Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Liquid crystalline tactoids in microscopic ordered-disordered interfaces : emergence of self-assembly… Wang, Pei-xi 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2018_november_wang_peixi.pdf [ 111.57MB ]
Metadata
JSON: 24-1.0372765.json
JSON-LD: 24-1.0372765-ld.json
RDF/XML (Pretty): 24-1.0372765-rdf.xml
RDF/JSON: 24-1.0372765-rdf.json
Turtle: 24-1.0372765-turtle.txt
N-Triples: 24-1.0372765-rdf-ntriples.txt
Original Record: 24-1.0372765-source.json
Full Text
24-1.0372765-fulltext.txt
Citation
24-1.0372765.ris

Full Text

LIQUID CRYSTALLINE TACTOIDS IN MICROSCOPIC ORDERED-DISORDEREDINTERFACES: EMERGENCE OF SELF-ASSEMBLY AND TOPOLOGICAL DEFECTSbyPEI-XI WANGB.Sc., Jilin University, 2014A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES(Chemistry)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)October 2018© PEI-XI WANG, 2018iiThe following individuals certify that they have read, and recommend to the Faculty of Graduateand Postdoctoral Studies for acceptance, the dissertation entitled:Liquid Crystalline Tactoids in Microscopic Ordered-Disordered Interfaces: Emergence ofSelf-Assembly and Topological Defectssubmitted by Pei-Xi Wang in partial fulfillment of the requirements forthe degree of Doctor of Philosophyin The Faculty of Graduate and Postdoctoral Studies (Chemistry)Examining Committee:Mark J. MacLachlanSupervisorMichael O. WolfSupervisory Committee MemberSupervisory Committee MemberScott RenneckarUniversity ExaminerDana GrecovUniversity ExaminerAdditional Supervisory Committee Members:Grenfell N. PateySupervisory Committee MemberJennifer A. LoveSupervisory Committee MemberiiiAbstractLiquid crystalline tactoids are discrete anisotropic microdroplets coexisting with continuousdisordered phases. In this thesis, an in-situ photopolymerization method was designed to rapidlycapture and solidify liquid crystalline tactoids in a crosslinked polymer matrix, which facilitatedthe direct observation of these fluid ordered microdroplets by scanning electron microscopy withthe resolution of individual liquid crystal mesogens. Different stages of the evolution of tactoidswere captured and examined, where the emergence of small-sized tactoids in initially disorderedphases, the coalescence of multiple tactoids, the generation of topological defects in coalescence,and the sedimentation of tactoids were directly observed by electron microscopy.The in-situ photopolymerization method was then extended to inverse emulsions, where thestructure and evolution of chiral nematic liquid crystalline tactoids in geometrical confinement ofmicrospheres were investigated by both optical and electron microscopy. This study revealed themicrostructures of topological defects of frustrated chiral nematic order in spherical confinement.Moreover, polymer and mesoporous silica microspheres with helical structures were obtained.The behavior of tactoids in the presence of colloidal doping nanoparticles was examined byelectron microscopy at the resolution of individual particles, which showed that liquid crystallinetactoids have size-selective exclusion effects on foreign nanoparticles. This principle was appliedto the separation of polymer nanospheres, gold nanoparticles, and paramagnetic nanoparticles bysize. These results suggest an approach to size-selectively separate nanoparticles using lyotropicliquid crystals, where nanoparticles smaller than a threshold size will be selectively collected intothe liquid crystalline tactoids and thus transferred from the disordered phase to the ordered phaseduring phase separation.ivThe phase separation of liquid crystals in the presence of paramagnetic doping nanoparticlesand gradient magnetic fields was studied. In this case, the disordered phases have higher volumemagnetic susceptibility than liquid crystalline tactoids due to the exclusion effects of tactoids onparamagnetic nanoparticles. Thus, the movement and orientation of tactoids could be controlledby gradient magnetic fields as weak as several hundred Gauss/cm. This approach enables controlof the phase separation rate and configuration, as well as the orientation of director fields in bothdiscrete tactoids and continuous macroscopic ordered phases.vLay SummaryWhen rod-shaped particles form a stable dispersion, due to their anisotropic geometry, theymay adopt a similar orientation above a critical concentration, which is termed a lyotropic liquidcrystal. When a lyotropic liquid crystal initially forms in a disordered system, microdroplets withwidths of several tens to hundreds of micrometers, in which the rod-shaped particles are orderlyarranged, may appear, and these microdroplets are called "tactoids". It was difficult to study themicroscopic structures of tactoids as they are soft and fluid. To address this issue, a method wasdeveloped in this thesis, by which tactoids can be rapidly captured in a jelly-like polymer matrixand further solidified. This approach enabled observation of the internal structures of tactoids atthe resolution of individual particles, and provided insights into the emergence of lyotropic liquidcrystals.viPrefaceAll the work presented in this thesis was carried out under the supervision of Prof. Dr. MarkJ. MacLachlan. All the cellulose nanocrystals were kindly provided by Dr. Wadood Y. Hamad atFPInnovations. I am the principal author of this thesis and I performed all of the experiments.Chapter 2: A version of this chapter, "Structure and transformation of tactoids in cellulosenanocrystal suspensions",53 has been published: Wang, P.-X.; Hamad, W. Y.; MacLachlan, M. J.Nat. Commun. 2016, 7, 11515. I performed all of the experiments and wrote the first draft of themanuscript, and contributed to the final version of the manuscript.Chapter 3: A version of this chapter has been published: "Polymer and Mesoporous SilicaMicrospheres with Chiral Nematic Order from Cellulose Nanocrystals",54 Wang, P.-X.; Hamad,W. Y.; MacLachlan, M. J. Angew. Chem. Int. Ed. 2016, 55, 12460-12464. I conducted all of theexperiments and wrote the first draft of the manuscript, and contributed to the final version of themanuscript.Chapter 4: A version of this chapter has been published: "Size-Selective Exclusion Effectsof Liquid Crystalline Tactoids on Nanoparticles: A Separation Method",55 Wang, P.-X.; Hamad,W. Y.; MacLachlan, M. J. Angew. Chem. Int. Ed. 2018, 57, 3360-3365. In this work, I performedall of the experiments, wrote the first draft of the manuscript, and contributed to the final versionof the manuscript.Chapter 5: A draft of this chapter has been submitted for publication: "Phase Separation inCompetitive Acceleration Fields Beyond Gravity: Movement and Orientation Control of LiquidCrystalline Tactoids" (2018), by Pei-Xi Wang, Wadood Y. Hamad, and Mark J. MacLachlan. Inviithis research, I designed and performed all of the experiments and then wrote the first draft of themanuscript, and contributed to the final version of the manuscript.In addition, a review article of the research work presented in Chapter 2 and Chapter 3 hasalready been published: "Liquid crystalline tactoids: ordered structure, defective coalescence andevolution in confined geometries", Wang, P.-X.; MacLachlan, M. J. Phil. Trans. R. Soc. A 2018,376, 20170042. I wrote the first draft of the manuscript of this review article, and contributed tothe final version of the manuscript.viiiTable of ContentsAbstract...........................................................................................................................................iiiLay Summary...................................................................................................................................vPreface............................................................................................................................................ viTable of Contents..........................................................................................................................viiiList of Tables................................................................................................................................ xiiiList of Figures...............................................................................................................................xivList of Symbols...........................................................................................................................xxxiList of Abbreviations................................................................................................................. xxxiiGlossary.................................................................................................................................... xxxiiiAcknowledgements...................................................................................................................xxxivDedication.................................................................................................................................xxxviChapter 1: Introduction.................................................................................................................... 11.1 Liquid Crystals...................................................................................................................11.1.1 History and Definition of Liquid Crystals.............................................................. 11.1.2 Classification of Liquid Crystalline Phases............................................................ 11.1.3 Applications of Liquid Crystals..............................................................................21.2 Liquid Crystalline Tactoids................................................................................................61.2.1 Tactoids in Lyotropic Liquid Crystals.................................................................... 61.2.2 Emergence of Tactoids in Disordered Phases.........................................................61.2.3 Geometries and Director Fields of Tactoids........................................................... 81.3 Cellulose Nanocrystals.....................................................................................................10ix1.3.1 Chiral Nematic Phases Formed by Cellulose Nanocrystals..................................101.3.2 Liquid Crystalline Tactoids of Cellulose Nanocrystals........................................ 111.4 Aims and Scope............................................................................................................... 14Chapter 2: Structure and Transformation of Liquid Crystalline Tactoids..................................... 162.1 Introduction......................................................................................................................162.2 Solidification of Liquid Crystalline Tactoids by In-Situ Photopolymerization...............162.3 Electron Microscopy Observations of Liquid Crystalline Tactoids................................ 182.3.1 Microstructures of Chiral Nematic Liquid Crystalline Tactoids.......................... 182.3.2 Coalescence of Tactoids and Formation of Topological Defects......................... 302.3.3 Gravitational Sedimentation of Liquid Crystalline Tactoids................................ 372.4 Conclusions......................................................................................................................422.5 Experimental Methods..................................................................................................... 462.5.1 Capture of Tactoids by In-Situ Photopolymerization........................................... 462.5.2 Concentration and Density of Ordered and Disordered Phases............................462.5.3 Materials............................................................................................................... 472.5.4 Characterization.................................................................................................... 47Chapter 3: Liquid Crystalline Tactoids in Geometrical Confinement........................................... 493.1 Introduction......................................................................................................................493.2 Microspheres with Chiral Nematic Order from Liquid Crystalline Tactoids.................. 503.2.1 Evolution of Liquid Crystalline Tactoids in Spherical Confinement................... 503.2.2 Solidification of Microdroplets by In-Situ Photopolymerization......................... 523.2.3 Structure of Liquid Crystalline Tactoids in Spherical Confinement.................... 543.2.4 Spherical Chiral Nematic Tactoids under Electron Microscopy.......................... 62x3.2.5 Tactoids in Geometrical Confinement with Different Dimensions...................... 673.3 Mesoporous Silica Microspheres with Chiral Nematic Order.........................................733.4 Conclusions......................................................................................................................783.5 Experimental Methods..................................................................................................... 793.5.1 Fabrication of Polymer Microspheres with Chiral Nematic Order.......................793.5.2 Preparation of Chiral Nematic Mesoporous Silica Microspheres.........................803.5.3 Materials............................................................................................................... 813.5.4 Characterization.................................................................................................... 81Chapter 4: Size-Selective Exclusion Effects of Liquid Crystalline Tactoids on Nanoparticles:Separation by Microscopic Ordered-Disordered Interfaces.......................................................... 834.1 Introduction......................................................................................................................834.2 Size-Selective Exclusion of Nanoparticles by Liquid Crystalline Tactoids.................... 844.2.1 Liquid Crystalline Tactoids in Mixtures of Nanorods and Nanospheres..............844.2.2 Exclusion of Large-Sized Nanoparticles by Liquid Crystalline Tactoids............ 894.2.3 Coalescence of Tactoids Traps Nanoparticles in Topological Defects................ 964.2.4 Existence of Small-Sized Nanoparticles in Liquid Crystalline Tactoids............1004.2.5 Size-Selective Particle Permeability of Ordered-Disordered Interfaces.............1044.3 Size-Selective Separation of Nanoparticles with Lyotropic Liquid Crystals................ 1084.3.1 Separation of Plasmonic Gold Nanoparticles by Size........................................ 1084.3.2 Multiple-Cycle Size-Selective Separation of Magnetic Nanoparticles.............. 1144.4 Conclusions....................................................................................................................1194.5 Experimental Methods................................................................................................... 1204.5.1 Photopolymerization of Liquid Crystals with Doping Nanoparticles.................120xi4.5.2 Diffusion of Nanoparticles through Ordered-Disordered Interfaces.................. 1204.5.3 Size-Selective Separation of Gold Nanoparticles by Liquid Crystals................ 1214.5.4 Recycling of Liquid Crystalline Phases for Multicycle Separations.................. 1214.5.5 Materials............................................................................................................. 1224.5.6 Characterization.................................................................................................. 123Chapter 5: Movement and Orientation Control of Liquid Crystalline Tactoids in CompetitiveAcceleration Fields: Phase Separations Beyond Gravity.............................................................1245.1 Introduction....................................................................................................................1245.2 Magnetic Buoyancy Forces on Tactoids with Lower Magnetic Susceptibility............. 