Aqueous and Organic Microgels for Soft Matter Manufacturing

AQUEOUS AND ORGANIC MICROGELS FOR SOFT MATTER MANUFACTURING

CHRISTOPHER S. O’BRYAN

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2019

To my family

ACKNOWLEDGMENTS

I would like to thank my advisor, Dr. Thomas E. Angelini, for all his guidance, support and discussion throughout my time at the University of Florida. I would like to thank my colleagues, lab-mates, and friends; especially Tapomoy Bhattacharjee, Christopher Kabb,

Cameron Morley, Tori Ellison, and Tristan Hormel for their helpful discussion and hard work.

Finally, I would like to thank my family, Peter, Susan, John, and Billy, for supporting me both inside and outside of the lab.

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES ...... 8

LIST OF FIGURES ...... 9

LIST OF ABBREVIATIONS ...... 11

ABSTRACT ...... 12

CHAPTER

1 SACRIFICIAL SUPPORT MATERIALS FOR SOFT MATTER MANUFACTURING ....13

Introduction ...... 13 Rheological Characterization of Support Materials ...... 14 Linear Elastic Solids and Viscous Fluids ...... 14 Maxwell Model ...... 15 Kelvin Voight Model ...... 17 3D Printing Instabilities in Sacrificial Support Materials ...... 18 Gravitational Instabilities ...... 18 Interfacial Instabilities ...... 19 Inertial Instabilities ...... 20 Crevasse Formation ...... 20 Sacrificial Support Material 3D Printing ...... 21 Density Matched Support Baths ...... 22 Packed Micelles ...... 22 Networks with Reversible Bonds ...... 23 Colloidal Clays ...... 23 Jammed Granular Microgels ...... 24

2 JAMMED GRANULAR MICROGELS AS SACRIFICIAL SUPPORT MATERIALS ...... 31

Introduction ...... 31 Results and Discussion ...... 32 Conclusions ...... 35

3 ORGANIC MICROGELS FOR SILICONE 3D PRINTING ...... 40

Introduction ...... 40 Results and Discussion ...... 41 Micro-Organogel Synthesis and Characterization ...... 41 Temperature Effects on Rheological Properties ...... 44

Precision Printing into Micro-Organogels ...... 44 Mechanical Robustness of 3D printed silicone structures ...... 46 Application of Micro-Organogels for 3D Silicone Printing ...... 47 Conclusions ...... 48

4 INTERFACIAL INSTABILITIES IN JAMMED MICROGELS ...... 58

Introduction ...... 58 Results and Discussion ...... 59 Rheological Characterization ...... 59 Interfacial Instabilities of Fluid Structures ...... 60 Interfacial Yielding of Soft Solids ...... 61 Interfacial Buckling of Soft Solids ...... 63 Conclusions ...... 66

5 POLYELECTROLYTE INTERACTIONS IN AQUEOUS MICROGELS ...... 75

Introduction ...... 75 Results and Discussion ...... 77 Polyelectrolyte Microgel Synthesis and Preparation ...... 77 Rheological Characterization of Polyelectrolyte Microgels ...... 77 Polyelectrolyte Scaling Laws ...... 80 Rheological Characterization of Uncharged Microgel ...... 83 Cell Viability in Polyelectrolyte Microgels ...... 84 Cell Proliferation in Polyelectrolyte Microgels ...... 85 Cell Metabolic Activity in Polyelectrolyte Microgels ...... 85 Conclusions ...... 87

APPENDIX

A PERFUSION AND FLOW THROUGH MICROGELS ...... 99

Introduction ...... 99 Theory ...... 100 Results and Discussion ...... 101 Characterizing size of microgel particles ...... 101 Permeability Measurements ...... 101 Conclusions ...... 103

B MATERIAL AND METHODS ...... 107

Organic Microgels for Silicone 3D Printing ...... 107 Preparing organogel support matrix ...... 107 Preparing silicone elastomer inks ...... 107 Silicone 3D printing ...... 108 Rheology ...... 108 Small-angle x-ray scattering ...... 109 Interfacial tension ...... 109

Gel permeation chromatography ...... 109 Interfacial Instabilities in Jammed Microgels ...... 110 Rheological Characterization ...... 110 3D Printing ...... 110 Image Acquisition and Data Analysis ...... 110 Polyelectrolyte Microgels for 3D Cell Culture ...... 110 Microgel Synthesis and Preparation ...... 110 Rheological Characterization ...... 112 Cell Culture and Short-Term Viability ...... 112 Cell Metabolic Activity ...... 113 PEG Microgel Synthesis ...... 114

LIST OF REFERENCES ...... 115

BIOGRAPHICAL SKETCH ...... 127

LIST OF TABLES

Table page

5-1 Experimental values for the synthesis of 10g of polyelectrolyte microgels composed of polyacrylamide with anionic (MAA), zwitterionic (CBMA), and cationic (qDMAEMA) comonomers...... 98

LIST OF FIGURES

Figure page

1-1 Soft Matter 3D Printing into Sacrificial Support Materials...... 26

1-2 Rheological Behavior of Elastic Solids and Viscous Fluids...... 26

1-3 The Maxwell model...... 27

1-4 The Kelvin-Voight model...... 27

1-5 Gravitational Instabilities in Sacrificial Support Materials...... 28

1-6 Inertial Instabilities in Sacrificial Support Materials...... 28

1-7 Static Crevasse Formation in Sacrificial Support Materials...... 29

1-8 Rheological characterization of F127 micelle samples...... 29

1-9 Rheology of Networks with Reversible Bonds...... 30

2-1 Rheological characterization of Chemically Distinct LLS Materials...... 37

2-2 Short-term Cell Viability Studies of Cells Dispersed in Microgels...... 38

2-3 Macroscopic and Confocal Imaging of 3D-Printed Tubes...... 39

3-1 Block copolymer self-assembly into micro-organogels...... 51

3-2 Rheological Characterization of Micro-Organogels...... 52

3-3 SAXS Characterization of Organogels...... 52

3-4 Effects of Elevated Temperature on Rheological Properties...... 53

3-5 Printing Precision of Silicone Elastomers into Micro Organogels...... 54

3-6 Layer to Layer Adhesion of Silicone Elastomer Structures...... 55

3-7 Surfaces and Mechanical Properties of Printed Structures...... 56

3-8 3D-Printed Silicone Structure in Micro-Organogels...... 57

4-1 Rheological Characterization of Interfacial Systems...... 68

4-2 Interfacial Instabilities of Neat Mineral Oil...... 69

4-3 Interfacial Instabilities of Jammed Organic Microgels...... 70

4-4 Interfacial Instabilities in Packed Micelle Beams...... 71

4-5 Buckling Instabilities in Aqueous Microgel beams...... 72

4-6 Wrinkling vs Buckling Instability in Microgel Beams...... 73

4-7 Shear Rate Analysis of Beam Buckling...... 74

5-1 Jammed Microgels as a Biomaterial...... 90

5-2 Phase contrasted micrographs of dilute samples confirm the presence of anionic (MAA), cationic (qDMAEMA), and zwitterionic (CBMA) microgels...... 90

5-3 Rheological characterization of charged microgels...... 91

5-4 Rheological Characterization of Polyelectrolyte Microgels...... 92

5-5 Multivalent ion interactions with charged microgels...... 93

5-6 Microgel rheology and polyelectrolyte scaling behavior...... 94

5-7 Rheological Characterization of Uncharged Microgels...... 95

5-8 Small amplitude oscillatory frequency sweeps of polyelectrolyte microgels swollen in MEGM cell growth media at 4 wt% polymer...... 95

5-9 Cell viability and proliferation...... 96

5-10 Fluorescence microscopy images of MCF-10A cells cultured in jammed polyelectrolyte microgels...... 96

5-11 Metabolic activity in polyelectrolyte microgels...... 97

A-1 Manufacturing Space Curves of Sacrificial Support 3D Printing Methods...... 104

A-2 Characterizing Microgel Size...... 105

A-3 Experimental Set-Up for Microgel Permeability Measurements...... 105

A-4 Permeability Measurements of Jammed Microgel Systems...... 106

LIST OF ABBREVIATIONS

3D Three Dimensional

ATP Adenosine Triphosphate

BIS N,N’-Methylenebisacrylamide

CBMA Carboxybetanine Methacrylate

GPC Gel Permeation Chromatography

LLS Liquid Like Solid

MAA Methacrylic Acid pAAM Poly(Acrylamide)

PDMS Polydimethylsiloxane

PEG Poly(Ethyelen Glycol)

PEGa Poly(Ethylene Glycol) Methyl Ether Acrylate

PEGda Poly(Ethylene Glycol) Diacrylate

PEO Poly(Ethylene Oxide)

PPO Poly(Propylene Oxide) qDMAEMA Quaternized dimethylaminoethyl methacrylate

SAXS Small Angle X-Ray Scattering

SEBS Styrene-block-ethyelen butylene-block-styrene

SEM Scanning Electron Microscopy

SEP Styrene-block-ehtylene propylene

SWLI Scanning White Light Interferometry

UV Ultraviolet

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

AQUEOUS AND ORGANIC MICROGELS FOR SOFT MATTER MANUFACTURING

Christopher S. O’Bryan

May 2019

Chair: Thomas E. Angelini Major: Mechanical Engineering

New capabilities in soft matter manufacturing have been developed to meet the increasing demand for complex three dimensional structures of soft solids and complex fluids.

Tissue and biomedical engineering now leverage additive manufacturing techniques to structure living cells, synthetic polymers, silicone elastomers, and other biomedical materials with high precision in 3D space. To support these printed structures during their solidification, a variety of sacrificial support materials have been developed. Common across these approaches is the use of soft materials in which small perturbations in the applied stress can manifest as drastic changes in the material properties. Here, we explore the application of jammed granular microgels as a sacrificial support material for soft matter manufacturing. We investigate the design and synthesis of both organic and aqueous microgels and explore their application for soft matter 3D printing, bioprinting, and 3D cell culturing applications. Additionally, we explore the destabilizing effect of interfacial tension on the deformation and failure of both liquids and soft solids in an immiscible jammed microgel environment.

CHAPTER 1 SACRIFICIAL SUPPORT MATERIALS FOR SOFT MATTER MANUFACTURING

Introduction

A defining characteristic of soft materials and complex fluids is their propensity to undergo drastic change in mechanical properties resulting from small mechanical or chemical perturbations.1 The moduli of soft materials can range from pascals to kilopascals, making them excellent surrogates for tissues and cellular constructs. However, difficulties can arise in

structuring these materials in their pre-solidified state; solidification times may range from

several seconds for hydrogels and silicone elastomers, to hours or days for biopolymers or

extracellular matrix produced by cells.2-4 Even after solidification, the final structures may be too

delicate to handle and unable to support their own weight. Until recently, it was practically

impossible to reproducibly shape soft materials into complex 3D structures with high spatial

resolution and small feature sizes.

To address these challenges, new methods for soft matter manufacturing have been

developed to leverage the material properties of soft materials to provide continuous support of

3D printed structures (Figure 1-1). Unique to these approaches is the ability to create complex

3D shapes out of liquids with little to no constraints on the solidification time of the printed

structure. After solidification, the printed structure can be either removed from the sacrificial

support material and handled, or left in the support material, providing a controlled environment

for biological studies.5 The quality of the printing process is independent of the rheological

properties of the printed ink but depends heavily on the material and rheological properties of the

support material. Thus, as new sacrificial support materials continue to be developed for soft

matter manufacturing, it is necessary to characterize their material properties and rheological

responses to stress perturbations.

Rheological Characterization of Support Materials

Recently developed sacrificial soft materials for 3D printing include jammed granular

particles, entangled polymer solutions, micelles packed into solid-like phases, and polymer

networks with reversible bonds.6-10 Often, these materials exhibit drastic changes to their rheological properties to facilitate the printing process; for example, jammed microgels will transition between the solid and fluid state under small stresses while polymer solutions are better described as shear thinning fluids. While simple rheological models cannot fully capture the complex behavior exhibited by many of these materials, they can provide the foundation for understanding the underlying behavior. Here, we present the fundamental models for linear

elastic solids and viscous fluids, as well as two classical viscoelastic models, the Maxwell and

Kelvin-Voight model.

Linear Elastic Solids and Viscous Fluids

The simplest method for characterizing a material’s behavior is to look at the linear

response to an applied stress. Linear elastic solids will deform proportionally to an applied stress

and can be modeled as a classical Hookean spring (Figure 1-2a). Here, the modulus of the

material is analogous to the spring constant such that τ = Gγ, where τ is the applied shear stress,

G is the shear modulus, and γ is the resulting shear deformation. By contrast, the deformation of

a fluid to an applied stress is proportional to the shear rate and can be modeled by a dashpot or

damper (Figure 1-2b). Here, the viscosity of the fluid can be thought of as the dampening

coefficient such that τ= ηγ , where τ is the applied shear stress, η is the viscosity of the fluid,

and γ is the shear rate of deformation.

While Newtonian fluids will have viscosity that remains constant, independent of the

applied shear stress, the viscosity of a non-Newtonian fluid is highly dependent on the shear

stress applied. Shear thinning materials exhibit lower viscosities with increasing shear stress. In

comparison, the viscosity of shear thickening materials increases with higher shear stresses

(Figure 1-2c). Traditionally, soft matter manufacturing has relied on the shear thinning behavior

of the printed inks to enable extrusion based 3D printing.11-12 As the printed material is extruded

through the nozzle, the shear stresses acting on the ink reduce the viscosity, increasing the

deposition rate of the material. Recently, these shear stresses have been shown to induce

alignment of liquid crystal polymer solutions to tailor the local stiffness and strength of the final

3D printed structure.13 Once the printed material has left the nozzle, the shear stresses are

removed and the viscosity increases, slowing the rate of deformation of the printed structure as it

solidifies.

In practice, the complex material often studied in the field of soft matter will exhibit a

combination of elastic and viscous responses to applied stresses and are thus referred to as

viscoelastic materials. For example, entangled polymer solutions will exhibit the elastic solid

like behavior on short time-scales (high frequency) while acting as a viscous fluid over long time

scales (low frequencies). Thus, it is important to look at the time and frequency depend

responses of a material to perturbations of stress and strain to understand the rheological

behavior of the material. The Maxwell and Kelvin-Voigt models, discussed below, are a first step in understanding the transient response of these materials.

Maxwell Model

The Maxwell model describes viscoelastic materials that at low frequencies behave like viscous fluids and at high frequencies behave like elastic solids. This behavior can be modeled by a spring and dashpot connected in series (Figure 1-3a). In the Maxwell model, the spring and

dashpot experience the same stress (σσ=sd = σ), however, the total strain is the cumulative

strain of each element (γγ=sd + γ ). Substituting in the stress strain/strain rate relationships and

solving the ordinary differential equation leads to an expression for the shear modulus of the

− t τ material as function of time, that is G= Ge0 , where τ is the relaxation time of the material.

The shear modulus of a material can be broken into two components, a shear storage

modulus (Gʹ) associated with the elastic response and the shear loss modulus (Gʺ) associated

with the viscous dissipation. By taking the frequency response function of the stress and strain,

we find that

∞ G′(ωω) = G( t ′)sin ( ω t ′′) dt ∫0

and

∞ G′′ (ωω) = G( t′′′)cos( ω t) dt . ∫0

Plugging in our relationship for G(t), we find

(ωτ )2 ′ GG(ω) = 0 2 1+ (ωτ )

and

(ωτ ) ′′ GG(ω) = 0 2 . 1+ (ωτ )

Thus, at the high frequency limit, a Maxwell material will exhibit solid like behavior, but at low

frequencies the viscous shear component is dominant, and the material behaves like a fluid; even

at the limit of zero applied stress and zero strain these materials will behave like a fluid and are

often referred to as “Maxwell Fluids” (Figure 1-3b).

There are several materials used in soft matter 3D printing, as either sacrificial supports

or as sacrificial inks, that can be considered Maxwell fluids, including concentrated micelles and

polymer networks with weak reversible bonds. Generally, these materials exhibit thermally

driven relaxations resulting from polymer entanglement, particle rearrangements, or spontaneous

bond-breaking. However, at the limit of long timescales, these sacrificial support materials will

behave as viscous fluids, and the printed structure will displace the sacrificial support materials

and deform. Thus, other materials may be necessary if the printed structure is to be supported

indefinitely in time.

Kelvin Voight Model

In direct contrast to the Maxwell model, the Kelvin-Voight model describes materials

that at low frequencies behave like elastic solids and at high frequencies behave like viscous

fluids. This behavior can be modeled by a spring and dashpot connected in parallel where the

spring and dashpot the same strain but different stresses (Figure 1-4a). Here, the spring and

dashpot experience the same strain (γγ=sd = γ) while the total stress on the system is the sum

of the stress experience by the spring and dashpot (σσ=sd + σ). Starting with a step strain

response and modeling the stress response, the shear modulus of a Kelvin-Voight material can be

expressed by Gt( ) = G0 +ηδ ( t) , where δ(t) is the Dirac function.

Using the previously discussed relations between the time and frequency responses of the

elastic material, we find that the elastic and viscous shear moduli can be expressed by as Gʹ(ω) =

G0 and Gʹʹ (ω) = ηω for Kelvin-Voight materials. At the low frequency limit, these materials will

behave like a linear-elastic solid as the elastic shear modulus remains flat and dominates the

viscous shear modulus (Figure 1-4b). Thus, we must now consider the stress necessary to yield

this class of materials when selecting a Kelvin-Voight material as a sacrificial support material

for soft matter manufacturing, including jammed granular microgels or solid gels with an

additional filler material.

3D Printing Instabilities in Sacrificial Support Materials

3D printing is a race between the destabilizing instabilities that may result in break-up or deformation of the printed structures, and the solidification of the printed material. Often, these instabilities are overcome by matching the solidification time of the printed material with the printing time, or by tailoring the rheological properties of the printed ink to provide additional support during the solidification process. Alternatively, sacrificial support materials may be used to overcome these instabilities with limited constraints on the printed material. Here, we discuss the instabilities related to soft matter manufacturing and the methods in which they are overcome in sacrificial support materials. In addition to the gravitational and interfacial instabilities associated with 3D printing, we discuss instabilities directly related to printing into sacrificial support materials, notably crevasse formation in the needle wake and inertial instabilities when printing at high speeds.

Gravitational Instabilities

Gravitational forces acting on printed structures can lead to sagging and deformation of the object before the solidification (Figure 1-5a). Various strategies have been developed to overcome these gravitational instabilities, including tailoring the rheological properties of the printed ink to provide mechanical stability, and tailoring the solidification time to align with printing speeds. Alternatively, sacrificial support baths may be used to provide mechanical support to printed structures during the solidification time.

Gravitational instabilities can be overcome in sacrificial support baths when the buoyancy forces acting on the printed structure is eliminated by matching the density of the support bath with that of the printed ink. Density requirements can be estimated by equating the destabilizing buoyancy forces acting on the printed structure to the resisting Stoke’s drag. For a spherical object, the speed that the object either moves or sinks is

∆ρVg v = 6πηr

where Δρ is the density mismatch, V is the volume of the sphere, g is gravitational acceleration, η

is the viscosity of the support material, and r is the radius of the sphere. In practice, perfect

density matching of the ink to support material is unobtainable and the printed structure will rise

or sink under the buoyancy forces. While viscous support materials may mask this instability

over the time scales of the print, alternative approaches may provide improved stability over

longer timescales.

An alternative approach to density matching is to 3D print the structures into a support

bath that exhibits a finite yield stress at relatively long time-lengths, including polymeric

networks and jammed granular microgels. Here, the stability of a printed structure can be

predicted by comparing the gravitational forces acting on the object to the yield stress of the

material, such that τρyhA> Vg where τy is the yield stress, Ah is the hydrodynamic surface area

of the printed structure, ρ is the density mismatch between the support bath and the printed ink, V is the volume of the printed object, and g is the acceleration due to gravity. Thus, printed structures with feature sizes

12τ r < y ∆ρg will remain stable indefinitely, allowing fluids to be freely structured in 3D with no limitation on solidification time.

Interfacial Instabilities

Immiscibility between the sacrificial support material and the printed inks can introduce instabilities of printed structures through mechanisms similar to the Rayleigh-Plateau instability.14-16 When destabilizing interfacial forces exceed the yield stress, printed structures

will begin to break up. A critical feature size in which feature remain stable (lc) can be predicted

γ such that lc = , where γ is the interfacial tension between the printing ink and the jammed τ y

microgels, and τy is the yield stress of the microgels. Additionally, immiscibility between printed structures and support material can lead to nanoparticles or microparticles accumulation at the interface, similar to a Pickering emulsion. While these particles may act as surfactants, stabilizing the interface between the ink and support material, they can prevent coalescence between subsequent printing layers. The influence of interfacial instabilities in 3D printing with jammed microgel supports is further explored in later chapters.

