CARBOHYDRATE-SILICA NANOPARTICLES FOR SIRNA DELIVERY:

SYNTHESIS, CHARACTERIZATION, AND GENE DELIVERY

By

Olivia Mariel Chesniak

A DISSERTATION

Michigan State University

in partial fulfillment of the requirements

Submitted to

for the degree of

Chemistry – Doctor of Philosophy

2019

ABSTRACT

CARBOHYDRATE-SILICA NANOPARTICLES FOR SIRNA DELIVERY:

SYNTHESIS, CHARACTERIZATION, AND GENE DELIVERY

By

Olivia Mariel Chesniak

RNA interference (RNAi) has long been pursued for its therapeutic potential. Sequence-specific
knockdown of gene expression requires that small interfering RNA (siRNA) gain access to cellular
cytoplasm, presenting difficulties for both the transport of nucleic acids to cells and their voyage
across cellular membranes. Numerous materials are under development as siRNA delivery vehicles
to address this need. The carbohydrate dextran has been incorporated into amine-functionalized sil-
ica nanoparticles (Dex-SiO2-NPs), enhancing their biocompatibility and success as siRNA delivery
vehicles. Inspired by the work of Stöber and others, reagent concentrations in the synthesis of Dex-
SiO2-NPs have been adjusted to tune nanoparticle diameter. The size, shape, and morphology of
Dex-SiO2-NPs have been characterized using transmission electron microscopy (TEM) and energy
dispersive x-ray spectroscopy (EDS). These methods have revealed that Dex-SiO2-NPs decrease in
silicon density toward their centers, when compared with SiO2-NPs. Thermal and porosity analysis
were used to profile Dex-SiO2-NPs both containing dextran and after its removal by calcination.
Having measured an increase in mesopores and decrease in micropores with calcination, it has
been concluded that dextran serves as a porogen in Dex-SiO2-NP synthesis. Not only does dextran
imbue these materials with unique morphology, it also enhances their function as delivery vehicles.
Dex-SiO2-NPs improve enhanced green fluorescent protein (EGFP) supression compared to silica
nanoparticles synthesized in the absence of dextran in human lung and kidney cells in vitro.

Copyright by
OLIVIA MARIEL CHESNIAK
2019

“As she waits for the future
she leans into the present,

for there are truly
beautiful things to
be gathered here."

Morgan Harper Nichols

This dissertation is the culmination of many years of work as both a scientist and a person.

I am honored to have grown through it.

iv

ACKNOWLEDGEMENTS

This work would not have been possible without the time and talents of numerous individuals. I
did not spend much time with my first advisor, Dr. Gregory Baker, but will always remember him
describing the microscopy images displayed outside his research laboratory as both science and art;
I knew immediately that I had to work with him. Unfortunate circumstances led me to Dr. Milton
R. Smith, III’s research group. He graciously joined a collaboration that neither of us were very
familiar with, and we learned many things about RNAi and silica together. Dr. Cristina Chan and
Dr. S. Patrick Walton have provided invaluable guidance in my research and, along with Daniel
Vocelle, have been skilled collaborators who have challenged me to become a truly interdisciplinary
scientist. I am also grateful to Dr. Xuefei Huang and Dr. John Frost, for serving on my committee.
Many have been incredibly generous with their time and equipment. I’m incredibly grateful to
Dr. T.J. Pinnavaia for his willingness to discuss my results and hypotheses with me, Dr. Xudong Fan
and Dr. Alicia Withrow for microscopy training, Dr. Daniel Holmes for assistance with solid state
NMR experiments, Dr. Kathy Severin for use of IR and UV-Vis instrumentation, Dr. Xuefei Huang
and his research group, Dr. Thomas Hamann and his research group for use of their furnace, Dr.
Gary Blanchard and his research group for use of their TGA, Dr. Michael Rich and the Composite
Materials and Structures Center for training and assistance with surface area and porosity analysis,
and Dr. Scott Calabrese Barton and his research group, particularly Yuanchao Liu, Alex Mirabal,
and Dr. Jim Woodward, for use of their surface area and porosity analysis equipment.

The Smith research group has been a wonderful environment to learn in, and for that I thank
past and present group members: Dr. Behnaz Ghaffari, Dr. Yu-Ling Lien, Dr. Kristin Shannon, Dr.
Tim Shannon, Dr. Susanne Miller, Dr. Shawn Preshlock, Dr. Dmitry Shabasov, Dr. Buddhadeb
Chattopadhyay, Po-Jen Hsiao, Reza Ghazfar, Mona Maleka Ashtiani, Alex O’Connell, Seokjoo Li,
and Ryan Fornwald. I also want to thank Aaliyah Jeter for her excellent work and enthusiasm as a
summer undergraduate researcher.

I am continually grateful for my friends who have grown with me and shared their invaluable

v

wisdom both personally and professionally. Dr. Georgina Comiskey, Dr. Nicole Hough, Dr.
Shannon Kraemer, and Dr. Monica Carey are the fiercest women I know and have been incredibly
generous with their time and talents. Many others have also made an impact on me in my time at
MSU.

My family has provided immense support and made significant sacrifices throughout my edu-
cation, and I can’t express my appreciation enough. I’m especially grateful for my father’s many
phone calls and jokes, chocolate and kabanos from my grandma, care packages from my brother,
sister-in-law, and nieces, and to the women of my family who inspired and encouraged me most in
pursuing my education: my mother Joyce Chesniak, my sister Deanna Chesniak, and my cousin Dr.
Laura Chesniak-Phipps. Lastly, I thank my husband, Dr. David Olson, for continually supporting
me, believing in my ability to accomplish my goals, and for helping make Lansing, Michigan
home.

vi

TABLE OF CONTENTS

.
.
.

.
.
.

.
.
.
.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.
LIST OF TABLES .
.
.
.
LIST OF FIGURES .
.
.
LIST OF SCHEMES .
.
.
KEY TO ABBREVIATIONS .

.
.
.
.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
INTRODUCTION .
1.1 RNA Interference and the Delivery Debacle . . . . . . . . . . . . . . . . . . . .
.
1.2 Silica Nanoparticle Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3 Porogens and Templates in Silica Nanoparticle Synthesis . . . . . . . . . . . . . .
1.4 Characterization of Silica Nanoparticles . . . . . . . . . . . . . . . . . . . . . . .

CHAPTER 1

1
1
3
6
8

ix

x
. xviii
. xix

.
.

.
.

.
.

.
.

.
.

CHAPTER 2 ESTABLISHING SIZE CONTROL OF DEXTRAN-SILICA NANOPAR-

TICLES .
.
. .
Introduction .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2 Synthetic Optimization of Dex-SiO2-NP . . . . . . . . . . . . . . . . . . . . . . . 18
2.3 Size Control in Modified Stöber Silica Nanoparticle Synthesis
. . . . . . . . . . . 23
2.4 Growth of Dex-SiO2-NPs .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.5 Conclusions .
CHAPTER 3 UNRAVELING DEXTRAN’S ROLE IN DEX-SIO2-NP MORPHOLOGY . 41
3.1
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.2 Characterization of Dex-SiO2-NP Morphology . . . . . . . . . . . . . . . . . . . 41
3.3 Verifying Incorporation of Dextran into Dex-SiO2-NPs . . . . . . . . . . . . . . . 46
3.4 Removal of Dextran and Post-Mortem Porosity Analysis . . . . . . . . . . . . . . . 49
3.5 Conclusions .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

Introduction .

. .

. .

. .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

. .

Introduction .

CHAPTER 4 DESIGN OF AMINE-COATED DEXTRAN-INCORPORATED SILICA

NANOPARTICLES AS SIRNA DELIVERY VEHICLES . . . . . . . . . . 61
4.1
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.2 Nanoparticle Delivery Vehicle Design and Characterization . . . . . . . . . . . . . 61
4.3 Nanoparticle Delivery Vehicle Performance . . . . . . . . . . . . . . . . . . . . . 64
4.4 Endocytosis of Dex-SiO2-NPs In Vitro . . . . . . . . . . . . . . . . . . . . . . . . 67
4.5 Conclusions .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
CHAPTER 5 FUTURE RESEARCH DIRECTIONS . . . . . . . . . . . . . . . . . . . . 70
. 70
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
CHAPTER 6 EXPERIMENTAL METHODS . . . . . . . . . . . . . . . . . . . . . . . . 71

5.1 Expanding Synthetic Directions . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.2 Silencing Implications .

. .

.

.

.

.

.

.

vii

.

. .

.
.
. .
6.1 Materials
.
.
6.2 Characterization .
.
6.3 Synthetic Procedures
.
.
6.4 Post-Synthetic Treatment
.
.

.

.

.

.

.

.
.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
.
NMR SPECTRA . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
SUPPLEMENTAL TEM MICROGRAPHS . . . . . . . . . . . . . 92
SUPPLEMENTAL FIGURES . . . . . . . . . . . . . . . . . . . . . 126
.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

.

.

.

.

.

APPENDICES .
APPENDIX A
APPENDIX B
APPENDIX C
REFERENCES .

.

.

.

.

.

viii

LIST OF TABLES

Table 2.1: Nanoparticle Diameter as a Function of Reagent Concentrations . . . . . . . . . 24

Table 2.2: Dex-SiO2-NP Diameters with Dextran Variation . . . . . . . . . . . . . . . .

. 32

Table 3.1: Percent Weight Lost of Dex-SiO2-NPs, SiO2-NPs, and Dextran by Thermo-

gravimetric Analysis .

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

Table 3.2: Percent Weight Lost of Dex-SiO2-NPs after Soxhlet Extraction by Thermo-

gravimetric Analysis .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

Table 3.3: Percent Weight Lost of Dex-SiO2-NPs after Calcination by Thermogravimetric

Analysis .

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

.

.

.

.

.

.

.

.

.

.

.

.

Table 3.4: Summary of Surface Area and Porosity of Dex-SiO2-NPs Before and After

Calcination .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

Table 4.1: Dex-SiO2-NPs Amine Quantification by Ninhydrin Assay . . . . . . . . . . . . 63

Table 4.2: Endocytotic Inhibitors and Action Mechanism . . . . . . . . . . . . . . . . . . . 68

Table C.1: Zeta Potentials of Dex-SiO2-NPs . . . . . . . . . . . . . . . . . . . . . . . . . . 128

ix

LIST OF FIGURES

Figure 1.1: Diagram of RNA interference Process

. . . . . . . . . . . . . . . . . . . . . .

Figure 1.2: Types of siRNA Delivery Vehicles in Clinical Trials . . . . . . . . . . . . . . .

Figure 1.3: Average Diameters of SiO2-NPs upon Variation of Reagents . . . . . . . . . . .

Figure 1.4: Carbohydrates Used in Biocompatible Materials . . . . . . . . . . . . . . . . .

Figure 1.5: Schematic Diagram of TEM and SEM . . . . . . . . . . . . . . . . . . . . . .

1

3

4

7

9

Figure 1.6: Schematic Diagram of Light Scattering by DLS . . . . . . . . . . . . . . . .

. 10

Figure 1.7: Schematic Diagram Describing Zeta Potential

. . . . . . . . . . . . . . . . .

. 11

Figure 1.8: Schematic Diagram of Gas Adsorption on a Porous Surface . . . . . . . . . . . 12

Figure 1.9: Physisorption Isotherm Classifications Recommended by IUPAC . . . . . . .

. 13

Figure 1.10: Classification of Hysteresis Loops Observed in Physisorption Isotherms Rec-

ommended by IUPAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

Figure 2.1: TEM Micrograph and Size Distribution of Bimodal Amine-Coated Dex-SiO2-

350 Nanoparticles .

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Figure 2.2: Diagram of Pressure Filtration Apparatus . . . . . . . . . . . . . . . . . . . . . 19

Figure 2.3: TEM Micrographs of Size Exclusion Membranes

. . . . . . . . . . . . . . . . 20

Figure 2.4: TEM Micrographs of Dex-SiO2-350 NPs Filtered Through 0.2 µm Polycar-

bonate Membrane .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 2.5: TEM Micrographs of Dex-SiO2-350 NPs Filtered Through 0.2 µm Polycar-

bonate Membrane .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 20

. 21

Figure 2.6: 1H-NMR Spectra of TEOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

Figure 2.7: TEM Micrographs and Number Distributions of Dex-SiO2-350 NPs

. . . . . . 23

Figure 2.8: TEM Micrographs and Number Distributions of Dex-SiO2-250 NPs
Figure 2.9: 1H-NMR Spectra of APTES . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . 24

. 25

x

Figure 2.10: 13C-NMR Spectra of APTES . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

Figure 2.11: TEM Micrographs and Number Distributions of Dex-SiO2-NPs . . . . . . . . . 27

Figure 2.12: TEM Micrographs of Dex-SiO2-NPs upon Variation of Reagents . . . . . . . . 28

Figure 2.13: Average Diameters of Dex-SiO2-NPs upon Variation of Reagents . . . . . . . . 29

Figure 2.14: TEM Micrographs and Number Distributions of Dex-SiO2-NPs . . . . . . . . . 30

Figure 2.15: TEM Micrographs and Number Distributions of Dex-Amine-SiO2-NPs . . . . . 30

Figure 2.16: TEM Micrographs and Number Distributions of Dex-SiO2-NPs . . . . . . . . . 31

Figure 2.17: Structures of Dextran and Cyclodextrins

. . . . . . . . . . . . . . . . . . . .

. 31

Figure 2.18: TEM Micrographs of Dex-SiO2-NPs Prepared with Different Dextrans . . . . . 33

Figure 2.19: TEM Micrographs of Purified Reaction Mixture during Dex-SiO2-350 NP

Synthesis .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 2.20: TEM Micrographs of Purified Reaction Mixture during Dex-SiO2-350 NP

.

.

.

.

.

.

.

.

.

.

.

.

.

.

. 34

. 35

Synthesis .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 2.21: TEM Micrograph of Dex-SiO2-350 NPs During Synthesis . . . . . . . . . . . . 36

Figure 2.22: Diagram Depicting Dex-SiO2-NP Synthesis and Dextran Solubility Hypothesis

36

Figure 2.23: Hydrodynamic Diameter of Aggregates During Dex-SiO2-NP Synthesis

. . . . 38

Figure 2.24: TEM Micrographs and Number Distributions of Dex-SiO2-NPs Prepared

Using Methanol, Ethanol, and Isopropanol

. . . . . . . . . . . . . . . . . . . . 39

Figure 2.25: TEM Micrographs and Number Distributions of SiO2-NPs Prepared Using

Methanol, Ethanol, and Isopropanol

. . . . . . . . . . . . . . . . . . . . . .

. 39

Figure 3.1: TEM Micrograph of SiO2-NPs and Dex-SiO2-NPs . . . . . . . . . . . . . . . . 42

Figure 3.2: TEM Micrographs of Dex-SiO2-NPs . . . . . . . . . . . . . . . . . . . . . . . 43

Figure 3.3: STEM Micrographs and EDS Line Scan of Dex-SiO2 and SiO2 Nanoparticles . 44

Figure 3.4: Hydrodynamic Diameter of Dextran in Water . . . . . . . . . . . . . . . . . . . 45

Figure 3.5: Hydrodynamic Diameter of Dextran Aggregates in Ethanol

. . . . . . . . . . . 45

xi

Figure 3.6: TGA Profile of Dex-SiO2-NPs

. . . . . . . . . . . . . . . . . . . . . . . . . . 46

Figure 3.7: TGA Profile of Dextran and SiO2 NPs

. . . . . . . . . . . . . . . . . . . . . . 47

Figure 3.8: TEM Micrographs of Dex-SiO2-200 Nanoparticles after TGA . . . . . . . . . . 48
Figure 3.9: Solid State CP-MAS 13C-NMR of Dex-SiO2-350 Nanoparticles . . . . . . . . . 49
Figure 3.10: 500 MHz 1H-NMR of Dextran . . . . . . . . . . . . . . . . . . . . . . . . . . 50

Figure 3.11: TEM Micrographs of Dex-SiO2-NPs After Soxhlet Extraction . . . . . . . . . . 51

Figure 3.12: TGA Profile of Dex-SiO2-NPs Before and After Soxhlet Extraction . . . . . . . 52

Figure 3.13: TGA Profile of Dex-SiO2-NPS Before and After Calcination . . . . . . . . .

. 53

Figure 3.14: TEM Micrographs of Dex-SiO2-NPs After Calcination . . . . . . . . . . . . . 54

Figure 3.15: Nitrogen Adsorption-Desorption Isotherms of Dex-SiO2-NPs . . . . . . . . . . 55

Figure 3.16: Differential Pore Volume of Dex-SiO2-200 NPs Before and After Calcination . 56

Figure 3.17: Differential Pore Volume of Dex-SiO2-250 NPs Before and After Calcination . 57

Figure 3.18: Differential Pore Volume of Dex-SiO2-350 NPs Before and After Calcination . 57

Figure 3.19: BET Fit of Dex-SiO2-NPs Before and After Calcination . . . . . . . . . . . . . 59

Figure 3.20: Schematic Diagram of Dex-SiO2-NPs Synthesis and Calcination . . . . . . . . 60

Figure 4.1: EGFP Silencing by Comiskey Dex-SiO2-NPs and SiO2-NPs . . . . . . . . . . . 62

Figure 4.2: Correlation between Dex-SiO2-NP zeta potential and amount of amine (mmol/mg)

detected by the ninhydrin assay for the NPs listed in Table 4.1 . . . . . . . . . . 63

Figure 4.3: EGFP Silencing of Dex-SiO2-NPs prepared using Dextran . . . . . . . . . . . . 65

Figure 4.4: EGFP Silencing of Dex-SiO2-NPs prepared using Cyclodextrin . . . . . . . . . 66

Figure 4.5: TEM Micrograph of 40% Amine Dex-SiO2-350 in H1299 Cells . . . . . . . . . 66

Figure 4.6: STEM Micrograph and EDS Line Scan of 40% Amine Dex-SiO2-350 in

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

H1299 Cells . .

. .

Figure 4.7: Silencing of EGFP in the Presence of Endocytotic Inhibitors

. . . . . . . . . . 68

xii

. . . . . . . . . . . . . . . . 82

. . . . . . . . . . . . . . . . 84

Figure A.1: 500 MHz 1H-NMR Spectrum of TEOS in CDCl3
Figure A.2: 125 MHz 13C-NMR Spectrum of TEOS in CDCl3 . . . . . . . . . . . . . . . . 83
Figure A.3: 500 MHz 1H-NMR Spectrum of TEOS in CDCl3
Figure A.4: 125 MHz 13C-NMR Spectrum of TEOS in CDCl3 . . . . . . . . . . . . . . . . 85
Figure A.5: 500 MHz 1H-NMR Spectrum of APTES in CDCl3
Figure A.6: 125 MHz 13C-NMR Spectrum of APTES in CDCl3 . . . . . . . . . . . . . . . 87
Figure A.7: 500 MHz 1H-NMR Spectrum of APTES in CDCl3
Figure A.8: 125 MHz 13C-NMR Spectrum of APTES in CDCl3 . . . . . . . . . . . . . . . 89
Figure A.9: 500 MHz 1H-NMR Spectrum of Dextran in D2O . . . . . . . . . . . . . . . . . 90
Figure A.10:500 MHz 1H-NMR Spectrum of Dextran in D2O/DMSO-d6.