1275.3 Unidirectional Phase Separations: Acceleration by Gradient Magnetic Fields............. 1305.3.1 Lyotropic Liquid Crystals Doped with Paramagnetic Nanoparticles................. 1305.3.2 Orientation of Liquid Crystalline Tactoids in Gradient Magnetic Fields........... 1365.3.3 Vertically Aligned Chiral Nematic Layers: Reversed Phase Separation............1445.4 Competitions between Gravitational and Magnetic Acceleration Forces......................1535.5 Conclusions....................................................................................................................1585.6 Experimental Methods................................................................................................... 1595.6.1 Synthesis of Superparamagnetic Magnetite (Fe3O4) Nanoparticles................... 1595.6.2 Solidification of CNC-Fe3O4 Mixtures by In-Situ Photopolymerization........... 1595.6.3 Materials............................................................................................................. 1605.6.4 Characterization.................................................................................................. 160Chapter 6: Conclusions and Future Perspectives.........................................................................1616.1 Conclusions....................................................................................................................1616.2 Future Perspectives........................................................................................................ 164xii6.2.1 Liquid Crystalline Tactoids in Other Soft Matter Systems.................................1646.2.2 Size-Selective Separation of Molecules by Liquid Crystalline Tactoids............164Bibliography................................................................................................................................ 166xiiiList of TablesTable 3.1. Preparation parameters for CNC-PAAm chiral nematic microspheres....................... 52Table 5.1. Phase separation times of CNC-Fe3O4 binary mixtures.............................................154Table 5.2. Distribution ratio of Fe3O4 NPs in isotropic and liquid crystalline phases................ 156xivList of FiguresFigure 1.1. Phase transitions of thermotropic liquid crystals as the temperature increases............3Figure 1.2. Phase transitions of lyotropic liquid crystals as the concentration decreases.............. 3Figure 1.3. Arrangement of rod-shaped mesogens in nematic liquid crystals. The mesogens havea long-range collective orientational order...................................................................................... 4Figure 1.4. Arrangement of rod-shaped mesogens in smectic liquid crystals, where in addition tothe long-range collective orientational order, the mesogens also have long-range positional orderin one dimension..............................................................................................................................4Figure 1.5. Arrangement of mesogens in chiral nematic (cholesteric) liquid crystals, which adopta helical superstructure.................................................................................................................... 5Figure 1.6. Nematic liquid crystalline tactoids formed in a dispersion of tobacco mosaic viruses,which are spindle-shaped birefringent microdroplets.12.................................................................. 9Figure 1.7. Chiral nematic liquid crystalline tactoids formed in a polypeptide solution, which arespherical microdroplets with periodically spaced birefringent layers.14..........................................9Figure 1.8. TEM micrographs of cellulose nanocrystals used in this research. Scale bars, (A) 200nm, (B) 100 nm, (C,D) 50 nm....................................................................................................... 12Figure 1.9. POM micrograph showing discrete chiral nematic liquid crystalline tactoids formedin an aqueous dispersion of cellulose nanocrystals. Scale bar, 50 μm.......................................... 13Figure 2.1. By adding the nonionic precursors of polyacrylamide into an aqueous suspension ofcellulose nanocrystals, this system could be captured in a cross-linked polymer matrix by in-situphotopolymerization, forming a CNC-PAAm composite hydrogel. The diameter of the Petri dishin this photograph is about 50 mm................................................................................................ 17xvFigure 2.2. Photograph showing a completely dried CNC-PAAm composite polymer matrix, inwhich the microstructures of tactoids have been solidified...........................................................17Figure 2.3. SEM micrographs of a newly emergent liquid crystalline tactoid formed by cellulosenanocrystals. Scale bars, (A) 2 μm, (B) 1 μm, (C) 300 nm, (D) 200 nm, (E,F) 100 nm............... 19Figure 2.4. Cross-sectional SEM micrographs showing the boundary of a newly emergent liquidcrystalline tactoid near its (A,B) left and (C,D) bottom regions. The boundary of this tactoid is amicroscopic ordered-disordered interface, which sharply separates the liquid crystalline domainfrom the isotropic phase surrounding it. Scale bars, (A,C) 200 nm, (B,D) 100 nm...................... 20Figure 2.5. High-resolution SEM micrographs showing the ordered arrangements of rod-shapedcellulose nanocrystal mesogens in a newly emergent liquid crystalline tactoid. The mesogens areunidirectionally aligned into a nematic phase since this tactoid is smaller than a half helical pitch,however, the director field is still slightly twisted in long ranges due to the chirality of this phase.Scale bars, (A-D) 100 nm.............................................................................................................. 21Figure 2.6. Cross-sectional SEM micrographs showing a liquid crystalline tactoid with only onechiral nematic layer. Scale bars, (A) 1 μm, (B,C) 300 nm, (D-F) 200 nm.................................... 22Figure 2.7. 3D model showing the left-handed chiral nematic order of CNC mesogens (depictedas rods in a golden brown color) in a newly formed liquid crystalline tactoid with only one chiralnematic band. The rod-shaped mesogens in the pseudo nematic layers rotate by 180 degrees fromone end to the other........................................................................................................................23Figure 2.8. Additional SEM micrographs showing liquid crystalline tactoids with (A,B) one, (C)zero, or (D-F) three chiral nematic layers. Scale bars, (A) 2 μm, (B) 400 nm, (C) 1 μm, (D) 3 μm,(E) 500 nm, (F) 200 nm................................................................................................................. 24xviFigure 2.9. Cross-section SEM micrographs showing a liquid crystalline tactoid with four chiralnematic layers. Scale bars, (A) 5 μm, (B,D) 2 μm, (C,E) 1 μm, (F) 300 nm................................ 25Figure 2.10. Cross-sectional SEM micrographs of a large-size liquid crystalline tactoid with ninechiral nematic layers. Scale bars, (A) 10 μm, (B,C) 2 μm, (D-F) 1 μm........................................ 26Figure 2.11. SEM micrographs showing several chiral nematic liquid crystalline tactoids sittingat a right-angled edge. Scale bars, (A) 200 μm, (B) 50 μm, (C,D) 10 μm, (E,F) 3 μm.................27Figure 2.12. SEM micrographs showing the (A-C) top and (D-F) front views of a tactoid sittingat a right-angled edge. Scale bars, (A) 10 μm, (B,D) 2 μm, (C,E) 1 μm, (F) 500 nm...................28Figure 2.13. SEM micrographs showing the (A-C) left and (D-F) right views of a chiral nematicliquid crystalline tactoid. Scale bars, (A,D) 500 nm, (B,E,F) 300 nm, (C) 200 nm...................... 29Figure 2.14. Three-dimensional model showing the contact between two discrete chiral nematicliquid crystalline tactoids in an aqueous dispersion of cellulose nanocrystals..............................32Figure 2.15. POM micrographs showing the coalescence of two chiral nematic liquid crystallinetactoids in a CNC dispersion......................................................................................................... 33Figure 2.16. SEM micrograph showing the initiation of coalescence between two chiral nematictactoids. Scale bar, 10 μm..............................................................................................................33Figure 2.17. SEM images showing the microstructures of the contact point between two tactoids.Scale bars, (A) 2 μm, (B) 1 μm, (C) 500 nm, (D) 300 nm.............................................................34Figure 2.18. SEM images showing the aggregation and contact between multiple chiral nematicliquid crystalline tactoids. Scale bars, (A) 20 μm, (B,C) 10 μm, (D-F) 5 μm............................... 35Figure 2.19. SEM images of liquid crystalline tactoids with topological defects of dislocated orfolded chiral nematic layers. Scale bars, (A) 5 μm, (B-F) 10 μm..................................................36xviiFigure 2.20. Cross-sectional SEM micrograph showing the gravitational sedimentation of chiralnematic liquid crystalline tactoids to the bottom of a CNC dispersion, where the tactoids coalescetogether to form a macroscopic continuous ordered phase. Scale bar, 20 μm.............................. 38Figure 2.21. SEM micrographs showing topological defects as an array of folded chiral nematiclayers. Scale bars, (A) 20 μm, (B,C) 2 μm, (D) 1 μm, (E,F) 500 nm............................................ 39Figure 2.22. SEM micrographs showing topological defects as an array of folded chiral nematiclayers. Scale bars, (A) 10 μm, (B) 3 μm, (C) 1 μm, (D) 500 nm...................................................40Figure 2.23. POM micrograph showing the fingerprint-like texture in a solid-state chiral nematicfilm formed by CNCs. Scale bar, 50 μm....................................................................................... 41Figure 2.24. 3D model showing the left-handed chiral nematic arrangement of rod-shaped CNCmesogens in a liquid crystalline tactoid.........................................................................................43Figure 2.25. 3D models showing the coalescence of liquid crystalline tactoids and the formationof topological defects as folded chiral nematic layers...................................................................44Figure 2.26. 3D model showing topological defects as an array of folded chiral nematic layers inthe continuous liquid crystalline phase at the bottom of a CNC dispersion..................................45Figure 3.1. POM micrographs showing the evolution of chiral nematic liquid crystalline tactoidsin spherical confinement. The evolution time and diameter of these microdroplets are as follows:(A) 0.5 h, 118 μm; (B) 1 h, 58 μm; (C) 3 h, 64 μm; (D) 6 h, 78 μm; (E) 9 h, 57 μm; and (F) 12 h,114 μm........................................................................................................................................... 51Figure 3.2. Sketch showing the in-situ photopolymerization of aqueous CNC microdroplets in aninverse emulsion. Using this approach, CNC-PAAm composite microspheres with chiral nematicstructures could be obtained.......................................................................................................... 53xviiiFigure 3.3. Optical micrographs of dried CNC-PAAm microbeads (A) observed under reflectedwhite light or (B) with transmitted light between crossed polarizers. Scale bars, 50 μm............. 56Figure 3.4. POM micrograph of a CNC-PAAm chiral nematic microgel that has swelled in water.Scale bar, 20 μm............................................................................................................................ 56Figure 3.5. Optical micrographs of CAMB-M microgels (after swelling in water) observed (A,C)between crossed polarizers or (B,D) under reflected white light. Scale bars, 50 μm....................57Figure 3.6. Orthogonal views of a chiral nematic microsphere obtained by confocal fluorescencemicroscopy. Here the imaging region is 178 μm in width and 130 μm in height. The sample wasstained by immersing the microspheres in an aqueous solution of rhodamine B (about 0.1 wt.%),then redispersed in pure water before imaging. This microsphere was flattened by gravity........ 58Figure 3.7. A three-dimensional reconstructed confocal fluorescence microscopy image showingthe chiral nematic layers in a microsphere as a series of concentric spherical shells. A topologicaldefect can be observed at the center of this microsphere, which was formed due to the frustrationof the chiral nematic liquid crystalline order in the spherical confinement...................................59Figure 3.8. POM images of CNC-PAAm microspheres showing chiral nematic tactoids capturedat different evolution stages. Diameter of microspheres, (A) 85 μm, (B) 136 μm, (C) 142 μm and(D) 141 μm.....................................................................................................................................60Figure 3.9. Three-dimensional models showing the evolution of chiral nematic liquid crystallinetactoids in spherical confinement.................................................................................................. 61Figure 3.10. A three-dimensional model showing the chiral nematic layers of a liquid crystallinetactoid in spherical confinement, which are organized into a series of concentric shells............. 61Figure 3.11. SEM micrographs showing the surface morphology of the solid-state CNC-PAAm(CAMB-M) composite microspheres. Scale bars, 20 μm..............................................................63xixFigure 3.12. SEM micrographs showing the onion-like structure of chiral nematic layers inside aCNC-PAAm microsphere, which was cracked in liquid nitrogen. As frost immediately formed onthe cold fracture surface, softened and smoothed it after melting (due to the hydrophilicity of thepolyacrylamide matrix), the arrangement of individual CNCs cannot be distinguished. Scale bars,(A) 10 μm, (B) 5 μm......................................................................................................................63Figure 3.13. SEM images showing a fracture surface of an epoxy resin containing CNC-PAAmmicrospheres. Scale bars, (A) 200 μm, (B) 50 μm........................................................................ 64Figure 3.14. SEM image showing the cross-section of a microsphere that has been split into twoparts by using an epoxy resin. Scale bar, 10 μm............................................................................64Figure 3.15. Cross-sectional SEM micrograph showing the spiral structure of the chiral nematicliquid crystalline tactoid inside a CNC-PAAm microsphere. This tactoid consists of two differentchiral nematic layers. The chiral nematic ordering of the liquid crystalline phase is frustrated dueto the spherical confinement, which results in the formation of a topological defect in the centerof this tactoid. Scale bar, 5 μm...................................................................................................... 65Figure 3.16. Cross-sectional SEM images showing the left-handed chiral nematic order of CNCsin a microsphere. Scale bars, (A) 2 μm, (B,C) 500 nm, (D) 200 nm.............................................66Figure 3.17. SEM micrographs showing the size distributions of completely dried CNC-PAAmmicrospheres. (A,B) CAMB-M, (C) CAMB-L, and (D) CAMB-S. These micrographs were takenat the same magnification. Scale bars, 200 μm..............................................................................69Figure 3.18. POM micrographs of CAMB-M microspheres. Typically, the chiral nematic liquidcrystalline tactoids inside a microsphere have been well integrated into a single liquid crystallinecore with a spherical geometry, which is located at the center of the microsphere and surroundedby an isotropic shell. Most of these integrated tactoids have highly-ordered onion-like structuresxxof multiple concentric spherical shells. These micrographs were taken at the same magnification.Scale bars, 50 μm...........................................................................................................................70Figure 3.19. POM micrographs showing CAMB-S microspheres in water, the diameters of thesesmall-size microspheres are about 2 to 3 times the helical pitch of the CNC chiral nematic phase.Typically, the liquid crystalline tactoids in a microsphere have also been well integrated, forminga series of concentric spherical chiral nematic shells near the boundary of the microsphere, whichleaves an isotropic core in the central region. Interestingly, in some cases there may be a discretetactoid left inside the isotropic core. All these micrographs were taken at the same magnification.Scale bars, 50 μm...........................................................................................................................71Figure 3.20. POM images showing CAMB-L microspheres. The chiral nematic liquid crystallinetactoids in these large-sized microspheres were not well integrated during the period of standing,which might be caused by the relatively long distances between them. These POM micrographswere taken at the same magnification. Scale bars, 50 μm............................................................. 72Figure 3.21. POM images of mesoporous silica microspheres. Scale bars, 10 μm...................... 75Figure 3.22. IR spectra of CNC-PAAm-SiO2 (blue) and SiO2 microspheres (red curve)............ 