Inertial Instabilities

Inertial instabilities arise at high translational printing speeds; competition between the viscous and inertial forces may result in unpredictable flow patterns in the wake of the printing nozzle and intermixing of the support material with the printed ink (Figure 1-6). The onset of these inertial instabilities can be predicted by the dimensionless Reynold’s number; for flow around a cylinder, unstable flow is predicted for Re = 10-15.17-19 The Reynold’s number for a

ρvd cylinder flowing through a material is Re = , where ρ and η are the density and viscosity of η

the support material, and d and v are the diameter and translational velocity of the printing

nozzle. Recent studies of high speed printing (v > 1 m/s) into jammed granular microgels found inertial instabilities arising at Re = 3.7-17, consistent with the predicted Reynold’s numbers for unstable flow around a cylinder.19-20

Crevasse Formation

When printing at high speeds into low yield stress microgels supports, a dynamic crevasse may form in the wake of the needle (Figure 1-6). This transient crevasse arises when the

vη hydrostatic pressure driving the reflow, ρgh , is resisted by the viscous stresses, , where v is d

the translational printing velocity, η is the viscosity of the support materials, and d is the

diameter of the printing nozzle. This dynamic crevasse may cause disruption to printed structures

when the depth of the crevasses is of the same length as the depth of the printing nozzle. Thus,

the maximum translational velocity for a given printing depth in microgel media can be

expressed as

ρghd v < , η

where h is the depth of the printing nozzle.

Even at the limit of low translational speeds, a static crevasse may form in the wake of

the printing nozzle as it traverses through the microgel support material if the hydrostatic

pressure is less than the yield stress of the support material, that is ρτgh < y ; here τy is the yield stress of the support material, ρ is the density of the support material, g is the acceleration due to gravity, and h is the depth of the needle (Figure 1-7). This instability is generally avoided by

preparing microgel support materials with relatively low yield stresses, but may also be avoided

through the use of a secondary filler material to flow into the crevasse as it forms.21

Sacrificial Support Material 3D Printing

To overcome instabilities encountered in soft matter 3D printing, a variety of materials have been developed and utilized as sacrificial supports for soft matter manufacturing. Below, we present several of these materials and compare their rheological performance to the previously mentioned models.

Density Matched Support Baths

Gravitational instabilities can be overcome when printing into a support bath with the

same density as the printed ink, eliminating buoyancy effects that can cause deformation of the

printed structure.22-24 This approach is generally used when employing an inkjet deposition

method for 3D printing as the translational movement of an extrusion nozzle can result in

currents in the support bath and lead to destruction of the printed structures. In one example,

alginate is 3D printed into a calcium chloride support bath.22 Here the calcium chloride both

provides the means to balance the density of the support material to match that of the printed

alginate, and to begin the solidification process once printed. However, as previously mentioned,

even small density mismatches can result in deformation of the printed structure over time that

must be accounted for.25

Packed Micelles

Nonionic triblock copolymers composed a hydrophobic poly(propylene oxide) (PPO)

mid-block and two hydrophilic poly(ethylene oxide) (PEO) end-blocks, commonly called

poloxamers, will form micelles when prepared at polymer concentrations above the critical

micelle concentration in water; the PPO mid-blocks will congregate to form hydrophobic cores while the PEO end-blocks will form highly swollen coronas. Packed micelle solutions made from Pluronic F127 have been demonstrated as a sacrificial support material for 3D printing fluorescently dyed Pluronic F127 inks.26-27 However, when we investigate the frequency response of packed micelles prepared from 25 wt% F127 Pluronic, we find that the elastic modulus continuously decreases with decreasing frequencies (Figure 1-8). If we were to

extrapolate the data, we expect a crossover in the elastic and viscous shear moduli around 10-4

Hz. Thus, over long time scales, sacrificial support materials made from F127 micelles will

exhibit fluid like behavior and may not be suitable for long-term support of printed structures.

Networks with Reversible Bonds

Self-healing hydrogel support materials have been shown to enable the 3D printing of

hydrogel materials, both as the sacrificial support material and as a stabilizing mechanism for the

printed ink.7, 28-29 Here, reversible non-covalent bonds formed through guest-host relationships

are based on modified hyaluronic acid with adamantine, or β-cyclodextrin. As the authors report

in their manuscript, these dynamic bonds lack the mechanical properties necessary for long term

stability or perfusion. This behavior is evident in the reported rheological characterization of the

materials as frequency sweeps show a cross-over of the elastic modulus (Gʹ) and the viscous shear modulus (Gʺ) (Figure 1-9a,b). It is interesting to note that while increasing the number of guest-host interactions along the polymer chain can increase the solid like behavior over intermediate time periods, a cross-over in the elastic and viscous shear moduli is still be observed at the limit of low frequencies (Figure 1-9b). While these materials may not be suitable for supporting printed structures for indefinite lengths of time, the authors do note that the hydrogel can be modified with a secondary, covalent crosslinking mechanism allowing the formation of a permanent gel network that exhibits solid-like behavior over all reported frequencies (Figure 1-

9c,d). After covalently crosslinking the support material, a transition can be seen in the rheological behavior of the gel as it transitions from fluid-like behavior to solid-like behavior.

Colloidal Clays

Laponite is a synthetic magnesium silicate clay with wide spread application as a rheological modifier. When dispersed in water, these nanometer-sized disc shaped particles will self-assemble through electrostatic interactions of their negatively changed faces and positively charged rims to form a thixotropic gel network with shear thinning rheological behavior.30-31

Laponite has found broad applications in 3D bioprinting and tissue engineering as rheological

modifiers to the bioinks and polymer solutions, providing stability to 3D printed structures

during the solidification processes.32-33 Recently, laponite solutions have been employed as

sacrificial support baths to enable the 3D printing of alginate and UV-curing epoxy resins.34

Over short time scales, laponite exhibits solid-like behavior with the elastic shear modulus

dominating the viscous shear modulus; however, crossovers in the shear moduli have been

observed at low frequencies presenting challenges in providing long-term support to printed

structures.35

Jammed Granular Microgels

Microgels have an extensive history as both a rheological modifier for industrial applications and as a platform for studying the jamming phenomenon.36-39 As the volume

fraction of microgels increases toward random close pack limit of hard spheres (Φ = 0.68) the system undergoes a jamming transition and exhibits solid-like behavior under low levels of applied stress.40-41 This transition is captured by the rheological behavior of the microgels,

notably a relatively flat elastic shear modulus that dominates the viscous shear modulus even at

the limit of low frequencies, and a shear rate independent shear stress at low shear rates

corresponding to the yield stress of the system.42-43 When the microgels are granular in size (d >

1-10 µm), gravitational body forces begin to dominate thermally driven Brownian motion and the particles remain stable in this jammed state.44

Recently, this jamming/unjamming transition has been leveraged by the soft matter

community to enable the 3D printing of colloidal particles, polymers, biopolymers, hydrogels,

cells, and silicone elastomers.6, 8, 45-48 3D printing into a sacrificial support bath of jammed microgels is possible because the stresses generated by a translating printing nozzle results in localized yielding of the microgel support without disturbing previously printed constructs.6 As

the injection tip leaves the printing region, the microgels will spontaneously and rapidly reflow,

trapping the printed structure in place and recovering their solid-like rheological properties. In

the following chapters we present a detailed exploration of the design and application of aqueous and organic microgels for 3D printing soft matter materials.

Figure 1-1. Soft Matter 3D Printing into Sacrificial Support Materials. a) Liquid inks are 3D- printed directly into sacrificial support materials. b) The sacrificial support material provide continued support to the printed structure during the solidification of the printed structure. Once cured, the printed structure can be either c) removed from the sacrificial support material and handled, or d) left within the sacrificial support material, providing a controlled environment for further studying or maturation.5

Figure 1-2. Rheological Behavior of Elastic Solids and Viscous Fluids. a) Elastic solids will undergo finite deformation under an applied stress. b) Under an applied stress, a viscous fluid will undergo continuous deformation until the stress is removed. c) A Newtonian fluid will exhibit a linear relationship between the shear stress and the shear rate. A shear-thinning material will experience a decrease in viscosity with increasing shear stresses. Likewise, a shear-thickening material will experience increased viscosity under higher shear stresses.

Figure 1-3. The Maxwell model. a) The Maxwell model can be described as a spring and dashpot in series. In this case, the spring and dashpot experience the same stress and the total strain is the sum of the strain of the spring and dashpot. b) Frequency sweeps of Maxwell fluids show a crossover of the elastic and viscous shear modulus. At low frequencies the viscous component dominates; at high frequencies the elastic component dominates.

Figure 1-4. The Kelvin-Voight model. a) The Kelvin-Voight model can be described by a spring and dashpot in parallel in which the strain of the dashpot and spring are equal, and the total stress is the sum of the individual stresses. b) Frequency sweeps of Kelvin-Voight materials show a constant elastic shear modulus and a viscous shear modulus that increases linearly.

Figure 1-5. Gravitational Instabilities in Sacrificial Support Materials. a) Gravitational instabilities can result in material sagging and deformation of printed structures when not properly supported. b) A material 3D-printed into a sacrificial support material will remain stable if the gravitational body forces are less than the force required to yield the support material.5

Figure 1-6. Inertial Instabilities in Sacrificial Support Materials. Inertial instabilities can arise when printing with low viscosity inks at high translational printing speeds. These inertial effects can lead to recirculation of the ink in the wake of the nozzle and intermixing of the support material into the printed ink. Additionally, transient crevasses may arise at high printing speeds.5

Figure 1-7. Static Crevasse Formation in Sacrificial Support Materials. Static crevasses can arise in the wake of the printing nozzle when the yield stress of the support material exceeds the hydrostatic pressure.5

Figure 1-8. Rheological characterization of F127 micelle samples. Small amplitude frequency sweeps of 25 wt% F127 Pluronic samples show a continuous decrease in the elastic modulus with decreasing frequencies.

Figure 1-9. Rheology of Networks with Reversible Bonds. a-b) Rheological characterization of adamantine and -cyclodextrin modified hyaluronic acid hydrogels exhibiting guest-host relationships with a 1:1 ratio at varying concentrations and degree of modification. c-d) Covalent crosslinking of modified hydrogels through photo- polymerization shows a shift from Maxwell behavior to Kelvin-Voight behavior.7

CHAPTER 2 JAMMED GRANULAR MICROGELS AS SACRIFICIAL SUPPORT MATERIALS

Introduction

Advanced tissue engineering now employs additive manufacturing techniques for 3D printing with living cells, natural extracellular matrix, and synthetic hydrogels.49-50 These

materials provide practically no physical support to the printed structures that they constitute,

which has led to the development of complex reinforcement strategies and specially designed

bio-inks.5, 51-53 Once printed, soft biomaterial structures are delicate and require extremely gentle

processing steps to be removed from their supporting matrix. Accordingly, such delicate

hydrogels and cell-laden scaffolds are often toughened chemically or mechanically to improve

their functionality.54-55 An alternative, purely physical strategy has recently been developed that

leverages the microgel jamming transition to preserve structures by trapping material in space

with a demonstrated precision of 1–2 cell diameters.6, 45, 47 This approach to structuring non-self-

supporting materials with high precision and few restrictions on material selection has been

tested with hydrogels, extracellular matrix, and living cells.

As the 3D printing technology has matured, a variety of new microgel systems have been

developed as support materials for soft matter manufacturing. One approach to manufacturing

microgel support baths is through the preparation of hydrogel slurries in which hydrogel slabs

are blended, cut, or otherwise mechanically fractured into micron sized particles. While this

approach has been demonstrated for the preparation of gellatin,8 gellan,56 and agarose microgel

particles,57-58 dispersity of the microgel size, the scalability of the synthesis process, and control

over the solvent and polymer concentrations impede the wide spread adoption of these methods

for 3D bioprinting and soft matter manufacturing. Alternatively, commercially available

microgel particles provide a fast and scalable means to acquire microgels in their unswollen

state, allowing for complete control over the solvent.6, 48 However, only a limited number of commercially available microgel system have been demonstrated for soft matter manufacturing.

Commercially available microgels can vary widely in their polymer-solvent interactions, and in their crosslinking and charge densities. If jammed granular microgel systems are to be useful throughout soft matter manufacturing and bioprinting, similar rheological behavior must be achievable in chemically distinct microgel systems.

Here, we investigate the effects of chemical composition of four chemically distinct, commercially available microgel systems on the precision of 3D printing into granular microgels. We find that the rheological behavior of each microgel system can be tuned to similar behavior through changes in the polymer concentration. In all cases, the resulting materials perform comparably in 3D printing tests, in which a 20–50 µm feature size is achieved.

Results and Discussion

To determine whether the rheology of chemically distinct LLS materials can be utilized for 3D printing, we investigate four different, commercially available microgel systems in search of rheological properties favorable for 3D bioprinting, notably a low elastic modulus (≈100 Pa) and yield stress (≈10 Pa), dominant elastic behavior, and rapid recovery of this elastic behavior after yielding (≈1 s). Here, we focus our investigation on Carbopol ETD 2020, Carbopol 980-NF,

Carbopol Ultrez 20, and Pemulen TR-2NF. These materials are composed of acrylic acid with varying charge densities, crosslinking densities, and functional groups that provide different swelling behavior. Briefly: ETD 2020 and Ultrez are copolymers consisting of acrylic acid and

C10-C30 alkyl acrylate that have been hydrophobically-modified with a PEG-block-alkyl acid ester copolymer whereas 980-NF is a crosslinked polyacrylic acid with no additional functional groups.59-60 Pemulen is a copolymer consisting of acrylic acid and C10-C30 alkyl acrylate crosslinked with allyl pentaerythritol and modified to provide emulsifying properties.61-62 We

hypothesize that these differences will change the swelling behavior of the microgels, thereby re-

scaling the concentration dependence rheological performance. To explore these changes in rheological performance, we test all four systems over a wide concentration range.

To characterize the rheological behavior of the various microgels, we perform low amplitude frequency sweeps, covering a range of 10-3 to 100 Hz, measuring the elastic modulus (

G′ ) and viscous modulus ( G′′ ). We find that for all four systems, G′ and G′′ remain relatively

flat and separated across the spectrum of oscillatory frequencies, behaving like linear solids with

damping (Figure 2-1a). Recently, differences between these materials in their charge density

were found through analyzing the scaling relationships between their moduli and

concentrations.63 Here, we find that Ultrez, 980-NF and Pemulen follow scaling concentration scaling behavior consistent with low charge density polymers (G′ ~ c9/4) whereas the ETD 2020

scaling is consistent with high charge density polymers (G′~ c1.0), where c is the microgel

polymer concentration (Figure 2-1b). We show data from samples that exhibited favorable 3D

printing performance, including a low elastic shear modulus (G′ of 20Pa – 100Pa) and dominant

elastic behavior over all frequencies (Figure 2-1a). Thus, the chemical differences between these

different microgels do not alter their qualitative rheological behavior and their linear behaviors

can be tuned by controlling the polymer concentration.

Low yield stress and short thixotropic time of LLS materials are essential for their use in

3D printing to localize yielding and promote rapid re-flow and solidification of the granular

microgel. The yield stress, τy of LLS is determined from a unidirectional shear rate sweep where

shear stress is measured at progressively decreasing shear strain rates. As the shear rate, γ is

ramped down to 10-3 s-1, the measured stress becomes independent of the applied shear rate. This

plateau stress is the yield stress of the LLS and can be determined by fitting the Herschel-

τ= τ + γγ p γ Bulkley model, yc(1 ( ) ) , where c is the crossover shear rate between regimes of

behavior, and p is a dimensionless constant of order 0.5 (Figure 2-1c).64 Here, we measure the

yield stress of each microgel system at various concentrations (Figure 2-1d). We find that all four

chemically distinct microgel systems have a narrow range of yield stresses between 7-10 Pa at

concentrations favorable for 3D printing (Figure 2-1c).

The time it takes for a yielded material to spontaneously re-solidify after the applied

stress drops below the yield stress is the thixotropic time. For 3D bioprinting into jammed

granular microgels, this sets the time-scale in which materials become trapped once the printing

nozzle has left the printing area. We measure the thixotropic time of jammed microgels by

observing the shear rate response as the shear stress is dropped from 100 Pa to 1/10 the yield

stress of the system; the point at which the shear rate plateaus and approaches zero is the

thixotropic time of the system. We find all four microgel systems exhibit thixotropic times of

order 1 s at the previously described concentrations (Figure 2-1e).

To explore the application of LLS as a support material for 3D bioprinting, we perform

short-term cell viability studies of cells cultured in microgel media. 3T3 fibroblast cells are

dispersed in microgels swollen in cell growth media with the rheological properties previously

described. We find that the LLS result in a reduced cell viability relative to cells cultured in

liquid cell growth media (Figure 2-2). Interestingly, microgels that have been hydrophobically

modified with additional functional groups, i.e. ETD 2020, Ultrez, and Pemulen, demonstrate

lower cell viability compared to the non-functionalized 980-NF. Previous studies have shown that viability is increased in cells cultured in ETD 2020 when extracellular matrix is included, so these results do not necessarily indicate a lack of suitability for any of these materials for bioprinting applications.45 However, as the use of LLS for 3D cell culturing grows, it may be

necessary to develop new microgels to improve cell viability and performance in the absence of

added matrix, potentially through a reduction in the polymeric charge density.

To demonstrate the potential of these commercially available microgel systems for 3D

bioprinting, we print a series of horizontal, vertical, and bent tubes out of a mixture of

polyvinylalcohol polymer and fluorescent microspheres (Fig. 2-3). Each tube is printed into a

different microgel system prepared at the concentrations present in figure 2-1, using different tip

designs, and path planning strategies designed to leverage the unique printing opportunities

available when printing into jammed granular microgels. While it is possible to print both

horizontal (Figure 2-3a) and vertical tubes (Figure 2-3c) with a single needle design, certain

geometries are better suited for a given orientation. Structures created from continuous helical

patterns (Figure 2-3b) are observed to have higher degrees of cross-sectional symmetry than those printed using more complex paths (Figure 2-3a). By utilizing a needle oriented at 45 degrees from vertical, we are able to follow a continuous helical pattern in both the horizontal and vertical directions without damaging the printed structure. This tip design enables bent tubular structures to be printed in a single path without interruption (Figure 2-3d). To directly test the level of precision achievable in the different microgel systems, single features are drawn and measured with confocal microscopy, demonstrating a feature size as small as 20 µm (Figure

2-3e).

Conclusions

Here we have studied several different liquid-like solid materials made from commercially available microgels of differing chemical composition. We have found that the rheological properties of these materials can be tuned and optimized to perform similarly in 3D printing. All four systems behave like soft elastic solids at low strains, yield at low stresses, and transition rapidly between the solid and fluidized states. These rheological behaviors differ

qualitatively from other complex fluids like entangled or dynamically bonding polymer networks, which exhibit apparent yielding and solid-like behaviors at short time-scales, yet evolve spontaneously over long times due to thermally driven relaxations.5 By contrast, jammed granular microgels are thermodynamic solids and can support 3D printed structures for long times. The performance of these materials for 3D bioprinting is demonstrated by the ability to generate fine, precise structures with numerous tip designs, feature orientations, and printing paths. The ability to print tubes in arbitrary orientations, in conjunction with the ability to join tubes with complex connections opens the door to creating complex 3D tubular networks, which is a huge challenge in tissue engineering applications. While preliminary short-term cell viability studies of cells cultured in LLS are presented here, further studies of long-term cell viability and metabolic activity are needed if these microgels are to be used in bioprinting applications; the results described here demonstrate their promise in terms of rheology and printing performance.

Figure 2-1. Rheological characterization of Chemically Distinct LLS Materials. Frequency sweeps are performed on four different microgel materials at 1% strain to measure the elastic (G′) and viscous (G′′ )shear moduli. a) At concentrations favorable for 3D printing, G′ remains flat and separated from the G′′ , a characteristic of soft elastic solids. b) For all four materials, G′ increases with increasing polymer concentration. Ultrez, 980-NF, and Pemulen exhibit scaling behavior consistent with low charge density polyelectrolytes (c9/4) while ETD 2020 follows the scaling behavior of high charge density (c1.0). To study the transition between solid and fluidized behaviors, a shear rate sweep is performed and the Hershel-Bulkley model is fit to the data. c) We find that all four commercially available microgels can be tuned to have a yield stress favorable for 3D bioprinting (~10 Pa). d) The yield stress of the material can be tuned through changes to the polymer concentration. e) The thixotropic time is determined by measuring the shear rate over time after dropping the applied stress to 1/10 the yield stress.48

Figure 2-2. Short-term Cell Viability Studies of Cells Dispersed in Microgels. Short-term cell viability studies of 3T3 fibroblast cells dispersed in commercially available microgel samples swollen in cell growth media relative. Cell viability is relative to cells cultured in liquid growth media.48

Figure 2-3. Macroscopic and Confocal Imaging of 3D-Printed Tubes. Hollow tube with 1mm diameter with wall thickness of order 125µm are 3D-printed into various microgel systems. a) Horizontal orientation in Ultrez with a vertical nozzle following a layer- by-layer printing path. b) Horizontal orientation printed into ETD 2020 with a bent nozzle and helical printing path. c) Vertical orientation in Pemulen with a vertical nozzle following a helical printing path. d) Confocal imaging of an “L”-shaped tube printed into 980-NF with a printing nozzle bent at 45 degrees following a helical printing path. e. An end-on view of single linear features printed the four microgel systems exhibit diameters ranging from 20-50µm (scale 100µm). This view of the features is generated by a maximum intensity projection along the linear feature, showing the absolute border of each written structure.48

CHAPTER 3 ORGANIC MICROGELS FOR SILICONE 3D PRINTING

Introduction

Silicone elastomers have found wide spread application in the industrial and biomedical fields, in part due to their low moduli, high extensibility and toughness, excellent thermal and oxidative stability, and chemical inertness.65-67 Although these materials properties have enabled

a multitude of industrial and medical applications, the challenges of handling the silicone

elastomer in its pre-vulcanized liquid state has limited its application in additive

manufacturing.68-72 Only recently have manufacturing methods emerged to enable the 3D

printing of liquid silicone elastomers. However, demonstrations of 3D printing liquid silicones

into jammed microgel supports have been marred by interfacial instabilities between the organic

inks and aqueous support materials; commercially available silicone-based microgels offer

limited rheological control of the support medium. An organic based jammed microgel system

with tunable rheological properties would enable the precise 3D printing of silicone elastomers and other organic soft materials by greatly reducing the interfacial instabilities between the printed material and their microgel support.