. . . . . . . . . . . . . . . 86

. . . . . . . . . . . . . . . 88

. . . . . . . . . . 91

Figure B.1: EDS Spectrum of Dex-SiO2-NPs . . . . . . . . . . . . . . . . . . . . . . . . . 93

Figure B.2: EDS Spectrum of SiO2-NPs . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 94

Figure B.3: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-

SiO2-200 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

Figure B.4: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-

SiO2-200 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

Figure B.5: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-

SiO2-200 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

Figure B.6: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-

SiO2-250 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

Figure B.7: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-

SiO2-250 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

Figure B.8: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-

SiO2-250 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

Figure B.9: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-

SiO2-350 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

xiii

Figure B.10: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-

SiO2-350 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

Figure B.11: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-

SiO2-350 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

Figure B.12: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and

Dex-SiO2-200 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

Figure B.13: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and

Dex-SiO2-200 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

Figure B.14: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and

Dex-SiO2-200 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

Figure B.15: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and

Dex-SiO2-250 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

Figure B.16: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and

Dex-SiO2-250 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

Figure B.17: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and

Dex-SiO2-250 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

Figure B.18: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and

Dex-SiO2-350 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

Figure B.19: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and

Dex-SiO2-350 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

Figure B.20: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and

Dex-SiO2-350 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

Figure B.21: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dex-

tran and Dex-SiO2-200 Conditions . . . . . . . . . . . . . . . . . . . . . . . . 105

Figure B.22: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dex-

tran and Dex-SiO2-200 Conditions . . . . . . . . . . . . . . . . . . . . . . . . 105

Figure B.23: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dex-

tran and Dex-SiO2-200 Conditions . . . . . . . . . . . . . . . . . . . . . . . . 106

Figure B.24: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dex-

tran and Dex-SiO2-250 Conditions . . . . . . . . . . . . . . . . . . . . . . . . 106

xiv

Figure B.25: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dex-

tran and Dex-SiO2-250 Conditions . . . . . . . . . . . . . . . . . . . . . . . . 107

Figure B.26: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dex-

tran and Dex-SiO2-250 Conditions . . . . . . . . . . . . . . . . . . . . . . . . 107

Figure B.27: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dex-

tran and Dex-SiO2-350 Conditions . . . . . . . . . . . . . . . . . . . . . . . . 108

Figure B.28: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dex-

tran and Dex-SiO2-350 Conditions . . . . . . . . . . . . . . . . . . . . . . . . 108

Figure B.29: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dex-

tran and Dex-SiO2-350 Conditions . . . . . . . . . . . . . . . . . . . . . . . . 109

Figure B.30: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and

Dex-SiO2-200 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 110

Figure B.31: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and

Dex-SiO2-200 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 110

Figure B.32: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and

Dex-SiO2-200 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 111

Figure B.33: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and

Dex-SiO2-250 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 111

Figure B.34: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and

Dex-SiO2-250 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 112

Figure B.35: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and

Dex-SiO2-250 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 112

Figure B.36: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and

Dex-SiO2-350 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 113

Figure B.37: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and

Dex-SiO2-350 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 113

Figure B.38: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and

Dex-SiO2-350 Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 114

Figure B.39: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-

SiO2-200 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

xv

Figure B.40: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-

SiO2-200 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

Figure B.41: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-

SiO2-200 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

Figure B.42: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-

SiO2-250 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

Figure B.43: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-

SiO2-250 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

Figure B.44: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-

SiO2-250 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

Figure B.45: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-

SiO2-350 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

Figure B.46: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-

SiO2-350 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

Figure B.47: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-

SiO2-350 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

Figure B.48: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-

SiO2-200 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

Figure B.49: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-

SiO2-200 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

Figure B.50: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-

SiO2-200 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Figure B.51: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-

SiO2-250 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Figure B.52: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-

SiO2-250 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

Figure B.53: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-

SiO2-250 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

Figure B.54: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-

SiO2-350 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

xvi

Figure B.55: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-

SiO2-350 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Figure B.56: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-

SiO2-350 Conditions

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124

Figure B.57: TEM Micrograph of 40% Amine Dex-SiO2-350 in HELA Cells . . . . . . . . . 125

Figure C.1: UV-Visible Spectrum of Rheumann’s Purple at Variable Concentrations

. . . . 127

Figure C.2: Calibration Curve of Rheumann’s Purple . . . . . . . . . . . . . . . . . . . . . 127

xvii

LIST OF SCHEMES

Scheme 1.1: Hydrolysis and Condensation of Alkyl Silicates to form SiO2

. . . . . . . . .

Scheme 1.2: Schematic Depiction of LaMer Monomer Addition Process for Nanoparticle

Growth .

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.

.

.

.

.

.

.

.

.

.

.

.

5

5

6

6

8

Scheme 1.3: Schematic Depiction of Aggregation and Growth Process for Nanoparticle

Synthesis .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Scheme 1.4: Schematic Depiction of Seeded Growth Process for Nanoparticle Synthesis .

.

Scheme 1.5: Schematic Depiction of Nanoparticle Synthesis with Organic Template . . . .

Scheme 4.1: Dex-SiO2-NP Synthetic Scheme . . . . . . . . . . . . . . . . . . . . . . . . . 62

Scheme 4.2: General Reaction of Primary Amine and Ninhydrin to Produce Ruhemann’s

Purple

.

.

.

.

.

.

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

xviii

KEY TO ABBREVIATIONS

°C Degree Celsius
α-CD alpha-Cyclodextrin
APTES 3-aminopropyl triethoxy silane
BET Brunauer, Emmet, and Teller
BJH Barrett, Joyner, and Halenda
CP-MAS Cross-polarization magic angle spinning
β-CD beta-Cyclodextrin
D2O Deuterium dioxide
d6-DMSO Deuterated dimethylsulfoxide
Da Dalton
Dex-SiO2-NP Dextran containing SiO2 nanoparticle
DFT Dynamic functional theory
DLS Dynamic light scattering
DNA Deoxyribonucleic acid
dsRNA Double stranded RNA
EDS Energy dispersive X-ray spectroscopy
eV Electron volts
FT Fourier transform
g Gram
EGFP Enhanced green fluorescent protein
h Hour
IR Infrared
L Liter
LF2K Lipofectamine 2000
Hz Hertz

xix

mol Mole
MWCO Molecular weight cutoff
NMR Nuclear magnetic resonance
NP Nanoparticle
psi Pounds per square inch
RH Hydrodynamic radius
RNA Ribonucleic acid
RNAi RNA interference
RISC RNA-induced silencing complex
rt Room temperature
S-TEM Scanning transmission electron microscopy
siRNA Small interfering RNA
TEM Transmission electron microscopy
TEOS Tetraethyl ortho silicate
TGA Thermogravimetric analysis
UV-Vis Ultraviolet visible

xx

CHAPTER 1

INTRODUCTION

1.1 RNA Interference and the Delivery Debacle

RNA interference (RNAi) is an endogenous pathway of gene suppression which can be exoge-
nously activated, giving rise to immense therapeutic potential. Fire and Mello first described RNAi
in C. elegans in 1998.1 Soon after, Elbashir and co-workers established siRNA-mediated gene
silencing in mammalian cells.2 Small interfering RNA (siRNA) are 21-23 nucleotides long and
initiate RNAi upon detection by proteins within the cellular cytoplasm. An assembly of proteins
known as the RNA-induced silencing complex (RISC) is activated by binding the guide strand of
siRNA. mRNA containing a sequence complimentary to the siRNA are targeted by RISC, which
are then degraded by the Argonaute RNase, leading to suppression of the target protein (Figure
1.1).3 In this way, an RNAi therapeutic addresses the origin of a protein-based disease.

Figure 1.1: Diagram of RNA interference Process. Image ©2017 Elsevier3

1

Nevertheless, the potential to exogenously activate RNAi using a tailored sequence of siRNA
brings with it several challenges. Nucleases in the blood can degrade siRNA, leading to low in vivo
half lives and poor drug performance. Additionally, the negative charge of siRNA limit their ability
to traverse cell membranes. These challenges have been addressed through the use of delivery
vehicles, which protect siRNA and assist in delivering it into cellular cytoplasm.4–6

When designing vehicles for siRNA delivery, size, surface functionalization, and mode of
siRNA binding are at the forefront of design considerations. Size and shape of nanoparticles (NPs)
play a role in determining their recognition by phagocytes and trafficking through endocytotic
pathways. NP surface functionality also plays a role in determining the toxicity of the delivery
vehicle, phagocytic recognition, and provides a convenient platform upon which to include moieties
to target specific tissues or cells. Further, the presence of different components on NP surfaces will
in turn impact surface charge, which plays a role in the complex’s biodistribution and ability to
cross cellular membranes. The binding and protection of siRNA can be accomplished via surface
attachment (electrostatic or covalent) or encapsulation within hollow NP pores or centers.7 Each of
these design considerations may be applied to different types of non-viral delivery vehicles, whether
polymeric, lipid-based, or solid. While lipid-based delivery vehicles can cross the membrane,
others gain access to the cellular cytoplasm via endocytosis. While the majority of clinical trials in
RNAi therapeutics have employed naked siRNA, lipid nanoparticles follow by only 10% (Figure
1.2). Polymer delivery vehicles and N-galactosamine (GalNAc)-siRNA conjugates have also been
investigated, each was represented in 6.6% of clinical trials as of 2016.8

The first drug candidate using RNAi recently obtained FDA approval to treat a hereditary
peripheral nerve disease, hereditary transthyretin-mediated amyloidosis (hATTR). Resulting from
build up of amyloid protein fibers in organs and tissues, treatment of hATTR by RNAi reduces the
production of the proteins which cause the disease. Onpattro (patisiran), developed by Alnylam
Pharmaceuticals, Inc., is a lipid nanoparticle which encapsulates siRNA and is intravenously
administered directly to the liver.9 Though the first of its kind, it will likely not be the last. There
are currently 16 active clinical trials for siRNA-based drugs.10

2

Figure 1.2: Types of siRNA Delivery Vehicles in Clinical Trials. Image ©2016 American Chemical
Society8

The breadth of siRNA delivery vehicles offers incredible opportunity, but also many questions
regarding the role of delivery vehicle characteristics within the RNAi pathway. This work extends
the understanding of the role of dextran in siRNA delivery vehicles by investigating its role in
silica nanoparticle synthesis.11,12 By extension, dextran’s location within delivery vehicles and
its interaction with biological environments may be further understood. Controlling size, shape,
and surface functionalization using silica nanoparticles, Comiskey found that dextran-infused silica
delivery vehicles elicited greater silencing of EGFP than non-dextran containing silica.11
1.2 Silica Nanoparticle Synthesis

Precise size control and straightforward one-pot conditions contribute to the widespread use
of sol-gel techniques for the synthesis of silica nanoparticles. Aqueous acid- or base-catalyzed
hydrolysis and condensation of alkoxysilanes produce monodispersed nanoparticles of specific
diameters.13–16 These methods were first reported by Kolbe and further developed by Stöber, Fink,
and Bohn in 1968, with the synthesis of uniform silica particles ranging from 0.2 to 2.0 µm in
diameter (Figure 1.3).17,18 The cornerstone of the so-called Stöber method is the ease with which
nanoparticle diameters are controlled via ammonium hydroxide concentration. Silica nanoparticles
ranging in size from 10 nm to 800 nm have been prepared using this method.19

Silica is formed via hydrolysis of an alkoxy silicate precursor, often tetraethyl orthosilicate

3

Figure 1.3: Average Diameters of SiO2-NPs upon Variation of Reagents. Image ©1968 Elsevier18

(TEOS), to form a silanol and subsequent condensation of silanols to form a network of siloxanes
(Scheme 1.1). Under basic conditions, the alkoxy silicate undergoes nucleophilic attack by a
2 fashion to produce a silanol and an ethoxide anion. Silanol proton
hydroxide ion in an SN
abstraction by hydroxide forms a siloxide ion and water in the first condensation step. Attack of
the siloxide ion on the silicon atom of a neighboring silanol results in siloxane formation. Loss
of hydroxide by the silanol regenerates the hydroxide catalyst.13,20,21 Continued hydrolysis and
condensation extend the siloxane network to form nanoparticles.

The hydrolysis rate increases with ammonium hydroxide concentration, as hydroxide ions are
better nucleophiles than water. The rate of condensation is greater than that of hydrolysis, as silanol
groups are more easily deprotonated than water, making silicon atoms of silanols more positively
charged and poised for nucleophilic attack. Silanol molecules therefore prefer to condense onto
larger siloxane aggregates as opposed to smaller aggregates or dimerization with other silanol
molecules.20,22,23

Nanoparticle growth has been widely studied. Stöber silica was initially thought to proceed
via the LaMer model for nanoparticle growth, which hypothesizes that an initial nucleation burst

4

Scheme 1.1: Hydrolysis and Condensation of Alkyl Silicates to form SiO2.

Scheme 1.2: Schematic Depiction of LaMer Monomer Addition Process for Nanoparticle Growth.

is followed by growth via monomer addition to aggregates (Scheme 1.2).24 This model is useful
for describing formation of silica nanoparticles at high concentrations of ammonium hydroxide.25
Bogush and Zukoski later reported that the concentration of hydrolyzed TEOS is greater than what
is required for nucleation until late in the reaction, suggesting that nanoparticle growth arises from
aggregation of siloxane networks (Scheme 1.3).26 In this model, nanoparticle size and polydisper-
sity would be governed by size-dependent aggregation rates governed by the colloidal stability of
particles. This model also accounts for the polydispersity often observed in Stöber silica popula-
tions, if large particles were to aggregate with other large particles instead of with smaller particles.
Building on the experimental results of Bogush and Zukoski, Lee constructed an aggregation-based
kinetic model which accurately represents diameters of Stöber silica nanoparticles at low ammonia
concentrations.25

Han and co-workers studied the direct impacts of TEOS hydrolysis and condensation kinetics

5

Scheme 1.3: Schematic Depiction of Aggregation and Growth Process for Nanoparticle Synthesis.

Scheme 1.4: Schematic Depiction of Seeded Growth Process for Nanoparticle Synthesis.

on nanoparticle size over time, reconciling the relationship between nucleation and aggregation in
silica nanoparticle synthesis. They concluded that two pathways exist: TEOS hydrolysis dominates
the first pathway, involving nucleation and early stages of silanol growth into small aggregates.
The second pathway is comprised of silanol condensation onto aggregates formed in the first
pathway (Scheme 1.4).20 This in situ seeded growth process builds on both the LaMer and Bogush
models and provides a more thorough understanding, leading to improvements in silica nanoparticle
synthesis by way of smaller diameters and narrower distributions.
1.3 Porogens and Templates in Silica Nanoparticle Synthesis

Functionalized silica nanoparticles have been shown to be successful drug and gene delivery
vehicles which exhibit low cytotoxicity.27–37 In this work, the carbohydrate dextran has been incor-
porated into the Stöber Method to enhance the biocompatibility of the resulting silica nanomaterials.
Dextrans are polysaccharides having a linear backbone made up of α(1-6) linked d-glucopyranosyl
units and shorter d-glucose side chains linked in an α(1-2), α(1-3), or α(1-4) manner.38 In a
therapeutic setting, dextrans are used as anticoagulants, whereas cross-linked dextran particles are
used in biochemical research settings for size exclusion chromatography.39 Dextran and other
polysaccharides have been used for delivery and enhancing biocompatibility of drugs and other
materials both independently and via nanoparticle incorporation (Figure 1.4).40–48

6

Figure 1.4: Carbohydrates Used in Biocompatible Materials.

Carbohydrates including some of those included in Figure 1.4 have previously been incorpo-
rated into silica nanostructures.49–53 Lin and Sinha have both incorporated dextran onto the surface
of mesoporous silica nanoparticles as gates, preventing payload release until certain biological con-
ditions were met.54,55 The hydrophilic/hydrophobic nature of cyclodextrins make them excellent
for guest-host systems has been used to protect drugs within mesopores of silica nanoparticles.56,57
Mansfiled and co-workers condensed a thiol-alkoxysilane in the presence of hydroxyethylcellulose,
resulting in amorphous nanoparticles ranging from 50 - 500 nm.58 The surface of silica nanopar-
ticles were functionalized with chitosan and hyaluronic acid by Salis and co-workers, exhibiting
high proliferation into cells and low cytotoxicity.59 Silvestri and co-workers condensed tetrapropyl
orthosilicate with heparin, resulting in hollow nanoparticles less than 100 nm in diameter.60

Surfactants including cetyltrimethylammonium bromide (CTAB) are used to template pores in
Stöber silica nanoparticles. The amphiphilic nature of the surfactant directs the hydrolysis and
condensation of an alkoxy silicate around surfactant assemblies, resulting in formation of pores
having precise size and shape according to the surfactant used (Scheme 1.5).13,61–64 Aside from
carbohydrates and surfactants, polymers and polymer-based nanoparticles have been incorporated
into silica structures.65 Organic materials have been used both as templates for silica nanopar-
ticle synthesis and as surface coatings. Polyamines and polylysine have been widely studied, as

7

Scheme 1.5: Schematic Depiction of Nanoparticle Synthesis with Organic Template.

their cationic character can electrostatically direct TEOS hydrolysis and condensation and enables
electrostatic binding of nucleic acids.66–68

As depicted in Scheme 1.5, porogens used in silica or ceramic synthesis are often removed
before materials are used. Organic porogens may be washed or rinsed out of the structures using
solvents or by thermal degradation under oxidative conditions. Ruckdeschel templated silica
onto polystyrene nanoparticles, producing hollow silica nanoparticles with uniform thickness and
dispersity after calcination.69 However, porogens may impart unique properties to materials and
are therefore not removed. This is the case for the first generation of Dex-SiO2-NPs synthesized
by Comiskey, the second generation of which are described herein.11,12 Dex-SiO2-NPs exhibit
a hollow core, which has been observed in silica nanoparticles synthesized using polymers and
aggregates as porogens.70–73 While studies in the literature suggest the location of dextran within
Dex-SiO2-NPs, thorough characterization of these materials has been pursued with the ultimate
goal of elucidating the role dextran plays in silica nanoparticle synthesis.
1.4 Characterization of Silica Nanoparticles

Silica nanoparticle characterization involves bridging methods from both chemistry and ma-
terials science. In this work, Dex-SiO2-NPs are characterized using an array of well-established
techniques.

Since Nuclear Magnetic Resonance (NMR) analysis of colloidal solutions is complicated due
to their inherent lack of solubility, solid state NMR has become a prevalent analysis method for

8

Figure 1.5: Schematic Diagram of TEM and SEM. Adapted from reference 77.

silica nanoparticles.74 13C cross-polarization magic angle spinning (CP-MAS) minimizes line
broadening often observed in solid state NMR by rapidly spinning the sample at 54.74 ◦relative
to the magnetic field.
In doing so, anisotropic interactions between nuclei which are typically
averaged by Brownian motion in solution NMR can be averaged to zero. Cross-polarization is
applied to enhance the signal from less abundant nuclei, including 13C.75,76

Electron microscopy (EM) provides visual details of nano-scale materials and are widely used
in characterization of silica nanoparticles and other materials. EM uses an accelerated electron
beam which interacts with a sample, resulting in emission of electrons, x-rays, and photons. The
contrast observed in EM originates from the different electron densities of elements comprising
the material under analysis.78–81 Depending on the arrangement of the sample relative to various
lenses and the power of the beam, different details are apparent. A thick sample will scatter
electrons rather than transmit them through the sample. With detectors situated above the sample,
these scattered electrons are collected and provide details of surface morphology characteristic of
scanning electron microscopy (SEM). A thinner sample, different lens placement, and detector on
the opposite side of the sample relative the beam collects transmitted electrons in transmission
electron microscopy (TEM) (Figure 1.5). Scanning Transmission Electron Microscopy (S-TEM)
brings with it the opportunity to use x-ray Energy Dispersive Spectroscopy (EDS) to study the
elemental profile of a defined sample area. S-TEM utilizes a convergent electron beam to scan
the user-defined sample area. Upon being scanned by a convergent electron beam, the elements

9

Figure 1.6: Schematic Diagram of Light Scattering by DLS. Adapted from reference 85.

within the sample emit characteristic x-rays which are analyzed by an electron detector. Analytical
software uses this data to construct a spectra of elements present in the sample.78,82–84 The size,
morphology, and elemental details obtained using electron microscopy provide significant insight
to the quality and characteristics of materials.

