Liquid Crystalline Phase Behavior and Fiber Spinning of Double-Stranded DNA  
Stabilized Single-Walled Carbon Nanotube Dispersions  
 
by 
 
Geyou Ao 
 
 
 
 
A dissertation submitted to the Graduate Faculty of 
Auburn University 
in partial fulfillment of the 
requirements for the Degree of 
Doctor of Philosophy 
 
Auburn, Alabama 
May 7, 2012 
 
 
 
 
Keywords: single-walled carbon nanotubes, double-stranded DNA, nematic 
 and cholesteric liquid crystal, phase behavior, macroscopic assembly  
 
 
Copyright 2012 by Geyou Ao 
 
 
Approved by 
 
Virginia A. Davis, Chair, Mary and John H. Sanders Associate Professor of Chemical 
Engineering 
Mark E. Byrne, Daniel F. & Josephine Breeden Associate Professor of Chemical Engineering 
Yoon Y. Lee, Uthlaut Family Professor of Chemical Engineering 
Aleksandr Simonian, Professor of Materials Engineering 
 
 
 
ii 
 
 
 
 
 
Abstract 
 
 
This dissertation describes the first discovery and characterization of lyotropic cholesteric 
single-walled carbon nanotube (SWNT) liquid crystals phase where the SWNT acts as a 
mesogens and is not simply inserted into an existing lyotropic dispersion. The liquid crystal 
phase was formed by concentrating mixtures of SWNT and aqueous double-stranded 
deoxyribonucleic acid (dsDNA). The characteristic cholesteric liquid crystal fingerprint texture 
with multi-colored regions was obtained due to the presence of the cholesterogenic forming 
biopolymer dsDNA. Depending on the dispersion methodology, the polydomain nematic phase 
previously reported for other lyotropic carbon nanotube dispersions could also be obtained. The 
phase behavior and dispersion microstructure were affected by the relative concentrations of 
dsDNA and SWNT and whether small bundles were removed prior to concentrating the 
dispersions. 
The phase behavior and rheology of aqueous dsDNA/SWNT dispersions were 
determined by a combination of rheology and optical microscopy. The results indicated that 
SWNT in dsDNA/SWNT dispersions behave as rigid rods. With increasing concentration the 
dispersions transitioned from a dilute phase with free rotational and translational rod motion, to a 
semidilute phase where rod rotation was inhibited, to a biphasic regime consisting of isotropic 
and liquid crystalline phases, and finally to a single liquid crystalline phase. The results of this 
research will enable new fundamental investigations comparing nematic and cholesteric liquid 
crystalline phase behavior and shear response. The simpler, aqueous cellulose nanocrystal (CNC) 
iii 
system was also investigated to provide greater insights into the more complex dsDNA/SWNT 
system. The CNC system also formed a cholesteric liquid crystal, but the effects of concentration 
on rheological properties were markedly different from both the dsDNA/SWNT system and 
classical lyotropic liquid crystalline polymers.  
The assembly of biopolymer stabilized SWNT dispersions into films and fibers was also 
investigated. In the case of cholesteric dsDNA/SWNT dispersions, casting films onto a substrate 
without any applied shear enabled retention of the helical microstructure. This finding may 
enable the production of films that possess not only SWNT strength and conductivity, but also 
with optical signatures. Shearing cholesteric dispersions during processing resulted in densely 
packed aligned films that have potential use as polarized films. Preliminary investigations of wet 
solution spinning of dsDNA/SWNT dispersions showed the significant impact the coagulant had 
on fiber microstructure. Investigation of lysozyme, (1-tetradecyl) trimethyl-ammonium bromide, 
SWNT dispersions (LSZ/TTAB/SWNT) fiber spinning resulted in promising microstructural 
characteristics and mechanical properties. As a result, fiber spinning optimization is being 
pursued by other researchers. 
The results of this research highlight the rich rheology and phase behavior of 
nanocylinder dispersions. These insights have provided a foundation for producing aligned bulk 
materials consisting of nanocylinder building blocks. In particular, the previously never achieved 
lyotropic cholesteric microstructure from SWNT dispersions and the potential enhanced 
biocompatibility of dsDNA/SWNT suggest that the range of applications that can be processed 
from liquid crystalline nanotube dispersions may be even broader than previously thought. 
 
 
 
 
iv 
 
 
 
 
 
Acknowledgments 
 
 
Pursuing my PhD degree in Auburn University has been one of the most rewarding and 
unforgettable journey of my life. However, I would not be able to accomplish this research 
without the contributions of numerous people. 
I am grateful to my advisor and dissertation committee: Virginia A. Davis for the vision 
and guidance which make this research possible and for her support and encouragement; Mark E. 
Byrne and Yoon Y. Lee for their classroom instruction and interest in my research; Aleksandr 
Simonian for generously agreeing to let me use lab instruments related to this research. 
I thank other researchers at Auburn University, especially: Vinod K. Radhakrishnan for 
training me on instruments at the beginning of this research; Matthew J. Kayatin for helping to 
solve experimental problems and useful discussions during the entire period of this project; 
Daniel W. Horn for fiber spinning experiments and his insights in developing the fiber spinning 
process; Esteban E. Ure?a-Benavides for useful discussions on rheology of liquid crystal SWNT 
dispersions; and Michelle Aono for characterizing the double-stranded nature of DNA used in 
this research. I also thank Philippe Poulin?s Group at CNRS-CRPP France for helping me 
understand fiber spinning techniques of SWNT dispersions. 
Last but not least, I thank my family and friends, especially my parents and my best friend 
Boelos Khereid for their constant support and encouragement. They make me believe that life is 
wonderful and full of happiness.  
 
 
v 
 
 
 
 
 
Table of Contents 
 
 
Abstract  ....................................................................................................................................... ii 
Acknowledgments ...................................................................................................................... iv  
List of Figures  .......................................................................................................................... viii  
List of Tables  ........................................................................................................................... xiii  
Chapter 1 ? Introduction  ............................................................................................................. 1 
Chapter 2 ? Background  ............................................................................................................. 6 
2.1 Carbon Nanotubes  ........................................................................................................... 7 
2.2 Deoxyribonucleic Acid (DNA)  ..................................................................................... 12 
2.3 Lysozyme  ...................................................................................................................... 15 
2.4 Cellulose Nanocrystals (CNC)  ...................................................................................... 17 
2.5 Biopolymer Stabilization of Carbon Nanotubes  ........................................................... 21 
2.6 Interaction of DNA with Single-Walled Carbon Nanotubes (SWNT)  ......................... 23 
2.7 Phase Behavior of Biopolymer-Stabilized Carbon Nanotube Dispersions ................... 26 
2.7.1 Onsager versus Flory Lattice Theory  ................................................................... 31 
2.7.2 Friedelian Classes of Liquid Crystals  .................................................................. 34 
2.7.3 Liquid Crystallinity of Biopolymer-Stabilized Carbon Nanotube Dispersions  ... 35 
2.7.4 Liquid Crystallinity of DNA-Stabilized SWNT Dispersions  .............................. 37 
2.8 Rheology of Liquid Crystalline Polymers  .................................................................... 39 
2.8.1 Rheology of Dilute and Semidilute Brownian Rods Dispersions  ........................ 40 
vi 
2.8.2 Rheological Signatures of Nematic Liquid Crystalline Polymers  ....................... 46 
2.8.3 Shear Effects on Liquid Crystal Rheology  .......................................................... 53 
2.8.4 Temperature Effects on Liquid Crystal Rheology  ............................................... 55 
2.8.5 Persistence Length Effects on Rheology  ............................................................. 57 
2.9 Macroscopic Assembly of Biopolymer-Stabilized Carbon Nanotube Dispersions  ...... 59 
2.9.1 Films/Coatings  ..................................................................................................... 60 
2.9.2 Fiber Spinning  ...................................................................................................... 61 
Chapter 3 ? Experimental Details  ............................................................................................. 65 
3.1 Materials  ....................................................................................................................... 65 
3.2 Preparation of SWNT Dispersions ................................................................................ 67 
3.3 Characterization of Double-Stranded DNA (dsDNA)  .................................................. 69 
3.4 Macroscopic Assembly of Films and Fibers  ................................................................. 70 
3.5 Mechanical and Electrical Testing of Fibers  ................................................................ 73 
3.6 Microscopy  ................................................................................................................... 74 
3.7 Spectroscopy  ................................................................................................................. 76 
3.8 Thermogravimetric Analysis  ........................................................................................ 76 
3.9 Rheology  ....................................................................................................................... 77 
Chapter 4 ? Rheology and Phase Behavior of Aqueous dsDNA/SWNT Dispersion  ............... 81 
4.1 Determining SWNT Concentration  .............................................................................. 81 
4.2 DNA Characterization  .................................................................................................. 88 
4.3 Phase Behavior and Rheology of dsDNA/SWNT Dispersions  .................................... 91 
4.3.1 Phase Behavior of dsDNA/SWNT Dispersions  ................................................... 92 
4.3.2 Rheology of dsDNA/SWNT Dispersions  .......................................................... 105 
vii 
4.4 Conclusions  ................................................................................................................. 150 
Chapter 5 ? Macroscopic Assembly of Films and Fibers  ....................................................... 152 
5.1 dsDNA/SWNT Films  .................................................................................................. 152 
5.2 dsDNA/SWNT Fiber Spinning  ................................................................................... 155 
5.3 Lysozyme/SWNT Fiber Spinning ............................................................................... 161 
5.4 Fiber Properties  ........................................................................................................... 162 
5.5 Conclusions  ................................................................................................................. 165 
Chapter 6 ? Phase Behavior of Lyotropic CNC Suspensions  ................................................. 166 
6.1 Preparation of CNC Suspensions  ................................................................................ 166 
6.2 Polarized Optical Microscopy of CNC Suspensions  .................................................. 167 
6.3 Rheology of CNC Suspensions  ................................................................................... 170 
6.4 Conclusions  ................................................................................................................. 176 
Chapter 7 ? Conclusions  ......................................................................................................... 177 
References  ............................................................................................................................... 180 
Appendix 1  .............................................................................................................................. 195 
Appendix 2  .............................................................................................................................. 199 
 
 
 
 
 
 
 
 
viii 
 
 
 
 
 
List of Figures 
 
 
Figure 2.1. Construction of carbon nanotube from single graphene sheet. . ................................ 9 
Figure 2.2. Schematic theoretical model for armchair, zig-zag and chiral SWNT. .................... 10 
Figure 2.3. Watson-Crick DNA double helix and base pairs.  .................................................. 14 
Figure 2.4. Ball-and-stick representation of lysozyme structure highlighting the disulfide 
bridges at the core.  .................................................................................................................... 17 
 
Figure 2.5. Chemical structure of sulfonated cellulose after acid hydrolysis.  .......................... 18 
 
Figure 2.6. Concentration regimes for rod-like molecules.  ...................................................... 27 
Figure 2.7. Phase diagram of SWNT in superacids and cross-polarized light micrographs 
showing the effect of solvent quality and SWNT concentration on microstructure.  ................ 29 
 
Figure 2.8. A diagrammatic representation of the excluded volume of two rod-like molecules, 
of length L and diameter d.  ....................................................................................................... 32 
 
Figure 2.9. Representations of molecular organizations in the nematic, cholesteric, and  
smectic Friedelian classes of liquid crystals. . ............................................................................ 35 
 
Figure 2.10. Universal shear thinning curve of monodisperse dilute Brownian rods. .............. 44 
Figure 2.11. Zero-shear viscosity versus concentration for PBLG.  .......................................... 45 
Figure 2.12. Viscosity versus shear rate curve showing a three region behavior.  .................... 48 
Figure 2.13. Shear viscosity and first normal stress difference versus shear rate for 17 %  
PBLG (MW=350,000) in m-cresol.  .......................................................................................... 51 
 
Figure 3.1. An illustration of a typical syringe extrusion process showing the fiber extrusion  
and collection method.  .............................................................................................................. 71 
 
Figure 3.2. Schematic drawing of the experimental set-up used to make nanotubes ribbons. ... 72 
Figure 3.3. The test frame for mechanical testing of SWNT fibers.  ......................................... 74 
ix 
Figure 3.4. The Two-Plate-Model for simple shear test with relative coordinate system.  ....... 78 
Figure 4.1. UV-vis calibration curves for dsDNA/SWNT dispersions with dsDNA:SWNT  
weight ratio of 7.5:1 without centrifugation using Unidym SWNT.  ........................................ 82 
 
Figure 4.2. UV-vis calibration curves for supernatant of dsDNA/SWNT dispersions with 
different initial volumes preparation using Unidym SWNT.  .................................................... 83 
 
Figure 4.3. UV-vis spectra for supernatant dsDNA/SWNT dispersions with dsDNA:SWNT 
weight ratio of 5.1:1.0 at different SWNT concentrations using Unidym SWNT.  .................. 83 
 
Figure 4.4. Thermogravimetric analysis (TGA ) of dsDNA.  .................................................... 85 
Figure 4.5. TGA profiles of 7.65 wt % dsDNA-1.02 wt % SWNT suspensions.  ..................... 86 
 
Figure 4.6. UV-Vis Absorbance spectra for DNA only solutions before and after 
sonication/centrifugation which were prepared from 0.75 wt % initial concentration of DNA 
in DI water. . ............................................................................................................................... 90 
 
Figure 4.7. Agarose gel electrophoresis of DNA samples in DI water.  .................................... 90 
Figure 4.8. Fluorescence quantitation of comparing DNA solutions before and after 
sonication/centrifugation using picogreen. ................................................................................. 91 
 
Figure 4.9. Optical microscopy of dsDNA/SWNT dispersions of (a) 13.6:1.0 by volume 
dsDNA:SWNT without centrifugation at a SWNT concentration of 0.24 vol % and (b) 
supernatant of 4.4:1.0 dsDNA:SWNT at a SWNT concentration of 0.69 vol %.  .................... 93 
 
Figure 4.10. Comparison of dsDNA/SWNT dispersions of (a) non-centrifuged 4.3:1.0 by  
volume dsDNA:SWNT at 1.7 vol % SWNT and (b) supernatant of 4.4:1.0 by volume 
dsDNA:SWNT at 2.3 vol % SWNT under cross-polarized microscopy.  ................................. 94 
 
Figure 4.11. Optical microscopy of concentrated bulk dispersion of dsDNA:SWNT = 8.5:1.0  
by wolume at 0.8 vol % SWNT.  ............................................................................................... 95 
 
Figure 4.12. Relationship between viscosity and concentration at a shear rate of 0.1 s-1 and a 
temperature of 10 oC for bulk (non-centrifuged) dsDNA/SWNT dispersions obtained from 
8.5:1.0 and 4.3:1.0 dsDNA:SWNT by volume.  ........................................................................ 97 
 
Figure 4.13. Relationship between viscosity and concentration for supernatants of 4.4:1 by 
volume dsDNA:SWNT at a shear rate of 0.1 s-1 and temperature of 10 oC.  ............................ 99 
 
Figure 4.14. Optical Microscopy of concentrated supernatant of 4.4:1.0 by volume 
dsDNA:SWNT (5.1:1.0 by weight) at 2.5 vol % SWNT. ....................................................... 100 
 
Figure 4.15. Plots of ln(1/P) of 4.4:1.0 by volume supernatant dsDNA/SWNT dispersions  
x 
against ln(c) of SWNT concentration at (a) 1.2, 1.3, 1.5, and 2.0 vol %, and (b) 2.0, 2.3, 2.5,  
2.6, and 3.2 vol %.  .................................................................................................................. 102 
 
Figure 4.16. Fingerprint texture in (a) 19.3 vol % of dsDNA solution and (b) bulk 6.4:1.0 by  
vol dsDNA/SWNT dispersion at a SWNT concentration of 1.3 vol % under cross-polarized 
microscopy.  ............................................................................................................................. 105 
 
Figure 4.17. (a) Steady shear viscosity of isotropic dsDNA/SWNT supernatant dispersions  
as a function of shear rate for various SWNT concentrations. (b) Zero-shear viscosity as a 
function of SWNT concentration.  ........................................................................................... 107 
 
Figure 4.18. (a) dsDNA only solvent viscosity ?s versus shear rate at concentrations 
corresponding to the same dsDNA concentrations in supernatant of dsDNA/SWNT at 388,  
705, 912, and 1576 ppm SWNT. (b) Intrinsic viscosity [?] = (? - ?s)/( ?s ?) versus shear  
rate for supernatant dsDNA/SWNT dispersions in the dilute regime (from 388 to 705 ppm 
SWNT) and semidilute regime (1576 ppm). ........................................................................... 110 
 
Figure 4.19. The steady shear viscosity of dsDNA/SWNT supernatant dispersions at  
different SWNT concentrations.  ............................................................................................. 112 
 
Figure 4.20. The complex viscosity of dsDNA/SWNT supernatant dispersions at different 
SWNT concentrations.  ............................................................................................................ 116 
 
Figure 4.21. The steady shear viscosity versus concentration at various shear rates for 
dsDNA/SWNT supernatant dispersions.  ................................................................................ 119 
 
Figure 4.22. The complex viscosity versus concentration curves at various frequencies for 
dsDNA/SWNT supernatant dispersions.  ................................................................................ 120 
 
Figure 4.23. The steady shear viscosity of bulk dsDNA/SWNT (noncentrifuged) dispersions  
with (a) 8.5:1, (b) 6.4:1, and (c) 4.3:1 of dsDNA/SWNT ratio by volume at various SWNT 
concentrations.  ........................................................................................................................ 122 
 
Figure 4.24. Time-concentration superposition of steady shear viscosities of dsDNA/SWNT 
supernatant dispersions.  .......................................................................................................... 124 
 
Figure 4.25. The effect of concentration on steady shear viscosities of dsDNA/SWNT 
supernatant dispersions.  .......................................................................................................... 126 
 
Figure 4.26. The steady shear viscosity versus shear stress plots of bulk dsDNA/SWNT 
(noncentrifuged) dispersions with (a) 8.5:1, (b) 6.4:1, and (c) 4.3:1 of dsDNA/SWNT ratio  
by volume at various SWNT concentrations.  ......................................................................... 128 
 
Figure 4.27. Cox-Merz rule alternatively (a) obeyed and (b) not obeyed at different SWNT 
concentrations for dsDNA/SWNT supernatant dispersions.  .................................................. 131 
 
xi 
Figure 4.28. Cox-Merz rule not obeyed at higher SWNT concentrations for dsDNA/SWNT 
noncentrifuged dispersions with 6.4:1 dsDNA/SWNT ratio by volume.  ............................... 132 
 
Figure 4.29. The ratio of the viscous to elastic modulus (tan (?) = G?/G?) versus frequency  
for dsDNA/SWNT supernatant dispersions at different SWNT concentrations.  ................... 134 
 
Figure 4.30. Elastic (storage) modulus G? as a function of frequency for dsDNA/SWNT 
supernatant dispersions at different SWNT concentrations.  ................................................... 136 
 
Figure 4.31. (a) The storage plateau modulus    , obtained at 0.1 s-1 angular frequency,  
versus the reduced volume fraction (? - ?*) of SWNT. (b) The value of ?*=0.2101 vol %  
gave the best linear fit with R2=0.980 and ? =3.62 ? 0.05 for     at 1 s-1 angular frequency 
versus (? - ?*) of SWNT. . ....................................................................................................... 138 
 
Figure 4.32. Cole-Cole plots show the storage modulus as a function of loss modulus and 
increasing SWNT concentration for supernatant 4.4:1.0 volume ratio of dsDNA:SWNT 
dispersions. .............................................................................................................................. 139 
 
Figure 4.33. Viscous (loss) modulus G? as a function of frequency for dsDNA/SWNT 
supernatant dispersions at different SWNT concentrations.  ................................................... 141 
 
Figure 4.34. The (a) G? and (b) G? as a function of concentration at different angular  
frequency for dsDNA/SWNT supernatant dispersions.  .......................................................... 143 
 
Figure 4.35. Time-temperature superposition of supernatant dsDNA/SWNT (Rice 187.2) 
dispersions with 4.4:1 dsDNA:SWNT ratio by volume at SWNT concentration of  
0.33 vol %; (a) steady shear viscosity versus shear rate, (b) complex viscosity versus  
angular frequency, and (c) storage modulus versus angular frequency at various temperatures  
of 10, 25, and 40 oC.  ............................................................................................................... 145 
 
Figure 4.36. Time-temperature superposition of supernatant dsDNA/SWNT (Rice 187.2) 
dispersions with 4.4:1 dsDNA:SWNT ratio by volume at SWNT concentration of 
 3.74 vol %; (a) steady shear viscosity versus shear rate, (b) complex viscosity versus  
angular frequency, and (c) storage modulus versus angular frequency at various temperatures  
of 10, 25, and 35 oC.  ............................................................................................................... 147 
 
Figure 4.37. Time-temperature superposition of bulk dsDNA/SWNT noncentrifuged  
dispersions with 6.4:1 dsDNA/SWNT ratio by volume at SWNT concentration of  
1.67 vol %; (a) steady shear viscosity versus shear rate, (b) complex viscosity versus  
angular frequency, and (c) storage modulus versus angular frequency at various temperatures  
of 10, 25, and 35 oC.  ............................................................................................................... 149 
 
Figure 5.1. (a) Translucent film obtained from a 4.4:1.0 by volume supernatant of 
dsDNA/SWNT dispersion (~2.0 vol % of SWNT) by shearing at approximately 500 s-1  
shear rate. (b-d) Cross-polarized optical microcopy images of films prepared without  
applied shear. The colors vary as the sample is rotated in (d) 0?, (e) 40?, and (f) 90?.  . .......... 153 
xii 
 
Figure 5.2. Scanning electron microscopy (SEM) images of translucent films obtained from  
a 4.4:1.0 by volume supernatant of dsDNA/SWNT dispersion (~2.0 vol % of SWNT) by 
shearing at approximately 500 s-1 shear rate.  .......................................................................... 154 
 
Figure 5.3. SEM of fibers extruded from cholesteric LC dispersion of dsDNA/SWNT  
(1.98 vol % of SWNT) in (a), (b), (d), (e) 2-propanol, (c) ethanol, and (f) tert-butanol.  ....... 157 
 
Figure 5.4. Raman Spectra at 785 nm excitation of fiber spun from cholesteric dispersion  
of dsDNA/SWNT (1.98 vol % SWNT) in ethanol.  ................................................................ 157 
 
Figure 5.5. SEM of fibers spun from noncentrifuged 6.4:1 by volume dsDNA/SWNT  
dispersions by (a) coagulating in 1-butanol and (b) its surface structure. For dry spinning  
fibers, surfaces of fibers (c) in contact with air and (d) Teflon have different structures. . ..... 159 
 
Figure 5.6. Scanning electron micrographs of cross-section structures of fibers produced  
from bulk 6.4:1 by volume dsDNA/SWNT dispersion with ~3.4 vol % SWNT by (a, b) dry 
spinning, and (c) coagulation in 1-butanl.  .............................................................................. 160 
 
Figure 5.7. SEM of (a) 7 strands and (b) 22 strands PVA coagulated LSZ/TTAB/SWNT  
fibers showing smooth structure and evidence of collapsing of multiple fibers. .................... 162 
 
Figure 6.1. Polarized light microscopy of CNC suspensions at various volume percent 
concentrations at a room temperature.  .................................................................................... 169 
 
Figure 6.2. Cross polarized microscopy images at 24, 35 and 40 ?C of a sample containing  
12.1 vol % CNC.  ..................................................................................................................... 170 
 
Figure 6.3. Transient behavior of CNC suspensions at various concentrations of biphasic,  
liquid crystal and gel regime.  .................................................................................................. 171 
 
Figure 6.4. Steady shear viscosity versus shear rate at 10, 25, 35, and 40 ?C of CNC  
suspensions with concentrations of a) 3.07 vol %, b) 6.99 vol %, and c) 12.1 vol %.  ........... 173 
 
Figure 6.5. Rheological properties versus concentration. a) Complex viscosity at 0.1 s-1,  
1.0 s-1, and steady shear viscosity at 0.1 s-1 and 1.0 s-1; b) storage modulus and c) loss  
modulus at 0.1 s-1, 1.0 s-1, 10 s-1 and 100 s-1.  .......................................................................... 175 
 
 
 
 
 
xiii 
 
 
 
 
 
List of Tables 
 
 
Table 3.1. Representative properties of SWNT used in this research.  ..................................... 66 
Table 4.1. Comparison of TGA data of dsDNA.  ...................................................................... 85 
Table 4.2. Percentage of dsDNA ash at 800 oC.  ....................................................................... 87 
Table 4.3. Consistency and rate index of steady shear and complex viscosities for dsDNA/SWNT 
supernatant dispersions behaving as power law fluids at high SWNT concentration.  ........... 117 
 
Table 4.4. Comparison of key rheological parameters of three different systems.  ................ 151 
Table 5.1. Electrical resistivity and conductivity of various SWNT fibers from both wet spun and 
dry spun assembly methods.  ................................................................................................... 164 
 
Table 5.2. Mechanical properties of various SWNT fibers from both wet spun and dry spun 
assembly methods.  .................................................................................................................. 165 
 
 
1 
 
 
 
 
 
Chapter 1 
Introduction 
Large area assemblies of single-walled carbon nanotubes (SWNT) have numerous 
potential applications including strong antimicrobial coatings,1-3 artificial muscles,4 chemical 
sensors5,6 electrical wires,7,8 hydrogen storage9 and high strength fibers.10-14 However, translating 
the remarkable mechanical, thermal, and electrical properties of individual SWNT into 
macroscopic objects is hindered by the strong van der Waals attraction between SWNT and the 
random entanglements resulting from many synthesis schemes. Lyotropic liquid crystalline 
SWNT dispersions are recognized as a key potential precursor for the fluid phase processing of 
carbon nanotubes into aligned materials with outstanding properties.10 Several research groups 
have achieved lyotropic carbon nanotube liquid crystalline phases; however, they have all been 
polydomain nematics.15-20 With the exception of SWNT in cholorosulfonic acid,20 the nematic 
phase has always had numerous defects compared to traditional rod-like polymer liquid crystals. 
While shear and other forces applied during processing can be used to anneal defects between 
domains, it has been shown that dispersions with larger domain size and fewer defects result in 
fewer defects in the assembled solid material.20 The lack of a smooth liquid crystal 
microstructure in most systems has been deemed a key impediment to producing macroscale 
materials with outstanding properties.21 Furthermore, while nematic liquid crystalline phases are 
useful for producing aligned films and fibers, the inability to produce cholesteric SWNT 
dispersions has limited the range of potential applications. An inherent property of the 
cholesteric microstructure is the selective reflection of visible light; there is growing interest in 
2 
using films produced from cholesteric phases in security papers and passive optical devices such 
as circular polarizers, notch filters, and reflective displays.22,23 The majority of cholesteric film 
research to date has focused on cellulose nanowhiskers; until now it has not been possible to 
achieve this microstructure with SWNT.   
In parallel to the ongoing advancements in bottom-up assembly of liquid crystalline 
nanocylinder (e.g. nanotubes, nanowires, nanowhiskers) dispersions,24 there has been growing 
interest in dispersing carbon nanotubes in solutions of biological molecules such as DNA and 
enzymes. These materials are some of the best known dispersants for carbon nanotubes due to 
their amphiphobic nature and potential for ????interactions;1,25,26 they also offer the advantages 
of green chemistry, potential biocompatibility and multi-functionality. Pioneering work by 
Philippe Poulin?s group at Centre de Recherch? Paul Pascal - Centre National de la Recherch? 
Scientifique (CRPP-CNRS) in Bordeaux, France demonstrated that evaporation of low 
concentration dispersions of SWNT in single-stranded DNA (ssDNA)16 and biological 
hyaluronic acid (HLA)17 resulted in liquid crystalline phase formation. In addition, Hobbie et al 
have shown that SWNT dispersed in bile salts can be assembled into fibrils as a result of the 
mesogenic nature of the solvent and geometrical confinement.27 It is noteworthy that, in addition 
to enabling carbon nanotube dispersion, many biological molecules are themselves mesogenic. 
For example, tobacco mosaic virus (TMV) was the material used in Onsager?s original work on 
liquid crystalline phase formation.28 More recently, the rod-like bacteriophage fd virus has been 
studied as a model system; this research includes fascinating work examining the phase behavior 
of dispersions containing both rods and spheres.29,30 In addition, the long and stiff helical 
structure of DNA molecules enables formation of multiple liquid crystalline phases;31,32 the 
critical concentration for cholesteric formation depends on factors such as the number of DNA 
3 
base pairs and the ionic strength.31-34 Cholesteric liquid crystalline phases of DNA are even 
evidenced to form in vivo.35-37   
Characterizing the phase behavior of lyotropic liquid crystalline dispersions of SWNT 
using rheology is in its infancy; most of the systems studied to date having been binary systems 
of carbon nanotubes in Newtonian solvents. The rheology of characterizing the liquid crystal 
SWNT dispersions is promising for understanding the flow microstructure and subsequent 
properties of macroscopic assemblies of nanomaterials. The well studied rheology of liquid 
crystalline polymers (LCPs) can be used as a guide for better understanding the complex and 
new field of rheology study on liquid crystal nanocylinders. With that being said, there is still a 
dilemma. Although the rheology of model LCPs such as hydroxypropylcellulose (HPC),38-40 poly 
(? benzyl-L-glutamate) (PBLG)41,42 or poly (? benzyl glutamate) (PBG)43,44 solutions have been 
widely studied, there are still struggling efforts to understand this field due to the lack of their 
universal rheological properties. In the case of SWNT, the flow behavior of the resulting liquid 
crystals became even more complicated due to the additional inherent 20 - 40 kBT/nm,45,46 van 
der Waals attraction between SWNT. In addition, the complex interplay between intermolecular 
interactions, molecular conformation, shear flow, and molecular orientation state all affect the  
rheological properties of LCPs solutions.40,43 The characteristic rheological properties of nematic 
LCPs that differ from ordinary polymers and low molecular weight liquid crystals (LMWLCs) 
have led to the development of new molecular theories such as those of Doi, Marrucci and 
Larson.47-49 However, the direct modeling of nematics with highly textured structures is so 
complicated that those theories adopted a simple geometry for calculation such as monodisperse 
rigid rods under an anisotropic excluded volume potential. However, potential polydispersity and 
4 
flexibility, textural defects, and various molecular interactions existing in the real LCPs systems 
lead to discrepancies between experimental results and theories.  
This work compares the rheology and phase behavior of dsDNA/SWNT dispersions 
produced by two different methods. The phase behavior and dispersion microstructure were 
affected by the dsDNA:SWNT ratio and the state of small bundles of SWNT existing in the 
dispersions. If there was not sufficient dsDNA to stabilize the dispersion, large aggregates 
formed at relatively low concentration. With sufficient dsDNA, evaporation of bulk dispersions, 
comprised of both bundled and individual SWNT, resulted in nematic phase formation. 
Evaporation of dsDNA/SWNT supernatants resulted in cholesteric phase formation. To the 
author?s knowledge, this is the first lyotropic cholesteric SWNT liquid crystal formed without 
integrating the SWNT into an existing liquid crystal matrix. Understanding the difference in 
phase behavior of nematic and cholesteric dsDNA/SWNT dispersions with concentration, 
textural effects and shear flow is important for improving the properties of their macroscopic 
assemblies. The origin of the cholesteric structure can be directly attributed to the majority of the 
dsDNA retaining its helical structure during dispersion preparation. This structure can be 
retained or eliminated during processing into aligned films. The potential enhanced 
biocompatibility of dsDNA/SWNT and the previously never achieved cholesteric microstructure 
suggest that the range of applications that can be processed from liquid crystalline nanotube 
dispersions may be even broader than previously thought.  
In addition, macroscopic assembly of fibers by solution spinning from different SWNT 
dispersions was investigated. For fiber spinning from bulk (non-centrifuged) dsDNA/SWNT 
dispersions, both wet spinning into 1-butanaol coagulant bath and dry spinning at ambient 
conditions were conducted. In addition, solution spinning from SWNT dispersion stabilized by 
5 
the enzyme lysozyme (LSZ) and cationic surfactant (1-tetradecyl) trimethyl-ammonium bromide 
(TTAB) was performed. Co-flow stream of polyvinyl alcohol (PVA) polymer coagulation bath 
was used to aide alignment. The goal of using lysozyme as a dispersant was to produce 
mechanically robust fibers with better electrical and thermal properties as well as possessing 
antibacterial properties due to the lytic activity of lysozyme against gram positive bacteria. These 
fibers are potentially useful for first responders and military clothing. In general, fibers produced 
from LSZ/TTAB/SWNT dispersions coagulated in PVA solution, showed better overall 
mechanical and electrical properties when compared to fibers obtained from dsDNA/SWNT 
dispersions either by wet or dry spinning.  
This dissertation summarizes research on the phase behavior and rheology of lyotropic 
dsDNA/SWNT and cellulose nanocrystals (CNC) aqueous dispersions and macroscopic 
assembly of film and fibers of SWNT in both dsDNA and lysozyme. Chapter 2 provides 
background information on materials used including SWNT and dsDNA, molecular interaction 
between dsDNA/SWNT, liquid crystals, phase behavior and rheology of rods dispersed in liquids, 
film coatings and fiber spinning. Chapter 3 provides information on instrumentation and 
experimental techniques. The results and discussions are divided into three chapters; Chapter 4 
provides information on phase behavior and rheology of aqueous dsDNA/SWNT dispersions, 
Chapter 5 describes film coatings from cholesteric dsDNA/SWNT liquid crystals, and fiber 
spinning of dsDNA/SWNT and LSZ/TTAB/SWNT dispersions, and Chapter 6 provides 
information on phase behavior and rheology of lyotropic sulfonated CNC suspensions. Finally, 
Chapter 7 summarizes the conclusions of this research and provides possible directions for future 
work in the field of SWNT liquid crystals.  
 
 
6 
 
 
 
 
 
Chapter 2 
Background 
The discovery of Buckminsterfullerene in 1985 by Sir Harry Kroto, Richard Smalley and 
Bob Curl, and coworkers at Rice University, which led to their being awarded the Nobel Prize in 
chemistry in 1996, has been the direct stimulus for the systematic study of carbon filaments of 
very small diameters in nano-scale size including carbon nanotubes (CNTs).50 The observation 
of multi-walled carbon nanotubes (MWNTs) by Iijima in 1991 and the subsequent identification 
of SWNT along with Bethune et al. two years later in 1993 launched the current explosion of 
interest in CNTs.51-53 Furthermore, Smalley and coworkers at Rice University successfully 
obtaining synthetic, high quality, bundles of aligned SWNT by laser vaporization of a graphite 
target in 1996 was a major breakthrough in the field which enabled providing large amount of 
high quality raw materials for scientific research.54 The subsequent studies on achieving 
synthetic nanotubes with higher yield and narrow diameter and chirality distributions have been 
actively conducted by different methods such as electric arc,55 chemical vapor deposition 
(CVD),56 and high-pressure carbon monoxide (HiPco).57 Synthesized nanotubes have finite 
lengths, and also contain impurities such as metal catalyst particles, amorphous carbons, and 
fullerenes. These impurities often impede their uniform assembly into macroscopic objects. Thus, 
purification after synthesis is often required, but this can lead to more defects in the nanotubes 
and the formation of highly packed bundles. Therefore, improved yields and properties of 
synthetic CNTs are a key need for their further application in new generation of materials.  
7 
However, the major obstacle for the current studies is turned out to be difficulty in 
obtaining dispersions of CNTs as individuals or small bundles into solvents or polymers. 
Ironically, this obstacle is due to the sp2 hybridized molecular perfection of the material which 
makes it less compatible with other materials; this is especially true for SWNT. Thus, in the past 
few years, chemical functionalization of CNTs, by both covalent interaction and non-covalent 
interaction, has become a main focus of this field. The goal is to improve the solubility of 
nanotubes and as a result translate their unique properties, to the largest extent into macroscopic 
applications. In particular, chemical modification of CNTs with biological molecules has started 
to attract significant attention ever since the early work done by Malcolm Green, Edman Tsang 
and coworkers in the mid-1990s on interacting biomolecules including enzymes and DNA 
oligomers with CNTs.58,59 This chapter provides general background information relevant to this 
dissertation including carbon nanotubes and their stabilization with biomolecules (especially 
DNA), phase behavior and rheology of isotropic rod dispersions and their liquid crytal phases, 
and macroscopic assembly into films and fibers.  
 
2.1. Carbon Nanotubes 
Carbon nanotubes (CNTs) are allotropes of carbon with the structure of seamless 
cylinders derived from rolling up graphene sheets of honeycomb lattice representing a single 
atomic layer of crystalline graphite. As synthesized CNTs have hemispherical fullerene like 
caps,60 but these are typically removed during purification due to the lower strain in the C-C 
bond. The bonding in CNTs is primarily sp2 hybridized carbon atoms covalently bonded to each 
other, although there may be some sp3 -bonded atoms at defect sites and in the regions of high 
curvature.61 CNTs can be classified as single-walled carbon nanotubes (SWNT), double-walled 
8 
carbon nanotubes (DWNT) or multi-walled carbon nanotubes (MWNT) according to their 
structure. Typically SWNT are single-shell cylinders with typical diameters of ~1 nm and 
lengths that typically range from 0.1 to 10 ?m, while DWNT and MWNT are nested 
arrangement of SWNT in two or several coaxial cylinders. The typical diameter of MWNTs is 
~30 nm and they usually have longer lengths than SWNT with a range of 0.5 to 50 ?m. The 
commercial availability of DWNT is not as abundant as SWNT and MWNT. The typical 
diameter of DWNT is ~4 nm with lengths smaller than 20 ?m. The simplest way of specifying 
the structure of an individual tube is in terms of its 1D unit cell, defined by the vectors Ch and T, 
where T is the translational vector of 1D unit cell and normal to Ch (Figure 2.1). The length of 
the chiral vector Ch = na1 +ma2 is used to define the circumference of any carbon nanotube, 
where a1 and a2 are the unit cell base vectors of the graphene sheet.62  
 
 
 
9 
 
Figure 2.1. Construction of carbon nanotube from single graphene sheet.63  
 
Each pair of integers (n, m) with 0 ? |m| ? n represents a possible nanotube structure. It 
can be seen from Figure 2.1 that, m = 0 denotes zigzag nanotubes, and the n = m denotes 
armchair nanotubes. All other vectors (n, m) correspond to chiral nanotubes. The chiral angle ? 
between the tube axis T and H, which indicates the direction of the closest hexagon rows to the 
tube axis, can also be used to describe three distinct types of nanotube structures. The zigzag and 
armchair nanotubes have chiral angles of 30o and 0o, respectively, while chiral nanotubes 
corresponds to chiral angles of 0o < ? < 30o. Figure 2.2 provides a description of the structure of 
SWNT with different chiralities.  
 
10 
  
Figure 2.2. Schematic theoretical model for a SWNT with the tube axis normal to: (a) the ? = 30o 
direction with (n,m) = (5,5) (an ?armchair? tube), (b) the ? = 0o direction with (n,m) = (9,0) (a 
?zigzag? tube), and (c) a general direction 0o < ? < 30o with (n,m) = (10,5) (a ?chiral? tube). 
Reproduced from Dresselhaus et al.64 
 
The angle at which a graphene sheet is rolled up determines not only the chirality, but 
also determines whether the nanotubes behave as a metal or semiconductor. Armchair nanotubes 
(n = m) are ballistic conductors, while other chiralities are considered to be metallic if (m ? n) is 
a multiple of 3, which corresponds to a very small bandgap.65 Other nanotubes are 
semiconducting when (m ? n) is not a multiple of 3. The nanotube diameter dt is given by  
?? nmmnaCd ccht ?????? ? 223                      (2.1) 
11 
Where      is the circumference length, and ac-c = 1.42 ? is the carbon-carbon (C-C) bond 
length. Furthermore, the molecular weight of SWNT with a given length Lt, and diameter dt, can 
be calculated from the definition of the 1D unit cell. The length of T is given as: 
                                        (2.2) 
where L is the circumference of nanotube,       , and dR is the highest common divisor of 
(2n+m, 2m+n). In addition, the number of carbon atoms per 1D unit cell is: 
                                         (2.3) 
For an individual tube of length Lt, the total number of carbon atoms N and further the 
molecular weight MW of the tube can be derived from the following equations: 
                                        (2.4) 
                   
    
                     (2.5) 
As a result, the molecular weight of a typical SWNT with 1nm in diameter and 1?m in 
length is 1.40?106 g/mol. 
As mentioned above, one of the surprising characteristics of CNTs is their electronic 
conductivity which can be either metallic, like copper, or semiconducting, like silicon. The 
ballistic conductance of armchair nanotubes can be as high as 1 ???cm at 25 oC.65 It is well 
established that, the experimental thermal conductivity ? of SWNT is more than 3000 W/ m?K 
which is equal or higher than that of diamond with ? value of 3320 W/ m?K.66-68 In addition, the 
thermal conductivity of as high as 6600 W/ m?K at room temperature was obtained for a (10,10) 
12 
SWNT by molecular dynamics simulation.69 Since the carbon-carbon bond in the graphene wall 
of nanotubes is one of the strongest in nature, CNTs are considered to be the stiffest and 
strongest materials ever occurring in nature or being synthetically produced. The tensile modulus 
of SWNT can be over 600 GPa, while tensile strength can be over 37 GPa, approximately 60 
times the strength of steel, while six times lighter than steel.65,70,71 These remarkable properties 
of nanotubes together with their low density ~1.4 g/cm3 provide these relatively new types of 
nanomaterials a huge potential in a wide range of applications.  
 
