc-shell proteins. Nature 1996, 381, (6577),
56-58.
13. Checa, A. G.; Jimenez-Lopez, C.; Rodriguez-Navarro, A.;
Machado, J. P., Precipitation of aragonite by calcitic bivalves in Mg-
enriched marine waters. Marine Biology 2007, 150, (5), 819-827.
14. Fu, G.; Qiu, S. R.; Orme, C. A.; Morse, D. E.; De Yoreo, J. J.,
Acceleration of calcite kinetics by abalone nacre proteins. Advanced
Materials 2005, 17, (22), 2678-+.
15. Imai, H.; Tatara, S.; Furuichi, K.; Oaki, Y., Formation of
calcium phosphate having a hierarchically laminated architecture
through periodic precipitation in organic gel. Chemical
Communications 2003, (15), 1952-1953.
16. Penn, R. L.; Banfield, J. F., Morphology development and
crystal growth in nanocrystalline aggregates under hydrothermal
203
conditions: Insights from titania. Geochimica Et Cosmochimica Acta
1999, 63, (10), 1549-1557.
17. Bolze, J.; Pontoni, D.; Ballauff, M.; Narayanan, T.; Colfen, H.,
Time-resolved SAXS study of the effect of a double hydrophilic block-
copolymer on the formation of CaCO3 from a supersaturated salt
solution. Journal of Colloid and Interface Science 2004, 277, (1), 84-
94.
18. Amos, F. F.; Sharbaugh, D. M.; Talham, D. R.; Gower, L. B.;
Fricke, M.; Volkmer, D., Formation of single-crystalline aragonite
tablets/films via an amorphous precursor. Langmuir 2007, 23, (4),
1988-1994.
19. Gehrke, N.; Nassif, N.; Pinna, N.; Antonietti, M.; Gupta, H. S.;
Colfen, H., Retrosynthesis of nacre via amorphous precursor particles.
Chemistry of Materials 2005, 17, (26), 6514-6516.
20. Politi, Y.; Arad, T.; Klein, E.; Weiner, S.; Addadi, L., Sea
urchin spine calcite forms via a transient amorphous calcium carbonate
phase. Science 2004, 306, (5699), 1161-1164.
21. Beniash, E.; Aizenberg, J.; Addadi, L.; Weiner, S., Amorphous
calcium carbonate transforms into calcite during sea urchin larval
spicule growth. Proceedings of the Royal Society of London Series B-
Biological Sciences 1997, 264, (1380), 461-465.
204
22. Levi-Kalisman, Y.; Falini, G.; Addadi, L.; Weiner, S., Structure
of the nacreous organic matrix of a bivalve mollusk shell examined in
the hydrated state using Cryo-TEM. Journal of Structural Biology
2001, 135, (1), 8-17.
23. Sumper, M., A phase separation model for the nanopatterning of
diatom biosilica. Science 2002, 295, (5564), 2430-2433.
24. Wucher, B.; Yue, W. B.; Kulak, A. N.; Meldrum, F. C.,
Designer crystals: Single crystals with complex morphologies.
Chemistry of Materials 2007, 19, (5), 1111-1119.
205