Dynamic light scattering (DLS) uses the speed of particles diffusing according to Brownian
motion to determine hydrodynamic size. The rate at which the intensity of scattered light fluctuates
is measured. These rates are dependent on the size of particles, with larger particles taking longer
to diffuse than smaller particles.85,86 The Zetasizer NanoZS uses a backscatter detector to measure
the scattering by particles interacting with the laser (Figure 1.6). The backscatter angle near
180◦decreases the effects of multiple scattering and larger dust contaminants, which are more
likely to scatter forward. Since the incident beam does not travel through the entire sample cell
and has a smaller path length, the concentration of sample measured increases while decreasing
the collection of light scattered by a particle by other particles. The scattering intensity collected
by the detector is interpreted by the correlator, which compares changes to the scattering pattern
over time in order to derive the rate of change. Since the movement of particles follows Brownian
motion, the Zetasizer software uses the Stokes-Einstein equation to calculate the particle size based
on these data.85,86

The electrophoretic mobility and velocity of nanoparticles, measured via laser doppler ve-
locimetry (LDV) while an electric field is applied to the suspension, are used to determine the

10

Figure 1.7: Schematic Diagram Describing Zeta Potential. Adapted from reference 85.

stability of particles in a suspension. Using the Henry equation, these measurements are used to
calculate the ζ potential.85–87 As depicted in Figure 1.7, a particle existing in a solution bears a
net surface charge, increasing the concentration of counter ions near the surface. These strongly
bound counter ions make up the Stern layer, while the outer regions are part of the diffuse layer. As
a particle moves in solution, ions within the slipping plane travel with it while those beyond this
region do not. The potential at the slipping plane is the ζ potential. Thus, ζ potential represents
the stability of particles in a suspension: if the magnitude of ζ potential is large, particles will repel
each other and remain well-dispersed, while a low-magnitude ζ potential suggests that there are
not enough ions in the slipping plane to prevent particle aggregation.85,86 As one may expect due
to the importance of ions in these measurements, ζ potential is pH-dependent.

DLS and ζ potential measurements have been widely used in biomedical research.88 Branda
and co-workers used DLS and ζ potential measurements to study the impact of growth media and
incubation conditions on Stöber silica nanoparticles. A decrease in hydrodynamic diameter and a

11

Figure 1.8: Schematic Diagram of Gas Adsorption on a Porous Surface.

decrease in ζ potential were observed after prolonged incubation, suggesting that aggregation of
biomolecules decrease the aggregation of nanoparticles with each other. The DLS results were
consistent with observations made in electron microscopy.89

Gas adsorption for surface area and porosity determination is an important and established part
of characterizing porous solids. Inert gas, often nitrogen at 77 K, is dosed into a sample of porous
solid to equilibrium, at which point the volume of gas is noted, and the dosing proceeds. These
data are represented in an adsorption-desorption isotherm, the shapes of which are indicative of the
porous nature of the analyzed materials (Figure 1.9).90,91

Microporous materials exhibit type I isotherms, with concave shapes relative to the x-axis and
plateaus at low relative pressure (p/p0). Micropores are less than 2 nm in diameter, Type I(a)
isotherms indicate the presence of micropores < 1 nm, while type I(b) indicates a broader pore
size distribution containing either larger micropores (< 2 nm) or a combination of large and small
micropores.91 Type II isotherms are characteristic of macroporous and non-porous materials, with
the gradual increase in volume adsorbed representing adsorption which is less restricted by pore
size. Point B represents the relative pressure at which monolayer coverage is complete. A less
distinctive point B, or no point B as seen in Type III, suggests that multilayer and monolayer

12

Figure 1.9: Physisorption Isotherm Classifications Recommended by IUPAC. Image ©2015 IUPAC
& De Gruyter91

coverage are indistinguishable. Mesoporous (2 - 50 nm) materials have Type IV isotherms, with
IV(a) depicting hysteresis arising from capillary condensation and IV(b) being reversible. While
the low p/p0 regions of Type IV isotherms resemble that of Type II isotherms, the larger size of
mesopores result in condensation within pores. The eventual plateau represents saturation. Type
V isotherms resemble Type III at low p/p0, attributed to weak interactions between the adsorbent
and adsorbate, often occurring in hydrophobic adsorbents having micro- and macropores. Type
VI isotherms are characteristically stepwise, representing continuous adsorption on uniform, non-
porous materials.90,91

As indicated above, some isotherms are reversible while, for others, the desorption branch
follows a different path than the adsorption branch. The different shapes of these variations,
known as hysteresis, provide further information on pore structures (Figure 1.10). Type H1 is

13

Figure 1.10: Classification of Hysteresis Loops Observed in Physisorption Isotherms Recom-
mended by IUPAC. Image ©2015 IUPAC & De Gruyter91

common in materials having uniform mesopores and ink-bottle-shaped pores, with the hysteresis
representing delayed condensation during adsorption. H2 hysteresis loops are characteristic of
complex networks of pores where percolation and pore blocking may occur. Type H3 and H4 loops
are somewhat similar, with the former resembling the early part of Type II isotherms. This loop is
common among malleable plate-like particles and materials having macropores. Type H4 shares
characteristics with Type I and Type II isotherms, and are present in materials containing micro-
and mesopores. Hysteresis loops of Type H5 are less common, and distinctive of structures having
mesopores that are both open and blocked.91

Adsorption data obtained between p/p0 0.05 - 0.3 are used in the Brunauer-Emmet-Teller method
(BET) to determine surface area. This region of the isotherm is typically linear, and application of
the BET equation to this region provides an indication of the surface area of a sample. The linear
form of the BET equation,

p/p0

n(1 − p/p0) =

1

nm ∗ C

+

C − 1
nm ∗ C

∗ (p/p0)

(1.1)

is applied, where n is the specific volume adsorbed at relative pressure p/ptextsuperscript0, and

14

nm is the specific monolayer capacity. The term C, related to the energy of monolayer adsorption,
provides an indication of the isotherm shape. The specific surface area as is calculated using

nm ∗ L ∗ σm

m

as =

(1.2)

where σm is the average molecular cross-sectional area of the adsorbate molecule (0.162 nm2 for
N2 at 77 K), and m is the mass of the adsorbent.90–95 In addition to characterization by isotherm
shapes, hysteresis loops, and BET surface area determination, density functional theory (DFT) is
useful in providing a more detailed view of porous structures. Since pore characteristics and their
correlation with different types of isotherms are well established, the pore size distribution may be
calculated by integrating the pore size distribution f(H) and theoretical isotherm ρ(P, H),

Þ ∞

0

f(H)ρ(P, H) dH

(1.3)

known as the integrated adsorption equation (IAE). The pore size distribution curve is calculated
by fitting IAE isotherms to experimental isotherms.96,97

Gu and co-workers recognized the variation of surface area measurements of silica nanoparticles
by nitrogen adsorption.79,98,99 Careful attention revealed differences in specific surface area as
much as 300 m2/g depending on post-synthetic treatment. The most significant differences arise as
a result of washing and thermal drying procedures. A decrease in dissolved silica was observed after
successive washing with water, but not with ethanol. In fact, an increase in carbon was measured
when silica nanoparticles were washed with ethanol alone. This suggests that water removes
unhydrolyzed ethoxy groups remaining in the silica structure, whereas ethanol exacerbates this
process by introducing more ethanol which is then esterified onto the silica surface.79 Previous
work identified that maximum surface areas and porosity measurements were obtained after heating
at 300◦C, with both metrics rapidly decreasing upon higher temperatures.100 The greatest surface
area measured by Gu and co-workers was when water-washed silica nanoparticles were dried under
vacuum at 120◦C. Low surface area (< 50 m2/g) was measured after thermal treatment ranging
from 50 - 200◦C at atmospheric pressure.98

15

The contents of this work builds upon the discoveries and developments of innumerable scien-
tists, some of which have been described here. Insights from the RNAi community have informed
functionalization dextran-containing silica nanoparticles prepared via a modified Stöber method.
Established chemical and physical characterization techniques have been implemented to further
understand the silica structure. As previously stated, some porogens used in silica nanoparticle
synthesis are removed via rinsing or calcination. Dextran, the porogen used herein, has been
removed via calcination to enable characterization of Dex-SiO2-NP structures in its absence. Thus
many of the characterization methods discussed here have been applied to Dex-SiO2-NPs both
before and after calcination, with surface area and porosity analysis being the most indicative of
the vacancies remaining in the silica structures in the absence of dextran. These studies have been
embarked upon with the intention that these changes may provide insight to the location of dextran
and, further, its role in Dex-SiO2-NP synthesis.

16

CHAPTER 2

ESTABLISHING SIZE CONTROL OF DEXTRAN-SILICA NANOPARTICLES

2.1 Introduction

Comiskey synthesized amine-coated dextran-containing silica nanoparticles that were success-
ful in vitro siRNA delivery vehicles.11,12 The bimodal distribution of nanoparticles begs the
question of which group of nanoparticles are active as delivery vehicles (Figure 2.1). Aside
from their activity as delivery vehicles, several questions about the physical characteristics of
these nanoparticles remained, including 1) the origins of and difference between large and small
nanoparticles, 2) the location of dextran within nanoparticles, and 3) the reason for hollow-looking
nanoparticle centers.

Figure 2.1: TEM Micrograph and Size Distribution of Bimodal Amine-Coated Dex-SiO2-350
Nanoparticles. Brightness enhanced by 20% to emphasize detail, 27kx magnification, Scale bar
0.2 µm.

17

2.2 Synthetic Optimization of Dex-SiO2-NP

Interest in whether the small and large nanoparticles function differently as delivery vehicles
led us to first pursue separation of the two different Dex-SiO2-NP size populations such that the
silencing of each population could be evaluated independent of the other. Others have separated
nanomaterials using density gradient centrifugation.101,102 Comiskey explored density gradient
centrifugation as a method for separating the differently sized Dex-SiO2-NPs, though all fractions
contained bimodal nanoparticle distributions.? Size exclusion filtration was therefore pursued as a
means of separating Dex-SiO2-NPs.

Though NP density varies within the large and small Dex-SiO2-NP populations, average diam-
eter remains consistent. Size exclusion membranes were explored as a means of separating the 100
nm Dex-SiO2-NPs from the 350 nm Dex-SiO2-NPs. All Dex-SiO2-NPs are separated from the
reaction mixture using a 30kDa MWCO regenerated cellulose membrane in a pressure filtration unit
(Figure 2.2). Following the initial concentration of the reaction suspension, the product is rinsed
three times with approximately 200 mL of MilliQ water. This method is problematic in several
ways, mainly in that the 30kDa MWCO is too small to allow 100 nm particles to pass through, as a
30kDa protein would have a diameter of approximately 5 nm. Additionally, the nanoparticles build
up on the membrane, making it such that even if the pore size of the membrane would allow the
particles through, the aggregation at the membrane surface may prevent effective filtration. The
use of 0.2 µm ultrafiltration membranes is a promising means of separating the 100 and 350 nm
nanoparticles within our most effective delivery vehicle. However, a 10x size difference is necessary
for effective separation of materials with these types of membranes. Therefore it is very possible
that this method of separation will not be effective for our nanoparticles. Low pressure filtration
was attempted with low-concentration (1 or 0.1 mg/mL) nanoparticle suspensions, collecting the
filtrate and rinsing the filter with MilliQ water. A variety of filter membranes have been used to this
point: polycarbonate, polyvinylidene fluoride (PVDF), nylon, and cellulose acetate. Polycarbonate,
cellulose, PVDF, and nylon membranes, all with pore sizes of 0.2 µm, were used within a pressure
filtration reservoir. While the polycarbonate membrane’s pores are laser-etched and therefore more

18

Figure 2.2: Diagram of Pressure Filtration Apparatus.

accurately 0.2 µm in size, the pores of the cellulose, PVDF, and nylon membranes are formed by a
polymer web, resulting in more variation in the actual pore size (Figure 2.3). At all concentrations
attempted, both the filtrate and the retentate contained large and small nanoparticles. Therefore,
studies at lower concentrations were not pursued for the PVDF, nylon, and cellulose membranes.
At 1 mg/mL, the polycarbonate membrane was likely fouling, as no particles were evident in the
filtrate and both small and large particles were observed in the retentate. Decreasing to 0.1 mg/mL
or 0.01 mg/mL resulted in less fouling, as hypothesized. Nevertheless, there was still a large amount
of filter fouling at these low concentrations.

In order to minimize filter fouling, suspending the nanoparticles in 10% ethanol or ethanol
alone in order to form a more stable suspension was studied. The process was first run on a small
scale, with 10 mL aliquots of 0.1 and 0.01 mg/mL nanoparticle. While some particles were able
to pass through the polycarbonate membrane when filtration was performed in 10% ethanol, filter

19

Figure 2.3: TEM Micrographs of Size Exclusion Membranes. Polycarbonate membrane (top
left) and mixed cellulose esters membrane (top right), PVDF membrane (bottom left) and nylon
membrane (bottom right), Image ©Millipore Sigma.103

Figure 2.4: TEM Micrographs of Dex-SiO2-350 NPs Filtered Through 0.2 µm Polycarbonate
Membrane. 50 mL of 0.01 mg/mL of Dex-SiO2-350 NPs suspended in 10% ethanol underwent
pressure filtration using a 0.2 µm Polycarbonate Membrane, before (left) and after (right) filtration.
27kx magnification, scale bar 0.2 µm. Brightness enhanced by 20% to emphasize detail.

fouling was still a problem (Figure 2.4). Performing the filtration process in 100% ethanol was even
more effective. While nanoparticles with both 100 and 200 nm diameter were observed (Figure
2.5), there was notably less aggregation than the filtration with 10% ethanol.

Despite the low concentration (0.01 mg/mL) and the stable suspension of our nanoparticles
in ethanol, aggregation on the membrane is still present, evidenced by both large and small

20

Figure 2.5: TEM Micrographs of Dex-SiO2-350 NPs Filtered Through 0.2 µm Polycarbonate
Membrane. 50 mL of 0.01 mg/mL of Dex-SiO2-350 NPs suspended in 100% ethanol underwent
pressure filtration using a 0.2 µm Polycarbonate Membrane, before (left) and after (right) filtration.
27kx magnification, scale bar 0.2 µm. Brightness enhanced by 20% to emphasize detail.

nanoparticles observed in the retentate sample by TEM. The original purpose of separating the large
and small nanoparticles within our delivery vehicle population was to investigate whether either
size is a more effective delivery vehicle than the other. Only a small amount of nanoparticle was
recovered from each filtration, which is attributed to filter fouling, which withholds a significant
portion of nanoparticle from the already dilute suspension. Considering its low efficiency and
effectiveness, filtration was not further pursued as a separation method.

In addition to attempts to separate the 100 nm and 350 nm diameter Dex-SiO2-NPs in order to
obtain a monomodal size distribution, the starting materials were purified immediately before use.
Other than dextran, the two primary reagents, tetraethylorthosilicate (TEOS) and 3-aminopropyl tri-
ethoxysilane (APTES) are both alkyl silicates, which readily hydrolyze in the presence of moisture.
Partial or complete hydrolysis of ethoxy groups present in the reagents would leave the more active
silicic acids as impurities. In the base-catalyzed hydrolysis and condensation of silicates, silicic
acid would have a different reaction rate than TEOS, ultimately impacting nanoparticle size. TEOS
and APTES were purified by distillation and evaluated by 1H-NMR and the distributions of NPs
by TEM. Despite producing bimodal Dex-SiO2-NPs, any existing impurities are not evident in the
1H-NMR spectrum of TEOS (Figure 2.6). Distilled TEOS produces monodisperse Dex-SiO2-NPs

21

Figure 2.6: 1H-NMR Spectra of TEOS. Before (bottom) and after (top) purification by distillation.

consistent with new reagents, as evidenced by TEM (Figure 2.8). APTES was similarly dis-
tilled and evaluated by 1H-NMR. The 1H-NMR spectrum of impure APTES contains one apparent
impurity peak, a quartet centered at 3.6 ppm (Figure 2.9), which is not present after distillation.
13C-NMR revealed two impurity peaks, which indicate more clearly than the 13C-NMR that the
impurity is ethanol (Figure 2.10). Alkyl silicates readily hydrolyze, which explains the presence of
ethanol in the APTES reagent.

Improper storage and lack of purification of the silicate starting materials has been identified
as the origin of the bimodal distribution of Comiskey’s Dex-SiO2-NPs. Though the lone impurity
identified via NMR was ethanol, it likely results from hydrolysis of the silicate to silicic acid based
on understanding of the reactivity of the reagents.104 These results, as well as the distribution
of resulting Dex-SiO2-NPs, support the hypothesis that APTES and TEOS hydrolyzed into more

22

Figure 2.7: TEM Micrographs and Number Distributions of Dex-SiO2-350 NPs. Dex-SiO2-350
NPs prepared with impure (left) and distilled (right) TEOS. 27kx magnification, scale bar 0.2 µm.
Brightness enhanced by 20% to emphasize detail.

reactive silicic acids under improper storage conditions. This decomposition was mitigated in the
future by storing silicon alkoxide reagents under inert gas and distilled periodically in order to
maintain reagent purity and produce uniform Dex-SiO2-NPs.
2.3 Size Control in Modified Stöber Silica Nanoparticle Synthesis

With purified starting materials in hand, the dextran-modified Stöber method was further
developed. The classical Stöber method adjusts the concentrations of TEOS, ethanol, water, and
ammonium hydroxide as a means to adjust the rates of hydrolysis and condensation, in turn tuning
the NP diameter. With the addition of 10 kDa dextran to the synthesis solution, resulting NP
diameters vary from those expected based on the studies of Stöber and others.
Building upon the optimization of Comiskey’s work described above and the precedent established
by Stöber and others, reagents were adjusted to afford Dex-SiO2-NPs with decreased average

23

Figure 2.8: TEM Micrographs and Number Distributions of Dex-SiO2-250 NPs. Dex-SiO2-250
NPs prepared with distilled (left) and new (right) TEOS. 27kx magnification, scale bar 0.2 µm.
Brightness enhanced by 20% to emphasize detail.

Table 2.1: Nanoparticle Diameter as a Function of Reagent Concentrations.
NP

NH4OH (M) H2O (M) TEOS (M) Average NP Diameter (nm)

Dex-SiO2-100
Dex-SiO2-200
Dex-SiO2-240
Dex-SiO2-350

0.6
0.7
0.6
0.6

10.6
5.5
10.6
15.6

0.001
0.01
0.01
0.01

100
200
240
350

diameters (Figure 2.11, Table 2.1). Decreasing the concentration of H2O from 15.6 M to 10.6
M afforded a decrease in average NP diameter by approximately 100 nm. Further decreasing the
concentration of H2O to 5.5 M, along with an increase in NH4OH to 0.7 M produced Dex-SiO2-NPs
with an average diameter of 200 nm. Dex-SiO2-NPs with average diameters of approximately 100
nm were obtained by decreasing the concentration of TEOS from 0.01 M to 0.001 M, with variation
of other reagents only resulting in 200 nm NPs. Generally, decreasing the concentration of H2O
and increasing the concentration of NH4OH decreases the average diameter of the Dex-SiO2-NPs

24

Figure 2.9: 1H-NMR Spectra of APTES. Before (bottom) and after (top) purification by distillation.
Arrow indicates impurity peaks attributed to ethanol.

most closely related to Comiskey’s Dex-SiO2-NP. In addition to the observed decrease in diameter,
Dex-SiO2-NPs have characteristic hollow centers, which are further investigated in Chapter 3.

Beyond Comiskey’s Dex-SiO2-NP synthetic configuration, NH4OH and H2O were further
studied to probe the effects of the different reagents on NP diameter. The average diameters of
Dex-SiO2-NPs were compared according to the concentrations of NH4OH and H2O used in their
preparation. Some anomalies exist, though a general trend may be extracted. While a greater
range of reagent concentrations would provide more generalizable conclusions, the current study
indicates that Dex-SiO2-NP diameters increase with increasing NH4OH and H2O concentrations
(Figure 2.12, Figure 2.13). Greater concentrations of NH4OH result in the opposite trend, with the
greatest diameters resulting from the least amount of H2O.

25

Figure 2.10: 13C-NMR Spectra of APTES. Before (bottom) and after (top) purification by distilla-
tion. Arrows indicate impurity peaks attributed to ethanol.