In this chapter, we develop a granular organic microgel system through the self-assembly of polystyrene block copolymers prepared at low polymer concentration in mineral oil. We characterize the rheological performance of the microgels in search of a formulation favorable for 3D printing and explore the limits of interfacial instability between the silicone oil and mineral oil. Furthermore, we demonstrate the capabilities of these organic based microgels as a sacrificial support material by 3D printing model trachea implants, perfusable tubular networks, functional fluid pumps with embedded features, and high precision silicone scaffolds.

Results and Discussion

Micro-Organogel Synthesis and Characterization

Block copolymer self-assembly is a well-established route for designing soft materials, offering control of nanostructure and rheological properties.73-75 To develop organic based

microgels for silicone 3D printing, we leverage the polymer self-assembly of block-copolymers

in an organic solvent. Here, we use a polystyrene-block-poly(ethylene/propylene) (SEP) diblock

copolymer and a polystyrene-block-poly(ethylene/butylene)-block-polystyrene (SEBS) triblock

copolymer; the polystyrene blocks are hydrophilic relative to the ethylene/propylene and

ethylene/butylene blocks (Figure 3-1a). We prepare these polymers at low polymer

concentrations (4-5 wt%) in light mineral oil and heat to 100 °C. The SEP and SEBS block

copolymers will self-assemble into structures with 1-2 nm glassy polystyrene cores with coronas

containing the oil-soluble blocks.75-77 In a pure diblock system, we find the SEP block copolymer

will assemble into discrete micelles (Figure 3-1b,c). By contrast, the pure triblock system will

assemble into a macroscopic network in which the ethylene/butylene blocks connect neighboring

polystyrene cores, analogous to crosslinks in a traditional hydrogel network (Figure 3-1d,e).78-79

Here, the spacing between the polystyrene cores is dictated by the swelling behavior of the

ethylene/butylene blocks. For both the triblock and diblock systems, the glassy polystyrene cores

prevent exchange of polymers between structures.

As previously discussed, gravitational instabilities are overcome by jammed granular

microgel support due to their low yield stress, low moduli, spontaneous reflow after yielding,

and rapid recovery of solid-like properties.6, 20, 45, 48 Here, 3D printing into a micelle suspension of diblock copolymers is inhibited by its fluid-dominated rheological properties; printed features will move under the buoyancy forces or break-up into droplets due to interfacial forces. By contrast, pure triblock networks are fully crosslinked networks, and will not spontaneous reflow

after yielding, resulting in irreversible damage to the gel as the printing nozzle traverses through

the gel. To develop a self-healing system capable of supporting printed structures, we blend

diblock and triblock copolymers in search of a formulation having insufficient bridging to create

a system-spanning network, but with sufficient bridging to form microscopic organogels. Here,

the addition of the diblock copolymers replaces the triblock bridges, promoting swelling without

creating additional bridges (Figure 3-1f-h). Similar to aqueous microgels, these micro-organogels locally fluidize as the printing nozzle translates through the support matrix, allowing the precise manufacturing of 3D printing silicone structures.

To identify a combination of diblock and triblock copolymers favorable for soft matter manufacturing, we perform a series of rheological measurements on samples prepared at various polymer concentrations with different ratios of diblock and triblock copolymer. To classify the time-dependent response of a material, we perform a low-amplitude frequency sweep and measure the elastic and viscous shear moduli (Figure 3-2a). We measure the yield stress of the material by performing unidirectional shear rate sweeps and measure the resulting shear stress

(Figure 3-2b). The yield stress is determined by fitting the Herschel-Bulkley model to the

τ= τ + γγ p resulting data, given by: yc(1 ( ) ) , where τ is the shear stress, τy is the bulk yield

64 stress, γ is the shear rate, γc is the cross-over shear rate, and p is a dimensionless constant.

Finally, we measure the thixotropic time response to quantify the rate of elastic recovery of the

material (Figure 3-2c). Here, we measure the time it takes for the shear rate to drop to zero after

a shear stress greater than the yield stress is rapidly removed.

We find that a formulation of 4.5 wt% global polymer concentration with an equal ratio

of diblock and triblock copolymer results in rheological properties consistent with jammed

microgel systems and are favorable for soft matter manufacturing. The frequency dependent

response of the 50:50 mixture is dominate by the elastic modulus and remains relatively flat over

a large range of frequencies, characteristics often associated with an elastic solid. Increasing the

concentration of diblock relative to triblock copolymer results in a crossover between the elastic

and viscous shear moduli, a characteristic consistent with a viscoelastic fluid as described by the

Maxwell model. Similarly, we find the 50:50 mixture has a yield stress between 3-4 Pa and a thixotropic response time on the order of one second. This short thixotropic reduces the duration over which the microgel support material is fluidized and mechanically unstable. To confirm whether this formulation produce micro-organogels, we dilute the sample in light mineral oil and take micrograph images with phase-contrasted illumination (Figure 3-2d). We observe granular structures on the order of 2-4 µm in size. Systems with higher triblock concentration behave like a fully crosslinked network, exhibiting high yield stress and permanent destruction of the network. We find that the rheological properties (i.e. Gʹ and τy) can be tuned through changes to

the global polymer concentration; increasing the polymer concentration results in a high yield

stress and higher shear moduli.

To provide insight into the micro- and nanostructure of the organogel systems, we

perform small-angle x-ray scattering (SAXS) experiments to measure the spacing between

polystyrene cores (Figure 3-3). We calculate the average core to core spacing, d, of organogel

samples prepared at global polymer concentrations of 4.5 wt% consisting of either 100%

diblock, 100% triblock, or a 50-50 blend of diblock and triblock by determining the location of

the first peak in the scatting intensity S(q). Here, the average core-to-core spacing is measured as

dq= 2π . Using these measurements in conjunction of stochiometric calculations of the

polymer chain density and gel permeation chromatography (GPC) to measure the molecular

weight of the block copolymers, we estimate the total number of polymer chains present in each

organogel unit cell. We find the pure triblock organogel consists of ~12 triblock chains/unit cell

core; the pure diblock organogel consists of ~6 diblock chains/unit cell core; and the blended

organogel consists of ~6 triblock chains and ~3-4 diblock chains/unit cell core. It is intriguing that the organo-microgels formulated here self-assemble when the coordination number between polystyrene cores lies at the limit of the Maxwell criterion for the stability of solids.80

Temperature Effects on Rheological Properties

To explore the efficacy of micro-organogel support materials for 3D printing silicone elastomers requiring elevated vulcanization temperature, we perform rheological characterizations of organogels prepared at global polymer concentrations of 4.5 wt% and 50-50

diblock to triblock at increasing temperatures. We measure the elastic and viscous shear moduli

at 1Hz and 1% strain for temperatures ramping from 25 °C to 80 °C (Figure 3-4a). At relatively

low temperature, the elastic and shear modulus appear to be independent of the temperature.

However, between 60 °C and 70 °C we observe a crossover in the elastic and viscous shear

modulus and organogel system exhibits liquid like behavior. We further investigate this

transition by performing small amplitude frequency sweeps at 1% strain from 101 – 10-3 Hz and

unidirectional shear rate sweeps from 500 – 0.005 s-1 to measure the elastic and viscous shear

moduli, and the yield stress of the organogels (Figure 3-4b,c). Consistent with the amplitude

sweep measurements, we observe a cross-over in the elastic and viscous shear moduli at 60 °C

and a transition from Hershel-Bulkley behavior of the shear rate sweeps to a purely viscous

response.

Precision Printing into Micro-Organogels

To demonstrate the control of feature size achievable with micro-organogel supports, we

3D print an array of lines out of a vinyl-terminated PDMS ink through various combinations of

translational speeds (v = 0.1 – 10 mm/s) and injection flow rates (Q = 10-10000 µL/hr). Printing

into the micro-organogel support is accomplished through a custom-made 3D printer consisting

of a syringe pump attached to three linear translational stages. The three translational stages

follow custom written trajectories at a specified speed while the syringe pump extrudes the ink at

the desired flow rate. To enable 3D confocal fluorescence imaging of the printed features, 1.0

µm fluorescent beads are dispersed in the PDMS ink prior to printing at low volume fractions.

Top-down and end-on projections of the printed structures show smooth features with near

circular cross-sectional areas (Figure 3-5a). We find that by reducing the flow rate of the pump

and increasing the translational speed, features as small as 30 µm are obtainable. Qualitative

analysis of the printed structures show the cross-sectional area follows volume conservation

Q principles, that is = A , where Ac is the cross-sectional area of the structure (Figure 3-5b). We v c note that for feature sizes ranging from 100 µm to 1 mm, the width and height are nearly identical.

As previously discussed, immiscibility between the silicone inks and the mineral oil organogels can lead to interfacial instabilities. End-on projections of printed features show the accumulation of fluorescent microbeads at the interface of the silicone lines and microgels; a phenomenon analogous to the Pickering emulsion in which micro- and nanoparticles will stabilize a fluid-fluid interface (Figure 3-5a).81 This phenomenon may introduce challenges in

achieving layer-to-layer adhesion between 3D printed silicone elastomer structures. Additionally,

break-up of printed structures may occur when the stresses generated from interfacial tension

exceed the yield stress of the supporting micro-organogels. To explore the role of interfacial

instabilities on the stability of printed structures, we print lines of low viscosity neat silicone oil

into micro organogels with varying yield stress and observe their behavior over time. We predict

γ the minimum feature size in which lines will remain stable to be lc = . Here, the interfacial τ y tension between neat mineral oil and neat silicone oil is taken to be γms = 13.1 ± 2.2 mN/m. We find that the minimum feature size for stable structure decreases with increasing yield stress, however, the scaling behavior does not follow the predicted relationship (Figure 3-5c). Increased polymer concentrations of block copolymers may act as stabilizing agents, altering the interfacial tension between the ink and support material and changing the scaling behavior. Further investigations on the role of interfacial instabilities on feature break-up is presented in the following chapter. Interestingly, we find that the time scale at which these break-up events occurs is dependent on the viscosity of the printed material; increasing viscosity of the printed silicone lines resulted in a longer time until breakup (Figure 3-5d).

Mechanical Robustness of 3D printed silicone structures

Strong adhesion between printed filaments in both the vertical and lateral direction is necessary for printing mechanically robust silicone structures. However, interfacial tension between the printed structures and the surrounding support material can lead to difficulties in achieving layer-to-layer adhesion.46 To investigate the ability to achieve both vertical and lateral adhesion when 3D printing silicone elastomer into micro organogels, we print both horizontal

sheets in the x-y plane (Figure 3-6a,b) and vertical sheets in the x-z plane (Figure 3-6d,e). Once

cured, we remove the sheets from the organogel support and serial wash. Visual inspection of the

horizontally printed sheet show no visual defects; similar inspections of the vertically printed

sheet find ridges in the surface that appear to correspond to the individual layers. Scanning white

light interferometry (SWLI) scans of the surface confirm these qualitative assessments; no visual

defects are observed in the horizontal SWLI scans while the vertical sheets show several sharp

valleys (Figure 3-6c,f).

While interfacial tension between the ink and support material can lead to issues in

adhesion between layers, it can also result in smooth surfaces as the interface evolves to a state

of minimum surface energy. SWLI scans of linear features printed from Momentive UV Electro

225 PDMS elastomer demonstrate that printed parts have surface roughness on the order of 150

nm (Figure 3-7a-c). Scanning electron microscopy (SEM) images of similar printed structures

reveal homogenous interiors of the printed structures with no mixing of the microgel and visually

smooth surfaces, consistent with the SWLI results (Figure 3-7e).

To investigate the mechanical integrity of 3D-printed silicone parts, “dog-bone” tensile test specimens are printed, cured, and removed from the organic support material by using the same protocols (Figure 3-7f). Extensional stress-strain tests on dog-bone samples demonstrate excellent mechanical integrity of 3D-printed silicone parts; 3D-printed silicone dog-bones were able to achieve approximately 700% strain before failure (Figure 3-7d,g). This physical robustness of the 3D-printed parts reveals that strong adhesion occurs between neighboring features during fabrication.

Application of Micro-Organogels for 3D Silicone Printing

The explore the application of micro-organogels as a support matrix in soft matter manufacturing, we 3D print a variety of silicone elastomer structures with varying size, complexity, and material. In one example, we print a model trachea implant out of room temperature vulcanizing silicone elastomer (Mold Max 10) (Figure 3-8a). Here, we fluctuate the diameter of the tube with increasing height. After letting the silicone cure, the structure is removed from the micro-organogel support. We measure a wall thickness of the structure to be

400 µm (Figure 3-8b). In a second example, we print a 3D scaffold structure out of a UV curing

silicone (Momentive UV Electro 225) with a sinusoidal pattern in both the x-y and x-z directions

with feature sizes on the order of 250 µm (Figure 3-8c,d). This structure was found to remain stable in the micro-organogel support but was too thin to be removed from the medium without destroying the scaffold.

To demonstrate the potential application of micro-organogels to 3D print silicone structure for biomedical applications, we create a perfusable, 3D network of hollow vessels out of Momentive UV Electro 224 silicone in which a single 25 mm diameter tube splits into six 3 mm diameter tubes (Figure 3-8e,f). After the structure has been cured, we remove it from the micro-organogel support and clean it through serial washing in a soapy bath. We find the structure is strong and flexible enough to connect pipe-fitting and pump fluid through all six openings at high flow rates (Figure 3-8g). In another example, we demonstrate the ability to manufacture silicone structures with multiple encapsulated parts in the micro-organogel support.

Here, we print a fluid pump out of Momentive UV Electro silicone containing two ball values to regulate the flow of fluid (Figure 3-8h). All three components are 3D printed in one continuous print job, cured, removed from the micro-organogel support, and washed to remove the micro- organgel. Through manual actuations of the lower chamber, we are capable of pumping fluid through the silicone pump from one reservoir to another, demonstrating the physical robustness of the printed structures (Figure 3-8i).

Conclusions

Microgels have been used for many years as model systems to study the fundamental physics of the jamming and glass transitions.43, 82-83 For an even longer period, dating back to the

1950s, microgels have been used in the industry to control the rheological behaviors of diverse products from engine oils to personal care products, such as lotions, shampoos, conditioners, and hand sanitizers.84-87 In contrast to aqueous microgels, organic microgel applications have remained limited by the challenges of microgel synthesis and swelling in nonpolar liquids. Here,

we have developed an oil-based microgel system by leveraging one of the most well-controlled and versatile tools in the soft matter research world: the self-assembly of block copolymers. We used commercially available components and demonstrated the tunability of the material’s rheological and interfacial properties. Our results demonstrate the ability of these microgels to meet the manufacturing requirements of soft silicone devices without significant constraints on ink formulation or fabrication machinery. Furthermore, we hope that this work will enable other researchers to investigate the huge diversity of possible block-copolymer chemistries, expanding the breadth of physical and chemical properties of micro-organogels for new technological applications, including improved soft matter manufacturing methods.

The minimum feature size, surface roughness of printed features, and the layer-to-layer spacing obtainable by printers are often used to describe the precision of rapid prototyping systems. Whereas the layer-to-layer spacing is driven by the mechanical precision of the printer, the minimum feature width and surface roughness are dependent on the material properties of the ink and support material; the appropriate measure of precision varies between different applications and the associated materials. Rapid prototyping systems are most frequently used to create large-scale, solid objects with marginal requirements for fine detail compared to precision machining, although increasing precision is a ubiquitous goal.50, 88 The level of precision

achieved here allows the fabrication of thin-walled tubes with a thickness of 450 mm that are robust enough to be handled and removed from the support material. Even smaller feature widths have been achieved, yet more robust materials and improved removal techniques are required if these delicate structures are to be removed from the support medium without failure.

3D bioprinting for tissue engineering has tremendous hurdles to overcome before implantable, functioning tissues and organs can be fabricated out of living cells.89 By contrast,

3D printing of implantable silicone devices is currently limited by the precision and strength of manufactured parts. The micro-organogel system developed here allows for the fabrication of devices suitable for biomedical applications that are robust enough to be handled, tested with standard industrial mechanical methods, and used in vigorous fluid pumping applications. The use of micro-organogel material to support printed parts allows multiple, nested components to be printed in a single step, opening the door for new methods and design strategies in additive silicone manufacturing. More generally, micro-organogels enable a wide breadth of potential applications where the unique rheology of the jamming transition is needed in oil-based systems.

Figure 3-1. Block copolymer self-assembly into micro-organogels. a) Organogel support materials are formulated with light mineral oil, polystyrene-block- polyethylene/propylene diblock copolymers, and polystyrene-block- polyethylene/butylene-block-polystyrene triblock copolymers (polymers drawn extended to illustrate their contour lengths). b) High concentrations of diblock copolymer results in a fluid phase of packed micelles, unable to support printed structures. c) Diblock micelles consist of polystyrene cores (red dots) surrounded by ethylene/propylene corneas. d) At high concentrations of triblock co-polymer, the support bath becomes globally crosslinked, and the printing nozzle causes permanent damage as it moves across the gel. e) Ethylene/butylene mid-blocks assemble into either crosslink “bridges” in which the polystyrene end-blocks are found in different cores, or into “loops” in which both polystyrene end-blocks are located in the same core. f-g) An equal blend of diblock and triblock copolymers results in closely packed microgels. Packed microgels provide a self-healing environment, allowing a printing nozzle to transverse the same region repeatedly while simultaneously supporting printed structures. h) The micro-organogels form when the diblock copolymers replace the triblock copolymers, reducing the number of ethylene/butylene bridges that form until the material is no longer a continuous network.47

Figure 3-2. Rheological Characterization of Micro-Organogels. a) Shear moduli measurements through small amplitude frequency sweeps show pure triblock and 50:50 blend solutions exhibit solid like behavior over long time scales. Pure diblock systems will behave like a viscous fluid. b) Unidirectional shear rate sweep measurements are performed to measure the yield stress of the block-copolymer systems. c) Thixotropic time measurements of the 50:50 blend show the recovery of the solid-like rheological properties within 1 s of removing the applied stress. d) Microscopic images taken with phase contrast illumination show the presence of microgels on the order of 2-4 µm in diameter.47

Figure 3-3. SAXS Characterization of Organogels. The average core-to-core spacing of organogel samples was found by determining the location of the first peak in the scattering intensity function from small angle x-ray scattering data.47

Figure 3-4. Effects of Elevated Temperature on Rheological Properties. a) Micro-organogels heated to 60 °C and higher behave as non-Newtonian viscous fluids, exhibiting a crossover in the elastic and viscous shear moduli. Measurements taken at 1 Hz and 1% strain. b) At 50 °C, the elastic modulus continues to dominate the viscous modulus; at 60 °C, we find a crossover in the moduli and the viscous modulus begins to dominate at low frequencies. c) We find similar results in the yield stress of the material; the yield stress remains relatively unchanged for temperatures up to 50 °C, but no discernable yield stress for materials heated to 60 °C and higher.47

Figure 3-5. Printing Precision of Silicone Elastomers into Micro Organogels. a) The width and height of printed silicone structures are measured using 3D fluorescence images. b) The width (a) and height (b) of printed structures can be predicted from the volume flow rate of the syringe pump (Q) and the tangential velocity (v) of the translating needle. c) Interfacial instabilities drive the breakup of neat silicone oil printed into the micro-organogel. The minimum feature size obtainable is controlled by the yield stress of the microgel support. d) The time in which a printed structure will break up can be increased by increasing the shear viscosity of the printed ink.47

Figure 3-6. Layer to Layer Adhesion of Silicone Elastomer Structures. a) Horizontal silicone structures are printed in the x-y plane to explore lateral adhesion between extruded filaments. b) Macroscopic image of the horizontal sheet being printed into the micro- organogel material. c) SWLI scans along the y-direction of the horizontally printed silicone sheets. d) Vertical silicone sheets are printed in the x-z plane to explore vertical adhesion of between layers. e) Macroscopic image of the vertical sheet being printed into the micro-organogel support material. f) SWLI scans along the z- direction of vertically printed silicone sheets.47

Figure 3-7. Surfaces and Mechanical Properties of Printed Structures. a-c), Scanning white light interferometry of a printed silicone surface show surface roughness of 150nm (A: 2D scan; b) slice along X-axis; c) slice along Y-axis). d) Stress-strain curve of printed silicone “dog-bone” specimens; printed silicone structures are capable of enduring upwards of 700% strain before mechanical failure. Tensile tests maintain a linear stress-strain relationship at low strains (inset). e) Scanning electron microscopy of the cross-section of a printed silicone structure demonstrates the uniformity of printed structures. f) Macro-photographic image of “dog-bone” specimen printed from silicone elastomer for tensile testing. g) Macro-photographic images of printed “dog-bone” structures in the relaxed and highly strained states.47

Figure 3-8. 3D-Printed Silicone Structure in Micro-Organogels. a) Model trachea implants are printed into micro-organogel supports using an RTV silicone. After curing, printed structures may be removed from the support material and handled. b) Cross- sectional views of model trachea demonstrate the ability to print structures with wall thickness of 400 µm. c-d) Silicone scaffolds printed with sinusoidal wave patterns in the x-y and x-z directions demonstrate the ability to print structures with 250 µm feature sizes. e-g) Macroscopic images of a perfusable tubular network printed into the micro-organogel support. After printing, the structure is cured through UV- irradiation, removed from the support, and connected to pipe fittings. The printed structure is robust enough to support high pressure fluid flow. h,i) A silicone pump containing encapsulated ball value is printed into the micro-organogel support in a single print job. After polymerizing and removing from the support material, water can be pumped through the valve through repetitive mechanical actuation of the bottom chamber.47

CHAPTER 4 INTERFACIAL INSTABILITIES IN JAMMED MICROGELS

Introduction

The boundary lines at the interface of three fluid phases will configure themselves into a state of minimum surface energy in which the contact angles of the interfaces can be found by solving Neumman’s triangle.90-91 Similarly, the contact angle between a liquid droplet and a rigid

solid substrate can be found by balancing the horizontal surfaces stresses through the Young-

90 Dupré equation γγls+= lv cos( αγ) vs . However, one can see that this simple force balance

between the interfacial terms neglects the out of plane contribution of the fluid-fluid interface;

these out of plane capillary forces can lead to elastic deformation of the substrate on the scale of

the elastocapillary length.92-94 The magnitude of this elastic deformation is set by the surface

stress acting on the surface and the modulus of the substrate such that λe = ϒ E . While the

elastocapillary effect is often negligible for rigid substrates, the deformation can be on the order

of micrometers to millimeters for soft substrates with moduli in the MPa to kPa range. In the

extreme case, in which the surfaces stresses exceed the yield stress of the soft solid, one could

predict the substrate to yield and undergo plastic deformation on the length of a ‘plastocapillary’

length λσpy= ϒ . While this plastocapillary length has been predicted, experimental exploration

of this length scale remains in its infancy.14, 95-96

Interfacial instabilities in jammed granular microgels have recently been explored by

printing PDMS toroids of varying size into jammed aqueous microgels and observing their

stability over time.15-16 Here, it was found the yield stress of the jammed microgels and the

interfacial tension between the ink and microgels determined the critical feature size for stability;

by increasing the yield stress of the microgels, smaller features are obtainable. This relationship

has been leveraged to 3D print silicone elastomer structures into aqueous microgel with high

yield stress; however, the large interfacial tension between the aqueous support and organic ink

resulted in large feature sizes and difficulty in achieving layer-to-layer adhesion.46 Alternatively,

interfacial instability may be avoided by choosing miscible sacrificial support material – printing

ink pairs.6 However, small interfacial mismatches may be beneficial in achieving consistent

cross-sectional area and preventing the exchange of solvents and other small molecules between

the printed ink and the support material. Here we explore the role of interfacial instabilities on

the deformation and yielding of 3D printed structures in immiscible support materials.