2.2. Deoxyribonucleic Acid  
The genetic molecule deoxyribonucleic acid (DNA), nature?s most important building 
block, was first described by James Watson and Francis Crick in 1953 and formed double helical 
structure. Their double helical model was based on X-ray diffraction data of DNA fibers which 
were provided by two different groups, one was Maurice Wilkins and colleagues and the other 
was Rosalind Franklin and Raymond Gosling. A representation of the model of Watson-Crick 
double helix is shown in Figure 2.3a. Understanding the double helical structure of DNA has 
been the great achievement in realizing the importance and significance of DNA as the genetic 
material. The repeating units of DNA are called nucleotides which are composed of nitrogen 
containing aromatic bases, which are purine or pyrimidine rings, five-carbon ribose sugars, and a 
backbone of phosphate groups. The molecular weight of double-stranded DNA (dsDNA) can be 
approximated as MW (dsDNA) = (number of nucleotides ? 607.4) + 157.9. Thus, the molecular 
weight of a dsDNA molecule with 750 base pairs is about 9.11?105 g/mol which is the 
approximate dsDNA length used in this work after sample preparation. The initial length of the 
native dsDNA used is more than 10,000 base pairs. The phosphate group of a nucleotide is 
13 
formed from a central phosphorous atom covalently bonded to 4 oxygen atoms. The 
electronegative oxygen atoms of phosphate groups are responsible for stabilizing DNA through 
ionic interactions with cations in the cell. The sugar-phosphate backbones helically rotate around 
the outer surface of DNA moving in opposite directions and form two different sized grooves of 
major and minor grooves. The major groove is approximately 50% wider than the minor. DNA 
helical structure is mainly stabilized by the base stacking of A-T (Adenine-Thymine) and G-C 
(Guanine-Cytosine) base pairs that promotes hydrophobic interactions between the non-polar 
portions of the bases (Figure 2.3c). The hydrogen bonding between complementary bases is 
another force contributor for holding the DNA double helix together with two hydrogen bonds 
between A-T base pairs and three hydrogen bonds between G-C base pairs (Figure 2.3b). 
Typically, the hydrogen bonds in DNA are 2-3 kcal/mol which is weaker than most hydrogen 
bonds (3-7 kcal/mol) due to geometric constraints within the double helix.72  
 
 
 
 
 
 
 
 
 
 
 
14 
 
 
Figure 2.3. Watson-Crick DNA double helix and base pairs.73 
 
DNA molecules can adopt different conformations depending on the environmental 
factors such as salt concentration and presence of alcohol. The main types of conformations are 
B-DNA, A-DNA, and Z-DNA. B-DNA is believed to the native conformation occurring in vivo. 
It has right-handed double helical structure of 10 base pairs per turn with well defined major and 
minor grooves and is found at low salt concentration. In fact, the DNA double helix structure 
discovered by Watson and Crick is known as the B form. The typical diameter of DNA helix is 
20 ? with small deviations according to the conformational changes. A-DNA is quite rare and is 
found in solutions with higher salt concentrations or with alcohol added. A-DNA has 
righ-handed helix very similar to B-DNA, but it has 11 base pairs per turn and a diameter of 26 
15 
?. As for Z-DNA, it has left-handed helix and occurs for DNA molecules with alternating poly 
(dG-dC) sequences in solutions with high salt concentrations or alcohol existence. Z-DNA has 
more compact cylinder-like structure and has 12 base pairs per turn with a diameter of 18 ?. The 
double-helical structure of DNA is remarkably stable with a relatively high melting temperature 
Tm. It is reported that the Tm of the DNA from salmon testes used in this research is 87.5oC in 
0.15M sodium choloride plus 0.015M sodium citrate solution.74 Melting or denaturation of DNA 
occurs when the external forces overcome the stabilizing forces of hydrogen bonding and 
base-stacking. Denaturation is usually accomplished by increasing the solution temperature or 
alkali treatment due to ionization of the base.74 Acid treatment is not an option because strong 
acid will degrade the DNA. However, under mild conditions like slow cooling, the two single 
strands of denatured DNA can find their complementary stands and base pairs with them to from 
hydrogen bonded double helix. In this work, DNA has been considered as a biopolymer, a 
specially sequenced copolymer that has been selected by evolution to have a particular property. 
One difference between random copolymers and DNA is that the former one can produce an 
ensemble of all possible arrangements of the repeating units.75  
 
2.3. Lysozyme 
In this research, lysozyme (LSZ) was also used as a dispersing agent for SWNT in 
producing robust and electrically conductive fibers by solution spinning. These fibers are 
potentially useful for first responders and military clothing. Lysozyme (LSZ) is a relatively small, 
globular enzyme catalyzing chemical reactions and can be found in many biological systems 
such as egg white, human tears, saliva and other secretions. It provides efficient antibacterial 
activity against gram positive bacteria by breaking down components of the cell wall 
16 
peptidoglycan of bacteria. Lysozyme is comprised of a single polypeptide chain of 129 amino 
acid residues with the approximate size of 30 ? in diameter and a molecular weight of 
approximately 14,000 g/mol (Figure 2.4).76,77 The polyampholytic nature of lysozyme, which 
contains multiple hydrophilic domains and more than 35 % of hydrophobic residues,78 makes it a 
competitive dispersant for hydrophilic or hydrophobic molecules. In its native state, LSZ 
contains 42% ?-helices and 6% ?-sheets of protein secondary structure which are folded within 
its polypeptide chain.78 Furthermore, the existence of four disulfide bridges formed between the 
thiol groups of cysteine residues provides crosslinks which prevent complete loss of tertiary 
globular structure.79 Due to the disulfide bridges, LSZ is relatively stable having the isoelectric 
point of 11 and a relatively high denaturation temperature of 76oC.80 When dispersing SWNT in 
LSZ solution, partial denaturation of LSZ occurs during sonication and leads to the exposure of 
the native internal hydrophobic domains to the surface which can further interact with the 
side-walls of SWNT through hydrophobic interactions. The outside hydrophilic groups of LSZ 
help stabilizing the LSZ/SWNT hybrids in aqueous dispersion through electrostatic forces.78 
 
17 
 
Figure 2.4. Ball-and-stick representation of lysozyme structure highlighting the disulfide bridges 
at the core. All protein atoms are shown as balls, and bonds between atoms are shown as a stick. 
Carbon, nitrogen, oxygen and sulfur atoms are colored grey, blue, red and yellow, respectively. 
Adapted from Szymanska et al.81 
 
2.4. Cellulose Nanocrystals 
The aim of studying the phase behavior of cellulose nanocrystals (CNC) aqueous 
suspensions in this work is to answer the following question; ?how does the simpler 
cholesterogenic aqueous cellulose nanorod systems compared to the complex ternary system 
(DNA, SWNT, water)? ? Cholesteric phase forming system, CNC in water, may give some 
insights on the rheology of cholesteric system since it avoids the complication of the DNA phase 
which is not truly rigid rods. Also, the comparatively simple binary system of CNC suspensions 
(CNC and water) is expected to provide better understanding in rheological behavior of 
cholesteric liquid crystals. However, the phase behavior and rheology of cholesteric liquid 
18 
crystal forming dsDNA/SWNT dispersions are the main focus of this work and will be discussed 
in detail in the following sections.  
Cellulose nanocrystals are considered to be the most abundant naturally occurring 
nanorods which can be produced from plants (e.g. cotton, ramie, cellulose), trees, bacteria and 
animals (tunicate) via the condensation polymerization of glucose. They have great potential to 
be used as fillers in renewable polymer matrices to create mechanically strong and optically 
smart nanocomposites and are also attractive material, because of their environmentally friendly 
nature. CNC are commonly isolated from their raw materials by acid hydrolysis process. Most 
typically, sulfuric acid is used resulting in the sulfonated CNC structure shown in Figure 2.5.  
 
 
 
Figure 2.5. Chemical structure of sulfonated cellulose after acid hydrolysis. 
 
 
19 
The properties of cellulose suspensions depend on both the material origin and the 
hydrolysis conditions.82,83 The lateral dimensions of cellulose nanocyrstals range from 3-30 nm84 
have various aspect ratios L/d (L being the length and d the diameter) from almost a particle L/d 
=1 to about 100 can be obtained by acid hydrolysis of different sources based cellulose.85 It has 
been observed that aqueous suspensions of sulfonated CNC can go through dilute, biphasic, and 
single phase liquid crystal regime with increasing concentration of dispersed rods in lyotropic 
systems. Revol et al.86 identified the type of liquid crystalline phase formed from sulfonatd 
aqueous CNC suspensions as chiral nematic (cholesteric). This particular ordered structure gives 
rise to many of the optical and mechanical properties of their resulting films. The crystal 
modulus of cellulose was first experimentally determined by Sakurada et al.87 using X-ray 
diffraction and gave a value of 138 GPa for cellulose-I. This value lies within the range of 
theoretical estimates of 100-160 GPa using molecular dynamics/mechanics methods.82,88 The 
tensile strength of cellulose crystals should be in the order of 10 GPa.85 Thus, CNC is considered 
as a possible alternative material of carbon nanotubes in terms of their relatively high mechanical 
properties and aspect ratios. However, much more research is needed on this new popular 
material. The optical properties of CNC films obtained from its liquid crystalline phase can be 
controlled by tuning its cholesteric pitch. In general, stronger interparticle interactions result in 
smaller pitch. With increasing suspension concentration across the biphasic regime, the pitch 
decreases from ~80 to 10 ?m.83 In addition, Pan et al.23 showed that the cholesteric pitch of 
cellulose films even fit the equation describing the correlation of concentration and pitch, 
established for chiral nematic polymer solutions. The effect of concentration on pitch is given in 
the following equation,23,89 
                                           (2.6) 
20 
where P is pitch, c is concentration, A is a constant dependant on the solvent, temperature, and 
the molecular weight of the solute polymer and k is solvent-dependant with typical value less 
than 2. Moreover, adjusting the ionic strength of suspensions by adding electrolytes could 
control the pitch and further control the wavelength of reflected light through subsequent 
films.83,85,90 Typically, with increasing concentration of electrolytes such as HCL, NaCl, and KCl, 
the cholesteric pitch both in the cellulose suspensions83 and films23 decreases due to reduction in 
pre-existing Coulombic repulsion between acid hydrolyzed CNC rods.22 The cholesteric pitch 
can also be affected by some other factors such as presence of magnetic field and its exposure 
time, and temperature.23 As a result, these properties provide promising potential application in 
security papers and passive optical devices such as circular polarizers, notch filters, and 
reflective displays.22,23 Furthermore, the cheap and abundant nature of CNC makes it a promising 
material for a wide range of future applications such as liquid crystal displays, and even can be 
used as a template of generating other nanoparticles to obtain products with improved properties. 
The ready dispersion of CNC in various solvents, even water, provides the possibility of 
obtaining liquid crystals and further applications into films with improved optical and 
mechanical properties. The presence of charge such as introduction of ?SO3H groups on their 
surface during the hydrolysis step promotes dispersibility in aqueous solvents due to electric 
repulsion. Hydrogen bonds created by dispersing CNC into aqueous solvents is another factor 
promoting the dispersibility as well as providing the basis of its mechanical strength.85,91 
However, the effect of pH due to the surface charge, on the phase behavior of CNC aqueous 
suspensions needs to be considered carefully too.  
21 
2.5. Biopolymer Stabilization of Carbon Nanotubes 
Interest in the biological functionalization of nanotubes has exploded in the last decade 
ever since the pioneering work of Malcolm Green, Edman Tsang, and coworkers58,59 who 
immobilized enzymes and DNA oligomers on the nanotube surfaces in the mid 1990s. In 
addition, biologically modified MWNT have been used as nanoprobes for high resolution 
imaging such as tips for AFM.92 Biomolecules, such as proteins and nucleic acids, can be bound 
to CNTs either by covalent or by non-covalent functionalization. The exciting part of this new 
emerging technology is that, it can combine nature?s toolbox with the unique properties of 
nanostructures,93 introducing molecular recognition and bio-directed self-assembly capabilities 
together with potential biocompatible characteristics to nanomaterials. It has been reported that 
SWNT can be readily coated with the protein, streptavidin, and be solubilized in it by 
sonication.94,95 Integration of enzymes to nanotubes was also studied broadly to render the 
nanosystems antimicrobial enzymatic activity as well as strong mechanical properties which are 
provide by the CNTs.1 While much of the biological functionalization of CNTs have been aimed 
at biosensor development, a large number of studies have been also aimed at finding possible 
therapeutic applications, such as targeted drug delivery, cancer treatment, artificial muscles and 
?feeling? artificial skin.4,96-99 Among all the biomolecules used to modify nanotubes, DNA and 
other nucleic acids are of great potential interest. At the basic level, nucleic acids could provide 
good solubilization of nanotubes enabling higher quality materials with more uniform properties. 
Nucleic acid functionalized CNTs might have great potential in biosensor applications due to 
their remarkable recognition abilities such as the perfect, well studied complementary base 
pairing mechanism of DNA and binding of proteins with specific mechanisms. Furthermore, 
DNA-guided macroscopic assembly of CNTs system is possible due to the DNA?s ability to 
22 
form various liquid crystalline phases at different conditions. Also, the prospect of nanotubes 
serving as strong carriers to deliver therapeutic DNA into cells is perhaps one of the most 
exciting future applications of biologically functionalized CNTs.100,101  
Although integration of biopolymers onto CNTs would improve the stability and 
biocompatibility of the system, the potential toxicity studies of nanotubes, which is still at an 
early research stage, have aroused concerns about potential safety issues. Several studies 
reported that direct injection of SWNT into the lungs of mice would cause inflammation of the 
lung tissue, or even lead to death.102,103 However, the authors pointed out that mechanical 
blockage of the large airways by aggregates of nanotubes might be the reason for causing death, 
rather than the toxicity of nanotubes.104 In order to provide a correct assessment of nanotubes? 
toxicity, inhalation toxicity studies rather than direct injection, and better initial material 
characterization of the nanotubes are needed, because synthetic nanotubes always have catalytic 
and amorphous carbon impurities. The safety and risk management of engineered carbon-based 
nanomaterials (CNM) has become increasingly demanding. The number of toxicology studies on 
carbon nanotubes is growing significantly in each year. Preliminary investigations conducted by 
several U.S. government research centers showed that engineered CNM including fullerenes and 
multi-walled carbon nanotubes can become airborne during handling and mixing in solution by 
sonication in a laboratory process.105 This indicated that laboratory workers may be at increased 
risk of exposure to engineered nanomaterials. The UK Health and Safety Executive (HSE) 
published an safety information sheet on the Risk Management of Carbon Nanotubes in 2009. It 
provides suggestions on precautionary approaches to the risk management of all CNTs including 
understanding the health and safety information of the materials before use, using protective 
equipment during handling, and appropriate disposal of waste carbon nanotube materials. A 
23 
regularly updated resource for the latest information is the website of Rice University-based 
International Council on Nanotechnology (ICON). It includes links to the Centers for Disease 
Control and Prevention?s (CDC) Public Health Grand Rounds Event ?Preventing Adverse Health 
Effects from Nanotechnology,? a safe handling guide, and excerpts from the latest toxicology 
and environmental impact studies. 
 
2.6. Interaction of DNA with Single-Walled Carbon Nanotubes 
The intriguing results of being able to obtain good dispersions of CNTs in aqueous 
solutions of both natural DNA106,107 and short, custom-synthezied oligonucleotides108,109 
necessitate the discussion of how the DNA molecules interact with CNTs on a molecular level. 
However, the molecular-level information on the thermodynamics and structure of DNA/SWNT 
hybrids is still not fully resolved.110 In fact, different types of molecular interactions between 
DNA and CNTs have been proposed based on experiments or molecular simulations, with most 
of them using SWNT. It is generally agreed that the main driving force of DNA/SWNT 
interaction is the ?-? stacking between the plane of aromatic nucleotide bases and the nanotube 
side walls which are oriented parallel to each other.97,107-109,111,112 When SWNT are stabilized by 
single-stranded DNA (ssDNA) in an aqueous environment, the strong adsorption of DNA onto 
SNWTs will lead to the wrapping mechanism of ssDNA around SWNT.109,112 The ionized 
phosphate backbone of the DNA renders hydrophilicity to the ssDNA/SWNT hybrids and 
prevents coated SWNT from aggregating by electrostatic repulsion. However, if there are 
complementary ssDNA molecules exist in the dispersion, SWNT precipitation could occur due 
to the formation of hydrogen bonds in the corresponding double-stranded DNA (dsDNA) 
product.111 Similarly, competing molecules with extended aromatic rings may also cause SWNT 
24 
precipitation through the formation of attractive ?-? stacking with SWNT walls by displacing the 
initial ssDNA coating from the SWNT.111 Depending on the application, this SWNT 
precipitation will be useful for either purification of sorted nanotubes,113,114 or releasing the gene 
for drug delivery.99,115 Besides stabilizing SWNT in aqueous environment, ssDNA with specific 
sequences can selectively recognize different chiralities of SWNT structure which provides the 
possibility of purifying synthetic mixtures of SWNT to yield single-chirality species with the 
same electronic type.109 In addition to ?-stacking, the physical interactions between DNA 
molecules and SWNT also involve van der Waals attractive forces, hydrophobic interaction 
forces, and electrostatic attraction between the DNA backbone and the charged SWNT.116,117 
Molecular dynamics simulations of short ssDNA segment with eight adenine bases even shows 
the spontaneous insertion of ssDNA into SWNT with appropriate diameter. The encapsulation of 
oligonucleotide into SWNT is due to the combined action of van der Waals attraction and 
hydrophobic interaction of exposed ssDNA bases to the interior of the SWNT wall.116 The 
tube-size is also one of the influencing factors of the insertion process of the oligonucleotide 
where the critical diameter of the SWNT is suggested to be 1.08 nm. The encapsulation of 
ssDNA into SWNT from solution is further observed by experimentally using a modified 
electrophoresis method.118  
Compared to the interaction of ssDNA with SWNT, there are fewer studies on the 
molecular interaction between dsDNA and SWNT. However, conflicting results make it hard to 
model real systems which involve many experimental variables. One type of interaction 
mechanism proposed for dsDNA/SWNT system is that, an array of SWNT with appropriate 
diameter can actually fit into the major grooves of the long helical dsDNA molecule by 
calculating the electronic structure of dsDNA/SWNT system in vacuum.119 Another molecular 
25 
simulation study used short dsDNA segment of 12 base pairs and SWNT with similar lengths as 
of dsDNA. It shows that the dsDNA molecule absorbed onto the SWNT wall by either 
hydrophobic or electrostatic interaction depending on whether the SWNT wall was charged or 
not.117 More specifically, for a B-DNA structure and uncharged SWNT system, only one end of 
the dsDNA segment binds to the SWNT surface through the hydrophobic driving force caused 
by the hydrophobic nature of the exposed dsDNA end base pair planes and the SWNT wall 
surface. However, in the case of A-DNA, both ends of the dsDNA molecule are able to adsorb 
onto the SWNT wall without inducing high bending energy. As for the positively charged 
SWNT, both forms of dsDNA (A-DNA and B-DNA) molecules interact with SWNT in the same 
way; the negatively charged dsDNA backbone binds to the nanotube wall due to the strong 
Coulombic interactions. In addition, the charged SWNT may promote the transition of DNA 
from the A to B form upon interaction with the DNA backbone. Different interaction 
mechanisms of long natural dsDNA (~10000 base pairs before sample preparation) with random 
sequences and SWNT has also been investigated. The initial dsDNA/SWNT dispersions 
involved both small bundles and individual SWNT.107 It was suggested that spontaneous 
unzipping of dsDNA on the SWNT surface would occur followed by an ordered helical 
wrapping of the SWNT surface during a long time period of around 35 days. This slow wrapping 
mechanism is considered to be affected by several factors like the length of the DNA sequence 
and the diameter of nanotubes. The authors believe that natural DNA bound on the SWNT 
surface turns into ssDNA and wraps around a nanotube via base-nanotube stacking 
interactions.113  
 
26 
2.7. Phase Behavior of Biopolymer-Stabilized Carbon Nanotube Dispersions 
Liquid crystalline (LC) phase behavior of CNTs dispersed in biological solutions has 
been studied in recent years. Liquid crystallinity is an alternative way to help align nanotubes by 
transferring the ordered, self-organizing characteristic of the LC media into the system. 
Fundamentally, a molecule is said to be mesogenic if it is able to form LC phases,120 and a liquid 
crystal is a mesophase that has the fluidity of a liquid and the order of a crystal. There are two 
types of liquid crystals: lyotropes and thermotropes. The phase transitions of lyotropic liquid 
crystals are driven primarily by changes in concentrations. While for the other system, 
thermotropic liquid crystals, the LC transitions occur with temperature changes only. A large 
amount of biological materials, such as proteins, carbohydrates, lipids and nucleic acids, have 
been found to form various LC phases in vivo as well as in vitro under well-defined 
conditions.121 The general idea is that well dispersed CNTs in small bundles or even in 
individuals can be obtained by using good dispersing agents of biopolymers at relatively low 
concentrations. Further, single LC phases of CNTs dispersions in biopolymers could be 
successfully obtained by increasing concentrations of both nanotubes and biomolecules by 
solvent evaporation16 or spontaneous phase separation.17,122 According to Onsager (1949), Flory 
(1956), and Doi?Edwards (1986) theories, before forming single phase liquid crystals, the phase 
behavior of fluid dispersed rods goes through different regimes with increasing concentration: 
dilute, semidilute, isotropic concentrated, and biphasic. In general, the Brownian motion of 
rotation and translation of rods is inhibited as the concentration increases. Rotation is inhibited in 
the semidilute regime and both rotation and translation are inhibited in the isotropic concentrated 
regime.47 At a critical concentration of I?  the system enters the biphasic regime: phases of 
randomly dispersed rods and oriented rods coexist. The fraction of the liquid crystalline phase 
27 
increases with concentration until the system turns into single LC phase at the critical 
concentration of LC? . The graphical description of these phase transitions are illustrated in 
Figure 2.6. The critical concentrations at which the ordered phase first appears and at which the 
fully LC single phase are achieved depend on both the solvent quality and aspect ratio 
distribution of rods. The solvent quality is a measure of the energetic interactions between rods 
and solvent.123  
 
 
Figure 2.6. Concentration regimes for rod-like molecules. Adapted from Davis et al 2004,19 and 
Doi and Edwards 1986.47  
   
 
 
 
 
 
 
28 
The formation of liquid crystals can be evidenced by several methods including imaging 
technology to directly observe the aligned morphology and examining the bulk properties by 
rheological study. Davis et al20 reported the only complete experimental phase diagram of 
SWNT in superacids. In general, there are few phase diagrams for nanocylinder liquid crystals. 
The center panel of Figure 2.7 shows the first phase diagram of SWNT dispersion in superacids, 
presenting the effects of both solvent quality and concentration.20 The solvent quality is 
measured by SWNT protonation. Traversing the phase diagram from left to right (low 
concentration to high concentration of rods, constant solvent quality) results in a transition from 
an isotropic (I) phase to a biphasic system in which the isotropic phase is in equilibrium with a 
liquid crystalline (LC) phase, and then to a single LC phase. 
 
 
 
 
 
29 
 
 
30 
Figure 2.7. Phase diagram of single-walled carbon nanotubes (SWNT) in superacids and 
cross-polarized light micrographs showing the effect of solvent quality and SWNT concentration 
on microstructure. Phases include isotropic (I), liquid-crystalline (LC), crystal solvate (CS) and 
solid (S). Solvent quality is quantified by fractional charge per carbon, measured by the shift dG 
of the Raman G peak of SWNT (514 nm laser). Black symbols denote experimental results and 
red symbols refer to model predictions.20 Circles designate the isotropic concentration i?  from 
experiment (filled circles) and model (open circles). Black and red diamonds indicate the initial 
system concentration before phase separation. Open red triangles indicate i?  and n?  in a 
system of monodisperse hard-rods (Onsager). Red lines are the model predictions of the isotropic 
and liquid-crystalline stability limits, that is, cloud curves. Black dotted lines connect the 
experimental data points to denote experimental phase boundaries such as the LC/LC+S vertical 
boundary, the LC+S/CS+S horizontal boundary, and the I+LC/I+CS diagonal boundary. An 
LC+CS regime may also exist within the shaded region, as suggested by Papkov124. a?d, Light 
microscopy shows morphology just inside the biphasic region at 1.21 vol % in ClSO3H (a), 
0.132 vol % in 120% H2SO4 (c), and in the liquid-crystalline phase at 12.1 vol % in ClSO3H (b) 
and 10.6 vol % in 120% H2SO4 (d). Liquid-crystalline domains are larger, defect density is lower, 
and domain orientation changes more gradually in ClSO3H (b) than in 120% H2SO4 (d). Scale 
bars: 50 ?m (a) and 20?m (b?d). 
 
 
 
 
31 
2.7.1. Onsager versus Flory Lattice Theory 
Onsager (1949) and Flory (1956) theories are considered to be historically important for 
expressing the phase transitions of molecules from an isotropic to a nematic liquid crystalline 
state. Both theories are built upon the basic idea of excluded volume interactions between hard 
rod-like molecules in a given volume. Rod-like molecules cannot interpenetrate and the excluded 
volume of adjacent rods would exclude the center of mass of any other rod from this particular 
region as shown in Figure 2.8. The excluded volume of two spherocylinders with particular 
orientations can be calculated from the following equation: 
                ?    ?       ?                    (2.7) 
where L is the length, d is the diameter, and the ? is the relative angle between two 
spherocylinders. For rods with high aspect ratios, the above equation is approximated as 
                   . Therefore, the excluded volume of two rods is proportional to the angle 
between their molecular long axes with a minimum value for parallel rods and the largest value 
for perpendicularly orientated rods. As a result, the formation of liquid crystalline phase with 
reduced free volume is caused by the domination of increasing translational entropy over 
lowering orientational entropy due to parallel alignments of rods.48,120 
For nematic liquid crystals, the parameter describing the degree of molecular orientation 
is given by the following equation. The orientational order parameter S equals to: 
            -                               (2.8)  
where ? is the angle between a rod and the nematic local director n, and the brackets ?   ?, denote 
an average over all rods. For a completely isotropic orientation of molecules S = 0, while for a 
32 
perfect parallel orientation of rod-like molecules S= 1. The orientational order parameter for 
liquid crystals will be always smaller than the value of unity. 
 
 
Figure 2.8. A diagrammatic representation of the excluded volume of two rod-like molecules, of 
length L and diameter d. ? is the angle between the molecular long axes. Adapted from Donald et 
al. 2006.120 
 
Onsager Theory. Onsager was the first to propose a theory in 1949 using tobacco 
mosaic virus (TMV) to describe the rod orientation distribution during the phase transition from 
isotropic to nematic liquid crystalline phase. Onsager?s approach has considered an ideal solution 
of monodisperse, long, hard rods interacting only by excluded volume forces in an athermal 
solvent and truncated the interaction free energy of rigid rods at the second viral coefficient. As a 
result, the phase boundaries of biphasic chimney obtained from Onsager theory are: 
?                                        (2.9) 
?                                      (2.10) 
33 
The order parameter at nematic phase boundary is S = 0.848. Onsager theory is most 
accurate for an ideal solution of rods with aspect ratio higher than 100,125 and at dilute 
concentrations where only pairwise excluded volume interactions are significant.48 However, in 
real systems, the biphasic region widens due to many factors including the polydispersity, the 
flexibility of rod-like molecules, and the existence of other types of attractive forces such as 
depletion interaction, sedimentation forces, and the balance of short-range electrostatic repulsion 
and long-range van der Waals attraction which is especially important for nanorods.120,126 Later 
adaptations to Onsager theory including attractive interactions127 and polydispersity126 enable it 
to be used for a broader range of real systems. Although Onsager theory has been less popular 
for rod-like polymers, it is gaining increasing application for nanocylinder systems, in addition to 
rigid rod-like viruses such as TMV, M13, and bacteriophage fd viruses.  
Flory Theory. In 1956, Flory?s lattice-based theory was proposed for predicting the 
nematic phase transition. Flory theory is better known because it works well for rod-like polymer 
liquid crystals. Flory theory can be applied for densely packed systems and predict the formation 
of nematic phase in a solvent-free melt when the aspect ratio L/d is greater than 6.4.128 
According to the Flory theory, rod-like molecules are divided into many fixed lattice cells and 
the partition function and free energy are obtained by calculating the packing effects of 
segmented rods. The critical concentration of rod-like particles in an athermal solvent are: 
?    
  
     
  
                                                             
?     
  
                                                                                
34 
Extension of the Flory lattice framework incorporated the effects of the Flory-Huggins 
interaction parameter ? and molecular flexibility which gave it having great popularity in 
modeling polymeric liquid crystals. The parameter ? is an inverse function of temperature with 
negative values of ? corresponding to a good solvent in which the reduction of free energy would 
occur due to the enthalpy decrease. However, the rod-solvent interactions introduced in Flory 
theory are only applicable at short ranges and indicate the compatibility between rod and 
solvent.126 The biphasic regions broadens with increase in positive values of ? (for less favorable 
solvents) and for increasing polydispersity or flexibility.129-131 When ? is positive, there is a 
tendency toward chemically driven segregation due to the insufficient entropy of mixing 
(T?Smix).  
 
2.7.2. Friedelian Classes of Liquid Crystals  
Friedel classified liquid crystals (LCs) into nematic, cholesteric, and smectic in 1922. 
Nematic LC phases are typically characterized by an orientational ordering of molecules which 
do not have positional order and the alignment of rod-like molecules is relative to the vector, 
known as the director n, as illustrated in Figure 2.9a. The ?threaded? texture observed under 
transmitted light microscope without polarizers and Schlieren texture with polarized light 
microscopy are characteristic optical properties of nematic phases. As for the cholesteric phase, 
locally the molecules are arranged in a nematic phase, but they follow a helical path about an 
axis perpendicular the director due to the chiral character of the constituting molecules as shown 
in Figure 2.9b. Fingerprint textures consisting of near-parallel sets of swirling lines are 
considered as the typical optical texture of cholesteric phase by polarized light microscopy study. 
Chirality could be a prominent factor on determining the liquid-crystalline phase behavior and 
35 
the material properties.132 However, the relationship between macroscopic cholesteric structure 
and the microscopic chiral properties of constituting molecules has not been well understood. 
Both long range orientational and positional order in smectic phase result in a layered structure 
(Figure 2.9c) and has a characteristic optical texture called focal-conic fan texture.  
 
 
Figure 2.9. Representations of molecular organizations in the three Friedelian classes of liquid 
crystals: (a) nematic; (b) cholesteric; and (c) smectic.120 
 
2.7.3. Liquid Crystallinity of Biopolymer-Stabilized Nanotube Dispersions 
Although the first carbon nanotube liquid crystal phases were not discovered less than ten 
years ago,15,16,19 there have already been several studies of biopolymer ? SWNT liquid crystals. 
Moulton et al.17 reported nematic LC phase formation for SWNT dispersed in biological 
36 
hyaluronic acid (HLA) solutions. HLA is one of the chief components of the extracellular matrix 
and contributes significantly to cell proliferation and migration. The dispersion of SWNT was 
obtained by sonication assistance and at a certain concentration range of both SWNT and HLA, 
and evolution of nematic LC phase occurred spontaneously over time in tightly sealed vials with 
phase separation occurring throughout the dispersions. The highest concentration of SWNT 
forming nematic LC was 0.5 wt %, and the HLA concentration was chosen as 0.5 wt % or 1.0 
wt %. The nematic phase was confirmed by strong birefringence under polarized light 
microscopy. When rotating the samples between cross polars, and the different domains turned 
bright and dark based on the relative alignment between their director and the polarization vector 
of the light.  
Highly concentrated suspensions of SWNT in gellan gum (a water-soluble 
polysaccharide which is used primarily as a gelling agent) were used to form LC phases by 
another research group.133 Gellan gum has been shown to a good dispersing agent for stabilizing 
SWNT in aqueous solution by non-covalent polymer wrapping. The authors could not detect a 
LC phase for the highly concentrated raw SWNT suspensions, which they assumed was due to 
the coexistence of large amount of spherical carbonaceous impurities. However, after a 
purification procedure, they observed large-scale bireflection from both the highly concentrated 
purified SWNT suspensions and their dried solid composite membranes using polarized light 
microscopy. Large domains of aligned structures in the dried membranes was further observed 
using scanning electron microscopy (SEM). The actual concentration of above mentioned highly 
concentrated purified SWNT suspension was 46 mg/ml of SWNT, which is about 3.2 vol % 
considering the density of SWNT is 1.45 g/cm3 and the mass ratio was in the range of 10 ~ 20 
mg gellan gum to 50 mg SWNT. In addition, Hobbie et al. have shown that SWNT dispersed in 
37 
bile salts can be assembled into fibrils as a result of the mesogenic nature of the solvent and 
geometrical confinement.134 Bile salts, which are produced in the liver, play a critical role in 
digestion by solubilizing the fat through the formation of mixed-lipid micelles.135 
 
2.7.4. Liquid Crystallinity of DNA-Stabilized SWNT Dispersions 
Nucleic acids dispersants, such as single-stranded DNA (ssDNA), have proven to be an 
excellent dispersing agents for SWNT through ?-stacking with the nanotube?s graphene side 
walls in aqueous solution with the aid of high power ultrasonication at low temperatures.97,108,113 
One important advantage of biopolymer stabilization of CNTs with sonication assistance is that 
it better preserves the intrinsic properties of nanotubes compared to other ways of covalent or 
oxidative functionalization15,136 which cause damage to the sp2 hybridized structure of nanotubes. 
However, some reduction in SWNT length and defect creation can still occur. Although another 
way of dispersing nanotubes, superacids protonation,19 would not cause damage to the nanotubes, 
handling in superacids is quite a difficult process and not as environmentally friendly as using 
aqueous solvents. Another potential advantage of using biological dispersants especially DNA is 
their unique properties of forming lyotropic liquid crystalline phase under certain conditions. 
Like some other biopolymers, such as rod-like bacteriophage ?d,137 the long and stiff helical 
structure of dsDNA molecules enables the formation of cholesteric phase in solution above a 
certain critical concentration which depends on the number of DNA base pairs, ionic strength 
and the experimental conditions used.31-34 This twisted LC cholesteric organization of DNA is 
are even evidenced to form in vivo.35,37,138 It was reported that natural DNA with relatively short 
molecular length of approximately 500 ? could also form precholesteric phase at lower 
concentration prior to forming cholesteric phase. At even higher concentrations, natural DNA 
38 
can form into multiple high-density phases.139-141 The first microscopic theory of the cholesteric 
phase was proposed by Straley in 1976 and predicted that the formation of a cholesteric phase 
was entirely due to the entropy driven excluded volume interactions which resembled the 
Onsager theory (1949) describing a nematic solution of rod-like molecules. One of observable 
features of DNA cholesterics is the cholesteric pitch in the typical fingerprint textures which may 
be directly dependant on the molecular length of DNA fragments as well as on the interaxial 
spacing of molecules and the solvent conditions like osmotic pressure and ionic strength.140,141  
Badaire et al.16 reported the first water-based nematic phase of non-covalently 
functionalized CNTs with nanotube concentrations up to above 4 wt % (~ 2.5 vol % for SWNT 
density of 1.3 g/cm3) by dispersing them in nearly equal concentration of denatured DNA. In 
their work, nematic phases of higher concentrations of SWNT stabilized by denatured DNA were 
obtained by solvent evaporation due to the unique phase diagram of DNA-nanotube suspensions 
which has linearly separated domains of dispersed and aggregated nanotubes. Unlike 
low-molecular-weight surfactants used to disperse nanotubes, denatured DNA does not cause 
depletion induced aggregation between nanotube bundles, making it an ideal candidate for 
acquiring highly concentrated systems.16,21 More ordered LC phases than the nematic phase had 
not yet been reported. However, given the premise that natural DNA can form cholesteric, and 
even higher density phases, above certain critical concentrations under well defined conditions, 
more ordered LC phase formation could be achieved at higher concentrations of CNTs by 
incorporating natural DNA with nanotubes in an appropriate way.   
 