75Figure 3.23. SEM images showing the chiral nematic order of a mesoporous silica microsphere.Scale bars, (A) 10 μm, (B) 2 μm, (C,E) 500 nm, (D) 1 μm, (F) 200 nm.......................................76Figure 3.24. Nitrogen adsorption-desorption isotherm (77 K) of mesoporous silica microspheres.The sample mass was 117.3 mg in this measurement. The desorption data are presented as emptycircles.............................................................................................................................................77Figure 4.1. SEM images for large polystyrene nanoparticles (denoted as L-PSNPs, diameters of265-280 nm) used in this work. Scale bars, (A) 200 nm, (B) 100 nm...........................................85xxiFigure 4.2. Optical microscopy images showing chiral nematic liquid crystalline tactoids formedin the presence of L-PSNPs (265 to 280 nm in diameter), which were observed between crossedpolarizers (A) or under reflected white light (B). Scale bars, 50 μm............................................ 85Figure 4.3. SEM images of small-size polystyrene nanoparticles (denoted as S-PSNPs, diametersof 30-57 nm) used in this work. Scale bars, 50 nm....................................................................... 86Figure 4.4. Optical microscopy images showing chiral nematic liquid crystalline tactoids formedin a binary mixture of CNCs and S-PSNPs (30-57 nm in diameter), which was observed betweencrossed polarizers (A) or under reflected white light (B). Scale bars, 50 μm............................... 86Figure 4.5. 3D model showing the exclusion effect of a liquid crystalline tactoid (mesogens aredepicted as golden brown rods) on large-sized doping nanoparticles (purple spheres)................ 87Figure 4.6. 3D model showing that small-sized nanoparticles (depicted as red spheres) can entera liquid crystalline tactoid..............................................................................................................88Figure 4.7. SEM image of a tactoid formed in a CNC/L-PSNP mixture. Scale bar, 2 μm.......... 91Figure 4.8. SEM images showing the inside area of a newly formed liquid crystalline tactoid in aCNC/L-PSNP binary mixture. Scale bars, (A) 1 μm, (B) 500 nm................................................ 91Figure 4.9. SEM images of (A) the upper left, (B,C) the left, and (D-F) the upper boundaries of anewly emergent chiral nematic liquid crystalline tactoid in a CNC/L-PSNP mixture. Scale bars,(A,C,E) 500 nm, (B,D) 1 μm, (F) 200 nm..................................................................................... 92Figure 4.10. SEM images showing L-PSNPs in disordered phases. Scale bars, 200 nm.............93Figure 4.11. 3D model for the emergence of a small tactoid in a CNC/L-PSNP mixture............93Figure 4.12. SEM micrographs of a liquid crystalline tactoid with about 10 chiral nematic layersformed in a CNC/L-PSNP mixture (A), as well as the boundary (B-D) and chiral nematic layersof this tactoid (E,F). Scale bars, (A) 5 μm, (B,C) 1 μm, (D,E) 500 nm, (F) 200 nm.....................94xxiiFigure 4.13. 3D model showing the coalescence of two liquid crystalline tactoids in the presenceof large-sized foreign nanoparticles...............................................................................................95Figure 4.14. (A) A few large-sized doping nanoparticles (depicted in purple) could be trapped atthe contact point between two coalescing tactoids. (B) A lot of large doping nanoparticles can besealed within the broad isotropic region surrounded by several merging tactoids. (C) Topologicaldefects generated during the coalescence of tactoids would eventually remain in the macroscopicliquid crystalline phases with a large number of doping nanoparticles sealed inside................... 96Figure 4.15. (A) Cross-sectional SEM micrograph showing the initiation of coalescence betweentwo liquid crystalline tactoids formed in a CNC/L-PSNP binary mixture. The boundaries of thesetwo tactoids can exclude the large-sized doping nanoparticles (B), but several nanoparticles weretrapped at the contact point between them (C,D), which would be sealed in the liquid crystallinephase. Scale bars, (A) 5 μm, (B,C) 1 μm, (D) 200 nm.................................................................. 97Figure 4.16. SEM images showing the contact between two tactoids in a CNC/L-PSNP mixture(A), the exclusion of L-PSNPs by tactoid boundaries (B,C), and L-PSNPs trapped in the contactregion (D-F). Scale bars, (A) 3 μm, (B,E) 500 nm, (C,D) 1 μm, (F) 200 nm................................98Figure 4.17. Cross-sectional SEM images of a CNC/L-PSNP mixture showing large-size dopingnanoparticles trapped in the topological defects of the macroscopic liquid crystalline phase (A-Band C-F). Scale bars, (A,D) 1 μm, (B,E) 500 nm, (C) 2 μm, (F) 200 nm......................................99Figure 4.18. SEM images showing the existence of small-sized doping nanoparticles (diametersof 30-57 nm) as about 50 nm hemispherical cavities in tactoids formed in a CNC/S-PSNP binarymixture (A-C and D-F). Scale bars, (A) 3 μm, (B,C,F) 100 nm, (D) 2 μm, (E) 200 nm.............101xxiiiFigure 4.19. (A-E) SEM images showing hemispherical cavities created by S-PSNPs in a "baby"tactoid in a CNC/S-PSNP mixture. (F) Image showing the coexistence of S-PSNPs and CNCs inthe disordered phase. Scale bars, (A) 2 μm, (B) 200 nm, (C,D,F) 100 nm, (E) 50 nm............... 102Figure 4.20. 3D models illustrating (A) the size-selective permeability of an ordered-disorderedinterface to foreign nanoparticles, and (B) the exclusive force from this interface on an invadingnanoparticle with a width larger than the gaps in the liquid crystalline lattice........................... 103Figure 4.21. Photographs showing that the downward diffusion of L-PSNPs was stopped by theordered-disordered interface of a phase-separated CNC dispersion............................................104Figure 4.22. (A) Cross-sectional SEM micrograph showing the ordered-disordered interface of aphase-separated CNC dispersion, which is the upper boundary of the macroscopically continuouschiral nematic liquid crystalline phase. (B-D) Expanded views near the interface showing that thedownward diffusion of the large-sized polystyrene nanoparticles (diameters of 265-280 nm) wasdramatically stopped by the upper boundary of the liquid crystalline phase. Scale bars, (A) 2 μm,(B) 1 μm, (C,D) 500 nm.............................................................................................................. 105Figure 4.23. High-resolution cross-sectional SEM image showing the microscopic structures ofthe interface between the disordered phase (upper) and the chiral nematic liquid crystalline phase(lower) in a completely phase-separated cellulose nanocrystal dispersion, where the organizationof individual liquid crystal mesogens can be directly observed. This ordered-disordered interfacedramatically blocked the downward diffusion of the large-sized doping nanoparticles (L-PSNPs),which could be observed in the upper disordered phase as either hemispherical cavities or partlystretched nanospheres. Scale bar, 200 nm................................................................................... 106Figure 4.24. SEM images showing that when small-size polystyrene nanoparticles (diameters of30 to 57 nm) were added on the top of a phase-separated cellulose nanocrystal suspension, thesexxivnanoparticles were able to pass through the isotropic-anisotropic interface by diffusion, resultingin hemispherical cavities (about 50 nm in diameter) in the chiral nematic liquid crystalline phase.Scale bars, (A) 500 nm, (B-D) 100 nm........................................................................................107Figure 4.25. Photographs (with transmitted white light) for phase-separated CNC-AuNP binarymixtures....................................................................................................................................... 109Figure 4.26. UV-Vis absorption of a phase-separated CNC/S-AuNP binary mixture............... 110Figure 4.27. UV-Vis absorption of a phase-separated CNC/L-AuNP binary mixture............... 110Figure 4.28. UV-Vis absorption spectra of the isotropic (Iso) and liquid crystalline (LC) phasesof a phase-separated CNC/L&S-AuNP ternary mixture............................................................. 111Figure 4.29. (A) UV-Vis absorption spectra of a phase-separated CNC dispersion, which exhibitno absorption peaks between 400 nm and 800 nm. (B) Additional UV-Vis absorption spectra ofthe small and large sized gold nanoparticles used in this study.................................................. 111Figure 4.30. (A,B) TEM micrographs showing the coexistence of the large and small sized goldnanoparticles in the disordered phase of a completely phase-separated CNC/L&S-AuNP ternarymixture. (C,D) Inside the liquid crystalline phase, only the small-sized gold nanoparticles couldbe observed. Scale bars, (A,B) 100 nm, (C,D) 200 nm............................................................... 112Figure 4.31. Particle size distributions of AuNPs in the isotropic (Iso) and liquid crystalline (LC)phases of a phase-separated CNC/L&S-AuNP ternary mixture, measured by TEM.................. 113Figure 4.32. A depiction showing the size-selective collection of the small foreign nanoparticles(small nanoparticles in red, large particles in purple) by chiral nematic liquid crystalline tactoidsof CNCs. Due to the higher density, tactoids will gradually settle to the bottom part of thedispersion and coalesce into a macroscopic ordered phase with a significantly higher ratio ofsmall nanoparticles...................................................................................................................... 113xxvFigure 4.33. Photographs showing phase-separated dispersions of cellulose nanocrystals mixedwith 8.7 nm (left), 8.7 nm & 107 nm (middle), and 107 nm (right) superparamagnetic magnetitenanoparticles, taken with transmitted white light........................................................................ 115Figure 4.34. Photographs showing phase-separated dispersions of cellulose nanocrystals mixedwith 8.7 nm (left), 8.7 nm & 107 nm (middle), and 107 nm (right) superparamagnetic magnetitenanoparticles. The photos were taken with reflected white light................................................ 116Figure 4.35. UV-Vis absorption spectra for phase-separated CNC/S-MNP (A) and CNC/L-MNP(B) binary mixtures......................................................................................................................116Figure 4.36. TEM images showing the changes in the size distribution of magnetic nanoparticlesduring the multiple-cycle size-selective separation process, which was conducted with a ternarymixture of S-MNPs, L-MNPs, and CNCs. (A) The initial mixture of S-MNPs and L-MNPs usedin the multicycle separation. (B) MNPs in the liquid crystalline phase of the 1st separation cycle.(C) MNPs isolated from the liquid crystalline phase of the 9th separation cycle. (D) MNPs in theisotropic phase of the 9th separation cycle. All the images were taken at the same magnification.Scale bars, (A-D) 100 nm............................................................................................................ 117Figure 4.37. Photographs showing the phase separations of a CNC/L&S-MNP ternary mixture inthe 1st, 3rd, 5th, 7th, and 9th separation cycles. The photographs were taken with reflected whitelight.............................................................................................................................................. 118Figure 4.38. The size distributions of MNPs in the initial mixture of L-MNPs and S-MNPs (redcolor), in the liquid crystalline phase of the 1st (green) and the 9th (cyan) separation cycles, andin the isotropic phase of the 9th separation cycle (blue)............................................................. 118Figure 5.1. 3D Model showing the exclusion effects of liquid crystalline tactoids (mesogens aredepicted as golden-brown rods) on superparamagnetic magnetite doping nanoparticles (depictedxxvias black spheres), thus the disordered phase has significantly higher magnetic susceptibility thanliquid crystalline tactoids. However, the tactoids have a higher density than isotropic phases dueto the ordered arrangements of mesogens................................................................................... 126Figure 5.2. When a paramagnetic-nanoparticle-doped lyotropic system is subjected to a gradientmagnetic field (H), the movement direction and velocity of a discrete liquid crystalline tactoid isdetermined by the vector sum of four external forces: the weight (Fg) in the vertically downwarddirection, the magnetic body force (Fm) along the gradient magnetic field to the high field region,the magnetic buoyancy force (Fbm) along the gradient magnetic field to the low field region (thisforce is exerted by the continuous isotropic phase surrounding the tactoid), and the gravitationalbuoyancy force (Fbg) in the vertically upward direction exerted by the continuous isotropic phaseas well. The unidirectional movement of tactoids (which is mainly driven by Fbm) results in shearforces, by which the chiral nematic layers are oriented parallel to the magnetic field................129Figure 5.3. TEM micrographs of the superparamagnetic magnetite (Fe3O4) nanoparticles used inthis study. Scale bars, (A,B) 30 nm, (C,D) 20 nm.......................................................................131Figure 5.4. When only affected by the gravity, a CNC-MNP binary mixture phase-separates intoa liquid crystalline phase (exhibiting higher brightness between two crossed polarizers) below adisordered phase over several hours. The CNC-MNP binary mixture has a volume of 1.0 mL andthe phase separation process was observed between two crossed polarizers (in the horizontal andvertical directions) with transmitted white light..........................................................................132Figure 5.5. Photographs of phase-separated CNC dispersions (colorless) and CNC-MNP binarymixtures (yellow colored). A grid background was used to distinguish the isotropic phase (whichis clear) from the liquid crystalline phase (which is birefringent) in (B).................................... 