Results and Discussion

Rheological Characterization

To explore the role of interfacial instabilities in soft matter manufacturing, we prepare

aqueous microgels samples using Ashland 980 carbomer across a range of polymer concentration

to achieve a range of yield stress and moduli. The yield stress of the microgel systems are

measured through unidirectional shear rate sweeps with applied shear rates between 500 – 0.01 s-

1. At high shear rates, the microgel systems exhibits a shear thinning behavior in which we

observe a non-linear increase in measured shear stress with increasing shear rate (Figure 4-1a).

At the limit of low shear rates, the shear stress plateaus to a finite stress, independent of the

applied shear rate. This plateau stress corresponds to the yield stress of the system and can be

τ= τ + γγ p measured by fitting the Hershel-Bulkley model yc(1 ( ) ) to the data where τ is the

measured shear stress, τy is the measured yield stress, γ is the applied shear rate, γc is the

crossover shear rate, and p is a dimensionless constant.64 The elastic and viscous shear moduli of

the aqueous microgels are measured using small amplitude frequency sweeps from 101 – 10-2 Hz

(Figure 4-1b). For all polymer concentrations explored here, the elastic shear remains relatively

flat and dominates the viscous shear modulus over the full range of frequencies, behavior

consistent with an elastic solid.

Here we explore the interfacial effects using three different oil-based inks, each with their own unique rheological properties. Neat mineral oil behaves like a Newtonian fluid in which the shear stress increases linearly with increasing shear rate; here we measure the viscosity of the neat mineral oil to be 46 mPa s (Figure 4-1c). By contrast, packed micelles made through the self-assembly of block copolymers in light mineral oil exhibit shear thinning behavior in which the viscosity decreases with increasing shear stresses (Figure 4-1c). Small amplitude frequency sweeps of the packed micelle samples show a crossover in the elastic and viscous shear moduli at low frequencies, consistent with this fluid-like behavior (Figure 4-1d). Jammed organic microgels exhibit solid-like behavior over the full range of frequencies and have a measurable yield stress corresponding with a plateau in the shear stress at low applied shear rates (Figure 4-

1c,d).

Interfacial Instabilities of Fluid Structures

To understand the role of interfacial instabilities in soft matter 3D printing, we consider the competing stabilizing and destabilizing forces acting on a fluid beam of radius r printed into an immiscible jammed microgel support bath with interfacial tension γ, and yield stress τy. Here, we expect the destabilizing interfacial stresses acting along the fluid beam to be in competition with the stress necessary to yield the microgel support surrounding the fluid beam. This

2 hydrodynamic area in which the microgels will yield can be approximated as Arhh= 2π ,

5 where the hydrodynamic radius ( rh ) is twice the radius of the structure (r). A simple force

balance between the interfacial force, Frγ = 2πγ, and the force necessary to yield the microgel

2 support material, Frmicrogel= πτ h y , predicts an inverse relationship between the radius of the

printed beam and the yield stress necessary to overcome the interfacial instabilities, that is:

τγ= rc .

To test this stability prediction, we print lines of neat mineral oil with increasing cross-

sectional areas into jammed aqueous microgel with varying yield stresses and observe the

stability of the fluid beam over time (Figure 3-2a). Fluid beams with small cross-sectional areas

are observed to break up into droplets, reminiscent of the Rayleigh-Plateau instability observe in

fluid jets. As we increase the cross-sectional area of the fluid beams, the resulting droplets begin

to take on a non-spherical orientations and eventually a transition from the break-up regime to a stable state in which no break-up is observed (Figure 4-2b). We generate stable vs break-up state diagrams for microgel supports with different yield stresses and find the boundary between stable structures and break-up shifts to higher values of 1/2rc with increasing τy. Furthermore, we

find the state boundary separating these two states scales linearly, consistent with our predicted

scaling between the interfacial stress and yield stress of the material (Figure 4-2c). We measure

the slope of the boundary between the stable and unstable regimes and find the interfacial tension

between the mineral oil beams and aqueous microgel support to be γ = 18.98 mN/m. While this

measurement is in good agreement with interfacial tension values reported in the literature,97 our measurements of interfacial tension between water and light mineral oil using the Pendant drop method find γ = 35.04 mN/m.

Interfacial Yielding of Soft Solids

To explore the role of interfacially driven plastic deformation of soft solids, we consider a solid beam of organic microgels printed into an immiscible support material. As with the liquid structures, we expect the interfacial forces acting on the solid beam driving the instabilities to be in competition with forces necessary to yield the surrounding microgel support. Here we

introduce a second stabilizing force preventing the break-up of the printed beam in the yield stress of the organic microgels; the interfacial forces acting on the beam must now overcome the yield stress of the solid beam before plastic deformation and failure can occur. We approximate

2 the force necessary to yield the solid beam as Frby= πσ, where σy is the elongation yield stress of the inner beam. For Hershel-Bulkley materials, the elongation yield stress can be related to the

98-99 shear yield stress by στyy= 3 .

To explore this additional stabilizing force, we 3D print beams of jammed organic microgels (τy = 5.2 Pa; σy = 9.0 Pa) into aqueous microgels with increasing yield stress and observe their stability over time (Figure 4-3a). We observe similar behavior in the jammed organic microgels as that seen in the neat mineral oil beams; jammed organic microgel beams above a critical feature size remained stable whereas smaller structures underwent plastic deformation and yielded into smaller drops (Figure 4-3b). We find this transition between stable and break-up states scales linearly between the yield stress of the supporting microgel bath and

1/2r (Figure 4-3c). We measure the apparent interfacial tension between the organic microgel beam and the aqueous microgel support by fitting a line to the boundary of the stable and unstable regime and find γ = 12.23 mN/m. Additionally, we observe an offset in the critical feature size for stability at the limit of τy = 0 Pa for the surrounding support material; this offset corresponds to the interfacial tension necessary to plastically yield the jammed organic microgels beam. By balancing the interfacial forces with the force necessary to yield the microgel beam,

4γ we predict the yield stress of the organic microgels as σ = . Using this prediction, we y 2r measure the yield stress of the organic microgel beam to be 50.3 Pa, approximately 5.5 times larger than rheologically measured yield stress of the organic microgels.

We are intrigued by the apparent change in the interfacial tension between the organic and aqueous microgels and hypothesize this change originates from the introduction of block copolymers into the oily phase. To explore these effects, we prepare packed micelle swollen in light mineral oil with similar polymer concentrations as our jammed organic microgels and print structure of increasing feature size into our aqueous microgel support material (Figure 4-4a,b).

Unlike the jammed organic microgel system, the packed micelles behavior is fluid-like over long time scales as evident by the crossover of the elastic and viscous shear moduli in small amplitude frequency sweeps and the purely viscous response to applied shear rates with no measurable yield stress. These difference in rheological behavior allow us to decouple the effects of a yield stress from the addition of block copolymers in the printed beams.

We find that the critical feature size of packed micelle beams follows a similar trend to the neat mineral oil, exhibiting a linear relationship between the support material yield stress and the 1/2r, with no offset in critical feature size at the limit of no support yield stress (Figure 4-4c).

Similar to the jammed organic microgel system, we find the apparent interfacial tension between the packed micelles and the aqueous microgel support as measured by the slope of the state boundary to be 10.65 mN/m. Thus, we suspect the changes in apparent interfacial tension seen in both the jammed organic microgels and the packed micelle systems to arise from the addition of block copolymers to the ink and not the presence of a finite yield stress. We hypothesize that the polystyrene cores act as a stabilizing agent at the oil-water interface in a mechanism similar to the Pickering emulsion in which nano-particles can stabilize the interface of emulsion droplets.81

Interfacial Buckling of Soft Solids

Solid beams with large aspect ratios will buckle when subjected to a sufficient large axial load. We hypothesize that similar buckling instabilities may occur when a soft solid is submerged in an immiscible support material. Here, the force driving the deformation of the

solid beam arises from interfacial tension between the solid beam and the support material. To examine this hypothesis, we print beams of jammed aqueous microgels into a support bath consisting of packed micelles (Figure 4-5a).

Similar to our previous findings, printed structures smaller than the critical feature size will undergo plastic deformation and break into smaller droplets with a linear relationship between the yield stress and 1/2r (Figure 4-5b). When the structure is sufficiently large, we observe stable behavior and no evolution in the shape over time. However, between these two extremes, a new instability immerges in which the printed beams appear to buckle at their ends as they axially contract inward, resulting in a ‘barbell’ shape. We find this behavior present in our jammed aqueous microgel beams across all concentrations and yield stresses explored.

We are intrigued by this behavior and hypothesize that the buckling follows classical

Euler-Bernoulli beam theory. To test this hypothesis, we print beams of aqueous microgels of varying yield stress and measure the resulting wavelength of the deformation. In Euler-Bernoulli beam theory, the wavelength of a solid beam buckling in an elastic medium is given as:

14 EI λπ= 2 b , Es where λ is the wavelength of the buckled beam, Eb is the elastic modulus of the embedded beam,

nd Es is the elastic modulus of the elastic medium, and I is the 2 moment of area of the beam; for a

π r 4 circular cross sectional beam with radius r, I = . Surprisingly, we find that the wavelength of 4 the microgel beam deformation is inversely related to the yield stress of the microgels; an increase in the yield stress results in a decrease in the measured wavelength (Figure 4-5c). The yield stress of jammed microgels have been shown to be linearly related to the elastic shear

modulus,63, 100 thus this observed behavior is in direct contrast with that predicted by Euler-

Bernoulli beam theory in which the wavelength scales like λ ∝ E14.

An alternative explanation of the observed behavior is that the surface of the jammed

microgel beam is undergoing a wrinkling instability in which the wavelength decreases with

increasing stiffness of the substrate. To test this hypothesis, we examine the behavior of the top

and bottom line profiles of the microgel beam as it undergoes deformation. For a wrinkling

instability, we expect the perturbations in the line profiles to be out of phase, whereas the line

profiles of a buckling instability will be in phase with one another (Figure 4-6a). The top and

bottom line profiles are measured by fitting Gaussian curves to the intensity profiles gathered

from macroscopic images of the microgel beams during the deformation process (Figure 4-6b).

A qualitative comparison between the top and bottom line profiles after the onset of deformation

suggests the deformations to be in phase with one another (Figure 4-6c).

To quantitatively confirm this observation, we compute cross-correlation functions of the

top and bottom line profiles for each time point (Figure 4-6d). Again, when we observe the

cross-correlation between the top and bottom line profiles at time points after the onset of

deformation, we find the two profiles to be in phase with one another as evident by a peak in the

cross-correlation function near τ = 0 (Figure 4-6e). These results strongly suggest that the

deformation seen in the jammed microgel beam is not a wrinkling instability, but rather some

form of transient buckling of the beam as it axially contracts.

Under certain conditions, a thin thread of a viscous fluid, such as honey, will undergo a

transient buckling instability driven by the balance of viscous, inertial, and gravitational

forces.101-103 Solid rope will undergo a similar coiling instability in which the viscous effects are replaced with the elastic behavior of the solid coil.104-105 In both these instances, a transient

buckling behavior is observed that remains localized at the end of the thread. We hypothesize a

similar behavior is occurring in our aqueous microgel beams, resulting in the buckling instability

observed. For our system, the microgel beams are printed horizontally, thus we eliminate

gravitational forces from consideration. To determine if the buckling instability is dominated by

viscous or elastic behavior, we measure the speed at which the microgel beams contract with

various initial feature sizes and rheological properties (Figure 4-7a). Surprisingly, we find very

little variation in the speed between samples with different initial feature sizes and yield stresses;

typical speeds range from 40-50 µm/s during the initial contraction to 1-10 µm/ followed by a steady decay from 10 µm/s to 1 µm/s. At these low speeds, the inertial contributions to the buckling instability are negligible. From our measurements of speed, we calculate the corresponding shear rates acting on the surface at various time points along the microgel beam

(Figure 4-7b). The shear rates and corresponding shear stresses suggest the microgel beam remains in the elastic regime. However, the calculated shear stresses are roughly 2 orders of magnitude lower than the yield stress of the microgel beam. Thus, continued exploration of the interfacial, viscous, and elastic forces acting on the microgel beam are necessary to understand the observed buckling phenomenon.

Conclusions

Here we have investigated the interfacially driven deformation of 3D printed structures into jammed granular microgel supports. The stability of the 3D-printed objects is governed by a

balance between the destabilizing interfacial forces and the stabilizing yield stress of the

surrounding microgel support. For a given yield stress, we find the critical feature size necessary

to maintain stability; this critical feature size is inversely related the yield stress of the microgel

support. When the printed structure has a yield stress, the destabilizing interfacial forces are now

balanced by the stabilizing forces of both the inner and outer yield stress, resulting in an offset in

the stability phase diagram. For sufficiently small structures, we observed interfacially driven plastic deformation of the solid beam. Understanding these instabilities and how they relate to the feature size are necessary to understand the limitation of soft matter manufacturing.

When printing aqueous microgel beams into an oily packed micelle support material, we observed an additional destabilizing phenomenon in which the beams exhibited a transient buckling behavior and under interfacial forces analogous to liquid rope buckling. The wavelength of the buckling instability does not appear to follow the classical Euler-Bernoulli beam theory prediction for wavelength. Further work is necessary to understand the onset of this buckling phenomenon and predict its behavior. One potential path forward is to compare the driving interfacial forces to the resisting viscous drag. Here, a simple balance between these

stresses would predict σγ = σσv + ys.., where σγ is the interfacial stress, σv is the viscous stress, and σy.s. is the yield stress of the microgel beam. In this instance, we could determine the viscous stresses on the beam by computing the viscous drag acting on a cylinder per unit length as given by

πη FL= 4 v , ln( 7.4 Re) where Re = ρηvr is the Reynold’s number, v is the velocity of the cylinder, η is the viscosity of the surrounding medium, and r is the radius of the cylinder.106

Figure 4-1. Rheological Characterization of Interfacial Systems. a) Jammed aqueous microgels are prepared with varying yield stresses by increasing the polymer concentration. b) All microgel samples exhibit solid like behavior over the full range of frequencies. c) Unidirectional shear rate sweeps of oil systems exhibit fundamentally different responses; neat mineral oil behave like a Newtonian fluid whereas packed micelles exhibit shear-thinning behavior and jammed microgels have a finite yield stress at low shear rates. d) Small amplitude frequency sweeps of packed micelles have a crossover in the elastic and viscous shear moduli. In contrast jammed microgels maintain their solid-like behavior at the limit of low frequencies.

Figure 4-2. Interfacial Instabilities of Neat Mineral Oil. a) Lines of neat mineral oil are 3D printed into support baths of jammed aqueous microgels with increasing yield stress. The cross-sectional area of the printed lines are controlled by changing the translational speed (vn) of the printing nozzle and the volumetric flow rate of the printed material (Q). b) Lines with small cross-sectional area undergo a break-up instability, whereas larger lines remain stable. c) Phase diagram of stable vs unstable features with increasing yield stress of the aqueous microgel support. The boundary between the stable and unstable regime corresponds to an apparent interfacial tension of 18.98 mN/m.

Figure 4-3. Interfacial Instabilities of Jammed Organic Microgels. a) Lines of jammed organic microgels (σy = 9.0 Pa) are 3D printed into support baths of jammed aqueous microgels with increasing yield stress. The cross-sectional area of the printed lines are controlled by changing the translational speed (vn) of the printing nozzle and the volumetric flow rate of the printed material (Q). b) Lines with small cross-sectional area undergo a break-up instability, whereas larger lines remain stable. c) Phase diagram of stable vs unstable features with increasing yield stress of the aqueous microgel support. The slope of the boundary between the stable and unstable regime corresponds to an apparent interfacial tension of 12.23 mN/m. Additionally, we observe an offset in the critical feature size for stability, resulting from the solid-like behavior of the jammed organic microgel beam.

Figure 4-4. Interfacial Instabilities in Packed Micelle Beams. a) Lines of oil-based packed micelles are 3D printed into support baths of jammed aqueous microgels with increasing yield stress. The cross-sectional areas of the printed lines are controlled by changing the translational speed (vn) of the printing nozzle and the volumetric flow rate of the printed material (Q). b) Lines with small cross-sectional area undergo a break-up instability, whereas larger lines remain stable. c) Phase diagram of stable vs unstable features with increasing yield stress of the aqueous microgel support. The boundary between the stable and unstable regime corresponds to an apparent interfacial tension of 10.65 mN/m.

Figure 4-5. Buckling Instabilities in Aqueous Microgel beams. a) 3D-printed beams of aqueous microgels into organic packed micelle supports will transition from a break- up to buckling to stable regime with increasing cross-sectional area. b) State diagrams capture the behavior of 3D-printed beams at various cross-sectional areas and with increasing yield stress. c) Wavelength of the buckled beams decreases with increasing yield stress.

Figure 4-6. Wrinkling vs Buckling Instability in Microgel Beams. a) The top and bottom line profiles of a beam undergoing a wrinkling instability will be out of phase with one another, whereas the line profiles will be in phase if the beam is undergoing a buckling instability. b) To determine if our jammed microgel beams are wrinkling or buckling, we measure the top and bottom line profiles from macroscopic images of a deforming beam. c) Qualitative comparison of the line profiles after deformation show similar beaks and valleys, suggesting the profiles are in phase with one another. d) We compute the cross-correlation between the top and bottom line profiles and plot their intensity for each time point. e) Examining the cross-correlation between the top and bottom line profiles (Ctb) show a peak in intensity near τ = 0, indicating the line profiles are in phase with one another.

Figure 4-7. Shear Rate Analysis of Beam Buckling. a) The speed at which the microgel beams contract appears to be independent of the initial feature size and the yield stress of the microgel beam. b) From our velocity profile and measurements of the beam diameter, we calculate the shear rate at various locations along the microgel beam (x) at 3 different time-points. We find the shear rate increases as you move out from the center of the beam, followed by a plateau in the shear rate at the critical location of buckling (x*). However, we find the shear rate at which the buckling instability occurs is not constant. Mapping these shear rates to our rheological data, we find the shear stresses to be of order O(10-1), roughly two orders of magnitude lower than the yield stress of the microgel beam.