39 
2.8. Rheology of Liquid Crystalline Polymers 
Rheology can probe the microstructure of dispersions and may even be able to 
qualitatively and perhaps quantitatively predict changes of microstructure due to various 
processing conditions including shear, temperature, and solvent removal in the assembly of 
macroscopic objects (e.g. films and fibers). Comparison of dsDNA/SWNT liquid crystal 
rheology to the rheological signatures of lyotropic nematic liquid crystalline polymers (LCPs) 
can provide better understanding of the phase behavior of nematic and cholesteric 
dsDNA/SWNT dispersions created in this research. Understanding the difference in phase 
behavior of nematic and cholesteric dsDNA/SWNT dispersions with concentration, textural 
effects, and shear flow is required to improve the properties of their macroscopic assemblies. The 
complex rheology of LCPs is very different from the ordinary polymers and low molecular 
weight liquid crystals (LMWLCs), and the subject of LCPs phase behavior is still somewhat 
controversial. Moreover, the insufficient rheological data of LCPs are mainly on lyotropic 
nematic LC phases and the flow properties of cholesterics have been studied little. Although, the 
rheology of model LCPs such as hydroxypropylcellulose (HPC),38-40 poly (? benzyl-L-glutamate) 
(PBLG)41,42 or poly (? benzyl glutamate) (PBG)43,44 solutions have been widely studied, there are 
still struggling efforts to understand this field due to the lack of universality in rheological 
properties of LCPs. Especially, when applied to nanoscale rods like SWNT, the flow behavior of 
the resulting liquid crystals became even more complicated due to the additional strong van der 
Waals attractions between SWNT and their polydispersity. The complex interplay between 
intermolecular interactions, molecular conformation, shear flow, and molecular orientation state 
are considered to be the factors responsible for the unique rheological properties of LCP 
40 
solutions.40,43 In general, the known evidence of lyotropic nematic LCPs include following and 
will be discussed in more details in Section 2.8.2: 
? Maximum in viscosity versus concentration curve 
? Three region viscosity versus shear rate curve 
? First normal stress difference changes sign with increasing shear rate 
? Long oscillatory transients after start up flow of steady shear 
? Cox-Merz rule not obeyed 
 
2.8.1. Rheology of Dilute and Semidilute Brownian Rods Dispersions 
The Brownian motion of rods is caused by the thermal fluctuation of the solvent 
molecules. As mentioned above, spontaneous alignment may occur for anisotropic mesogens like 
rods and platelets with increasing concentration, or with external shear force and electric field. 
However, at low concentrations, rods are distributed with random orientations especially in 
dilute or semidilute dispersions. Understanding the rheological behavior of isotropic system is 
therefore important and can provide the first step towards the development of rheological models 
for rods materials including cylindrical nanomaterials.142,143 If rods (length L and diameter d) can 
be considered to have dilute behavior in suspensions, they need to satisfy the following 
conditions; the rods are able to rotate freely both translational and rotational without being 
impeded by the neighboring rods, and the volume swept by a single rod rotation about its mass of 
center is large around L3. When the number concentration of rods ?  reaches a value 
proportional to L-3, rod-rod interaction is expected; this is the semidilute regime where the 
rotational motion of Brownian rods will be hindered. In Doi-Edwards theory (1986), dilute 
behavior is defined when the number concentration of Brownian rods is 
31L??
 and the 
41 
semidilute behavior occurs in the range of
23 11 dLL ?????
. Since the average distance between 
neighboring rods is very large compared to the rod diameter in the dilute and semidilute regimes, 
rods can be treated as thin line particles in these dispersions states.144 With increasing 
concentration, the onset of the concentrated isotropic regime occurs around 2/1 dL??  where 
the dimension of rod diameter becomes comparable to the distances between neighboring rods. 
In the isotropic concentrated regime, excluded-volume interactions take effect and isotropic 
packing of rods starts to be difficult. Rod motion is confined to straw-like volumes. At even 
higher concentrations, the rods are spontaneously oriented into anisotropic state, and Brownian 
motion of rods no longer exists.145 However, Mori et al.146 found that, experimentally, the 
transition from dilute to semidilute behavior occurs more than 30 times the theoretical estimation 
giving 303 ?Lo? . Since number concentration of rods is related to the volume fraction ?  of 
rods as ,4/2 ??? Ld? the critical concentration of semidilute behavior occurs when 
.)/(244/30 222 LdLdo ?? ??  This corresponds to 0.24 vol % of the dispersed phase for rods 
with aspect ratio of 100.48  
Characterization of size and length to diameter ratio (aspect ratio) distribution of 
nanocylinders dispersed in fluids, is significantly important in better understanding the phase 
behavior of the dispersions in this study. Based on Onsager (1949) and Flory (1956) theories of 
rods in fluids, the aspect ratio of rods together with their polydispersity, and rigidity, and the 
solvent quality are considered to be the major factors affecting the formation of lyotropic liquid 
crystals. Particularly, the biphasic region should be broadened by polydispersity and decreasing 
solvent quality. Fortunately, the average aspect ratio of dilute dispersions of Brownian nanorods 
can be rheologically determined using the Kirkwood-Auer-Batchelor (KAB) Equation. The 
42 
zero-shear viscosity of a (monodisperse) solution or dispersion of Brownian rigid rods can be 
used to calculate the average aspect ratio. The zero shear viscosity, o? , of a dilute suspension of 
rods is147 
??????? ? TBso 54lim 0 ????????? ???                       (2.13) 
where s?  is the solvent viscosity, B?  is Boltzmann?s constant, T  is temperature, ?  is the 
number of rods per unit volume, and ? ?TB??? 3/?  is the rotational relaxation time of the rods, 
and ?  is the perpendicular drag coefficient of the rods. The number concentration of rods is 
? ?LR2/ ??? ?  where R is the rod radius. The perpendicular drag coefficient ?  on a slender 
cylinder 148 is )(6 3 ????? fLS?  where,  
)/ln( 1 RL??
  and  2659.1
5.11 64.01)( ???? ????f
 
This yields the KAB relationship indicating that the intrinsic viscosity ???  of dilute 
dispersions of (monodisperse) Brownian rods is: 
? ? ? ?? ?? ? ? ? ? ?? ?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
???
? 2
2
0 /ln
659.1
)/(ln
5.11
/ln
64.01
/ln45
/2lim
RL
RL
RL
RL
RL
s
so
??
???
?
       (2.14) 
Therefore, L/R (and aspect ratio L/d where d is the rod diameter) can be directly obtained 
from the low shear rate plateau of a master curve generated from dilute dsDNA/SWNT 
dispersions of different concentrations (multiple concentrations reduce error). In the polydisperse 
43 
case, the viscosity contributions of each subpopulation of length are additive and the viscosity 
average length is related to the ratio of the third and first moments of the distribution of length: 
L
L
ddLvisc
31?                             (2.15) 
where,  NLL
N
i
ni
n ??? 1 , L  and 3L  are the first and third moments of the length distribution, 
N is the total number of rods, and n  is the moment of interest.149  
Applicability of this theory to new systems requires Brownian rigid rod behavior of dilute 
dispersions or the collapse of viscosity versus shear rate curves from samples of different 
concentrations onto a single master curve. In the latter case, the Newtonian plateau should 
transition toward the shear thinning region at a Weissenberg number Wi of approximately 0.2 
( ?? ??rotiW ) and the slope in the shear thinning region should be -0.5 and commence at Wi = 1 
as shown in Figure 2.10. Fortunately, rotational rheology of SWNT up to 4 micron in length in 
superacids ( s? ~ 30 mPa? s)19 and single particle tracking of SWNT in surfactants ( s? ~ 1 
mPa? s)150 revealed the Brownian rods behavior of the dilute SWNT. In addition, it was reported 
that even germanium nanowires with long lengths (up to 8 ?m) and high aspect ratios, can 
undergo Brownian motion in a dilute suspension.143 
 
44 
 
Figure 2.10. Universal shear thinning curve of monodisperse dilute Brownian rods.151  
 
In the semidilute regime, Doi?Edwards theory predicted the zero-shear viscosity of 
semidilute solutions of rods using the tube model where the movement of a test rod is confined 
by the surrounding rods: 
36 )10/( ???? roBo DTL?                       (2.16) 
where ?61?
roD
is the dilute-solution rotary diffusivity of the particle, and ?  is a 
dimensionless constant. Rearranging the equation 2.16 gives the relationship 3???o . Thus, 
when zero-shear viscosities are plotted against concentration on a logarithmic scale for 
semidilute solutions of rods, the slope of the curve is 3. An example of experimental data in 
qualitative agreement with the semidilute theory is given in the Figure 2.11 below. However, the 
window of semidilute behavior with slop of 3 appears to be very narrow in the range of 2.5 - 5 
Red
uced 
v
i
scosi
ty
 
W ei ssen b er g  n u mb er  (W
i
)
5
45 
wt % which is almost nonexistent. The narrowness of semidilute window can be attributed to 
small molecular length, polydispersity, and molecular flexibility.144 As a comparison, the 
Kirkwood-Auer-Batchelor equation predicts that, the slope in the dilute regime is related only to 
the rod aspect ratio and is smaller than 1 when the o?  is plotted against concentration. 
Specifically, the larger the rod aspect ratio is, the higher the slope which approaches closer to 1.  
 
 
Figure 2.11. Zero-shear viscosity versus concentration for PBLG.144 
 
In this study, the ternary system of dsDNA/SWNT dispersion is relatively complicated. 
Due to the excess amount of viscous dsDNA used to stabilize SWNT either in non-centrifuged 
(dsDNA:SWNT=7.5:1 by weight) or supernatant (dsDNA:SWNT=5.1:1 by weight) dispersions, 
the resulting rheology of dilute SWNT (<0.03 wt % or 500 ppm of SWNT) in dsDNA solution 
might reflect the phase behavior of dsDNA rather than the SWNT. Based on the UV-vis analysis 
performed in this study, the dispersibility of SWNT in dsDNA solutions is also affected by the 
46 
volume of samples prepared. The volume of starting 0.75 wt % dsDNA-0.1 wt % SWNT 
dispersions used to get concentrated liquid crystal is typically 150 mL. In order to ensure the 
consistency in experimental data, same volume of 150 mL samples with identical weight ratios 
of dsDNA to SWNT were used. More specifically, the steady shear viscosities of both 
non-centrifuged and supernatant dispersions of dsDNA/SWNT with various concentrations in 
dilute regime were tested in the shear rate range from 0.01 to 100 1/s given the condition that 
they are in accessible instrument torque transducer range. If the viscosity versus shear rate curves 
from samples of different concentrations collapse onto a single master curve as described above, 
the difference in average aspect ratio in the dilute regime between the bulk (non-centrifuged) 
dispersion and the supernatant of dsDNA/SWNT dispersion can be detected by rheology. Since 
this approach has only been used for dispersions of rods in Newtonian fluids, it is not known if 
the presence of the dsDNA will prevent applicability of this method. 
 
2.8.2. Rheological Signatures of Nematic Liquid Crystalline Polymers  
Although LCPs have very different rheological properties when compared to ordinary 
polymers, their isotropic phases have similar rheological behavior to ordinary polymers.152 To 
determine the complexity of the microstructure in a flowing LCP, one need only look at the 
steady state viscosity behavior.41 Rheological signatures of LCPs in lyotropic nematic phase are 
described in the sections below. 
Maximum in Viscosity versus Concentration Curve. A well-known rheological 
characteristic of lyotropic nematic LCPs is the nonmonotonic relationship between viscosity and 
concentration at a constant temperature. In lyotropic LCPs, the low shear viscosity goes through 
maximum in the biphasic regime and the magnitude of this maximum decreases with increasing 
47 
shear rate.48 The viscosity increases as a function of increasing concentration as long as the 
system is predominantly isotropic. However, the increasing proportion of the anisotropic phase 
with concentration results in a decrease in viscosity as a lower shear stress is required for 
orientation. The sharp decrease in viscosity in the biphasic regime may be due to the relative 
ease of molecules sliding past each other in the partially ordered phase than isotropic 
concentrated state. This is supported by the fact that the alignment in liquid crystal is due to the 
gain in translational entropy compensating for the loss of orientational entropy. With further 
increase in concentration, viscosity decreases with concentrations within the liquid crystal phase. 
The viscosity decreasing sharply on the transition to liquid crystal phase is a complex 
characteristic which implies a reverse trend to the common knowledge obtained from ordinary 
polymers that viscosity increases monotonically with increasing polymer concentrations. Once 
the order can no longer be increased, the viscosity once again keeps increasing with 
concentration. 
Three Region Viscosity versus Shear Rate Curve. The rheological behavior of 
lyotropic nematics also has a typical shape of viscosity as a function of shear rate. It often has 
three viscosity regions of shear thinning, almost constant viscosity, and additional shear thinning 
with increasing shear rate. The three distinct regions occurred in viscosity versus shear rate curve 
in some of nematic LCPs are shown in Figure 2.12. Region I shear thinning is the most 
distinguishing feature for lyotropic nematic LCPs and have been observed in several types of 
model systems at high concentrations such as HPC and PBLG.41,153 Even for SWNT dispersed in 
superacids solutions, possible three region behavior was reported with a very narrow region II 
plateau between the two shear thinning regions.19 Regions analogous to a low shear rate 
?Newtonian plateau? in Region II and shear thinning at high shear rate in Region III are common 
48 
for isotropic solutions of flexible or rigid polymers. The polydomain structure of nematic phase 
is considered to be preserved at the first shear thinning region and part of the second region with 
a constant viscosity both appearing at low shear rate.  
 
 
Figure 2.12. Viscosity versus shear rate curve showing a three region behavior. For some LCPs, 
the low-shear-rate Region 1 is not found.150,152 
 
Region I has generally been observed only in high viscosity, ?tumbling? nematics where 
the shear viscosity in Region II exceeds ~10 Pa? s. Typically, region I shear thinning is believed 
to be related to director tumbling in liquid crystalline phase upon shear flow. However, Region I 
is not seen in all lyotropic nematic polymeric liquid crystals and depends on many factors such 
as specific materials used, flexibility of rods, defect texture, or the range of shear rates 
covered.153 It is also suggested that Region I of LCPs can be associated with cholestericity. At 
the low shear rate of Region I some LCPs like HPC show diffraction characteristics of the 
49 
cholesteric phase and the cholestericity is further broken down by the higher shear flow in 
Region II.48 
In this work, the high viscosities resulting from the long rod length and high aspect ratio 
of SWNT (L ~ 500 nm, D ~ 1 nm) coupled with liquid crystalline phase formation when 
stabilized by dsDNA, could be expected to enable measurement of Region I behavior in 
dsDNA/SWNT dispersion. However, due to the following reasons, it is not apparent whether a 
three region behavior can be expected in the LC phase of dsDNA/SWNT dispersions. 
a) In the bulk, both thermotropic and lyotropic materials contain so many orientational defect 
that homogeneity typically persist over domains no larger than about one micron.154 Region 
I behavior is also associated with small characteristic texture length scales on the order of 
one micron. The defect texture in the dsDNA/SWNT nematic and cholesteric liquid crystals 
and the presence of excess or free dsDNA may have an effect on the polydomain liquid 
crystal texture leading to a complex ternary system (SWNT, dsDNA, and water).  
b) 40% PBLG (MW=238,000) solution which showed Region I shear thinning also suffered 
from complex hysteresis consisting of two distinct viscosity braches for steady shear 
viscosity measurement with different preshearing.41 This behavior may be due to the texture 
elasticity and the change in defect texture upon shear flow indicating that Region I behavior 
is extremely sensitive to the sample preshear history. Thus, an exploration of preshear 
protocols for ?wiping off? the flow history is required to testing for Region I in 
dsDNA/SWNT liquid crystal.  
c) Not all LCPs exhibited Region I shear thinning. Due to the fact that all rheometers have 
detection limits on shear stress, it is relatively difficult to test Region I experimentally. 
Further, with the low shear rate involved in shear thinning of Region I, several other 
50 
problems are have to be considered such as sample evaporation and extremely long time 
needed for testing which again highly demanding a reasonable preshear protocol.  
 
First Normal Stress Difference Changes Sign. The unique rheological properties of 
nematic LCPs have led to the development of new molecular theories such as Doi, Marrucci and 
Larson.47-49 However, the direct modeling of nematics with highly textured structures are so 
complicated that those theories adopted a simple geometry for calculation. For example, the Doi 
model assumes only monodisperse rigid rods and an anisotropic excluded volume potential when 
predicting the two sign changes in the first normal force difference N1 for lyotropic LCPs. Due to 
the polydispersity and flexibility of rods, textural defects, and various molecular interactions 
existing in the real LCPs systems, the experimental results would have discrepancies with theory. 
Figure 2.13 shows the N1 changing signs with increasing shear rate for lyotropic nematic LCPs. 
The negative N1 was first reported by Kiss and Porter in 1978.155 Since then, few LCP solutions 
showed two sign changes in N1 where at lower shear rates N1 was positive and increased with 
shear rate due to the tumbling of average molecular orientation (i.e., the ?director?), and at 
intermediate shear rate N1 suddenly dropped and changes to negative due to the director wagging 
which would cause low molecular order. At even higher shear rates, the N1 became positive 
again due to the flow alignment of rods.48,152 This change in N1 also corresponded to the 
universal three region steady-shear viscosity behavior in flow curve. The origin of negative N1 is 
still controversial in spite of considerable research.48,156 It is widely believed that the occurrence 
of a negative relative minimum in N1 is associated with a flow transition between director 
tumbling and flow-aligning behavior.156,157 Correspondingly, the sign change observed in some 
LCPs occurred near the onset of shear thinning Region III in the viscosity versus shear rate curve. 
51 
On the other hand, for conventional polymeric liquids, or for nontumbling liquid crystals, N1 
increased monotonically with increasing shear rate (only positive N1).158 Along with relative 
minimum in the N1 flow curve, even if the minimum was positive, the oscillatory response in 
transient flow were considered to be indirect evidences of director tumbling.39,159 Theories also 
indicate that tumbling is affected by polymer flexibility which would increase as the temperature 
increases.38,160,161  
The rheology of nanocylinder liquid crystals would be an extension of well understood 
flow behavior of model LCPs such as HPC and PBG at moderate concentrations. For cholesteric 
liquid crystals of dsDNA/SWNT dispersions, it might be possible that the typical curve of two 
sign changes of N1 would occur outside the experimental shear-rate range, or cholesterics may 
have different character in negative N1 behavior as compared to nematic LCPs. These 
possibilities are discussed in Chapter 4.  
 
 
Figure 2.13. Shear viscosity and first normal stress difference versus shear rate for 17 % PBLG 
(MW=350,000) in m-cresol. The circled triangles are negative N1 values.48,155  
52 
Long Oscillatory Transients. Another striking behavior of LCPs in a shear flow is long 
oscillatory transients achieved before reaching steady state. Typically for liquid crystals, the 
transient response is complex and much longer in duration than 100 shear units. While for 
isotropic suspensions and ordinary polymer solutions, the oscillatory transients last a few shear 
units at most.49 Moreover, in Region II of the viscosity versus concentration curve, the 
oscillation period is fairly regular and roughly inversely proportional to shear rate. As a result, 
when the transient stress response obtained at different shear rates in Region II are plotted 
against shear units the maximum and the minimum of the curves will overlap.162 However, to the 
author?s knowledge, it has never before been possible to achieve definitive Region II transient 
stress plots for nanocylinder liquid crystals including SWNT. Varying the shear rate range 
coupled with evaluating the distortion texture due to transient flow by shear cell will be a 
feasible approach to investigate the rheological properties of dsDNA/SWNT liquid crystals 
occurring in Region II 
Cox-Merz Rule Not Obeyed. The Cox-Merz rule states that, the modulus of complex 
viscosity ?*( ?), a linear viscoelastic material property, equals to steady shear viscosity ?(?? ) at 
equal values of angular frequency ? (s-1) and shear rate ??  (s-1). However, unlike ordinary 
polymers, lyotropic nematic LCPs have a common characteristic of not obeying the empirical 
Cox-Merz rule. Basically, polymeric fluids having a complex deformation-dependant 
microstructure would fail to obey the Cox-Merz rule.163,164 Although many studies have been 
contributed to the well-established Cox-Merz rule relating the linear viscoelastic material 
properties to the nonlinear flow curve ever since its introduction in 1958,165 the first molecular 
theory successfully predicting this rule for monodisperse polymer was proposed by Marrucci late 
in 1996.166,167 Molecular theory predicting the Cox-Merz rule was later extended to a 
53 
polydisperse polymers by Mead et al.168,169 using the same molecular mechanism proposed in 
Marrucci?s work for fast flows which is convective constraint release (CCR). The CCR is the 
nonlinear relaxation mechanism and the longest relaxation time in fast flows scales with ?? -1 
independent of molecular weight. According to above molecular theories, the Cox-Merz 
relationship is an unphysical coincidence indicating no physical similarities in the dynamics of 
polymers or molecular relaxation mechanisms between linear and nonlinear flows. The 
Cox-Merz rule is due to the CCR process and the similar mathematical relationship between 
steady shear and complex dynamic viscosities at identical high dimensionless shear rates or 
frequencies which has the same relaxation mechanisms transition from linear to nonlinear flow 
regions.169  
 
2.8.3. Shear Effects on Liquid Crystal Rheology  
In this section, the change in polydomain texture of nematic LCPs upon shear will be 
discussed since it is widely believed that the rheological phenomena are associated with texture 
evolution during shear. The texture of a bulk sample of an LCP is defined as the spatial variation 
in the director, which is the direction of average molecular orientation.170 The polydomain 
texture in LCPs is relatively stable under shear when compared to small molecule organic liquid 
crystals due to the high aspect ratio of rods and long relaxation time in LCPs. Furthermore, 
Marrucci171 defines the texture length scale a , which can be thought as the diameter of a 
domain, a volume over which uniformity persist in the orientation,170 as below 
2/1
??
?
?
?
??
?
?
??
???
Ka                               (2.17) 
54 
where K is a typical Frank elastic constant, ???  is viscous energy density. This relation indicates 
that the refinement of texture would occur as the flow rate increases. In fact, the defects of 
polydomain liquid crystals can be annealed with the application of external force. Shear flow 
alignment could be a feasible approach to achieving higher degree of rod alignment since it is 
used widely in films or fiber processing. However, effectively controlling the shear rate is an 
important task considering the characteristic three region viscosity curve as a function of shear 
rate which is discussed before. Except for the shear rate range of Region III where flow 
alignment occurs, shear can also result in other behaviors such as low shear aggregation, yielding, 
jamming, vorticity alignment, and shear banding. Shear banding is one of the regularly observed 
textures in LCPs which is considered to be caused by the distortions of nematic director field 
under shear. Usually, shear banding refers to dispersion segregating into birefringent bands of 
high and low concentration as a result of shear.48 Interestingly, the same sequence of texture 
evolution with shear was reported for both a low-molecular-weight polyester melt38 in the 
nematic range and PBLG solution.172 For start-up of shearing at lower stain, shear bands 
perpendicular to shear flow quickly form. With increasing strain, stripes parallel to flow begin to 
appear and the perpendicular bands started to disappear until only the parallel stripes remain. The 
difference between ?stripes? and ?bands? is that, the former are parallel to the flow direction 
while the latter are perpendicular to it. At even higher strain, the stripes would break down into a 
fine-scale ?speckled? texture. However, bands orthogonal to the previous shear direction appear 
again after the cessation of shear. Although the phenomena of this particular sequence are not 
completely understood, it is highly considered to be related to the tumbling nature of LCPs.  
Furthermore, bands have been observed for almost all LCPs, their origins still have been 
an ongoing source of research.173 The shear banding is a quasi-periodic texture with the 
55 
annealing or healing time, ta, for the spontaneous disappearance of a disclination texture from a 
nematic of viscosity ? and thickness h as 
Khta 2??                                (2.18) 
where K is the characteristic Frank constant.48 Due to the high viscosity in LCPs with 
comparison to small molecule nematics, the textures are much more persistent in LCPs. The 
effective ?director diffusivity?, K/?, of small molecule nematics is on the order of 10-6 cm2/s, 
while for polymeric nematics K/? ~ 10-8 to 10-10 cm2/s. Since LCPs have much long relaxation 
time, preshear protocol is needed to obtain reliable dynamic measurements by ?wiping off? the 
previous flow history.  
In this work, the effect of small shear rate applied for a few seconds on the cholesteric 
pitch of dsDNA/SWNT liquid crystal was investigated by polarized light microscopy coupled 
with shear cell. The phase behavior of dsDNA/SWNT liquid crystals upon shear was also 
conducted by rheology including dynamic and steady shear test.  
 
2.8.4. Temperature Effects on Liquid Crystal Rheology  
The motivation for investigating the temperature effects on the rheology in this work is 
seeing if there is a difference in nematic and cholesteric system of dsDNA/SWNT. 
Time-temperature superposition (TTS) is a valuable tool for describing the viscoelastic behavior 
of lyotropic polymers where the phase transition is only affected by concentration rather than 
temperature. It can show the phase behavior at a ?reference temperature? by obtaining data at 
several temperatures to generate a master curve covering many decades of time or frequency 
whereas only a range of three or four decades can be obtained at a single temperature.174 Shift 
56 
factors for the magnitudes of measured stresses (vertical shift factor, Tb ) and for time or 
frequency (horizontal shift factor, Ta ) can be used to create the master curve. Usually, the shift 
along the vertical axis is often small or negligible. While the temperature dependence of shift 
factor Ta  can be expressed as the empirical WLF (Williams-Landel-Ferry) equation;175,176 
Trrr rr aTTC TTCTT l o g)(l o g 2 1 ??? ????? ??
                    (2.19)      
where C1 and C2 are two material specific coefficients, and Tr is the reference temperature. The 
change in temperature times density (vertical shift factor) ?? TTb rrT /?  is small and can be 
ignored in most cases. The WLF equation reveals that shifting of viscosity, which is composed 
of both stress and time, requires the application of both shift factors such as )(T
abTT?
 versus 
Ta?? . The WLF relation is usually appropriate for temperature changes in the range of Tg to 
Tg+100 K for melts of glassy polymers where Tg is the glass transition temperature. 
The TTS principle only applies to thermorheologically simple materials where the 
material structures do not fundamentally change in the temperature range considered. Thus, this 
technique would fail for polymers with complex relaxation mechanism and with phase transition 
or chemical change occurring within the temperature range measured. However, with some 
exceptions, TTS principle was also reported to hold for some polymer blends with individual 
components having different relaxation mechanisms.177 Moreover, time-temperature 
superposition is not always applicable for rod-like polymers. The solvent quality of rod-like 
polymers can be tuned by simply changing temperature which further resulted in the change of 
phase boundaries. On the other hand, the phase behavior of SWNT in superacids19 and even low 
57 
aspect ratio CdSe-cylcohexane178 has been found to be temperature insensitive. However, little 
rheology has been done for the latter case. For SWNT in superacids, although the phase 
transition is independent of temperature, LC phase behavior was obtained with improved solvent 
quality by using superacids with different protonating abilities.  
 
2.8.5. Persistence Length Effects on Rheology 
Recent reports on SWNT in superacids19 showed all five rheological signatures of LCPs 
listed above and come to the conclusion that SWNT can form liquid crystals when appropriately 
dispersed. Possible three region behavior was observed for SWNT in 102% H2SO4. Also, both 
the shear viscosity and the first normal stress difference oscillated for over 800 shear units for 7.6 
vol % SWNT in superacids. However, the new system of dsDNA/SWNT dispersions appears to 
be very different and has the added complexity of DNA which is not really a rigid rod. A 
molecule is effectively a rigid rod, if the rod contour length L is much shorter than the 
persistence length, P?  (L << P? ). Some molecules, such as helical DNA, collagen, or PBLG, 
are called ?semiflexible? with L ? P? . They are stiffer than ordinary flexible polymers like 
polystyrene but are not rigid rods.179 Persistence length is the length scale over which a filament 
shows significant curvature induced by thermal fluctuations. This is because the dynamics of 1D 
rods in a viscous medium are essentially dominated by the balance of Brownian forces, which 
tend to bend them, and elastic forces, which oppose this curvature.180 The simplest model for 
molecules with uniform distribution of flexibility is the inextensible elastic beam with bending 
rigidity EI?? , where E is the elastic modulus and I is the area moment of inertia about the 
filament axis. A hollow cylinder of inner radius ir  and outer radius or , has a moment of inertia 
of ).( 44 io rrI ??? 181 As a result, the persistence length can be expressed as the ratio of bending 
58 
rigidity to thermal energy, TkBP /?? ? , where Bk  is the Boltzmann constant, and T is the 
absolute temperature. The persistence length of DNA measured by light scattering is around 
0.053 ?m, which is about the length of 150 base pairs.182 However, under low-salt conditions, the 
effective persistence length of DNA is expected to be longer due to the electrostatic repulsion of 
charges along the molecules.48 The prediction of the bending rigidity of SWNT according to the 
theory based on a continuum model for a hollow cylinder of radius R is 3CR??? , where the 
?in-plane stiffness? C is estimated as ? 345 J/m2.183,184 This results in P? =134 ?m for SWNT 
with typical diameter of 1 nm. This model indicates a diameter dependent bending rigidity and 
persistence length of nanotubes. Considering that the length of the majority of commercial 
SWNT is L ~ 500 nm, which gives P? /L on the order of 100 and they are effectively rigid rods 
( P? >>L). The individual shape fluctuations of SWNT can be measured by fluorescence 
microscopy. The experimental values of P?  obtained from fluorescence microscopy range from 
26 to 138 ?m for SWNT of diameter from 0.77 to 1.15 nm and length larger than 3?m.180 The 
persistence length of each SWNT can also be determined rheologically from the dynamic data. 
The theoretical relationship between persistence length and bending relaxation times is;181,185,186 
? ? 421
???
?
???
? ?
?
L
nT
Bn
P ?
??
??                          (2.20) 
where )/ln(/4 dL??? ?  is the friction coefficient, and 4/ qn ??? ?  is the relaxation time of 
mode n, with Lnq /)2/1( ??? . The model and experimental measurements show P? ? 26 ~ 
138 ?m for SWNT indicating that SWNT is rigid rod up to L ~ 3 ?m. 
 
59 
2.9. Macroscopic Assembly of Biopolymer-Stabilized Carbon Nanotube Dispersions 
The remarkable properties of nanomaterials can often be only realized if they are 
processed into useful macroscopic formats. Generally, aligned CNTs could promise enhanced 
mechanical, optical, electrical, thermal, magnetic properties for the resulting macroscopically 
assembled objects. The recent achievements in obtaining nematic LC phase from the dispersion 
of SWNT in superacids19, in biological hyaluronic acid and denatured DNA solutions16,17 and 
aqueous dispersion of acid treated MWNTs15 not only opened up the rich interest in LC study 
with nanomaterials but also provided a novel approach to process CNTs nanomaterials with 
large-scale of alignment of nanotubes. However, obtaining a better dispersion of CNTs at the 
really first stage still remains to be a challenge due the inherent van der Waals attraction between 
the tubes, which is on the order of 20 - 40 kBT/nm.45,46 Alignment on a larger scale over 
macroscopic areas is also an obstacle which is important for potential applications of CNTs. 
Several methods have been developed to obtain a good dispersion of nanotubes in solvent 
including ultrasonication, surface modification of nanotubes,187 surfactant addition21,188 and 
further using external forces of electrical, magnetic and shear flow to obtain large-scale ordered 
structure of CNTs. Typically, macroscopic assemblies of CNTs materials have been achieved in 
the formats of films, coatings and fibers by different methods, while fluid phase possessing is 
considered to be a promising route. 
 
 
 
60 
2.9.1. Films/Coatings 
Several methods of fluid phase processing of CNTs into films and coatings were reported 
recently, including drop drying,189,190 spin coating,191 shear alignment,192 and layer-by-layer 
assembly.1 Alan H. Windle?s group demonstrated films of CNTs obtained by drop drying of the 
biphasic or LC aqueous dispersions of both MWNT and SWNT with oxidative acid 
treatment.189,190 It was confirmed that the typical textures of the LC dispersion was retained in 
the resulting thin films including birefringence and ordered structure on drying by evaporation. 
However, predominant alignment of the nanotubes parallel to the edge of the dried sample was 
also found. This segregation effect of significantly greater quantity of nanotubes on the periphery 
than in the center was explained to be due to the ?ring effect? in dried coffee stains.193 The 
driving force for this effect is considered to be minimizing energy in the field by orientation of 
the nanotubes.  
Integrating biomolecules with CNTs for direct alignment during macroscopic assembly is 
becoming more attractive due to increased recognition of the benefits of incorporating nature?s 
toolbox with nanomaterials. Making use of the desirable features of biological systems, such as 
the abilities of precise self-assembly and molecular recognition, is a feasible route to achieve 
broadly aligned nanomaterials with highly improved properties and biocompatible capability. A 
possible approach to obtain a uniform macroscopically aligned film from lyotropic nematic 
suspension of SWNT stabilized by denatured DNA was reported by mechanical shearing.192 It 
was important to note that alignment in their study is achieved when only started with a nematic 
LC material not isotropic suspension, which further proved that alignment is also due to the 
intrinsic properties of the liquid crystal and not only to the flow. However, the alignment of 
nanotubes in the films was weaker than the above discussed systems by stabilizing the MWNT 
61 
and SWNT with oxidative acid treatment. This result was attributed to the following reasons like 
polydispersity and waviness of nanotubes materials, residual catalysts impurities and poor 
dispersion of tubes. One should keep in mind that stabilizing pristine nanotubes with denatured 
DNA could promise retaining the unique properties of nanotubes without damage. It is also 
possible that, due to the viscous properties of DNA, the presence of it in the final film product 
especially when DNA coated on the film surface, hindered direct and clear observation of 
nanotubes alignment in the system, while it is more straightforward to observe the alignment in 
acid treated tubes dispersed only in water.  
Besides macroscopic assembly of films from LC phase of CNTs dispersions, large scale 
biomimetic SWNT coatings from DNA and lysozyme (LSZ) stabilized isotropic dispersions 
were fabricated by layer-by-layer by assembly with electrostatic interaction between layers.1 The 
resulting coatings of 68 layers with 205 nm thickness, showed relatively strong mechanical 
properties of high hardness of 1 GPa and Young?s modulus of 22 GPa, while retaining 
significant antibacterial activity of LSZ. Uniform orientation of SWNT within each layer was 
achieved by applying directed air stream between each deposition step.  
 
2.9.2. Fiber Spinning 
Fiber spinning is also one of the viable routes of fluid processing to produce scalable and 
industrial quantities of CNTs fibers. Fiber spinning methods include melt spinning and solution 
spinning. Melt spinning is not an option due to the nature of CNTs; they decompose prior to 
melting around 750 oC in air and 2000 oC in an inert atmosphere. Furthermore, the melting and 
decomposition temperature of biopolymers are relatively low. For example, DNA denatures in 
boiling solvents and undergoes a significant decomposition at 175 oC and followed by a more 
62 
dramatic change starting at 205 oC under nitrogen. The maximum decomposition rate occurs at 
220 oC, and above 205 oC DNA pyrolysis occurs.194  Thus, solution spinning is the most 
feasible approach. However, the combination of various steps involved in solution spinning such 
as solvent removal, shaping, and the handling of the resulting fiber makes solution spinning a 
very complex process.195 Generally, there are two types of solution spinning: dry-spinning and 
wet-spinning. Both methods involve pressure-driven extrusion of macromolecular dispersions 
through small orifices into a coagulation region. In the absence of drawing, the cross section of 
the fiber is on the order of orifice diameter. However, the exact dimensions depend on the 
particular solidifying conditions of the extruding fiber. During dry-spinning, the dispersion, 
typically called a dope, is extruded into a heated chamber, and the coagulation of the fiber occurs 
by solvent evaporation. For wet-spinning, the dispersion is extruded into a coagulation bath. 
Fiber formation is achieved by the solvent leaving the fiber by diffusing into the bath. The 
ultimate fiber properties of these methods are affected by many factors, including properties of 
initial dispersion, extrusion speeds, and post treatments such as drawing. For wet-spinning, 
additional parameters include the diffusion mechanism of solvent, the temperature and flow 
behavior of the bath, and potential chemical reactions between the solvent and the bath.  
Vigolo and Poulin et al.14 demonstrated the first SWNT fiber spinning process by 
introducing flow-induced alignment. Isotropic surfactant stabilized SWNT dispersions were 
injected into the co-flowing steam of a polymer solution with certain contents of polyvinyl 
alcohol (PVA). PVA provides a strong bridging attraction between the nanotube bundles. The 
resulting fibers did not show a good alignment, and the mechanical and electrical properties of 
the fibers were not very promising with Young?s modulus in the range of 10-20 GPa, which is 
far weaker than the modulus of individual nanotubes,196 and a high resistivity of about 0.1 ??cm. 
63 
However, this spinning process provided a simple and feasible approach to scaled-up production. 
It is noteworthy to mention the systematic fiber spinning processes developed at Rice University 
which successfully produced well-aligned continuous macroscopic fibers composed solely of 
SWNT which are originally dispersed in superacids. Several spinning techniques including wet- 
jet wet spinning, dry-jet wet spinning, and coagulant co-flow extrusion provided a feasible 
approach to industrial scale-up process of fiber spinning with large-scale alignment of 
nanotubes.10,20,197 The neat SWNT fibers produced via solution spinning of the original 
dispersion of high ordered superacids/SWNT with 8 wt% SWNT, showed much improved 
properties than those of SWNT fibers involving polymers (PVA) mentioned previously. The 
Young?s modulus of neat SWNT fibers was 120 ? 10 GPa and the tensile strength obtained was 
116 ? 10 MPa. In addition, electrical resistivity of 0.2 m??cm and thermal conductivity of 21 
W/K?m were obtained for neat SWNT fibers.10   
Poulin?s group also demonstrated fiber spinning from denatured DNA-stabilized 
dispersions, and achieved stronger mechanical properties compared to their previously 
characterized fibers from nanotubes dispersed in conventional surfactants, such as sodium 
dodecyl sulfate (SDS).198 SDS is an anionic surfactant, and stabilize the nanotubes through 
electrostatic repulsion which overcome the van der Waals interactions. Due to the presence of 
non-conducting DNA chains forming uniform coating by wrapping around the CNTs, the 
electrical properties of fibers were lowered. Although DNA molecules have a negatively charged 
phosphate backbone, the intrinsic conductive property of DNA is still an unsolved problem with 
its conductivity being influenced by many factors, such as DNA sequence, molecular length, 
interaction between molecules, microstructure and the environment of DNA, to name a few.199 
However, with thermal annealing to remove impurities in DNA fibers, the conductivity was 
64 
considerably increased while compromising the mechanical properties due to partial removal of 
DNA and the coagulant, which was PVA in this case. As an alternative, annealing the CNTs 
before fiber production led to smaller improvements in conductivity without damaging the 
mechanical properties. In general, doubling of Young?s modulus and tensile strength was 
obtained with respect to the reference SDS fibers. Although stretching, or drawing, is considered 
to be one possible approach to achieve better alignment of nanotubes in the tube, both the 
electrical and mechanical properties of DNA fibers in this study did not show any improvement 
in these properties upon stretching. These behaviors were not well understood yet, but they 
indicate the specific influence of DNA. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
65 
 
 
 
 
 
Chapter 3 
Experimental Details 
3.1. Materials 
The representative properties of different sources of single-walled carbon nanotubes 
(SWNT) used in this research are shown in Table 3.1. Two batches of SWNT were used for 
characterizing the phase behavior, films, and preliminary investigations on fiber spinning.  
High-Pressure carbon monoxide process (HiPco) SWNT (Batch 183.6) from Rice University 
(Houston, TX) were purified by a thermal oxidation-acid extraction cycle.200 A TA instruments 
(New Castle, DE) Q500 thermo gravimetric analyzer (TGA) and Zeiss EM 10C 10CR 
high-resolution transmission electron microscope (TEM) confirmed a purity of 98-99 %. Unidym 
SWNT, also produced by the HiPco process, were supplied by Unidym, Inc. (Menlo Park, CA). 
The batch Lot# P0900 used had 92 % purity which was confirmed by TGA. In the later part of 
fiber spinning, Rice HiPco SWNT of batch 187.2 were used. According to TGA, the purity of 
purified SWNT used in PVA coagulated LSZ/SWNT fibers was 99 %, while the raw SWNT 
without purification had a purity of 64.8 %. The purity of SWNT in fiber spinning from 
dsDNA/SWNT bulk dispersions by both dry and wet spinning methods was 99 %. The physical 
dimension of SWNT, length and diameter, were measured by atomic force microscopy (AFM). 
 
 
 
 
66 
Table 3.1. Representative properties of SWNT used in this research. 
Source Batch Purity  (%) 
Average Aspect 
Ratio  
        
Raman D:G ratio 
514 & 785 nm laser 
Rice 183.6 98-99 ~ 550 ~0.06 (514 nm) ~0.04 (785 nm) 
Unidym P0900 92 ~ 600 ~ 
Rice 187.2 64.8 ~ 826 0.1 (514 nm) 0.1 (785 nm) 99 
 
Salmon testes double-stranded DNA (dsDNA) with 5.4 % sodium salt was purchased 
from Sigma-Aldrich, Inc. (Saint Louis, Missouri). The % G-C content for dsDNA from salmon 
testes was reported to be 41.2 % and the Tm (melting temperature) was reported to be 87.5 oC in 
0.15 M sodium chloride plus 0.015 M sodium citrate.74 The UV-Vis absorbance profile 
(Ultrospec 2100Pro UV-Vis spectrophotometer) proved a pure preparation of DNA with the 
absorbance ratio at 260 and 280 nm wavelength A260/A280 = 1.86. The A260/A280 ratio of 1.8 - 2.0 
is generally an indication of a pure DNA where the A280 corresponds to the maximum UV light 
absorbance of protein.72 Dialyzed, lyophilized LSZ from chicken egg white was purchased from 
Sigma-Aldrich, Inc. (Saint Louis, Missouri). Cationic surfactant (1-tetradecyl) 
trimethyl-ammonium bromide, 98% (TTAB) was obtained from Alfa Aesar. Sulfonated cellulose 
nanocrystal (CNC) suspensions were prepared at Clemson University. The representative 
properties of CNC were the following; average length of 107 ? 55 nm, average width of 20.5 ? 
6.1 nm, and average hight of 8.9 ? 2.8 nm with a rectangular cross section. 
 