133xxviiFigure 5.6. In a horizontally oriented gradient magnetic field (1050 Gauss/cm) from a permanentmagnet placed beside the vial, the CNC-MNP binary mixture phase-separated into a macroscopicliquid crystalline phase (with a lighter yellow color) in the low-magnetic-field region as well as adisordered phase in the high-magnetic-field region (A), while the pure CNC suspension was notsignificantly affected (B)............................................................................................................. 133Figure 5.7. In a vertical gradient magnetic field (intensity about 1050 Gauss/cm) from a magnetunder the vial, phase separation of the CNC-MNP binary mixture results in a macroscopic liquidcrystalline phase (with a higher brightness between crossed polarizers) above a disordered phase.This process is much faster than in the absence of the gradient magnetic field..........................134Figure 5.8. In a vertical gradient magnetic field from a permanent magnet placed under the vial,the CNC-MNP binary mixture phase-separated into a macroscopic liquid crystalline phase (witha lighter yellow color) above a disordered phase (A). However, the pure CNC dispersion was notsignificantly influenced by the magnetic field (B)...................................................................... 135Figure 5.9. Photographs of phase-separated CNC-MNP binary mixtures captured in cross-linkedpolyacrylamide networks by in-situ photopolymerization. The phase separation of these systemsoccurred in (A) horizontally or (B) vertically oriented gradient magnetic fields........................137Figure 5.10. POM images of a CNC-MNP dispersion (no magnets). Scale bars, 50 μm...........138Figure 5.11. Cross-sectional SEM micrographs for a phase-separated CNC-MNP dispersion (nomagnetic fields). Chiral nematic layers in the macroscopically continuous ordered phase adopt ahorizontal orientation, while discrete liquid crystalline tactoids in the continuous isotropic phaseare nearly randomly oriented. Scale bars, (A) 10 μm, (B) 1 μm, (C) 200 nm, (D) 100 nm........ 139Figure 5.12. 3D model of a CNC-MNP binary mixture (no magnetic fields)............................140Figure 5.13. SEM images of a pure CNC dispersion. Scale bars, (A) 40 μm, (B) 20 μm..........140xxviiiFigure 5.14. SEM images showing the ordered-disordered interfaces of CNC-MNP suspensionsthat phase-separated in a horizontal gradient magnetic field. The interfaces are tilted. Scale bars,(A) 10 μm, (B) 2 μm, (C,F) 1 μm, (D) 500 nm, (E) 3 μm........................................................... 141Figure 5.15. 3D model showing the unidirectional orientation of chiral nematic layers in both thecontinuous ordered phase and discrete liquid crystalline tactoids of a phase-separated CNC-MNPbinary mixture. The phase separation of this system occurred in a horizontally oriented gradientmagnetic field, which resulted in a nearly vertical ordered-disordered interface........................142Figure 5.16. POM micrographs showing the unidirectional orientation of chiral nematic layers ofliquid crystalline tactoids in a CNC-MNP suspension subjected to an external gradient magneticfield (in the direction from left to right with respect to these images). The system was captured ina crosslinked polyacrylamide matrix by photopolymerization during the phase separation process(after a standing time of 60 minutes). The orientation of chiral nematic layers in these tactoids isparallel to the external gradient magnetic field. From (A,B) to (C) and (D), the imaging area wasincreasingly further from the high-magnetic-field region at the magnetic pole. These micrographsindicate that tactoids were driven by magnetic buoyancy forces (from the continuous disorderedphase that has a higher magnetic susceptibility) and move to lower magnetic field regions, wherethey aggregate and coalesce into continuous liquid crystalline phases. Scale bars, 50 μm.........143Figure 5.17. POM micrograph of a CNC-MNP binary mixture that phase-separated in a verticalgradient magnetic field, where the chiral nematic layers in both the continuous ordered phase anddiscrete liquid crystalline tactoids are vertically oriented. Scale bar, 50 μm.............................. 145Figure 5.18. SEM images showing the vertical orientation of chiral nematic layers in continuousordered phases and discrete liquid crystalline tactoids. Scale bars, (A) 15 μm, (B) 10 μm........ 145xxixFigure 5.19. High-resolution SEM micrographs revealing the ordered arrangements of cellulosenanocrystal mesogens in the continuous liquid crystalline phase of a CNC-MNP suspension thatphase-separated in a vertical gradient magnetic field. The pseudo nematic layers formed by theseCNC mesogens are in the vertical direction, which are parallel to the external gradient magneticfield. Scale bars, (A) 300 nm, (B,C) 200 nm, (D) 100 nm.......................................................... 146Figure 5.20. Cross-sectional POM micrographs showing the continuous liquid crystalline phaseof a CNC-MNP binary mixture, which phase-separated in a vertical gradient magnetic field. Thechiral nematic layers are aligned in a vertical orientation throughout this liquid crystalline phase,from the liquid-air interface at the top of the dispersion (A) to the ordered-disordered interface atthe bottom of the anisotropic phase (B-D). Scale bars, 50 μm....................................................147Figure 5.21. SEM images showing the horizontal ordered-disordered interface and vertical chiralnematic layers in a CNC-MNP binary system that phase-separated in a vertical gradient magneticfield. Scale bars, (A) 5 μm, (B) 3 μm, (C) 2 μm, (D) 1 μm, (E) 500 nm, (F) 200 nm.................148Figure 5.22. SEM images showing the horizontal ordered-disordered interface and vertical chiralnematic layers in a CNC-MNP binary system that phase-separated in a vertical gradient magneticfield. Scale bars, (A) 30 μm, (B) 10 μm, (C) 3 μm, (D) 1 μm, (E) 300 nm, (F) 150 nm.............149Figure 5.23. Additional SEM images showing the vertical orientation of chiral nematic layers ina phase-separated CNC-MNP binary mixture. These chiral nematic layers are aligned parallel tothe external gradient magnetic field, perpendicular to the horizontal ordered-disordered interface.In this case, the cross-sectional appearance of chiral nematic structures is also determined by thefracture angle when the cross-section surface was created from the CNC-MNP-PAAm compositepolymer matrix. Scale bars, (A) 5 μm, (B) 2 μm, (C) 1 μm, (D) 400 nm....................................150xxxFigure 5.24. 3D model showing the cross-sectional structure of a CNC-MNP binary mixture thatphase-separated in a vertical gradient magnetic field from a magnet placed under the suspension.In this system, chiral nematic layers in both the continuous ordered phase (the upper region) anddiscrete liquid crystalline tactoids (in the lower region) are vertically oriented, while the interfacebetween the ordered and disordered phases is in the horizontal direction...................................151Figure 5.25. SEM images showing the liquid-air interface at the top of a CNC-MNP suspensionthat phase-separated in a vertical gradient magnetic field. The chiral nematic layers are verticallyoriented near this interface. Scale bars, (A) 100 μm, (B) 30 μm.................................................152Figure 5.26. SEM micrographs showing the biphasic region in a CNC-MNP binary mixture. Thesystem was captured by photopolymerization while the phase separation process was ongoing ina vertical gradient magnetic field from a magnet placed under the dispersion. The chiral nematiclayers in both the continuous ordered phase and discrete liquid crystalline tactoids are verticallyoriented. Scale bars, (A) 20 μm, (B) 10 μm.................................................................................152Figure 5.27. Phase separation times of CNC-MNP dispersions (volume of 1.0 mL) with differentdoping concentrations of Fe3O4 nanoparticles (0 to 140.5 ppm by weight) and in vertical gradientmagnetic fields with different strengths (0, 600, 1050, and 3050 Gauss/cm). The striped columnsrepresent reversed phase separations, where the liquid crystalline phase forms above the isotropicphase. (Errors in the determination of phase separation times are +/− 10 minutes)....................155Figure 5.28. UV-Vis absorption spectra of the isotropic phase (red curve) and liquid crystallinephase (blue curve) of a phase-separated pure CNC dispersion. The two phases were isolated anddiluted by water to 10 times their original volumes.................................................................... 156Figure 5.29. Absorbance of MNPs in isotropic (red) and liquid crystalline phases (blue)........ 157xxxiList of Symbols∇ the Nabla operator° degrees°C degrees Celsiusμ0 vacuum permeability, or magnetic constantμL microlitersμm micrometers, or micronsρ densityχm volume magnetic susceptibilityA Amperesd differential of a functionF force vectorg gramsg gravitational acceleration vectorH magnetic field vectorM magnetization vector, or magnetic moment per unit volumeN NewtonspH power of hydrogenppm parts per millionV volumewt weightxxxiiList of Abbreviations3D three dimensionalCAMB cellulose nanocrystal and polyacrylamide composite microbeadCNC cellulose nanocrystalFT Fourier transformHPLC high-performance liquid chromatographyIR infraredIso isotropicLC liquid crystallineMNP magnetite nanoparticleNP nanoparticlePAAm polyacrylamidePOM polarized optical microscopyRPM revolutions per minuteSEM scanning electron microscopyTEM transmission electron microscopyUV ultravioletVis visiblexxxiiiGlossaryFg weightFbg gravitational buoyancy forceFm magnetic body forceFbm magnetic buoyancy forcexxxivAcknowledgementsI would like to gratefully acknowledge my outstanding supervisor, Prof. Mark MacLachlan,for his excellent guidance over the past four years. His wisdom, encouragement, and unshakableoptimism helped me to overcome all the challenges and difficulties in my PhD research. It is alsoa great pleasure to work under his supervision, not only because he gives me the full freedom topursue my interests, but because I can always get timely and expert assistance when I encounterproblems. It would be impossible for me to complete my studies without his guidance.I would like to thank Dr. Vitor Zamarion, a visiting scientist that generously helped me withmy first research project about the hard-templating of Prussian blue analogues. I may not be ableto complete this difficult project during the first year of my PhD without his kind assistance andadvice. I was lucky to have the help of Vitor at the early stages of my doctoral studies.I would also like to acknowledge Dr. Thanh-Dinh Nguyen, Dr. Jing Xu, Dr. Mostofa Khan,Dr. Georg Meseck, Dr. Takayuki Hiratani, and Dr. Yuanyuan Cao for their valuable suggestionsand comments on my research. I have got a lot of experience and skills from these intelligent andkind-hearted postdoctoral researchers.I would like to thank all my current and former colleagues in the MacLachlan group, for thememorable time we shared together over the past four years: Zhengyu, Susan, Hessam, Andrea,Veronica, Guillaume, Rebecca, Debbie, Lev, Katarina, Andy, Christopher, Yitao, Yiling, Osamu,Miguel, Erlantz, Francesco, Charlotte, Orla, Ingrid, Camille, and Guus. Specifically, I would liketo thank Becky for her fiddle music and old time square dance parties, Lev for his assistance withImageJ, and Debbie for her lovely cats.xxxvI would like to thank Dr. Wadood Hamad for providing all the cellulose nanocrystals I used.I am also grateful to Derrick Horne and Brad Ross for their professional assistance with scanningand transmission electron microscopy. I would also like to acknowledge Dr. Saeid Kamal for hishelp with confocal laser scanning microscopy. And I thank Anita Lam for her assistance with theX-ray powder diffractometry.Finally, I would like to acknowledge my loving parents for their continued support.xxxviDedicationto my parents“为学日益 为道日损”道德经 第四十八章1Chapter 1: Introduction1.1 Liquid Crystals1.1.1 History and Definition of Liquid CrystalsLiquid crystals are a phase of matter in the form of a liquid while exhibiting anisotropy dueto the ordered arrangement of molecules or particles (termed mesogens) resembling crystals. Thefirst observation of liquid crystalline phases dates back to 1888, when Friedrich Reinitzer tried todetermine the melting point of cholesteryl benzoate.1,2 The crystalline solid initially melted into acloudy liquid, which then became clear and transparent at a higher temperature. Compounds likecholesteryl benzoate are thermotropic liquid crystals, which undergo phase transitions in specifictemperature ranges (Figure 1.1).3 Liquid crystalline phases can also form in specific dispersionsof molecules or nanoparticles exceeding a critical concentration, and this type is called lyotropicliquid crystals (Figure 1.2).41.1.2 Classification of Liquid Crystalline PhasesLiquid crystals are generally divided into three classes: nematic, cholesteric (chiral nematic),and smectic. The nematic phases are characterized by long-range collective orientational order ofthe mesogens (Figure 1.3), while smectic phases additionally have long-range positional order inone dimension (Figure 1.4). The cholesteric phases, which are also known as the chiral nematicphases, have twisted structures (Figure 1.5), and this type of liquid crystals was first observed incholesterol derivatives.21.1.3 Applications of Liquid CrystalsLiquid crystals have many applications due to their unique optical, electrical, and magneticproperties. The most important industrial application is liquid crystal displays, which is based onthe twisted nematic effect first discovered by Schadt and Helfrich in 1971.5 Some types of chiralnematic liquid crystals could be used in temperature sensing devices since the helical pitch (andtherefore the color) of these materials changes with temperature.6 Lyotropic liquid crystals, suchas those formed by the self-assemble of surfactants or amphiphilic block-copolymers, have beenwidely used in the template-synthesis of mesoporous materials.7−103Crystalline Liquid Crystalline IsotropicIncreasing TemperatureFigure 1.1. Phase transitions of thermotropic liquid crystals as the temperature increases.Crystalline Liquid Crystalline IsotropicAdding SolventFigure 1.2. Phase transitions of lyotropic liquid crystals as the concentration decreases.4Figure 1.3. Arrangement of rod-shaped mesogens in nematic liquid crystals. The mesogens havea long-range collective orientational order.Figure 1.4. Arrangement of rod-shaped mesogens in smectic liquid crystals, where in addition tothe long-range collective orientational order, the mesogens also have long-range positional orderin one dimension.5Figure 1.5. Arrangement of mesogens in chiral nematic (cholesteric) liquid crystals, which adopta helical superstructure.61.2 Liquid Crystalline Tactoids1.2.1 Tactoids in Lyotropic Liquid CrystalsLiquid crystalline tactoids are discrete ordered microdroplets (sizes from tens to hundreds ofmicrometers) coexisting with continuous disordered phases. Since the pioneering work of Zocheron vanadium pentoxide sols in 1925,11 tactoids have been observed in various types of lyotropicliquid crystals such as tobacco mosaic virus,12 iron oxyhydroxide nanorods,13 polypeptides,14 andcellulose nanocrystals.15 Small tactoids (several micrometers in size) spontaneously emerge fromdisordered phases; these liquid