CHAPTER 5 POLYELECTROLYTE INTERACTIONS IN AQUEOUS MICROGELS

Introduction

Micro-scale hydrogel particles, commonly called microgels, can be packed together to form jammed solids that exhibit dominantly elastic responses to small levels of shear deformation.41, 83, 107-108 Unlike rigid granular particles, individual microgels are capable of large

deformations and volume changes, allowing for packing fractions to exceed the random close

packing limit of hard spheres.40-41 Most often, microgels are composed of charged polymers, which strongly drives their swelling and enables the use of extremely long chain lengths between crosslinks; most commercially available and laboratory synthesized microgels rely on anionic charge species to promote swelling and achieve jamming at relatively low polymer concentrations.38, 48, 60 While microgels have been used extensively for studying the role of

particle elasticity in the glass transition and jamming phenomena,108-111 they remain practically

untapped as a biomaterial with limited applications largely focused on drug delivery and cell

encapsulation.112-113

Recently, jammed microgels have been employed as sacrificial support materials for 3D printing structures made from a diversity of soft matter components including polymers, hydrogels, silicone elastomers, and living cells.6, 8, 45, 47, 114 In contrast to other sacrificial support

materials such as polymers with reversible bonds,7, 115 packed micelles,21 and buoyancy matched

salt solutions,22-23 jammed granular microgels exhibit solid-like behavior over long time scales, providing stability to 3D printed structures.5 When swollen in cell growth media, microgels can

be employed as a support material for 3D bioprinting, allowing for controlled, systematic studies

of micro-tissue and tumor development (Figure 5-1a).45, 116 Alternatively, individual cell

behavior may be examined through the dispersion of cells in this 3D cell culture medium (Figure

5-1b).117 However, cell culture media contains several monovalent and divalent cations including

Na+, K+, Ca2+, and Mg2+ which are involved in numerous cellular processes including cell-cell

adhesion, cell signaling, and regulation of internal osmotic pressure.118 While the sensitivity of

charged microgels to added salt is well understood in the case of monovalent ions such as Na+

and Cl-,63, 119-120 the changes in the rheological properties of packed polyelectrolyte microgels in

the presence of multivalent ions, such as those within cell growth media, have not been

thoroughly investigated. Understanding how multivalent ion – polyelectrolyte interactions

influence the rheological properties of microgel systems is necessary to guide the development

and application of new microgels for use in biomaterial systems.

Here, we synthesize anionic, cationic, and zwitterionic microgels and systematically

investigate the role of multivalent ion - polyelectrolyte interactions on the rheological properties

of microgel systems. We design polyelectrolyte microgels with either anionic (MAA), cationic

(qDMAEMA), or zwitterionic (CBMA) charged species at varying charge densities and study their changes in rheological properties in the presence of increasing concentrations of Ca2+ and

Cl- ions. In the high-salt limit, the rheological behavior of anionic and cationic microgels follow

polyelectrolyte scaling laws while rheological properties of zwitterionic microgels become

independent of added salt. In addition, we explore the application of these polyelectrolyte

microgels as biomaterials by culturing cells in microgel environments. Cell performance in the

presence of polyelectrolyte microgels is quantified through short-term viability, proliferation,

and metabolic activity assays. We find that cell performance is dependent on the chemical

composition of the microgels and remains relatively independent of the rheological behavior;

anionic and zwitterionic microgels have minimal effect on the short-term viability and metabolic activity of cultured cells while our cationic microgels appear cytotoxic. Thus, the rheological

properties of these jammed microgel packed can be tuned and optimized for bioprinting and 3D

cell culturing applications independent of cell performance.

Results and Discussion

Polyelectrolyte Microgel Synthesis and Preparation

To explore the effects of polyelectrolyte – ion interactions on the rheological properties of jammed granular microgels, we synthesize charged microgels composed of polyacrylamide

(pAAm) polymers containing ionized co-monomers crosslinked with 1 mol% poly(ethylene

-1 glycol) diacrylate (PEGda, Mn = 700 g mol ) through a precipitation polymerization in ethanol.

Each microgel system is prepared with a charged species at varying polymer charge density; here

we use methacrylic acid (MAA) for anionic microgels, quaternized 2-(dimethylamino)ethyl

methacrylate (qDMAEMA) for cationic microgels, and carboxybetaine methacrylate (CBMA)

for zwitterionic microgels (Figure 5-1d).

Phase contrast micrographs of dilute samples confirm the formation of microgel particles

for each of the three charged species explored here (Figure 5-2). We measure the average size of each microgel in the dilute state and find the mean particle diameters to be 4.76±1.49 µm for

MAA microgels, 5.17±1.94 µm for qDMAEMA microgels, and 5.21±2.14 µm for the CBMA microgels (± intervals correspond to one standard deviation about the mean).

Rheological Characterization of Polyelectrolyte Microgels

To characterize the rheological properties of our polyelectrolyte microgels, we prepare sample at varying polymer concentration in ultrapure water and record the shear stress response to unidirectional shear rate sweeps (Figure 5-3a). At polymer concentrations below the jamming concentration, the shear stress exhibits a purely shear-thinning response to an applied stress reflecting dominantly fluid-like behavior (Figure 5-3a,b). As the polymer concentration is increased above the jamming concentration, the shear stress develops a plateau at low shear

rates, which corresponds to the yield stress (Figure 5-3a,b). This yield stress of the jammed microgels can be measured by fitting the Hershel-Bulkley model to the unidirectional shear rate sweep data:

τ= τ + γγ p yc(1 ( ) ) ,

where τ is the measured shear stress, τy is the yield stress, γ is the applied shear rate, γc is the cross-over shear rate, and p is a dimensionless constant.41, 64̇ We find the jamming concentrationṡ

of the microgels to be 0.3, 0.45, and 0.9 wt% for microgel particles containing 17 mol% of

MAA, qDMAEMA, and CBMA, respectively.

The yield stress of microgel packs may be tuned through changes in either the polymer

charge density of the microgels, or in the overall polymer concentration. As we increase the

polymer charge density at a fixed polymer concentration, the yield stress increases (Figure 5-3d).

Likewise, increasing the total polymer concentration results in an increase in the measured yield

stress (Figure 5-3c). The shear thinning behavior of jammed microgel packs at high shear rates is

captured by the dimensionless constant p in the Hershel-Bulkley model. Here, we find that p covers a range between 0.4 and 0.55, consistent with previously reported values for similar systems.83 Interestingly, we find that this shear thinning exponent follows a weak logarithmic

scaling with the measured yield stress, independent of the monomer charge species present

(Figure 5-4a).

To understand the role of salts on the rheological performance of jammed microgels, we

prepare microgel samples with increasing concentrations of calcium chloride, a prevalent

multivalent salt in cell growth media. As we increase the ionic charge concentration, we observe

a decrease in the measured yield stress; for anionic and cationic microgels prepared with high

concentrations of added salt, we observe a purely shear-thinning response to an applied shear

rate suggesting the system is no longer in a jammed state (Figure 5-5a). Similar rheological

behavior is exhibited in microgels with varying charged species, polymer charge densities, and

polymer concentrations (Figure 5-5b).

To compare rheological behavior across the various samples, we normalize the data by

plotting the ratio of the yield stress at a given salt concentration, τy, to the yield stress with no

0 added salt, τy , versus the molar ratio of added ionic charge to polymer charge. For anionic

microgels we consider the ionic charge of added Ca2+ ions; for cationic microgels we consider

the ionic charge of added Cl- ions. We find that cationic and anionic microgels collapse to a

single curve in which the yield stress decreases with increasing molar charge ratio (Figure 5-4b;

Figure 5-5c). As the total charge of added ions approaches the number of polymeric charges

(molar charge ratio = 1:1), the yield stress decreases to approximately 10% of the initial, zero-

0 salt yield stress, τy . Further increasing the molar charge ratio leads to a continued decrease in the measured yield stress until a yield stress can no longer be measured and the system exhibits a purely viscous response to an applied shear rate. Surprisingly, zwitterionic microgels exhibit an

initial decrease in the measured yield stress followed by a plateau at molar charge ratios greater

than 1:1 (Figure 5-4c; Figure 5-5d). This initial decrease in measured yield stress with increasing

salt concentrations deviates from the antipolyelectrolyte effect observed in charge neutral

polyzwitterions, in which the addition of low molecular weight salts results in an increase in

swelling.121 This deviation likely arises from the multivalent nature of Ca2+ investigated here.122

For the cationic and anionic microgels, we investigate the underlying mechanisms that drive the

changes in rheological properties with increasing ion concentration by applying polyelectrolyte

gel scaling laws to jammed microgel systems, below.

Polyelectrolyte Scaling Laws

Neutral hydrogels swell to an equilibrium concentration in which the driving osmotic

pressure (Π) generated from the random motion of the polymer chains is balanced by an elastic

65 restoring force (Fel) generated from stretching the polymer chains. In polyelectrolyte hydrogels,

dissociated counterions near the polymer backbone contribute to the osmotic pressure in addition

to the polymeric contribution. When fully swollen, the elastic shear modulus of the

polyelectrolyte gel, Gʹ, is approximately equal to the osmotic pressure, which can be expressed

in terms of a polymeric contribution and an ionic contribution, given by

1 c2 G′ = kT + p , B ξ 3 + A( cp4 Ac ion )

where kB is the Boltzmann constant, ξ is the polymer mesh-size, cp is the polymer

concentration, A is the average distance between uncondensed charges, and cion is the

123 concentration of added salt. At the high salt limit in which cp >> 4Acion, the ionic contributions

dominate the polymeric contributions. Here, we assume our polyelectrolyte microgels are fully

ionized, and therefore the ratio of cp /A can be approximated as the polymer charge concentration

(ccharge). Thus, at the low-polymer, high-salt limit where the ionic contribution dominates the

polymeric contribution, we can simplify this expression to

2 ccharge G′ = kTB . 4cion

Previous investigations have found polyelectrolyte theory may be applied to microgel

systems at concentrations just above jamming.63 Here, we measure the elastic and viscous shear

moduli, Gʹ and Gʺ, of the jammed microgels through low amplitude frequency sweeps spanning

a range between 10-3 and 101 Hz (Figure 5-6a). We find that the elastic component remains

relatively flat and dominates the viscous component across the full range of frequencies,

consistent with viscoelastic solid behavior. Similar to the yield stress of microgels, the elastic

shear modulus decreases with increasing concentrations of calcium chloride. Comparing these

two rheological properties, we find that the yield stress of the microgels and the elastic shear

modulus at 1Hz are linearly correlated such that

τ y = 0.14G′

for all samples exhibiting a yield stress under all conditions reported here (Figure 5-6b). This relationship between the yield stress and elastic shear modulus of the jammed microgels arises from the elastic energy necessary to deform frictionless particles as they rearrange and slide past one another and is consistent with previous results, indicating that interfacial polymer interactions like entanglements do not dominate yielding in the systems tested here.63 Thus, we

can relate the polyelectrolyte scaling laws for Gʹ to the yield stress of the microgels, given by

2 ccharge τ yB= 0.14kT . 4cion

To determine if this polyelectrolyte scaling law prediction for added salt applies to

0 jammed granular microgels, we plot the normalized yield stress of the microgels, τy /τy as a

function of the inverse of the ionic charge concentration. In the high salt limit, where 1/cion is small, we find the normalized yield stress follows the predicted scaling behavior for polyelectrolytes with added salt for both the anionic and cationic microgels at various polymer concentrations and charge densities (Figure 5-6c). At the low salt limit, where 1/cion is large, the data deviates from this predicted scaling behavior and we observe a plateau in the normalized

* yield stress. We determine a critical ionic charge concentration (c ion) in which this transition

from the high salt, polyelectrolyte scaling behavior to the low salt, plateau regime is observed.

* Rescaling the rheological curves by c ion, we find the data collapses to a single curve in which

the yield stress follows polyelectrolyte scaling at high salt concentrations (Figure 5-6d). This rheological behavior is observed for both the addition of divalent counter ions to anionic microgels (MAA) and monovalent ions to cationic microgels (qDMAEMA). Furthermore, we find the critical ionic charge concentration scales linearly with the charge concentration of the microgel particles and is in close agreement with the limits of the high-salt regime as predicted by polyelectrolyte theory.123 We further test our hypothesis that the rheological response can be

predicted by polyelectrolyte theory by examining anionic microgels in the presence of a

monovalent salt (NaCl) and find similar scaling behavior. Similar attempts to rescale the yield

stress of zwitterionic microgels do not show the same scaling behavior. At high added salt

concentrations, the zwitterionic microgels exhibit a plateau in the yield stress. At low added salt

concentrations, the yield stress follows a scaling behavior weaker than that predicted by

polyelectrolyte theory (Figure 5-6e,f).

Previous investigations show that microgels follow polyelectrolyte scaling laws for

polymer concentration within a small range of concentrations above jamming.63 At higher

concentrations, jammed microgel behavior deviates from that of fully swollen gels and follows

the scaling laws of rubber elasticity.40, 63 Here, we observe similar behavior in our anionic and

cationic microgels. As the polymer concentration of the microgels increases above the jamming

concentration, individual microgel particles will deform without osmotically driven deswelling,

resulting in volume fraction in excess of the random close packing limit of hard spheres.41, 109

However, increasing the concentration of added salts appears to drive deswelling of the microgel

particles, leading to a transition of the system from the jammed to unjammed state. In this limit

of high added salts, the decrease in the ionic contribution to the osmotic pressure can be

predicted by polyelectrolyte scaling laws for added salts. We find that this scaling law is

dependent on the ionic charge concentration of the added salt and appears independent of the ion valency. Similar behavior has been reported for polyelectrolyte brushes in the presence of mono- and multivalent ions.124-125 At low salt concentrations, polyelectrolyte brushes are in an “osmotic brush” regime in which the uncondensed counterions contribute to the driving osmotic pressure and the brushes maintain an extended configuration; as the concentration of added salts increases, the polyelectrolyte brushes transition to a “salted brush” regime in which a decrease in the osmotic pressure leads to a collapse of the brush structure.124 Zwitterionic microgel behavior deviates from these polyelectrolyte theory scaling predictions, suggesting the continued contribution to the osmotic pressure, even in the limit of high added salts.

Rheological Characterization of Uncharged Microgel

As we have shown, the changes in rheological properties of polyelectrolyte microgels follows polyelectrolyte scaling laws for added salt at the limit of high salts; electrostatic screening from added salts reduce the ionic contribution to osmotic pressure driving the swelling.

For an uncharged hydrogel, the driving osmotic pressure is solely dependent on the polymeric

KT contribution, that is Π= b , where ξ is the correlation length. Thus, we expect minimal effects ξ 3 of added salts on the rheological performance of uncharged microgels.

To test our prediction, we synthesize uncharged pAAM microgels containing 2 mol%

N,N-Methylenebisacrylamide (BIS) crosslinker through a precipitation reaction and characterize their rheological performance at various concentrations through unidirectional shear rate sweeps.

Similar to our polyelectrolyte microgels, the yield stress of uncharged microgels increases with increasing polymer concentration (Figure 5-7a). We note that above the jamming concentration, the yield stress follows similar scaling law behavior as polymeric hydrogels with increasing

94 126 polymer concentration, that is τ y ∝ c . To explore the effects of added salt on the rheological

performance of uncharged microgels, we prepare a microgel samples with CaCl2 salt and

measure the yield stress. As expected, we observe relatively no change in the measured yield

stress of the uncharged microgels when prepared with no added salt (0mM CaCl2) and at high

levels of added salt (250mM CaCl2) (Figure 5-7b).

Cell Viability in Polyelectrolyte Microgels

To explore the applicability of polyelectrolyte microgels in biomaterial applications such

as 3D bioprinting and 3D cell culture, we investigate the behavior of human mammary epithelial

cells (MCF-10A) in microgel packs swollen in cell growth media and determine cell health and performance through measurements of short-term cell viability, cell proliferation rate, and cell metabolic activity. For these experiments, microgel samples are swollen in cell growth media at a polymer concentration of 4 wt%. At these polymer concentrations, all microgels behave like elastic solids when swollen in cell growth media having shear moduli between 2 Pa and 60 Pa

(Figure 5-8).

Short-term cell viability measurements are performed on cells dispersed in microgel growth media after 3 h and 24 h to determine the population of viable cells (Figure 5-9a;

Appendix B). We find anionic and zwitterionic microgels maintain 90-95% cell viability after 24 h. These results are consistent with previous investigations of cell viability in microgel packs.45,

127 P(qDMAEMA) is a known cytotoxin, often used as an antifouling agent.128-129 Accordingly, for cells cultured in qDMAEMA microgels we observe increased levels of cell death; while microgels containing 5 mol% qDMAEMA maintain 75% cell viability after 24 hours, microgels containing 9 mol% and 17 mol% qDMAEMA exhibit greater than 70% and 90% cell death after

24 h, respectively. While these results demonstrate that microgels containing qDMAEMA do not constitute a suitable cell culture environment, they may find use in biomaterial applications where targeted cytotoxicity is desired.

Cell Proliferation in Polyelectrolyte Microgels

To investigate cell proliferation in microgel media relative to standard conditions in liquid media, we culture plated MCF-10A cells in microgel culture media and measure the cell population density at the 0 h, 24 h, and 48 h time points (Figure 5-9b; Appendix B). After 24 h, the percent population change of cells cultured in zwitterionic microgel media containing 17 mol% CBMA matched that of liquid media, while both 5 mol% and 17 mol% anionic microgels exhibit a small but statistically insignificant decrease in cell density. We note that this apparent decrease in cell population is consistent with short-term viability results. After 48 h, we observe a considerable deviation in population changes between cells cultured in liquid media and cells cultured in microgel media; cell population changes in microgels are lower than in liquid media, but statistically the same as earlier time-points. Fluorescence images of cells cultured in 17 mol% CBMA microgels for 48 hours reveal the formation of discrete cell islands similar to those seen in cells cultured in liquid growth media (Figure 5-10). In contrast, cells cultured in both 5 mol% MAA and 17 mol% MAA form smaller clusters (Figure 5-10). While this result shows that anionic and zwitterionic microgel media may not suitable for contexts where cell growth is needed, arresting cell proliferation may be advantageous in applications where rapid cell growth is undesirable, particularly if cells remain viable and metabolically active. For example, a reduced proliferation rate may be beneficial in the future when 3D printing tissues with complex vasculature requiring the precise placement of multiple cell types over long working times.

Cell Metabolic Activity in Polyelectrolyte Microgels

To study how microgel-based biomaterials may affect the metabolic activity of cells, we measure adenosine triphosphate (ATP) levels in MCF-10A cells cultured in microgel media for

24 h using the CellTiter-Glo 3D assay. The CellTiter-Glo 3D assay utilizes a luciferase reaction that produces a luminescence signal proportional to the levels of intracellular ATP.130 For these

experiments, cells are plated into 96-well culture dishes and culture in microgel media for 24 h

(Appendix B).

We perform a series of control experiments to determine the effectiveness of the

CellTiter-Glo 3D assay to measure ATP levels in a representative set of our jammed microgel systems. To determine the effectiveness of the luciferase reaction in the microgels, we prepare microgels and liquid media with known concentrations of adenosine 5′-triphosphate disodium salt (Figure 5-11a). We find the luminescence intensity to be linearly correlated with ATP concentration for our anionic and zwitterionic microgels across the concentration range tested here. Furthermore, the luminescence intensities measured in the microgel media is consistent with the luminescence intensity of ATP measured in liquid cell growth media. These results suggest the luciferase reaction is not hindered by the presence of the polyelectrolyte microgels.

To test the effectiveness of the assay’s detergent in lysing the cells cultured in a microgel environment, we dye MCF-10A cells with Cell Tracker Green CMFDA, a live-cell dye, and culture the cells in microgel media containing 1.1 µM ethidium homodimer, a dead-cell dye that enters the membrane after cell lysis. Fluorescence images taken 30 min after the addition of the

CellTiter-Glo 3D assay show an uptake in ethidium homodimer entering the membrane of the cells and a decrease in the CMFDA signal (Figure 5-11b). These results qualitatively confirm that the CellTiter-Glo 3D detergent successfully lyses the cells in the microgel environment. To quantify the effectiveness of this process, we compare luminescence intensity values of cells cultured in identical conditions but assayed in different media; cells are cultured under identical culture conditions to produce comparable metabolic activity and the microgel culture media is introduced immediately prior to measuring the ATP of the cell (Appendix B). Surprisingly, we find a drastic decrease in the luminescence signal produced by the CellTiter-Glo 3D assay in

comparison to the signal produced by cells cultured in liquid media (Figure 5-11c). We find a

25% and 50% decrease in cell luminescence intensity when measured in our 5 mol% and 17

mol% anionic microgels, as well as a 55% decrease in intensity when measured in our 17 mol%

zwitterionic microgels. These decreases in luminescence intensity occur despite the cells being

cultured under identical conditions. Thus, we attribute this decrease in luminescence intensity to

the effectiveness of the assay reagents to react with the ATP released by the lysed cells, rather

than the cell metabolic activity. Our findings highlight the importance in establishing the proper

baseline measurements when developing future assays to measure cell activity in microgel

environments.

To explore the changes in metabolic activity of cells in a microgel environment, we

culture cells for 24 h in microgels with varying charged species and polymer charge density. We

find that the relative luminescence intensity decreases by 30%, 50% and 60% relative to cells

cultured in liquid growth media for both the anionic and zwitterionic containing 5 mol%, 9

mol%, and 17 mol% polymer charge density (Figure 5-11d). However, when we determine the

ATP levels corrected by calibration, and the control experiments of luminescence intensity, and

accounting for proliferation, we find the adjusted relative ATP produced remains statistically

unchanged in comparison to liquid growth media (Figure 5-11e). Thus, we conclude the short-

term metabolic activity of MCF-10A cells cultured in the microgel environment is unhindered by the presence of polyelectrolyte. Further studies are needed to investigate the long-term viability

and metabolic activity of cells cultured in polyelectrolyte microgels.