67 
3.2. Preparation of SWNT Aqueous Dispersions 
All dispersions of SWNT in dsDNA were achieved by sonication assistance (Sonics VC 
750, Newtown, CT) using a standard probe of 13 mm diameter with a replaceable tip (made of 
titanium alloy Ti-6AL-4V, # 630-0406) in de-ionized water (DI water). The probe tip of the 
sonicator causes a cavitation field in the liquid through rapid vibration and the subsequent 
collapse of violently formed microscopic bubbles releasing significant energy which affects the 
objects in the surrounding cavitation filed. The mechanical vibration is provided by transforming 
the electrical signal produced by ultrasonic generator through a convertor/transducer and is 
transmitted into bath or down horn. The strong shear force produced from the cavitation process 
can exfoliate the SWNT bundles during sonication, and may also cause damage or cutting of the 
tubes due to the fluid friction at the surface of nanotubes.201,202 The radial elongation flow 
resulting from the collapsing of a cavitation bubble can produce theoretically predicted strain 
rates d?/dt of up to 109 s-1 causing a large relative velocity between nanotubes and liquid. 
Furthermore, the resulting friction force is scaled approximately as ? d?/dt L where ? is fluid 
viscosity and L is the length of the nanotubes. Nanotubes will break when this applied shear 
stress exceeds the tensile strength of tubes which is about 37 GPa.201,203 It has been also reported 
that the dispersibility of SWNT in different organic solvents during a cavitation process is 
affected by the sonication time, and output power, as well as solvent qualities such as viscosity, 
vapor pressure, surface tension, and density.202 
  Two types of initial dsDNA/SWNT dispersions were prepared for further evaporation 
to get higher concentrations of SWNT. When using dsDNA as the dispersing agent for SWNT, 
both cholesteric and nematic phase formation is possible, but which phase forms depends on the 
preparation method. In a typical experiment, dsDNA was dissolved in water at 35 oC for about 
68 
45 min by stirring with a magnetic bar. SWNT powders were mixed with aqueous the solution of 
dsDNA, and the mixtures were sonicated for 30 min at power level of 50 W in an ice bath to 
obtain a fine black dispersion. Both the low temperature used in making dsDNA solution and the 
ice bath applied during sonication provided a favorable condition for keeping the DNA as double 
stranded. The ice batch is also an effective way to prevent solvent evaporation caused by the 
high energy released during sonication process. The phase behavior of dispersions of 1.0 wt % 
dsDNA/0.1 wt % SWNT (10.0:1.0, Rice SWNT 183.6), 0.75 wt % dsDNA/0.1 wt % SWNT 
(7.5:1.0, Unidym SWNT) and 0.5 wt % dsDNA/0.1 wt % SWNT (5.0:1.0, Rice SWNT 183.6) 
were investigated using cross-polarized microscopy and rheology. Though 0.1 wt % dsDNA/ 0.1 
wt % SWNT (1.0:1.0, Rice SWNT 183.6) were also prepared, obvious agglomeration was 
observed during slow evaporation of this dispersion. This was due to the fact that insufficient 
dsDNA was present to stabilize the SWNT, which resulted in aggregate formation with 
increasing concentration. Hence, former three sets of dispersions were used for further analysis. 
Based on a SWNT density of 1.45 g/cm3, dsDNA density of 1.7 g/cm3, and a water density of 1.0 
g/cm3, these ratios corresponded to 0.59 vol % dsDNA/0.07 vol % SWNT (8.5:1.0), 0.44 vol % 
dsDNA/0.07 vol % SWNT (6.4:1.0), and 0.29 vol % dsDNA/0.07 vol % SWNT (4.3:1.0). The 
conversion of weight percent to volume percent is given in the Equation 2 in Appendix 1. In 
addition to studying the phase behavior of the bulk dispersions, dispersions of 0.44 vol % 
dsDNA/0.07 vol % SWNT were centrifuged at 17,000 g for 3 hours to remove bundles and the 
resulting supernatants of individual SWNT1 were evaporated. This resulted in a cholesteric phase 
of dsDNA/SWNT dispersion. Interestingly, repeated thermogravimetric analyses and mass 
balances on various concentrations of centrifuged dsDNA/SWNT samples showed that the 
dsDNA:SWNT volume ratio decreased from 6.4:1.0 to 4.4:1.0 (weight ratio of 5.1:1.0) in the 
69 
supernatant. Each of above dispersions were collected separately and used for slow evaporation 
at room temperature by rotating on an orbital shaker. Samples were collected at different time 
intervals and the concentrations of nanotubes were monitored by TGA for gel-like suspension 
obtained from Unidym SWNT dispersion and UV-Vis spectrophotometer for dilute dispersions 
of all kinds.  
For dsDNA/SWNT fibers, both 4.4:1.0 by volume dsDNA:SWNT supernatant 
dispersions with ~2.0 vol % SWNT and 6.4:1.0 by volume dsDNA:SWNT bulk dispersion with 
~3.4 vol % SWNT were used which were prepared by the same method given above. As for 
lysozyme/SWNT fiber spinning, dispersions of 0.4 wt % LSZ/0.4 wt % TTAB/0.4 wt % SWNT 
(non-purified Rice SWNT 187.2 with purity of ~ 64.8 % SWNT) and 0.4wt % LSZ/0.4 wt % 
TTAB/0.6 wt % SWNT (purified Rice SWNT 187.2 with purity of 99 %) were used. They were 
prepared by sonication using 3 mm diameter probe at the power level of 25 W with 0.5 s pulse 
on and 0.2 s pulse off setting. The time span used for sonicating LSZ/TTAB/SWNT 
(non-purified SWNT) is 1 hr 15 min, while for LSZ/TTAB/SWNT (purified SWNT) is 1 hr 45 
min. In addition, LSZ/TTAB/SWNT dispersions were centrifuged at 1,000 g for half an hour 
prior to use to precipitate the big bundles of SWNT which are not well dispersed.  
 
3.3. Characterization of Double-Stranded DNA  
In order to determine the length and double-stranded characteristic of DNA used, three 
sets of DNA only solutions in DI water with initial concentration of 0.75 wt % were prepared 
accordingly. Those were DNA aqueous solution before sonication, after sonication, and after 
sonication and centrifugation, respectively, using the same experimental protocol as preparing 
the DNA/SWNT dispersions. DNA fragments were determined to be 750 bp and smaller in size 
70 
after sonication by running DNA samples on 1 % agarose gel in 0.5x TAE buffer with Bionexus 
Hi-Lo DNA marker. Ethidium bromide was stained for visualization. DNA fragments were 
analyzed to be double stranded even after sonication/centrifugation by both UV-vis absorbance 
spectrophotometer and fluorometric quantification. DNA samples with starting concentrations of 
7.5 ?g/?L were diluted in TE to create different concentrations of solutions. After plating 100 ?L 
of DNA samples in duplicate on a 96 well plate, 100 ?L of picogreen working solution (25?L 
picogreen stock in 5 ml TE) was added to each well. The fluorescence of DNA assayed with 
picogreen reagent (Invitrogen, Cat#P11495) was measured by SPECTRAFluor Plus fluorescence 
spectroscopy (TECAN, Instrument serial number 94385) in relative fluorescent units (RFU). 
RFU was normalized by subtracting off TE only (blank) RFU from all measurements before 
graphing.  
 
3.4. Macroscopic Assembly of Films and Fibers 
Wet-jet solution spinning was used for fiber spinning from the cholesteric phase 
dsDNA/SWNT dispersions (~1.98 vol % of SWNT). The extrusion process was controlled by 
Harvard Apparatus (Holliston, MA) PHD 4400 Hpsi high force/high pressure programmable 
syringe pump which has average linear force of >200 lbs. A high pressure 8 mL stainless steel 
(#316 SS) syringe (0.11 in3 cross sectional area) with a 1/16? SwagelokTM fitting was secured to 
the syringe pump. Pressure in excess of 1500 psi can be achieved when this syringe is used in 
PHD 4400 Hpsi. Upchurch Scientific (Oak Harbor, WA) 1/16? outer diameter of 2? long pre-cut 
stainless steel tubing with small orifice diameters (125 ?m, 180?m) were adapted to the syringe. 
The end of tubing was submerged ~ 1 cm below the liquid surface of coagulant bath which was 
positioned below the tubing. Different coagulants were used for preliminary fiber spinning 
71 
including various alcohols and acetone coagulants, but fibers were only obtained in some of 
them such as ethanol (95 %), 2-propanol (99.8 +%), and tert-butanol. The fiber spinning process 
is depicted in Figure 3.1.  
 
 
Figure 3.1. An illustration of a typical syringe extrusion process showing the fiber extrusion and 
collection method. 
 
As for polyvinyl alcohol (PVA) coagulated LSZ/TTAB/SWNT fibers, the same fiber 
spinning setup reported in Vigolo et al.14 was used in this research (Figure 3.2). The 
LSZ/TTAB/SWNT dispersions were tangentially injected into the circular trajectory of the 5 wt% 
PVA (MW=195,000) coagulation bath rotated at 100 rpm speed. The tip of the syringe needle 
(0.5 mm diameter) was submerged into the polymer solution with 1-2 cm deep and located at a 
distance of ~2 cm from the rotation axis of the polymer solution. The alignment of the nanotubes 
F
A . S y ri n g e  P u m p
B . S ta i n l e ss S t e e l Sy ri n g e
C. E x trusio n  N e e d l e
D. Co a g u l a ti o n  B a th
E . S W N T  Fi b e r
F. I ron  W ir e  
72 
took place at the tip in the direction of fluid velocity due to the shear contribution of polymer 
flow. The PVA coated SWNT ribbon is later transferred and rinsed in DI water bath. Finally, the 
dry fibers were obtained by hang drying the ribbons in the room temperature.  
 
 
Figure 3.2. Schematic drawing of the experimental set-up used to make nanotubes ribbons. 
Adapted from Vigolo et al.14 
 
Carbon nanotube films were also prepared from the cholesteric phase of dsDNA/SWNT 
dispersions (~1.98 vol % of SWNT) by simply drying droplets of samples in ambient condition 
or vacuum oven. Films were pretreated with linear shear or no shear, on 1 cm wide copper tape. 
Nice translucent films were obtained by peeling them off the copper tape substrates after drying 
(Chapter 5). 
 
73 
3.5. Mechanical and Electrical Testing of Fibers 
Both Zwick/Roell 72.5 and Instron 5565 Calibration Lab machines were used to 
determine the mechanical properties of the fibers produced in this research. Fiber samples were 
glued into a paper testing frame with a gauge length of 26 mm as shown in Figure 3.3. The load 
cell used for this testing was 100 cN with a cross-head separation speed of 0.54 mm/min (2 % of 
initial fiber length). The load and gauge length were balanced and zeroed before each run. The 
raw test data collected for each sample were load force, extension, and time. The cross sectional 
area of fiber at break was measured by directly imaging the cross section under optical 
microscopy and calculating the irregular area using image analysis by drawing the contour line 
of the fiber cross section. The contour line of a fiber cross section is not a perfect circle, so the 
method of calculating the area by assuming an ordered shape of circle would cause a larger error 
in the mechanical properties results. The cross sectional area of fiber at each individual time was 
given by the following relationship St = Lf ? Sf / Lt based on the assumption of constant fiber 
volume and uniform deformation; Li ? Si = Lf ? Sf = Lt ? St where L is the length, S is the cross 
sectional area, i is the initial value, f is the final value, and t is the value at a particular time.    
 
74 
 
Figure 3.3. The test frame for mechanical testing of SWNT fibers. The fiber sample was secured 
onto the paper frame by droplets of super glue (Loctite? 431Adehesif Instantane, #234D67 or 
Gorilla wood glue, Cincinnati, OH). 
 
The electrical resistance of the processed fibers was measured by the 2-wire Ohms 
method using an HP 34401A Multimeter (Santa Clara, CA). The fibers were placed on slides and 
Pelco? colloidal silver paste from Ted Pella, Inc. (United States) was used to create the lead 
contacts which were separated about ~ 1mm apart. The fiber cross sectional area was obtained 
from optical microscopy.  
 
3.6. Microscopy 
Optical Microscopy. Optical anisotropy is one of the fundamental properties of liquid 
crystals, as well as crystals. The propagation of a ray of light through the medium depends on its 
orientation.120 Isotropic and anisotropic materials can be distinguished by polarized light 
The in n er sid e ed g e of  
u p p er an d  lo wer p ieces 
m u st b e align ed  
f ib er
sm all d rop let o f  glu e 
in  th e m id d le
44  mm
2 6  mm
15
 mm
75 
microscopy as the specimen is rotated relative to the crossed polarizers. The intensity of the 
polarization of isotropic materials remains permanently zero, when the circular stage is rotated 
through 360 degrees in crossed polarizers. This is called extinction, due to the samples remaining 
as dark. For anisotropic materials, typical birefringence would be observed by rotating the 
sample between crossed polarizers with the intensity of polarization vary cyclically, from zero up 
to a maximum after 45 degrees and back down to zero after a 90 degree rotation. A specimen 
will be in extinction, when the permitted vibration directions of light passing through are parallel 
with those of two polarizing filters of either the polarizer of analyzer.  
In this experiment, optical microscopy was performed on a Nikon (Melville, NY) Eclipse 
80i optical microscope with transmitted light using precleaned glass slides and coverslips for 
liquid crystal samples. Polarizing microscopy was also used for examining the birefringence 
character of translucent films obtained from cholesteric phase of dsDNA/SWNT dispersions with 
transmitted light. Both reflected and transmitted light can be used for polarized microscopy, and 
reflected light is mostly used for the study of opaque materials, for example, metals, silicon 
wafers, mineral oxides and sulphides. A Linkam (Tadworth, UK) optical rheology system ? 
CSS450 assembled with the Nikon microscope was also used for investigating the shear and 
temperature effects on the sample microstructure. The Linkam shear cell can perform steady 
shear, step rate, and oscillatory tests with shear rate and angular frequency changing in the range 
of 0.001 to 10 rads/sec. 
Scanning Electron Microscopy. The surface morphology of macroscopically assembled 
films and fibers from cholesteric dsDNA/SWNT dispersions was monitored by JEOL 7000F 
FE-scanning electron microscope (SEM) (Tokyo, Japan) with energy dispersive X-ray 
spectroscopy (EDS) after sputter coating the films and fibers with gold. 
76 
3.7. Spectroscopy 
UV-visible Spectroscopy. An Ultrospec 2100Pro UV-vis spectrophotometer was used to 
obtain UV-vis calibration curves for dsDNA/SWNT dispersions (Unidym SWNT) with 
dsDNA:SWNT ratio of 7.5:1. Samples were placed in a Starna quartz cell and pure water was 
taken as a background. The typical wavelength range was from 200-900 nm. The absorbance 
peak at 745 nm wavelength was chosen to obtain the extinction coefficient ?  of Unidym 
SWNT dispersion by checking dsDNA/SWNT with different concentrations. But for 
dsDNA/SWNT dispersions obtained from Rice SWNT, the ?  was determined by absorbance at 
763 nm wavelength. According to the Beer-Lambert Law cA ??? , the absorbance A  has 
linear relationship with the concentration c (mg/L or ppm). The concentration of the substance 
can be deduced by measuring the absorbance with known path length ?  (cm) and the molar 
absorptivity (extinction coefficient, L?mg-1?cm-1).  
Raman Spectroscopy. The spatial distribution of fibers spun from cholesteric 
dsDNA/SWNT dispersions was evaluated by Raman scattering studies which were carried out 
with Renishaw (Hoffman Estates, IL) inVia Raman microscope with a Leica 50x/0.75 NA 
?N-plan? objective and Prior (Rockland, MA) scanning stage. A Spectra-Physics (Irvine, CA) 
263 air-cooled ion laser was used for 514 nm wavelength irradiation and a SDL (San Jose, Ca) 
8530 high power laser was used for 785 nm irradiation. All measurements were taken under 
ambient conditions. 
 
3.8. Thermogravimetric Analysis 
The concentration of nanotubes in dsDNA/SWNT dispersions (Unidym SWNT) was 
measured by TA instruments (New Castle, DE) Q500 thermo gravimetric analyzer (TGA) for 
77 
concentrated gel-like suspensions. The same TGA standard profile under nitrogen environment 
was used for all samples by purging the chamber with sample gas at a constant flow rate of 60 
cm3/min with a constant nitrogen balance protection at a flow rate of 40 cm3/min. Samples 
underwent a 10 oC /min ramp to 120 oC and were isothermal for 15 min, then ramped again with 
the same speed of 10 oC /min up to 800 oC and then kept isothermal for 45 min.  
 
3.9. Rheology 
Rheological data was acquired in an Anton Paar (GmbH) Physica MCR 301 rheometer, 
which is equipped with TruGap and Toolmaster to detect and control the measuring gap in the 
geometry. Generally, for dsDNA/SWNT dispersions, a double gap geometry (i.d. = 24.644 mm, 
o.d. = 26.675 mm) was used for lower concentrations below 0.35 wt %. For the higher 
concentrations above 0.60 wt %, a parallel plate fixture (25 mm diameter) was used. For 
lyotropic cellulose nanocrystals (CNC) suspensions, Mooney-Ewart (o.d.= 22.192 mm) or 
double gap geometries were used for the low viscosity samples; 25 mm parallel plates and cone 
and plate (2o nominal angle) geometries were used for the high viscosity samples. The 
temperature of the samples was controlled by Peltier temperature control device and Peltier 
upper oven204 counter-cooled at 20 oC of circulating water bath. Samples were loaded when the 
fixtures were at 10 oC if not indicated otherwise. Care was taken during each measurement to 
ensure that the torques, deformations and speeds recorded were well within the capabilities of the 
instrument as quoted by the manufacturer. Some of the fundamental rheological parameters can 
be defined by a Two-Plate-Model as shown in Figure 3.4. The shear force F  is applied to the 
upper plate with the shear area A  and lead to the resulting movement with velocity v  and a 
deflection of s. The lower plate is stationary ( 0?v ). The distance between the plates is h . For 
78 
simple shear, motion is along one coordinate direction (x1) with velocity varying in a second 
coordinate (x2) orthogonal to the shear geometry resulting in a laminar flow in the form of 
layers.205 The shear stress, shear rate and shear viscosity at steady state are calculated as: 
AF/??                                 (3.1) 
2
1xv?????                                 (3.2) 
?
??
? ????? )()(                               (3.3) 
where, 
 ??  Shear stress (Pa) 
 ???  Shear rate (1/s) 
 ??  Steady shear viscosity (Pa?s) 
and the magnitude of the shear rate equals to hv??? . 
 
 
Figure 3.4. The Two-Plate-Model for simple shear test with relative coordinate system. 
?
h
s
Ar ea
?
1
?
3
?
2
F , ?
?
?
?
?
79 
For oscillatory tests, the upper plate is moved back and force by the shear force ?F and 
the deformation is ?? tan/ ????? hs  where ?  is shear strain (%) and ?  is deflection 
angle. Other important parameters obtained from oscillatory test are storage modulus G? and loss 
modulus G?. The G? value is a portion of the shear energy applied on a sample stored as the 
deformation energy representing the elastic behavior of a sample. While the G? value is a portion 
of energy lost to the sample as shear heat representing the viscous behavior of a sample. The 
notations of G? and G? are given below: 
)cos()( ?????G                               (3.4) 
)sin()( ??????G                               (3.5) 
where, ?  is the phase angle.  
Moreover, the overall resistance to deformation of a material during the shear process is 
measured by complex modulus G* and the overall resistance to flow under oscillatory shear is 
measured by complex viscosity ?*. The magnitudes of above parameters are defined by 
following equations: 
? ? ? ?22)( )( GGttG ??????? ??                        (3.6) 
2/122
)(
)(
???
?
???
? ?
?
??
?
? ????
?
??
?
? ???
?
? ??
?
?? GG
t
t                      (3.7) 
where, ?  is angular frequency (1/s). 
80 
When shearing a viscoelastic material between two parallel surfaces not only viscous 
shear stress T12 is applied on the fluid, but also normal stress differences N1 and N2 are present. 
The tensor Tii is defined as the sum of pressure and the extra stress as shown by Equation 3.8.  
                                iiii PT ????                             (3.8)    
since most liquids are considered incompressible, the pressure term is eliminated in the following 
expressions for the first and second normal stress differences N1 and N2; 
                            221122111 ?? ???? TTN                        (3.9) 
                            332233222 ?? ???? TTN                       (3.10) 
typically, at a negative N1, the liquid tends to force the plates together and at the positive N1 the 
liquid pushes up on the plates. Also, N1 is always positive in sign for isotropic materials and N2 is 
usually found to be negative and smaller than N1 in magnitude. Moreover, N2 tends to bulge the 
fluid at the edge of the plates.206,207 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
81 
 
 
 
 
 
Chapter 4 
Rheology and Phase Behavior of Aqueous dsDNA/SWNT dispersions  
This chapter describes the primary focus of this research and summarizes the results of 
extensive investigation into the phase behavior and rheology of aqueous dsDNA/SWNT 
dispersions in a broad SWNT concentration range from isotropic to biphasic to fully liquid 
crystalline phases. In addition, basic characterizations of materials used are also included such as 
UV-vis absorbance of SWNT in aqueous dispersions and double-stranded nature of DNA. 
 
4.1. Determining SWNT Concentration 
The equation for the conversion of SWNT concentration measured in units of milligram 
per liter or weight percent to volume percent is provided in Appendix-A. The concentration of 
SWNT (in mg/L) in dilute dispersions was determined by UV-vis spectrophotometer. In this 
experiment, for dilute dsDNA/SWNT (Unidym SWNT and Rice SWNT 187.2) dispersions 
without centrifugation, UV-vis absorbance at 745 nm wavelength with extinction coefficient (? ) 
of 0.0257 L/(mg?cm) was determined. But for the dsDNA/SWNT dispersion obtained from Rice 
SWNT 183.6, ?  0.0247L/(mg?cm) at 763 nm wavelength was used due to the occurrence of 
Van Hove singularity peak at this particular wavelength (Figure 4.1). However, ?  of the 
supernatant differed dramatically from that of the bulk (non-centrifuged) dispersions due to the 
elimination of SWNT bundles.  
82 
 
Figure 4.1. UV-vis calibration curves for dsDNA/SWNT dispersions with dsDNA:SWNT weight 
ratio of 7.5:1 without centrifugation using Unidym SWNT. 
 
The extinction coefficient was also found to be dependent on the volume of sample 
sonicated. Figure 4.2 illustrates the calibration curves obtained from supernatant of 
dsDNA/SWNT dispersions prepared from different initial volumes. These results highlight the 
importance of verifying extinction coefficients, particularly if the results are to be extrapolated to 
samples that were concentrated by orders of magnitude. The concentration of SWNT in 
supernatant was determined by a mass balance assuming dsDNA:SWNT ratio in the sediment 
after centrifugation is the same as in initial dispersion which is 7.5:1. Figure 4.3 shows the full 
UV-vis spectra of supernatant dispersions with increasing SWNT concentrations. The van Hove 
singularity peaks obtained between 500 to 800 nm of wavelength indicate that SWNT were 
dispersed predominantly as individuals.1   
 
A t 7 4 5  n m , 
y  =  0 .0 25 7x
R
2
=  0 .9 9 9 8
A t 7 6 3  n m ,
y  = 0 .0 2 4 7 x
R
2
=  0 .9 9 9 8
0
30
60
90
120
150
180
0 2000 4000 6000 8000
A
bsorb
anc
e 
S W NT Con c e nt ra tion (mg/L)
74 5 nm
7 6 3  n m
83 
 
Figure 4.2. UV-vis calibration curves for supernatant dsDNA/SWNT dispersions with different 
initial volumes preparation using Unidym SWNT. The concentration of SWNT in supernatant 
was calculated by mass balance assuming the dsDNA:SWNT ratio in sediment is 7.5:1. 
 
 
Figure 4.3. UV-vis spectra for supernatant of dsDNA/SWNT dispersions with dsDNA:SWNT 
weight ratio of 5.1:1.0 at different SWNT concentrations using Unidym SWNT. 
10 0 m l  i ni t i al  di spe r si on ,
y  = 0.0009 x
R
2
= 1
50  ml i ni t i al  di sp ersi on ,
y  = 0.0041 x
R
2
= 1
0
1
2
3
4
0 200 400 600 800 1000
A
bs
or
ba
nc
e
 
a
t 
7
4
5
 nm
S W NT Con c e nt ra tion (mg/L)
1 0 0  m l
5 0  m l
0
0 .1
0. 2
0 .3
0 .4
0 .5
300 500 700 900
A
bsor
banc
e
W a v e le ngt h (n m)
0 .0 6  v o l %
0 .1 0  v o l %
0 .1 2  v o l %
0 .1 3  v o l %
0 .1 8  v o l %
0 .2 1  v o l %
0 .2 9  v o l %
0 .3 2  v o l %
84 
The concentration of SWNT (in wt %) in evaporated samples of dsDNA/SWNT 
dispersions with relatively high concentration was determined by TGA analysis. In order to 
quantify the SWNT concentration with low error, the TGA of dsDNA raw material without 
dissolving in water was also measured (Figure 4.4). The weigh loss of dsDNA at 124.3 oC was 
~22.6 % and at 206.0 oC was ~24.3 %, respectively, under the standard profile described in 
section 3.8. Previous researchers194 showed that the relative weight losses of their dsDNA 
sample under nitrogen from salmon sperm at 124.2 oC was -8.0 % and at 206.8 oC it was -10.6 %. 
It was shown that dsDNA undergoes a significant decomposition at 175 oC and followed by a 
more dramatic change starting at 205 oC. The maximum decomposition rate occurs at 220 oC and 
above 205 oC the DNA pyrolysis occurs.194 In this research, the initial weight loss at 124.3 oC 
was -22.6 % which is much higher than -10.6 %; this might be due to the dsDNA sample having 
more water absorption. Also, the sample was held isothermal at 120 oC for 15 min at which point 
all the water would be vaporized. However, from 124-206 oC the weight loss percentages have 
small differences both are around ~2 %. All the comparison data are summarized in Table 4.1. 
 
85 
 
Figure 4.4. Thermogravimetric analysis (TGA ) of dsDNA. 
 
Table 4.1. Comparison of TGA data of dsDNA. 
Samples adsDNA bdsDNA 
Weight loss at 124 oC (%) -22.6 -8.0 
Weight loss at 206 oC (%) -24.3 -10.6 
Ash at 600 oC (%) 41.2 48 
Ash at 800 oC (%) ~38 ~ 
adsDNA used in this experiment. 
bdsDNA used in reference paper.194 
 
 
 
86 
As shown in Table 4.1, approximately 38 % of dsDNA ash remained at 800 oC. However, 
with sonication assistance during the preparation of dsDNA/SWNT dispersion, the dsDNA ash at 
800 oC in the system could differ from the value of 38 % for dsDNA raw material. Therefore, 
mass balance of dsDNA/SWNT dispersion by tracking the masses of initial samples before 
centrifugation, sediment and supernatant after centrifugation, showed that the dsDNA ash at 800 
oC in the supernatant of concentrated dsDNA/SWNT suspension should be lower than 38 %. In 
order to determine the percentage of dsDNA ash at the end of TGA, a 0.75 wt % dsDNA-0.1 wt % 
SWNT (Unidym SWNT) dispersion was prepared by only sonication and evaporated to obtain 
7.65 wt % dsDNA-1.02 wt % SWNT suspension. The resulting concentrated suspension was 
analyzed by TGA under nitrogen with several runs as shown in Figure 4.5 to indicate the final 
weight percents remaining.  
 
 
Figure 4.5. TGA profiles of 7.65 wt % dsDNA-1.02 wt % SWNT suspensions. 
 
0
20
40
60
80
100
0 200 400 600 800 1000
W
e
ight 
%
T empera t ure ( ? )
Run  1  - 3 .7 1  w t.  %
Run  2  - 2 .8 6  w t.  %
Run  3  - 2 .6 1  w t.  %
Run  4  - 2 .8 3  w t.  %
R un  5 - 2. 67  w t. %
Run  6  - 2 .8 1  w t.  %
87 
Therefore, three variables and three equations were used to determine the percentage of 
dsDNA ash at 800 oC. At 150 oC, the mass is sum of dsDNA and SWNT, since water is all 
evaporated. At 800 oC, the mass is sum of remaining dsDNA and SWNT. SWNT mass does not 
change during the whole process. Again, the dsDNA:SWNT weight ratio is 7.65:1.02. Example 
calculation is shown below; 
dsDNA mass=x (mg), SWNT mass=y (mg), percent of dsDNA remaining=z % 
For run 2: total sample put in=26.6549 mg; final weight % left=2.86 %; 
at 150 oC, sample weight %=8.01 %. 
So,    at 150 oC     x+y=26.6549?8.01 %=2.14 
      at 800 oC    ( z %)?x+y=26.6549?2.86 %=0.76 
                          x=7.48?y 
After calculation;            x =1.88; y =0.25; z % =27.08 % 
All the TGA data of 7.65 wt % dsDNA-1.02 wt % SWNT dispersion were calculated 
with the same method and the results are shown in Table 4.2. The results for run 1 were 
significantly dfferent and not included. The average dsDNA ash percentage at 800 oC calculated 
from run 2-6 were chosen and took the average as z %=27.0 %. Thus, SWNT concentrations of 
gel-like dsDNA/SWNT (Unidym SWNT) dispersions were calculated by assuming 27.0 % of 
dsDNA remaining at the end of TGA burning. 
 
Table 4.2. Percentage of dsDNA ash at 800 oC. 
7.65 wt % dsDNA-1.02 wt % SWNT (Unidym SWNT) dispersion 
Run 1 2 3 4 5 6 
z (%) 41.10 27.08 26.40 27.22 25.13 29.20 
88 
4.2. DNA Characterization 
The majority of DNA/SWNT dispersion research to date has been performed using single 
stranded DNA (ssDNA). As for the dispersion of SWNT in ssDNA, dispersion of SWNT in 
dsDNA requires sonication; this results in high localized temperatures and shear. In this work, an 
ice bath was used to prevent significant heating of the bulk dispersion. In order to investigate 
whether sonication and/or subsequent centrifugation affected the dsDNA structure, dispersions 
of dsDNA before and after these processing steps were analyzed using fluorescence spectroscopy 
and gel electrophoresis. DNA has a characteristic UV absorbance peak in a band centered around 
260 nm. In order to determine the length and double-stranded characteristic of DNA used, three 
sets of DNA solutions (in DI water) with initial concentration of 0.75 wt % were prepared 
accordingly: DNA aqueous solution before sonication, after sonication, and after sonication and 
centrifugation. The absorbance of DNA solutions at 260 nm was almost the same before and 
after sonication/centrifugation (Figure 4.6). Since the hyperchromicity of ssDNA results in a 40% 
increase in absorbance relative to dsDNA,72 the absence of any change in the absorbance spectra 
supports that the solutions contained primarily dsDNA. The absorbance ratio at 260 nm and 280 
nm was determined for the DNA raw material. This is a typical test for purity; the measured ratio 
A260/A280 = 1.86 was in the range of 1.8 - 2.0 which is generally considered an indication of pure 
DNA.72 The A280 corresponds to the maximum UV light absorbance of protein.208 However, 
DNA molecules were found to be cut into smaller pieces of less than 750 bp during sonication by 
agarose gel electrophoresis (Figure 4.7). Typically, one base pair equals to 3.4 ?s. So, the 
molecular length L of DNA after sonication is around 2550 ?, which is 255 nanometers. The 
aspect ratio L/d of DNA after sonication is 127.5, considering the diameter d is 2 nm. Further 
indication that DNA molecules remained double stranded after sonication/centrifugation was 
89 
given by fluorescence spectroscopy. It has been shown that picogreen reagent preferentially 
binds dsDNA as opposed to ssDNA.209 If the DNA processing was creating a large amount of 
ssDNA, there should be a significant decrease in the picogreen relative fluorescence units (RFU) 
before and after the processing. However, no substantial change occurred in the picogreen RFU 
between the DNA samples before and after sonication/centrifugation (Figure 4.8). The Coleman 
group has shown that dsDNA coverage of SWNT bundles occurred over time and most of the 
nanotube could be covered in about 21 days after preparation of dsDNA/SWNT with sonication 
assistance and almost all of the nanotube walls were coated with a thin monolayer of DNA by 
day 35. During this progressive debundeling and coating process, SWNT binding dsDNA may 
denature on the nanotube surface, however, this is a relatively slow process.107 Considering that 
our total evaporation process and characterization of dsDNA/SWNT dispersion were 
accomplished in a period of four weeks, it is reasonable to conclude that the majority of the DNA 
was retained as double-stranded DNA.  
 
 
 
 
90 
 
Figure 4.6. UV-Vis Absorbance spectra for DNA only solutions before and after 
sonication/centrifugation which were prepared from 0.75 wt % initial concentration of DNA in 
DI water. 
 
 
Figure 4.7. Agarose gel electrophoresis of DNA samples in DI water including (1) 7.5 ?g salmon 
sperm genomic DNA (gDNA) before sonication, (2) after sonication, (3) after 
sonication/centrifugation, (4) 10 ?L HiLo DNA marker (Bionexus) and (5) 15 ?g salmon sperm 
gDNA before sonication, (6) after sonication, (7) after sonication/centrifugation. 
0 .0  
0 .5  
1 .0  
200 250 300 350 400
A
bs
or
ba
nc
e
W av el engt h ( nm)
n o  so n icat ion
so n icat e d
so n icat e d  &  ce n trif u g e d
91 
 
Figure 4.8. Fluorescence quantitation of comparing DNA solutions before and after 
sonication/centrifugation using picogreen. 
 
4.3. Phase Behavior and Rheology of dsDNA/SWNT Dispersions 
In this section, the phase behavior of dsDNA/SWNT dispersions produced by two 
different methods are compared. Evaporation of bulk dispersions, comprised of both bundled and 
individual SWNT, resulted in nematic phase formation while evaporation of dsDNA/SWNT 
supernatants resulted in cholesteric phase formation. To the author?s knowledge, this is the first 
lyotropic cholesteric SWNT liquid crystal with interaction of a biopolymer. The origin of the 
cholesteric structure can be directly attributed to the majority of the dsDNA retaining its helical 
structure during processing. This structure can be retained or eliminated during processing into 
aligned films. The potential enhanced biocompatibility of dsDNA/SWNT and the previously 
never achieved cholesteric microstructure suggest that the range of applications that can be 
processed from liquid crystalline nanotube dispersions may be even broader than previously 
thought. 
1
10
100
1000
10000
100000
3 . 8 3 7 . 5 3 7 5 . 0
E st i m a t e d  C o n cen t r a t i o n  o f  S e r i a l  D i l u t i o n s ( n g / m l )  
R
F
U
N a t i v e
S o n i c a t e d
S o n i c a t e d  &  C e n t r i f u g e d
10
1 
10
0 
10
5 
10
3 
10
2 
10
4 
.
i a t e d  C o n cen t r a t i o n  o f  S e r i a l  D i l u t i o n s ( n g / m l )  
R
F
U
t i v e
i c a t e d
i c a t e d  &  C e n t r i f u g e d
R
F
U
R
F
U
92 
4.3.1. Phase Behavior of dsDNA/SWNT Dispersions  
Slow evaporation of the dispersions increased the concentration of both dsDNA and 
SWNT. In addition, marked transitions from isotropic to concentrated anisotropic dispersions 
were observed. Similar to the findings of Badaire et al.16 for ssDNA/SWNT, the phase behavior 
of the bulk (non-centrifuged) dispersions was dependent on the dsDNA/SWNT ratio; if 
insufficient dsDNA was present to stabilize the SWNT, increasing concentration resulted in 
aggregate formation. In this work, evaporation of a dispersion of 1.0:1.0 dsDNA/SWNT by 
weight (0.9:1.0 by volume) resulted in the formation of visible aggregates tens to hundreds of 
microns in diameter due to the presence of an insufficient amount of dsDNA to counteract the 
inherent 20 - 40 kBT/nm,45,46 van der Waals attraction between SWNT. In contrast, Badaire et al. 
(2005) observed sufficient stabilization to enable liquid crystalline phase behavior for 1:1 by 
weight dispersions of ssDNA/SWNT. Since impurities are known to stabilize SWNT dispersions 
by creating physical impediments to reagglomeration, this discrepancy is largely attributed to the 
present work using SWNT batches with purity ranging from 92.0 to 99.7 % while Badaire et al. 
used SWNT with 80 % purity.16 Differences between the type of DNA, the dispersion 
preparation, and the SWNT size distribution and surface chemistry also likely contributed to 
differences in phase behavior.   
Evaporation of the dispersions containing an excess of dsDNA resulted in the phase 
behavior typical of dispersions of rods.47 The dispersions were isotropic at low concentration. 
With increasing concentration (evaporation) the systems became biphasic, a liquid crystalline 
phase formed in equilibrium with the isotropic phase. The fraction of birefringent liquid crystal 
domains increased with concentration until the systems became completely liquid crystalline at a 
critical volume fraction designated ?LC. Figure 4.9a shows that in the biphasic region, these 
93 
dispersions exhibited a strand-like spaghetti structures similar to that observed for SWNT in 
superacids.19 These strands are nematic liquid crystalline domains where the SWNT are free to 
translate along the length of the strands. In contrast to the bulk dispersions, the initial 
birefringent morphology for the dsDNA/SWNT supernatant was narrow moving streams of 
near-parallel lines indicative of the initial stage of cholesteric liquid crystal formation (Figure 
4.9b).   
 
 
Figure 4.9. dsDNA/SWNT dispersions of (a) 13.6:1.0 by vol. dsDNA:SWNT (15.9:1.0 by 
weight) without centrifugation at a SWNT concentration of 0.24 vol % and (b) supernatant of 
4.4:1.0 dsDNA:SWNT at a SWNT concentration of 0.69 vol % under cross-polarized light on a 
Nikon Eclipse 80i with Plan Apo VC 60x/1.4 NA oil immersion objective with 60x 
magnification while image (a) has 2.0 magnification in front of the camera. The scale bars in 
images are (a) 10 ?m and (b) 30 ?m. 
 
 
94 
As shown in Figure 4.10, the differences in morphology between the concentrated bulk 
dispersions and concentrated supernatants become even more dramatic at higher concentrations. 
The polydomain nematic structure shown in Figure 4.10a for a concentrated bulk 
(non-centrifuged) dispersion of 4.3:1.0 dsDNA:SWNT at a concentration of 1.7 vol % SWNT 
has a similar microstructure to that previously observed for liquid crystals of SWNT/102% 
H2SO4,19,20 MWNTox/H20,15,190 and ssDNA/SWNT.16 The SWNT are locally aligned within 
each domain but are randomly oriented on a larger length scale.  
 
 
Figure 4.10. Comparison of dsDNA/SWNT dispersions of (a) non-centrifuged 4.3:1.0 by volume 
dsDNA:SWNT at a SWNT concentration of 1.7 vol % SWNT and (b) supernatant of 4.4:1.0 by 
volume dsDNA:SWNT at 2.3 vol % SWNT under cross-polarized light on a Nikon Eclipse 80i 
with Plan Apo VC 60x/1.4 NA oil immersion objective with 60? magnification. 
 
 
 
95 
The individual domains of polydomain nematic liquid crystal of bulk dsDNA.SWNT 
dispersion became bright and dim as the samples were rotated between cross polarizers as shown 
in Figure 4.11. However, these domains are randomly oriented on larger length scale (~100 ?m). 
For these dispersions, the critical concentration ?LC for the biphasic to liquid crystal transition 
depended on the initial dsDNA:SWNT ratio. Cross-polarized optical microscopy was initially 
used to approximate this transition by determining the lowest concentration at which no isotropic 
domains were observed. Since this method is somewhat dependent on the magnification and the 
concentrations tested, rheology was also used for further confirmation of liquid crystalline phase 
behavior and to substantiate the estimated value of ?LC.  
 
 
Figure 4.11. Concentrated bulk dispersion of dsDNA:SWNT = 8.5:1.0 (10.0:1.0 by weight) at 
0.8 vol % SWNT under cross-polarized light using a Nikon Eclipse 80i with Plan Apo VC 
60x/1.4 NA oil immersion objective with 60? magnification. The domains become bright and 
dark as the sample is rotated at different angles of (a) 0o, (b) 30o, (c) 90o, respectively. The scale 
bar in the figures is 30 ?m. 
 
 
96 
In a normal polymer solution or colloidal dispersion, the low shear viscosity increases 
with concentration. However, for liquid crystalline dispersions, viscosity does not change 
monotonically with increasing concentration. The viscosity increases with concentration as long 
as the system is predominantly isotropic. However, the increasing fraction of anisotropic 
domains eventually results in a decreased resistance to flow. Viscosity therefore decreases with 
concentration until increasing rod concentration no longer results in increased order. Once the 
order can no longer be increased, the viscosity once again increases with concentration. 
Therefore, in the biphasic region, the viscosity versus concentration curve goes through a 
maximum and then reaches a minimum at approximately ?LC. Based on the minima shown in 
Figure 4.12, bulk dispersions evaporated from a 8.5:1 dsDNA:SWNT have ?LC ~ 1.2 vol % while 
those from 4.3:1.0 dsDNA:SWNT have ?LC ~ 1.9 vol %. This may indicate the presence of more 
bundles in the 4.3:1.0 dsDNA:SWNT; since a lower average aspect ratio would result in a higher 
value of ?LC. 
 
 
 
97 
 
Figure 4.12. Relationship between viscosity and concentration at a shear rate of 0.1 s-1 and a 
temperature of 10 oC for bulk (non-centrifuged) dsDNA/SWNT dispersions obtained from 
8.5:1.0 (solid squares) and 4.3:1.0 (open squares) dsDNA:SWNT by volume. Error bars are 
given for all points, but are smaller than the data markers in many cases. The largest error bars 
are in the biphasic region at the peak in viscosity; this is consistent with previous research on 
lyotropic rod-like polymers and SWNT in superacids.  
 