crystalline microdroplets gradually grow larger by coalescence asthe system relaxes, and eventually merge into macroscopically continuous ordered phases. Thus,tactoids mediate the transitions between isotropic and liquid crystalline phases in these lyotropicsystems.161.2.2 Emergence of Tactoids in Disordered PhasesWhen an originally isotropic dispersion (below the critical concentration) is concentrated byevaporation, the formation of tactoids as birefringent microdroplets can be observed by polarizedoptical microscopy (POM). While the direct and real-time monitoring of liquid crystal mesogensduring the emergence of a tactoid is still difficult, this process has been theoretically investigatedsince the 1930s.17 The formation of a liquid crystalline tactoid requires both an attractive force togather together the mesogens into a microdroplet and a repulsive force to arrange these mesogensinto liquid crystalline order. The interactions between mesogens are typically simulated using theDLVO (Derjaguin-Landau-Verwey-Overbeek) theory, where the electrostatic repulsion betweenthe charged mesogens is balanced by the van der Waals attractive forces.18,19 However, Langmuir7demonstrated that the electrostatic attractive forces between the liquid crystalline mesogens andoppositely charged intervening counterions are enough to balance the repulsive interactions, suchas the osmotic pressures and hydration.17,20Although the above thermodynamic simulations can partly explain how tactoids are createdby competitive repulsive and attractive forces, it remains unclear whether these liquid crystallinemicrodroplets have any unique intrinsic characteristics that distinguish them from the disorderedenvironment. Are the mesogens in a tactoid somehow different from those in the nearby isotropicphase? This is a difficult issue to address due to the lack of direct microscopic observations, butstudies on phase-separated lyotropic liquid crystals might give some hints since phase separationin these systems is usually mediated by the coalescence and sedimentation of tactoids.In a recent study based on phase-separated cellulose nanocrystal dispersions, the rod-shapedmesogens in liquid crystalline phases were found to have significantly longer lengths and higheraspect ratios than those in the isotropic phases.21 This discovery is consistent with the theoreticalsimulations based on the Onsager theory,22 which reveals that when phase separation occurs in adispersion of rod-shaped nanoparticles with different aspect ratios, there will be a higher fractionof the longer nanorods in the liquid crystalline phase than in the disordered phase. As it has beenconfirmed that the phase transitions in cellulose nanocrystal dispersions are mediated by tactoids,which kinetically transfer liquid crystal mesogens from one phase to the other, the enrichment ofhigh-aspect-ratio mesogens in the macroscopic liquid crystalline phases may have occurred sincethe nucleation of tactoids.According to the Onsager theory,23 the critical concentration for phase separation in a roddispersion decreases as the aspect ratio of the rods increases. Because the emergence of a liquidcrystalline tactoid is intrinsically a process of microscopic phase separation that occurs when the8local critical concentration is reached, if an initially isotropic dispersion of rod-shaped mesogensis slowly and homogeneously concentrated, liquid crystalline tactoids will first emerge from themicrodomains that are rich in high-aspect-ratio rod-shaped mesogens because of the lower localcritical concentrations, and thus selectively collect these mesogens into the macroscopic orderedphase by coalescence and sedimentation.1.2.3 Geometries and Director Fields of TactoidsAs demonstrated by a series of studies on vanadium pentoxide tactoids,24−26 the geometry ofa large-sized nematic tactoid is determined by the competition between the surface energy fromthe ordered-disordered interfacial tension,27 which favors a spherical shape, and the orientationalelastic energy of the nematic liquid crystalline phase, which tends to elongate the tactoid. On theother hand, a small-sized tactoid is shaped by the competition between the surface energy and theanchoring energy at the boundary of the tactoid, where the latter is caused by the deviation of thedirector field from the tangential orientation at the ordered-disordered interface.Experimentally, the geometry and director field configurations of a tactoid can be examinedby polarized optical microscopy as the alignment of mesogens results in birefringence. Nematictactoids formed in lyotropic liquid crystals of rod-shaped particles, such as vanadium pentoxidenanoribbons,28 aluminum oxyhydroxide nanorods,29 or tobacco mosaic viruses (Figure 1.6),30 areelongated spindle-shaped microdroplets with circular arc boundaries. Tactoids in smectic liquidcrystals, such as fd virus dispersions, are spindle-shaped microdroplets with a number of smecticrings periodically spaced along the long axes.31 Chiral nematic liquid crystalline tactoids, such asthose formed in polypeptide solutions and cellulose nanocrystal suspensions (Figure 1.7),14,15 arespherical or ellipsoidal-shaped microdroplets with periodic birefringent chiral nematic layers.9Figure 1.6. Nematic liquid crystalline tactoids formed in a dispersion of tobacco mosaic viruses,which are spindle-shaped birefringent microdroplets.12Figure 1.7. Chiral nematic liquid crystalline tactoids formed in a polypeptide solution, which arespherical microdroplets with periodically spaced birefringent layers.14101.3 Cellulose Nanocrystals1.3.1 Chiral Nematic Phases Formed by Cellulose NanocrystalsCellulose nanocrystals (CNCs) are rod-shaped nanoparticles obtained by partial hydrolysisof natural cellulose in sulfuric acid. They have lengths of 100 to 400 nm, and widths of 10 to 30nm (Figure 1.8). Due to the negatively charged sulfate ester groups on the surface, CNCs couldform a stable colloidal suspension in water. In 1959, Marchessault and coworkers first observedthe formation of permanently birefringent gels from colloidal dispersions of cellulose crystallitesafter acid hydrolysis.32 In 1992, Gray and Marchessault demonstrated that cellulose nanocrystalscould self-assemble into chiral nematic liquid crystalline phases in aqueous suspensions above acritical concentration.15As a sustainable nanomaterial that can be easily extracted from forests and fields, cellulosenanocrystals have attracted widespread interest in recent years for their optical, mechanical, andelectromagnetic properties that are being developed for a variety of applications.33 The lyotropicliquid crystalline behavior of CNCs is especially valuable since it could be used for the templatesynthesis of many organic and inorganic materials. In 2010, MacLachlan and coworkers reportedthe fabrication of mesoporous silica films with chiral nematic structures.34 Afterwards, a series ofother materials with chiral nematic ordering and photonic properties were synthesized.35−46Despite the growing interest in the solution and solid-state properties of CNCs, there is stilla large gap in our understanding of the liquid crystallinity and microstructures of these materials.One hypothesis is that the chirality originates from a helical twist of the crystals,47,48 and severalresearchers have observed twisting of cellulose microfibrils or CNCs by electron microscopy andatomic force microscopy,49,50 but the twisting is quite subtle. Bergström and coworkers reported11that the chiral nematic phase has already been established when the CNCs are about 50 nm apartin water.51 It is not clear how chirality is mediated between the CNCs over these distances. Someresearchers postulated that there would be a chiral charge distribution of the sulfate ester groupson the surface resulting from the sulfuric acid hydrolysis process.15 The transition from isotropicto lyotropic liquid crystalline phases in the case of CNCs have been difficult to study, but criticalunderstanding might come from the microstructures of tactoids, which are key components in theevolution of liquid crystallinity in CNC suspensions.1.3.2 Liquid Crystalline Tactoids of Cellulose NanocrystalsAbove a critical concentration, tactoids will spontaneously emerge in the isotropic phases ofCNC dispersions, which can be observed by polarized optical microscopy (POM) as birefringentellipsoidal microdroplets with periodically spaced chiral nematic layers (Figure 1.9). As tactoidscoalesce and settle, the system phase-separates into a liquid crystalline phase below an isotropicphase. When dried, a solid chiral nematic film will be formed.The significance of tactoids in improving the properties of CNC-derived materials has beennoticed. Lagerwall and coworkers reported that applying shear forces to CNC dispersions as theydried can improve the alignment of the tactoids and, consequently, improve the optical propertiesof the chiral nematic films.52 While hypothetical models of tactoids have been built, the structureof tactoids and the arrangement of liquid crystalline mesogens within them had not been directlyobserved with electron microscopy due to the difficulty of capturing these soft microdroplets in asolid-state matrix while maintaining their ordered structures, yet this information can be valuablebecause tactoids are the earliest evolution stages and smallest units of liquid crystalline phases.12A BC DFigure 1.8. TEM micrographs of cellulose nanocrystals used in this research. Scale bars, (A) 200nm, (B) 100 nm, (C,D) 50 nm.13Figure 1.9. POM micrograph showing discrete chiral nematic liquid crystalline tactoids formedin an aqueous dispersion of cellulose nanocrystals. Scale bar, 50 μm.141.4 Aims and ScopeThe research presented in this thesis focuses on the microscopic structures and fundamentalphysical properties of liquid crystalline tactoids. While most experiments herein were conductedwith cellulose nanocrystals, these results might help to understand the evolution and dynamics ofother lyotropic liquid crystals and soft matter systems.In Chapter 2, a generalized method was developed to enable direct observations of tactoidsby electron microscopy. Using an in-situ photopolymerization approach, a cross-linked polymermatrix could be rapidly formed to capture and solidify liquid crystalline tactoids in a fluid systemwithout significant distortions. This approach allowed to directly visualize, for the first time, thearrangement of liquid crystal mesogens within individual tactoids by electron microscopy.53In Chapter 3, the structure and evolution of chiral nematic liquid crystalline tactoids in thegeometrical confinement of spherical microdroplets were examined by both optical and electronmicroscopy. Polymer microspheres with left-handed chiral nematic order were obtained throughin-situ photopolymerization in inverse emulsions. By using a double-matrix templating method,mesoporous silica microspheres with chiral nematic structures were fabricated, which may havepotential applications in optical devices and chiral separations.54In Chapter 4, the structure and evolution of tactoids in the presence of doping nanoparticleswere directly investigated by scanning electron microscopy at the resolution of individual liquidcrystal mesogens, which revealed the size-selective exclusion effects of liquid crystalline tactoidson foreign nanoparticles. This phenomenon indicates a new approach to size-selectively separatenanoparticles using lyotropic liquid crystals, where the nanoparticles with widths smaller than athreshold size will be selectively transferred from the disordered phase into the ordered phase by15liquid crystalline tactoids during the phase separation process. This principle was applied to theseparation of polymer nanospheres, gold nanoparticles, and magnetic nanoparticles by size.55In Chapter 5, the phase separations of lyotropic liquid crystals in competitive gravitationaland magnetic acceleration fields were investigated. A significant difference in volume magneticsusceptibility was created between liquid crystalline tactoids and disordered phases based on theexclusion effects of tactoids on superparamagnetic doping nanoparticles, which enabled positionand orientation control of liquid crystalline tactoids by static gradient magnetic fields as weak asseveral hundred Gauss/cm. The movement of tactoids is determined by the competition betweenthe magnetic and gravitational acceleration fields. This approach allowed control of the rate andconfiguration of phase separation, as well as the orientation of the director fields in both discretetactoids and continuous macroscopic ordered phases.16Chapter 2: Structure and Transformation of Liquid Crystalline Tactoids2.1 IntroductionUntil a few years ago, structural studies of liquid crystalline tactoids were mainly based onoptical microscopy or other non-destructive testing methods due to the highly deformable natureof these fluid microscopic structures. Unfortunately, optical techniques are usually not capable ofrevealing the arrangement of mesogens due to their limited resolution. A possible solution is tocapture and solidify these liquid crystalline microdroplets by introducing a cross-linked polymermatrix into the fluid system, during which the ordered microstructures of tactoids should be wellmaintained without distortions.2.2 Solidification of Liquid Crystalline Tactoids by In-Situ PhotopolymerizationAqueous dispersions of cellulose nanocrystals were combined with the non-ionic precursorsof polyacrylamide (PAAm), which are acrylamide (monomer), a crosslinker, and a photoinitiator.After homogenization, the dispersion was allowed to stand in the dark for a prescribed period oftime (1 to 12 hours). Afterwards, ultraviolet radiation (wavelength of 300 nm) was applied to thesystem, and a crosslinked polymer matrix rapidly formed to capture the liquid crystalline tactoidsin the fluid system (Figure 2.1). After removal of solvent, the polyacrylamide matrix turned intoa solid-state plastic with the microstructures of tactoids stabilized inside the matrix (Figure 2.2).This results in a "fossil record" that represents a transition stage of the evolution of this lyotropicliquid crystalline phase. The solid polymer matrix can be cracked to give fresh fracture surfaces,where the microscopic structures of tactoids and other liquid crystalline entities could be directlyobserved by electron microscopy at the resolution of individual mesogens.17Figure 2.1. By adding the nonionic precursors of polyacrylamide into an aqueous suspension ofcellulose nanocrystals, this system could be captured in a cross-linked polymer matrix by in-situphotopolymerization, forming a CNC-PAAm composite hydrogel. The diameter of the Petri dishin this photograph is about 50 mm.Figure 2.2. Photograph showing a completely dried CNC-PAAm composite polymer matrix, inwhich the microstructures of tactoids have been solidified.182.3 Electron Microscopy Observations of Liquid Crystalline Tactoids2.3.1 Microstructures of Chiral Nematic Liquid Crystalline TactoidsAs revealed by the cross-sectional scanning electron microscopy (SEM) observations of thesolid-state samples captured by in-situ photopolymerization, liquid crystalline tactoids formed inCNC dispersions are discrete ordered microdroplets with spherical or ellipsoidal boundaries thatare clearly distinguishable from the isotropic phase. Small tactoids seem to be unwound nematicdue to the boundary conditions since cellulose nanocrystal mesogens in them are unidirectionallyaligned (Figure 2.3). As a discrete liquid crystalline microdomain, this tactoid is separated fromthe surrounding disordered phase by an arc-shaped sharp boundary (Figure 2.4). CNC mesogensinside this liquid crystalline tactoid are arranged into nematic layers (Figure 2.5), while outsidethis domain they are substantially disordered, consistent with an isotropic phase. Larger tactoidsadopt left-handed chiral nematic order appearing as a series of periodically spaced parallel layers(Figure 2.6 to Figure 2.10), where each of these periodic layers represents a half-helical pitch ofthe chiral nematic phase, i.e., the mesogens twist by 180 degrees from one end to the other.To obtain three-dimensional microstructures of liquid crystalline tactoids, an intersection oftwo fracture surfaces that are perpendicular to each other was created from a dried CNC-PAAmcomposite matrix (Figure 2.11 and Figure 2.12), where the left-handed chiral nematic order in atactoid was clearly confirmed by direct electron microscopy observations from multiple angles atthe resolution of individual CNC mesogens (Figure 2.13).19A BC DE FFigure 2.3. SEM micrographs of a newly emergent liquid crystalline tactoid formed by cellulosenanocrystals. Scale bars, (A) 2 μm, (B) 1 μm, (C) 300 nm, (D) 200 nm, (E,F) 100 nm.20A BC DFigure 2.4. Cross-sectional SEM micrographs showing the boundary of a newly emergent liquidcrystalline tactoid near its (A,B) left and (C,D) bottom regions. The boundary of this tactoid is amicroscopic ordered-disordered interface, which sharply separates the liquid crystalline domainfrom the isotropic phase surrounding it. Scale bars, (A,C) 200 nm, (B,D) 100 nm.21A BC DFigure 2.5. High-resolution SEM micrographs showing the ordered arrangements of rod-shapedcellulose nanocrystal mesogens in a newly emergent liquid crystalline tactoid. The mesogens areunidirectionally aligned into a nematic phase since this tactoid is smaller than a half helical pitch,however, the director field is still slightly twisted in long ranges due to the chirality of this phase.Scale bars, (A-D) 100 nm.22A BC DE FFigure 2.6. Cross-sectional SEM micrographs showing a liquid crystalline tactoid with only onechiral nematic layer. Scale bars, (A) 1 μm, (B,C) 300 nm, (D-F) 200 nm.23Figure 2.7. 