Conclusions

Jammed granular microgels allow for 3D printing of fluids, polymer solutions, and cells with few constraints on solidification time or rheological properties of the “ink.”6, 45 Here, we

have designed new polyelectrolyte microgels and characterized their performances as

biomaterials for 3D printing and cell culture applications. The rheological properties of the

microgels can be tuned by changing polymer concentration or polymer charge density of the

polymeric microparticles. We find that interactions between the polyelectrolyte microgels and

ions in the solvent result in a decrease in the ionic contribution to the osmotic swelling of the

microgels and consequently drives changes to rheological properties. For anionic and cationic

microgels in the high-salt limit, these rheological changes follow the scaling laws describing

polyelectrolyte gels with added salt. Further studies on very highly charged polyelectrolyte

microgels may be necessary to investigate the effects of salt bridging on their rheological

behaviors; the collapse of highly charged polyelectrolyte brushes can be exaggerated in the

presence of multivalent ions as a result of electrostatic bridging between chains.131-132 By

contrast, simple polyelectrolyte scaling laws do not capture the rheological behavior of

zwitterionic microgels. Instead, zwitterionic microgels exhibit a plateau in rheological properties

in the high salt limit. This unique behavior may be advantageous when swelling microgels in

salt-rich solvents, such as cell growth media. However, interactions between the zwitterionic microgels and biological zwitterionic molecules, including amino acids and proteins, may result in unforeseen changes in rheological properties beyond the scope of this work.133-134 Further

development of charge-neutral microgels may circumvent these interactions, providing

opportunities for further biomaterial applications using microgels.

We have characterized the performance of MCF-10A cells cultured in charged microgel

environments. Short-term viability studies of anionic and zwitterionic microgels showed greater

than 90% cell viability after 24 h, while metabolic studies showed that ATP production in the

cells remains relatively unchanged. In contrast, cationic microgels containing the qDMAEMA

charged species were found to be cytotoxic. These results were independent of the rheological

properties of the microgels tested, suggesting that cell performance in jammed microgels is governed by the chemical composition of the microgels and are less sensitive to their rheological properties. Thus, the material properties of the microgels may be tuned to optimize for 3D cell culture or bioprinting applications with limited effects on cell performance. Further assessment is necessary to study the long-term viability of cells cultured in 3D microgel environments. For example, the metabolic activity and hepatic function of in vitro liver constructs are generally characterized through albumin secretion and urea synthesis.135-136 As new assay protocols are developed to measure these markers of long-term cell function in microgels, proper baselines must be established to account for the effects of the microgel environment.

Figure 5-1. Jammed Microgels as a Biomaterial. a) Jammed granular microgels can act as sacrificial support materials for 3D printing cellular constructs, allowing the study of collective cell mechanics and micro-tissue growth in a controlled 3D environment. b) Similarly, jammed granular microgels may be used to study the behavior of individual cells in 3D environments. c) Swelling of charged microgels is driven by the osmotic pressure of counterions associated with polyelectrolyte backbones. d) Here, we investigate microgel particles containing methacrylic acid (MAA), quaternized 2-(dimethylamino)ethyl methacrylate (qDMAEMA), or carboxybetaine methacrylate (CBMA) charged monomer species at varying charge densities.100

Figure 5-2. Phase contrasted micrographs of dilute samples confirm the presence of anionic (MAA), cationic (qDMAEMA), and zwitterionic (CBMA) microgels. We measure the average size of each microgel in the dilute state and find the mean particle diameters to be 4.76±1.49 µm for MAA microgels, 5.17±1.94 µm for qDMAEMA microgels, and 5.21±2.14 µm for the CBMA microgels. Scale bar: 200 µm100

Figure 5-3. Rheological characterization of charged microgels. a) Unidirectional shear rate sweeps are performed on microgel samples prepared at increasing polymer concentrations to determine the jamming concentration. b) In the dilute regime, in which the polymer concentration is below the jamming concentration, the shear stress exhibits a purely viscous response to the applied shear rate. As the polymer concentration increases toward the jamming concentration, the microgel particles become packed and the system exhibits a finite yield stress. Further increasing the polymer concentration leads to elastic deformation of the microgel particles, enabling the packing factor to exceed that of random close packed hard spheres. The yield stress of microgel packs prepared at concentrations above the jamming concentration is determined by fitting the Hershel-Bulkley model to the data. c) Increasing the polymer concentration leads to an increase in the measured yield stress for microgels prepared with constant polymer charge density (shown here: 17 mol% charged groups; no added CaCl2; Lines drawn to guide the eye). d) Similarly, the yield stress increases with increasing polymer charge density for microgels prepared at the same polymer concentration (MAA, qDMAEMA: 5 wt% polymer; CBMA: 4 wt%; no 100 added CaCl2). Error bars are smaller than graphical symbols.

Figure 5-4. Rheological Characterization of Polyelectrolyte Microgels. a) The yield stress of microgel samples is measured by fitting the Hershel-Bulkley model, = p 1 + / τy γ γc , to unidirectional shear rate sweeps. Here we find that the scaling exponent p falls between 0.4 and 0.55, indicating shear thinning behavior𝜏𝜏 at high shear� rates.� ̇ Interestingly,̇ � � this scaling exponent follows a weak logarithimic scaling with the measured yield stress. b) To compare rheological results of microgels with added salt across samples, we plot the ratio of the yield stress at a given salt 0 concentration to the yield stress with no added salt (τy / τy ) as a function of the molar ratio of added ionic charge to polymer charge. The rheological results for anionic and cationic microgels collapse to a single curve in which the yield stress decreases with increasing salt concentrations. At high molar charge ratios, the microgels exhibit a purely viscous response to an applied shear rate and the yield stress is 0. c) Zwitterionic microgels exhibit an initial, steeper collapse followed by a non-zero plateau in the measured yield stress at high salt concentrations.100

Figure 5-5. Multivalent ion interactions with charged microgels. a) Increasing ion concentration through the addition of calcium chloride results in a decrease in the yield stress. (Lines drawn as visual aids) b) Rescaling the rheological measurements by the yield stress with no added salt and plotting versus the molar charge ratio collapses the data to a single scaling curve for anionic and cationic microgels. c) Plotting data from zwitterionic microgel systems shows an ion-independent plateau in the rheological properties at molar charge ratios in excess of 1:1.100

Figure 5-6. Microgel rheology and polyelectrolyte scaling behavior. a) The elastic (Gʹ) and viscous (Gʺ) shear moduli of the microgels are measured through low amplitude frequency sweeps. Over a wide range of frequencies, Gʹ remains relatively flat and dominates Gʺ, characteristic of a viscoelastic solid. Increasing the concentration of calcium chloride concentration leads to a decrease in Gʹ and Gʹʹ. b) We find the yield stress of microgel packs to be linearly correlated to the elastic shear modulus for all systems under all conditions reported here. c) To test if polyelectrolyte scaling laws can predict the rheological performance of our anionic and cationic microgels, we plot their normalized yield stress by 1/cion. At high concentrations of added salt, the rheological behavior follows the scaling prediction; at low concentrations of added salt, we observe a plateau in the measured yield stress. We identify the critical ion * concentration c ion in which this transition occurs. d) Rescaling our rheological curves * by c ion results in a collapse to a single scaling curve for anionic and cationic microgels. e) By contrast, the scaling behavior of zwitterionic microgels with the added salt concentration is weaker than the predicted scaling from polyelectrolyte theory and exhibits a plateau at the limits of high added salt. f) We find the transition from the weak scaling behavior at low salts and the plateau at high salt corresponds to a critical ion concentration comparable to the charge density of the zwitterionic microgels. Error bars are smaller than graphical symbols.100

Figure 5-7. Rheological Characterization of Uncharged Microgels. a) The Measured yield stress of uncharged microgels in aqueous solutions increases with increasing polymer concentration; above the jamming concentration, the yield stress of uncharged microgels follows polymer scaling laws. b) There are no changes in the rheological properties of uncharged microgels with added salt.

Figure 5-8. Small amplitude oscillatory frequency sweeps of polyelectrolyte microgels swollen in MEGM cell growth media at 4 wt% polymer. The elastic shear modulus (Gʹ) remains relatively flat across the full range of frequencies, confirming the microgels are in the jammed state and ranges from 2 – 60 Pa.100

Figure 5-9. Cell viability and proliferation. a) Short-term cell viability measurements of MCF-10A suspended in 3D microgel media. Anionic and zwitterionic microgels maintain 90-95% cell viability after 24 h whereas cells cultured in cationic microgels exhibit increased levels of cell death with increasing charge concentration; cells cultured in 17 mol% qDMAEMA microgels exhibit greater than 90% cell death after 24 hours. b) Relative changes in MCF-10A cell populations cultured in the presence of anionic and zwitterionic microgel media after 24 h are statistically equivalent to liquid culture media (7.5 ± 12.35 %). After 48h, cell population changes in microgels are lower than in liquid media, but statistically the same as earlier time-points. (All microgel samples are prepared at 4 wt%; molar concentration of charged species are shown).100

Figure 5-10. Fluorescence microscopy images of MCF-10A cells cultured in jammed polyelectrolyte microgels. MCF-10A cells are dyed with CMFDA cell tracker green, plated on a glass bottom 12-well plates and cultured in microgel support media. Cell population growth rates from the cell population density determined by counting the number of cells for a given culture area. Scale bar: 200 µm.100

Figure 5-11 Metabolic activity in polyelectrolyte microgels. a) CellTiter-Glo 3D calibration measurements of luminescence intensity at known ATP concentrations for liquid and microgel media. b) Fluorescence microscopy of MCF-10A cultured in microgel media after adding CellTiter-Glo 3D visually confirm cell lysis. (17mol% MAA; cells dyed Cell Tracker Green CMFDA (green: live) and with ethidium homodimer (red: dead); scale bar: 100µm) c) Relative luminescence intensity of MCF-10A cells cultured in liquid media for 24 h. Liquid media is replaced with microgels immediately prior to cell lysis (t < 5min). The decrease in luminescence intensity results from microgels hindering the effectiveness of the CellTiter-Glo 3D assay, not a drop in ATP levels. d) Relative ATP production without correction from MCF-10A cells cultured in microgel media for 24 h relative to cells cultured in liquid media for 24 h. e) Adjusting the relative ATP levels for the decrease in luminescence intensity (Fig. 5-11c) and accounting for cell proliferation rates (Fig. 5-9b) show no significant change in ATP levels for cells cultured in microgels relative to liquid media. (all microgel samples are prepared at 4 wt%; molar concentration of charged species are shown).100

Comonomer AAm Comonomer PEG700da AIBN EtOH Comonomer Incorporation (mol%) (g) (g) (g) (mg) (mL) MAA 5 8.55 0.55 0.90 100 112 MAA 9 8 1 1 100 112 MAA 17 7.27 1.82 0.91 100 112 CBMA 5 7.85 1.33 0.82 100 112 CBMA 9 7.01 2.23 0.76 100 112 CBMA 17 5.59 3.74 0.67 100 112 qDMAEMA 5 7.52 7.69 0.79 100 112 qDMAEMA 9 6.54 2.75 0.71 100 112 qDMAEMA 17 5.02 4.38 0.60 100 112 Table 5-1. Experimental values for the synthesis of 10g of polyelectrolyte microgels composed of polyacrylamide with anionic (MAA), zwitterionic (CBMA), and cationic (qDMAEMA) comonomers.

APPENDIX A PERFUSION AND FLOW THROUGH MICROGELS

Introduction

A ubiquitous goal of tissue engineering is the development of functional tissues and

organs for implantation.50, 89 However, current limitations in printing capabilities, include speed

and precision, limit the size of a structure that can be created in a given time (Figure A-1).5 For

example, printing a tissue the size of an organ (~1L) with feature sizes of 1 mm at modest

printing speeds (v = 20 mm/s) would require printing times on the order of a day; reducing the

feature size to 100 µm would require in excess of 70 days to accomplish the same feat. Thus,

new methodologies and support material are needed to increase the speed of fabrication at high

speeds if fully functional tissues and organs are to be manufactured in a timely fashion. In the

intermediate, current 3D printing capabilities enable the printing of micro and mesoscale to study

the collective cell behavior in a 3D environment.50, 137

The mass transport of nutrients, waste, and oxygen to 3D-printed cellular constructs presents an additional challenge that must be overcome to enable the bioprinting of organs and macroscopic tissues.89 While vascular networks will be needed to overcome diffusion limits

within the microtissues; perfusion of fresh growth media through the support matrix will be

necessary to provide nutrients and drugs to the microstructure, while removing waste and bile.89,

138 To understand the scope of this challenge, we consider cell culture guidelines for feeding cells

grown in 2D. One well of a 6 well plate generally require 2.5 mL of fresh media every 48 h. This reduces to the relatively modest feeding rate of 1 mL of media per 106 cells every day. However,

when we scale these feeding rates up to meet the scale of an organ (109 cells), it is now necessary

to flow over 1 L of fresh media per day. Even when considering micro-sized tumor structures,

significant flow is necessary to maintain cellular performance; a 500 µm diameter spheroid of

cells requires approximately 0.5 mL of fresh media a day, or 20 µL/hr. Thus, if jammed granular microgels are to be used as support materials for 3D cell culturing, it will be necessary to develop perfusion methods to provide a continuous source of fresh nutrients to the cells while removing harmful waste.

Here, we take the preliminary steps to understand the permeability of jammed granular microgels. We determine the effective mesh size of jammed microgel systems through permeability studies and compare it to the measured size of the microgel particles.

Theory

The flow of an incompressible Newtonian fluid can be described by the linear expression of the Navier-Stokes Equation:

2 η∇uFii + −∇ P =0 , where η is the viscosity of the fluid, ui is the velocity vector, Fi are the forces acting on the fluid, and P is the driving pressure. If we assume that the viscous forces are linearly related to the velocity of the fluid, Stokes equation can be simplified to

uPi =−∇κ(),

(i.e. Darcy’s Law) where κ is the permeability of the medium.139 Solving Darcy’s law for the unidirectional flow with driving pressure ΔP, the volumetric flow (Q) through a material with cross sectional area A and length L can be expressed as the more familiar

−∆κ AP Q = . L

For semi-dilute polymer networks, the permeability can be related to the mesh size of the network such that ξ= κη where η is the viscosity of the solvent.140 Thus, for a 1 mm tall polyacrylamide gel with ξ = 10 nm, the driving pressure necessary to flow 100 µL through a 96

100

well plate in 24 hours is on the order of 300 MPa. Microgel systems provide a means to lower

the required driving pressure. Here, pores between microparticles increase the permeability of

the system, promoting flow and decreasing the required pressure. An order of magnitude

increase in the effective mesh size can result in a two orders of magnitude decrease in the

required driving pressure.

Results and Discussion

Characterizing size of microgel particles

In this inverstigation, we explore the permeability of three microgel systems with varying

size microparticles. Polyacrylamide microbeads are purchased from Biorad with advertized sizes

ranging between 90-180 µm and <45 µm. In additional, we synthesize uncharged poly(ethylene

glycol) methyl ether acrylate (PEGa) microgel beads crosslinked with poly(ethylene glycol)

diacrylate (PEGda) through an inverse emulsion reaction. To characterize the size of these

microparticles, we capture brightfield and phase-contrasted micrographs of dilute microgel samples (Figure A-2a-c). We measure the size distribution of the microgel particles from these micrographs and find the mean and standard devation of their diameters to be µ = 142.2 µm and

σ = 24.7 µm for the large Biorad beads, µ = 35.5 µm and σ = 6.48 µm for the small Biorad beads, and µ = 7.58 µm and σ = 1.35 µm for the PEG microgels (n > 150) (Figure A-2d).

Permeability Measurements

To determine the permeability of our microgel samples, we measure the flow of water through a column of jammed microgels (Figure A-3). Here, the unidirectional volumetric flow through the porous medium is described by Darcy’s law:

−∆κ AP Q = , L

101

were κ is the permeability of the system, A is the cross-sectional area of the column, ΔP is the

pressure driving the flow, and L is the length over which the water flows through the porous

medium. The driving pressure of the water through the jammed microgels is generated by the

hydrostatic pressure of the water column, that is ∆=Pρ gh , where ρ is the density of water, g is

the graviational constant, and h is the height of the water. Likewise, the volumetric flow through

∂h the microgels can be expressed asQA= . Thus, we can express Darcy’s law as a first order ∂t

differential equation and model the height of the water column as an exponential decay such that

−t L h() t= Ce τ , where τ = . κρ g

For each microgel system, we measure the height of the water column as a function of

time as if flows through the jammed microgels. To negate evaporation and capillary effects

during this process, we submerge the bottom of the test column in a bath of water and place a

thin layer of mineral oil on top of the water column. As expected, the height of the water column

decays exponentially with time (Figure A-4a). From the decay coefficient, we calculate the

permeability of each microgel system and find κ = 12573 µm2/Pa s, κ = 552 µm2/Pa s, and κ =

13.8 µm2/Pa s, for the large Biorad, small Biorad, and PEG microgel respectively. We note that

the permeability of the microgels scales quadratically with the average size of the microgel beads

(Figure A-4b). Similarily, we calculate the effective mesh size ξ for the microgel systems from polymer scaling laws, that is ξ= κη, and find ξ scales linearly with the size of the microgel

particles. Here we measure the effective mesh sizes as ξ = 3.54 µm, ξ = 0.74 µm, and ξ = 0.12

µm for the large Biorad, small Biorad, and PEG microgel respectively.

102

Conclusions

Here, we have explored the permeability of jammed microgel systems with varying microgel particle sizes. We find the permeability of the system increases quadratically with the average size of the microgel particles. These results represent the first step in developing a perfusion system for long-term cell culture in jammed microgel materials. Further work is necessary to apply these findings to develop a cell culture dish, to understand the spatial flow patterns within these culture containers, and how they translate to cell health and metabolic activity.

In addition, we have synthesized uncharged PEG microgel particles with an average microgel size 7.58 µm through an inversion emulsion reaction. Images of the PEG microgels taken through phase-contrasted microscopy show relatively spherical particles. Continued work is necessary to evaluate the biocompatibility of these PEG microgel particles and their efficacy as a support material for soft matter 3D printing.

103

Figure A-1. Manufacturing Space Curves of Sacrificial Support 3D Printing Methods. Manufacturing curves of current 3D printing methods using sacrificial support materials; continued advancements are necessary to achieve the ubiquitous goal of large structures with small features printed in short time periods.5

104

Figure A-2. Characterizing Microgel Size. Microgel size was measured by taking micrograph images of a) Biorad 90-180 µm pAAM beads b) Biorad <45 µm pAAM beads, and c) PEG micro beads. Scale bars 100 µm. d) Probability distribution function of microbead size for the three systems explored here.

Figure A-3. Experimental Set-Up for Microgel Permeability Measurements. A fluid column is placed over a column of microgels and the height is measured over time. The permeability is measured by fitting an exponential decay to the resulting data.

105

Figure A-4. Permeability Measurements of Jammed Microgel Systems. a) The permeability of the jammed microgels is determined by measuring the height of the fluid column over time and fitting an exponential decay to the data. b) The permeability of the jammed microgels scale quadratically with the average size of the microgel particles. Likewise, the effective mesh size of the jammed microgels scales linearly with the average size of the microgel particles (inset).

106

APPENDIX B MATERIAL AND METHODS

Organic Microgels for Silicone 3D Printing

Preparing organogel support matrix

Organic microgels were prepared from the SEP copolymer S607-bEP1561 (molecular weight, 172.6 kg/mol; polydispersity, 1.03), consisting of 28 mole percent (mol %) polystyrene

(KRATON G1702); the SEBS copolymer S208.5-b-EB974-b-S208.5 (molecular weight, 98.1 kg/mol; polydispersity, 1.03), consisting of 30 mol % polystyrene (KRATON G1650); and light mineral oil [National Formulary/Food Chemicals Codex (NF/FCC)–grade] (Fisher Scientific).

The various solubility parameters have been reported as 14.1, 17.1, 17.53, 16.2, and 20.1 MPa1/2 for mineral oil, butylene, ethylene, propylene, and polystyrene, respectively (46, 47). Block copolymer mixtures were prepared at 2.25 wt % diblock copolymer, 2.25 wt % triblock copolymer, and 95.5 wt % light mineral oil. Samples were heated to 150 °C and continuously mixed using a Scilogex Overhead Stirrer set to 250 rpm for 4 to 6 hours. Other self-assembled phases were prepared by using the same protocols, but with the compositions specified in the manuscript body.

Preparing silicone elastomer inks

UV-curing silicone elastomer inks were prepared by mixing Momentive UV Electro 225 at a 50:1 base–to–curing agent ratio. Low-viscosity silicone oil (Sigma-Aldrich) was added at 25 wt % to lower the viscosity of the printing ink. Samples were homogeneously mixed at 3500 rpm for 30 s in a FlackTek DAC 150 SpeedMixer before they were degassed and loaded into a

Hamilton GASTIGHT syringe. After printing, the silicone structures were cured under a 400-W,

320- to 390-nm UV flood curing lamp (Sunray) for 10 to 20 min. RTV silicone elastomer inks were prepared by mixing Smooth-On Mold Max 10 at a 10:1 base–to–curing agent ratio.

107

Samples were homogeneously mixed at 3500 rpm for 30 s in a speed mixer and loaded into a BD plastic syringe. After removal from the micro-organogel system, printed structures were washed first with warm soapy water (Alconox) and then with ethanol. Samples for testing feature size were printed using a UV-curing silicone elastomer formed from a vinyl-terminated PDMS base

(Gelest DMS-V31) and a [2 to 3% (mercaptopropyl)methylsiloxane)]-dimethylsiloxane copolymer cross-linker (Gelest SMS-022) and were prepared at a 3:1 base–to–crosslinker ratio

(gel fraction, 0.938). A stock solution of 2,2-dimethoxy-2- phenylacetophenone photoinitiator

(Sigma-Aldrich) was prepared at 0.125 g/1 ml of ethanol and added to the vinyl-terminated silicone elastomer ink at 0.05 wt %. Finally, 1-mm fluorescent microspheres were added to the silicone elastomer ink at ~0.1 wt % for confocal imaging.