Rates of hydrolysis and condensation of alkyl silicates are governed by the concentrations
of hydroxide and water within the reaction solution. The results observed here are consistent
with those of Stöber and others,18,105 NP diameter increases with increasing H2O and NH4OH
concentrations to a maximum, followed by a decrease in diameter at the highest concentrations.

Among the conditions studied, there was persistent difficulty obtaining monodispersed Dex-
SiO2-NPs with diameters less than 200 nm, which would be ideal for gaining entry to cells
for therapeutic delivery applications. Decreasing the concentration of TEOS was successful in
obtaining Dex-SiO2-NPs having an average diameter of approximately 100 nm (Figure 2.14).
Adjusting the concentration of TEOS within Dex-SiO2-250 by 1
8 (Figure 2.14 left and
left-center, respectively) afforded very similar monodisperse distributions of NPs. Despite the

4 and 1

26

Figure 2.11: TEM Micrographs and Number Distributions of Dex-SiO2-NPs. From left: Dex-
SiO2-100, Dex-SiO2-200, Dex-SiO2-250, Dex-SiO2-350. 27kx magnification, scale bar 0.2 µm.
Brightness enhanced by 20% to emphasize detail.

similarities in their distributions, these two NPs vary in their morphology, with Dex-SiO2-250-
4TEOS having a more spherical appearance and Dex-SiO2-250-1
1
8TEOS having a rougher surface.
Similarly, Dex-SiO2-350-1
8TEOS have similar average diameters but
different apparent surface morphologies (Figure 2.14 right-center and right, respectively).

4TEOS and Dex-SiO2-350-1

APTES is a key component in transforming Dex-SiO2-NPs into effective delivery vehicles,
as the amine moiety can electrostatically bind nucleic acids.
Integration of APTES into these
smaller NPs at a decreased concentration proportional to TEOS does not significantly alter the size
distribution or morphology of any of the studied conditions (Figure 2.15).

Decreasing the amount of TEOS and dextran in the reaction mixture did not produce significantly
smaller Dex-SiO2-NPs (Figure 2.16). Nevertheless, these NPs do appear to have a smoother shell,
which is to be expected as the amount of dextran and TEOS are scaled evenly from the original
amounts in both populations.

The Dex-SiO2-NPs discussed to this point were prepared with 9-11 kDa dextran and have unique

27

Figure 2.12: TEM Micrographs of Dex-SiO2-NPs upon Variation of Reagents. With 0.011 M
TEOS, in ethanol. 27kx magnification, scale bar 0.2 µm. Brightness enhanced by 20% to
emphasize detail.

28

Figure 2.13: Average Diameters of Dex-SiO2-NPs upon Variation of Reagents. With 0.011 M
TEOS, in ethanol.

morphology and have been shown to enhance silencing by siRNA in vitro.12 Further exploration
of size, morphology, and function as delivery vehicles was pursued via incorporation of different
dextrans and carbohydrates (Scheme 2.17). In each case, the alternative carbohydrate was used in
place of 9-11 kDa dextran as indicated in Table 2.2.

TEM micrographs of alternative carbohydrate NPs prepared according to Dex-SiO2-250 con-
ditions are shown in Figure 2.18. NPs prepared using 6kDa dextran and 40 kDa dextran closely
resemble those prepared using 9-11 kDa dextran in size and appearance, as both populations con-
tain hollow centers. NPs prepared using α-cyclodextrin and β-cyclodextrin (CDex-SiO2-NPs) are
more bimodal in diameter compared to the 6 kDa, 9-11 kDa, and 40 kDa NPs. Additionally,
CDex-SiO2-NPs do not appear to have a hollow center. While dextrans and cyclodextrins share

29

Figure 2.14: TEM Micrographs and Number Distributions of Dex-SiO2-NPs. From left: Dex-
4TEOS, Dex-SiO2-350-1
SiO2-250-1
8TEOS. 27kx
magnification, scale bar 0.2 µm. Brightness enhanced by 20% to emphasize detail.

4TEOS, Dex-SiO2-250-1

8TEOS, Dex-SiO2-350-1

Figure 2.15: TEM Micrographs and Number Distributions of Dex-Amine-SiO2-NPs. From left:
4TEOS, Dex-Amine-SiO2-250-1
Dex-Amine-SiO2-250-1
4TEOS,
Dex-Amine-SiO2-350-1
8TEOS. 27kx magnification, scale bar 0.2 µm. Brightness enhanced by
20% to emphasize detail.

8TEOS, Dex-Amine-SiO2-350-1

30

Figure 2.16: TEM Micrographs and Number Distributions of Dex-SiO2-NPs. From left: 1
8Dex-SiO2-250-1
SiO2-250-1
enhanced by 20% to emphasize detail.

4Dex-
8TEOS. 27kx magnification, scale bar 0.2 µm. Brightness

4TEOS, 1

Figure 2.17: Structures of Dextran and Cyclodextrins. Dextrans of varying molecular weight and
α-cyclodextrin and β-cyclodextrin.

31

Table 2.2: Dex-SiO2-NP Diameters with Dextran Variation.

Dex-SiO2-NP

Synthesis Conditions

Hydrodynamic Diameter
of Dextran in MQ H2O

Weighted Average
NP Diameter (nm)

Dex-SiO2-200
Dex-SiO2-250
Dex-SiO2-350
Dex-SiO2-200
Dex-SiO2-250
Dex-SiO2-350
Dex-SiO2-200
Dex-SiO2-250
Dex-SiO2-350
Dex-SiO2-200
Dex-SiO2-250
Dex-SiO2-350
Dex-SiO2-200
Dex-SiO2-250
Dex-SiO2-350

146
194
318
211
245
276
187
242
312
168
311
63
259
391
63

Carbohydrate
6 kDa Dextran
6 kDa Dextran
6 kDa Dextran
10 kDa Dextran
10 kDa Dextran
10 kDa Dextran
40 kDa Dextran
40 kDa Dextran
40 kDa Dextran
α-Cyclodextrin
α-Cyclodextrin
α-Cyclodextrin
β-Cyclodextrin
β-Cyclodextrin
β-Cyclodextrin

3
3
3
6
6
6
7
7
7
-
-
-
-
-
-

the glucose monomer, the greater structure varies from a branched polymer to a wreath of discrete
glucose units which could certainly impact NP assembly.
2.4 Growth of Dex-SiO2-NPs

Pursuant to understanding the formation of Dex-SiO2-NPs, samples of reaction solution were
extracted at designated points in the synthesis process, purified, and studied by TEM. The synthetic
procedure employed here, as described in Chapter 6, first dissolves dextran in MilliQ water. After
stirring ethanol, NH4OH and MilliQ water, dextran is added and allowed to stir for 10 minutes.
TEOS is then added, and the reaction solution is stirred 10 minutes before the addition of APTES.
Samples were extracted using a syringe, diluted, and centrifuged. After removal of the supernatant
liquid, the remaining solids are rinsed three times with MilliQ water before suspension in ethanol
for deposition on a TEM grid. Key points in this process at which samples were extracted include
soon after addition of TEOS, immediately before APTES addition, and after APTES addition.
Figure 2.19 includes several TEM micrographs from this process.

After TEOS addition, dark, amorphous aggregates are evident, as well as a strand-like network

32

Figure 2.18: TEM Micrographs of Dex-SiO2-NPs Prepared with Different Dextrans. Top, from
left: 6 kDa, 9-11 kDa, 40 kDa. Bottom, from left: α-cyclodextrin, β-cyclodextrin prepared using
Dex-SiO2-250 conditions. 27kx magnification, scale bar 0.2 µm. Brightness enhanced by 20% to
emphasize detail.

of a lighter shade. The TEM micrograph collected after several more minutes of stirring TEOS in
the dextran-containing reaction solution shows more spherical aggregates, though smaller, lighter
objects are still evident. With APTES addition (Figure 2.19, bottom), spherical nanoparticles
quickly took shape. There is no observable change in the observed NPs between 4 minutes and
8 minutes post APTES addition. 20 hours after APTES addition a bimodal size distribution is
observed. Extraction of more samples within that time frame may have provided information per-
taining to the origins of the bimodal distribution. Further, it should be noted that these experiments
were conducted prior to Dex-SiO2-NP synthetic optimization, resulting in the bimodal distribution
of nanoparticles.

Curious about the process in the absence of APTES, the same experiment was conducted
without the addition of APTES (Figure 2.20). In this case, after preparation of the grid for TEM
analysis, a solution of 5% periodic acid was applied to the grid in order to oxidize dextran, followed
by a silver methanamine stain which would stain the dextran by precipitating silver. While the

33

staining process could provide valuable insight into the location of dextran, in practice it raised
questions as to which of the dark crystalline regions result from stained dextran and which result
from SiO2-NP or Dex-SiO2-NP precipitation.

Aggregates similar to those apparent in Figure 2.19 were evident at 4, 8, and 20 minutes
after TEOS addition (Figure 2.20). Spherical nanoparticle structures were evident in the TEM
micrographs beginning at 30 minutes post TEOS addition. One hour after TEOS addition, both
spherical nanoparticles and aggregates were evident. These observations are consistent with the
nucleation and aggregation mechanism for SiO2-NP assembly discussed by Han and others.20
Closer examination of Dex-SiO2-NPs 60 minutes after the addition of TEOS evidences abundant
granules on the surface of the two nanoparticles shown in the micrograph (Figure 2.21). Upon
cursory analysis, one may speculate that this suggests that dextran is accessible from the surface both
by periodic acid and silver methanamine, which would certainly an exciting prospect. Nevertheless,
the nature of this experiment is such that one cannot eliminate the possibility that these granules

Figure 2.19: TEM Micrographs of Purified Reaction Mixture during Dex-SiO2-350 NP Synthesis.
Top, from left: 5.45 minutes post TEOS, 8.3 minutes post TEOS. Bottom, from left: 4 minutes post
APTES, 8.3 minutes post APTES, 20 hours post APTES. Scale bar 500 nm. Brightness enhanced
by 20% to emphasize detail.

34

Figure 2.20: TEM Micrographs of Purified Reaction Mixture during Dex-SiO2-350 NP Synthesis.
Top, from left: 4 minutes post TEOS, 8 minutes post TEOS. Bottom, from left: 20 minutes post
TEOS, 30 minutes post TEOS, 60 minutes post TEOS. Stained using silver methanamine to detect
dextran after NPs were deposited onto TEM grid. Scale bar 500 nm. Brightness enhanced by 20%
to emphasize detail.

originate from excess dextran in the reaction solution having adhered to the nanoparticle surface
whilst the sample upon the TEM grid dried.

Though CDex-SiO2-NPs were not studied in this way, one could hypothesize that the structural
differences between branched dextran and cyclodextrin would influence the formation of Dex-SiO2-
NPs compared to CDex-SiO2-NPs. Additionally, while these microscopic observations provide
insight to the process of Dex-SiO2-NP formation, it is important to acknowledge the effect of
purification procedures and sample preparation, including drying, may have on the results.

A new hypothesis of Dex-SiO2-NP formation based on dextran solubility emerged. Polymer
solubility is known to be a complex relationship between polymers and non-polymeric solvents. In
simplest terms, the θ solvent is one in which a polymer behaves as an ideal random coil.106 The
process of Dex-SiO2-NP synthesis is described in Figure 2.22. 9-11 kDa dextran dissolved in water
is added to the reaction solution of ethanol, water, and ammonium hydroxide while stirring.

35

Figure 2.21: TEM Micrograph of Dex-SiO2-350 NPs During Synthesis. 60 minutes post TEOS.
Stained using silver methanamine to detect dextran after NPs were deposited onto TEM grid. Scale
bar 100 nm. Brightness enhanced by 20% to emphasize detail.

It is possible that this solution is not ideal for dextran solubility, which may result in aggregates of
dextran existing within aqueous solvation spheres (Figure 2.22, top center). Upon TEOS addiiton
to this solution, condensation around these localized water-rich aggregates could result in the

Figure 2.22: Diagram Depicting Dex-SiO2-NP Synthesis and Dextran Solubility Hypothesis.

36

hollow center morphology observed. Alternatively, in the presence of more "good solvent" (i.e. a
higher concentration of H2O, Dex-SiO2-350 compared to Dex-SiO2-200) dextran will dissolve to
a greater extent, possibly existing as a random coil and resulting in greater dispersion throughout
NPs.107 Alternatively, it’s possible that the increased concentration will expand the solvation
spheres resulting in larger aggregates, and larger Dex-SiO2-NPs (Figure 2.22).

This hypothesis was probed by measuring the hydrodynamic radii of aggregates when dextran
dissolved in MilliQ water is mixed in reaction solutions containing methanol, ethanol, and iso-
propanol as a function of time (Figure 2.23). Presently, water is added to dextran and the solution
is manually agitated and given the opportunity to dissolve over a period of ten minutes prior to
addition to the reaction mixture. Greater precautions could be taken to ensure that dextran is
well-dissolved in water before addition to the reaction solution. The greater polarity and decreased
steric effects of methanol compared to ethanol and isopropanol increase its ability to solvate water.
Since dextran is water-soluble, the increased solubility of water in methanol is expected to enlarge
the solvation sphere. The aggregates in methanol observed by DLS are expected be larger than
those of the other two solvents. This hypothesis was validated by DLS results, which evidenced
that aggregates arising from the solution containing methanol and dextran were the largest, with
the isopropanol solution giving rise to the smallest aggregates.

In solutions where dextran and TEOS were added, which most resemble Dex-SiO2-NP synthetic
conditions, methanol aggregates were the largest, followed by isopropanol and then ethanol, though
the last two were similar. In conditions that excluded dextran (Figure 2.23, bottom), methanol again
produced the largest aggregates, while isopropanol and ethanol solutions were very similar. There
was no significant change in aggregate size over the course of 140 minutes.

Dex-SiO2-NPs and SiO2-NPs resulting from these experiments are shown in Figures 2.24 and
2.25, respectively. From the starting point of Dex-SiO2-NPs synthesized in ethanol solution, both
methanol and isopropanol-produced NPs exhibit a decrease in diameter. The SiO2-NPs produced in
the absence of dextran produced monomodal size distributions. The distributions of NPs produced
in methanol and ethanol were approximately the same size, though the breadth of distributions

37

Figure 2.23: Hydrodynamic Diameter of Aggregates During Dex-SiO2-NP Synthesis. Solution of
Dex-SiO2-250 conditions with methanol (left), ethanol (center), and isopropanol (right), containing
9-11 kDa dextran (top), dextran and TEOS (middle), and TEOS (bottom).

varied among the larger and smaller populations. NPs produced in the isopropanol solution were
larger than those in ethanol or methanol.

DLS results indicate that time does not play a significant role in aggregation in the process of NP
synthesis. Based on these results, correlation of dextran aggregate size and resulting NP diameter
is only observed in cases using ethanol. The different morphologies observed in Dex-SiO2-NPs
synthesized in methanol and isopropanol suggests that dextran may be incorporated differently.
Further studies varying polarity and sterics via different solvents may provide further insight into
dextran aggregation and NP formation.

38

Figure 2.24: TEM Micrographs and Number Distributions of Dex-SiO2-NPs Prepared Using
Methanol, Ethanol, and Isopropanol. 67kx Magnification, scale bar 100 nm. Brightness enhanced
by 20% to emphasize detail.

Figure 2.25: TEM Micrographs and Number Distributions of SiO2-NPs Prepared Using Methanol,
Ethanol, and Isopropanol. 67kx Magnification, scale bar 100 nm. Brightness enhanced by 20% to
emphasize detail.

39

2.5 Conclusions

Though effective delivery vehicles for silencing, the bimodal size distribution of Comiskey’s
Dex-SiO2-NPs hindered more thorough understanding of delivery vehicle construction and activity.
Though size exclusion membranes were ineffective at separating the two populations of Comiskey’s
Dex-SiO2-NPs, purification of the alkyl silicate starting materials yielded monodispersed size
distributions. Building on the work of Stöber and others, concentrations of ammonium hydroxide,
water, and TEOS were varied to produce Dex-SiO2-NPs of varying diameter. The effects of
different dextrans and cyclodextrins in NP synthesis were also examined. TEM was used to
investigate Dex-SiO2-NP formation at different time points before and after the addition of TEOS
and APTES, revealing amorphous aggregates that grow into spherical nanoparticles. Lastly, DLS
and TEM were used to probe the presence and behavior of aggregates in Dex-SiO2-NP and SiO2-
NP reaction solutions. Methanol and isopropanol solutions produced Dex-SiO2-NPs smaller than
those prepared in ethanol, though further study is necessary to draw conclusions about the role of
dextran in the reaction solution and, ultimately, the nanoparticle structure.

40

CHAPTER 3

UNRAVELING DEXTRAN’S ROLE IN DEX-SIO2-NP MORPHOLOGY

3.1 Introduction

Dextran-containing silica nanoparticles (Dex-SiO2-NPs) have shown promise as delivery vehi-
cles for RNAi therapeutics. This, along with their unique appearance by TEM, prompted further
study and characterization. This study was approached with the aim of determining the location of
dextran in the NPs, which could ultimately result in understanding dextran’s role in NP synthesis.
The complexity of Dex-SiO2-NPs presents several characterization challenges. Therefore, a
combination of chemical and physical characterization techniques are used herein. As insoluble
structures, solution NMR of Dex-SiO2-NPs is ineffective. CP-MAS 13C-NMR, a method for
measuring the magnetic resonances of materials in the solid state, is used herein to determine the
organic components present in Dex-SiO2-NPs. Thermogravimetric analysis is used to determine
the water and organic content of NPs. Dextran was removed from Dex-SiO2-NPs by Soxhlet
extraction and calcination, followed by porosity and surface area analysis before and after dextran
removal from NPs.
3.2 Characterization of Dex-SiO2-NP Morphology

When compared to SiO2-NPs prepared in the absence of dextran, the apparent core-shell
structure exhibited by Dex-SiO2-NP in TEM images (Figure 3.1) prompted further study. While
the outer shells of Dex-SiO2-NPs appear smooth like the nanoparticles prepared without dextran,
the interior is consistently less dense.

Upon examination at higher magnification, several differences among the Dex-SiO2-NPs are
evident apart from size. The core and shell regions of Dex-SiO2-200 (Figure 3.2, left) appear
distinct from one another. Dex-SiO2-200 NPs also have a smooth center region, in contrast with
Dex-SiO2-250 and Dex-SiO2-350 NPs. A gradual transition from the outer shell to the NP center,

41

Figure 3.1: TEM Micrograph of SiO2-NPs and Dex-SiO2-NPs. SiO2-NPs (left) and Dex-SiO2-
250 NPs (right), exhibiting less dense centers. 27kx magnification, Scale bar 0.2 µm. Brightness
enhanced by 20% to emphasize detail.

along with fine, grainy center details set Dex-SiO2-350 NPs apart (Figure 3.2, right). Dex-SiO2-
250 NPs fall between Dex-SiO2-200 and Dex-SiO2-350 NPs in both size and appearance. Distinct
center and shell regions resemble Dex-SiO2-200 NPs, while the grainy center (Figure 3.2, center)
is akin to that of Dex-SiO2-350 NPs.

Having spatially defined the area via scanning transmission electron microscopy, a series of
successive EDS spectra build the profile observed in Figure 3.3. The sample area of a Dex-SiO2-
350 nanoparticle is shown on the left side of Figure 3.3, with the white line indicating the area
scanned. The elemental profile (Figure 3.3, right) of silicon and oxygen were examined along this
line. The y-axis presents the counts of x-rays of the specified energy detected, and is therefore
indicative of elemental abundance. The low electron density of carbon (Kα = 0.277 compared to
1.739 (Si) and 0.525 (O)), as well as the use of microscopy grids containing carbon films does not
allow for reliable measurement of the carbon content within the sample. Another limitation of the
EDS method used here is its qualitative nature. Nevertheless, the results obtained by EDS confirm
that Dex-SiO2-NPs decrease in density of SiO2 toward the center.