3D model showing the left-handed chiral nematic order of CNC mesogens (depictedas rods in a golden brown color) in a newly formed liquid crystalline tactoid with only one chiralnematic band. The rod-shaped mesogens in the pseudo nematic layers rotate by 180 degrees fromone end to the other. This is an idealized model, while in reality the CNCs might not be perfectlylayered and aligned.24A DB EC FFigure 2.8. Additional cross-sectional SEM micrographs showing liquid crystalline tactoids with(A,B) one, (C) zero, or (D-F) three chiral nematic layers. Scale bars, (A) 2 μm, (B) 400 nm, (C)1 μm, (D) 3 μm, (E) 500 nm, (F) 200 nm.25A DB EC FFigure 2.9. Cross-section SEM micrographs showing a liquid crystalline tactoid with four chiralnematic layers. Scale bars, (A) 5 μm, (B,D) 2 μm, (C,E) 1 μm, (F) 300 nm.26A BC DE FFigure 2.10. Cross-sectional SEM micrographs of a large-size liquid crystalline tactoid with ninechiral nematic layers. Scale bars, (A) 10 μm, (B) 2 μm, (C-F) 1 μm.27A BC DE FFigure 2.11. SEM micrographs showing several chiral nematic liquid crystalline tactoids sittingat a right-angled edge. Scale bars, (A) 200 μm, (B) 50 μm, (C,D) 10 μm, (E,F) 3 μm.28A DB EC FFigure 2.12. SEM micrographs showing the (A-C) top and (D-F) front views of a tactoid sittingat a right-angled edge. Scale bars, (A) 10 μm, (B,D) 2 μm, (E) 1 μm, (C,F) 500 nm.29A DB EC FFigure 2.13. SEM micrographs showing the (A-C) left and (D-F) right views of a chiral nematicliquid crystalline tactoid. Scale bars, (A,D) 500 nm, (B,E,F) 300 nm, (C) 200 nm.302.3.2 Coalescence of Tactoids and Formation of Topological DefectsAs the total concentration of a CNC suspension increases or as the lyotropic system relaxes,bigger tactoids with more chiral nematic layers appear. By tracking the evolution of tactoids in adrying dispersion by POM for hours, the coalescence of liquid crystalline tactoids was observed,during which time several smaller tactoids merged together to form a larger tactoid (Figure 2.14and Figure 2.15). Because the resulting tactoid favors a spherical geometry, the driving force forthis coalescence mechanism is the ordered-disordered interfacial tension on the boundaries of thefusing tactoids. The interfacial tension is of the order of magnitude of 10−4 mN/m as measured byGray and coworkers;27 for comparison, the interfacial tension between hexane and water is about50 mN/m.56 These data could help explain the relatively slow phase separation rates of cellulosenanocrystal dispersions. It should be noted that other growth mechanisms, for example, Ostwaldripening, might be simultaneously active.In some situations, especially when two tactoids come together with the helical axes of theirchiral nematic structures perpendicular to each other, the coalescence between tactoids will resultin the formation of topological defects in the director field of the resulting tactoid, as this processis the combination of multiple liquid crystalline microdomains with initially random orientations.The chiral nematic layers would bend, fold, elongate or dislocate during the coalescence process.Interestingly, the layers seem to be soft and flexible enough to deform while accommodating theunion. The topological defects could be healed by the reorientation of the director fields becausethey are thermodynamically unstable, as planar chiral nematic layers are favored structures underweak confinement and surface anchoring conditions, which are respectively caused by the weakordered-disordered interfacial tension at the boundary of a tactoid and the relatively large sizes of31CNC mesogens. In some cases, however, the defective structures remain in the fused tactoid fora long time due to the slow relaxation rates of the liquid crystalline mesogens.By capturing the CNC suspensions in crosslinked polyacrylamide matrices, the coalescenceof liquid crystalline tactoids was further investigated by scanning electron microscopy, where theencounter and direct contact between multiple tactoids that have random helical axis orientations(Figure 2.16 to Figure 2.18) as well as tactoids with fusion defects of dislocated or folded chiralnematic layers (Figure 2.19) were observed. These results indicate a new method to examine themicrostructures of topological defects in lyotropic liquid crystals, where the in-situ formation ofa crosslinked polymer matrix not only provides high structural stability for electron microscopyanalysis, but also permanently captures a thermodynamically metastable transition state from analways-changing liquid crystalline system.32Figure 2.14. Three-dimensional model showing the contact between two discrete chiral nematicliquid crystalline tactoids in an aqueous dispersion of cellulose nanocrystals.330 s 8 s 10 s 12 s 14 s18 s 24 s 40 s 56 s 94 sFigure 2.15. POM micrographs showing the coalescence of two chiral nematic liquid crystallinetactoids in a CNC dispersion.Figure 2.16. SEM micrograph showing the initiation of coalescence between two chiral nematictactoids. Scale bar, 10 μm.34A BC DFigure 2.17. SEM images showing the microstructures of the contact point between two tactoids.Scale bars, (A) 2 μm, (B) 1 μm, (C) 500 nm, (D) 300 nm.35A BC DE FFigure 2.18. SEM images showing the aggregation and contact between multiple chiral nematicliquid crystalline tactoids. Scale bars, (A) 20 μm, (B,C) 10 μm, (D-F) 5 μm.36A BC DE FFigure 2.19. SEM images of liquid crystalline tactoids with topological defects of dislocated orfolded chiral nematic layers. Scale bars, (A) 5 μm, (B-F) 10 μm.372.3.3 Gravitational Sedimentation of Liquid Crystalline TactoidsThe CNC mesogens would be more closely packed inside liquid crystalline tactoids as theyadopt a similar orientation in each pseudo nematic layer, which enables them to occupy the spacemore efficiently. This may lead to a higher density of CNCs inside tactoids than in the isotropicphases. Quantifying the density of the discrete tactoids would be difficult due to their small sizesand unstable structures, but this value could be estimated from the density of macroscopic liquidcrystalline phases as they are formed by the same mesogens adopting the same ordered structures.In this study, an aqueous dispersion of CNCs (4.08 wt.%) was allowed to stand in a sealedseparatory funnel for one week to reach phase equilibrium, afterwards, the concentration of theCNCs in the ordered phase was measured to be 4.46 wt.%, while the CNC concentration in thedisordered phase was 3.75 wt.%. The densities of the liquid crystalline and isotropic phases are1.0165 g/cm3 and 1.0150 g/cm3, respectively.As tactoids have a higher density than the isotropic phase due to a more efficient packing ofthe liquid crystalline mesogens, they gradually settle to the bottom of the dispersion as driven bygravity, and eventually coalesce together to form a macroscopically continuous liquid crystallinephase, in which the chiral nematic layers are horizontally stacked (Figure 2.20).Fusions could also occur between discrete liquid crystalline tactoids and continuous orderedphases when tactoids encounter the horizontal ordered-disordered interface at the bottom of thedispersion, and join the macroscopic chiral nematic phase by coalescence. This process may alsoresult in the formation of topological defects due to the random orientation of the helical axis ofdiscrete tactoids.As a result of the coalescence between tactoids with different director fields, a large numberof kinetically arrested topological defects could be observed in the macroscopic liquid crystalline38phase (Figure 2.21 and Figure 2.22), which have significantly longer periodicity (about 5 to 10micrometers) than the half helical pitch of the chiral nematic structure of CNCs (shorter than onemicrometer). These topological defects may be one of the possible origins of the fingerprint-liketextures observed in solid-state chiral nematic films formed by CNCs (Figure 2.23).Figure 2.20. Cross-sectional SEM micrograph showing the gravitational sedimentation of chiralnematic liquid crystalline tactoids to the bottom of a CNC dispersion, where the tactoids coalescetogether to form a macroscopic continuous ordered phase. Scale bar, 20 μm.39A BC DE FFigure 2.21. SEM micrographs showing topological defects as an array of folded chiral nematiclayers. Scale bars, (A) 5 μm, (B,C) 2 μm, (D) 1 μm, (E,F) 500 nm.40A BC DFigure 2.22. SEM micrographs showing topological defects as an array of folded chiral nematiclayers. Scale bars, (A) 10 μm, (B) 3 μm, (C) 1 μm, (D) 500 nm.41Figure 2.23. POM micrograph showing the fingerprint-like texture in a solid-state chiral nematicfilm formed by CNCs. Scale bar, 50 μm.422.4 ConclusionsWith the aid of an in-situ photopolymerization method, liquid crystalline tactoids formed inCNC dispersions were rapidly captured and solidified in crosslinked polymer matrices, enablingdirect electron microscopy observations of these metastable soft matter entities at the resolutionof individual liquid crystal mesogens. The results reveal that the tactoids themselves have chiralnematic liquid crystalline structures and they are already highly ordered (Figure 2.24).According to optical and electron microscopy observations, tactoids grow by a coalescencemechanism, where several smaller tactoids merge together to form a larger one. This coalescenceprocess is driven by the ordered-disordered interfacial tension, and usually generates topologicaldefects in the director field of the resulting tactoid since this is the combination of multiple liquidcrystalline microdomains with initially random orientations. The defects could be healed by thereorientation of the director field as they are thermodynamically unstable, but, in some cases, thedefective structures remain in the fused tactoid for a long time because of the slow relaxation rateof the mesogens.As tactoids have a higher density than the isotropic phase due to a more efficient packing ofthe mesogens, they will gradually settle to the bottom of the suspension and eventually coalesceinto a macroscopic continuous liquid crystalline phase containing kinetically arrested topologicaldefects (Figure 2.25 and Figure 2.26).43Figure 2.24. 3D model showing the left-handed chiral nematic arrangement of rod-shaped CNCmesogens in a liquid crystalline tactoid.44A BFigure 2.25. 3D models showing the coalescence of liquid crystalline tactoids and the formationof topological defects as folded chiral nematic layers.45Figure 2.26. 3D model showing topological defects as an array of folded chiral nematic layers inthe continuous liquid crystalline phase at the bottom of a CNC dispersion.462.5 Experimental Methods2.5.1 Capture of Tactoids by In-Situ PhotopolymerizationThe in-situ photopolymerization capture of tactoids was conducted with nonionic precursorsof polyacrylamide, which are compatible with the self-assembly of CNCs in aqueous dispersions.In a typical experiment, acrylamide (1.0 mg, monomer), N,N′-methylenebisacrylamide (100 mg,cross-linker) and 2-hydroxy-4′-(2-hydroxyethoxy)-2-methylpropiophenone (5 mg, photoinitiator;can be replaced by 50 mg of 2,2-diethoxyacetophenone) were mixed with an aqueous dispersionof CNCs (10 mL, 4 wt.%, conductivity = 2.19 mS/cm, pH = 2.44). After sonication for about 10minutes in cold water, the homogeneous mixture (total volume of about 11.2 mL) was placed ina polystyrene Petri dish (diameter of about 50 mm) to evaporate under ambient conditions in thedark for a prescribed period of time (1, 3, 6, 9, 12 or 24 hours). Afterwards, an ultraviolet-B lightsource (wavelength of 300 nm, 8 W) was used to initiate the in-situ photopolymerization process.Ultraviolet irradiation was applied to the system for about 30 minutes, a cross-linked hydrogel ofpolyacrylamide was rapidly generated to capture the liquid crystalline tactoids and other orderedmicroscopic structures formed by the self-assembly of CNCs (in general, robust hydrogels couldbe obtained in 20 minutes).Samples for SEM observations were obtained by heating the hydrogels in air at 60 °C for 12hours, then breaking the resulting hard and brittle plastic blocks into small pieces with a hammer.2.5.2 Concentration and Density of Ordered and Disordered PhasesTo determine the concentration and density of the isotropic and liquid crystalline phases, anaqueous dispersion of CNCs (volume of about 200 mL) was placed in a sealed separatory funnel47to stand at 8 °C for 7 days. The two phases were then separated and the concentrations of CNCswere determined as follows: the mass of a CNC suspension (10.0 mL) in a pre-weighed vial wasmeasured with an analytical balance, which was then completely dried by heating at 75 °C for 24hours, cooled and weighed again. Densities of the two phases were determined by measuring themass of a suspension in a volumetric flask (volume of 10.0 mL) and comparing with the mass ofdeionized water of the same volume.2.5.3 MaterialsCNC suspensions (4 wt.% in water, conductivity = 2.19 mS/cm, pH = 2.44) were providedby FPInnovations. The CNCs were obtained from hydrolysis of wood pulp in sulfuric acid usinga literature procedure. TEM micrographs of these crystals gave dimensions of 19 ± 9 nm by 245± 135 nm (based on the measurements of 98 cellulose nanocrystals). Acrylamide (Aldrich, 98%),N,N′-methylenebisacrylamide (Aldrich, 99%), 2,2-diethoxyacetophenone (Acros, 98%), as wellas 2-hydroxy-4′-(2-hydroxyethoxy)-2-methylpropiophenone (Aldrich, 98%) were used withoutfurther purification.2.5.4 CharacterizationOptical microscopy was performed with an Olympus BX41 microscope. Scanning electronmicroscopy (SEM) experiments were conducted on a Hitachi S4700 electron microscope at anacceleration voltage of 10 kV. Samples were sputter-coated with 8.0 nm of platinum-palladium(80 to 20) alloy before imaging. Transmission electron microscopy (TEM) images were obtainedwith a Hitachi H7600 electron microscope at an acceleration voltage of 62 kV. Samples for TEMmeasurements were prepared by diluting an aqueous suspension of CNCs to about 0.002 wt.