As shown in Figure 4.10b, the liquid crystal formed by concentrating the dsDNA/SWNT 
supernatants has a dramatically different microstructure than previously obtained lyotropic 
SWNT liquid crystals. The classical fingerprint texture120 consisting of dark and light strips 
under cross-polarized light microscope indicate a cholesteric structure in which the molecules are 
respectively normal to the plane of polarizer and in the plane of the polarizer.33,120 Since the 
helical cholesteric pitch is the distance when the director rotates by 2?, with the range of 
few-hundred ?s to micrometers, it can be determined easily by measuring the distance between 
either two bright or dark lines in the micrograph.137 Figure 4.13 shows the evolution of the 
0
1000
2000
3000
4000
0 0 .5 1 1 .5 2 2 .5 3 3 .5
V
is
c
osi
ty
 (Pa
.s
)
S W NT (v ol % )
98 
cholesteric microstructure and dispersion viscosity with increasing concentration. A cholesteric 
oily streak texture of thin near-parallel lines started to appear at 0.7 vol % of SWNT (3.6 vol % 
dsDNA) due to the amount of excess, or free, dsDNA which was not strongly interacting with 
the SWNT. Large birefringent domains could be observed at 1.5 vol % SWNT where the 
viscosity was at its maximum value of 2165     . Still further increasing the concentration, 
resulted in the domains of near-parallel set of swirling lines growing larger and the appearance of 
many compactly distributed small spherical swirling lines resembling fish-scales. Areas without 
obvious fingerprint textures resembled oil paint and also exhibited strong birefringence under 
polarized light. Further increasing concentration resulted in extending birefringent domains; 
iridescent colors appeared at the concentrations near the viscosity minimum at 2.2 vol % SWNT. 
Moreover, when the polarizer was extracted all these textures disappeared and the sample was a 
uniform gray. This indicated uniform dispersion throughout the sample without any noticeable 
aggregates or clusters. When rotating the specimens relative to the polarizers, different domains 
of the cholesteric phases changed colors resulting white, beige, orange, purple, light green, and 
blue regions (Figure 4.14). This was due to Pseudo Bragg reflections which are one of the 
characteristic optical properties of a cholesteric phase.120  
 
99 
 
Figure 4.13. Relationship between viscosity and concentration for supernatants of 4.4:1 by 
volume dsDNA:SWNT at a shear rate of 0.1 s-1 and temperature of 10 oC. Images were taken 
under cross-polarized light using a Nikon Eclipse 80i with Plan Apo VC 60x/1.4 NA oil 
immersion objective with 60? magnification. The SWNT concentrations were (a) 0.69 vol %, (b) 
1.5 vol %, (c) 2.0 vol %, (d) 2.3 vol %, (e) 2.5 vol %, and (f) 3.4 vol %. The scale bars are 30 
?m.   
100 
 
Figure 4.14. Concentrated supernatant of 4.4:1.0 by volume dsDNA:SWNT (5.1:1.0 by weight) 
at 2.5 vol % SWNT under cross-polarized light on a Nikon Eclipse 80i with Plan Apo VC 
60x/1.4 NA oil immersion objective with 60? magnification. The domain colors vary as the 
sample is rotated at different angles of (a) 0o, (b) 40o, (c) 100o, respectively. The scale bars are 30 
?m.   
 
The color variation became more obvious with increasing concentration in response to 
variations in the helical pitch. The pitch of the dsDNA/SWNT dispersions decreased linearly 
from 3.45 ? 0.25 ?m to 2.65 ? 0.15 ?m when the concentration of SWNT increased from 1.2 vol % 
to 2.0 vol %; this is in the transition range from biphasic to single-phase liquid crystal, and 
indicates a more ordered alignment of molecules in the cholesteric phase with increasing 
concentration. With a further increase in concentration to 3.2 vol %, the pitch increased to 3.38 ? 
0.53 ?m. Moreover, Figure 4.15 shows the relationship between pitch and concentration of 
cholesteric dsDNA/SWNT dispersion which fits well into the Equation 2.6 given in the section 
2.4. Due to the nonmonotonic feature, two different sets of constants A and k were obtained. For 
dsDNA/SWNT dispersions at lower SWNT concentrations, k = 0.498; A = -1.328 were obtained, 
while for samples with higher SWNT concentrations, k = -0.494; A = -0.657 were obtained, 
respectively. It is difficult to measure the cholesteric pitch at lower concentrations due to moving 
101 
streams of near-parallel lines. At even higher concentrations, the fingerprint texture was absent 
due to the development of a more solid-like phase; this is supported by a plateau in rheological 
measurements of storage modulus versus frequency for these dispersions. Kornyshev and 
Leikin?s theory for cholesteric liquid crystals of dsDNA,141 predicts that cholesteric unwinding 
would occur at the cholesteric to isotropic and cholesteric to hexagonal transition due to the 
weakening of biaxial correlation between dsDNA molecules. Van Winkle et al.33 observed that 
the cholesteric pitch of aqueous solutions of dsDNA unwinds from 2.2 ?m to at least 10 ?m 
before forming the hexagonal phase. Thus, considering the relatively small increase in 
dsDNA/SWNT pitch, the LC phase was still in cholesteric region at 2.9 vol % of SWNT. 
 
102 
 
Figure 4.15. Plots of ln(1/P) of 4.4:1.0 by volume supernatant dsDNA/SWNT against ln(c) of 
SWNT concentration at (a) 1.2, 1.3, 1.5, and 2.0 vol %, and (b) 2.0, 2.3, 2.5, 2.6, and 3.2 vol %.  
 
 
 
y  = 0 .4 9 7 7 x  - 1 .3 2 8 3
R?  =  0 .9 6 8 5
- 1. 3
- 1 .2
- 1 .1
- 1
- 0 .9
- 0 .8
0 0 .2 0 .4 0 .6 0 .8
ln 
(1/P
) 
(?
m)
ln (c) (SW NT v ol %)
y  = - 0 .4 9 4 3 x  - 0 .6 5 6 5
R?  =  0 .9 38 7
- 1 .3
- 1 .2
- 1 .1
- 1
- 0 .9
- 0 .8
0. 6 0. 8 1 1. 2 1. 4
ln 
(1
/P
) 
(?
m
)
ln (c) (SW NT v ol %)
(a)
(b)
103 
The surprising difference between the microstructures obtained from concentrated bulk 
dispersions and concentrated supernatants is related to the cholesterogenic nature of dsDNA 
(Figure 4.16a) and the initial dispersion state of the SWNT. Aqueous dsDNA solutions prepared 
by the same procedure, but without the addition of SWNT, exhibited cholesteric phase behavior. 
They became biphasic at ????vol % and fully liquid crystalline at ?LC ~ 12.6 vol %. These values 
are markedly higher than those obtained in the presence of SWNT; this indicates that the 
presence of the higher aspect ratio more rigid SWNT facilitates liquid crystalline phase 
formation. The pitch of the dsDNA solutions monotonically decreased from 4.35?  0.38 to 3.02 
?  0.20 ?m while the concentration increased from 8.1 to 19.3 vol %; these values are slightly 
higher than those for the concentrated dsDNA/SWNT. The structure and pitch of dsDNA 
cholesterics are dependent on many factors such as molecular length of dsDNA fragments, 
interaxial spacing of molecules and the solvent condition like osmotic pressure and ionic strength. 
In addition, steric, van der Waals, and hydration forces may also contribute to macroscopic 
properties of dsDNA cholesterics.140,141,210 In the ?d virus cholesteric system, the authors showed 
the typical behavior of cholesteric pitch decreasing with increasing concentration in a power law 
function until saturates at a certain value in solutions with different ionic strength. Slight 
unwinding of the pitch occurred as the concentration further increased to approach smectic 
transition. Rapid unwinding of the pitch was also observed as the temperature was increased.137 
As for dsDNA, different results were provided for cholesteric pitch measurement in previous 
studies. On the one hand, cholesteric unwinding at the cholesteric to isotropic and cholesteric to 
even higher density phase-hexagonal were reported.140,141 On the other hand, Van Winkle at al. 
reported that the pitch was relatively independent of the dsDNA concentration for a wide range 
104 
(150-270 mg/ml), but increased with concentration approaching the cholesteric to high-density 
phase transition.139 
In the presence of SWNT, there appears to be a competition between the tendency of the 
dsDNA to form a cholesteric liquid crystal and the tendency of the SWNT to form a nematic 
liquid crystal. When the initial dispersion is a dsDNA/SWNT supernatant consisting of 
individual SWNT,1 the natural tendency of dsDNA to form cholesteric phases dominates; this 
indicates that the majority of dsDNA retains its helicity over time. The majority of bulk 
dispersions exhibited the previously described polydomain nematic texture that has been 
characteristic of all lyotropic carbon nanotube liquid crystals to date. In the case of some bulk 
dsDNA/SWNT dispersions, such as the 6.4:1.0 dispersion concentrated to 1.3 vol % shown in 
Figure 4.16b, small fingerprint regions were observed due to the excess amount of free dsDNA. 
However, these small regions did not form continuous domains and were not visible at higher 
concentration. We believe that in the bulk dispersions, the presence of SWNT bundles physically 
impedes the formation of continuous cholesteric domains. In addition, Cathcart et al.107 have 
suggested that dsDNA denatures in the process of debundling SWNT over a time period of 
approximately thirty days. Therefore, the presence of SWNT bundles in the bulk dispersion may 
serve to result in some dsDNA denaturation during the evaporation process and reduce the 
driving force for cholesteric phase behavior.    
 
 
 
105 
 
Figure 4.16. Fingerprint texture in (a) 19.3 vol % of dsDNA solution and (b) bulk 6.4:1.0 by 
volume dsDNA/SWNT dispersion at a SWNT concentration of 1.3 vol % under cross-polarized 
light on a Nikon Eclipse 80i with Plan Apo VC 60x/1.4 NA oil immersion objective with 60? 
magnification. The scale bars are 30 ?m.   
 
4.3.2. Rheology of dsDNA/SWNT Dispersions 
Steady Shear and Dynamic Viscosities. Before running the test of flow curve for steady 
shear viscosity, a step rate test was performed to determine the time required to achieve 
steady-state flow. The duration of transient obtained from this experiment is relatively short 
when compared to those of characteristic liquid crystalline polymers which are usually as large 
as 100 shear units, or even more. The shear unit is equal to the product of multiply of shear rate 
and time     . The start up of shear flow of both dsDNA/SWNT supernatant and bulk dispersions 
and dsDNA only solutions used as references were provided in the Appendix-B. The rheological 
behavior of dsDNA/SWNT supernatant dispersions at isotropic SWNT concentrations is shown 
in Figure 4.17. In the dilute regime with SWNT concentration smaller than 912 ppm, the 
zero-shear viscosity increases slightly as a function of concentration (Figure 4.17a). When the 
106 
SWNT concentration enters into semidilute regime around 912 ppm, the zero-shear viscosity 
increases rapidly as a function of concentration in a relationship of    ?  (Figure 4.17b). 
This result agrees with the prediction proposed by the Doi-Edwards theory for semidilute 
solutions of rods.47 Above a SWNT concentration of 4956 ppm, the zero-shear viscosity 
increases more significantly with concentration and deviates from the theory prediction given for 
semidilute rods. This can be an indication of the isotropic concentrated regime. According to 
steady shear viscosity obtained for higher concentration dispersions (Figure 4.19), the range of 
SWNT concentration where the isotropic concentrated regime exists would be starting from 4956 
ppm to a concentration value much less than 8540 ppm.      
 
 
 
 
107 
 
Figure 4.17. (a) Steady shear viscosity of isotropic dsDNA/SWNT (Rice 187.2)supernatant 
dispersions as a function of shear rate for various SWNT concentrations ranging from dilute ( ? 
912 ppm) to semidilute (912 ? 3140 ppm) to concentrated isotropic (4956 to   8540 ppm). (b) 
Zero-shear viscosity as a function of SWNT concentration. 
1
10
100
1000
10000
0 .0 1 0 .1 1 10 100
? 
(
mP
a
.s
)
S hea r Rate  (1/s )
3 8 8  ppm
4 4 1  ppm
5 5 6  ppm
7 0 5  ppm
9 1 2  ppm
1 5 7 6  ppm
2 0 1 2  ppm
2 5 0 5  ppm
3 1 4 0  ppm
4 9 5 6  ppm
(a)
(b)
1
10
100
1000
10000
100 1000 10000
V
is
c
osi
ty
 (mP
a
.s
)
Con c e nt ra tion (p pm)
y  = 9E - 09x
3 . 0 4 9 6
R ?  = 0. 9955
Sl o p e = 3
108 
In addition, the viscosities of aqueous dsDNA solutions are shown in Figure 4.18a. These 
concentrations correspond to those used in the dsDNA/SWNT dispersions. This data therefore 
provides insight into the effect of the aqueous dsDNA solvent viscosity on the dsDNA/SWNT 
supernatants. The viscosity of dsDNA only solvents increased slightly with increasing dsDNA 
concentration from 0.33 to 1.36 vol %. In addition, dsDNA only solvents in this particular 
concentration range behaved as Newtonian fluids in the measured shear rate range from 0.1 to 
100 s-1. Figure 4.18b shows the plots for intrinsic viscosity [?] = (? - ?s)/( ?s ?) of supernatant 
dsDNA/SWNT dispersions at various SWNT concentrations as a function of shear rate; ? is the 
viscosity of supernatant dsDNA/SWNT dispersion, ?s is the  relative dsDNA only solvent 
viscosity, and ? is the volume fraction of SWNT. For supernatant of dsDNA/SWNT dispersions 
at dilute regime from 388 to 912 ppm SWNT, the intrinsic viscosities were close to each other. 
However, significant decrease in intrinsic viscosity was observed at higher SWNT concentration 
of 1576 ppm, which again confirmed that this particular concentration was in the semidilute 
regime. According to Kirkwood-Auer-Batchelor (KAB) relationship for a dilute dispersion of 
Brownian rods, which is given in the Equation 2.14 in the section 2.8.1, the average aspect ratio 
      of rods in dilute dispersions can be calculated from their intrinsic viscosity. Generally, the 
intrinsic viscosity was determined from a master curve created by dilute dispersions at various 
concentrations to minimize the error. As in Figure 4.18b, although no shear thinning behavior 
existed in the measured shear rate range, by making an assumption that the viscosity plateau of 
dsDNA/SWNT dilute dispersions was reflective of KAB rod behavior, a SWNT aspect ratio 
      of 233 was obtained according to KAB equation. However, there is a marked discrepancy 
between this calculated average aspect ratio value and that measured by atomic force microscopy 
(AFM). The       of SWNT (Rice 187.2), which is used in this part of experiment, was 
109 
determined to be 826 by AFM. The significant difference in       can be caused by the 
inability to access the shear thinning region at high shear rates, which made it impossible to 
know if samples were following KAB rod behavior. In addition, just changing a couple of tubes 
counted by AFM can dramatically change the average aspect ratio measured.149  
 
 
 
 
 
 
110 
 
Figure 4.18. (a) dsDNA only solvent viscosity ?s versus shear rate at concentrations 
corresponding to the same dsDNA concentrations in supernatant of dsDNA/SWNT (Rice 187.2) 
dispersions at 388, 705, 912, and 1576 ppm SWNT. (b) Intrinsic viscosity versus shear rate for 
supernatant dsDNA/SWNT dispersions in the dilute regime (from 388 to 705 ppm SWNT) and 
semidilute regime (1576 ppm). 
1
10
100
0 .1 1 10 100
?
s
(
mP
a
.s
)
S hea r Rate  (1/s )
0 .3 3  v o l % 0 .6 0  v o l %
0 .7 8  v o l % 1 .3 6  v o l %
( a)
( b)
100
1000
10000
0 .1 1 10 100
Intrins
ic
 
V
is
c
osi
ty
S hea r Rate  (1/s )
3 8 8  p p m 7 0 5  p p m
9 1 2  p p m 1 5 7 6  p p m
111 
The steady shear viscosities of cholesteric phase forming dsDNA/SWNT dispersions with 
increasing SWNT concentration are shown in Figure 4.19. Generally, the viscosity kept 
increasing monotonically with SWNT concentration from 0.26 to 1.47 vol % where the local 
viscosity maximum occurred in the biphasic regime (Figure 4.19 a, b). With further increase in 
concentration to 2.22 vol % SWNT, the viscosity started to decrease due to the formation of 
more ordered alignment of the mesogens into liquid crystals. After the system transitioned into a 
fully liquid crystalline phase, the viscosity once again increased continuously with SWNT 
concentration (Figure 4.19c). This nonmonotonic relationship of viscosity as a function of 
concentration is a typical rheological signature of lyotropic LCPs. As shown in Figure 4.19a, at a 
low concentration of SWNT at 0.26 vol %, the dispersion behaved as Newtonian fluid. When the 
concentration increased to 0.35 vol % SWNT, the sample started to show shear thinning behavior 
at a shear rate of 2.0 s-1 after a wide viscosity plateau. At 0.60 vol % SWNT concentration, the 
steady shear viscosity showed a Newtonian plateau up to 0.03 s-1 at low shear rates which was 
followed by a wide region of shear thinning. Although the first two points are slightly lower in 
viscosity, within the experimental error, they can be considered as being in the plateau value 
range. With a further increase in SWNT concentration, the shear thinning regions of each sample 
extended throughout the overall experimental range. Interestingly, at 0.82 vol % SWNT 
concentration, the viscosity shear thins at lower shear rates which are followed by an inflection 
plateau at intermediate shear rates, and shear thins again at high shear rates. This particular 
concentration is the starting point of viscosity increase in the viscosity versus concentration 
curve as shown in Figure 4.13.  
 
112 
 
Figure 4.19. The steady shear viscosity of dsDNA/SWNT supernatant dispersions at different 
SWNT concentrations. 
0.01
0.1
1
10
100
1000
10000
0.01 0.1 1 10 100
? 
(
P
a
.s
)
S h ea r  Rat e ( 1/s )
0.2 6 v ol  %
0.3 5 v ol  %
0.6 0 v ol  %
0.8 2 v ol  %
0.8 5 v ol  %
1
10
100
1,000
10,000
100,000
0.01 0.1 1 10 100
?
 
(
P
a.s
)
S h ea r  Rat e (1/s )
0.9 3 v ol  %
1.0 4 v ol  %
1.2 1 v ol  %
1.4 7 v ol  %
1.9 3 v ol  %
(c)
(a)
(b )
1
10
100
1000
10000
100000
0.01 0.1 1 10 100
? 
(
P
a
.s
)
S h ea r  Rat e ( 1/s )
2 .2 2  v o l  %
2.3 2 v ol  %
2.6 0 v ol  %
3.3 7 v ol  %
1
10
100
1, 00 0
1 0 ,0 0 0
0 .0 1 0 .1 1 10 100
? 
(
P
a
.s
)
S hea r Rate  (1/s )
1 .9 3  v o l %
1
10
100
1000
0 .0 1 0 .1 1 10 100
? 
(
Pa
.s
)
S hea r Rate  (1/s )
2. 22  v ol %
113 
After 0.85 vol % SWNT concentration, samples showed the shear thinning behavior of 
power law fluids; 1nK???? ; where ?  is the shear rate, K is the consistency coefficient related 
to temperature and n is the rate index with n<1 related to shear thinning and n>1 related to shear 
thickening behavior. However, for several samples with SWNT concentrations higher than 1.47 
vol %, which were above the concentration at the local viscosity maximum, the viscosity flow 
curves showed an inflection point in the shear rate range measured showing slightly different two 
shear thinning regions. This inconspicuous but reproducible viscosity ?kink? (or hesitation) is 
also observed for many LCPs at a shear rate where the first normal stress difference N1 is in the 
negative value range.153,159,162 For dsDNA/SWNT supernatant samples at higher concentrations, 
negative N1 was obtained within the experimental shear rate range. However, the typical curve of 
N1 in LCPs which change signs from positive to a negative minimum and become positive again 
with increasing shear rate is absent in this system. Rheological models for polydomain flow of 
tumbling rod-like liquid crystals suggest that the kink phenomenon in a shear thinning viscosity 
is related to a tumbling to oscillating or wagging transition.157,211  Thus, the steady shear 
viscosity hesitation observed in this cholesteric phase forming dsDNA/SWNT system may be an 
indication of the transition from Region 1 to Region II behavior.  
The possible three region behavior was also reported for SWNT dispersed in superacids 
solutions with a very narrow Region II plateau between the two shear thinning regions.19 
Typically, Region I shear thinning is believed to be related to director tumbling in liquid 
crystalline phase upon shear flow. However, Region I is not seen in all lyotropic nematic 
polymeric liquid crystals and depends on many factors such as specific materials used, flexibility 
of rods, defect texture, or the range of shear rates covered.153 The lack of obvious characteristic 
Region I behavior in the complex ternary system of dsDNA/SWNT may be caused by several 
114 
reasons including the orientational defect textures in this system. Both the polydispersity of 
nanotubes and semiflexibility nature of dsDNA molecules may lead to very complex 
microstructures in dsDNA/SWNT dispersions which can result in unusual rheological behaviors 
differing from LCPs. Even 40% PBLG solutions which showed Region I shear thinning also 
suffered from complex hysteresis consisting of two distinct viscosity braches for steady shear 
viscosity measurement with different preshearing.41 This behavior was attributed to the texture 
elasticity and the change in defect texture upon shear flow indicating that Region I behavior is 
extremely sensitive to the sample preshear history.  
The power law coefficients of steady shear and complex viscosity were given in the 
Table 4.3. At high enough concentrations of 3.90 and 5.50 vol %, only dynamic measurements 
are accessable. As for steady shear viscosity test, samples were forced to come out of the 
measuring plates at a shear rate of 2.5 s-1 where the first normal stress N1 had a significant jump 
from negative N1 values to positive values of over 3000 Pa.    
Figure 4.20 shows the complex viscosity as a function of frequency for dsDNA/SWNT 
supernatant dispersions at various SWNT concentrations. At low concentrations of 0.26 to 0.60 
vol % SWNT, the samples were mainly Newtonian with slight shear thinning occurring at high 
shear rates (Figure 4.20a). After 0.82 vol % of SWNT concentration, samples started to show 
shear thinning behavior of power law fluids. However, unlike steady shear viscosity, the 
dynamic viscosity versus frequency curves at high SWNT concentrations did not show viscosity 
hesitation in the experimental frequency range. Table 4.3 summarizes the change in power law 
coefficients with increasing SWNT concentration of several dsDNA/SWNT supernatant samples. 
Marked transitions in power law coefficients occurred at 1.93, 2.22, and 3.37 vol % of SWNT, 
respectively. Interestingly, the maximum in complex viscosity versus concentration was obtained 
115 
near 1.93 vol % SWNT which will be discussed in further in Figure 4.22. Although the steady 
shear viscosity maximum was at 1.47 vol % SWNT at 0.01 and 0.1 s-1 shear rate (Figure 4.21a), 
at a higher shear rate of 1.0 s-1, the maximum in steady shear viscosity was obtained near 1.93 
vol % SWNT (Figure 4.21b). The fully liquid crystalline phase was determined to occur near 
2.22 vol % by both rheology and optical light microscopy as previously shown in Figure 4.13. 
The marked changes in power law coefficient from 2.60 to 3.37 vol % SWNT, especially for 
complex viscosity, provided an evidence of a liquid crystalline to gel transition in the system.   
 
116 
 
Figure 4.20. The complex viscosity of dsDNA/SWNT supernatant dispersions at different 
SWNT concentrations. 
0.01
0 .1
1
10
100
1000
0.1 1 10 100
?
*
 
(
P
a.s
)
A n g u lar  F r equ ency  ( 1/s )
0.2 6 v ol  %
0.3 5 v ol  %
0.6 0 v ol  %
0 .8 2  v o l  %
0.8 5 v ol  %
1
10
100
1000
10000
0.1 1 10 100
?
*
 
(
P
a.s
)
A n g u lar  F r equ ency  ( 1/s )
0.9 3 v ol  %
1.0 4 v ol  %
1.2 1 v ol  %
1.4 7 v ol  %
1.9 3 v ol  %
1
10
100
1000
10000
100000
0.1 1 10 100
?
*
 
(
P
a.s
)
A n g u lar  F r equ ency  ( 1/s )
2.2 2 v ol  %
2.3 2 v ol  % 
2.6 0 v ol  %
3.3 7 v ol  %
(a) 
(b ) 
(c) 
117 
Table 4.3. Consistency and rate index of steady shear and complex viscosities for dsDNA/SWNT 
supernatant dispersions behaving as power law fluids at high SWNT concentrations.  
SWNT 
Concentration 
(vol %) 
Steady Shear Viscosity Complex Viscosity 
K n K n 
0.82 9.25 0.373 15.4 0.288 
0.85 56.5 0.201 29.4 0.349 
0.93 112 0.150 87.4 0.245 
1.04 148 0.110 108 0.219 
1.21 206 0.114 202 0.186 
1.47 298 0.143 476 0.170 
1.93 281 0.223 1381 0.144 
2.16 306 0.262 1353 0.138 
2.22 30.8 0.434 167 0.186 
2.32 39.9 0.457 483 0.187 
2.60 80.9 0.410 521 0.190 
3.37 363 0.211 2371 0.121 
3.90 ~ ~ 12674 0.111 
5.50 ~ ~ 200229 0.069 
 
 
 
 
118 
Figure 4.21 shows the viscosity versus concentration curves at different shear rates from 
0.01 to 100 s-1 for dsDNA/SWNT supernatant dispersions. Similar behavior of abrupt drop in 
viscosity was obtained at various shear rates of 0.01, 0.1, and 1 s-1 at the transition from isotropic 
to cholesteric state. The value of viscosity maximum was higher at lower shear rates measured. 
For viscosities at 10 s-1 shear rate, the nonmonotonic trend is not so obvious, while at high shear 
rates of 100 s-1, the viscosity maximum was erased. According to Kiss and Porter, the viscosity 
maximum of liquid crystalline polymer PBG in m-cresol also decreased and slightly shifted to 
lower concentration with increasing shear rate.155  
The complex viscosity ?* also showed the same nonmonotonic behavior as steady shear 
viscosity as a function of concentration at various angular frequencies from 0.1 to 100 1/s as 
shown in Figure 4.22. The maximum in ?* was obtained near 1.93 vol % SWNT which is 
slightly higher than the concentration, 1.47 vol %, where the steady shear viscosity maximum 
was obtained. Also, the value of ?* maximum decreased with increasing frequency. For complex 
viscosity, even at high frequencies of 10 and 100 1/s, the curves showed an obvious maximum in 
the biphasic regime.  
 
 
 
 
 
 
 
 
119 
 
Figure 4.21. The steady shear viscosity versus concentration at various shear rates for 
dsDNA/SWNT supernatant dispersions. 
 
 
 
 
0
100
200
300
400
500
600
0 1 2 3 4
?
 
(
P
a.s
)
? %  ( S W NT s)
1 s - 1
10 s - 1
100 s - 1
0
3000
6000
9000
12000
15000
0 1 2 3 4
?
 
(
P
a.s
)
? %  ( S W NT s)
0.01 s - 1
0.1 s - 1
(a)
(b )
120 
 
Figure 4.22. The complex viscosity versus concentration curves at various frequencies for 
dsDNA/SWNT supernatant dispersions. 
 
 
 
(a)
(b )
0
2000
4000
6000
8000
10000
0 1 2 3 4
?
*
 
(
P
a.s
)
? %  ( S W NT s)
0 .1  rad /s
1 r ad /s
0
50
100
150
200
250
300
0 1 2 3 4
?
*
 
(
Pa.s)
? %  ( S W NT s )
10  r ad /s
10 0 rad/s
121 
The viscosity kinks in steady shear viscosities were also observed for various nematic 
dsDNA/SWNT (noncentrifuged) dispersions with different dsDNA:SWNT volume ratios (Figure 
4.23). Generally, the viscosity kinks were more obvious when the SWNT concentrations were 
near the ?LC value where the corresponding viscosity is at its local minimum in the viscosity 
versus concentration curves. According to the nonmonotonic relationship of viscosity versus 
concentration for bulk dsDNA/SWNT dispersions, the ?LC values are ~1.13, 3.91, and 1.92 vol % 
of SWNT for 8.5:1, 6.4:1, and 4.3:1 volume ratio of dsDNA:SWNT, respectively. The relatively 
large ?LC of 3.91 vol % SWNT obtained for 6.4:1 dsDNA/SWNT by volume may be a 
consequence of using SWNT with different aspect ratios than the 8.5:1 and 4.3:1 volume ratio 
dispersions used the same type of SWNT. Additional data on steady shear viscosity of 6.4:1 by 
volume of dsDNA:SWNT bulk dispersions and dsDNA only solution as references are provided 
in the Appendix-B.     
122 
 
Figure 4.23. The steady shear viscosity of bulk dsDNA/SWNT (noncentrifuged) dispersions with 
(a) 8.5:1, (b) 6.4:1, and (c) 4.3:1 of dsDNA/SWNT ratio by volume at various SWNT 
concentrations. 
0.01
0.1
1
10
100
1000
10000
0 .0 1 0 .1 1 10 100 1000
?
 
(
P
a.s
)
S h ea r  Rat e (1/s)
8 .5 : 1 0.6 0 v ol  %
1.0 2 v ol  %
1 .1 3  v o l  %
0.01
0.1
1
10
100
1000
10000
100000
0 .0 1 0 .1 1 10 100 1000
?
 
(
P
a.s
)
S h ea r  Rat e (1/s)
4 .3 : 1 1.0 1 v ol  %
1.5 7 v ol  %
1 .9 2  v o l  %
1
10
100
1000
10000
100000
0.01 0.1 1 10 100
?
 
(
Pa.s)
S h ea r  Rat e (1/s)
6 .4 : 1
1.2 9 v ol  %
1.6 4 v ol  %
3.9 1 v ol  %
(a) 
(b ) 
(c) 
123 
Time-Concentration Superposition. Figure 4.24 shows the viscosity master curve for 
the dsDNA/SWNT supernatant created by shifting with respect to stress and inverse time shift 
factors to the reference concentration. For each sample, reduced steady shear viscosity, ?r = (? 
bc)/ac, was plotted against the reduced shear rate, ? r =? ac, where ? is the steady shear viscosity 
and ?  is the shear rate. The stress and inverse time shift factors are defined as bc = cref/c and ac 
= (?0/?0,ref)(cref/c), respectively where c is the SWNT concentration. The subscript ref indicates 
the reference concentration which was chosen to be 1.21 vol % SWNT, while 0 and r refer to the 
zero shear viscosity and reduced variable, respectively. The zero shear viscosities were derived 
from Carreau-Yasuda model by fitting the existing data of Power Law fluids into the model. 
Even though the time - concentration superposition is not perfect, the master curve of 
dsDNA/SWNT supernatant dispersions at 10 oC showed two slight inflection points indicating 
the possible characteristic three region flow behavior of LCPs.48,212   
 
 
124 
 
Figure 4.24. Time - concentration superposition of steady shear viscosities of dsDNA/SWNT 
supernatant dispersions. 
 
The effects of concentration on steady shear viscosities were also analyzed by plotting 
the viscosity against the shear stress which is shown in Figure 4.25. When the samples were 
isotropic, the viscosity increased dramatically with a slight increase in concentration from 0.26 to 
0.35 vol % SWNT and showed Newtonian behavior with the onset of shear thinning only at high 
shear rates (Figure 4.25a). At 0.60 vol % SWNT, the shear thinning behavior dominates with 
little observable plateau indicating an isotropic to anisotropic alignment transition in the system. 
In fact, the lowest concentration where the near-parallel lines of fingerprint texture were detected 
by the optical microscope is 0.69 vol % SWNT. With further increase in concentration, the 
0. 1
1
10
100
1000
10000
100000
1000000
0 .0 0 0 0 1 0 .0 0 1 0 .1 10 1000Re
duc
e
d 
S
te
a
dy
 
S
he
a
r 
V
i
s
c
os
i
ty
 
(
P
a
.
s
)
R educe d Shea r R ate ( 1/ s)
0 .8 5  v o l % 0 .9 3  v o l % 1 .0 4  v o l % 1 .2 1  v o l %
1 .2 7  v o l % 1 .4 7  v o l % 1 .9 3  v o l % 2 .1 6  v o l %
2. 22  v ol % 2. 32  v ol % 2. 60  v ol % 3. 37  v ol %
125 
sample transition from significant shear thinning to possible yielding behavior especially at 1.21 
vol % SWNT where the yield stress is estimated to be about 215 Pa (Figure 4.25b). The stress at 
which the viscosity decreases by greater than an order of magnitude is considered as the yield 
stress. It can be also taken as the maximum stress below which no flow occurs. Many 
commercially important materials such as emulsions and concentrated suspensions exhibit yield 
stress and the performance of the materials are largely affected by the value of the yield stress.213 
When the concentration increased to the vicinity of fully liquid crystal phase around 2.22 vol % 
or above, the viscosity curves showed two distinct shear thinning regions with different slopes 
separated by a seemingly Newtonian plateau, especially at 2.22 and 2.60 vol % SWNT 
concentration. The shear thinning Region I at lower shear rates may become more evident if even 
lower shear rates were accessible in the experiment.  
 
 
 
 
 
 
 
126 
 
Figure 4.25. The effect of concentration on steady shear viscosities of dsDNA/SWNT 
supernatant dispersions.  
 
 
 
( a)
(b )
0.01
0.1
1
10
100
0.0001 0.01 1 100
? 
(
P
a
.s
)
S hea r  S t r e s s  ? ( P a )
0.2 6 v ol  % 0.3 0 v ol  % 0.3 5 v ol  % 0.6 0 v ol  %
1
10
100
1,000
10,000
100,000
1 10 100 1000
? 
(
P
a
.s
)
S h ea r  S t r es s ? ( P a)
0. 85  v ol  % 1. 21  v ol  % 2. 16  v ol  % 2. 22  v ol  % 2. 60  v ol  %
127 
Figure 4.26 shows the steady shear viscosity change as a function of shear stress for 
various bulk dsDNA/SWNT dispersions with different dsDNA:SWNT volume ratios. When the 
SWNT concentrations were in the biphasic regime, the viscosity showed similar behavior with 
increasing shear stress for bulk dispersions of nematic liquid crystals when compared to 
cholesteric dsDNA/SWNT dispersions. However, at SWNT concentrations of ?LC, where the 
viscosities were at minimum, such as 1.13 (Figure 4.26a), 3.91(Figure 4.26b), and 1.92 (Figure 
4.26c) vol % of SWNT respectively, the viscosity versus shear stress curves showed different 
behavior than their cholesteric counterparts. More specifically, at SWNT concentrations of ?LC 
for nematic bulk dispersions, the viscosity drops more rapidly with more than four orders of 
magnitude within a narrow range of increasing shear stress much smaller than an order of 
magnitude. Also, at ?LC of nematic bulk dispersions, samples showed obvious yielding behavior 
instead of shear thinning regions with different slopes when compared to cholesteric dispersions 
at ?LC of near 2.22 vol % SWNT. Near the critical concentration of ?LC for cholesteric 
dispersions, the viscosity shear thins over shear stress range more than an order of magnitude. 
 
 
128 
 
Figure 4.26. The steady shear viscosity versus shear stress plots of bulk dsDNA/SWNT 
(noncentrifuged) dispersions with (a) 8.5:1, (b) 6.4:1, and (c) 4.3:1 of dsDNA/SWNT ratio by 
volume at various SWNT concentrations. 
0.01
0.1
1
10
100
1000
10000
100000
1 10 100 1000
?
 
(
P
a.s
)
S h ea r  S t r es s ? ( P a)
4 .3 : 1
1.0 1 v ol  %
1 .3 8  v o l  %
1.9 2 v ol  %
0.01
0.1
1
10
100
1000
10000
1 10 100
?
 
(
P
a.s
)
S h ea r  S t r es s ? ( P a)
8 .5 : 1
0.5 3 v ol  %
0.6 0 v ol  %
1.1 3 v ol  %
(a) 
( b)  
(c) 
0.1
1
10
100
1000
10000
100000
1 10 100 1000
?
 
(
P
a.s
)
S h ea r  S t r es s ? ( P a)
6 .4 : 1
1.0 1 v ol  %
2.8 5 v ol  %
3 .9 1  v o l  %
129 
Cox-Merz Rule. Ordinary polymers follow the empirical Cox-Merz rule that the steady 
shear viscosity and complex viscosity at the same shear rate or angular frequencies overlap very 
well for the entire concentration range. However, this rule is not obeyed for lyotropic nematic 
LCPs except in the isotropic regime. Interestingly, for cholesteric phase forming dsDNA/SWNT 
supernatant dispersions, Cox-Merz rule is alternately obeyed and disobeyed in the anisotropic 
regimes with increasing SWNT concentration. When the Cox-Merz rule is not obeyed, the steady 
state value drops either below or rise above the oscillatory values. The origin of this atypical 
behavior of alternately obeying and disobeying the Cox-Merz rule is not clear. However, this 
may be due to the more sensitive nature of dynamic measurement to shear history which caused 
deviation from steady shear values especially at lower angular frequencies for lower SWNT 
concentrations in the biphasic region. Liquid crystal samples in biphasic region tend to have 
complex flow behaviors with large errors between different runs. In addition, when the 
Cox-Merx rule was not obeyed in some of the samples in biphasic region, the steady shear and 
complex viscosities only deviated from each other slightly. At higher shear rates and angular 
frequencies, both viscosities seemed to overlap with each other. More specifically, at low 
concentration of 0.35 vol % SWNT, the complex viscosity overlaps to the steady shear viscosity 
at equal values of shear rate and angular frequency in the entire range of 0.1-100 s-1 as shown in 
Figure 4.27a. When the concentration increased to 0.60 vol % SWNT, the complex viscosity 
started to deviate more obviously from steady shear viscosity at low shear rate/frequency which 
is shown in Figure 4.27b. This particular concentration is close to that of 0.69 vol % SWNT 
where the first cholesteric oily streak texture of thin parallel lines were captured by polarized 
light optical microscopy. The deviation to Cox-Merz rule at lower shear rate/frequency 
continued to apply for samples when the SWNT concentration increased up to 0.85 vol % which 
130 
is the onset of apparent viscosity increase in the viscosity versus concentration curve. 
Unexpectedly, with further increase of the SWNT concentration in the biphasic regime from 0.93 
to 1.21 vol %, the complex and steady shear viscosities again overlapped each other within the 
error range (Figure 4.27a). This concentration range is typically in the left shoulder of the 
viscosity maximum in the viscosity versus concentration curve where the viscosity increases 
with concentration. For the next two samples with higher concentrations of 1.27 and 1.47 vol % 
SWNT, the complex and steady shear viscosities slightly deviated from each other in the 
measured shear rate/frequency range. The 1.47 vol % SWNT is the concentration where the 
steady shear viscosity reaches its maximum value of 2165 Pa.s. After this point, when the 
concentration increased further to transition from biphasic to liquid crystal phase, the deviation 
from Cox-Merz rule became much obvious through the entire measured shear rate/frequency 
range (Figure 4.27b). Not obeying the empirical Cox-Merz rule is additional evidence of 
cholesteric liquid crystalline phase formation in the supernatant of dsDNA/SWNT dispersions.  
 
 
 
 
 
 
 
 
 
131 
 
Figure 4.27. Cox-Merz rule alternately (a) obeyed and (b) not obeyed at different SWNT 
concentrations for dsDNA/SWNT supernatant dispersions. Several concentrations were selected 
as examples in this figure.  
 