The sizes of the central cavities visible by TEM in Dex-SiO2-NPs range from 10 - 100 nm
(Figure 3.2), and is dependent on the type of Dex-SiO2-NP being analyzed. Dex-SiO2-350 NPs,

42

Figure 3.2: TEM Micrographs of Dex-SiO2-NPs. Dex-SiO2-200 (left), Dex-SiO2-250 (center),
and Dex-SiO2-350 (right) at 27kx magnification, scale bar 0.2 µm (top) and 67kx magnification,
scale bar 100 nm (bottom). Brightness enhanced by 20% to emphasize detail.

though larger than the 200 and 250 series, generally has smaller cavities. Since the cavernous
centers are the most distinguishable features of Dex-SiO2-NPs and SiO2-NPs, it was hypothesized
that the central cavity may be composed of dextran. Diameters of the hollow centers visible by
TEM were measured and compared with the hydrodynamic diameter of dextran dissolved in water
as measured by DLS. Three different size domains were evident by DLS (Figure 3.4). The most
narrow peak was centered at 10 nm, followed by a less intense, broader peak at 200 nm, and the least
intense peak at 5000 nm. The radius of gyration of a 9-11 kDa unit of dextran has been determined
to exist between 10 - 15 nm,108–110 larger than the diameter of a 10kDa protein (approximately
3 nm).111 The less precise folding of dextran, as well as its branched structure may contribute to
its greater hydrodynamic diameter. The DLS results represent different dextran aggregates which

43

Figure 3.3: STEM Micrographs and EDS Line Scan of Dex-SiO2 and SiO2 Nanoparticles. S-TEM
micrographs of Dex-SiO2-NPs (left) and SiO2-NPs (right), with area scanned for EDS analysis
indicated by fuschia line (top), and silicon and oxygen elemental abundance profile along indicated
area (bottom). Scale bar 300 nm. Y-axis indicates abundance as measured by X-ray counts.

are solubilized in water to differing degrees and ranging in diameter. Dissolving dextran in water
and then stirring in ethanol, simulating synthesis conditions, revealed aggregates with diameters
consistent with those observed within Dex-SiO2-NP cavities (Figure 3.5). These results indicate
that dextran aggregates may be templates around which the SiO2 shell forms.

Microscopy and EDS have elucidated the SiO2 component of the various Dex-SiO2-NPs, as
well as structural variations from SiO2-NPs synthesized without dextran.
It can be concluded
from these results that the morphological differences between Dex-SiO2-NPs and SiO2-NPs result
from the use of dextran during synthesis, though its presence in the final NP structure cannot be
established from the microscopy or EDS results shown here.

44

Figure 3.4: Hydrodynamic Diameter of Dextran in Water. Distribution of Hydrodynamic Diam-
eters of 9-11 kDa dextran dissolved in MilliQ Water at 0.01 g/mL and 0.001 g/mL obtained using
Dynamic Light Scattering.

Figure 3.5: Hydrodynamic Diameter of Dextran Aggregates in Ethanol. Distribution of Hydrody-
namic Diameters of at 0.01 g/mL 9-11 kDa dextran dissolved in MilliQ water and stirred in ethanol
for the indicated time obtained using Dynamic Light Scattering.

45

Figure 3.6: TGA Profile of Dex-SiO2-NPs. TGA of Dex-SiO2-200 (left), Dex-SiO2-250 (center),
and Dex-SiO2-350 (right).

Table 3.1: Percent Weight Lost of Dex-SiO2-NPs, SiO2-NPs, and Dextran by Thermogravimetric
Analysis.

25 - 100 °C 100-800 °C 25 - 800 °C

NPs

Dex-SiO2-200
Dex-SiO2-250
Dex-SiO2-350

SiO2-NPs
Dextran

7.09
5.95
3.94
9.30
8.70

15.98
16.40
12.59
8.30
89.06

23.08
22.34
16.53
17.6
97.76

3.3 Verifying Incorporation of Dextran into Dex-SiO2-NPs

The high temperatures used in the thermogravimetric analysis of Dex-SiO2-NPs allowed for
removal and determination of percent incorporation of dextran within NPs. An initial loss from 25
°C to 100 °C was observed for all Dex-SiO2-NPs, due to loss of water from within the nanoparticles
(Figure 3.6, Table 3.1). Dex-SiO2-200 NPs lost 7% of its weight to water, while Dex-SiO2-250
NPs lost approximately 6%, and Dex-SiO2-350 NPs lost 4%. Silica nanoparticles are known to be
hygroscopic, therefore the initial loss of water is not surprising. SiO2 NPs lost 9.3% of mass due
to water, while dextran itself contains 8.7% of water by weight (Figure 3.7, Table 3.1).

The second loss begins at 300 °C and plateaus at 600 °C. Consistent with observations of
pure dextran (Figure 3.7), this loss represents the loss of dextran from NPs. Accounting for the
loss of water, Dex-SiO2-200 NPs lost 16% due to dextran. Dex-SiO2-250 NPs lost 16%, and
Dex-SiO2-350 NPs lost approximately 13%. The TGA of Dex-SiO2-200 NPs exhibits a stepwise
loss of weight at 500 °C, also observed in the TGA of dextran alone (Figure 3.7), and attributed

46

Figure 3.7: TGA Profile of Dextran and SiO2 NPs. TGA of 9-11 kDa Dextran and SiO2 NPs
synthesized without dextran.

to complex decomposition processes of dextran. If all dextran included in the reaction mixture is
incorporated into nanoparticles, it would be 24% of the weight of the nanoparticles. Considering
the moisture present in dextran and silica, loss of 16 - 23% of NP weight by TGA indicates that
nearly all dextran included in synthesis is incorporated into nanoparticles. After loss of water,
SiO2 NPs lost only 83% of mass, which is attributed to residual ethanol. Dextran itself loses
nearly all of it’s remaining mass, an additional 89% after loss of water (Figure 3.7, Table 3.1).
Interestingly, NPs with different diameters, produced by variation of ethanol and water, exhibit
variable dextran incorporation. There is also a correlation between decreased water content with
increasing nanoparticle diameter. This may be due to the difference in dextran content across the
different sizes of nanoparticles. This system uses lyophilized dextran dissolved separately in water,
therefore it is reasonable to conclude that the correlation is due to different dextran content.

The hollow appearance of Dex-SiO2-NPs varies significantly from the more solid-looking
products of the analogous reaction produced in the absence of dextran (Figure 3.1). These ob-

47

Figure 3.8: TEM Micrographs of Dex-SiO2-200 Nanoparticles after TGA. Dex-SiO2-200 before
(left) and after (right) TGA. Decreased density in the nanoparticle shell post-TGA suggests frac-
turing of SiO2 surface with thermal removal of dextran. 67kx magnification, scale bar 0.1 µm.
Brightness enhanced by 20% to emphasize detail.

servations are consistent with other reports in the literature of hollow SiO2-NPs prepared using
templates.73,112 In order to confirm the presence of dextran in Dex-SiO2-NPs, solid state CP-MAS
13C-NMR was performed. The resulting NMR spectra shows a singlet at 97.8 ppm corresponding
to C1 of dextran, and a broad singlet centered at 71.5 ppm corresponds to C2-C6 (Figure 3.9).
These resonances are consistent with the literature and solution NMR of dextran, confirming that
dextran is incorporated in the Dex-SiO2-NPs. The low intensity resonances at 30 and 40 ppm result
from the presence of adamantane, used as a reference, while the resonance at 8 ppm ethoxy groups
originating from the TEOS starting material and/or the ethanol used as solvent.

Solid state NMR results verify the presence of dextran in the Dex-SiO2-NPs. TEM and EDS
line scan results show monodispersed Dex-SiO2-NPs bearing a dense shell and less dense core.
Nevertheless, neither method provides direct evidence of the location of dextran on the surface
or within the Dex-SiO2-NPs. Attempts to identify dextran’s location within Dex-SiO2-NPs via
chemical derivatization were unsuccessful. The inability to quantify dextran through chemical
derivatization may provide an indication of the accessibility of dextran from the NP surface.

48

Figure 3.9: Solid State CP-MAS 13C-NMR of Dex-SiO2-350 Nanoparticles. Dextran appears in
solid state CP-MAS 13C-NMR of Dex-SiO2-NPs, confirming its presence in nanoparticles after
synthesis and purification.

3.4 Removal of Dextran and Post-Mortem Porosity Analysis

While TEM, EDS, TGA, and CP-MAS 13C-NMR have provided valuable information con-
firming the amount and presence of dextran, as well as variation of silica content across NPs,
the location of dextran within Dex-SiO2-NPs has not been informed by these means. The present
approach removes dextran from Dex-SiO2-NPs, which are then subjected to surface area and poros-
ity analysis by gas adsorption. This so-called "post-mortem" analysis, when compared with gas
adsorption results from intact Dex-SiO2-NPs, can reveal the volume vacated by dextran, as well as
its location within the NPs.

Soxhlet extraction was used to remove dextran from Dex-SiO2-NPs using water, followed by
characterization by TEM for observation of structural changes, while TGA reveals how much
dextran is removed by the extraction process. Soxhlet extraction was performed using water as

49

Figure 3.10: 500 MHz 1H-NMR of Dextran. 10 kDa Dextran in D2O (top), and recovered from
Dex-SiO2-NPs by Soxhlet extraction in D2O/DMSO-d6 (bottom).

a solvent to enhance the solubility of dextran. Nanoparticles were recovered from the cellulose
thimble and analyzed using TEM and TGA. The filtrate was lyophilized to obtain a white solid,
confirmed as dextran via 1H-NMR (Figure 3.10).

The electron density of carbon is too low to be detected by TEM, thus it is interesting that
structural changes arising from Soxhlet extraction are observed by TEM (Figure 3.11, compared
to Figure 3.2), indicating that the silica structure itself is compromised. Overall, the surfaces of
Dex-SiO2-NPs become more rough after Soxhlet extraction and the centers appear more hollowed.
Dex-SiO2-200 NPs have smaller amorphous, granular pieces within the NP population, which may
be pieces of silica that were removed from the NPs along with dextran. Dex-SiO2-250 NPs bear a
significantly less dense center, a more grainy appearance, and a less gradual progression from the

50

Figure 3.11: TEM Micrographs of Dex-SiO2-NPs After Soxhlet Extraction. Dex-SiO2-200 (left),
Dex-SiO2-250 (center), and Dex-SiO2-350 (right) at 27kx magnification, scale bar 0.2 µm (top)
and 67kx magnification, scale bar 100 nm (bottom). Brightness enhanced by 20% to emphasize
detail.

center to the core. The most apparent change to Dex-SiO2-350 is the more grainy center and rough
exterior shell.

Another noticeable feature of the Dex-SiO2-NPs following Soxhlet extraction of dextran is the
variation in appearance of different nanoparticles within one sample. While many appear more
hollow, other nanoparticles retain a solid appearance, with little transmission. This may be attributed
to varying rates of condensation and hydrolysis during nanoparticle synthesis, potentially resulting
in non-uniform incorporation of dextran. VanBlaaderen and co-workers noted that decreased rates
of hydrolysis results in persistence of ethoxy groups within SiO2-NPs.? Since the ethoxy group is
larger than a hydroxyl group, a less dense silica network arises from such conditions. The overall
decrease in density at the center of the particles after extraction may be explained by small amounts

51

Figure 3.12: TGA Profile of Dex-SiO2-NPs Before and After Soxhlet Extraction. Dex-SiO2-200
(left), Dex-SiO2-250 (center), and Dex-SiO2-350 (right)

Table 3.2: Percent Weight Lost of Dex-SiO2-NPs after Soxhlet Extraction by Thermogravimetric
Analysis.

NPs

Dex-SiO2-200
Dex-SiO2-250
Dex-SiO2-350

7.86
5.23
4.83

25 - 100 °C 100-800 °C 25 - 800 °C

15.07
15.09
11.74

22.93
21.13
16.57

of silica that are fused to dextran being removed during the extraction.

TGA revealed a minimal loss of dextran by Soxhlet extraction (Figure 3.12, Table 3.2). The
different Dex-SiO2-NPs exhibited decreasing water content with increasing molecular weight,
consistent with data collected on NPs before extraction (Table 3.1). After Soxhlet extraction, the
amount of dextran extracted from Dex-SiO2-200 and Dex-SiO2-250 were both approximately 15%,
while Dex-SiO2-350 NPs lost 11.74%. Compared with the amount of dextran initially incorporated
into Dex-SiO2-NPs, the amount of dextran extracted by Soxhlet ranged from 0.85% - 1.31%.
Incomplete removal of dextran by Soxhlet extraction may be due to its intimate incorporation
within the molecular structure of SiO2 nanoparticles. Considering the low efficiency of dextran
extraction, the change in appearance of Dex-SiO2-NPs after Soxhlet extraction is more likely due
to repeated rinsing of water eroding the silica structure than due to removal of dextran.

The low dextran removal efficiency of Soxhlet extraction led to exploration of different methods
of dextran removal. Since removal of organic components from ceramic structures has been
accomplished via calcination at temperatures ranging from 500°C - 700°C, we proceeded with
calcination as the means of removing dextran.113 TEM and TGA were used to determine the

52

Figure 3.13: TGA Profile of Dex-SiO2-NPS Before and After Calcination. Dex-SiO2-200 before
(left) and after (right) TGA.

Table 3.3: Percent Weight Lost of Dex-SiO2-NPs after Calcination by Thermogravimetric Analysis.

NPs

25 - 100 °C 100-800 °C 25 - 800 °C

Dex-SiO2-200
Dex-SiO2-250
Dex-SiO2-350

1.80
3.06
1.88

1.06
1.43
0.66

2.86
4.50
2.54

efficacy of dextran removal and assess the physical characteristics of Dex-SiO2-NPs. TGA of Dex-
SiO2-NPs after calcination at 350°C over four days evidenced a minor loss of water, and small loss
of organic content (<2%) indicating that the majority of dextran was removed during the heating
process (Figure 3.13, Table 3.3). Microscopy revealed that, unlike Soxhlet extraction, calcination
did not compromise the structure of Dex-SiO2-NPs (Figure 3.14). No significant erosion of the
silica surface was evident, and the size distributions did not vary as a result of calcination.

With successful dextran removal by calcination, nitrogen adsorption-desorption analysis was
performed to analyze changes to the surface area and porosity of Dex-SiO2-NPs arising in the
absence dextran. The obtained Dex-SiO2-NPs isotherms are shown in Figure 3.15. The three
isotherms in the top row represent Dex-SiO2-200, Dex-SiO2-250, and Dex-SiO2-350 before calci-
nation. The presence of a plateau classifies these as Type I, according to IUPAC classification.91
Type I isotherms indicate early completion of monolayer adsorption and microporosity. The in-
crease in adsorption nearing a partial pressure of one bears similartities to the macroporous region
of the Type II isotherm. The presence of characteristics suggests that both micro- and macropores
exist within Dex-SiO2-NPs. Type II character is present in all isotherms obtained, and may rep-

53

Figure 3.14: TEM Micrographs of Dex-SiO2-NPs After Calcination. TGA of Dex-SiO2-200 NPs
(left), Dex-SiO2-250 NPs (center), and Dex-SiO2-350 NPs (right) at 27kx magnification, scale bar
0.2 µm (top) and 67kx magnification, scale bar 100 nm (bottom). Brightness enhanced by 20% to
emphasize detail.

resent the hollow centers within Dex-SiO2-NPs and/or the interstitial spaces between NPs in the
sample. After calcination, the isotherm shapes vary more significantly among the different types
of Dex-SiO2-NPs. Calcined Dex-SiO2-350 NPs exhibit a Type IV isotherm, which is typical of
mesoporous materials. The isotherm shapes of calcined Dex-SiO2-200 and Dex-SiO2-250 NPs do
not differ much from those of the non-calcined NPs.

The significant hysteresis loop observed after calcination of Dex-SiO2-350 NPs resembles
H2(a) the most, indicating complex pore shapes. The steep desorption branch characteristic
of H2(a) hysteresis loops are attributed to pore blocking in pore necks or cavitation-induced
evaporation.91 The hysteresis loops of Dex-SiO2-200 and Dex-SiO2-250 are more pronounced in
isotherms obtained before calcination than after (Figure 3.15 ). Unlike Dex-SiO2-350, the hysteresis

54

Figure 3.15: Nitrogen Adsorption-Desorption Isotherms of Dex-SiO2-NPs. Adsorption (top)
and desorption (bottom) isotherms of Dex-SiO2-200 NPs (left), Dex-SiO2-250 NPs (center), and
Dex-SiO2-350 NPs (right).

of the smaller NPs is more consistent with type H4. Signatures of H4 hysteresis loops include the
Type I/II character of the adsorption branch and the lower limit of the desorption branch existing at
the partial pressure where cavitation occurs.

Insights gained from isotherm shapes are illustrated by the differential pore volume distributions
obtained via DFT (Figure 3.16, Figure 3.17, and Figure 3.18). Non-calcined NPs contain more
micropores (< 2 nm) than mesopores (2 - 50 nm). Following calcination, a decrease in micropores is
accompanied by an increase in mesopores (Table 3.4). It can be inferred from these results that the
removal of dextran creates larger pores within the Dex-SiO2-NPs and results in the loss of smaller
pores. These observations are consistent with the behavior of organic templates and/or porogens,
which are widely used to customize pore size and shape within materials, including SiO2 NPs.114
The presence of micropores and some mesopores before dextran removal suggests that, though its
removal from the NPs results in larger pore formation, even with dextran intact Dex-SiO2-NPs are
porous.

Before calcination, the porosity of Dex-SiO2-200 and Dex-SiO2-250 NPs are consistent with

55

Figure 3.16: Differential Pore Volume of Dex-SiO2-200 NPs Before and After Calcination. Dif-
ferential Pore Volume of Dex-SiO2-200 NPs before and after calcination.

Dex-SiO2-350. Post-calcination, the microporosity of Dex-SiO2-200 and Dex-SiO2-350 decrease
along with an increase in mesoporosity (Figure 3.16 and Figure 3.18), while Dex-SiO2-250 loses
microporosity but does not gain significant mesoporosity (Figure 3.17).

56

Figure 3.17: Differential Pore Volume of Dex-SiO2-250 NPs Before and After Calcination. Dif-
ferential Pore Volume of Dex-SiO2-250 NPs before and after calcination.

Figure 3.18: Differential Pore Volume of Dex-SiO2-350 NPs Before and After Calcination. Dif-
ferential Pore Volume of Dex-SiO2-350 NPs before and after calcination.

57

Table 3.4: Summary of Surface Area and Porosity of Dex-SiO2-NPs Before and After Calcination.

NPs

Dex-SiO2-200
Dex-SiO2-200
Dex-SiO2-250
Dex-SiO2-250
Dex-SiO2-350
Dex-SiO2-350

Heat
Treatment
120 °C, 2h
300 °C, 96h
120 °C, 2h
300 °C, 96h
120 °C, 2h
300 °C, 96h

BET Surface
Area (m2/g)

183
20
183
14
125
80

1.5-2 nm
Pore Volume
(cm3/g)
0.019
0.0013
0.018

0

0.016
0.006

2-10 nm
Pore Volume
(cm3/g)
0

0.003
0.004

0

0.003
0.021

10-25 nm
Pore Volume
(cm3/g)
0.005
0.005
0.006
0.005
0.005
0.007

Nitrogen adsorption data for Dex-SiO2-NPs ranging from a partial pressure of 0 to 0.3 correlate
well with the BET equation (Figure 3.19), with correlation coefficients > 98%. Dex-SiO2-200 and
Dex-SiO2-250 NPs exhibit a large decrease in surface area after calcination from 183 cm3/g to 20
cm3/g (Dex-SiO2-200) and 183 cm3/g to 14 cm3/g (Dex-SiO2-250) (Table 3.4). Each variation of
Dex-SiO2-NPs lose surface area following calcination, though Dex-SiO2-200 and Dex-SiO2-250
lose more than Dex-SiO2-350. This is hypothesized to arise as a result of the decreased structural
integrity at the surface of Dex-SiO2-NPs as the surface area to volume ratio increases. Here, the loss
of structural integrity is manifest as pore collapse and/or nanoparticle fusion. Gu and co-workers
and Li and co-workers made similar observations of low SiO2-NP surface areas as a function of
purification and heat treatment.79,98,99,115
3.5 Conclusions

Dex-SiO2-NPs present a challenge for typical chemical characterization techniques and have
herein been studied using both chemical and physical techniques. TEM was used to study the
physical differences of SiO2-NPs and Dex-SiO2-NPs, revealing hollow core-shell structures bearing
variable gradients and diameters for NPs synthesized in the presence of dextran. Elemental details
across a two dimensional plane of the NPs were obtained via S-TEM equipped with EDS capabilities,
confirming that the increased light transmitted through the centers of NPs were due to a decrease
in SiO2. While carbonaceous materials cannot be detected by EDS, solid state CP-MAS 13C-
NMR provided spectral evidence of dextran incorporation into Dex-SiO2-NPs. TGA revealed that

58

Figure 3.19: BET Fit of Dex-SiO2-NPs Before and After Calcination. Brunauer, Emmett, and
Teller Fit of Nitrogen Adsorption-Desorption Data for Dex-SiO2-200 NPs (left), Dex-SiO2-250
NPs (center), and Dex-SiO2-350 NPs (right) before (top) and after (bottom) calcination.