%,48sonicating the diluted suspension for 3 hours in cold water, then drop-casting the suspension ontocarbon-coated copper grids.A Durasonix 3 Litre Ultrasonic Cleaner (frequency of 40 kHz, power of 120 W) was used inthis study for sonication treatments of CNC dispersions.49Chapter 3: Liquid Crystalline Tactoids in Geometrical Confinement3.1 IntroductionThe behavior of liquid crystals in confined geometries is significant both for understandingthe topological properties of these ordered soft matters and for the development of novel opticaldevices.57,58 However, since most previous studies were based on molecular thermotropic liquidcrystals with fast phase transition rates, the intermediate states in these systems were difficult todistinguish and elucidate. New insights would be provided by lyotropic liquid crystals becausethey generally have much slower kinetics that would enable tracking of the phase transitions ingeometrical confinement during a relatively long period of time.Herein, using an inverse-emulsion photopolymerization method, the structure and evolutionof chiral nematic liquid crystalline tactoids confined in spherical microdroplets were investigatedby both optical and electron microscopy. Subsequently, solid-state microspheres with left-handedchiral nematic structures were fabricated, the diameters of which could be tuned between tens tohundreds of micrometers. Mesoporous silica microspheres with chiral nematic order were furthersynthesized through a double-matrix method. These novel polymer and silica microspheres withchiral nematic structures may offer new opportunities for HPLC and other applications. As well,the observations in this study might provide a new approach to understand the evolution of chiralnematic liquid crystals in geometrical confinement, especially spherical confinement, which hasattracted considerable attention due to its potential applications in optics and sensing.59−61503.2 Microspheres with Chiral Nematic Order from Liquid Crystalline Tactoids3.2.1 Evolution of Liquid Crystalline Tactoids in Spherical ConfinementIn this research, the microdroplets of lyotropic liquid crystals were prepared by emulsifyinga homogenized aqueous dispersion of cellulose nanocrystals in a non-polar organic solvent suchas cyclohexane. The evolution of chiral nematic tactoids in these microdroplets was tracked as afunction of time, where the samples were taken from the emulsion system at 30-minute intervalsand immediately examined by POM.Initially, newly emergent tactoids are small enough to avoid the confinement effect from themicrodroplet boundary, therefore they have planar parallel chiral nematic layers similar to thoseformed in bulk dispersions (Figure 3.1 A). As time passes, more tactoids emerge in the isotropicphase, they coexist in the same microdroplet and grow by coalescence (Figure 3.1 B,C). When atactoid is large enough to be affected by the spherical confinement of the microdroplet boundary,its chiral nematic layers would be curved to accommodate the water-oil interface (Figure 3.1 D),which leads to the formation of a concentric spherical multi-shell structure with radially orientedhelical axes (Figure 3.1 E). The chiral nematic cores may keep growing and finally occupy mostof the space in the aqueous microdroplets (Figure 3.1 F).Although the confinement of chiral nematic liquid crystals in spherical geometries has beenextensively investigated,62−65 in most cases only thermotropic liquid crystals were studied, whichare molecular substances that can self-assemble very quickly. As large rod-shaped nanoparticles,CNCs have much slower dynamics. Therefore, the evolution of liquid crystalline phases in CNCdispersions can take many hours, providing a good opportunity to study the formation of a chiralnematic phase in a confined geometry starting from an isotropic dispersion.51A B CD E FFigure 3.1. POM micrographs showing the evolution of chiral nematic liquid crystalline tactoidsin spherical confinement. The evolution time and diameter of these microdroplets are as follows:(A) 0.5 h, 118 μm; (B) 1 h, 58 μm; (C) 3 h, 64 μm; (D) 6 h, 78 μm; (E) 9 h, 57 μm; and (F) 12 h,114 μm.523.2.2 Solidification of Microdroplets by In-Situ PhotopolymerizationIn a standard experiment, a CNC dispersion (4.0 wt.%) was homogeneously mixed with thenonionic precursors of polyacrylamide (i.e., acrylamide, a crosslinker, and a photoinitiator), thenemulsified under nitrogen or argon protection in cyclohexane (1:10 v/v) facilitated by surfactantSpan-80 (0.8% w/v) and moderate stirring (400 rpm) to give liquid microspheres. Chiral nematictactoids of CNCs were allowed to grow in the microdroplets for a prescribed period of time (6 to9 hours) with the system sealed, then photopolymerization was initiated by ultraviolet irradiation(wavelength of 300 nm, 8W) to capture the tactoids in crosslinked polyacrylamide microspheres(Figure 3.2). After the removal of solvents, solid-state CNC-PAAm composite microbeads withchiral nematic structures were obtained. The microspheres prepared in this way have an averagediameter of 125 μm (denoted as CAMB-M, Table 3.1).Table 3.1. Preparation parameters for CNC-PAAm chiral nematic microspheres.SampleNameCyclohexane(mL)Span-80(g)CNCDispersion(mL)StirringSpeed(RPM)AverageDiameter(μm)CAMB-M 80 0.6 6.0 400 125CAMB-S 80 1.2 6.0 600 30CAMB-L 80 0.4 6.0 300 22053Figure 3.2. Sketch showing the in-situ photopolymerization of aqueous CNC microdroplets in aninverse emulsion. Using this approach, CNC-PAAm composite microspheres with chiral nematicstructures could be obtained.543.2.3 Structure of Liquid Crystalline Tactoids in Spherical ConfinementThe as-prepared CNC-PAAm composite microspheres initially dried from ethanol are whiteand opaque, but transparent and colorless microbeads could be obtained by redispersing them inwater then drying again under ambient conditions (Figure 3.3 A). The chiral nematic tactoids ofCNCs can be observed in the microspheres by POM between crossed polarizers (Figure 3.3 B).They showed concentric retardation lines with a periodic spacing of about 5 μm, which indicatesradially oriented helical axes. Many of the microspheres showed a distinct Maltese cross patternwhen observed by POM, indicative of their spherical symmetry.These microspheres can swell into transparent microgels in water due to the hydrophilicityof polyacrylamide, in which the chiral nematic layers of the captured CNC tactoids can be moreclearly observed by POM both between crossed polarizers (Figure 3.4) and under reflected whitelight (Figure 3.5). The structures of chiral nematic layers in these microspheres could be furtherrevealed by confocal laser scanning microscopy after staining the samples with a fluorescent dye(e.g., rhodamine B) aqueous solution (Figure 3.6 and Figure 3.7). Here the contrast might arisefrom the uneven scattering of the emitted fluorescence due to the optical anisotropy of the chiralnematic layers, rather than an uneven distribution of the dye. It is obvious that the chiral nematiclayers are organized into a series of concentric spherical shells (an onion-like structure), which isconsistent with the concentric circle patterns observed by POM. However, the overall geometryof the layers is a spiral shape, leaving a topological defect at the center.Interestingly, in some CNC-PAAm microspheres, the liquid crystalline tactoids have planarparallel chiral nematic layers (Figure 3.8 A), which are similar to those formed in the bulk CNCdispersions. Tactoids with curved (arc-shaped) chiral nematic layers were also observed in somemicrospheres (Figure 3.8 B,C). Nevertheless, in most cases, the chiral nematic layers of tactoids55are well organized as concentric spherical shells (Figure 3.8 D). These structures may representdifferent stages of the evolution of liquid crystalline tactoids in spherical microdroplets. Initially,a newly formed tactoid in the center of an aqueous microdroplet has planar chiral nematic layersdue to the long distance between its edge and the water-oil interface (Figure 3.9 A). The size ofthis tactoid gradually increases over time, accompanied by the generation of new chiral nematiclayers, and eventually the growth of this tactoid will be constrained by the boundary of the watermicrosphere. The chiral nematic layers would bend to fit the spherical geometry of the water-oilinterface (during this process an S = +1/2 disclination is formed), they keep extending along theinterface (Figure 3.9 B,C), which finally results in a concentric spherical multi-shell structure (inan ideal situation) or a spiraling structure (Figure 3.9 D and Figure 3.10).56A BFigure 3.3. Optical micrographs of dried CNC-PAAm microbeads (A) observed under reflectedwhite light or (B) with transmitted light between crossed polarizers. Scale bars, 50 μm.Figure 3.4. POM micrograph of a CNC-PAAm chiral nematic microgel that has swelled in water.Scale bar, 20 μm.57A BC DFigure 3.5. Optical micrographs of CAMB-M microgels (after swelling in water) observed (A,C)between crossed polarizers or (B,D) under reflected white light. Scale bars, 50 μm.58Figure 3.6. Orthogonal views of a chiral nematic microsphere obtained by confocal fluorescencemicroscopy. Here the imaging region is 178 μm in width and 130 μm in height. The sample wasstained by immersing the microspheres in an aqueous solution of rhodamine B (about 0.1 wt.%),then redispersed in pure water before imaging. This microsphere was flattened by gravity.59Figure 3.7. A three-dimensional reconstructed confocal fluorescence microscopy image showingthe chiral nematic layers in a microsphere as a series of concentric spherical shells. A topologicaldefect can be observed at the center of this microsphere, which was formed due to the frustrationof the chiral nematic liquid crystalline order in the spherical confinement.60A BC DFigure 3.8. POM images of CNC-PAAm microspheres showing chiral nematic tactoids capturedat different evolution stages. Diameter of microspheres, (A) 85 μm, (B) 136 μm, (C) 142 μm and(D) 141 μm.61A B C DFigure 3.9. Three-dimensional models showing the evolution of chiral nematic liquid crystallinetactoids in spherical confinement.Figure 3.10. A three-dimensional model showing the chiral nematic layers of a liquid crystallinetactoid in spherical confinement, which are organized into a series of concentric shells.623.2.4 Spherical Chiral Nematic Tactoids under Electron MicroscopyAs revealed by SEM observations (Figure 3.11), most of the CNC-PAAm microbeads havea spherical geometry, where the surface patterns may represent the turbulent flow of the liquid inthese aqueous microdroplets during the polymerization process. After smashing (with a hammer)some microbeads flash frozen in liquid nitrogen, tactoids in the center of the cracked microbeadscould be examined by SEM, which have onion-like microstructures of concentric spherical shells(Figure 3.12). These preliminary results are consistent with the polarized optical microscopy andconfocal microscopy observations.Because the helical pitch of the chiral nematic ordering in these microspheres is much largerthan the wavelength of visible light, the chirality of these structures cannot be characterized withcircular dichroism (CD) spectroscopy. Fortunately, the large sizes of CNC mesogens allow theirorientations in the microspheres to be directly observed by electron microscopy. Although theseCNC-PAAm spheres could be briefly cracked in liquid nitrogen, frost immediately forms on thecold fracture surface, softens and smooths it after melting. However, high-quality cross-sectionswith intact morphology could be obtained by cracking these microspheres after embedding themin epoxy resins (Figure 3.13 and Figure 3.14), thereafter the organization of CNC mesogens canbe directly examined by SEM.The chiral nematic layers of the tactoids generally form a spiral structure with a topologicaldefect at the center rather than perfect concentric spherical shells (Figure 3.15). The left-handedchiral nematic ordering of mesogens is apparent at high magnifications (Figure 3.16). As shownby POM observations, outside the liquid crystalline tactoid is the disordered phase, where CNCsare randomly arranged. This is the result of the phase separation of CNC lyotropic liquid crystalsin geometrical confinement.63A BFigure 3.11. SEM micrographs showing the surface morphology of the solid-state CNC-PAAm(CAMB-M) composite microspheres. Scale bars, 20 μm.A BFigure 3.12. SEM micrographs showing the onion-like structure of chiral nematic layers inside aCNC-PAAm microsphere, which was cracked in liquid nitrogen. As frost immediately formed onthe cold fracture surface, softened and smoothed it after melting (due to the hydrophilicity of thepolyacrylamide matrix), the arrangement of individual CNCs cannot be distinguished. Scale bars,(A) 10 μm, (B) 5 μm.64A BFigure 3.13. SEM images showing a fracture surface of an epoxy resin containing CNC-PAAmmicrospheres. Scale bars, (A) 200 μm, (B) 50 μm.Figure 3.14. SEM image showing the cross-section of a microsphere that has been split into twoparts by using an epoxy resin. Scale bar, 10 μm.65Figure 3.15. Cross-sectional SEM micrograph showing the spiral structure of the chiral nematicliquid crystalline tactoid inside a CNC-PAAm microsphere. This tactoid consists of two differentchiral nematic layers. The chiral nematic ordering of the liquid crystalline phase is frustrated dueto the spherical confinement, which results in the formation of a topological defect in the centerof this tactoid. Scale bar, 5 μm.66A BC DFigure 3.16. Cross-sectional SEM images showing the left-handed chiral nematic order of CNCsin a microsphere. Scale bars, (A) 2 μm, (B,C) 500 nm, (D) 200 nm.673.2.5 Tactoids in Geometrical Confinement with Different DimensionsThe average diameter of the microspheres could be easily controlled from 30 μm to 220 μmby changing the amount of surfactant and stirring speed used. A larger amount of surfactant anda higher stirring speed resulted in smaller microspheres, as confirmed by SEM measurements ofthe dried samples (Table 3.1 and Figure 3.17). After swelling into highly transparent microgelsin water, the existence of chiral nematic tactoids in these microspheres could be clearly observedunder POM (Figure 3.18, Figure 3.19, and Figure 3.20 for CAMB-M, CAMB-S, and CAMB-L,respectively).Not surprisingly, although most of the tiny microspheres (about 30 μm in diameter, denotedas CAMB-S) show liquid crystalline tactoids with chiral nematic layers organized into concentricspherical shells, generally they have only one or two half-helical-pitches near the boundary, withan isotropic core in the center (Figure 3.18). On the other side, liquid crystalline tactoids in largemicrospheres (about 220 μm in diameter, named as CAMB-L) were not efficiently integrated dueto the rather long distances between them (Figure 3.19); thus, the multiple discrete tactoids werenot forced to merge into a single one. To obtain single and integrated chiral nematic tactoids, thesize of the microdroplets should be comparable to that of individual tactoids, which correspondsto the situation of CAMB-M (about 125 μm in diameter) (Figure 3.20).According to the literature,66 the geometrical confinement of a liquid in a limited space mayresult in enhanced viscosity and slow relaxation; when the space is sufficiently narrow, the liquidwould be solidified, where shear flows will not be allowed until a critical shear stress is exceeded.This solidification phenomenon occurs especially when the spatial confinement is narrower thanfour particle dimensions thick. Nevertheless, in the present experiments, the sizes of the aqueousmicrodroplets are always significantly larger than the lengths of cellulose nanocrystal mesogens,68therefore the influence of confinements on the viscosity of CNC dispersions should be negligiblein these studies.In the present studies, the liquid crystalline tactoids