0 .1
1
10
100
1000
10000
0 .1 1 10 100
?, 
?
* 
(
Pa
.s
)
S hea r Rate , A ngula r Fre quenc y  ( 1 /s )
0.3 5 v ol  % - ? 0.3 5 v ol  % - ?* 0.9 3 v ol  % - ?
0.9 3 v ol  % - ?* 1.2 1 v ol  % - ? 1.2 1 v ol  % - ?*
0. 1
1
10
100
1000
10000
0 .1 1 10 100
?, 
?
* 
(
Pa
.s
)
S hea r Rate , A ngula r Fre quenc y  ( 1 /s )
0.6 0 v ol  % - ? 0.6 0 v ol  % - ?* 0.8 5 v ol  % - ? 0.8 5 v ol  % - ?*
1.2 7 v ol  % - ? 1.2 7 v ol  % - ?* 2.2 2 v ol  % - ? 2.2 2 v ol  % - ?*
(a)
(b )
132 
Similarly, for nematic bulk dsDNA/SWNT dispersions of 6.4:1 volume ratio, the 
empirical Cox-Merz rule was not followed for several anisotropic samples. In Figure 4.28, the 
Cox-Merz rule was not obeyed for the biphasic sample of 3.31 vol % SWNT. However, at lower 
SWNT concentration of 1.01 vol %, the steady shear and complex viscosities overlap with each 
other at the same shear rates and angular frequencies. Additional results for both dsDNA/SWNT 
supernatant and bulk dispersions and dsDNA only solution on not obeying the empirical 
Cox-Merz rule are given in the Appendix-B.     
 
 
Figure 4.28. Cox-Merz rule not obeyed at higher SWNT concentrations for bulk dsDNA/SWNT 
(noncentrifuged) dispersions with 6.4:1 dsDNA/SWNT ratio by volume. 
 
 
 
0.1
1
10
100
1000
10000
100000
0.1 1 10 100
?, 
?
*
 
(
P
a
.s
)
S h ea r  Rat e, A n g u lar  F r equ ency  ( 1/s)
1.0 1 v ol  % - ? 1.0 1 v ol  % - ?* 3.3 1 v ol  % - ? 3.3 1 v ol  % - ?*
133 
Dynamic Modulus. The dynamic data in this research were obtained within the linear 
viscoelastic region (LVE) as determined by amplitude sweep measurements at an angular 
frequency of 10 1/s. Amplitude sweep data can be found in Appendix-B. Figure 4.29 shows the 
loss factor tan (?) which is the ratio of the viscous to elastic modulus, as a function of angular 
frequency of dsDNA/SWNT supernatant dispersions at different SWNT concentrations. The loss 
factor tan ? obtained from frequency sweep is a measurement of sample rigidity which indicates 
its deformation behavior. Thus, a sample has a solid-like behavior when its tan ? value is smaller 
than one indicating that the elastic behavior dominates over the viscous behavior and vice versa 
if tan ? value is larger than one. When the SWNT concentration reached to 0.82 vol %, tan ? 
became smaller than one and exhibited solid-like behavior which indicates that the elastic 
behavior dominates over the viscous behavior. While for dsDNA/SWNT dispersions with SWNT 
concentrations lower than 0.82 vol %, the viscous behavior dominated over the elastic behavior 
and tan ? was larger than one.  
 
 
134 
 
Figure 4.29. The ratio of the viscous to elastic modulus (tan (?) = G?/G?) versus frequency for 
dsDNA/SWNT supernatant dispersions at different SWNT concentrations. 
 
The storage modulus versus frequency of dynamic data for dsDNA/SWNT supernatant 
dispersions at various SWNT concentrations is shown in Figure 4.30. In general, the storage 
modulus G? showed the similar nonmonotonic relationship with concentration as the steady shear 
viscosity. The G? kept increasing when the SWNT concentration increased from isotropic to 
biphasic phase, then started to decrease during the transition from biphasic to fully liquid crystal 
phase. Once the molecular orientation in the system was not significantly affected by increasing 
concentration, the G? once again increased as a function of increasing concentration. The 
dynamic data of samples at less than 0.26 vol % SWNT was not experimentally accessible. In 
Figure 4.30a, two apparent jumps of more than an order of magnitude increase in G? at low 
frequency were detected at low SWNT concentrations; one was the increase from 0.26 to 0.35 
vol %, and the other was from 0.60 to 0.82 vol %, respectively. In fact, from 0.60 to 0.82 vol %, 
0.1
1
10
100
0.1 1 10 100
T
a
n 
(
?
)
A n g u lar  F r equ ency  ( 1/s )
0.26  v ol  %
0.35 v ol  %
0.60 v ol  %
0.82 v ol  %
0.85 v ol  %
1.27 v ol  %
1.93 v ol  %
2.22 v ol  %
2.60 v ol  %
3.37  v ol  %
135 
the G? increased by almost two orders of magnitude and the slope of the curve decreased to level 
up at low frequency. A plateau in G? at low frequencies indicates a possible formation of a 
network while the apparent jump in low frequency modulus G? more than an order of magnitude 
is indicative of a percolation threshold.214 Generally, rheological percolation is identified by the 
formation of a plateau in G? at low frequencies in dynamic tests resulting from an apparent yield 
stress.215 Two apparent jumps observed in this system may indicate an existence of two 
percolation thresholds. It is hypothesized that one of them is related to excess dsDNA network 
and another one is related to SWNT percolation. With further increase in concentration, the slope 
of the curves decreased indicating the formation of a stiffer structure. At high SWNT 
concentrations, larger than 3.37 vol %, the G? became independent of frequency indicating the 
development of more solid-like phase (Figure 4.30b).  
 
 
 
 
 
 
136 
 
Figure 4.30. Elastic (storage) modulus G? as a function of frequency for dsDNA/SWNT 
supernatant dispersions at different SWNT concentrations.   
 
 
10
100
1000
10000
100000
1000000
0.1 1 10 100
G
' 
(
P
a)
A n g u lar  F r equ ency  ( 1/s )
2.1 6 v ol  %
2.2 2 v ol  %
2.6 0 v ol  %
3.3 7 v ol  %
3.9 0 v ol  %
5.5 0 v ol  %
(a)
(b)
0.001
0.01
0.1
1
10
100
1000
10000
0.1 1 10 100
G
' 
(
P
a
)
A n g u lar  F r eq u en cy  ( 1/s )
0.2 6 v ol  %
0.3 5 v ol  %
0.6 0 v ol  %
0.8 2 v ol  %
0.8 5 v ol  %
1.0 4 v ol  %
1.2 1 v ol  %
1.2 7 v ol  %
1 .9 3  v o l  %
137 
Furthermore, the data in Figure 4.30a were used to estimate the percolation concentration 
by varying ?* to obtain the best linear fit of the double log plot of     versus (? - ?*) where     
is the plateau modulus, ? is the volume fraction of nanotubes, and ?* is the percolation volume 
fraction. As mentioned earlier, the plateau in G? and more than an order of magnitude increase in 
G? at low frequency were considered to be the indication of percolation threshold. Near the 
percolation threshold,      ?  ?   , where ? is the percolation exponent. Two groups of data 
were considered; one group of data is from 0.82 ? 1.27 vol % SWNT samples (Figure 4.31a), 
and another one is from 0.26 ? 1.27 vol % SWNT samples (Figure 4.31b). The     values at 
higher SWNT concentrations were not included. Because at high SWNT concentrations, samples 
contained a high fraction of anisotropic phase and G? began to decrease as a function of ?. In 
Figure 4.31a, the best linear fit of R2= 0.945 was obtained for ?*=0.62 vol % and the percolation 
exponent ? = 2.78 ? 0.16 from the slope of the double log plot of     at 0.1 1/s angular 
frequency versus (? - ?*). The percolation concentration 0.62 vol % may be related to the 
network formation of rigid SWNT. While for the other group of data in Figure 4.31b, the best 
linear fit was obtained for ?*=0.21 vol % with ? = 3.62 ? 0.05 and R2= 0.980. The lower 
percolation concentration 0.21 vol % may be caused by the dominating network of semiflexible 
free dsDNA. A percolation threshold of ?*=0.11 vol % SWNT and ? = 1.6 were obtained 
experimentally for SWNT superacids system.151 It was reported that a percolation exponent of 
2.1 ? 0.2 obtained by simulation corresponds to percolation bonds that resist stretching but can 
rotate freely, while percolation bonds which resist both stretching and rotation have a percolation 
exponent of 3.75 ? 0.11.214,216 The percolation exponent obtained in this work indicates that the 
SWNT network in the dsDNA/SWNT system was composed of associating rods.  
 
138 
 
Figure 4.31. (a) The storage plateau modulus    , obtained at 0.1 s-1 angular frequency, versus 
the reduced volume fraction (? - ?*) of SWNT. The value of ?*=0.6175 vol % gave the best 
linear fit with R2=0.945 and ? =2.78 ? 0.16. (b) The value of ?*=0.2101 vol % gave the best 
linear fit with R2=0.980 and ? =3.62 ? 0.05 for     at 1 s-1 angular frequency versus (? - ?*) of 
SWNT.  
 
1
10
100
1000
0 .1 1
G
0
' 
(Pa
)
? - ? *
s lope  = 2 .7 8   0 .1 6
0 .0 0 1
0. 01
0 .1
1
10
100
1000
0 .0 1 0 .1 1 10
G
0
' 
(Pa
)
? - ? *
s lope  = 3 .6 2   0 .0 5
(a)
(b )
139 
In addition, the Cole-Cole plot of relationship between G? and G? is shown in Figure 4.32. 
With increasing SWNT concentration from 0.60 to 0.85 vol %, the increase in storage modulus 
G? at a given value of loss modulus G? indicated an evolution of structural change occurring in 
supernatant dsDNA/SWNT dispersions. In addition to having increased G? for a given G? value, 
the slops of G? against G? curves increased consistently with increasing SWNT concentration, 
indicating the formation of a more complex structure. The Cole-Cole plots of bulk dispersions 
are provided in Appendix-B.  
 
 
Figure 4.32 Cole-Cole plots show the storage modulus as a function of loss modulus and 
increasing SWNT concentration for supernatant 4.4:1.0 volume ratio of dsDNA:SWNT 
dispersions. 
 
0 .0 0 1
0 .0 1
0 .1
1
10
100
1000
10000
0 .0 0 1 0 .1 10 1000
S
to
ra
ge 
M
odulus,
 G
' 
(Pa
)
Loss  M odulus,  G "  (Pa )
0 .2 6  v o l. %
0 .3 5  v o l. %
0 .6 0  v o l. %
0 .8 5  v o l. %
1 .0 4  v o l. %
1 .2 7  v o l. %
1 .9 3  v o l. %
2 .3 2  v o l. %
2 .6 0  v o l. %
3 .3 7  v o l. %
140 
The loss modulus G? of dsDNA/SWNT dispersions at various SWNT concentrations as a 
function of frequency are shown in Figure 4.33. The G? jumped up to more than an order of 
magnitude at lower angular frequencies from 0.26 to 0.35 vol % of SWNT concentration. At the 
same time, the slope of the curve kept decreasing as the SWNT concentration increased from 
0.26 to 1.93 vol % SWNT and showed a low frequency plateau at 1.93 vol % of SWNT 
concentration (Figure 4.33a). With further increase in concentration at 2.22 and 2.60 vol %, 
respectively, the G? curves showed steep increases in slopes at higher angular frequencies, 
although plateaus were obtained at lower angular frequencies (Figure 4.33b). It was determined 
from the steady state viscosity versus concentration curve that fully cholesteric liquid crystalline 
phase was formed at around a SWNT concentration of 2.22 vol %. The fully liquid crystalline 
phase tend to flow align easier than biphasic phase at high angular frequencies causing the 
increase of viscous behavior of the liquid crystal dispersion. This may be the reason the steep 
increase in the G? curve slope at higher angular frequency for liquid crystal samples. When the 
concentration increased further to become larger than 3.37 vol %, the G? showed a plateau again 
indicating a solid phase formation.  
 
 
 
 
 
 
 
 
141 
 
 
Figure 4.33. Viscous (loss) modulus G? as a function of frequency for dsDNA/SWNT 
supernatant dispersions at different SWNT concentrations.   
 
 
10
100
1000
10000
100000
0.1 1 10 100
G
"
 
(
P
a)
A n g u lar  F r equ ency  ( 1/s )
2.1 6 v ol  %
2.2 2 v ol  %
2.6 0 v ol  %
3.3 7 v ol  %
3.9 0 v ol  %
5.5 0 v ol  %
0.001
0.01
0.1
1
10
100
1000
0.1 1 10 100
G
"
 
(
P
a)
A n g u lar  F r equ ency  ( 1/s )
0.2 6 v ol  %
0.3 5 v ol  %
0.6 0 v ol  %
0.8 2 v ol  %
0.8 5 v ol  %
1.0 4 v ol  %
1.2 1 v ol  %
1.2 7 v ol  %
1.9 3 v ol  %
(a)
(b)
142 
Both the G? and G? at a certain angular frequencies showed a nonmonotonic relationship 
with increasing SWNT concentration in dsDNA/SWNT supernatant dispersions as shown in 
Figure 4.34. Initially, the G? at 0.1 1/s kept increasing rapidly with concentration and reached a 
maximum near 1.93 vol % SWNT. The maximum in G? at 0.1 1/s was obtained at similar SWNT 
concentration as the dynamic viscosity maximum at the same frequency. An equally rapid 
decrease of G? occurred as the amount of anisotropic phase increased and showed a local 
minimum at 2.22 vol % SWNT. After this point, the G? started to continuously increase again as 
a function of concentration. The same trends were observed at different frequencies, yet the 
value of the G? maximum increased with increasing frequency. The maximum in G? was also 
shifted to slightly lower SWNT concentration when increasing frequency from 0.1 to 100 1/s. 
Similarly, the local maximums and minimums obtained for loss modulus G? at 0.1 to 10 1/s 
frequencies with increasing concentration were at the very similar SWNT concentrations of 
storage modulus G? versus concentration curves (Figure 4.34b). However, there are no apparent 
changes in the G? maximum values at different frequencies, and no obvious shift in 
concentration was observed. The curve for G? versus concentration at 100 1/s frequency was not 
included due to a data point fluctuation.  
 
 
 
 
 
 
143 
 
Figure 4.34. The (a) G? and (b) G? as a function of concentration at different angular frequency 
for dsDNA/SWNT supernatant dispersions. 
 
  
 
(a)
(b )
0
200
400
600
800
0 1 2 3 4
G
"
 
(
P
a)
? %  ( S W NT s )
0.1  r ad /s
1 r ad /s
10  r ad /s
0
500
1000
1500
2000
2500
3000
0 1 2 3 4
G
' 
(
P
a)
? %  ( S W NT s)
0.1  r ad /s
1 r ad /s
1 0  rad /s
10 0 rad/s
144 
Temperature Study. Studies of temperature effects on the rheology of supernatant 
dsDNA/SWNT dispersions at isotropic and higher concentration and concentrated bulk 
dispersions were investigated by rheology. Temperature was increased in the range of 10 to 35 
oC for concentrated samples. At an even higher temperature of 40 oC for concentrated 
dispersions, both evaporation and structural change introduced large error in measurements. The 
time-temperature superposition (TTS) was obtained by shifting collected data with respect to 
both horizontal and vertical shift factors. For steady shear and complex viscosity, ?r = ?/aT (or 
         ), was plotted against the reduced shear rate or angular frequency, ? r = ?  aT (or 
       ). The inverse time shift factor was defined as aT = ?0(T)/?0(T0), where T was the 
measuring temperature and T0 is the reference temperature which was chosen as 25 oC. Usually, 
the stress shift factor is taken as a unity. Figure 4.35 shows the TTS of an isotropic dispersion of 
supernatant dsDNA/SWNT at 0.33 vol % SWNT concentration. Both the steady shear and 
complex viscosities decreased with increasing temperature from 10 to 40 oC and overlapped to a 
single master curve in the measured shear rate and angular frequency range, respectively. As for 
storage modulus, there were deviations at high angular frequencies with increasing temperature. 
This may have been caused by the fact that at higher temperature, the sample was less viscous 
and it was difficult to measure storage modulus at higher angular frequencies due to relatively 
high inertia force of rotating bob.  
 
 
145 
 
Figure 4.35. Time-temperature superposition of supernatant dsDNA/SWNT (Rice 187.2) 
dispersions with 4.4:1 dsDNA:SWNT ratio by volume at SWNT concentration of 0.33 vol %; (a) 
steady shear viscosity versus shear rate, (b) complex viscosity versus angular frequency, and (c) 
storage modulus versus angular frequency at various temperatures of 10, 25, and 40 oC. 
0 .1
1
10
0 .0 1 0 .1 1 10 100 1000
R
educe
d 
?* 
(
Pa
.s
)
Reduc e d A ngula r Fre quenc y  (1/s )
1 0  o C
2 5  o C
4 0  o C
0 .0 0 1
0. 01
0 .1
1
10
0 .1 1 10 100 1000
R
educe
d 
G
' 
(
Pa
)
Reduc e d A ngula r Fre quenc y  (1/s )
1 0  o C
2 5  o C
4 0  o C
0 .1
1
10
0 .0 0 1 0 .1 10 1000
R
educe
d 
? 
(
Pa
.s
)
Reduc e d S hea r Rate  (1/s )
1 0  o C
2 5  o C
4 0  o C
146 
Figure 4.36 shows the TTS of concentrated supernatant dsDNA/SWNT dispersions with 
3.74 vol % SWNT at different concentration from 10 to 35 oC. Similarly, viscosities and storage 
modulus decreased with increasing temperature. In general, the TTS of concentrated supernatant 
dispersions gave overlapped master curves. When plotting the reduced steady shear and complex 
viscosities as a function of reduced shear rate and angular frequencies, viscosity master curves 
were obtained, though viscosities at 10 oC were slightly deviated from master curves. For TTS of 
the storage modulus, the G? at 10 oC had more deviation from the master curve. The shifted 
values of viscosities and storage modulus at 25 and 35 oC overlapped nearly perfectly.  
 
147 
 
Figure 4.36. Time-temperature superposition of supernatant dsDNA/SWNT (Rice 187.2) 
dispersions with 4.4:1 dsDNA:SWNT ratio by volume at SWNT concentration of 3.74 vol %; (a) 
steady shear viscosity versus shear rate, (b) complex viscosity versus angular frequency, and (c) 
storage modulus versus angular frequency at various temperatures of 10, 25, and 35 oC. 
10
100
1000
0 .0 1 1 100
R
educe
d 
G? 
(
Pa
)
Reduc e d A ngula r Fre quenc y  (1/s )
1 0  o C
2 5  o C
35  oC
1
10
100
1000
0 .0 0 1 0 .1 10 1000
R
educe
d 
? (
Pa
.s
)
Reduc e d S hea r Rate  (1/s )
1 0  o C
2 5  o C
3 5  o C
1
10
100
1000
0 .0 1 1 100
R
educe
d 
?* (
Pa
.s
)
Reduc e d A ngula r Fre quenc y  (1/s )
1 0  o C
2 5  o C
3 5  o C
148 
Studies of temperature effects on the rheology of 6.4:1 by volume bulk dsDNA/SWNT 
dispersions at SWNT concentration of 1.67 vol % were also investigated by rheology. When 
increasing the temperature from 10 to 35 oC, a slight decrease in the steady shear viscosity of 
bulk dsDNA/SWNT dispersion was observed. The TTS of   fall onto a master curve showing a 
rapid shear shear thinning region at lower shear rate and a slower shear thinning region at higher 
shear rate as shown in Figure 4.37a. More specifically, at a higher temperature of 35 oC, it is 
more obvious that the   versus shear rate curve changes from power law shear thinning to 
detectable two shear thinning regions with different slopes separated by a transition region more 
like a plateau. This behavior may be due to a structural change or phase transition occurring in 
the system at higher temperature of 35 oC. Further evidence of structural change at 35 oC can be 
seen from the TTS of the dynamic measurement of both complex viscosity and storage modulus 
as shown in Figure 4.37b,c. Both the  * and G? at 35 oC drifted away from the master curve 
obtained at 10 and 25 oC.  
 
149 
 
Figure 4.37. Time-temperature superposition of bulk dsDNA/SWNT (Rice 187.2) dispersions 
with 6.4:1 dsDNA:SWNT ratio by volume at SWNT concentration of 1.67 vol %; (a) steady 
shear viscosity versus shear rate, (b) complex viscosity versus angular frequency, and (c) storage 
modulus versus angular frequency at various temperatures of 10, 25, and 35 oC. 
0 .1
1
10
100
1000
10000
0 .0 1 1 100 10000
Re
duc
e
d 
?
* (
P
a
.s
)
Rdu c e d A ngual r Fre quenc y  (1/s )
1 0  o C
2 5  o C
3 5  o C
10
100
1000
10000
0 .0 1 1 100 10000
R
educe
d 
G
' 
(
Pa
)
Reduc e d A ngula r Fre quenc y  (1/s )
1 0  o C
2 5  o C
3 5  o C
0. 1
1
10
100
1000
10000
0 .0 0 1 0 .1 10 1000
R
educe
d 
? 
(
Pa
.s
)
Reduc e d S hea r Rate  ( 1 /s )
1 0  o C
2 5  o C
3 5  o C
(a)
(b )
(c)
150 
4.4. Conclusions 
The phase behavior and rheology of aqueous dsDNA/SWNT dispersions were 
determined by a combination of polarized optical microscopy. Concentration of bulk 
(noncentrifuged) dsDNA/SWNT dispersions consisting of both bundled and individual SWNT 
led to the formation of polydomain nematic phase. While evaporating the supernatant dispersions 
resulted in cholesteric phase formation. The summary of rheological data obtained from both 
bulk and supernatant dsDNA/SWNT dispersions and dsDNA only solutions is shown in Table 
4.4. Rheology can detect both the biphasic ? liquid crystal transition at the critical concentration 
?N and the liquid crystal ? gel transition at even higher concentration. In addition, the nematic 
and cholesteric phases of dsDNA/SWNT dispersions showed different rheological behavior, 
especially with different shear thinning behavior in viscosity versus shear stress plots. 
Reproducible viscosity kinks in both nematic and cholesteric phases may be an indication of 
director movement transition in the polydomain texture. The empirical Cox-Merz rule was not 
obeyed in the dsDNA/SWNT system indicating the formation of liquid crystals. Moreover, the 
percolation threshold can be calculated from the storage modulus in oscillatory test. Temperature 
effects on rheological characteristics of both nematic and cholesteric dsDNA/SWNT dispersions 
were investigated and corresponding time ? temperature superposition was obtained.  
 
151 
Table 4.4. Comparison of key rheological parameters of three different systems. 
Samples 
~ P?  
(vol %) 
max?  
(Pa.s) 
)( max1 ??  
(vol %) 
min?  
(Pa.s) 
)( min2 ??  
(vol %) 
'maxG  
(Pa) 
)( 'max3 G?  
(vol %) 
? ?3??C  
(%) 
? ?3tan ??  'minG
 
(Pa) 
)( 'min4 G?  
(vol %) 
? ?4??C  
(%) 
? ?4tan ??  
Non-centrifuged 
dsDNA:SWNT=6.4:1 
1.01 2175 3.31 1500 3.91 3270 3.31 3.95 0.20 2070 3.91 6.31 0.23 
Supernatant 
dsDNA:SWNT=4.4:1 
0.82 2165 1.47 106.4 2.22 901.5 2.16 10 0.35 110 2.22 3.98 0.35 
dsDNA only 7.1 66.55 4.53 27.35 6.17         
1) Max. and Min. are the values at 0.1 s-1 shear rate or angular frequency on the nonmonotonic phase transition plots versus volume fraction 
? . Maximum value is in the biphasic region, and minimum value usually corresponds to the single LC phase formation. 
2) P?  is the volume fraction where percolation occurs; ? is bulk viscosity; G? is storage modulus; C?  is critical strain; ?tan  is damping 
factor. 
* No max. and min. values were obtained from G? versus volume fractions plots for dsDNA only in this preliminary experiments. 
 
 
152 
 
 
 
 
 
Chapter 5 
Macroscopic Assembly of Films and Fibers 
The remarkable properties of single-walled carbon nanotube (SWNT) can be realized 
when they are translated into macroscopic objects such as SWNT films and fibers. However, the 
strong van der Waals interaction and random entanglement of synthesized SWNT remains a key 
scientific challenge that limits their real world applications. It has previously been shown that 
processing nematic liquid crystalline dispersions is a promising route for producing aligned solid 
materials such as films and fibers.10,20,217 Considerably less is known about processing lyotropic 
cholesteric dispersions.48 In this chapter, various SWNT dispersions were assembled into films 
and fibers by fluid phase processing. In addition, the optical, electrical, and mechanical 
properties of films and fibers were determined.  
 
5.1. dsDNA/SWNT Films 
Research on cellulose nanowhisker liquid crystals has shown that retention of helicity 
during processing can enable films with controllable optical properties that can be used for such 
circular polarizers, reflective displays and security papers.22,23 Spreading cholesteric 
dsDNA/SWNT dispersions (2.0 vol % of SWNT) onto copper tape by roller coating, allowing 
the coating to dry, and removing the substrate resulted in approximately 20 ?m thick films which 
were translucent under ambient light (Figure 5.1a). Under cross polarized light they appeared 
either bright or dark depending on the orientation of the polarizer. SEM of these films (Figure 
5.2) revealed a aligned densely packed structure comprised of dsDNA/SWNT bundles ranging 
 
153 
from 8 to 22 nm in diameter. The alignment and packing in this microstructure is far superior to 
previously reported aligned dsDNA/SWNT films produced from dilute dispersions using 
layer-by-layer assembly and drying under a directed air stream.1 These results may indicate the 
shear applied during processing induced in a cholesteric to nematic transition such as that 
previously proposed for rod-like polymer solutions.  
 
 
Figure 5.1. (a) Translucent film obtained from a 4.4:1.0 by volume supernatant of 
dsDNA/SWNT dispersion (~2.0 vol % of SWNT) by shearing at approximately 500 s-1 shear rate. 
(b-d) Cross-polarized optical microcopy images of films prepared without applied shear. The 
colors vary as the sample is rotated in (d) 0?, (e) 40?, and (f) 90?. Nikon Eclipse 80i with LU 
Plan Fluor 20x/0.45 objective with 20? magnification and 1.5 magnification in front of the 
camera was used. The scale bars in the figures are 50 ?m.  
 
154 
 
Figure 5.2. SEM images of translucent films obtained from a 4.4:1.0 by volume supernatant of 
dsDNA/SWNT dispersion (~2.0 vol % of SWNT) by shearing at approximately 500 s-1 shear rate. 
The scale bar in (a) is 10?m and in (b) is 200 nm. 
 
Producing the films by simply dropping the cholesteric dispersions onto the tape and not 
applying any shear also resulted in materials that were translucent under ambient light. However, 
under cross polarized light these films showed bright birefringent bands indicative of the 
retention of cholesteric structure (Figure 5.1 b-d). These films are similar in appearance to those 
obtained by Lee et al.218 for films produced from a helical liquid crystalline phase of 
Au-antistreptavidin M13 virus. Since the effects of shear on cholesteric liquid crystals are not as 
well understood as the effects of shear on nematic liquid crystal, rheo-optical characterization 
was used to provide insights into the shear induced unwinding of cholesteric phases into nematic 
phases. Preliminary investigations of shear effects on the cholesteric pitch of dsDNA/SWNT 
(Rice 187.2) dispersion at 2.32 vol % SWNT using a Linkam shear cell at a 30 ?m gap at a room 
temperature were performed. When shearing the sample at a small shear rate of 0.25 s-1 for a few 
 
155 
second ~10s, the cholesteric pitch showed a slight increase from 1.59 ? 0.18 to 1.84 ? 0.24 ?m 
after cessation of shear. 
 
5.2. dsDNA/SWNT Fiber Spinning 
At the preliminary stage of the fiber spinning research, cholesteric dsDNA/SWNT 
supernatant dispersions with higher concentrations of SWNT were quenched into various alcohol 
and acetone coagulants. Fibers were only obtained in some coagulants such as ethanol (95 %), 
2-propanol (99.8 +%), and tert-butanol. The surface of these fibers showed various textures 
including spherical (with diameter 70-150 nm) and rod-like (60-80 nm in width and several 
hundred nm in length) particles, wavy or smooth structures (Figure 5.3). These different 
structures could be caused by many factors such as the different coagulants used, residence time, 
orifice diameter of the syringe needle, and extruding speed which are important in effecting the 
boundary conditions of fiber coagulating process. Many studies have been conducted to use 
alcohols (especially ethanol, 2-propanol, and tert-butanol), sometimes adding multivalent cations 
into the alcohols/water mixtures, to induce DNA condensation which is caused by the charge 
neutralization of DNA phosphate charge. The dehydrating action of alcohols reduces the water 
activity through the reduction of dielectric constant of the solvent medium.219-222 Considering 
that high purity alcohols were used in this work, DNA molecules on the surface were further 
aggregated into a fibrous network and transitioned into a more compact DNA structure as shown 
in Figure 5.3b, c.219 The formation of relatively smooth and wavy surfaces, it might have been 
due to the shorter residence time in coagulants in which the spine of hydration stabilizing the 
DNA was not affected seriously. Moreover, when the residence time in a coagulant was 
comparatively short, the fiber surface showed continuous fibril structure (Figure 5.3e). 
 
156 
Preliminary EDS elemental analysis showed that the carbon content in fibers decreased along the 
radius radiation from 72.98-36.84 C wt % in center to 17.88-36.78 C wt % on the surface. In 
contrast, the phosphorous content increased from 2.91-7.81 P wt % to 6.36-16.58 P wt %. Since 
the samples were sputter coated with gold for better imaging under SEM, the gold content of 
20.98-63.92 Au wt % was also detected on the surface. This could cause error in the elemental 
percentages. The relative degree of SWNT alignment in dsDNA/SWNT fiber was quantified by 
the difference in G band at different angles in Raman spectroscopy using 785 nm wavelength 
laser. The G band corresponds to sp2 tangential mode of the nanotube and was located near 1595 
cm-1 wavenumber in this fiber sample. The ratio of G peak intensities, with the laser aligned 
parallel to the fiber?s long axis and perpendicular to the long axis, was in the range of 1.9 to 5, 
which indicated relatively low alignment of nanotube as shown in Figure 5.4. It is possible that, 
the complex surface morphology and the DNA coating on the fiber surface prevented signaling 
the alignment of nanotube in the inner side of fiber sample. 
 
 
 
157 
 
Figure 5.3. Scanning electron microscope images of fibers extruded from cholesteric LC 
dispersion of dsDNA/SWNT (1.98 vol % of SWNT) in (a), (b), (d), (e) 2-propanol, (c) ethanol, 
and (f) tert-butanol. The fiber diameter is ~81 ?m in (a). The scale bar of the first image is 10 ?m 
and the rest of them are 1 ?m. 
 
 
Figure 5.4. Raman Spectra at 785 nm excitation of fiber spun from cholesteric dispersion of 
dsDNA/SWNT (1.98 vol % SWNT) in ethanol.  
0 . 0 E + 0 0
5 . 0 E + 0 4
1 . 0 E + 0 5
1 . 5 E + 0 5
2 . 0 E + 0 5
2 . 5 E + 0 5
3 . 0 E + 0 5
3 . 5 E + 0 5
1200 1300 1400 1500 1600 1700
R a m a n  S h i f t  ( c m - 1 )
I
n
t
e
n
s
i
t
y
0  D e g r e e
4 5  D e g r e e
9 0  D e g r e e
 
158 
In addition, a bulk (noncentrifuged) dispersion with 6.4:1 volume ratio of dsDNA:SWNT 
at ~ 3.4 vol % SWNT was also assembled into fibers via two separate paths; dry spinning and 
wet spinning. A 22 G syringe needle (0.41 mm inner diameter) was used and slower injection 
rate of 0.2 ml/min was chosen for fiber spinning from the bulk dsDNA/SWNT dispersion. The 
wet spun fibers of dsDNA/SWNT dispersion were coagulated into 1-butanol bath. For dry 
spinning, a Teflon plate was used to collect the SWNT dispersion extruded from a syringe. 
Mechanical and electrical properties of the dry spun fibers were determined after they were dried 
at ambient condition overnight and further vacuum dried at room temperature for 4 hr. The fiber 
spinning from various SWNT dispersions is still in the preliminary stage toward the development 
of an improved fiber spinning scheme. Optimization of initial SWNT dispersions and their fiber 
spinning process is ongoing.  
As shown in Figure 5.5, although having an uniform shape in general, the surface 
structures of fibers produced from bulk (noncentrifuged) dsDNA/SWNT dispersions by either 
wet spinning in 1-butanol or dry spinning, were still not as smooth as those of PVA coagulated 
lysozyme/SWNT fibers which are discussed in the next section. Especially for fibers coagulated 
in 1-butanol, more irregular structures were observed. This is may be due to DNA condensation 
in the alcohol environment (Figure 5.5b). Interestingly, the surface morphology of dry spun 
fibers differed depending on whether their surfaces were in contact with air or the Teflon plate 
(Figure 5.5c, d). The fiber surface in contact with the Teflon plate when drying showed much 
smoother structures. It is difficult to see the nanotube alignment on the surface due to the DNA 
coating. However, large amount of small SWNT bundles pulled out from the dsDNA interacted 
SWNT fiber can be observed at the cross section of fibers (Figure 5.6a). The SWNT bundle size 
was determined to be around ~15 nm in diameter from Figure 5.6b at higher magnification. In 
 
159 
Figure 5.6c, the cross section of fiber also showed that, dsDNA/SWNT fibers produced by 
coagulation in 1-butanol had a loose structure overall with many voids and holes inside.   
 
 
Figure 5.5. Scanning electron micrographs of fibers spun from bulk (noncentrifuged) 6.4:1 by 
volume dsDNA/SWNT dispersion with ~3.4 vol % SWNT by (a) coagulating in 1-butanol and (b) 
its subsequent surface structure. For dry spinning fibers, surfaces of fibers (c) in contact with air 
and (d) Teflon have different structures. The scale bars are (a) 100 ?m and (b, c, d) 1 ?m.   
  
 
160 
 
Figure 5.6. Scanning electron micrographs of cross-section structures of fibers produced from 
bulk 6.4:1 by volume dsDNA/SWNT dispersion with ~3.4 vol % SWNT by (a, b) dry spinning, 
and (c) coagulation in 1-butanol. The scale bars are (a) 1 ?m, (b) 100 nm, and (c) 10 ?m. 
 
161 
5.3. Lysozyme/SWNT Fiber Spinning 
Wet solution spinning of lysozyme/SWNT dispersions into the co-flow stream of a 
polymer solution was performed in collaboration with Dr. Philippe Poulin?s group at Centre de 
Recherch? Paul Pascal - Centre National de la Recherch? Scientifique (CRPP-CNRS) in 
Bordeaux, France. The focus of this work was to improve the dispersion and fiber spinning of 
single-walled carbon nanotubes in aqueous lysozyme (LSZ). Lysozyme is an enzyme protein 
with antibacterial activity and its ability to disperse SWNT as individuals has been previously 
shown by Nepal et al.1 The addition of cationic surfactant (1-tetradecyl) trimethyl-ammonium 
bromide (TTAB) into the LSZ/SWNT dispersion plays an important role of stabilizing SWNT 
through electrostatic repulsion which counteracts with the van der Waals attraction of 
SWNT.14,21 TTAB has a cationic detergent with a positively charged ammonium bromide head 
group and a 14 carbon tail. Although it would be difficult to deconvolute molecular interactions 
in the complex system of LSZ/TTAB/SWNT aqueous dispersions, the main interactions are 
considered to be hydrophobic in nature and exist between the sidewall of SWNT, hydrophobic 
core of LSZ, and hydrophobic alkyl-tail of TTAB.78,223-225 Both purified and non-purified SWNT 
were used to prepare LSZ/TTAB/SWNT dispersions for macroscopic assembly into fibers in a 5 
wt % polymer PVA coagulation bath. The injection rate of LSZ/TTAB/SWNT dispersion was 50 
ml/hr.  
The PVA coagulated fibers, shown in Figure 5.7, had conical geometries and smooth 
surfaces due to PVA coating on the outer side of fibers. Fibers composed of more than one strand, 
were made through collapsing wet PVA coagulated individual SWNT ribbons on top of one 
another during hang-drying to produce a single, larger fiber. Looking closely at Figure 5.7a, 
small ridges near the conical edge of the fiber can be seen. These ridges are due to the collapsing 
 
162 
of individual fibers onto one another creating discontinuities in the structure. This phenomenon 
can be seen much more clearly in Figure 5.7b where the ridges occur not only at the edges but 
throughout the fiber. The increased discontinuity in the 22 strand fiber is simply due to the 
increased number of strands that were collapsed. According to SEM, the diameter of a single 
strand is about 10-15 ?m. 
 
 
Figure 5.7. Scanning electron micrographs of (a) 7 strands and (b) 22 strands PVA coagulated 
LSZ/TTAB/SWNT fibers showing smooth structure and evidence of collapsing of multiple 
fibers. 
 
5.4. Fiber Properties 
Mechanical and electrical properties were measured on dry SWNT fibers. Table 5.1 
shows the electrical resistivity and conductivity of PVA coagulated LSZ/SWNT/TTAB fibers 
using purified and non purified SWNT and dsDNA/SWNT fibers produced from dry and wet 
spinning methods. Comparison of the electrical properties shows that dsDNA/SWNT fibers 
showed much higher resistivity than PVA coagulated LSZ/TTAB/SWNT fibers. Since the initial 
bulk dispersion of dsDNA/SWNT has high weight ratio of 7.5:1 (6.4:1 by volume) 
 
163 
dsDNA:SWNT, the existence of large content of non conducting dsDNA polymer caused high 
electrical resistivity in the resulting dried fibers. Especially for dsDNA/SWNT fibers coagulated 
in 1-butanol the resistivity was very high at 145 ? 13.17 ??cm. This could be caused by DNA 
condensation in alcohol resulting in a more electrical resistant structure. As a comparison, the 
electrical conductivity of PVA coagulated fibers from isotropic dispersion of 0.4 wt% 
ssDNA-0.4 wt% SWNT was reported to be 0.042 S/cm, which is 23.8 ??cm in resistivity.198 
Although the electrical resistivity of LSZ/TTAB/SWNT fibers is relatively lower than that of 
ssDNA/SWNT fibers, it is still very large when compared to reported values of PVA coagulated 
SDS/SWNT fibers14 and fibers consisting solely of SNWTs.10 When the insulating polymer PVA 
was used to coagulate the wet spun fibers, it naturally inhibited the flow of electrical current 
thereby reducing the conductivity of the resulting fibers.  
The mechanical properties tested were Young?s modulus, tensile strength, and strain at 
break. Table 5.2 shows the mechanical test data of various SWNT fibers. In general, the 
mechanical test data shows that PVA coagulated LSZ/TTAB/SWNT fibers were more 
mechanically robust than dsDNA/SWNT fibers. The reasoning for improved mechanical 
properties is the incorporation of PVA within the fiber during the coagulation process. The 
addition of the PVA not only increases the tensile properties of the fiber but also increases the 
stiffness by providing a strong bridging effect between SWNT bundles.11,14 Again, the 
mechanical properties of fibers obtained in this work is still inferior to those of previously 
reported fibers, including PVA coagulated ssDNA/SWNT fibers which had Young?s modulus of 
14.1 GPa, and tensile strength of 110 MPa. In order to improve the performance of SWNT fibers, 
both the initial SWNT dispersion used and the process of fiber spinning need to be optimized in 
future work.   
 
164 
Table 5.1. Electrical resistivity and conductivity of various SWNT fibers from both wet spun and 
dry spun assembly methods. 
Sample 
Electrical Properties 
Electrical Resistivity, ? 
(??cm) 
Electrical Conductivity, c 
(S?m-1) 
PVA coagulated 
LSZ/TTAB/SWNT(purified 
SWNT) fiber 
11.70 ? 2.09 8.74 ? 1.40 
PVA coagulated 
LSZ/TTAB/SWNT (non 
purified SWNT) fiber 
21.56 ? 4.61 4.89 ? 1.45 
1-butanol coagulated 
dsDNA/SWNT fibers 145 ? 13.17 0.69 ? 0.07 
Dry spun dsDNA/SWNT 
fibers 52.37 ? 4.46 1.92 ? 0.17 
 
 
 
 
 
 
 
 
 
 
 
 
 
165 
Table 5.2. Mechanical properties of various SWNT fibers from both wet spun and dry spun 
assembly methods. 
Sample 
Mechanical Properties 
Young's Modulus                
(GPa) 
Tensile Strength                
(MPa) 
Strain at Failure        
(%) 
PVA coagulated 
LSZ/TTAB/SWNT
(purified SWNT) 
fiber 
3.48 ? 0.89 49.12 ? 10.95 6.30 ? 7.48 
PVA coagulated 
LSZ/TTAB/SWNT 
(non purified 
SWNT) fiber 
5.23 ? 1.00 67.08 ? 16.13 18.84 ? 13.11 
1-butanol 
coagulated 
dsDNA/SWNT 
fibers 
0.77 ? 0.32 16.15 ? 4.92 8.30 ? 5.01 
Dry spun 
dsDNA/SWNT 
fibers 
1.28 ? 0.25 24.42 ? 5.53 8.33 ? 3.81 
 
5.5. Conclusions 
In general, PVA coagulated LSZ/TTAB/SWNT fibers showed better mechanical and 
electrical properties than dsDNA/SWNT fibers produced by both wet and dry spinning methods. 
The PVA polymer provided a strong bridging effect between SWNT bundles and improved the 
mechanical properties of fibers with a compromise in lowering the electrical properties. The 
electrical resistivities of dsDNA/SWNT fibers were extremely high especially for wet spinning 
fibers in 1-butanol. In order to improve the electrical properties of dsDNA/SWNT, one possible 
approach is to decrease the dsDNA amount in the initial dispersion, but it needs to be sufficient 
enough to stabilize SWNT and prevent aggregation of nanotubes. Moreover, optimization of 
both intial SWNT dispersions and fiber spinning schemes is highly required. 
 