Silicone 3D printing

All 3D printing was performed using a linear stage as a syringe pump (Physik

Instrumente) attached to three linear translation stages (Newport). Silicone elastomer inks were loaded into syringes equipped with either a disposable blunt dispensing needle (Vita Needle) or a custom-made glass needle. Glass needles were created by drawing a glass microcapillary (1 mm outer diameter) to a desired diameter using a pipette puller (David Kopf Instruments). Typical needle inner diameters vary over a range of 150 mm to 1 mm. Both the syringe pump and translation stages were controlled through custom-written MATLAB script functions and trajectory files.

Rheology

All rheological measurements were taken on a Malvern Kinexus pro+ rheometer using a roughened 40-mm upper cone with an angle of 4° and a roughened 40-mm lower plate.

Frequency sweeps were taken at 1% strain from 101 to 10−4 Hz. The yield stress of the material was determined by applying a shear rate sweep from 500 to 10−3 s−1 and measuring the shear

108

stress. The thixotropic time was measured by first applying a shear stress greater than the yield

stress of the organic microgel system. The applied stress was then dropped below the yield stress of the material, and the shear rate was measured as a function of time.

Small-angle x-ray scattering

SAXS was performed using a Bruker NANOSTAR SAXS system. Samples were placed

into quartz capillary tubes (diameter, 1.5 mm; wall thickness, 10 mm) and flame-sealed. SAXS measurements were taken over the course of 18 hours using a 2D wire detector with 1024 × 1024 pixels. The data were then integrated over the azimuthal angle.

Interfacial tension

Surface tensions were determined using the pendant drop method, in which a drop is suspended from a needle into air. Images of the droplet were taken and analyzed in MATLAB to determine the interfacial tension between the drop and air. Interfacial tension measurements between silicone oil and mineral oil were determined by measuring the contact angles formed by placing a drop of light mineral oil (NF/FCC-grade; Fisher Scientific) in a bath of 100 cSt oil

(Sigma-Aldrich) and solving Young’s equation. The contact angles were measured using FIJI

ImageJ software.

Gel permeation chromatography

Molecular weight and polydispersity were determined by gel permeation chromatography in tetrahydrofuran at 40 °C and a flow rate of 1.0 ml/min (Agilent isocratic pump, degasser, and autosampler; columns: Waters Styragel 5-mm guard + two Waters Styragel HR 4E columns; molecular weight range, 101 to 2 × 105 g/mol). Detection consisted of an Agilent 1260 Infinity refractive index detector, and the system was calibrated using polystyrene standards.

109

Interfacial Instabilities in Jammed Microgels

Rheological Characterization

Rheological measurements are performed on an Anton Paar MCR 702 Rheometer with a

25 mm, roughened plate on plate configuration. Unidirectional shear sweeps are performed by

ramping the shear rate from 500 s-1 to 10-3 s-1 and measuring the resulting shear stress. Frequency

sweeps are performed at 1% strain from 101 Hz to 10-2 Hz using the same geometric

configuration.

3D Printing

All 3D printing was performed using a linear stage as a syringe pump (Physik

Instrumente) attached to three linear translation stages (Newport). Printing inks are loaded into

Hamilton Gas Tight syringes equipped with blunt needle tips of varying gauge size. All prints

were performed at translational speeds of 10 mm/s while the feature size was varied through

changes in the volumetric flow of the ink. Both the syringe pump and translational stages were

controlled through custom-written MATLAB scripts.

Image Acquisition and Data Analysis

Macroscopic images of printed structures were taken with a Nikon D3X using DigiCam

Capture image acquisition software. Image and data analysis was performed using ImageJ software and custom written Matlab code.

Polyelectrolyte Microgels for 3D Cell Culture

Microgel Synthesis and Preparation

Cationic, anionic, and zwitterionic microgels are synthesized as previously reported with some modifications.63 Briefly, a solution of acrylamide (Alfa-Aesar), ionizable comonomer (see

-1 below), poly(ethylene glycol) diacrylate (Mn = 700 g mol , Sigma Aldrich), and

azobisisobutyronitrile (Sigma Aldrich) in ethanol is prepared. The solution is sparged with

110

nitrogen for 30 min, then placed into a preheated oil bath set at 60 °C. After approximately 30 min, the solution becomes hazy and a white precipitate begins to form. The reaction mixture is heated for an additional 4 h. At this time, the precipitate is collected by vacuum filtration and rinsed with ethanol on the filter. The microparticles are triturated with 500 mL of ethanol overnight. The solids are collected by vacuum filtration and dried on the filter for ~10 min. The particles are dried completely in a vacuum oven set at 50 °C to yield a loose white powder.

Specific details for the various compositions are included in Table 5-1.

Anionic microparticles are synthesized using methacrylic acid (Sigma-Aldrich) as comonomer. Zwitterionic microparticles are synthesized using carboxybetaine methacrylate

(CBMA). CBMA is synthesized using a modified version of a previously reported procedure.141

Briefly, DMAEMA (20.0 g) is added to a round-bottom flask equipped with a magnetic stirrer and cooled to 0 °C. Acrylic acid (18.3 g) is added dropwise at 0 °C, and the mixture is allowed to stir at 0 °C for 30 min, then 4 h at room temperature. At this time, anhydrous tetrahydrofuran (25 mL) is added and the mixture is stirred for 16 h. Triethylamine (25 mL) is added to deprotonate the resulting monomer. Anhydrous THF and anhydrous diethyl ether are added to precipitate the zwitterionic monomer. The monomer is collected by vacuum filtration, dried under vacuum, and stored in a desiccator. Cationic microparticles are synthesized using quaternized 2-

(dimethylamino)ethyl methacrylate (qDMAEMA). qDMAEMA is synthesized as previously described.142 Briefly, DMAEMA (18.7 g) is mixed in anhydrous THF (30 mL). Methyl iodide

(20.2 g) in anhydrous THF (30 mL) is added dropwise at 0 °C. The reaction mixture is warmed to room temperature and stirred for 24 h. At this time, the monomer is collected by vacuum filtration and rinsed on the filter with anhydrous THF. The white solid is dried under vacuum and stored in a desiccator.

111

Microgel samples are prepared by dispersing the microparticles in ultrapure Millipore

water to the desired polymer concentration. Samples are subsequently speed mixed (FlakTec)

and allowed to equilibrate overnight. NaOH is added to the anionic microgels to deprotenate the

methacrylic acid charged species until a final pH of ~6.4 is reached. Calcium chloride dihydrate

(Fisher) is added to the microgel solutions to study the effects of ion-polyelectrolyte interactions

on the rheological behavior of microgel packs.

Rheological Characterization

Rheological measurements are performed on a Malvern Kinexus Pro Rheometer with a

40 mm 1 degree roughened cone on plate configuration and an Anton Paar MCR 702 Rheometer

with a 50 mm, 1.0 degree cone on plate configuration. Unidirectional shear sweeps are

performed by ramping the shear rate from 500 s-1 to 10-3 s-1 and measuring the resulting shear

stress. Frequency sweeps are performed at 1% strain from 10 Hz to 10-2 Hz using the same

geometric configuration.

Cell Culture and Short-Term Viability

Human mammary epithelial cells (MCF-10A) are cultured in MEBM cell culture medium

supplemented with MEGM BulletKit supplements (Lonza) and incubated at 37 °C, 5% CO2, and

95% humidity. Microgels for cell culturing are swollen in MEGM cell growth media to a final polymer concentration of 4 wt%. NaOH is added to the medium to adjust the pH to 7.4 under incubation conditions. The salt concentrations of MEGM cell growth media are as follows: 2 mM CaCl2, 112 mM NaCl, 2.5 mM KCl, and 1.5 mM MgSO4.

Short-term cell viability assays are performed by dispersing MCF-10A cells in microgel prepared at 4 wt% polymer in MEGM media. After passaging, the cells are pelleted and re- suspended at a high density in liquid growth media before subsequently being pipetted into the microgel growth media. The cells and microgels are gently pipette mixed to evenly disperse the

112

cells throughout the suspension and incubated at 37 °C and 5% CO2. ReadyProbes Cell Viability

Blue/Green Imaging Kit (Thermo Fisher) is added to the microgel media per the manufacturer’s

instructions at the 3 h and 24 h time points and gently pipette mixed. The cells are then incubated

for an additional 30 minutes before cell viability images are taken using confocal microscopy.

Cell proliferation measurements are performed by plating MCF-10A cells in glass

bottomed petri dishes with liquid growth media. 24 h after plating, cells are dyed with 10 µM

CMFDA Cell Tracker Green (Thermo Fisher) for 30 minutes. After dyeing, cells are culture in

liquid cell growth media for 2 h. The liquid media is subsequently replaced with microgel media

and cells are incubated at 37 °C and 5% CO2. Fluorescence images are taken at the 0 h, 24 h, and

48 h time points after the microgel media has been introduced. Cell density measurements are

determined by counting the number of cells present in the fluorescence images field of view.

Cell Metabolic Activity

Cell metabolic activity studies are performed using CellTiter-Glo 3D Cell Viability

Assay (Promega). After passaging, cells are plated in a tissue-treated 96-well plate with liquid cell growth media. Cells are allowed to settle and attached to the well plate for 2 h before the liquid media is replaced with 100 µL of microgel media and cultured for 24 h. After 24 h, 100

µL of CellTiter-Glo 3D assay is added to each well. The media is pipette mixed, and then placed on an orbital shaker for 25 min. After mixing, all 200 µL are transferred to an opaque 96 well plate (Corning) and luminescence measurements are taken on a BioTek Synergy HTX microplate reader. Luminescence calibration curves are performed using known concentrations of adenosine

5’-triphosphate (ATP) disodium salt hydrate (Sigma).

113

PEG Microgel Synthesis

Uncharged poly(ethylene glycol) (PEG) microgels are synthesized through an inverse emulsion reaction; aqueous emulsion droplets containing poly(ethylene glycol) methyl ether acrylate (PEGa) and poly(ethylene glycol) diacrylate (PEGda) are dispersed in continuous organic phase. The Organic phase is prepared by combining 500 mL of Kerosene (Sigma) and

3.5 g of PGPR surfactant (Palsgaard) in a 1L beaker and stirring for 20 min at 0 °C. Separately, we prepare an aqueous phase containing 36.5 g PEGa, 0.5 g PEGda, 0.225g APS, and 112.775 g water and stir for 30 minutes. After stirring, the aqueous phase is added to the oil phases and homogenized at 8000 RPMs for 5 min at 0 °C to form aqueous emulsion droplets. The system is then purged for 30 min with nitrogen gas. After purging, we add 2mL of TEMED to the system and continue mixing for 4 h as the reaction is brought up to room temperature. After crosslinking, samples are centrifuged to separate the microgel particles from the oil phase and subsequently washed with methanol. After washing, the PEG microgels are titrated into a bath of diethyl ether to collapse the microgels and dried overnight in a vacuum oven.

114

LIST OF REFERENCES

1. de Gennes, P. G., Soft matter (Nobel lecture). Angewandte Chemie International Edition in English 1992, 31 (7), 842-845.

2. Wood, G.; Keech, M., The formation of fibrils from collagen solutions 1. The effect of experimental conditions: kinetic and electron-microscope studies. Biochemical Journal 1960, 75 (3), 588.

3. Yang, Y.-l.; Kaufman, L. J., Rheology and confocal reflectance microscopy as probes of mechanical properties and structure during collagen and collagen/hyaluronan self- assembly. Biophysical journal 2009, 96 (4), 1566-1585.

4. Yang, Y.-l.; Motte, S.; Kaufman, L. J., Pore size variable type I collagen gels and their interaction with glioma cells. Biomaterials 2010, 31 (21), 5678-5688.

5. O’Bryan, C. S.; Bhattacharjee, T.; Niemi, S. R.; Balachandar, S.; Baldwin, N.; Ellison, S. T.; Taylor, C. R.; Sawyer, W. G.; Angelini, T. E., Three-Dimensional Printing with Sacrificial Materials for Soft Matter Manufacturing. MRS Bulletin 2017, 42 (8), 571-577.

6. Bhattacharjee, T.; Zehnder, S. M.; Rowe, K. G.; Jain, S.; Nixon, R. M.; Sawyer, W. G.; Angelini, T. E., Writing in the Granular Gel Medium. Science advances 2015, 1 (8), e1500655.

7. Highley, C. B.; Rodell, C. B.; Burdick, J. A., Direct 3D Printing of Shear‐Thinning Hydrogels into Self‐Healing Hydrogels. Advanced Materials 2015, 27 (34), 5075-5079.

8. Hinton, T. J.; Jallerat, Q.; Palchesko, R. N.; Park, J. H.; Grodzicki, M. S.; Shue, H.-J.; Ramadan, M. H.; Hudson, A. R.; Feinberg, A. W., Three-Dimensional Printing of Complex Biological Structures by Freeform Reversible Embedding of Suspended Hydrogels. Science advances 2015, 1 (9), e1500758.

9. Kolesky, D. B.; Homan, K. A.; Skylar-Scott, M. A.; Lewis, J. A., Three-dimensional bioprinting of thick vascularized tissues. Proceedings of the National Academy of Sciences 2016, 113 (12), 3179-3184.

10. Kolesky, D. B.; Truby, R. L.; Gladman, A. S.; Busbee, T. A.; Homan, K. A.; Lewis, J. A., 3D bioprinting of vascularized, heterogeneous cell‐laden tissue constructs. Advanced materials 2014, 26 (19), 3124-3130.

11. Ouyang, L.; Highley, C. B.; Rodell, C. B.; Sun, W.; Burdick, J. A., 3D printing of shear- thinning hyaluronic acid hydrogels with secondary cross-linking. ACS Biomaterials Science & Engineering 2016, 2 (10), 1743-1751.

115

12. Siqueira, G.; Kokkinis, D.; Libanori, R.; Hausmann, M. K.; Gladman, A. S.; Neels, A.; Tingaut, P.; Zimmermann, T.; Lewis, J. A.; Studart, A. R., Cellulose nanocrystal inks for 3D printing of textured cellular architectures. Advanced Functional Materials 2017, 27 (12), 1604619.

13. Gantenbein, S.; Masania, K.; Woigk, W.; Sesseg, J. P.; Tervoort, T. A.; Studart, A. R., Three-dimensional printing of hierarchical liquid-crystal-polymer structures. Nature 2018, 561 (7722), 226.

14. Mora, S.; Phou, T.; Fromental, J.-M.; Pismen, L. M.; Pomeau, Y., Capillarity driven instability of a soft solid. Physical review letters 2010, 105 (21), 214301.

15. Pairam, E.; Le, H.; Fernández-Nieves, A., Stability of toroidal droplets inside yield stress materials. Physical Review E 2014, 90 (2), 021002.

16. Pairam, E.; Vallamkondu, J.; Koning, V.; van Zuiden, B. C.; Ellis, P. W.; Bates, M. A.; Vitelli, V.; Fernandez-Nieves, A., Stable nematic droplets with handles. Proceedings of the National Academy of Sciences 2013, 110 (23), 9295-9300.

17. Coutanceau, M.; Defaye, J.-R., Circular cylinder wake configurations: A flow visualization survey. Applied Mechanics Reviews 1991, 44 (6), 255-305.

18. Taneda, S., Experimental investigation of the wake behind a sphere at low Reynolds numbers. Journal of the Physical Society of Japan 1956, 11 (10), 1104-1108.

19. Thom, A., The flow past circular cylinders at low speeds. Proceedings of the Royal Society of London. Series A, Containing Papers of a Mathematical and Physical Character 1933, 141 (845), 651-669.

20. LeBlanc, K. J.; Niemi, S. R.; Bennett, A. I.; Harris, K. L.; Schulze, K. D.; Sawyer, W. G.; Taylor, C.; Angelini, T. E., Stability of High Speed 3D Printing in Liquid-Like Solids. ACS Biomaterials Science & Engineering 2016, 2 (10), 1796-1799.

21. Wu, W.; DeConinck, A.; Lewis, J. A., Omnidirectional Printing of 3D Microvascular Networks. Advanced Materials 2011, 23 (24), H178-H183.

22. Xu, C.; Chai, W.; Huang, Y.; Markwald, R. R., Scaffold‐Free Inkjet Printing of Three‐ Dimensional Zigzag Cellular Tubes. Biotechnology and bioengineering 2012, 109 (12), 3152-3160.

23. Landers, R.; Hübner, U.; Schmelzeisen, R.; Mülhaupt, R., Rapid Prototyping of Scaffolds Derived from Thermoreversible Hydrogels and Tailored for Applications in Tissue Engineering. Biomaterials 2002, 23 (23), 4437-4447.

116

24. Landers, R.; Pfister, A.; Hübner, U.; John, H.; Schmelzeisen, R.; Mülhaupt, R., Fabrication of soft tissue engineering scaffolds by means of rapid prototyping techniques. Journal of materials science 2002, 37 (15), 3107-3116.

25. Christensen, K.; Zhang, Z.; Xu, C.; Huang, Y., Deformation Compensation During Buoyancy-Enabled Inkjet Printing of Three-Dimensional Soft Tubular Structures. Journal of Manufacturing Science and Engineering 2018, 140 (1), 011011.

26. Homan, K. A.; Kolesky, D. B.; Skylar-Scott, M. A.; Herrmann, J.; Obuobi, H.; Moisan, A.; Lewis, J. A., Bioprinting of 3D convoluted renal proximal tubules on perfusable chips. Scientific reports 2016, 6.

27. Kolesky, D. B.; Truby, R. L.; Gladman, A.; Busbee, T. A.; Homan, K. A.; Lewis, J. A., 3D bioprinting of vascularized, heterogeneous cell‐laden tissue constructs. Advanced materials 2014, 26 (19), 3124-3130.

28. Song, K. H.; Highley, C. B.; Rouff, A.; Burdick, J. A., Complex 3D‐Printed Microchannels within Cell‐Degradable Hydrogels. Advanced Functional Materials 2018, 28 (31), 1801331.

29. Rodell, C. B.; Kaminski, A. L.; Burdick, J. A., Rational design of network properties in guest–host assembled and shear-thinning hyaluronic acid hydrogels. Biomacromolecules 2013, 14 (11), 4125-4134.

30. Dawson, J. I.; Kanczler, J. M.; Yang, X. B.; Attard, G. S.; Oreffo, R. O., Clay gels for the delivery of regenerative microenvironments. Advanced Materials 2011, 23 (29), 3304- 3308.

31. Ruzicka, B.; Zaccarelli, E., A fresh look at the Laponite phase diagram. Soft Matter 2011, 7 (4), 1268-1286.

32. Ahlfeld, T.; Cidonio, G.; Kilian, D.; Duin, S.; Akkineni, A.; Dawson, J.; Yang, S.; Lode, A.; Oreffo, R.; Gelinsky, M., Development of a clay based bioink for 3D cell printing for skeletal application. Biofabrication 2017, 9 (3), 034103.

33. Gladman, A. S.; Matsumoto, E. A.; Nuzzo, R. G.; Mahadevan, L.; Lewis, J. A., Biomimetic 4D printing. Nature materials 2016, 15 (4), 413.

34. Jin, Y.; Compaan, A.; Chai, W.; Huang, Y., Functional nanoclay suspension for printing- then-solidification of liquid materials. ACS applied materials & interfaces 2017, 9 (23), 20057-20066.

35. Sollich, P.; Lequeux, F.; Hébraud, P.; Cates, M. E., Rheology of soft glassy materials. Physical review letters 1997, 78 (10), 2020.

117

36. Bonham, J.; Faers, M.; van Duijneveldt, J., Non-aqueous microgel particles: synthesis, properties and applications. Soft matter 2014, 10 (47), 9384-9398.

37. Saatweber, D.; Vogt-Birnbrich, B., Microgels in organic coatings. Progress in organic coatings 1996, 28 (1), 33-41.

38. Saunders, B. R.; Laajam, N.; Daly, E.; Teow, S.; Hu, X.; Stepto, R., Microgels: From Responsive Polymer Colloids to Biomaterials. Advances in colloid and interface science 2009, 147, 251-262.

39. Thorne, J. B.; Vine, G. J.; Snowden, M. J., Microgel applications and commercial considerations. Colloid and Polymer Science 2011, 289 (5-6), 625.

40. Menut, P.; Seiffert, S.; Sprakel, J.; Weitz, D. A., Does Size Matter? Elasticity of Compressed Suspensions of Colloidal-and Granular-Scale Microgels. Soft Matter 2012, 8 (1), 156-164.

41. Nordstrom, K. N.; Verneuil, E.; Arratia, P.; Basu, A.; Zhang, Z.; Yodh, A. G.; Gollub, J. P.; Durian, D. J., Microfluidic rheology of soft colloids above and below jamming. Physical review letters 2010, 105 (17), 175701.

42. Coussot, P.; Tocquer, L.; Lanos, C.; Ovarlez, G., Macroscopic vs. local rheology of yield stress fluids. Journal of Non-Newtonian Fluid Mechanics 2009, 158 (1), 85-90.