5-7% of Dex-SiO2-NP sample mass is due to water and 12-16% originates from dextran. Full
incorporation of dextran into Dex-SiO2-NPs would account for 24% of sample mass. Surface area
and porosity measurements revealed that Dex-SiO2-NPs contain both micro- and mesopores and
have surface areas ranging from 183 cm3/g for smaller NPs to 125 cm3/g for larger NPs.

While Soxhlet extraction was an inefficient method for removing dextran, TEM images of
NPs after extraction suggested that NPs within the same population may be heterogeneous in
dextran incorporation. The different appearance of Dex-SiO2-NPs by TEM after Soxhlet extraction
has been attributed to erosion by continual extraction with water. Calcination was successfully
implemented to remove dextran from Dex-SiO2-NPs. Surface area and porosity analysis following
calcination showed a decrease in surface area resulting from partial pore collapse and NP fusion.
TEM micrographs provide no evidence indicating pore collapse or fusion of NPs after calcination,
though this method is limited by the inability to observe details at the surface of Dex-SiO2-NPs.

DLS evidence indicates that, when dissolved in MilliQ water at 25 °C, a range of hydrodynamic
diameters of dextran and dextran aggregates exist. These data and other evidence in the literature

59

Figure 3.20: Schematic Diagram of Dex-SiO2-NPs Synthesis and Calcination.

suggest that the increased concentration of water around these aggregates results in higher rates of
TEOS hydrolysis and condensation than elsewhere in the reaction mixture, resulting in the observed
hollow structures. BET indicates the presence of micropores (< 2 nm) and mesopores (2 - 50 nm).
The change in isotherm shape and DFT pore volume analysis indicate that calcination results in a
decrease in microporosity and an increase in mesoporosity for Dex-SiO2-200 and Dex-SiO2-350
NPs. Based on these results, it can be concluded that dextran exists primarily throughout the silica
shell and acts as a porogen in the synthetic process (Figure 3.20). The range of pore sizes apparent
by pore size analysis and TEM are hypothesized to originate from dextran and larger aggregates of
dextran present in the synthesis solution.

60

CHAPTER 4

DESIGN OF AMINE-COATED DEXTRAN-INCORPORATED SILICA

NANOPARTICLES AS SIRNA DELIVERY VEHICLES

4.1 Introduction

Taking advantage of the RNA interference (RNAi) pathway for therapeutic gain requires the use
of a delivery vehicle in order to protect siRNA, transport it to the target location, and release it for
use in the RNAi pathway. Numerous variations of delivery vehicles have been developed to date.
Here, carbohydrate-silica nanoparticles with exposed amines are studied in order to understand the
role of diameter, carbohydrate incorporation, and amine content in silencing. Dex-SiO2-NPs are
effective at delivering siRNA and inducing silencing, as exhibited via in vitro sudies using enhanced
green fluorescent protein (EGFP) (Figure 4.1).11,12 Herein the design and characterization of Dex-
SiO2-NPs as delivery vehicles are described. The performance of select variations as siRNA
delivery vehicles is presented in terms of EGFP fluorescence in vitro.
4.2 Nanoparticle Delivery Vehicle Design and Characterization

Dex-SiO2-NP delivery vehicles were designed to be biocompatible, bind siRNA, and be of
precise size and shape. The carbohydrate dextran, known for its biocompatibility, is incorporated
into the nanoparticle. Amine moieties are incorporated into Dex-SiO2-NPs using APTES (Scheme
4.1). At physiological pH, at which the delivery vehicles would operate, exposed amines will be
protonated and therefore positively charged, electrostatically binding negatively charged siRNA.

Zeta potential measurements were used to verify that amine groups were present on the surface,
as indicated by a positive zeta potential. Comiskey varied the concentration of APTES in the
synthesis of from 0.05 - 99 mol % Dex-SiO2-NPs and SiO2-NPs (Figure 4.1). Comiskey observed
increasing silencing (decreased fluorescence of EGFP) from 0.05% amine to 40% amine in NPs
containing dextran. An increase in fluorescence, indicative of less effective silencing, was observed

61

Figure 4.1: EGFP Silencing by Comiskey Dex-SiO2-NPs and SiO2-NPs. Silencing of EGFP
expressed as relative fluorescence by Dex-SiO2-NPs (top) and SiO2-NPs (bottom) of varying
amine content.11

for the 50%, 80%, and 99% Dex-SiO2-NPs. SiO2-NPs synthesized without dextran did not elicit
significant silencing regardless of amine content.11

Scheme 4.1: Dex-SiO2-NP Synthetic Scheme.

The ninhydrin reagent was used to quantify accessible amines on Dex-SiO2-NPs according to
literature precedent.116,117 Reaction of a primary amine with ninhydrin yields a purple complex,

62

Scheme 4.2: General Reaction of Primary Amine and Ninhydrin to Produce Ruhemann’s Purple.

Table 4.1: Dex-SiO2-NPs Amine Quantification by Ninhydrin Assay. Zeta potential and Ninhydrin
Amine Quantification of Dex-SiO2-NPs with varying amine content, synthesized using Dex-SiO2-
350 conditions.

Zeta Potential in H2O

Amine Quantity

(mmol NH2/mg NP)

APTES
(mmol)

1.2
0.3
0.075
0.038

(mV)
+17
+25
+9
-4

0.0037
0.0079
0.0028
0.0013

Figure 4.2: Correlation between Dex-SiO2-NP zeta potential and amount of amine (mmol/mg)
detected by the ninhydrin assay for the NPs listed in Table 4.1.

known as Ruhemann’s purple, having a λmax of 588 nm (Scheme 4.2). A calibration curve was
constructed after reacting APTES with ninhydrin (Figure C.2). Table 4.1 summarizes the amine
quantification results for a series of Dex-SiO2-NPs synthesized with varying amine content.

With decreasing APTES, a decrease in zeta potential was observed (Table 4.1). With the
exception of the second entry in Table 4.1, a decrease in APTES is correlated with decreased amine

63

content measured via the ninhydrin assay. The second entry in Table 4.1 lies outside of this trend.
This may result from amines “buried” within the silica structure in the Dex-SiO2-NPs of the first
entry, hence inaccessible by the ninhydrin reagent and resulting in a smaller detectable amount by
the assay. Nonetheless, a correlation exists (R2 = 0.9269) between amine quatity and zeta potential
of the nanoparticles (Figure 4.2).
4.3 Nanoparticle Delivery Vehicle Performance

Dex-SiO2-NPs were evaluated on their ability to silence the expression of EGFP in H1299,
HEK293, and HELA cell lines. siRNA containing the complimentary sequence to the mRNA of
EGFP were transfected to cells using amine-functionalized Dex-SiO2-NPs as described in Chapter
6. A decrease in fluorescence is indicative of silencing. The data here are normalized against
untreated cells, such that a relative fluorescence less than 1 is indicative of silencing. Silencing
results for 40% amine-functionalized Dex-SiO2-NPs prepared using a variety of dextrans and
cyclodextrins in H1299 and HEK293 in vitro are presented herein.

Among the Dex-SiO2-NPs prepared using different molecular weights of dextran, Dex-SiO2-
350 NPs synthesized using 9-11 kDa dextran elicited the most significant silencing (Figure 4.3).
Of the Dex-SiO-350 NPs Comiskey prepared, the 40% amine with 9-11 kDa dextran also achieved
the best silencing (Figure 4.1).

The cyclodextrin-incorporated Dex-SiO2-NPs did not elicit significant silencing, with perfor-
mance slightly improved in HEK293 cells compared with H1299 (Figure 4.4). There was no
significant difference in silencing among α-cyclodextrin NPs and β-cyclodextrin NPs. Despite the
mediocre silencing results, none of the Dex-SiO2-NPs were cytotoxic.

In addition to the silencing results for the various Dex-SiO2-NPs described in this work, further
evidence of successful in vitro transfection was obtained via TEM. 24 hours following transfection,
H1299 and HELA cells were suspended in a resin. The resin was sectioned, and a thin film was
placed on a TEM grid. The resulting micrographs depict NPs existing within cellular vesicles
and surrounding cells (Figures 4.5 and B.57). Some Dex-SiO2-NPs appear smaller and less

64

Figure 4.3: EGFP Silencing of Dex-SiO2-NPs prepared using Dextran. In vitro silencing quanti-
fied via fluorescence of EGFP in H1299 (dark) and HEK293 (light) cell lines prepared using 6 kDa
dextran (top left), 9-11 kDa dextran (bottom left), and 40 kDa dextran (right). Experiments per-
formed using 200 µg/mL NP and 50 nM siRNA. Fluorescence measured 24 hours post-transfection.
Error bars represent the standard deviation of three data points.

dense, suggesting that Dex-SiO2-NPs may be degraded by the transfection process and/or cellular
components. S-TEM and EDS were used to verify the presence of silicon in objects assumed to
be Dex-SiO2-NPs. A variety of structures were studied, ranging from large, dark NPs to smaller,
more hollow NPs. As the line scans in Figure 4.6 illustrate, EDS confirmed that these structures
include silicon. The decreased intensity of the smaller NPs (Figure 4.6, center) indicates a decrease
in silicon abundance compared with the other structures.

65

Figure 4.4: EGFP Silencing of Dex-SiO2-NPs prepared using Cyclodextrin. In vitro silencing
quantified via fluorescence of EGFP in H1299 (dark) and HEK293 (light) cell lines prepared using,
from left, α-cyclodextrin and β-cyclodextrin. Experiments performed using 200 µg/mL NP and 50
nM siRNA. Fluorescence measured 24 hours post-transfection. Error bars represent the standard
deviation of three data points.

Figure 4.5: TEM Micrograph of 40% Amine Dex-SiO2-350 in H1299 Cells. H1299 cells suspended
in resin 24 hours post-transfection with siRNA-loaded Dex-SiO2-NPs. Brightness enhanced by
20% to emphasize detail, scale bar 1000 nm. Image from reference 12.

66

Figure 4.6: STEM Micrograph and EDS Line Scan of 40% Amine Dex-SiO2-350 in H1299
Cells. S-TEM micrographs of H1299 cells suspended in resin 24 hours post-transfection with
siRNA-loaded Dex-SiO2-NPs with area scanned for EDS analysis indicated by fuschia line (top).
Silicon elemental abundance profile along indicated area (bottom). Brightness enhanced by 20%
to emphasize detail, scale bar 1 µm.

4.4 Endocytosis of Dex-SiO2-NPs In Vitro

The origins of improved silencing of Dex-SiO2-NPs compared to SiO2-NPs were studied.
Different endocytotic inhibitors were incubated with cells prior to transfection (Table 4.2). After
24 hours, the fluorescence of EGFP was measured in cells transfected with naked siRNA (control),
LF2K, and Dex-SiO2-350 NPs functionalized with 40% APTES (Figure 4.7). No silencing was
observed when naked siRNA was used regardless of inhibitor.
Inhibition of ATP-dependent
processes and the scavenger receptor pathway was observed when LF2K was used for transfection.
Transfection using Dex-SiO2-NP was not inhibited by chloropromazine, filipin, or the two used
together. Inhibition was observed when cytochalasin D, low temperatures, and dextran sulfate were
used. This indicates that actin-dependent, energy-dependent, and scavenger receptor pathways are
involved in the endocytosis of these delivery vehicles.

67

Table 4.2: Endocytotic Inhibitors and Action Mechanism. Table adapted from reference 12.

Cytochalasin D
Temperature
Dextran Sulfate

Actin dependent processes
Energy dependent processes

Scavenger receptors

Inhibitor

Target

Chloropromazine Clathrin-mediated endocytosis
Caveolae-mediated endocytosis

Filipin

Figure 4.7: Silencing of EGFP in the Presence of Endocytotic Inhibitors. Silencing evaluated in
H1299 cells 24 hours post-transfection. Results normalized to cells containing nanoparticles only.
Error bars represent ±1 standard deviation; n = 3. *Significant difference (p < 0.05) as compared to
conditions without inhibitors. Image reproduced from reference 12 ©World Scientific Publishing
Co./Imperial College Press 2016.

68

4.5 Conclusions

Building on Comiskey’s work preparing Dex-SiO2-NPs with varying amine content, this work
has taken the 40% amine variation containing dextran and expanded the types of carbohydrate
incorporated in the NP. In addition to using 9-11 kDa dextran, studied by Comiskey, 6 kDa and
40 kDa dextran and α- and β-cyclodextrin were incorporated into NPs using Dex-SiO2-350, Dex-
SiO2-250, and Dex-SiO2-200 conditions. The silencing ability of this range of Dex-SiO2-NPs
were studied via EGFP knockdown in H1299 and HEK293 cells. While no significant insights
were gained from the variation of delivery vehicle size and dextran content, further studies may
be conducted varying amine content and synthetic conditions within these new Dex-SiO2-NP
classifications. Resin-suspended cells were inspected by TEM 24 hours post-transfection, revealing
Dex-SiO2-NPs within cellular vesicles and around the perimeter of cells. Both intact NPs and some
appearing to have been degraded were present in both regions, suggesting that the Dex-SiO2-NPs
are subject to degradation within the slightly acidic cellular environment.

69

CHAPTER 5

FUTURE RESEARCH DIRECTIONS

5.1 Expanding Synthetic Directions

The results and experiments herein have provided insight to the synthesis of Dex-SiO2-NPs.
Nevertheless, questions remain which merit further investigation. This work has shown that
monodispersed Dex-SiO2-NPs are easily accessible when purified silicate starting materials are
used. Expanding upon the knowledge that dextran aggregation gives rise to the unique morphology,
the diameters of accessible Dex-SiO2-NPs can be expanded. Further optimization of the post-
synthetic treatment of Dex-SiO2-200 and Dex-SiO2-250 NPs should also be explored to facilitate
characterization. Integration of different carbohydrates into SiO2-NPs is another direction that can
be further explored using the methods and techniques established herein.
5.2 Silencing Implications

The size variation and polydispersity of Comiskey’s nanoparticles add uncertainty to the role of
delivery vehicle size. While Comiskey’s work showed that silencing using Dex-SiO2-NPs was better
than SiO2-NPs, a better comparison may now be accomplished with monodispersed Dex-SiO2-NPs
and SiO2-NPs in a comparable size range. This would provide a more robust comparison of the
effect of delivery vehicle size on silencing ability. Since dextran can be removed by calcination,
re-functionalization of calcined Dex-SiO2-NPs with APTES or APTES and surface-bound dextran
would provide valuable control experiments.

70

CHAPTER 6

EXPERIMENTAL METHODS

6.1 Materials

Tetraethyl ortho silicate (TEOS) and 3-aminopropyl triethoxy silane (APTES) were obtained
from Millipore Sigma and distilled prior to use.118 9-11 kDa dextran was obtained from Millipore
Sigma and was used without further purification. Ethanol (200 proof), ammonium hydroxide (ap-
proximately 30% as NH3), and Milli-Q water (18MΩ) were used as received. 30 kDa molecular
weight cut off regenerated cellulose membranes (47 mm diameter) were purchased from Milli-
pore Sigma. MilliQ water was obtained from a Millipore Synergy UV water purification system,
RNAase free water was purchased from Fisher Scientific. 300 mesh carbon-coated copper TEM
grids and carbon-coated nickel grids were purchased from Ted Pella, Inc.

6.2 Characterization

Transmission Electron Microscopy

Nanoparticle powder was suspended in ethanol and sonicated until well dispersed. 5 µL of the
suspension was pipetted onto a copper TEM grid. Images were acquired using a JEOL 100CXII
transmission electron microscope operating at an accelerating voltage of 100 keV and equipped
with either a GATAN Orius digital camera or an Olympus MegaView III digital camera.
Distributions of nanoparticle diameter were obtained after measuring at least 100 nanoparti-
cles20,119 using CellProfiler 2.2.0. The CellProfiler pipeline first inverted the image and applied
a grayscale threshold such that dark pixels below the threshold are set to zero. Objects were
identified based on their intensity and their areas and diameters were measured. Object diameters
were separated into 10 nm bins based on diameter (converted from pixels to nanometers based on
the number of pixels comprising the scale bar) and eccentricity (< 0.4, more circular than oblong).

71

Histograms of object diameters were generated from no less than 100 individual measurements.
TEM micrographs of NP-containing cells were prepared as described in reference 12. 24 hours
post-transfection, H1299 or HELA cells were suspended in resin, allowed to solidify, and then
sectioned. A sectioned film was placed on a TEM grid and studied via TEM or S-TEM.

Silver Methanamine Staining for TEM

100 mL solutions of 3% hexamethylene tetramine (methenamine), 5 mL 5% silver nitrate, 0.5%
periodic acid and 5% sodium borate were prepared in MilliQ water. The silver methenamine stock
solution was prepared by adding the silver nitrate solution dropwise to the methenamine solution.
The solution was filtered through a 0.2 µm PTFE filter into an amber bottle. The stock solution
was stored at 4 ◦C until immediately prior to use. The silver methenamine working solution was
prepared by mixing 50 mL of silver methenamine stock with 5 mL of 5% sodium borate solution.
A 5 µL sample of Dex-SiO2-NPs was deposited onto a nickel TEM grid and air dried. Following
treatment with 0.5% periodic acid solution for 30 minutes, the grid was washed three times by
scooping 5 µL droplets of MQ water and absorbing it with the tip of a filter paper wedge. After
rinsing, 5 µL of silver methenamine working solution was deposited on the grid and air dried in
the dark. The grid was rinsed three times with MilliQ water and air dried prior to analysis.

Energy Dispersive X-Ray Spectroscopy

EDS was performed on a JEOL 2200FS scanning transmission electron microscope operating at
an accelerating voltage of 200 keV. EDS results, including line scans, were collected and analyzed
using INCA software.

Thermogravimetric Analysis

TGA measurements were obtained using a TA Q500 instrument under air. The sample was heated
with a thermal ramp rate of 15 °C per minute to 800 °C.

72

Surface Area and Porosity Analysis

Dried Dex-SiO2-NP powder was degassed at 120 °C for 12 hours prior to surface area and porosity
analysis. Nitrogen adsorption/desorption isotherms were obtained using a Micromeritics ASAP
2020 instrument. Brunauer, Emmett and Teller (BET), Barrett, Joyner and Halenda (BJH), and
Density Functional Theory (DFT) calculations were used to determine the surface area, porosity,
and porosity distributions.

CP-MAS 13C NMR
Dry Dex-SiO2-NP powder was loaded into a 6 mm zirconia rotor. Solid State CP-MAS 13C NMR
was performed on a Varian Infinity Plus spectrometer operating at 400 MHz. The sample was spun
at 5 kHz and was referenced to adamantane.

Dynamic Light Scattering

Hydrodynamic diameters and ζ potentials were measured using a Malvern Zetasizer NanoZS. Hy-
drodynamic radius measurements by dynamic light scattering were performed using disposable UV
cuvettes purchased from Brandtech®and ζ potential measurements were performed using a folded
capillary zeta cell purchased from Malvern Panalytical. A 500 µL stock solution of 1.0 mg/mL was
measured, diluted two-fold, and measured again. The sample was measured after every dilution
until the zeta potential and hydrodynamic diameter were stabilized near 0.05 mg/mL.