166 
  
 
 
 
 
Chapter 6 
Phase Behavior of Lyotropic CNC Suspensions 
This section reported the results of a separate collaborative research with Dr. Christopher 
L. Kitchens at Clemson University. Sulfonated cellulose nanocrystal (CNC) suspensions were 
prepared at Clemson University. The phase behavior of CNC suspensions were analyzed by 
rheology and optical microscopy by this work. The rheology of lyotropic CNC suspensions in 
the wide concentration range from isotropic to fully liquid crystalline regime was investigated to 
help understand the phase behavior of cholesteric phase forming systems. In this work, the 
temperature and concentration effects on the phase behavior of lyotropic CNC suspensions were 
studied by rheology, polarized light microscopy, and rheo-optical measurements using the 
Linkam shear cell.  
 
6.1. Preparation of CNC Suspensions 
The CNC used were isolated from cotton cellulose via acid hydrolysis using 64 % 
sulfuric acid at 45 oC for 50 min. The final products were washed with deionized water for 
several days until the pH remained constant. Aqueous CNC suspensions at different 
concentrations were prepared by dividing the same batch into several samples and slowly 
evaporating the water under a nitrogen stream. CNC density of 1.64 g/ml and water density of 
0.9970 g/ml were used to convert the unit of CNC weight percent into volume percent. The size 
distribution of samples determined by AFM showed the average length of 107 nm with a 
standard deviation (s) of 55 nm, the average width of 20.5 nm (s=6.1 nm), and the average height 
 
167 
of 8.9 nm (s=2.8 nm) which has rectangular cross sections. Although washed with deionized 
water many times, the CNC suspensions are still acidic due to the introduction of ?SO3H group 
on the nanocrystals surface during the acid hydrolysis process. For example, the pH value of a 
suspension of 1.59 vol % is 2.9 and become more acidic with increasing concentration. The 
acidic nature of CNC suspensions used in this work will affect its phase behavior. A detailed 
description of material preparation is given in Ure?a-Benavides et al. 2011.91  
 
6.2. Polarized Optical Microscopy of CNC Suspensions 
The microstructure of liquid crystalline CNC suspensions was studied by polarized light 
microscopy. As shown in Figure 6.1, at 5.00 vol % only a small amount of liquid crystalline 
droplets were randomly dispersed in a continuous isotropic phase. The droplets of the anisotropic 
phase turned bright and dark when rotating the sample under polarized light microscopy while 
the continuous isotropic phase remained unchanged. With increasing concentration, the volume 
of liquid crystal phase increased as can be seen in 6.30 vol % sample. At 9.03 vol % the biphasic 
system consisted of almost uniformly dispersed isotropic and anisotropic phases at a larger 
region showing a strong birefringent texture. With further increase in concentration to 9.91 vol %, 
the sample exhibited colorful discontinuous stripes which are a characteristic texture of 
precholesteric ordering. The presence of an isotropic phase was still detected at the edge of 
microscopic samples up to 10.4 vol % of CNC, although most of the regions appeared to the 
birefringent and changed colors when rotated between cross-polarizers. When the concentration 
was increased to 12.1 vol %, the suspension was entirely birefringent including the edge. 
Moreover, several domains of uniform stripes resembling fingerprint textures were observed 
throughout the microscopic sampling indicating a cholesteric liquid crystal formation. As a 
 
168 
comparison, the fingerprint texture observed in cholesteric dsDNA/SWNT dispersion was much 
more obvious. Based on these cross polarized optical microscopy results, the critical 
concentration of liquid crystal formation lies in the range of 10.4 vol % < LC? <12.1 vol % which 
is much higher compared to ~2.22 vol % SWNT for cholesteric dsDNA/SWNT. According to 
Onsager theory, rods with low aspect ratio form liquid crystal phase at higher concentration. It is 
reasonable that CNC formed liquid crystal at higher concentration because the aspect ratio of 
CNC was much smaller compared to that of dsDNA interacted SWNT. When the concentration 
reached 14.5 vol %, the texture was more random although different domains change colors 
when rotated under cross-polarizers. However, the system behaved as a gel and no evidence of 
the cholesteric stripe texture was observed. The change in microstructure was further evidenced 
by the frequency independence of the storage modulus measured at 10 oC.  
 
 
 
 
169 
 
Figure 6.1. Polarized light microscopy of CNC suspensions at various volume percent 
concentrations at a room temperature. Scale bars are 50 ?m. 
 
The microstructural evolution of CNC liquid crystals with temperature was investigated 
using polarized light microscopy coupled with a Linkam optical rheology system. Figure 6.2 
shows temperature effects on the phase behavior of a representative CNC liquid crystal sample 
of 12.1 vol %. The microscopic images were taken after 12.1 vol % CNC was heated and held 
for an hour at 35 and 40 oC, respectively. The microstructure of CNC at 35 oC remained similar 
to the one at ambient temperature during the experimental period. However, after heating at 40 
oC, the liquid crystal phase behavior appeared to be changing based on the vanishing of the 
sample colors under polarized optical microscopy. The change from colorful to bright and dark 
samples provides evidence that the single phase CNC liquid crystal transitioned into the biphasic 
regime after heating at 40 oC. The steady shear rheology of 12.1 vol % CNC at 40 oC also 
 
170 
supports this idea; the viscosity increased noticeably at 40 oC, while at 10, 25, and 35 oC, the 
viscosity remained unchanged. More details on concentration and temperature effects on CNC 
rheology are discussed in the following section.  
 
 
Figure 6.2. Cross polarized microscopy images at 24, 35 and 40 ?C of a sample containing 12.1 
vol % CNC. The scale bar represents 50 ?m. 
 
6.3. Rheology of CNC Suspensions 
The phase behavior of lyotropic CNC suspensions as a function of concentration and 
temperature were studied by rheology. In steady shear measurements, the transient behavior of 
various CNC suspensions were measured before steady shear viscosity to determine the time 
period required for becoming steady state. Figure 6.3 shows the start up of flow of viscosity 
versus time for CNC suspensions at different concentrations in biphasic, liquid crystal, and gel 
regime. Similar to both bulk and supernatant of dsDNA/SWNT dispersions, the anisotropic CNC 
suspensions also showed short transient behavior. All oscillatory shear data of CNC suspensions 
were measured in the linear viscoelastic (LVE) region.  
 
 
171 
 
Figure 6.3. Transient behavior of CNC suspensions at various concentrations of biphasic, liquid 
crystal and gel regime.    
 
Effects of Temperature on Viscosity. As mentioned earlier, many of the nanocylinder 
liquid crystal dispersions including SWNT in superacids19 and low aspect ratio 
CdSe-cylcohexane178 showed temperature insensitivity in phase behavior. The rheological 
behavior of CNC suspensions as a function of temperature was investigated at three different 
CNC concentrations which were representative of the isotropic, biphasic and liquid crystal 
phases. As shown in Figure 6.4a, at the isotropic concentration of 3.07 vol %, the steady shear 
viscosity decreased with increasing temperature; this is a common behavior for predominantly 
isotropic dispersions. Furthermore, the wide Newtonian plateau in viscosity was observed at all 
temperatures measured at this concentration. The inception of the shear thinning region at 30 s-1 
shear rate was apparent at both 10 and 25 oC.  
     For the biphasic sample of 6.99 vol % CNC, clear evidence of three region flow behavior, 
 
172 
a typical rheological signature for lyotropic LCPs systems, was observed as shown in Figure 6.4b. 
However, for cholesteric dsDNA/SWNT dispersions, only slight deflection points separating two 
shear thinning regions with different slopes were observed. At the lower shear rate region, all 
viscosities at three different temperatures showed Region I shear thinning and followed by 
apparent Region II plateaus starting from about 0.5 s-1 shear rate with the most obvious plateau at 
10 oC. The intermediate Region II resembling low shear rate ?Newtonian plateau? is generally 
associated with director wagging48 or vorticity alignment.226 At even higher shear rates, samples 
showed Region III shear alignment. Overall, the viscosities at 6.99 vol % CNC concentration 
were still affected by temperature which decreased as a function of increasing temperature. 
With a further increase in concentration to a fully liquid crystalline sample of 12.1 vol % 
CNC, the viscosity was temperature independent between 10 and 35 oC with shear thinning 
behavior throughout the measured shear rate range (Figure 6.4c). This indicates that, even for 
low aspect ratio CNC nanorods, the viscosity is insensitive to temperature when it becomes a 
liquid crystal. However, at 40oC, a marked increase in viscosity was observed especially at lower 
shear rates. This was initially thought to be due to the evaporation of solvent during the 
measurements. Another possible reason for the viscosity increase was that, with the temperature 
increase from 35 to 40 oC, the microstructure of liquid crystal CNC may transition into biphasic 
phase leading to an increase in viscosity. Hot stage optical microscopy revealed a pronounced 
increase in isotropic domains when heating from 35 to 40 oC, and the color texture vanished with 
time showing only bright and dark domains. This suggests that there was a change in phase 
behavior between 35 and 40 oC. Similar behavior has been observed for rod like polymers since 
in those systems increasing temperature resulted in a changed solvent quality. 
 
173 
 
Figure 6.4. Steady shear viscosity versus shear rate at 10 (red diamonds), 25 (blue squares), 35 
(orange triangles) and 40 ?C (purple circles) of CNC suspensions with concentrations of (a) 3.07 
vol %, (b) 6.99 vol %, and (c) 12.1 vol %. 
 
174 
Viscosity and Modulus versus Concentration. In a normal polymer solution or 
colloidal dispersion, the low shear viscosity increases with concentration. However, for liquid 
crystalline dispersions, one of the hallmarks is that the low shear viscosity versus concentration 
curve goes through a maximum in the biphasic regime. Similar behavior has also been reported 
for the elastic and viscous modulii.48 Figure 6.5a shows the complex and steady shear viscosity 
change as a function of increasing CNC concentration at 0.1 and 1.0 s-1 shear rate or angular 
frequency. Although no maximum was obtained for steady shear viscosity through the entire 
concentration range measure, a change in slope occurred in the biphasic regime at 7.69 vol % 
CNC concentration. This contrast to the viscosity maximum observed for both nematic and 
cholesteric dsDNA/SWNT dispersions and the typical rheological behavior of LCPs. On the 
other hand, slight maximum peaks were obtained for complex viscosity ?* and storage modulus 
G? at 12.1 vol % CNC concentration, which is fully liquid crystal according to optical 
microscopy (Figure 6.5ab). No maxima were observed in the biphasic regime. With further 
increase in concentration, the samples evolved to gel state where viscosities and storage and loss 
modulii became independent of concentration. The origin of the discrepancy between this system 
with LCPs is not clear. However, it is interesting to note that another micron length CNC system 
also showed unexpected behavior where only a tiny maximum was obtained for steady shear 
viscosity versus concentration curve at the critical concentration I?  where the sample became 
biphasic.227 Considering the acidic nature of the CNC suspensions, it is hypothesized that the 
change in the ionic strength with increasing CNC concentration caused changes in attractive 
interactions of molecules and resulted in this anomalous behavior. The impact of ionic strength 
on the phase behavior and attractive interactions needs to be explored in the future research. 
 
175 
 
Figure 6.5. Rheological properties versus concentration; (a) complex viscosity at 0.1 s-1 (red 
diamonds), 1.0 s-1 (blue squares), and steady shear viscosity at 0.1 s-1 (orange triangles), 1.0 s-1 
(purple circles); (b) storage and (c) loss modulus at 0.1 s-1 (red diamonds), 1.0 s-1 (blue squares), 
10 s-1 (orange triangles) and 100 s-1 (purple circles). The solid vertical lines indicate a phase 
transition; the discontinuous vertical line indicates an approximate liquid crystal to gel transition. 
(a)
(b)
(c)
 
176 
6.4. Conclusions 
Aqueous dispersions of sulfonated CNC produced from cotton show lyotropic phase 
behavior. Based on cross-polarized optical microscopy, a liquid crystalline phase was obtained in 
the concentration range of 10.4 vol % < ?LC < 12.1 vol % and exhibited the fingerprint texture 
characteristic of a cholesteric liquid crystal. At even higher concentration, the system 
transitioned to a birefringent gel. The temperature effect on the change in phase behavior 
between 35 to 40 oC was also detected. Although the typical nonmonotonic relationship of 
viscosity and modulus in LCPs was not observed for CNC suspensions, the three-region 
rheological behavior of LCPs was obtained for biphasic and liquid crystalline samples of CNC 
suspensions. The discrepancy of phase behavior between CNC nanorod suspensions and LCPs 
may have been caused by polydispersity and introduction of electrostatic interactions in the CNC 
system. This hypothesis provides a foundation for future research.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
177 
 
 
 
 
 
Chapter 7 
Conclusions 
Integrating carbon nanotubes (CNTs) with natural biopolymers, especially dsDNA, have 
broadened the potential applications of remarkable nanomaterials to biofunctional materials as 
well as other engineering applications. The characteristics of nature?s toolbox could be exploited 
to help understand and process nanomaterials to achieve a breakthrough in nanotechnology. 
Carbon nanotubes are considered to be one of the most promising materials in nanoscale science 
and technology with great potential applications. However, translating the remarkable 
mechanical, thermal and electrical properties of individual nanotubes into macroscopic objects of 
neat films and fibers is scientifically challenged by the isotropic random entanglement of pristine 
SWNT. Understanding the phase behavior of biopolymers-stabilized CNTs is important for the 
development of nanotube based biomaterials as well as other types of nanomaterials. Since many 
biopolymers, especially DNA, are proven to be good dispersing agents for CNTs, uniform 
dispersion of nanotubes in small bundles or even in individuals at the first stage can be 
successfully obtained by binding biomolecules to CNTs either by covalent or non-covalent 
functionalization.  
In this study, both cholesteric and nematic liquid crystal phases of pristine SWNT 
stabilized in aqueous dsDNA solvent were observed. When dsDNA/SWNT liquid crystals were 
produced from a bulk dispersion, the presence of SWNT bundles appeared to inhibit cholesteric 
phase formation and the nematic texture typical of other lyotropic SWNT liquid crystals resulted. 
When the liquid crystal was produced from a supernatant comprised only of individual SWNT, 
 
178 
the helicity of the dsDNA was able to drive the assembly of a cholesteric liquid crystal. To the 
author?s knowledge, this is both the first report of biopolymer stabilized lyotropic cholesteric 
SWNT liquid crystalline phase and one of a very few systems capable of forming either 
cholesteric or nematic phases. The ultimate microstructure of films produced from the 
cholesteric dispersions was found to be shear dependent. These results will enable new 
fundamental investigations comparing nematic and cholesteric liquid crystalline phase behavior 
and shear response. In addition, they potentially provide a foundation for previously 
unachievable applications that utilize both the intrinsic properties of SWNT and the controlled 
optical properties of cholesteric films. Furthermore, preliminary investigation of the macroscopic 
assembly of fibers from dsDNA/SWNT and LSZ/TTAB/SWNT has shown sufficient promise to 
warrant additional research.   
Significant insights were gained into the rheology and phase behavior of aqueous 
dsDNA/SWNT. In the semidilute regime, the supernatant of dsDNA/SWNT dispersion followed 
the predicted rheological behavior given by Doi-Edwards theory. In the fully liquid crystal 
regime, the nematic and cholesteric dsDNA/SWNT dispersions showed different rheological 
behavior in steady shear viscosities at SWNT concentrations near critical concentration of ?LC. 
In addition, slow evaporation led to phase transitions from isotropic to liquid crystalline phase 
which were accompanied by the characteristic nonmonotonic relationship of rheology as a 
function of increasing SWNT concentration. As expected, the empirical Cox-Merz rule was not 
obeyed for the dsDNA/SWNT liquid crystals. However, not all of the rheological signatures 
anticipated based on lyotropic liquid crystal polymers (LCPs) were observed. Although plots of 
steady shear viscosity versus shear rate for several liquid crystal samples showed two distinct 
shear thinning regions, they lacked the obvious three region behavior characteristic of many 
 
179 
LCPs. In addition to dsDNA/SWNT system, the phase behavior and microstructure of another 
cholesterogenic system, cellulose nanocrystal (CNC) suspensions, were investigated as a 
function of concentration and temperature. For sulfonated CNC suspensions, only slight maxima 
were obtained in liquid crystal region for complex viscosity and storage modulus as a function of 
increasing concentration. However, steady shear viscosity and loss modulus did not show 
maximum values throughout the entire concentration range investigated. Interestingly, typical 
three-region rheological behavior of LCPs was observed for CNC suspensions in biphasic and 
liquid crystalline phases. In addition, temperature study of CNC suspensions showed a possible 
phase transition in liquid crystal phases at higher temperature of 40 oC.  
These results of this research highlight that while nanorod dispersions have many 
parallels with rodlike polymer solutions, the complicance of polydispersity and various 
molecular interactions such as van der Waals forces and electrostatic interactions can result in 
unusual phase behaviors. Understanding the rheology and phase behavior of the dispersions is 
critical to controlling the properties of bulk materials such as films and fibers comprised of 
nanoscale building blocks.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
180 
 
 
 
 
 
References 
 
1 Nepal, D., Balasubramanian, S., Simonian, A. L. & Davis, V. A. Strong antimicrobial 
coatings: Single-walled carbon nanotubes armored with biopolymers. Nano Letters 8, 
1896 (2008). 
2 Kang, S., Pinault, M., Pfefferle, L. D. & Elimelech, M. Single-walled carbon nanotubes 
exhibit strong antimicrobial activity. Langmuir 23, 8670 (2007). 
3 Pangule, R. C. et al. Antistaphylococcal nanocomposite films based on enzyme- 
nanotube conjugates. ACS Nano, 874 (2010). 
4 Shin, S. R. et al. DNA-wrapped single-walled carbon nanotube hybrid fibers for 
supercapacitors and artificial muscles. Advanced Materials 20, 466 (2008). 
5 Collins, P. G., Bradley, K., Ishigami, M. & Zettl, A. Extreme oxygen sensitivity of 
electronic properties of carbon nanotubes. Science 287, 1801 (2000). 
6 Li, J. et al. Carbon nanotube nanoelectrode array for ultrasensitive DNA detection. Nano 
Letters 3, 597 (2003). 
7 Tans, S. J. et al. Individual single-wall carbon nanotubes as quantum wires. Nature 386, 
474 (1997). 
8 Cao, J., Wang, Q., Wang, D. & Dai, H. Suspended carbon nanotube quantum wires with 
two gates. Small 1, 138 (2005). 
9 Dillon, A. C. et al. Storage of hydrogen in single-walled carbon nanotubes. Nature 386, 
377 (1997). 
10 Ericson, L. M. et al. Macroscopic, neat, single-walled carbon nanotube fibers. Science 
305, 1447 (2004). 
11 Poulin, P., Vigolo, B. & Launois, P. Films and fibers of oriented single wall nanotubes. 
Carbon 40, 1741 (2002). 
12 Dalton, A. B. et al. Super-tough carbon-nanotube fibres. Nature 423, 703 (2003). 
13 Zhang, M., Atkinson, K. R. & Baughman, R. H. Multifunctional carbon nanotube yarns 
by downsizing an ancient technology. Science 306, 1358 (2004). 
14 Vigolo, B. et al. Macroscopic fibers and ribbons of oriented carbon nanotubes. Science 
290, 1331 (2000). 
 
181 
15 Song, W. H., Kinloch, I. A. & Windle, A. H. Nematic liquid crystallinity of multiwall 
carbon nanotubes. Science 302, 1363 (2003). 
16 Badaire, S. et al. Liquid crystals of DNA-stabilized carbon nanotubes. Advanced 
Materials 17, 1673 (2005). 
17 Moulton, S. E., Maugey, M., Poulin, P. & Wallace, G. G. Liquid crystal behavior of 
single-walled carbon nanotubes dispersed in biological hyaluronic acid solutions. Journal 
of the American Chemical Society 129, 9452 (2007). 
18 Islam, M. F. et al. Nematic nanotube gels. Physical Review Letters 92, 88303 (2004). 
19 Davis, V. A. et al. Phase behavior and rheology of swnts in superacids. Macromolecules 
37, 154 (2004). 
20 Davis, V. A. et al. True solutions of single-walled carbon nanotubes for assembly into 
macroscopic materials. Nature Nanotechnology 4, 830 (2009). 
21 Zakri, C. & Poulin, P. Phase behavior of nanotube suspensions: From attraction induced 
percolation to liquid crystalline phases. Journal of Materials Chemistry 16, 4095 (2006). 
22 Orts, W. J., Godbout, L., Marchessault, R. H. & Revol, J. F. Enhanced ordering of liquid 
crystalline suspensions of cellulose microfibrils: A small angle neutron scattering study. 
Macromolecules 31, 5717 (1998). 
23 Pan, J. H., Hamad, W. & Straus, S. K. Parameters affecting the chiral nematic phase of 
nanocrystalline cellulose films. Macromolecules 43, 3851 (2010). 
24 Davis, V. A. Liquid crystalline assembly of nanocylinders. Journal of Materials 
Research (2011). 
25 Nepal, D. & Geckeler, K. E. Proteins and carbon nanotubes: Close encounter in water. 
Small 3, 1259 (2007). 
26 Nepal, D., Sohn, J. I., Aicher, W. K., Lee, S. & Geckeler, K. E. Supramolecular 
conjugates of carbon nanotubes and DNA by a solid-state reaction. Biomacromolecules 6, 
2919 (2005). 
27 Hobbie, E. K. et al. Self-assembly of ordered nanowires in biological suspensions of 
single-wall carbon nanotubes. ACS Nano 3, 189 (2009). 
28 Onsager, L. The effects of shape on the interaction of colloidal particles. Annals of the 
New York Academy of Sciences 51, 627 (1949). 
29 Adams, M., Dogic, Z., Keller, S. L. & Fraden, S. Entropically driven microphase 
transitions in mixtures of colloidal rods and spheres. Nature 393, 349 (1998). 
 
182 
30 Dogic, Z., Frenkel, D. & Fraden, S. Enhanced stability of layered phases in parallel hard 
spherocylinders due to addition of hard spheres. Physical Review E 62, 3925 (2000). 
31 Strzelecka, T. E., Davidson, M. W. & Rill, R. L. Multiple liquid-crystal phases of DNA 
at high-concentrations. Nature 331, 457 (1988). 
32 Nakata, M. et al. End-to-end stacking and liquid crystal condensation of 6-to 20-base pair 
DNA duplexes. Science 318, 1276 (2007). 
33 Vanwinkle, D. H., Davidson, M. W. & Rill, R. L. Terraces in the cholesteric phase of 
DNA liquid-crystals. Journal of Chemical Physics 97, 5641 (1992). 
34 Zakharova, S. S., Jesse, W., Backendorf, C. & van der Maarel, J. R. C. Liquid crystal 
formation in supercoiled DNA solutions. Biophysical Journal 83, 1119 (2002). 
35 Sipski, M. L. & Wagner, T. E. Probing DNA quaternary ordering with circular-dichroism 
spectroscopy - studies of equine sperm chromosomal fibers. Biopolymers 16, 573 (1977). 
36 Livolant, F. Cholesteric organization of DNA invivo and invitro. European Journal of 
Cell Biology 33, 300 (1984). 
37 Rill, R. L., Livolant, F., Aldrich, H. C. & Davidson, M. W. Electron-microscopy of 
liquid-crystalline DNA - direct evidence for cholesteric-like organization of DNA in 
dinoflagellate chromosomes. Chromosoma 98, 280 (1989). 
38 Baek, S. G., Magda, J. J., Larson, R. G. & Hudson, S. D. Rheological differences among 
liquid-crystalline polymers. Ii. Disappearance of negative n in densely packed lyotropes 
and thermotropes. Journal of Rheology 38, 1473 (1994). 
39 Huang, C. M., Magda, J. J. & Larson, R. G. The effect of temperature and concentration 
on n and tumbling in a liquid crystal polymer. Journal of Rheology 43, 31 (1999). 
40 Hongladarom, K., Secakusuma, V. & Burghardt, W. R. Relation between molecular 
orientation and rheology in lyotropic hydroxypropylcellulose solutions. Journal of 
Rheology 38, 1505 (1994). 
41 Walker, L. M., Wagner, N. J., Larson, R. G., Mirau, P. A. & Moldenaers, P. The 
rheology of highly concentrated pblg solutions. Journal of Rheology 39, 925 (1995). 
42 Larson, R. G. & Mead, D. W. Time and shear-rate scaling laws for liquid crystal 
polymers. Journal of Rheology 33, 1251 (1989). 
43 Baek, S. G., Magda, J. J. & Larson, R. G. Rheological differences among 
liquid-crystalline polymers. I. The first and second normal stress differences of pbg 
solutions. Journal of Rheology 37, 1201 (1993). 
 
183 
44 Hongladarom, K. & Burghardt, W. R. Molecular alignment of polymer liquid crystals in 
shear flows. 2. Transient flow behavior in poly (benzyl glutamate) solutions. 
Macromolecules 26, 785 (1993). 
45 O'Connell, M. J. et al. Band gap fluorescence from individual single-walled carbon 
nanotubes. Science 297, 593 (2002). 
46 Girifalco, L. A., Hodak, M. & Lee, R. S. Carbon nanotubes, buckyballs, ropes, and a 
universal graphitic potential. Physical Review B 62, 13104 (2000). 
47 Doi, M. & Edwards, S. F. The theory of polymer dynamics. 1986. (Oxford University 
Press, Clarendon). 
48 Larson, R. G. The structure and rheology of complex fluids. 1999. (Oxford University 
Press, New York). 
49 Marrucci, G. Rheology of nematic polymers. In Liquid Crystallinity in Polymers: 
Principles and Fundamental Properties, A. Ciferri, ed. 1991(VCH, New York). 
50 Kroto, H. W., Heath, J. R., O'Brien, S. C., Curl, R. F. & Smalley, R. E. C 60: 
Buckminsterfullerene. Nature 318, 162 (1985). 
51 Iijima, S. Helical microtubules of graphitic carbon. Nature 354, 56 (1991). 
52 Iijima, S. & Ichihashi, T. Single-shell carbon nanotubes of 1-nm diameter (vol 363, pg 
603, 1993). Nature 364, 737 (1993). 
53 Bethune, D. S. et al. Cobalt-catalyzed growth of carbon nanotubes with 
single-atomic-layerwalls. Nature 363, 605 (1993). 
54 Thess, A. et al. Crystalline ropes of metallic carbon nanotubes. Science 273, 483 (1996). 
55 Journet, C. et al. Large-scale production of single-walled carbon nanotubes by the 
electric-arc technique. Nature 388, 756 (1997). 
56 Liu, J. et al. Fullerene pipes. Science 280, 1253 (1998). 
57 Scott, C. D. et al. Iron catalyst chemistry in modeling a high-pressure carbon monoxide 
nanotube reactor. Journal of Nanoscience and Nanotechnology 3, 63 (2003). 
58 Tsang, S. C. et al. Immobilization of small proteins in carbon nanotubes: High-resolution 
transmission electron microscopy study and catalytic activity (september, pg 1803, 1995). 
Journal of the Chemical Society-Chemical Communications, 2579 (1995). 
59 Tsang, S. C. et al. Immobilization of platinated and iodinated oligonucleotides on carbon 
nanotubes. Angewandte Chemie-International Edition in English 36, 2198 (1997). 
 
184 
60 Fujita, M., Saito, R., Dresselhaus, G. & Dresselhaus, M. S. Formation of general 
fullerenes by their projection on a honeycomb lattice. Physical Review B 45, 13834 
(1992). 
61 Moniruzzaman, M. & Winey, K. I. Polymer nanocomposites containing carbon 
nanotubes. Macromolecules 39, 5194 (2006). 
62 Dresselhaus, M. S., Dresselhaus, G. & Saito, R. Carbon fibers based on c60 and their 
symmetry. Physical Review B 45, 6234 (1992). 
63 Wildoer, J. W. G., Venema, L. C., Rinzler, A. G., Smalley, R. E. & Dekker, C. Electronic 
structure of atomically resolved carbon nanotubes. Nature 391, 59 (1998). 
64 Dresselhaus, M. S., Dresselhaus, G. & Saito, R. Physics of carbon nanotubes. Carbon 33, 
883 (1995). 
65 Baughman, R. H., Zakhidov, A. A. & de Heer, W. A. Carbon nanotubes - the route 
toward applications. Science 297, 787 (2002). 
66 Che, J. W., Cagin, T. & Goddard, W. A. Thermal conductivity of carbon nanotubes. 
Nanotechnology 11, 65 (2000). 
67 Hone, J. et al. Electrical and thermal transport properties of magnetically aligned single 
walt carbon nanotube films. Applied Physics Letters 77, 666 (2000). 
68 Anthony, T. R. et al. Thermal diffusivity of isotopically enriched 12c diamond. Physical 
Review B 42, 1104 (1990). 
69 Berber, S., Kwon, Y. K. & Tomanek, D. Unusually high thermal conductivity of carbon 
nanotubes. Physical Review Letters 84, 4613 (2000). 
70 Krishnan, A., Dujardin, E., Ebbesen, T. W., Yianilos, P. N. & Treacy, M. M. J. Young's 
modulus of single-walled nanotubes. Physical Review B 58, 14013 (1998). 
71 Gao, G. H., Cagin, T. & Goddard, W. A. Energetics, structure, mechanical and 
vibrational properties of single-walled carbon nanotubes. Nanotechnology 9, 184 (1998). 
72 Sinden, R. R. DNA structure and function. 1994. (Academic Press). 
73 Vanoss, C. J., Smeenk, R. J. T. & Aarden, L. A. Inhibition of association vs dissociation 
of high-avidity DNA anti-DNA complexes - possible involvement of secondary 
hydrogen-bonds. Immunological Investigations 14, 245 (1985). 
74 Marmur, J. & Doty, P. Determination of base composition of deoxyribonucleic acid from 
its thermal denaturation temperature. Journal of Molecular Biology 5, 109 (1962). 
75 Jones, R. A. L. Soft condensed matter. 2002. (Oxford University Press, USA). 
 
185 
76 Nepal, D. & Geckeler, K. E. Ph-sensitive dispersion and debundling of single-walled 
carbon nanotubes: Lysozyme as a tool. Small 2, 406 (2006). 
77 da Silva, M. A., Farhat, I. A., Areas, E. P. G. & Mitchell, J. R. Solvent-induced lysozyme 
gels: Effects of system composition and temperature on structural and dynamic 
characteristics. Biopolymers 83, 443 (2006). 
78 Matsuura, K. et al. Selectivity of water-soluble proteins in single-walled carbon nanotube 
dispersions. Chemical Physics Letters 429, 497 (2006). 
79 White, J. S. & White, D. C. Sourcebook of enzymes. 1997. (CRC). 
80 da Silva, M. A. & Areas, E. P. G. Solvent-induced lysozyme gels: Rheology, fractal 
analysis, and sol-gel kinetics. Journal of Colloid and Interface Science 289, 394 (2005). 
81 Szymanska, A., Hornowski, T., Kozak, M. & Slosarek, G. The saxs and rheological 
studies of hewl amyloid formation. Acta Physica Polonica A 114, 447 (2008). 
82 Eichhorn, S. J. Cellulose nanowhiskers: Promising materials for advanced applications. 
Soft Matter 7, 303. 
83 Dong, X. M., Kimura, T., Revol, J. F. & Gray, D. G. Effects of ionic strength on the 
isotropic-chiral nematic phase transition of suspensions of cellulose crystallites. 
Langmuir 12, 2076 (1996). 
84 Elazzouzi-Hafraoui, S. et al. The shape and size distribution of crystalline nanoparticles 
prepared by acid hydrolysis of native cellulose. Biomacromolecules 9, 57 (2007). 
85 Samir, M. A. S. A., Alloin, F. & Dufresne, A. Review of recent research into cellulosic 
whiskers, their properties and their application in nanocomposite field. 
Biomacromolecules 6, 612 (2005). 
86 Revol, J. F. et al. Chiral nematic suspensions of cellulose crystallites; phase separation 
and magnetic field orientation. Liquid Crystals 16, 127 (1994). 
87 Sakurada, I., Nukushina, Y. & Ito, T. Experimental determination of the elastic modulus 
of crystalline regions in oriented polymers. Journal of Polymer Science 57, 651 (1962). 
88 Tashiro, K. & Kobayashi, M. Theoretical evaluation of three-dimensional elastic 
constants of native and regenerated celluloses: Role of hydrogen bonds. Polymer 32, 
1516 (1991). 
89 Uematsu, Y. & Uematsu, I. Mesomorphic Order in Polymers and Polymerization in 
Liquid Crystalline Media. ed. A. Blumstein ACS Symposium Series 74, 136 (1978). 
90 Guo, J. X. & Gray, D. G. In: Gilbert rd (ed) cellulosic polymers, blends and composites. 
1994. (Hanser Publishers, New York). 
 
186 
91 E. E. Ure?a-Benavides, G. A., V. A. Davis, C. L. Kitchens. Rheology and phase behavior 
of concentrated cellulose nanocrystal aqueous suspensions. Macromolecules 44, 8990 
(2011). 
92 Wong, S. S., Joselevich, E., Woolley, A. T., Cheung, C. L. & Lieber, C. M. Covalently 
functionalized nanotubes as nanometre-sized probes in chemistry and biology. Nature 
394, 52 (1998). 
93 Weiss, P. S. A conversation with prof. Angela belcher: Leader in biotemplated 
nanomaterials. ACS Nano 2, 1508 (2008). 
94 Shim, M., Kam, N. W. S., Chen, R. J., Li, Y. M. & Dai, H. J. Functionalization of carbon 
nanotubes for biocompatibility and biomolecular recognition. Nano Letters 2, 285 (2002). 
95 Bottini, M. et al. Adsorption of streptavidin onto single-walled carbon nanotubes: 
Application in fluorescent supramolecular nanoassemblies. Journal of Nanoscience and 
Nanotechnology 6, 3693 (2006). 
96 Sanchez-Pomales, G., Santiago-Rodriguez, L. & Cabrera, C. R. DNA-functionalized 
carbon nanotubes for biosensing applications. Journal of Nanoscience and 
Nanotechnology 9, 2175 (2009). 
97 Albertorio, F., Hughes, M. E., Golovchenko, J. A. & Branton, D. Base dependent 
DNA-carbon nanotube interactions: Activation enthalpies and assembly-disassembly 
control. Nanotechnology 20 (2009). 
98 Anand, S. V. & Mahapatra, D. R. Quasi-static and dynamic strain sensing using carbon 
nanotube/epoxy nanocomposite thin films. Smart Materials & Structures 18 (2009). 
99 Liu, Z. et al. Drug delivery with carbon nanotubes for in vivo cancer treatment. Cancer 
Research 68, 6652 (2008). 
100 Pantarotto, D. et al. Functionalized carbon nanotubes for plasmid DNA gene delivery. 
Angewandte Chemie 116, 5354 (2004). 
101 Bianco, A., Kostarelos, K. & Prato, M. Applications of carbon nanotubes in drug delivery. 
Current Opinion in Chemical Biology 9, 674 (2005). 
102 Shvedova, A. A. et al. Unusual inflammatory and fibrogenic pulmonary responses to 
single-walled carbon nanotubes in mice. American Journal of Physiology-Lung Cellular 
and Molecular Physiology 289, L698 (2005). 
103 Chou, C. C. et al. Single-walled carbon nanotubes can induce pulmonary injury in mouse 
model. Nano Letters 8, 437 (2008). 
104 Warheit, D. B. et al. Comparative pulmonary toxicity assessment of single-wall carbon 
nanotubes in rats. Toxicological Sciences 77, 117 (2004). 
 
187 
105 Johnson, D. R., Methner, M. M., Kennedy, A. J. & Steevens, J. A. Potential for 
occupational exposure to engineered carbon-based nanomaterials in environmental 
laboratory studies. Environmental Health Perspectives 118, 49. 
106 Ao, G., Nepal, D., Aono, M. & Davis, V. A. Cholesteric and nematic liquid crystalline 
phase behavior of double-stranded DNA stabilized single-walled carbon nanotube 
dispersions. ACS Nano 5, 1450 (2011). 
107 Cathcart, H. et al. Ordered DNA wrapping switches on luminescence in single-walled 
nanotube dispersions. Journal of the American Chemical Society 130, 12734 (2008). 
108 Hughes, M. E., Brandin, E. & Golovchenko, J. A. Optical absorption of DNA-carbon 
nanotube structures. Nano Letters 7, 1191 (2007). 
109 Tu, X. M., Manohar, S., Jagota, A. & Zheng, M. DNA sequence motifs for 
structure-specific recognition and separation of carbon nanotubes. Nature 460, 250 
(2009). 
110 Pividori, M. & Alegret, S. DNA adsorption on carbonaceous materials. Immobilisation of 
DNA on Chips I, 1 (2005). 
111 Chen, R. J. & Zhang, Y. Controlled precipitation of solubilized carbon nanotubes by 
delamination of DNA. The Journal of Physical Chemistry B 110, 54 (2006). 
112 Meng, S., Maragakis, P., Papaloukas, C. & Kaxiras, E. DNA nucleoside interaction and 
identification with carbon nanotubes. Nano Letters 7, 45 (2007). 
113 Zheng, M. et al. DNA-assisted dispersion and separation of carbon nanotubes. Nature 
Materials 2, 338 (2003). 
114 Arnold, M. S., Green, A. A., Hulvat, J. F., Stupp, S. I. & Hersam, M. C. Sorting carbon 
nanotubes by electronic structure using density differentiation. Nature Nanotechnology 1, 
60 (2006). 
115 Feazell, R. P., Nakayama-Ratchford, N., Dai, H. & Lippard, S. J. Soluble single-walled 
carbon nanotubes as longboat delivery systems for platinum (iv) anticancer drug design. 
Journal of the American Chemical Society 129, 8438 (2007). 
116 Gao, H., Kong, Y., Cui, D. & Ozkan, C. S. Spontaneous insertion of DNA 
oligonucleotides into carbon nanotubes. Nano Letters 3, 471 (2003). 
117 Zhao, X. & Johnson, J. K. Simulation of adsorption of DNA on carbon nanotubes. 
Journal of the American Chemical Society 129, 10438 (2007). 
118 Okada, T., Kaneko, T., Hatakeyama, R. & Tohji, K. Electrically triggered insertion of 
single-stranded DNA into single-walled carbon nanotubes. Chemical Physics Letters 417, 
288 (2006). 
 