43. Liu, A. J.; Nagel, S. R., Nonlinear dynamics: Jamming is not just cool any more. Nature 1998, 396 (6706), 21-22.

44. Evans, D. F.; Wennerstrom, H., Colloidal domain. Wiley-Vch: 1999.

45. Bhattacharjee, T.; Gil, C. J.; Marshall, S. L.; Urueña, J. M.; O’Bryan, C. S.; Carstens, M.; Keselowsky, B.; Palmer, G. D.; Ghivizzani, S.; Gibbs, C. P., Liquid-Like Solids Support Cells in 3D. ACS Biomaterials Science & Engineering 2016, 2 (10), 1787-1795.

46. Hinton, T. J.; Hudson, A.; Pusch, K.; Lee, A.; Feinberg, A. W., 3D Printing PDMS Elastomer in a Hydrophilic Support Bath via Freeform Reversible Embedding. ACS Biomaterials Science & Engineering 2016, 2 (10), 1781-1786.

47. O’Bryan, C. S.; Bhattacharjee, T.; Hart, S.; Kabb, C. P.; Schulze, K. D.; Chilakala, I.; Sumerlin, B. S.; Sawyer, W. G.; Angelini, T. E., Self-Assembled Micro-Organogels for 3D Printing Silicone Structures. Science Advances 2017, 3 (5), e1602800.

48. O’Bryan, C. S.; Bhattacharjee, T.; Marshall, S. L.; Sawyer, W. G.; Angelini, T. E., Commercially Available Microgels for 3D Bioprinting. Bioprinting 2018, e00037.

118

49. Jakab, K.; Norotte, C.; Marga, F.; Murphy, K.; Vunjak-Novakovic, G.; Forgacs, G., Tissue engineering by self-assembly and bio-printing of living cells. Biofabrication 2010, 2 (2), 022001.

50. Murphy, S. V.; Atala, A., 3D bioprinting of tissues and organs. Nature biotechnology 2014, 32 (8), 773.

51. Ferris, C. J.; Gilmore, K. J.; Beirne, S.; McCallum, D.; Wallace, G. G., Bio-ink for on- demand printing of living cells. Biomaterials Science 2013, 1 (2), 224-230.

52. Rutz, A. L.; Lewis, P. L.; Shah, R. N., Toward next-generation bioinks: Tuning material properties pre-and post-printing to optimize cell viability. MRS Bulletin 2017, 42 (8), 563-570.

53. Williams, D.; Thayer, P.; Martinez, H.; Gatenholm, E.; Khademhosseini, A., A Perspective on the Physical, Mechanical and Biological Specifications of Bioinks and the Development of Functional Tissues in 3D Bioprinting. Bioprinting 2018.

54. Bakarich, S. E.; Gorkin III, R.; in het Panhuis, M.; Spinks, G. M., Three-dimensional printing fiber reinforced hydrogel composites. ACS applied materials & interfaces 2014, 6 (18), 15998-16006.

55. Hong, S.; Sycks, D.; Chan, H. F.; Lin, S.; Lopez, G. P.; Guilak, F.; Leong, K. W.; Zhao, X., 3D printing of highly stretchable and tough hydrogels into complex, cellularized structures. Advanced materials 2015, 27 (27), 4035-4040.

56. Compaan, A. M.; Song, K.; Huang, Y., Gellan Fluid Gel as a Versatile Support Bath Material for Fluid Extrusion Bioprinting. ACS applied materials & interfaces 2019.

57. Cooke, M. E.; Jones, S. W.; ter Horst, B.; Moiemen, N.; Snow, M.; Chouhan, G.; Hill, L. J.; Esmaeli, M.; Moakes, R. J.; Holton, J., Structuring of hydrogels across multiple length scales for biomedical applications. Advanced Materials 2018, 30 (14), 1705013.

58. Moxon, S. R.; Cooke, M. E.; Cox, S. C.; Snow, M.; Jeys, L.; Jones, S. W.; Smith, A. M.; Grover, L. M., Suspended manufacture of biological structures. Advanced Materials 2017, 29 (13), 1605594.

59. Carbopol Polymer Products. https://www.lubrizol.com/Life-Sciences/Products/Carbopol- Polymer-Products (accessed October 1, 2018).

60. Carl, J. L. I.; Amjad, Z.; Masler III, W. F.; Wingo, W. H. Easy to Disperse Polycarboxylic Acid Thickeners. 1994.

61. Pemulen Polymeric Emulsifiers. https://www.lubrizol.com/Life- Sciences/Products/Pemulen-Polymeric-Emulsifiers (accessed October 1, 2018).

119

62. Guerrero, A. A.; Vargas, A. Cosmetic composition containing emulsifying copolymer and anionic sulfosuccinate. 1993.

63. Bhattacharjee, T.; Kabb, C. P.; O’Bryan, C. S.; Urueña, J. M.; Sumerlin, B. S.; Sawyer, W. G.; Angelini, T. E., Polyelectrolyte Scaling Laws for Microgel Yielding near Jamming. Soft matter 2018, 14 (9), 1559-1570.

64. Herschel, W. H.; Bulkley, R., Konsistenzmessungen von Gummi-Benzollösungen. Colloid & Polymer Science 1926, 39 (4), 291-300.

65. Flory, P. J., Principles of Polymer Chemistry. Cornell University Press: 1953.

66. Morton, M., Rubber technology. Springer Science & Business Media: 2013.

67. Polmanteer, K. E., Silicone rubber, its development and technological progress. Rubber chemistry and technology 1988, 61 (3), 470-502.

68. Bhattacharjee, N.; Urrios, A.; Kang, S.; Folch, A., The upcoming 3D-printing revolution in microfluidics. Lab on a Chip 2016, 16 (10), 1720-1742.

69. Derby, B., Inkjet printing of functional and structural materials: fluid property requirements, feature stability, and resolution. Annual Review of Materials Research 2010, 40, 395-414.

70. Femmer, T.; Kuehne, A. J.; Wessling, M., Print your own membrane: direct rapid prototyping of polydimethylsiloxane. Lab on a Chip 2014, 14 (15), 2610-2613.

71. Mannoor, M. S.; Jiang, Z.; James, T.; Kong, Y. L.; Malatesta, K. A.; Soboyejo, W. O.; Verma, N.; Gracias, D. H.; McAlpine, M. C., 3D printed bionic ears. Nano letters 2013, 13 (6), 2634-2639.

72. Stringer, J.; Derby, B., Formation and stability of lines produced by inkjet printing. Langmuir 2010, 26 (12), 10365-10372.

73. Alexandridis, P.; Olsson, U.; Lindman, B., A record nine different phases (four cubic, two hexagonal, and one lamellar lyotropic liquid crystalline and two micellar solutions) in a ternary isothermal system of an amphiphilic block copolymer and selective solvents (water and oil). Langmuir 1998, 14 (10), 2627-2638.

74. Bates, F. S.; Fredrickson, G., Block copolymers-designer soft materials. Physics today 2000, 52.

75. Sliozberg, Y.; Strawhecker, K.; Andzelm, J.; Lenhart, J., Computational and experimental investigation of morphology in thermoplastic elastomer gels composed of AB/ABA blends in B-selective solvent. Soft Matter 2011, 7 (16), 7539-7551.

120

76. Halperin, A., Polymeric micelles: a star model. Macromolecules 1987, 20 (11), 2943- 2946.

77. Izzo, D.; Marques, C., Formation of micelles of diblock and triblock copolymers in a selective solvent. Macromolecules 1993, 26 (26), 7189-7194.

78. Laurer, J. H.; Bukovnik, R.; Spontak, R. J., Morphological characteristics of SEBS thermoplastic elastomer gels. Macromolecules 1996, 29 (17), 5760-5762.

79. Mortensen, K., Structural properties of self-assembled polymeric micelles. Current opinion in colloid & interface science 1998, 3 (1), 12-19.

80. Maxwell, J. C., L. on the calculation of the equilibrium and stiffness of frames. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science 1864, 27 (182), 294-299.

81. Pickering, S. U., Cxcvi.—emulsions. Journal of the Chemical Society, Transactions 1907, 91, 2001-2021.

82. Bi, D.; Zhang, J.; Chakraborty, B.; Behringer, R., Jamming by shear. Nature 2011, 480 (7377), 355-358.

83. Pellet, C.; Cloitre, M., The glass and jamming transitions of soft polyelectrolyte microgel suspensions. Soft matter 2016, 12 (16), 3710-3720.

84. Barry, B.; Meyer, M., The rheological properties of carbopol gels II. Oscillatory properties of carbopol gels. International Journal of Pharmaceutics 1979, 2 (1), 27-40.

85. Kim, J.-Y.; Song, J.-Y.; Lee, E.-J.; Park, S.-K., Rheological properties and microstructures of Carbopol gel network system. Colloid and Polymer Science 2003, 281 (7), 614-623.

86. Saunders, B. R.; Vincent, B., Microgel particles as model colloids: theory, properties and applications. Advances in colloid and interface science 1999, 80 (1), 1-25.

87. Singla, A. K.; Chawla, M.; Singh, A., Potential applications of carbomer in oral mucoadhesive controlled drug delivery system: a review. Drug development and industrial pharmacy 2000, 26 (9), 913-924.

88. Chia, H. N.; Wu, B. M., Recent advances in 3D printing of biomaterials. Journal of biological engineering 2015, 9 (1), 1.

89. Miller, J. S., The billion cell construct: will three-dimensional printing get us there? PLoS biology 2014, 12 (6), e1001882.

121

90. De Gennes, P.-G.; Brochard-Wyart, F.; Quéré, D., Capillarity and wetting phenomena: drops, bubbles, pearls, waves. Springer Science & Business Media: 2013.

91. Neumann, F., Vorlesungen über die Theorie der Capillarität gehalten an der Universität Königsberg von Dr. Franz Neumann,... Herausgegeben von Dr. A. Wangerin. BG Teubner: 1894.

92. Long, D.; Ajdari, A.; Leibler, L., Static and dynamic wetting properties of thin rubber films. Langmuir 1996, 12 (21), 5221-5230.

93. Andreotti, B.; Bäumchen, O.; Boulogne, F.; Daniels, K. E.; Dufresne, E. R.; Perrin, H.; Salez, T.; Snoeijer, J. H.; Style, R. W., Solid capillarity: when and how does surface tension deform soft solids? Soft Matter 2016, 12 (12), 2993-2996.

94. Style, R. W.; Dufresne, E. R., Static wetting on deformable substrates, from liquids to soft solids. Soft Matter 2012, 8 (27), 7177-7184.

95. Style, R. W.; Jagota, A.; Hui, C.-Y.; Dufresne, E. R., Elastocapillarity: Surface tension and the mechanics of soft solids. Annual Review of Condensed Matter Physics 2017, 8, 99-118.

96. Style, R. W.; Isa, L.; Dufresne, E. R., Adsorption of soft particles at fluid interfaces. Soft Matter 2015, 11 (37), 7412-7419.

97. Posocco, P.; Perazzo, A.; Preziosi, V.; Laurini, E.; Pricl, S.; Guido, S., Interfacial tension of oil/water emulsions with mixed non-ionic surfactants: comparison between experiments and molecular simulations. RSC Advances 2016, 6 (6), 4723-4729.

98. Coussot, P.; Gaulard, F., Gravity flow instability of viscoplastic materials: The ketchup drip. Physical Review E 2005, 72 (3), 031409.

99. German, G.; Bertola, V., Formation of viscoplastic drops by capillary breakup. Physics of fluids 2010, 22 (3), 033101.

100. O'Bryan, C. S.; Kabb, C. P.; Sumerlin, B. S.; Angelini, T. E., Jammed Polyelectrolyte Microgels for 3D Cell Culture Applications: Rheological Behavior with Added Salts. ACS Applied Bio Materials 2019.

101. Habibi, M.; Rahmani, Y.; Bonn, D.; Ribe, N., Buckling of liquid columns. Physical review letters 2010, 104 (7), 074301.

102. Maleki, M.; Habibi, M.; Golestanian, R.; Ribe, N.; Bonn, D., Liquid rope coiling on a solid surface. Physical review letters 2004, 93 (21), 214502.

103. Cruickshank, J.; Munson, B. R., Viscous fluid buckling of plane and axisymmetric jets. Journal of Fluid Mechanics 1981, 113, 221-239.

122

104. Rahmani, Y.; Habibi, M.; Javadi, A.; Bonn, D., Coiling of yield stress fluids. Physical Review E 2011, 83 (5), 056327.

105. Habibi, M.; Ribe, N.; Bonn, D., Coiling of elastic ropes. Physical review letters 2007, 99 (15), 154302.

106. Batchelor, C. K.; Batchelor, G., An introduction to fluid dynamics. Cambridge university press: 1967.

107. Heyes, D.; Brańka, A., Interactions between microgel particles. Soft Matter 2009, 5 (14), 2681-2685.

108. Mattsson, J.; Wyss, H. M.; Fernandez-Nieves, A.; Miyazaki, K.; Hu, Z.; Reichman, D. R.; Weitz, D. A., Soft Colloids Make Strong Glasses. Nature 2009, 462 (7269), 83.

109. Cloitre, M.; Borrega, R.; Monti, F.; Leibler, L., Glassy Dynamics and Flow Properties of Soft Colloidal Pastes. Physical review letters 2003, 90 (6), 068303.

110. Pelton, R.; Chibante, P., Preparation of Aqueous Latices with N-isopropylacrylamide. Colloids and Surfaces 1986, 20 (3), 247-256.

111. Pusey, P. N.; Van Megen, W., Phase Behaviour of Concentrated Suspensions of Nearly Hard Colloidal Spheres. Nature 1986, 320 (6060), 340.

112. Jiang, Y.; Chen, J.; Deng, C.; Suuronen, E. J.; Zhong, Z., Click Hydrogels, Microgels and Nanogels: Emerging Platforms for Drug Delivery and Tissue Engineering. Biomaterials 2014, 35 (18), 4969-4985.

113. Oh, J. K.; Drumright, R.; Siegwart, D. J.; Matyjaszewski, K., The Development of Microgels/Nanogels for Drug Delivery Applications. Progress in Polymer Science 2008, 33 (4), 448-477.

114. Kabb, C. P.; O’Bryan, C. S.; Deng, C. C.; Angelini, T. E.; Sumerlin, B. S., Photoreversible Covalent Hydrogels for Soft-Matter Additive Manufacturing. ACS applied materials & interfaces 2018, 10 (19), 16793-16801.

115. Song, K. H.; Highley, C. B.; Rouff, A.; Burdick, J. A., Complex 3D‐Printed Microchannels within Cell‐Degradable Hydrogels. Advanced Functional Materials 2018, 1801331.

116. Scott, E. A.; Nichols, M. D.; Kuntz-Willits, R.; Elbert, D. L., Modular Scaffolds Assembled Around Living Cells Using poly (ethylene glycol) Microspheres with Macroporation via a Non-Cytotoxic Porogen. Acta biomaterialia 2010, 6 (1), 29-38.

117. Bhattacharjee, T.; Angelini, T. E., 3D T cell motility in jammed microgels. Journal of Physics D: Applied Physics 2018, 52 (2), 024006.

123

118. Ham, R. G.; McKeehan, W. L., Media and Growth Requirements. In Methods in enzymology, Elsevier: 1979, pp 44-93.

119. Fernández-Nieves, A.; Fernández-Barbero, A.; De las Nieves, F., Salt Effects Over the Swelling of Ionized Mesoscopic Gels. The Journal of Chemical Physics 2001, 115 (16), 7644-7649.

120. López-León, T.; Ortega-Vinuesa, J. L.; Bastos-González, D.; Elaïssari, A., Cationic and Anionic poly (N-isopropylacrylamide) Based Submicron Gel Particles: Electrokinetic Properties and Colloidal Stability. The Journal of Physical Chemistry B 2006, 110 (10), 4629-4636.

121. Lowe, A. B.; McCormick, C. L., Synthesis and Solution Properties of Zwitterionic Polymers. Chemical reviews 2002, 102 (11), 4177-4190.

122. de Grooth, J.; Ogieglo, W.; de Vos, W. M.; Gironès, M.; Nijmeijer, K.; Benes, N. E., Swelling Dynamics of Zwitterionic Copolymers: The Effects of Concentration and Type of Anion and Cation. European polymer journal 2014, 55, 57-65.

123. Rubinstein, M.; Colby, R. H.; Dobrynin, A. V.; Joanny, J.-F., Elastic Modulus and Equilibrium Swelling of Polyelectrolyte Gels. Macromolecules 1996, 29 (1), 398-406.

124. Farina, R.; Laugel, N.; Pincus, P.; Tirrell, M., Brushes of strong polyelectrolytes in mixed mono-and tri-valent ionic media at fixed total ionic strengths. Soft Matter 2013, 9 (44), 10458-10472.

125. Yu, J.; Mao, J.; Yuan, G.; Satija, S.; Chen, W.; Tirrell, M., The effect of multivalent counterions to the structure of highly dense polystyrene sulfonate brushes. Polymer 2016, 98, 448-453.

126. De Gennes, P.-G.; Gennes, P.-G., Scaling concepts in polymer physics. Cornell university press: 1979.

127. Nguyen, P. K.; Snyder, C. G.; Shields, J. D.; Smith, A. W.; Elbert, D. L., Clickable Poly (ethylene glycol)‐Microsphere‐Based Cell Scaffolds. Macromolecular chemistry and physics 2013, 214 (8), 948-956.

128. Álvarez-Paino, M.; Muñoz-Bonilla, A.; López-Fabal, F. t.; Gómez-Garcés, J. L.; Heuts, J. P.; Fernández-García, M., Effect of Glycounits on the Antimicrobial Properties and Toxicity Behavior of Polymers Based on Quaternized DMAEMA. Biomacromolecules 2014, 16 (1), 295-303.

129. Jones, R. A.; Poniris, M. H.; Wilson, M. R., pDMAEMA Is Internalised by Endocytosis but Does Not Physically Disrupt Endosomes. Journal of controlled release 2004, 96 (3), 379-391.

124

130. Hannah, R.; Beck, M.; Moravec, R.; Riss, T., CellTiter-Glo™ Luminescent Cell Viability Assay: A Sensitive and Rapid Method for Determining Cell Viability. Promega Cell Notes 2001, 2, 11-13.

131. Brettmann, B. K.; Laugel, N.; Hoffmann, N.; Pincus, P.; Tirrell, M., Bridging contributions to polyelectrolyte brush collapse in multivalent salt solutions. Journal of Polymer Science Part A: Polymer Chemistry 2016, 54 (2), 284-291.

132. Farina, R.; Laugel, N.; Yu, J.; Tirrell, M., Reversible adhesion with polyelectrolyte brushes tailored via the uptake and release of trivalent lanthanum ions. The Journal of Physical Chemistry C 2015, 119 (26), 14805-14814.

133. Vaisocherova, H.; Yang, W.; Zhang, Z.; Cao, Z.; Cheng, G.; Piliarik, M.; Homola, J.; Jiang, S., Ultralow Fouling and Functionalizable Surface Chemistry Based on a Zwitterionic Polymer Enabling Sensitive and Specific Protein Detection in Undiluted Blood Plasma. Analytical chemistry 2008, 80 (20), 7894-7901.

134. Zhang, Z.; Chen, S.; Jiang, S., Dual-Functional Biomimetic Materials: Nonfouling poly (carboxybetaine) with Active Functional Groups for Protein Immobilization. Biomacromolecules 2006, 7 (12), 3311-3315.

135. Shulman, M.; Nahmias, Y., Long-Term Culture and Coculture of Primary Rat and Human Hepatocytes. In Epithelial Cell Culture Protocols, Springer: 2012, pp 287-302.

136. Sistare, F. D.; Mattes, W. B.; LeCluyse, E. L., The Promise of New Technologies to Reduce, Refine, or Replace Animal Use While Reducing Risks of Drug Induced Liver Injury in Pharmaceutical Development. ILAR journal 2016, 57 (2), 186-211.

137. Mironov, V.; Visconti, R. P.; Kasyanov, V.; Forgacs, G.; Drake, C. J.; Markwald, R. R., Organ printing: tissue spheroids as building blocks. Biomaterials 2009, 30 (12), 2164- 2174.

138. Jain, R. K., Transport of molecules, particles, and cells in solid tumors. Annual review of biomedical engineering 1999, 1 (1), 241-263.

139. Darcy, H., Les fontaines publiques de la ville de Dijon: exposition et application. Victor Dalmont: 1856.

140. Brochard, F.; De Gennes, P., Dynamical scaling for polymers in theta solvents. Macromolecules 1977, 10 (5), 1157-1161.

141. Li, Y.; Xue, H.; Song, Y., Production and Purification of Carboxylic Betaine Zwitterionic Monomers. Google Patents: 2014.

125

142. Semsarilar, M.; Ladmiral, V.; Blanazs, A.; Armes, S., Cationic Polyelectrolyte-Stabilized Nanoparticles via RAFT Aqueous Dispersion Polymerization. Langmuir 2012, 29 (24), 7416-7424.

126

BIOGRAPHICAL SKETCH

Christopher S. O’Bryan received his B.S. degree in aerospace engineering and his M.S. in mechanical engineering from the University of Florida. In 2019, he received his Ph.D. in mechanical engineering for his work in developing aqueous and organic microgels for soft matter 3D printing applications, working with Dr. Thomase E. Angelini. In 2017, he received the

Student Research Award from the University of Florida for his work in developing micro- organogels for silicone 3D printing applications. Chris’s research continues to focus on understanding the destabilizing effects of interfacial tension on structures 3D printed into sacrificial support materials, as well as understanding the role of polyelectrolyte – ion interactions on the rheological behavior of jammed microgel systems.

127