In vitro Gene Silencing Experiments
siRNA was bound to Dex-SiO2-NPs suspended in MilliQ H2O and evaluated for silencing in H1299
cells expressing EGFP as described in reference 12.

73

6.3 Synthetic Procedures

Synthesis of Dextran-SiO2 Nanoparticles
Preparation of Dex-SiO2-100
190 mL of absolute ethanol and 10 mL of MilliQ water were added to a 500 mL round bottom
flask. Dextran (9-11 kDa, 2.4 x 10-6 mol, 24 mg) was dissolved in 10 mL of MilliQ water and
added to the reaction solution, followed by 10 mL of NH4OH. TEOS (0.024 mmol, 0.053 mL)
was added dropwise by syringe and the mixture was stirred at room temperature for 16 hours. The
reaction solution was purified via pressure filtration under nitrogen using a 30 kDa molecular weight
cut off regenerated cellulose membrane. The solid Dex-SiO2-NP powder was transferred to a cen-
trifuge tube and rinsed three times with ethanol and twice with MilliQ water. Yields averaged 0.13 g.

Preparation of Dex-SiO2-200
190 mL of absolute ethanol and 10 mL of MilliQ water were added to a 500 mL round bottom
flask. Dextran (9-11 kDa, 2.4 x 10-6 mol, 24 mg) was dissolved in 10 mL of MilliQ water and
added to the reaction solution, followed by 10 mL of NH4OH. TEOS (2.4 mmol, 0.53 mL) was
added dropwise by syringe and the mixture was stirred at room temperature for 16 hours. The
reaction solution was purified via pressure filtration under nitrogen using a 30 kDa molecular weight
cut off regenerated cellulose membrane. The solid Dex-SiO2-NP powder was transferred to a cen-
trifuge tube and rinsed three times with ethanol and twice with MilliQ water. Yields averaged 0.13 g.

Preparation of Dex-SiO2-240
170 mL of absolute ethanol and 30 mL of MilliQ water were added to a 500 mL round bottom
flask. Dextran (9-11 kDa, 2.4 x 10-6 mol, 24 mg) was dissolved in 10 mL of MilliQ water and
added to the reaction solution, followed by 10 mL of NH4OH. TEOS (2.4 mmol, 0.53 mL) was
added dropwise by syringe and the mixture was stirred at room temperature for 16 hours. The
reaction solution was purified via pressure filtration under nitrogen using a 30 kDa molecular weight
cut off regenerated cellulose membrane. The solid Dex-SiO2-NP powder was transferred to a cen-

74

trifuge tube and rinsed three times with ethanol and twice with MilliQ water. Yields averaged 0.13 g.

Preparation of Dex-SiO2-350
150 mL of absolute ethanol and 50 mL of MilliQ water were added to a 500 mL round bottom
flask. Dextran (9-11 kDa, 2.4 x 10-6 mol, 24 mg) was dissolved in 10 mL of MilliQ water and
added to the reaction solution, followed by 10 mL of NH4OH. TEOS (2.4 mmol, 0.53 mL) was
added dropwise by syringe and the mixture was stirred at room temperature for 16 hours. The
reaction solution was purified via pressure filtration under nitrogen using a 30 kDa molecular weight
cut off regenerated cellulose membrane. The solid Dex-SiO2-NP powder was transferred to a cen-
trifuge tube and rinsed three times with ethanol and twice with MilliQ water. Yields averaged 0.13 g.

Preparation of Amino-Dextran-SiO2 Nanoparticles
150 mL of absolute ethanol and 10, 30, or 50 mL of MilliQ water were added to a 500 mL round
bottom Schlenk flask. Dextran (9-11 kDa, 2.4 x 10-6 mol, 24 mg) was dissolved in 10 mL of MilliQ
water and added to the reaction solution, followed by 10 mL of NH4OH. TEOS (2.4 mmol, 0.53
mL) was added dropwise by syringe and the mixture was stirred at room temperature for 10 minutes
before the addition of APTES (40 mol%, 0.96 mmol, 0.224 mL). The reaction solution was stirred
for 16 hours at room temperature and purified via pressure filtration under nitrogen using a 30 kDa
molecular weight cut off regenerated cellulose membrane. The solid Dex-SiO2-NP powder was
transferred to a centrifuge tube and rinsed three times with ethanol and twice with MilliQ water.
The filtered solid was resuspended in MilliQ water and sonicated until well dispersed. Yields
averaged 0.13 g.

Preparation of Amine-Coated SiO2 Nanoparticles Using Various Carbohydrates
150 mL of absolute ethanol and 10, 30, or 50 mL of MilliQ water were added to a 500 mL round
bottom Schlenk flask. The selected carbohydrate (6 kDa dextran, 4 x 10-6 mol, 24 mg; 9-11 kDa
dextran, 2.4 x 10-6 mol, 24 mg; 40 kDa dextran, 6 x 10-7 mol, 24 mg; α-cyclodextrin, 2.3 x 10-5

75

mol, 24 mg; β-cyclodextrin, 2.1 x 10-5 mol, 24 mg) was dissolved in 10 mL of MilliQ water and
added to the reaction solution, followed by 10 mL of NH4OH. TEOS (2.4 mmol, 0.53 mL) was
added dropwise by syringe and the mixture was stirred at room temperature for 10 minutes before
the addition of APTES (40 mol%, 0.96 mmol, 0.224 mL). The reaction solution was stirred for
16 hours at room temperature and purified via pressure filtration under nitrogen using a 30 kDa
molecular weight cut off regenerated cellulose membrane. The solid Dex-SiO2-NP powder was
transferred to a centrifuge tube and rinsed three times with ethanol and twice with MilliQ water.
The filtered solid was resuspended in MilliQ water and sonicated until well dispersed. Yields
averaged 0.13 g.

Preparation of SiO2 Nanoparticles
46 mL of absolute ethanol and 10 mL of NH4OH were added to a 100 mL round bottom flask
affixed with a Schlenk adapter. In a 20 mL scintillation vial, 4 mL of absolute ethanol was mixed
with TEOS (5.6 x 10-4 mol, 0.125 mL), this solution was added to the reaction flask dropwise via
syringe and stirred for 2 hours under nitrogen. The mixture was purified by repeated centrifugation
(30 min at 3000 rpm) and rinsing with ethanol three times. The filtered solid was suspended in
ethanol and sonicated until well dispersed. Yield 18 mg.

Process for Amine Coating of Nanoparticles

50 mL of absolute ethanol, 12.5 mL of MilliQ water, and 3.3 mL of NH4OH were added to a 250
mL Schlenk flask. SiO2 NPs or Dex-SiO2-NPs suspended in ethanol (0.0117 g/mL, 0.08775 g,
7.5 mL) were added to the reaction mixture, followed by APTES (concentration varied as listed in
Table 2), and stirred for 3 hours under nitrogen. For smaller concentrations of APTES (the final
two entries in Table 2), a 15 mM solution of APTES in ethanol was prepared and used to maintain
accuracy. The amount of ethanol initially added to the reaction mixture was reduced such that the
total volume remained the same. The mixture was purified by repeated centrifugation (30 min at
3000 rpm) and rinsing with ethanol three times. The filtered solid was suspended in ethanol and

76

sonicated until well dispersed.

Ninhydrin Amine Quantification Assay

100 mM ninhydrin solution in absolute ethanol was freshly prepared. Standard solutions of APTES
were prepared in a concentration range of 0.001 mM to 0.5 mM. Functionalized silica nanoparticles
were dried under vacuum at 100 °C for 4-6 hours, and then 20 mg of particle was placed in a capped
vial with 0.9 mL of absolute ethanol and sonicated until well dispersed. Alternatively, ethanol-
suspended nanoparticle was placed in an Eppendorf tube such that 10 mg of particle was present,
and centrifuged to remove the ethanol. 0.9 mL of absolute ethanol was added, sonicated until well
suspended, and transferred to a capped vial. To each vial, 0.1 mL of pH 5 sodium acetate/acetic acid
buffer and 1 mL of 100 mM ninhydrin solution was added, followed by sonication for 10 minutes.
Samples were incubated in an oil bath at 65 °C for 30 minutes, and then allowed to come to room
temperature. Following transfer to Eppendorf tubes, the sample solutions were centrifuged for 10
minutes at 6000 rpm. The supernatant was analyzed using an ATI Unicam UV2 spectrophotometer
with ethanol as blank.

Study of Dex-SiO2-NP Assembly
0.1 mL of reaction solution was extracted at designated time points of the synthesis of Dex-SiO2-
350 NPs. After removal using a syringe, each sample was diluted, centrifuged, and washed three
times with MilliQ water before suspension in ethanol for deposition upon a formvar-coated copper
or nickel TEM grid and, in some cases, stained using silver methanamine for dextran detection.
Method A: TEOS and 9-11 kDa dextran were stirred for 10 minutes, after which APTES was added
and the solution was stirred overnight. Method B: TEOS and 9-11 kDa dextran were stirred for 20
hours, after which APTES was added and the solution was stirred for one hour.

Dextran Solubility Study

The process for synthesizing Dex-SiO2-250 was used, varying the alcoholic solvent among

77

methanol, ethanol, or isopropanol and alternating the inclusion of dextran and TEOS such that
dextran only, TEOS only, and dextran and TEOS were included. Samples of 0.1 mL of reaction
solution were extracted periodically after the addition of Dextran or TEOS and injected into a
cuvette. Hydrodynamic diameter was measured by DLS. Products of experiments using TEOS or
TEOS and dextran were purified according to Dex-SiO2-250 and NPs were analyzed by TEM.

6.4 Post-Synthetic Treatment

Filtration

A 10 mL suspension of nanoparticle in water or 10% ethanol in water was prepared at 1.0, 0.1, and
0.01 mg/mL. A pressure filtration reservoir was equipped with the indicated membrane and secured
in place by a Teflon O-ring. The suspension was filtered under 40 psi of N2. The membrane was
removed, agitated and rinsed thoroughly with ethanol. The resulting suspension was labeled as the
retentate. Both the filtrate and the retentate were concentrated for TEM analysis.

Iterative Filtration

A 50 mL suspension of nanoparticle in ethanol was prepared at 0.01 mg/mL. A pressure filtration
reservoir was equipped with a 0.2 µm polycarbonate membrane beneath a 0.45 µm mixed cellulose
esters membrane, secured in place by a Teflon O-ring. The suspension was filtered under 40 psi of
N2. Both membranes were removed, agitated and rinsed thoroughly with ethanol. The resulting
suspension was labeled as the retentate. The filtrate was collected and kept separate from the
retentate. The pressure filtration device was reassembled with the same filters in the same order,
and the retentate was filtered once again. This process was initially performed four times, although
later repetitions increased that number in order to further separate the nanoparticles. Both the
filtrate and the retentate were concentrated and rinsed several times with MilliQ water, ultimately
being resuspended in MilliQ water in preparation for in vitro use.

78

Soxhlet Extraction
Dry nanoparticle powder (2 g) was weighed into a cellulose thimble (Whatman, SIZE). 250 mL of
MilliQ water was brought to a boil in a 500 mL round bottom flask under nitrogen with stirring. The
loaded thimble was placed into the Soxhlet extractor and the corresponding condenser was attached.
Extraction was performed over 2 days. The solution was filtered under nitrogen pressure using
a 30kDa MWCO regenerated cellulose membrane, concentrated, and lyophilized to recover dextran.

Calcination

Calcination was performed under air in a Thermo Scientific Thermolyne F48025-60 muffle furnace.
Dry nanoparticle powder was pulverized, placed in a crucible, and heated at 120 °C for 2 hours in
air, followed by a constant temperature between 350 °C and 900 °C for 2 - 100 hours.

79

APPENDICES

80

APPENDIX A

NMR SPECTRA

81

Figure A.1: 500 MHz 1H-NMR Spectrum of TEOS in CDCl3.

82

Figure A.2: 125 MHz 13C-NMR Spectrum of TEOS in CDCl3.

83

Figure A.3: 500 MHz 1H-NMR Spectrum of TEOS in CDCl3. Purified by distillation.

84

Figure A.4: 125 MHz 13C-NMR Spectrum of TEOS in CDCl3. Purified by distillation.

85

Figure A.5: 500 MHz 1H-NMR Spectrum of APTES in CDCl3.

86

Figure A.6: 125 MHz 13C-NMR Spectrum of APTES in CDCl3.

87

Figure A.7: 500 MHz 1H-NMR Spectrum of APTES in CDCl3. Purified by distillation.

88

Figure A.8: 125 MHz 13C-NMR Spectrum of APTES in CDCl3. Purified by distillation.

89

Figure A.9: 500 MHz 1H-NMR Spectrum of Dextran in D2O.

90

Figure A.10: 500 MHz 1H-NMR Spectrum of Dextran in D2O/DMSO-d6.. Dextran recovered from Soxhlet Extraction.

91

APPENDIX B

SUPPLEMENTAL TEM MICROGRAPHS

92

Figure B.1: EDS Spectrum of Dex-SiO2-NPs.

93

Figure B.2: EDS Spectrum of SiO2-NPs.

94

Figure B.3: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-SiO2-200
Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to emphasize
detail.

Figure B.4: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-SiO2-200
Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to emphasize
detail.

95

Figure B.5: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-SiO2-200
Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to emphasize
detail.

Figure B.6: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-SiO2-250
Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to emphasize
detail.

96

Figure B.7: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-SiO2-250
Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to emphasize
detail.

Figure B.8: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-SiO2-250
Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to emphasize
detail.

97

Figure B.9: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-SiO2-350
Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to emphasize
detail.

Figure B.10: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-SiO2-
350 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to emphasize
detail.

98

Figure B.11: TEM Micrograph of Dex-SiO2-NPs Prepared using 6 kDa Dextran and Dex-SiO2-
350 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to emphasize
detail.

99

Figure B.12: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and Dex-
SiO2-200 Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to
emphasize detail.

Figure B.13: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and Dex-
SiO2-200 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to
emphasize detail.

100

Figure B.14: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and Dex-
SiO2-200 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to
emphasize detail.

Figure B.15: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and Dex-
SiO2-250 Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to
emphasize detail.

101

Figure B.16: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and Dex-
SiO2-250 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to
emphasize detail.

Figure B.17: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and Dex-
SiO2-250 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to
emphasize detail.

102

Figure B.18: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and Dex-
SiO2-350 Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to
emphasize detail.

Figure B.19: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and Dex-
SiO2-350 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to
emphasize detail.

103

Figure B.20: TEM Micrograph of Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and Dex-
SiO2-350 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to
emphasize detail.

104

Figure B.21: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and
Dex-SiO2-200 Conditions . 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20%
to emphasize detail.

Figure B.22: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and
Dex-SiO2-200 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to
emphasize detail.

105

Figure B.23: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and
Dex-SiO2-200 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20%
to emphasize detail.

Figure B.24: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and
Dex-SiO2-250 Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to
emphasize detail.

106

Figure B.25: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and
Dex-SiO2-250 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to
emphasize detail.

Figure B.26: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and
Dex-SiO2-250 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20%
to emphasize detail.

107

Figure B.27: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and
Dex-SiO2-350 Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to
emphasize detail.

Figure B.28: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and
Dex-SiO2-350 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to
emphasize detail.

108

Figure B.29: TEM Micrograph of Calcined Dex-SiO2-NPs Prepared using 9-11 kDa Dextran and
Dex-SiO2-350 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20%
to emphasize detail.

109

Figure B.30: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and Dex-SiO2-
200 Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to emphasize
detail.

Figure B.31: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and Dex-SiO2-
200 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to emphasize
detail.

110

Figure B.32: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and Dex-SiO2-
200 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to emphasize
detail.

Figure B.33: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and Dex-SiO2-
250 Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to emphasize
detail.

111

Figure B.34: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and Dex-SiO2-
250 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to emphasize
detail.

Figure B.35: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and Dex-SiO2-
250 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to emphasize
detail.

112

Figure B.36: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and Dex-SiO2-
350 Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to emphasize
detail.

Figure B.37: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and Dex-SiO2-
350 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to emphasize
detail.

113

Figure B.38: TEM Micrograph of Dex-SiO2-NPs Prepared using 40 kDa Dextran and Dex-SiO2-
350 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to emphasize
detail.

114

Figure B.39: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-SiO2-
200 Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to emphasize
detail.

Figure B.40: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-SiO2-
200 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to emphasize
detail.

115

Figure B.41: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-SiO2-
200 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to emphasize
detail.

Figure B.42: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-SiO2-
250 Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to emphasize
detail.

116

Figure B.43: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-SiO2-
250 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to emphasize
detail.

Figure B.44: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-SiO2-
250 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to emphasize
detail.

117

Figure B.45: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-SiO2-
350 Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to emphasize
detail.

Figure B.46: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-SiO2-
350 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to emphasize
detail.

118

Figure B.47: TEM Micrograph of Dex-SiO2-NPs Prepared using α-cyclodextrin and Dex-SiO2-
350 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to emphasize
detail.

119

Figure B.48: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-SiO2-
200 Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to emphasize
detail.

Figure B.49: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-SiO2-
200 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to emphasize
detail.

120

Figure B.50: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-SiO2-
200 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to emphasize
detail.

Figure B.51: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-SiO2-
250 Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to emphasize
detail.

121

Figure B.52: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-SiO2-
250 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to emphasize
detail.

Figure B.53: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-SiO2-
250 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to emphasize
detail.

122

Figure B.54: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-SiO2-
350 Conditions. 14kx magnification, Scale bar 0.5 µm. Brightness enhanced by 20% to emphasize
detail.

Figure B.55: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-SiO2-
350 Conditions. 27kx magnification, Scale bar 0.2 µm. Brightness enhanced by 20% to emphasize
detail.

123

Figure B.56: TEM Micrograph of Dex-SiO2-NPs Prepared using β-cyclodextrin and Dex-SiO2-
350 Conditions. 67kx magnification, Scale bar 100 nm. Brightness enhanced by 20% to emphasize
detail.

124

Figure B.57: TEM Micrograph of 40% Amine Dex-SiO2-350 in HELA Cells. HELA cells sus-
pended in resin 24 hours post-transfection with siRNA-loaded Dex-SiO2-NPs. Brightness enhanced
by 20% to emphasize detail, scale bar 5000 nm.

125

APPENDIX C

SUPPLEMENTAL FIGURES

126

Figure C.1: UV-Visible Spectrum of Rheumann’s Purple at Variable Concentrations. APTES
reacted with ninhydrin reagent, ethanol used as blank

Figure C.2: Calibration Curve of Rheumann’s Purple. Generated from the absorbance at 588 nm
in the UV-Visible spectrum of APTES reacted with ninhydrin reagent, ethanol used as blank. Error
bars represent the standard deviation over 5 independent measurements.

127

Table C.1: Zeta Potentials of Dex-SiO2-NPs.

Synthesis Conditions Zeta Potential in H2O

(mV)
11.2
11.7
7.0
30.2
4.5
4.2
14.9
5.6
0.3
19.4
8.3
2.3
8.6
6.5
4.7

Carboyhdrate
6 kDa Dextran
6 kDa Dextran
6 kDa Dextran
9-11 kDa Dextran
9-11 kDa Dextran
9-11 kDa Dextran
40 kDa Dextran
40 kDa Dextran
40 kDa Dextran
α-cyclodextrin
α-cyclodextrin
α-cyclodextrin
β-cyclodextrin
β-cyclodextrin
β-cyclodextrin

Dex-SiO2-350
Dex-SiO2-250
Dex-SiO2-200
Dex-SiO2-350
Dex-SiO2-250
Dex-SiO2-200
Dex-SiO2-350
Dex-SiO2-250
Dex-SiO2-200
Dex-SiO2-350
Dex-SiO2-250
Dex-SiO2-200
Dex-SiO2-350
Dex-SiO2-250
Dex-SiO2-200

128

REFERENCES

129

REFERENCES

[1]

[2]

[3]

Fire, A.; Xu, S. Q.; Montgomery, M. K.; Kostas, S. A.; Driver, S. E.; Mello, C. C. Nature,
1998, 391, 806–811.
Elbashir, S. M.; Harborth, J.; Lendeckel, W.; Yalcin, A.; Weber, K.; Tuschl, T. Nature, May
2001, 411(6836), 494–498.
Cavallaro, G.; Sardo, C.; Craparo, E. F.; Porsio, B.; Giammona, G. Int. J. Pharm., June
2017, 525(2), 313–333.
Adesina, S. K.; Akala, E. O. Mol. Pharm., December 2015, 12(12), 4175–4187.