188 
119 Lu, G., Maragakis, P. & Kaxiras, E. Carbon nanotube interaction with DNA. Nano 
Letters 5, 897 (2005). 
120 Donald, A. M., Windle, A. H. & Hanna, S. Liquid crystalline polymers. 2006. 
(Cambridge University Press). 
121 Rizvi, T. Z. Liquid crystalline biopolymers: A new arena for liquid crystal research. 
Journal of Molecular Liquids 106, 43 (2003). 
122 Xin, X. et al. Incorporation of carbon nanotubes into a lyotropic liquid crystal by phase 
separation in the presence of a hydrophilic polymer. Langmuir 26, 3562. 
123 Rai, P. K. et al. Isotropic-nematic phase transition of single-walled carbon nanotubes in 
strong acids. Journal of the American Chemical Society 128, 591 (2006). 
124 Papkov, S. Liquid crystalline order in solutions of rigid-chain polymers. Liquid Crystal 
Polymers I, 75 (1984). 
125 Frenkel, D. Structure of hard-core models for liquid crystals. The Journal of Physical 
Chemistry 92, 3280 (1988). 
126 Green, M. J., Parra-Vasquez, A. N. G., Behabtu, N. & Pasquali, M. Modeling the phase 
behavior of polydisperse rigid rods with attractive interactions with applications to 
single-walled carbon nanotubes in superacids. Journal of Chemical Physics 131 (2009). 
127 Khokhlov, A. R. & Semenov, A. N. On the theory of liquid-crystalline ordering of 
polymer chains with limited flexibility. Journal of Statistical Physics 38, 161 (1985). 
128 Flory, P. J. & Ronca, G. Theory of systems of rodiike particles: Ii. Thermotropic systems 
with orientation-dependent interactions. Molecular Crystals and Liquid Crystals 54, 311 
(1979). 
129 Flory, P. J. Phase equilibria in solutions of rod-like particles. Proceedings of the Royal 
Society of London. Series A. Mathematical and Physical Sciences 234, 73 (1956). 
130 Flory, P. J. Statistical thermodynamics of mixtures of rodlike particles. 6. Rods connected 
by flexible joints. Macromolecules 11, 1141 (1978). 
131 Flory, P. J. & Abe, A. Statistical thermodynamics of mixtures of rodlike particles. 1. 
Theory for polydisperse systems. Macromolecules 11, 1119 (1978). 
132 Barry, E., Hensel, Z., Dogic, Z., Shribak, M. & Oldenbourg, R. Entropy-driven formation 
of a chiral liquid-crystalline phase of helical filaments. Physical Review Letters 96 
(2006). 
133 Lu, L. H. & Chen, W. Large-scale aligned carbon nanotubes from their purified, highly 
concentrated suspension. ACS Nano 4, 1042. 
 
189 
134 Hobbie, E. K. et al. Self-assembly of ordered nanowires in biological suspensions of 
single-wall carbon nanotubes. ACS Nano 3, 189 (2008). 
135 Russell, D. W. The enzymes, regulation, and genetics of bile acid synthesis. Annual 
Review of Biochemistry 72, 137 (2003). 
136 Tasis, D., Tagmatarchis, N., Bianco, A. & Prato, M. Chemistry of carbon nanotubes. 
Chemical Reviews 106, 1105 (2006). 
137 Dogic, Z. & Fraden, S. Cholesteric phase in virus suspensions. Langmuir 16, 7820 
(2000). 
138 Livolant, F. Cholesteric organization of DNA invivo and invitro. European Journal of 
Cell Biology 33, 300 (1984). 
139 Vanwinkle, D. H., Davidson, M. W., Chen, W. X. & Rill, R. L. Cholesteric helical pitch 
of near persistence length DNA. Macromolecules 23, 4140 (1990). 
140 Stanley, C. B., Hong, H. & Strey, H. H. DNA cholesteric pitch as a function of density 
and ionic strength. Biophysical Journal 89, 2552 (2005). 
141 Kornyshev, A. A., Leikin, S. & Malinin, S. V. Chiral electrostatic interaction and 
cholesteric liquid crystals of DNA. European Physical Journal E 7, 83 (2002). 
142 Wierenga, A. M. & Philipse, A. P. Low-shear viscosity of isotropic dispersions of 
(brownian) rods and fibres; a review of theory and experiments. Colloids and Surfaces A: 
Physicochemical and Engineering Aspects 137, 355 (1998). 
143 Marshall, B. D., Davis, V. A., Lee, D. C. & Korgel, B. A. Rotational and translational 
diffusivities of germanium nanowires. Rheologica Acta 48, 589 (2009). 
144 Mead, D. W. & Larson, R. G. Rheooptical study of isotropic solutions of stiff polymers. 
Macromolecules 23, 2524 (1990). 
145 Marceau, S., Dubois, P., Fulchiron, R. & Cassagnau, P. Viscoelasticity of brownian 
carbon nanotubes in pdms semidilute regime. Macromolecules 42, 1433 (2009). 
146 Mori, Y., Ookubo, N., Hayakawa, R. & Wada, Y. Low frequency and high frequency 
relaxations in dynamic electric birefringence of poly ( benzyl l glutamate) in m cresol. 
Journal of Polymer Science: Polymer Physics Edition 20, 2111 (1982). 
147 Kirkwood, J. G. & Plock, R. J. Non newtonian viscoelastic properties of rod like 
macromolecules in solution. The Journal of Chemical Physics 24, 665 (1956). 
148 Batchelor, G. K. The stress system in a suspension of force-free particles. Journal of 
Fluid Mechanics 41, 545 (1970). 
 
190 
149 Parra-Vasquez, A. N. G. et al. Simple length determination of single-walled carbon 
nanotubes by viscosity measurements in dilute suspensions. Macromolecules 40, 4043 
(2007). 
150 Duggal, R. & Pasquali, M. Dynamics of individual single-walled carbon nanotubes in 
water by real-time visualization. Physical Review Letters 96, 246104 (2006). 
151 Davis, V. A. Phase behavior and rheology of single-walled carbon nanotubes (swnts) in 
superacids with application to fiber spinning. Dissertation of Doctor of Philosophy, Rice 
University (2006). 
152 Ciferri, A. Liquid crystallinity in polymers: Principles and fundamental properties. 1991 
(Wiley-VCH). 
153 Walker, L. & Wagner, N. Rheology of region-i flow in a lyotropic liquid-crystal polymer 
- the effects of defect texture. Journal of Rheology 38, 1525 (1994). 
154 Mackley, M. R., Pinaud, F. & Siekmann, G. Observation of disclinations and optical 
anisotropy in a mesomorphic copolyester. Polymer 22, 437 (1981). 
155 Kiss, G. & Porter, R. S.   193 (John Wiley & Sons). 
156 Marrucci, G. & Maffettone, P. L. A description of the liquid-crystalline phase of rodlike 
polymers at high shear rates. Macromolecules 22, 4076 (1989). 
157 Larson, R. G. Arrested tumbling in shearing flows of liquid-crystal polymers. 
Macromolecules 23, 3983 (1990). 
158 Larson, R. G. Constitutive equations for polymer melts and solutions. Butterworths, 364 
(1988). 
159 Grizzuti, N., Cavella, S. & Cicarelli, P. Transient and steady-state rheology of a liquid 
crystalline hydroxypropylcellulose solution. Journal of Rheology 34, 1293 (1990). 
160 Semenov, A. N. Rheological properties of a liquid crystal solution of rod-like molecules. 
Soviet Physics JETP 66, 321 (1983) 
161 Maffettone, P. L. & Marrucci, G. The nematic dumbbell model. Journal of Rheology 36, 
1547 (1992). 
162 Moldenaers, P. & Mewis, J. Transient behavior of liquid crystalline solutions of poly 
(benzylglutamate). Journal of Rheology 30, 567 (1986). 
163 Sakamoto, K., MacKnight, W. J. & Porter, R. S. Dynamic and steady-shear melt rheology 
of and ethylene-methacrylic acid copolymer and its salts. Journal of Polymer Science 
Part A-2: Polymer Physics 8, 277 (1970). 
 
191 
164 Earnest Jr, T. R. & MacKnight, W. J. Effect of hydrogen bonding and ionic aggregation 
on the melt rheology of an ethylene-methacrylic acid copolymer and its sodium salt. 
Journal of Polymer Science: Polymer Physics Edition 16, 143 (1978). 
165 Cox, W. P. & Merz, E. H. Correlation of dynamic and steady flow viscosities. Journal of 
Polymer Science 28, 619 (1958). 
166 Marrucci, G. Dynamics of entanglements: A nonlinear model consistent with the 
cox-merz rule. Journal of Non-Newtonian Fluid Mechanics 62, 279 (1996). 
167 Ianniruberto, G. & Marrucci, G. On compatibility of the cox-merz rule with the model of 
doi and edwards. Journal of Non-Newtonian Fluid Mechanics 65, 241 (1996). 
168 Mead, D. W., Larson, R. G. & Doi, M. A molecular theory for fast flows of entangled 
polymers. Macromolecules 31, 7895 (1998). 
169 Mead, D. W. Analytic derivation of the cox-merz rule using the mld "toy" model for 
polydisperse linear polymers. Rheologica Acta 50, 837 (2011). 
170 Larson, R. G. & Doi, M. Mesoscopic domain theory for textured liquid crystalline 
polymers. Journal of Rheology 35, 539 (1991). 
171 Marrucci, G. Rheology of liquid crystalline polymers. Pure Applied Chemistry 57, 1545 
(1985). 
172 Larson, R. G. & Mead, D. W. Development of orientation and texture during shearing of 
liquid-crystalline polymers. Liquid Crystals(UK) 12, 751 (1992). 
173 Harrison, P. & Navard, P. Investigation of the band texture occurring in 
hydroxypropylcellulose solutions using rheo-optical, rheological and small angle light 
scattering techniques. Rheologica Acta 38, 569 (1999). 
174 Dealy, J. & Plazek, D. Time-temperature superposition-A users guide. Rheology Bulletin 
78, 16 2009. 
175 Williams, M. L., Landel, R. F. & Ferry, J. D. The temperature dependence of relaxation 
mechanisms in amorphous polymers and other glass-forming liquids. Journal of the 
American Chemical Society 77, 3701 (1955). 
176 Ferry, J. D. Viscoelastic properties of polymers.1980. (John Wiley & Sons Inc). 
177 Van Gurp, M. & Palmen, J. Time-temperature superposition for polymeric blends. 
Rheology Bulletin 67, 5 (1998). 
178 Li, L. S., Marjanska, M., Park, G. H. J., Pines, A. & Alivisatos, A. P. Isotropic-liquid 
crystalline phase diagram of a cdse nanorod solution. The Journal of Chemical Physics 
120, 1149 (2004). 
 
192 
179 Odijk, T. Theory of lyotropic polymer liquid crystals. Macromolecules 19, 2313 (1986). 
180 Fakhri, N., Tsyboulski, D. A., Cognet, L., Weisman, R. B. & Pasquali, M. 
Diameter-dependent bending dynamics of single-walled carbon nanotubes in liquids. 
Proceedings of the National Academy of Sciences 106, 14219 (2009). 
181 Gittes, F., Mickey, B., Nettleton, J. & Howard, J. Flexural rigidity of microtubules and 
actin filaments measured from thermal fluctuations in shape. Journal of Cell Biology 120, 
923 (1993). 
182 Smith, D. E., Perkins, T. T. & Chu, S. Dynamical scaling of DNA diffusion coefficients. 
Macromolecules 29, 1372 (1996). 
183 Kudin, K. N., Scuseria, G. E. & Yakobson, B. I. C_ {2} f, bn, and c nanoshell elasticity 
from ab initio computations. Physical Review B 64, 235406 (2001). 
184 Yakobson, B. I. & Couchman, L. S. Persistence length and nanomechanics of random 
bundles of nanotubes. Journal of Nanoparticle Research 8, 105 (2006). 
185 Gittes, F. & MacKintosh, F. C. Dynamic shear modulus of a semiflexible polymer 
network. Physical Review E 58, 1241 (1998). 
186 Shankar, V., Pasquali, M. & Morse, D. C. Theory of linear viscoelasticity of semiflexible 
rods in dilute solution. Journal of Rheology 46, 1111 (2002). 
187 Hirsch, A. Functionalization of single-walled carbon nanotubes. Angewandte 
Chemie-International Edition 41, 1853 (2002). 
188 Krause, B., Petzold, G., Pegel, S. & Potschke, P. Correlation of carbon nanotube 
dispersability in aqueous surfactant solutions and polymers. Carbon 47, 602 (2009). 
189 Zhang, S. J., Li, Q. W., Kinloch, I. A. & Windle, A. H. Ordering in a droplet of an 
aqueous suspension of single-wall carbon nanotubes on a solid substrate. Langmuir 26, 
2107. 
190 Song, W. H. & Windle, A. H. Isotropic-nematic phase transition of dispersions of 
multiwall carbon nanotubes. Macromolecules 38, 6181 (2005). 
191 Schmidt, R. H., Kinloch, I. A., Burgess, A. N. & Windle, A. H. The effect of aggregation 
on the electrical conductivity of spin-coated polymer/carbon nanotube composite films. 
Langmuir 23, 5707 (2007). 
192 Zamora-Ledezma, C. et al. Anisotropic thin films of single-wall carbon nanotubes from 
aligned lyotropic nematic suspensions. Nano Letters 8, 4103 (2008). 
193 Deegan, R. D. et al. Capillary flow as the cause of ring stains from dried liquid drops. 
Nature 389, 827 (1997). 
 
193 
194 Cataldo, F. DNA degradation with ozone. International Journal of Biological 
Macromolecules 38, 248 (2006). 
195 Gupta, V. B. & Kothari, V. K. Manufactured Fibre Echnology. 1997 (Springer). 
196 Treacy, M. M. J., Ebbesen, T. W. & Gibson, J. M. Exceptionally high young's modulus 
observed for individual carbon nanotubes. Nature 381, 678 (1996). 
197 Pasquali, M. et al. Neat carbon nanotube articles processed from super acid solutions and 
methods for production thereof. 2009. (WO Patent WO/2009/058,855). 
198 Barisci, J. N. et al. Properties of carbon nanotube fibers spun from DNA-stabilized 
dispersions. Advanced Functional Materials 14, 133 (2004). 
199 Endres, R. G., Cox, D. L. & Singh, R. R. P. Colloquium: The quest for high-conductance 
DNA. Reviews of Modern Physics 76, 195 (2004). 
200 Xu, Y. Q., Peng, H. Q., Hauge, R. H. & Smalley, R. E. Controlled multistep purification 
of single-walled carbon nanotubes. Nano Letters 5, 163 (2005). 
201 Lucas, A. et al. Kinetics of nanotube and microfiber scission under sonication. The 
Journal of Physical Chemistry C 113, 20599 (2009). 
202 Cheng, Q., Debnath, S., Gregan, E. & Byrne, H. J. Ultrasound-assisted swnts dispersion: 
Effects of sonication parameters and solvent properties. The Journal of Physical 
Chemistry C 114, 8821. 
203 Hilding, J., Grulke, E. A., Zhang, Z. G. & Lockwood, F. J. Dispersion of carbon 
nanotubes in liquids. Journal of Dispersion Science Technology 24, 2003, 1. 
204 Sato, J. & Breedveld, V. Evaporation blocker for cone-plate rheometry of volatile 
samples. Applied Rheology 15, 390 (2005). 
205 Gupta, R. K. Polymer and composite rheology. 2000. (CRC). 
206 Macosko, C. W. & Larson, R. G. Rheology: Principles, measurements, and applications. 
1994. (VCH New York). 
207 Kumar, A. & Gupta, R. K. Fundamentals of polymer engineering. 2003 (CRC). 
208 Sancar, A. Structure and function of DNA photolyase. Biochemistry 33, 2 (1994). 
209 Singer, V. L., Jones, L. J., Yue, S. T. & Haugland, R. P. Characterization of picogreen 
reagent and development of a fluorescence-based solution assay for double-stranded 
DNA quantitation. Analytical Biochemistry 249, 228 (1997). 
210 Kornyshev, A. A. & Leikin, S. Electrostatic interaction between long, rigid helical 
macromolecules at all interaxial angles. Physical Review E 62, 2576 (2000). 
 
194 
211 Marrucci, G. & Maffettone, P. L. Nematic phase of rodlike polymers. Ii. Polydomain 
predictions in the tumbling regime. Journal of Rheology 34, 1231 (1990). 
212 Onogi, S. & Asada, T. Rheology and rheo-optics of polymer liquid crystals. Rheology 1, 
127 (1980). 
213 Yoshimura, A. S., Prud'Homme, R. K., Princen, H. M. & Kiss, A. D. A comparison of 
techniques for measuring yield stresses. Journal of Rheology 31, 699 (1987). 
214 Hough, L. A., Islam, M. F., Janmey, P. A. & Yodh, A. G. Viscoelasticity of single wall 
carbon nanotube suspensions. Physical Review Letters 93, 168102 (2004). 
215 Huang, Y. Y., Ahir, S. V. & Terentjev, E. M. Dispersion rheology of carbon nanotubes in 
a polymer matrix. Physical Review B 73, 125422 (2006). 
216 Sahimi, M. & Arbabi, S. Mechanics of disordered solids. Ii. Percolation on elastic 
networks with bond-bending forces. Physical Review B 47, 703 (1993). 
217 Vollrath, F. & Knight, D. P. Liquid crystalline spinning of spider silk. Nature 410, 541 
(2001). 
218 Lee, S. W., Wood, B. M. & Belcher, A. M. Chiral smectic c structures of virus-based 
films. Langmuir 19, 1592 (2003). 
219 Roy, K. B., Antony, T., Sakena, A. & Bohidar, H. B. Ethanol-induced condensation of 
calf thymus DNA studied by laser light scattering. Journal of Physical Chemistry B 103, 
5117 (1999). 
220 Matzeu, M., Onori, G. & Santucci, A. Condensation of DNA by monohydric alcohols. 
Colloids and Surfaces B-Biointerfaces 13, 157 (1999). 
221 Knight, J. D. & Adami, R. C. Stabilization of DNA utilizing divalent cations and alcohol. 
International Journal of Pharmaceutics 264, 15 (2003). 
222 Bloomfield, V. A. DNA condensation. Current Opinion in Structural Biology 6, 334 
(1996). 
223 Hayashi, K., Kugimiya, M., Imoto, T., Funatsu, M. & Bigelow, C. C. Inhibitory 
interaction of cationic detergents with the active center of lysozyme. Ii. Ph dependence of 
the interaction. Biochemistry 7, 1467 (1968). 
224 Bordbar, A. K., Hosseinzadeh, R. & Norozi, M. H. Interaction of a homologous series of 
n-alkyl trimethyl ammonium bromides with eggwhite lysozyme. Journal of Thermal 
Analysis and Calorimetry 87, 453 (2007). 
225 Sun, Z. et al. Quantitative evaluation of surfactant-stabilized single-walled carbon 
nanotubes: Dispersion quality and its correlation with zeta potential. The Journal of 
Physical Chemistry C 112, 10692 (2008). 
 
195 
226 Montesi, A., Pena, A. A. & Pasquali, M. Vorticity alignment and negative normal 
stresses in sheared attractive emulsions. Physical Review Letters 92, 58303 (2004). 
227 Bercea, M. & Navard, P. Shear dynamics of aqueous suspensions of cellulose whiskers. 
Macromolecules 33, 6011 (2000). 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
196 
 
 
 
 
 
Appendix 1 
The conversion of SWNT concentration from a unit of milligram per liter (1 mg/L=1 ppm) 
measured by UV-Vis spectrophotometer into volume fraction ?  (vol %) is given as follows: 
10010010 6 ??? ?S W N T sS W N T sm ??                             (1) 
where, 
SWNTsm  = Mass of SWNT in dsDNA/SWNT dispersion (mg/L) 
SWNTs?  = Density of SWNT (1.45 g/cm
3) 
 
The conversion of SWNT concentration from mass fraction ?  (wt %) measured by 
Thermogravimetric analysis into volume fraction ?  (vol %) is given as follows: 
 
100100
2
2
?
??
?
OH
OH
D N A
D N A
S W N T s
S W N T s
S W N T s
S W N T s
mmm
m
???
??                 (2) 
where,  
SWNTsm  = Mass of SWNT in dsDNA/SWNT dispersion (g) 
DNAm  = Mass of dsDNA in dsDNA/SWNT dispersion (g) 
OHm2  = Mass of water in dsDNA/SWNT dispersion (g) 
SWNTs?  = Density of SWNT (1.45 g/cm
3) 
 
197 
DNA?  = Density of dsDNA (1.7 g/cm
3) 
OH2?  = Density of water (1.0 g/cm
3) 
 
Table 1. The conversion of SWNT concentration into volume fraction for evaporated samples 
starting from 0.75 wt % dsDNA-0.1 wt % SWNT dispersion without centrifugation. 
Samples: 
non-centrifuged 
dsDNA/SWNT 
dispersion with 
dsDNA:SWNT = 7.5:1 
SWNTs?  
(g/cm3) 
DNA?  
(g/cm3) 
OH2?  
(g/cm3) 
?   
(wt %) 
?  
(vol %) 
1 1.45 1.7 1.0 0.70 0.49 
2 1.45 1.7 1.0 1.10 0.77 
3 1.45 1.7 1.0 1.25 0.88 
4 1.45 1.7 1.0 1.44 1.01 
5 1.45 1.7 1.0 1.82 1.29 
6 1.45 1.7 1.0 2.30 1.64 
7 1.45 1.7 1.0 3.01 2.18 
8 1.45 1.7 1.0 3.88 2.85 
9 1.45 1.7 1.0 4.47 3.31 
10 1.45 1.7 1.0 5.21 3.91 
11 1.45 1.7 1.0 6.07 4.62 
 
 
198 
Table 2. The conversion of SWNT concentration into volume fraction for evaporated samples of 
supernatant of 0.75 wt % dsDNA-0.1 wt % SWNT (original dispersion) after centrifugation. 
Samples: supernatant of 
dsDNA/SWNT 
dispersion with 
dsDNA:SWNT = 5.1:1 
SWNTs?  
(g/cm3) 
DNA?  
(g/cm3) 
OH2?  
(g/cm3) 
?   
(wt %) 
?  
(vol %) 
1 1.45 1.7 1.0 0.38 0.26 
2 1.45 1.7 1.0 0.49 0.35 
3 1.45 1.7 1.0 0.85 0.60 
4 1.45 1.7 1.0 1.15 0.82 
5 1.45 1.7 1.0 1.19 0.85 
6 1.45 1.7 1.0 1.30 0.93 
7 1.45 1.7 1.0 1.46 1.04 
8 1.45 1.7 1.0 1.68 1.21 
9 1.45 1.7 1.0 1.76 1.27 
10 1.45 1.7 1.0 2.03 1.47 
11 1.45 1.7 1.0 2.62 1.93 
12 1.45 1.7 1.0 2.91 2.16 
13 1.45 1.7 1.0 2.99 2.22 
14 1.45 1.7 1.0 3.12 2.32 
15 1.45 1.7 1.0 3.45 2.60 
16 1.45 1.7 1.0 4.37 3.37 
17 1.45 1.7 1.0 4.97 3.90 
18 1.45 1.7 1.0 6.69 5.50 
 
 
199 
Table 3. The conversion of dsDNA concentration into volume fraction for evaporated samples of 
dsDNA only solution with starting concentration of 0.75 wt % after sonication and 
centrifugation. 
Samples: dsDNA solution DNA?  
(g/cm3) 
OH2?  
(g/cm3) 
?  (wt %) ?  (vol %) 
1 1.7 1.0 0.80 0.47 
2 1.7 1.0 5.34 3.21 
3 1.7 1.0 6.44 3.90 
4 1.7 1.0 7.46 4.53 
5 1.7 1.0 8.12 4.94 
6 1.7 1.0 8.78 5.36 
7 1.7 1.0 10.06 6.17 
8 1.7 1.0 11.49 7.09 
9 1.7 1.0 12.63 7.84 
10 1.7 1.0 13.08 8.13 
11 1.7 1.0 15.53 9.76 
12 1.7 1.0 19.63 12.56 
13 1.7 1.0 24.75 16.21 
14 1.7 1.0 28.88 19.28 
 
 
200 
 
 
 
 
 
Appendix 2 
Step Rate Test 
 
 
Figure 1. The start up of a shear flow indicated by viscosity versus time at 0.1 s-1 shear rate for 
bulk 6.4:1.0 volume ratio of dsDNA:SWNT dispersions. 
 
0
1000
2000
3000
4000
5000
0 . 0 1 0 . 1 1 10 100 1000
T i m e  ( s )
V
i
s
c
o
s
i
t
y
 
(
P
a
.
s
)
0 . 7 7  v o l .  %
0 . 8 8  v o l .  %
1 . 0 1  v o l .  %
1 . 2 9  v o l .  %
1 . 6 4  v o l .  %
2 . 1 8  v o l .  %
2 . 8 5  v o l .  %
3 . 3 1  v o l .  %
3 . 9 1  v o l .  %
4 . 6 2  v o l .  %
 
201 
 
Figure 2. The start up of a shear flow indicated by viscosity versus time at 0.1 s-1 shear rate for 
supernatant 4.4:1.0 volume ratio of dsDNA:SWNT dispersions. 
 
 
Figure 3. The start up of a shear flow indicated by viscosity versus time at 0.1 s-1 shear rate for 
dsDNA only solutions after sonication and centrifugation. 
 
0
1000
2000
3000
4000
0 . 0 1 0 . 1 1 10 100 1000
T i m e  ( s )
V
i
s
c
o
s
i
t
y
 
(
P
a
.
s
)
0 . 6 0  v o l .  %
0 . 8 5  v o l .  %
1 . 0 4  v o l .  %
1 . 2 1  v o l .  %
1 . 2 7  v o l .  %
1 . 9 3  v o l .  %
2 . 3 2  v o l .  %
2 . 6 0  v o l .  %
3 . 3 7  v o l .  %
0
100
200
0 . 1 1 10 100 1000
T i m e  ( s )
V
i
s
c
o
s
i
t
y
 
(
P
a
.
s
)
4 . 9  v o l .  %
5 . 4  v o l .  %
6 . 2  v o l .  %
7 . 1  v o l .  %
8 . 1  v o l .  %
9 . 8  v o l .  %
1 2 . 6  v o l .  %
 
202 
Steady-Shear Viscosity 
 
Figure 4. The steady-state shear viscosity with increasing SWNT concentration for bulk 6.4:1.0 
volume ratio of dsDNA:SWNT dispersions. 
 
 
Figure 5. The steady-state shear viscosity with increasing dsDNA concentration for dsDNA only 
solutions after sonication and centrifugation. 
1 . 0 E - 0 2
1 . 0 E - 0 1
1 . 0 E + 0 0
1 . 0 E + 0 1
1 . 0 E + 0 2
1 . 0 E + 0 3
1 . 0 E + 0 4
1 . 0 E + 0 5
0 . 0 1 0 . 1 1 10 100
S h e a r  R a t e  ( 1 / s )
V
i
s
c
o
s
i
t
y
 
(
P
a
.
s
)
0 . 4 9  v o l .  %
0 . 7 7  v o l .  %
0 . 8 8  v o l .  %
1 . 0 1  v o l .  %
1 . 2 9  v o l .  %
1 . 6 4  v o l .  %
2 . 1 8  v o l .  %
2 . 8 5  v o l .  %
3 . 3 1  v o l .  %
3 . 9 1  v o l .  %
4 . 6 2  v o l .  %
1 . 0 E - 0 1
1 . 0 E + 0 0
1 . 0 E + 0 1
1 . 0 E + 0 2
1 . 0 E + 0 3
0 . 0 1 0 . 1 1 10 100
S h e a r  R a t e  ( 1 / s )
V
i
s
c
o
s
i
t
y
 
(
P
a
.
s
)
3 . 2  v o l .  %
3 . 9  v o l .  %
4 . 5  v o l .  %
4 . 9  v o l .  %
5 . 4  v o l .  %
6 . 2  v o l .  %
7 . 1  v o l .  %
8 . 1  v o l .  %
9 . 8  v o l .  %
1 2 . 6  v o l .  %
 
203 
 
Figure 6. Relationship between viscosity and concentration for dsDNA only solutions at a shear 
rate of 0.1 s-1 and temperature of 10 oC. 
 
 
 
 
 
 
 
 
 
 
 
 
0
40
80
120
160
200
0 5 10 15
V
is
c
osi
ty
 (
P
a
.s
)
dsDNA  (v ol % )
 
204 
Cox-Merz Rule Not Obeyed 
 
(a) 
 
(b) 
Figure 7. The steady shear viscosity and complex viscosity as a function of shear rate and 
angular frequency with increasing SWNT concentration for bulk 6.4:1.0 volume ratio of 
dsDNA:SWNT dispersions. 
1 . 0 E-0 1
1 . 0 E+ 0 0
1 . 0 E+ 0 1
1 . 0 E+ 0 2
1 . 0 E+ 0 3
1 . 0 E+ 0 4
0 . 1 1 10 100
Sh e a r R a t e ,  A n g u l a r F re q u e n cy  (1 / s )
?
,
 
?
*
 
(Pa
.
s)
? -   0 . 7 7  v o l .  %
? * -  0 . 7 7  v o l .  %
? -   0 . 8 8  v o l .  %
? * -  0 . 8 8  v o l .  %
? -   1 . 0 1  v o l .  %
? * -  1 . 0 1  v o l .  %
? -   1 . 2 9  v o l .  %
? * -  1 . 2 9  v o l .  %
? -   1 . 6 4  v o l .  %
? * -  1 . 6 4  v o l .  %
1 . 0 E + 0 0
1 . 0 E + 0 1
1 . 0 E + 0 2
1 . 0 E + 0 3
1 . 0 E + 0 4
1 . 0 E + 0 5
0 . 1 1 10 100
S h e a r  R a t e ,  A n g u l a r  Fr e q u e n c y  ( 1 / s )
?
,
 
?
*
 
(
P
a
.
s
)
? -   2 . 1 8  v o l .  %
? * -  2 . 1 8  v o l .  %
? -   2 . 8 5  v o .  l %
? * -  2 . 8 5  v o l .  %
? -   3 . 3 1  v o l .  %
? * -  3 . 3 1  v o l .  %
? -   3 . 9 1  v o l .  %
? * -  3 . 9 1  v o l .  %
? -   4 . 6 2  v o l .  %
? * -  4 . 6 2  v o l .  %
 
205 
 
(a) 
 
(b) 
Figure 8. The steady shear viscosity and complex viscosity as a function of shear rate and 
angular frequency with increasing SWNT concentration for supernatant 4.4:1.0 volume ratio of 
dsDNA:SWNT dispersions. 
1 . 0 E - 0 1
1 . 0 E + 0 0
1 . 0 E + 0 1
1 . 0 E + 0 2
1 . 0 E + 0 3
1 . 0 E + 0 4
0 . 1 1 10 100
S h e a r  R a t e ,  A n g u l a r  Fr e q u e n c y  ( 1 / s )
?
,
 
?
*
 
(
P
a
.
s
)
? -   0 . 3 5  v o l .  %
? * -  0 . 3 5  v o l .  %
? -   0 . 6 0  v o l .  %
? * -  0 . 6 0  v o l .  %
? -   0 . 8 5  v o l .  %
? * -  0 . 8 5  v o l .  %
? -   1 . 0 4  v o l .  %
? * -  1 . 0 4  v o l .  %
? -   1 . 2 1  v o l .  %
? * -  1 . 2 1  v o l .  %
1 . 0 E+ 0 0
1 . 0 E+ 0 1
1 . 0 E+ 0 2
1 . 0 E+ 0 3
1 . 0 E+ 0 4
1 . 0 E+ 0 5
0 . 1 1 10 100
Sh e a r R a t e ,  A n g u l a r F re q u e n cy  (1 / s )
?
,
 
?
*
 
(Pa
.
s)
? -  1 . 2 7  v o l .  %
? * - 1 . 2 7  v o l .  %
? -  1 . 9 3  v o l .  %
? * - 1 . 9 3  v o l .  %
? -  2 . 3 2  v o l .  %
? * - 2 . 3 2  v o l .  %
? -  2 . 6 0  v o l .  %
? * - 2 . 6 0  v o l .  %
? -  3 . 3 7  v o l .  %
? * - 3 . 3 7  v o l .  %
 
206 
 
(a) 
 
(b) 
Figure 9. The steady shear viscosity and complex viscosity as a function of shear rate and 
angular frequency with increasing DNA concentration for dsDNA only solutions after sonication 
and centrifugation. 
 
1 . 0 E - 0 1
1 . 0 E + 0 0
1 . 0 E + 0 1
1 . 0 E + 0 2
1 . 0 E + 0 3
0 . 1 1 10 100
S h e a r  R a t e ,  A n g u l a r  Fr e q u e n c y  ( 1 / s )
?
,
 
?
*
 
(
P
a
.
s
)
? -   4 . 5  v o l .  %
? * -  4 . 5  v o l .  %
? -   4 . 9  v o l .  %
? * -  4 . 9  v o l .  %
? -   5 . 4  v o l .  %
? * -  5 . 4  v o l .  %
? -   6 . 2  v o l .  %
? * -  6 . 2  v o l .  %
1 . 0 E - 0 1
1 . 0 E + 0 0
1 . 0 E + 0 1
1 . 0 E + 0 2
1 . 0 E + 0 3
1 . 0 E + 0 4
0 . 1 1 10 100
S h e a r  R a t e ,  A n g u l a r  Fr e q u e n c y  ( 1 / s )
?
,
 
?
*
 
(
P
a
.
s
)
? -   7 . 1  v o l .  %
? * -  7 . 1  v o l .  %
? -   8 . 1  v o l .  %
? * -  8 . 1  v o l .  %
? -   9 . 8  v o l .  %
? * -  9 . 8  v o l .  %
? -   1 2 . 6  v o l .  %
? * -  1 2 . 6  v o l .  %
 
207 
Amplitude Sweep Test 
 
Figure 10. The linear viscoelastic (LVE) region of bulk 6.4:1.0 volume ratio of dsDNA:SWNT 
dispersions at different SWNT concentrations indicated by the linear function of storage modulus 
with strain at a fixed angular frequency of 10 s-1. 
 
 
Figure 11. The linear viscoelastic (LVE) region of supernatant 4.4:1.0 volume ratio of 
dsDNA:SWNT dispersions at different SWNT concentrations indicated by the linear function of 
storage modulus with strain at a fixed angular frequency of 10 s-1. 
1 . 0 E+ 0 0
1 . 0 E+ 0 1
1 . 0 E+ 0 2
1 . 0 E+ 0 3
1 . 0 E+ 0 4
0 . 0 1 0 . 1 1 10 100
S t r a i n  ( % )
S
t
o
r
a
g
e
 
M
o
d
u
l
u
s
,
 
G
'
 
(
P
a
)
0 . 7 7  v o l .  %
0 . 8 8  v o l .  %
1 . 0 1  v o l .  %
1 . 2 9  v o l .  %
1 . 6 4  v o l .  %
2 . 1 8  v o l .  %
2 . 8 5  v o l .  %
3 . 3 1  v o l .  %
3 . 9 1  v o l .  %
4 . 6 2  v o l .  %
1 . 0 E - 0 2
1 . 0 E - 0 1
1 . 0 E + 0 0
1 . 0 E + 0 1
1 . 0 E + 0 2
1 . 0 E + 0 3
1 . 0 E + 0 4
0 . 0 1 0 . 1 1 10 100
S t r a i n  ( % )
S
t
o
r
a
g
e
 
M
o
d
u
l
u
s
,
 
G
'
 
(
P
a
)
0 . 2 6  v o l .  %
0 . 3 5  v o l .  %
0 . 6 0  v o l .  %
0 . 8 5  v o l .  %
1 . 0 4  v o l .  %
1 . 2 1  v o l .  %
1 . 2 7  v o l .  %
1 . 9 3  v o l .  %
2 . 3 2  v o l .  %
2 . 6 0  v o l .  %
3 . 3 7  v o l .  %
 
208 
 
Figure 12. The linear viscoelastic (LVE) region of dsDNA only solutions after sonication and 
centrifugation at different DNA concentrations indicated by the linear function of storage 
modulus with strain at a fixed angular frequency of 10 s-1.  
 
 
 
 
 
 
 
 
 
 
 
 
1 . 0 E + 0 0
1 . 0 E + 0 1
1 . 0 E + 0 2
1 . 0 E + 0 3
0 . 0 1 0 . 1 1 10 100
S t r a i n  ( % )
S
t
o
r
a
g
e
 
M
o
d
u
l
u
s
,
 
G
'
 
(
P
a
)
4 . 5  v o l .  %
4 . 9  v o l .  %
5 . 4  v o l .  %
6 . 2  v o l .  %
7 . 1  v o l .  %
8 . 1  v o l .  %
9 . 8  v o l .  %
1 2 . 6  v o l .  %
 
209 
Frequency Sweep Test 
 
Figure 13. The storage modulus response as a function of angular frequency and increasing 
SWNT concentration for bulk 6.4:1.0 volume ratio of dsDNA:SWNT dispersions.  
 
 
Figure 14. The loss factor tan ? as a function of angular frequency and increasing SWNT 
concentration for bulk 6.4:1.0 volume ratio of dsDNA:SWNT dispersions. 
1 . 0 E - 0 1
1 . 0 E + 0 0
1 . 0 E + 0 1
1 . 0 E + 0 2
1 . 0 E + 0 3
1 . 0 E + 0 4
0 . 1 1 10 100
A n g u l a r  F r e q u e n c y  ( 1 / s )
S
t
o
r
a
g
e
 
M
o
d
u
l
u
s
,
 
G
'
 
(
P
a
)
0 . 7 7  v o l .  %
0 . 8 8  v o l .  %
1 . 0 1  v o l .  %
1 . 2 9  v o l .  %
1 . 6 4  v o l .  %
2 . 1 8  v o l .  %
2 . 8 5  v o l .  %
3 . 3 1  v o l .  %
3 . 9 1  v o l .  %
4 . 6 2  v o l .  %0
1
2
3
0 . 1 1 10 100
A n g u l a r  Fr e q u e n c y  ( 1 / s )
Ta
n
 
?
0 . 8 8  v o l .  %
1 . 0 1  v o l .  %
1 . 2 9  v o l .  %
1 . 6 4  v o l .  %
2 . 1 8  v o l .  %
2 . 8 5  v o l .  %
3 . 3 1  v o l .  %
3 . 9 1  v o l .  %
4 . 6 2  v o l .  %
 
210 
 
Figure 15. Cole-Cole plots show the storage modulus as a function of loss modulus and 
increasing SWNT concentration for bulk 6.4:1.0 volume ratio of dsDNA:SWNT dispersions. 
 
 
Figure 16. The storage modulus response as a function of angular frequency and increasing DNA 
concentration for dsDNA only solutions after sonication and centrifugation.  
1 . 0 E - 0 1
1 . 0 E + 0 0
1 . 0 E + 0 1
1 . 0 E + 0 2
1 . 0 E + 0 3
1 . 0 E + 0 4
0 1 10 100 1000 10000
L o s s  M o d u l u s ,  G ' '  ( P a )
S
t
o
r
a
g
e
 
M
o
d
u
l
u
s
,
 
G
'
 
(
P
a
)
0 . 8 8  v o l .  %
1 . 0 1  v o l .  %
1 . 2 9  v o l .  %
1 . 6 4  v o l .  %
2 . 1 8  v o l .  %
2 . 8 5  v o l .  %
3 . 3 1  v o l .  %
3 . 9 1  v o l .  %
4 . 6 2  v o l .  %
1 . 0 E + 0 0
1 . 0 E + 0 1
1 . 0 E + 0 2
1 . 0 E + 0 3
0 . 1 1 10 100
A n g u l a r  Fr e q u e n c y  ( 1 / s )
S
t
o
r
a
g
e
 
M
o
d
u
l
u
s
,
 
G
'
 
(
P
a
)
4 . 5  v o l .  %
4 . 9  v o l .  %
5 . 4  v o l .  %
6 . 2  v o l .  %
7 . 1  v o l .  %
8 . 1  v o l .  %
9 . 8  v o l .  %
1 2 . 6  v o l .  %
 
211 
 
Figure 17. The loss factor tan ? as a function of angular frequency and increasing DNA 
concentration for dsDNA only solutions after sonication and centrifugation. 
 
 
Figure 18. Cole-Cole plots show the storage modulus as a function of loss modulus and 
increasing DNA concentration for dsDNA only solutions after sonication and centrifugation. 
 
0
1
2
3
0 . 1 1 10 100
A n g u l a r  Fr e q u e n c y  ( 1 / s )
Ta
n
 
?
4 . 5  v o l .  % 4 . 9  v o l .  %
5 . 4  v o l .  % 6 . 2  v o l .  %
7 . 1  v o l .  % 8 . 1  v o l .  %
9 . 8  v o l .  % 1 2 . 6  v o l .  %
1 . 0 E + 0 0
1 . 0 E + 0 1
1 . 0 E + 0 2
1 . 0 E + 0 3
0 1 10 100 1000
L o s s  M o d u l u s ,  G "  ( P a )
S
t
o
r
a
g
e
 
M
o
d
u
l
u
s
,
 
G
'
 
(
P
a
) 4 . 5  v o l .  %
4 . 9  v o l .  %
5 . 4  v o l .  %
6 . 2  v o l .  %
7 . 1  v o l .  %
8 . 1  v o l .  %
9 . 8  v o l .  %
1 2 . 6  v o l .  %