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

Polyelectrolytes in the Synthesis of Composite Materials

Copyright © 2007

by

Brandon J. McKenna

iii

This dissertation is dedicated to my parents:

Anne-Marie and John J. McKenna.

iv

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

Chapter I

Complex Coacervation for the Synthesis of

New Materials

1

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

Chapter IV

Calcium Carbonate Mineralization

via Complex Coacervation

110

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

148

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

169

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

Chapter VI

Towards a Model for Nacre Formation

183

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