[4]
[5] Wittrup, A.; Lieberman, J. Nat. Rev. Genet., August 2015, 16(9), 543–552.

[6]

[7]

[8]

[9]

Bobbin, M. L.; Rossi, J. J. Annu. Rev. Pharmacol. Toxicol., 2016, 56(1), 103–122.
Angart, P.; Vocelle, D.; Chan, C.; Walton, S. Pharmaceuticals, March 2013, 6(4), 440–468.
Cooper, B. M.; Putnam, D. ACS Biomaterials Science & Engineering, 2016, 2(11), 1837–
1850.

Office of the Commissioner. Press announcements - FDA approves first-of-its kind tar-
geted RNA-based therapy to treat a rare disease. https://www.fda.gov/NewsEvents/
Newsroom/PressAnnouncements/ucm616518.htm. Accessed: 2018-8-15.

[10] Search of:

siRNA |

recruiting, not yet

rolling by invitation studies
https://www.clinicaltrials.gov/ct2/results?term=siRNA&cntry=US&
Search=Apply&recrs=b&recrs=a&recrs=f&recrs=d&age_v=&gndr=&type=&rslt=.
Accessed: 2019-1-20.

| united states

recruiting,
list

-

active, not
results

recruiting,

en-
- ClinicalTrials.gov.

[11] Comiskey, G. SYNTHESIS AND CHARACTERIZATION OF POLYMERIC AND INOR-
GANIC MATERIALS FOR BIOMEDICAL APPLICATIONS. PhD thesis, Michigan State
University, East Lansing, MI, 2013.

[12] Vocelle, D.; Chesniak, O. M.; Malefyt, A. P.; Comiskey, G.; Adu-Berchie, K.; Smith, M. R.;

Chan, C.; Walton, S. P. Technology, 2016, 04(01), 42–54.

[13] ALOthman, Z. Materials, December 2012, 5(12), 2874–2902.
[14] Gorrepati, E. A.; Wongthahan, P.; Raha, S.; Fogler, H. S. Langmuir, 2010, 26(13), 10467–

10474.

[15] Silva, C.; Airoldi, C. J. Colloid Interface Sci., November 1997, 195(2), 381–387.
[16] Zhang, J. H.; Zhan, P.; Wang, Z. L.; Zhang, W. Y.; Ming, N. B. J. Mater. Res., 2003, 18(3),

649–653.

130

[17] Kolbe, G. Friedrich-Schiller-Universitat Jena, 1956.
[18] Stöber, W.; Fink, A.; Bohn, E. J. Colloid Interface Sci., 1968, 69, 62–69.
[19] Van Blaaderen, A.; Van Geest, J.; Vrij, A. J. Colloid Interface Sci., 1992, 154(2).
[20] Han, Y.; Lu, Z.; Teng, Z.; Liang, J.; Guo, Z.; Wang, D.; Han, M.-Y.; Yang, W. Langmuir,

June 2017, 33(23), 5879–5890.

[21] Brinker, C. J.; Scherer, G. W. in Sol-Gel Science, Brinker, C. J.; Scherer, G. W., Eds., pp

96–233. Academic Press, San Diego, 1990.

[22] Bogush, G. H.; Tracy, M. A.; Iv, C. F. Z. J. Non-Cryst. Solids, 1988, 104, 95–106.
[23] Hartlen, K. D.; Athanasopoulos, A. P. T.; Kitaev, V. Langmuir, March 2008, 24(5), 1714–

1720.

[24] LaMer, V. K.; Dinegar, R. H. J. Am. Chem. Soc., November 1950, 72(11), 4847–4854.
[25] Lee, K.; Sathyagal, A. N.; McCormick, A. V. Colloids and Surfaces A: Physicochemical

and Engineering Aspects, 1998, 144, 115–125.

[26] Bogush, G. H.; Zukoski, C. F. J. Colloid Interface Sci., 1991, 142(1).
[27] Owens, G. J.; Singh, R. K.; Foroutan, F.; Alqaysi, M.; others. Prog. Mater Sci., 2016.
[28] Cha, W.; Fan, R.; Miao, Y.; Zhou, Y.; Qin, C.; Shan, X.; Wan, X.; Li, J. Molecules, May

2017, 22(5).

[29] Vallet-Regí, M.; Colilla, M.; Izquierdo-Barba, I.; Manzano, M. Molecules, December 2017,

23(1).

[30] Liberman, A.; Mendez, N.; Trogler, W. C.; Kummel, A. C. Surf. Sci. Rep., September 2014,

69(2-3), 132–158.

[31] Cheng, K.; Blumen, S. R.; MacPherson, M. B.; Steinbacher, J. L.; Mossman, B. T.; Landry,

C. C. ACS Appl. Mater. Interfaces, 2010, 2(9), 2489–2495.

[32] Steinbacher, J. L.; Landry, C. C. Langmuir, April 2014, 30(15), 4396–4405.
[33] Nampi, P. P.; Mohan, V. S.; Sinha, A. K.; Varma, H. Mater. Res. Bull., 2012, 47(6),

1379–1384.

[34] Li, L.; Liu, T.; Fu, C.; Meng, X.; Liu, H. J. Nanosci. Nanotechnol., 2016, 16, 6766–6772.
[35] Bobo, D.; Robinson, K. J.; Islam, J.; Thurecht, K. J.; Corrie, S. R. Pharm. Res., 2016.
[36] Li, J.-L.; Cheng, Y.-J.; Zhang, C.; Cheng, H.; Feng, J.; Zhuo, R.-X.; Zeng, X.; Zhang, X.-Z.

ACS Appl. Mater. Interfaces, February 2018, 10(6), 5287–5295.

[37] Solberg, S. M.; Landry, C. C. J. Phys. Chem. B, August 2006, 110(31), 15261–15268.

131

[38] Naessens, M.; Cerdobbel, A.; Soetaert, W.; Vandamme, E. J. J. Chem. Technol. Biotechnol.,

August 2005, 80(8), 845–860.

[39] Dextran

and

related

polysaccharides.

technical-documents/articles/biofiles/dextran-and-related.html.
cessed: 2019-1-23.

https://www.sigmaaldrich.com/
Ac-

[40] Mehvar, R. J. Control. Release, 2000, 69(1), 1–25.
[41] Larsen, C. Adv. Drug Deliv. Rev., 1989, 3(1), 103–154.
[42] Gil, E. C.; Colarte, A. I.; El Ghzaoui, A.; Durand, D.; Delarbre, J. L.; Bataille, B. Eur. J.

Pharm. Biopharm., 2008, 68(2), 319–329.

[43] Posocco, B.; Dreussi, E.; De Santa, J.; Toffoli, G.; Abrami, M.; Musiani, F.; Grassi, M.;

Farra, R.; Tonon, F.; Grassi, G.; Dapas, B. Materials, May 2015, 8(5), 2569–2615.

[44] Chen, X.; Yao, X.; Chen, L. Polym. Int., 2015, 64(3), 430–436.
[45] Zafar, R.; Zia, K. M.; Tabasum, S.; Jabeen, F.; Noreen, A.; Zuber, M. Int. J. Biol. Macromol.,

2016, 92, 1012–1024.

[46] Davis, M. E.; Brewster, M. E. Nat. Rev. Drug Discov., December 2004, 3(12), 1023–1035.
[47] Zhang, J.; Ma, P. X. Adv. Drug Deliv. Rev., 2013, 65(9), 1215–1233.
[48] O’Mahony, A. M.; O’Driscoll, C. M.; Caitriona, M.; O’Neill, M. J.; Godinho, B.; Darcy, R.;

Cryan, J. F. Pharmaceutical Nanotechnology, 2013, 1(1), 6–14.

[49] Lacatusu, I.; Badea, N.; Bojin, D.; Iosub, I.; Meghea, A. Mol. Cryst. Liq. Cryst., 2010,

521(October), 272–278.

[50] Numata, M.; Li, C.; Bae, A.-H.; Kaneko, K.; Sakurai, K.; Shinkai, S. Chem. Commun.,

2005, (37), 4655–4657.

[51] Coradin, T.; Lopez, P. J. Chembiochem, 2003, 4(4), 251–259.
[52] Hao, N.; Chen, X.; Jeon, S.; Yan, M. Adv. Healthc. Mater., 2015, 4(18), 2797–2801.
[53] Xie, M.; Lei, H.; Zhang, Y.; Xu, Y.; Shen, S.; Ge, Y.; Li, H.; Xie, J. RSC Adv., 2016, 6,

9328–9337.

[54] Lin, Z.; Li, J.; He, H.; Kuang, H.; Chen, X.; Xie, Z.; Jing, X.; Huang, Y. RSC Adv., 2015,

5(13), 9546–9555.

[55] Sinha, A.; Chakraborty, A.; Jana, N. R. ACS Appl. Mater. Interfaces, 2014, 6(24), 22183–

22191.

[56] Guo, M.; Jiang, M.; Pispas, S.; Yu, W.; Zhou, C. Macromolecules, 2008, 41(24), 9744–9749.
[57] Zafar, N.; Fessi, H.; Elaissari, A. Int. J. Pharm., 2014, 472(1-2), 118–129.

132

[58] Mansfield, E. D. H.; Pandya, Y.; Mun, E. A.; Rogers, S. E.; Abutbul-Ionita, I.; Danino, D.;

Williams, A. C.; Khutoryanskiy, V. V. RSC Adv., 2018, 8(12), 6471–6478.

[59] Salis, A.; Fanti, M.; Medda, L.; Nairi, V.; Cugia, F.; Piludu, M.; Sogos, V.; Monduzzi, M.

ACS Biomater. Sci. Eng., May 2016, 2(5), 741–751.

[60] Silvestri, B.; Pezzella, A.; Luciani, G.; Costantini, A.; Tescione, F.; Branda, F. Materials

Science and Engineering C, 2012, 32(7), 2037–2041.

[61] Krasucka, P.; Stefaniak, W.; Kierys, A.; Goworek, J. Microporous Mesoporous Mater., 2016,

221, 14–22.

[62] Krasucka, P.; Goworek, J.; Kierys, A. Adsorption, 2016, 22(4-6), 663–671.
[63] Wang, X.; Zhang, Y.; Luo, W.; Elzatahry, A. A.; Cheng, X.; Alghamdi, A.; Abdullah, A. M.;

Deng, Y.; Zhao, D. Chem. Mater., 2016, 28(7), 2356–2362.

[64] Gao, Z.; Zharov, I. Chem. Mater., March 2014, 26(6), 2030–2037.
[65] Xing, Z.; Tay, S.-W.; Ng, Y. H.; Hong, L. ACS Appl. Mater. Interfaces, May 2017, 9(17),

15103–15113.

[66] Wang, Y.; Wang, J.; Yang, Y.; Sun, Y.; Yuan, Y.; Li, Y.; Liu, C. Colloids Surf. B Biointerfaces,

May 2017, 153, 272–279.

[67] van Bommel, K. J. C.; Friggeri, A.; Shinkai, S. Angew. Chem. Int. Ed Engl., March 2003,

42(9), 980–999.

[68] van Bommel, K.; Jung, J. H.; Shinkai, S. Adv. Mater., 2001, 13(19), 1472–1476.
[69] Ruckdeschel, P.; Dulle, M.; Honold, T.; Förster, S.; Karg, M.; Retsch, M. Nano Res., 2016,

9(5), 1366–1376.

[70] Liberman, A.; Martinez, H. P.; Ta, C. N.; Barback, C. V.; Mattrey, R. F.; Kono, Y.; Blair,
S. L.; Trogler, W. C.; Kummel, A. C.; Wu, Z. Biomaterials, July 2012, 33(20), 5124–5129.

[71] Zhou, H.; Luo, J.; Gao, Q.; Yang, T. J. Appl. Polym. Sci., 2016, 44200, 1–8.
[72] Hu, J.; Chen, M.; Fang, X.; Wu, L. Chem. Soc. Rev., 2011, 40(40), 5472–5491.
[73] Park, S.-J.; Kim, Y.-J.; Park, S.-J. Langmuir, November 2008, 24(21), 12134–12137.
[74] Faulkner, R.; DiVerdi, J.; Yang, Y.; Kobayashi, T.; Maciel, G. Materials, December 2012,

6(1), 18–46.

[75] Alia, A.; Ganapathy, S.; de Groot, H. J. M. Photosynth. Res., November 2009, 102(2-3),

415–425.

[76] Polenova, T.; Gupta, R.; Goldbourt, A. Anal. Chem., June 2015, 87(11), 5458–5469.

133

[77]

Inkson, B. J. in Materials Characterization Using Nondestructive Evaluation (NDE) Meth-
ods, Hübschen, G.; Altpeter, I.; Tschuncky, R.; Herrmann, H.-G., Eds., pp 17–43. Woodhead
Publishing, January 2016.

[78] Ayache, J.; Beaunier, L.; Boumendil, J.; Ehret, G.; Laub, D.

in Sample Preparation
Handbook for Transmission Electron Microscopy: Methodology, Ayache, J.; Beaunier, L.;
Boumendil, J.; Ehret, G.; Laub, D., Eds., pp 33–55. Springer New York, New York, NY,
2010.

[79] Li, S.; Wan, Q.; Qin, Z.; Fu, Y.; Gu, Y. Langmuir, September 2016, 32(36), 9180–9187.
[80] Li, W.; Xu, Y.; Zhou, Y.; Ma, W.; Wang, S.; Dai, Y. Nanoscale Res. Lett., August 2012,

7(1), 485.

[81] Park, K.; Kittelson, D. B.; Zachariah, M. R.; McMurry, P. H. J. Nanopart. Res., June 2004,

6(2), 267–272.

[82] Yang, F.; Yang, W.; Zhu, L.; Chen, Y.; Ye, Z. Prog. Org. Coat., 2016, 95, 1–7.
[83] Zhong, G.; Guo, W.; Liu, Y.; Wei, Y.; Meng, X.; Hu, Z.; Liu, F. Colloids Surf. A Physicochem.

Eng. Asp., 2015, 485, 63–72.

[84] Chen, W.; Takai, C.; Khosroshahi, H. R.; Fuji, M.; Shirai, T. Colloids Surf. A Physicochem.

Eng. Asp., 2015, 481, 375–383.

[85] Malvern Panalytical Ltd. Zetasizer User Manual, January 2007.
[86] Hunter, R. J. Zeta Potential in Colloid Science: Principles and Applications. Academic

Press, September 2013.

[87] Moayedi, H.; Kazemian, S.; Kassim, K. A.; Nazir, R.; Raftari, M. J. Dispersion Sci. Technol.,

September 2013, 34(9), 1273–1279.

[88] Stetefeld, J.; McKenna, S. A.; Patel, T. R. Biophys. Rev., December 2016, 8(4), 409–427.
[89] Branda, F.; Silvestri, B.; Costantini, A.; Luciani, G. J. Sol-Gel Sci. Technol., January 2015,

73(1), 54–61.

[90] Sing, K. S. W.; Williams, R. T. Adsorpt. Sci. Technol., December 2004, 22(10), 773–782.

[91] Matthias Thommes, Katsumi Kaneko, Alexander V. Neimark, James P. Olivier, Francisco
Rodriguez-Reinoso, Jean Rouquerol and Kenneth S. W. Sing. Pure and Applied Chemistry,
2015, 87(9-10), 1051–1069.

[92] Kruk, M.; Jaroniec, M. Chem. Mater., October 2001, 13(10), 3169–3183.
[93] Gregg, S. J.; Sing, K. S. W. Adsorption, surface area, and porosity. Academic Press, 1991.
[94] Sing, K. S. W.; Williams, R. T. Adsorpt. Sci. Technol., 2005, 23(10), 839–853.
[95] Sing, K. Colloids Surf. A Physicochem. Eng. Asp., 2001, 187-188, 3–9.

134

[96] Olivier, J. P. J. Porous Mater., July 1995, 2(1), 9–17.
[97] Lastoskie, C. M.; Gubbins, K. E. Stud. Surf. Sci. Catal., 2000, 128, 41–50.
[98] Li, S.; Wan, Q.; Qin, Z.; Fu, Y.; Gu, Y. Langmuir, January 2015, 31(2), 824–832.
[99] Li, S.; Wan, Q.; Qin, Z.; Fu, Y.; Gu, Y. J. Therm. Anal. Calorim., November 2018.
[100] Li, Z.-J.; Liu, C.-R.; Zhao, Q.-S. J. Non-Cryst. Solids, March 2000, 265(1), 189–192.
[101] Kuang, Y.; Song, S.; Huang, J.; Sun, X. J. Solid State Chem., 2015, 224, 120–126.
[102] Xiong, B.; Cheng, J.; Qiao, Y.; Zhou, R.; He, Y.; Yeung, E. S. J. Chromatogr. A, June 2011,

1218(25), 3823–3829.

[103] Microfiltration membranes | chromatography sample prep - life science research | EMD
http://www.emdmillipore.com/US/en/life-science-research/

millipore.
chromatography-sample-preparation/microfiltration-membranes/FqCb.qB.
7ZkAAAFBDQZlvzJE,nav. Accessed: 2018-12-31.

[104] Chen, S.-L.; Dong, P.; Yang, G.-H.; Yang, J.-J. Ind. Eng. Chem. Res., January 1996, 35(12),

4487–4493.

[105] Van Blaaderen, A.; Vrij, A. J. Colloid Interface Sci., 1993, 156, 1–18.
[106] Su, W.-F. in Principles of Polymer Design and Synthesis, Su, W.-F., Ed., pp 9–26. Springer

Berlin Heidelberg, Berlin, Heidelberg, 2013.

[107] Antoniou, E.; Tsianou, M. J. Appl. Polym. Sci., August 2012, 125(3), 1681–1692.
[108] de Belder, a. N. Amersham Bioscience, 2003, p 60.
[109] Granath, K. A. J. Colloid Sci., 1958, 13(4), 308–328.
[110] Kasaai, M. R. Carbohydr. Polym., 2012, 88(1), 373–381.
[111] Erickson, H. P. Biol. Proced. Online, 2009, 11(1), 32–51.
[112] Bao, Y.; Shi, C.; Wang, T.; Li, X.; Ma, J. Microporous Mesoporous Mater., June 2016, 227,

121–136.

[113] Guo, Y.; Andrew Davidson, R.; Peck, K. A.; Guo, T. J. Colloid Interface Sci., May 2015,

445, 112–118.

[114] Hu, J.; Chen, M.; Fang, X.; Wu, L. Chem. Soc. Rev., 2011, 40(40), 5472–5491.
[115] Li, Z.-J.; Liu, C.-R.; Zhao, Q.-S. J. Non-Cryst. Solids, March 2000, 265(1), 189–192.
[116] Soto-Cantu, E.; Cueto, R.; Koch, J.; Russo, P. S. Langmuir, April 2012, 28(13), 5562–5569.
[117] Szegedi, A.; Popova, M.; Goshev, I.; Klébert, S.; Mihály, J. J. Solid State Chem., 2012, 194,

257–263.

135

[118] Shlyakhtenko, L. S.; Gall, A. A.; Lyubchenko, Y. L. in Cell Imaging Techniques: Methods
and Protocols, Taatjes, D. J.; Roth, J., Eds., Methods in Molecular Biology, pp 295–312.
Humana Press, Totowa, NJ, 2 edition, 2013.

[119] Nagao, D.; Satoh, T.; Konno, M. J. Colloid Interface Sci., December 2000, 232(1), 102–110.

136