of CNCs seemed not to be significantlyinfluenced by the confinement of large-size microdroplets, which have diameters about 15 timeslarger than the helical pitch (about 14 micrometers) of the chiral nematic phase formed by CNCs.However, it has been reported that chiral nematic liquid crystals are able to form highly orderedconcentric-spherical-shell structures in droplets with diameters of more than 30 times the helicalpitch.59 Therefore the influence of the geometrical confinement on the structure of chiral nematicliquid crystals may vary depending on their physical properties, such as the relaxation rate of theliquid crystalline mesogens.69A BC DFigure 3.17. SEM micrographs showing the size distributions of completely dried CNC-PAAmmicrospheres. (A,B) CAMB-M, (C) CAMB-L, and (D) CAMB-S. These micrographs were takenat the same magnification. Scale bars, 200 μm.70A BC DFigure 3.18. POM micrographs of CAMB-M microspheres. Typically, the chiral nematic liquidcrystalline tactoids inside a microsphere have been well integrated into a single liquid crystallinecore with a spherical geometry, which is located at the center of the microsphere and surroundedby an isotropic shell. Most of these integrated tactoids have highly-ordered onion-like structuresof multiple concentric spherical shells. These micrographs were taken at the same magnification.Scale bars, 50 μm.71A BC DFigure 3.19. POM micrographs showing CAMB-S microspheres in water, the diameters of thesesmall-size microspheres are about 2 to 3 times the helical pitch of the CNC chiral nematic phase.Typically, the liquid crystalline tactoids in a microsphere have also been well integrated, forminga series of concentric spherical chiral nematic shells near the boundary of the microsphere, whichleaves an isotropic core in the central region. Interestingly, in some cases there may be a discretetactoid left inside the isotropic core. All these micrographs were taken at the same magnification.Scale bars, 50 μm.72A BFigure 3.20. POM images showing CAMB-L microspheres. The chiral nematic liquid crystallinetactoids in these large-sized microspheres were not well integrated during the period of standing,which might be caused by the relatively long distances between them. These POM micrographswere taken at the same magnification. Scale bars, 50 μm.733.3 Mesoporous Silica Microspheres with Chiral Nematic OrderBased on these results, this inverse emulsion method was further extended to the fabricationof mesoporous silica microspheres with chiral nematic ordering. In a typical procedure, an acidicdispersion of CNCs was mixed with the nonionic precursors of polyacrylamide (i.e., acrylamide,a crosslinker, and a photoinitiator) and tetramethyl orthosilicate (TMOS), then homogenized bysonication, and used as the aqueous phase. Other experimental conditions were kept the same asthose used for the preparation of CAMB-M samples. TMOS underwent a sol-gel transition in theacidic environment during the growth of liquid crystalline tactoids in the aqueous microdroplets.After a period of time (9-12 hours), photopolymerization was initiated by UVB irradiation, and acrosslinked polyacrylamide matrix rapidly formed to capture the chiral nematic order of tactoidsin solid microspheres. An ethanolic solution of ammonium hydroxide was then quickly added toaccelerate the condensation of silanol groups of the silica gels inside the microspheres, thereforeforming an additional matrix of silica to support the chiral nematic structures of CNCs, which isa double-matrix method. After complete drying, the microspheres were calcined at 540 °C in airto remove all the organics including CNCs and PAAm, yielding mesoporous silica microsphereswith chiral nematic order.After calcination, the chiral nematic ordering of liquid crystalline tactoids remained intact inmany of the silica microspheres, which could be observed by POM (Figure 3.21). The completeremoval of cellulose nanocrystals, polyacrylamide and other organic compounds was confirmedby IR spectroscopy and elemental analysis. The IR spectrum (Figure 3.22) of CNC-PAAm-silicacomposite microspheres showed characteristic peaks of cellulose (1161 cm−1 for pyranose ringsand 3339 cm−1 for hydroxy groups) and polyacrylamide (1450 cm−1 for methylene bridges, 1617cm−1 for amino groups, and 1656 cm−1 for carbonyl groups);67,68 however, after calcination, only74the Si-O-Si bending and stretching vibration bands (806 cm−1 and 1056 cm−1 peaks) of the silicamatrix remained.69 Elemental analysis of the CNC-PAAm-silica composite microspheres showed8.27 wt.% nitrogen, 31.02 wt.% carbon, and 5.26 wt.% hydrogen, but these three elements wereall lower than the detection limit in the silica microspheres after calcination.For electron microscopy examination, cross-sections of the mesoporous silica microsphereswere obtained by cracking an epoxy resin with the samples embedded inside, and then the chiralnematic structures in these silica microspheres could be directly observed by SEM (Figure 3.23).As revealed by high magnification images, the silica microspheres are composed of grain-shapednanoparticles which adopt a left-handed helical superstructure (Figure 3.23 D).The mesoporous properties of the silica microspheres were evaluated by nitrogen adsorptionand desorption measurements (Figure 3.24). The samples showed a BET specific surface area of257 m2/g, a BJH adsorption cumulative pore volume of 0.993 cm3/g, as well as a BJH adsorptionaverage pore width of 14.6 nm. With high surface area, mesoporosity, good thermal stability, andchiral nematic ordering, these silica microspheres may have potential applications in asymmetriccatalysis or chiral separations.75A BFigure 3.21. POM images of mesoporous silica microspheres. Scale bars, 10 μm.Figure 3.22. IR spectra of CNC-PAAm-SiO2 (blue) and SiO2 microspheres (red curve).76A DB EC FFigure 3.23. SEM images showing the chiral nematic order of a mesoporous silica microsphere.Scale bars, (A) 10 μm, (B) 2 μm, (C,E) 500 nm, (D) 1 μm, (F) 200 nm.77Figure 3.24. Nitrogen adsorption-desorption isotherm (77 K) of mesoporous silica microspheres.The sample mass was 117.3 mg in this measurement. The desorption data are presented as emptycircles.783.4 ConclusionsIn this study, cellulose nanocrystals were allowed to self-assemble in aqueous microdropletsof an inverse emulsion system. Liquid crystalline tactoids with chiral nematic order were formedand further captured in crosslinked polyacrylamide microspheres by in-situ photopolymerization.The chiral nematic structures of CNCs are preserved within the polymer matrix, as characterizedby both optical and electron microscopy. The size of the microspheres can be easily adjusted bychanging the stirring speed and the amount of surfactant used. Furthermore, it was found that, toobtain single, integrated chiral nematic cores with concentric spherical multi-shell structures, thesize of the water microdroplets in the inverse emulsion should be similar to that of discrete CNCtactoids.This method was then extended to the fabrication of silica microspheres with chiral nematicorder, which exhibit high surface area, mesoporosity, and high thermal stability. These materialsmay have potential applications in sensing, optics and chiral separation.In this work, microdroplets of cellulose nanocrystals were generated by an inverse emulsionmethod, which is low-cost and scalable. However, the microspheres obtained in this approach areusually polydispersed. Monodispersed microspheres of cellulose nanocrystals can be synthesizedby using microfluidics devices, and the chiral nematic order in those microspheres could be wellcontrolled by adjusting the concentration of CNCs, the diameters of the microdroplets, and otherexperimental parameters.70−72793.5 Experimental Methods3.5.1 Fabrication of Polymer Microspheres with Chiral Nematic OrderIn a standard experiment, sorbitane monooleate (0.60 g, also known as Span-80, surfactant)and cyclohexane (80 mL, non-polar phase) were placed in a 250 mL single-necked round-bottomflask, which was equipped with a magnetic stir bar and a rubber stopper. The stirring speed wasset at 400 revolutions per minute (rpm), and the system was deoxygenated by bubbling argon ornitrogen through the solution for 15 minutes (via a long stainless-steel needle).Afterward, a homogeneous mixture of a cellulose nanocrystal dispersion (4.0 wt.% in water,6.0 mL), acrylamide (1.2 g, monomer), N,N′-methylenebisacrylamide (120 mg, crosslinker), and2-hydroxy-4′-(2-hydroxyethoxy)-2-methylpropiophenone (6 mg, photoinitiator) which had beensonicated for about 5 minutes was slowly injected into the cyclohexane solution. The system wasdeoxygenated for another 15 minutes, then kept stirring under an argon (or nitrogen) atmospherefor about 9 hours to achieve a stable water-in-oil emulsion, during which time the chiral nematicliquid crystalline tactoids of CNCs were allowed to evolve in the aqueous microspheres, and thesystem was sealed to prevent the evaporation of cyclohexane or water. Then, a UVB light source(wavelength of 300 nm, power of 8 W) was used in the photopolymerization process, and the UVirradiation was applied to the system for about 2 hours.After adding ethanol (150 mL), the CNC-PAAm composite microspheres could be collectedas white powders by filtration. To remove insoluble oligomers from the products, the as-preparedmicrospheres were redispersed in fresh ethanol (50 mL) and sonicated for about 5 minutes. Aftersedimentation of the microspheres (which are heavier than the oligomers), the turbid supernatantsuspension was removed. This washing process was repeated for several cycles until the freshly80added ethanol remained clear after sonication and sedimentation. The cleaned microspheres werecollected by filtration and finally dried in air at 60 °C for 12 hours. In a typical experiment, about1.1 g of microspheres could be obtained.The above experimental conditions give CAMB-M microspheres with an average diameterof 125 μm. For preparation of CAMB-S samples, 1.2 g Span-80 and a stirring speed of 600 rpmwere used to give an average diameter of about 30 μm; the CAMB-L microspheres, which havean average diameter of about 220 μm, were fabricated with 0.4 g Span-80 and a stirring speed of300 rpm (Table 3.1).3.5.2 Preparation of Chiral Nematic Mesoporous Silica MicrospheresThe concentration of an acidic CNC suspension (initial pH = 2.8) was raised from 4.0 wt.%to 6.1 wt.% by heating at about 60 °C in air. Such a concentrated CNC suspension (5.0 mL) wasmixed with acrylamide (600 mg, monomer), N,N′-methylenebisacrylamide (60 mg, crosslinker),2-hydroxy-4′-(2-hydroxyethoxy)-2-methylpropiophenone (3 mg, photoinitiator), and tetramethylorthosilicate (2.0 mL, TMOS, silica precursor), then the mixture was homogenized by sonicationfor 5 minutes, and used as the aqueous phase to form the water-in-oil emulsion system. All otherexperimental conditions were kept exactly the same as those used for the synthesis of CAMB-Mmicrospheres. After UVB irradiation, a mixture of 30.0 mL ammonium hydroxide (28.0 wt.% inwater) and 120.0 mL ethanol was added, and the products were allowed to stand for 2 hours, thenwashed and further dried as previously described. In a typical synthesis, about 1.60 g compositemicrospheres could be obtained after drying. The mesoporous silica microspheres were preparedby calcination in air. An appropriate amount of completely dried CNC-PAAm-Silica compositemicrospheres were first heated to 90 °C for 8 hours, then the oven temperature was increased to81540 °C during a period of 4 hours, and kept at 540 °C for another 4 hours to burn off all organiccomponents. In a representative experiment, 495 mg of the dried CNC-PAAm-Silica compositemicrospheres were calcined to give 211 mg of mesoporous silica microspheres.3.5.3 MaterialsCellulose nanocrystal suspensions (4.0 wt.% in water, pH = 2.8, conductivity = 2.16 mS/cm)were provided by CelluForce, which were obtained from hydrolysis of wood pulp in sulfuric acidusing the literature method. TEM measurements of the cellulose nanocrystals gave dimensions ofabout 20 nm in width and 245 nm in length. Cyclohexane (Anachemia, 99%), Span-80 (sorbitanemonooleate, Croda), acrylamide (Aldrich, 98%), N,N′-methylenebisacrylamide (Aldrich, 99%),2-hydroxy-4′-(2-hydroxyethoxy)-2-methylpropiophenone (Aldrich, 98%), ammonium hydroxide(28.0-30.0 wt.% ammonia in water, Fisher), tetramethyl orthosilicate (TMOS, Aldrich, 98%) andRhodamine B (Aldrich, 95%) were used without further purification.3.5.4 CharacterizationConfocal laser scanning microscopy (CLSM) was performed on a Zeiss LSM 510 confocalmicroscope. Dried CNC-PAAm microspheres were allowed to swell in a 0.1 wt.% Rhodamine Baqueous solution for 1 hour, then collected by filtration and briefly washed with water to removeresidual dye. The stained microspheres (i.e., microgels) were dispersed in pure water again, andplaced onto a glass slide for CLSM characterizations. The samples were excited at 543 nm with aHelium-Neon green laser beam, then a BP 560-615 IR LSM emission filter was used in front ofthe detector. Confocal Z-stack micrographs of the chiral nematic microspheres were acquired at1.0 micron intervals.82Other characterization methods, such as scanning electron microscopy (SEM), transmissionelectron microscopy (TEM) and polarized optical microscopy (POM), are the same as those usedin the previous chapters.83Chapter 4: Size-Selective Exclusion Effects of Liquid Crystalline Tactoids onNanoparticles: Separation by Microscopic Ordered-Disordered Interfaces4.1 IntroductionNanomaterials have great potential for applications in catalysis,73 pharmaceuticals,74 energyharvesting,75 and sensing.76 Many of these applications require the separation of nanoparticles bysize and/or shape. Reported techniques include centrifugation,77 diafiltration,78 electrophoresis,79size exclusion chromatography,80 and field flow fractionation,81 each is applicable to a differentrange of particles, and some depend on the charge of the nanoparticles.82 Thus, the developmentof new methods for separating nanoparticles is an important challenge for this field.Here, the interactions between tactoids and doping nanoparticles were studied. By scanningelectron microscopy (SEM), it was observed that liquid crystalline tactoids have a size-selectiveexclusion effect on foreign nanoparticles, where tactoids exclude particles larger than a thresholdsize. This principle was then successfully applied to the separation of gold nanoparticles by size.Afterwards, the recycling of liquid crystalline phases for multiple-cycle size-selective separationwas investigated with superparamagnetic magnetite (Fe3O4) nanoparticles, where the isolation ofthe small nanoparticles from the cellulose nanocrystal mesogens makes it possible to completelyseparate small particles from large ones.844.2 Size-Selective Exclusion of Nanoparticles by Liquid Crystalline Tactoids4.2.1 Liquid Crystalline Tactoids in Mixtures of Nanorods and NanospheresThese studies are concerned with the structure and evolution of liquid crystalline tactoids inthe presence of doping nanoparticles, where the nanoparticles have widths larger or smaller thanthe average separation distance between adjacent liquid crystal mesogens in the ordered phases.For cellulose nanocrystals used in the following experiments, this distance was reported to rangefrom 25 to 51 nm, depending on the total concentration of the system.83Aqueous dispersions of CNCs (4.0 wt.%) and carboxyl-functionalized (negatively charged)polystyrene nanoparticles (PSNPs, about 10 wt.%, white in color) were homogeneously mixed ina 9:1 volume ratio. The binary mixture of CNCs and large-sized PSNPs (L-PSNPs, diameters of265-280 nm, Figure 4.1) was initially isotropic and white, then chiral nematic tactoids graduallyappeared as birefringent microdroplets with periodic layers that can be observed between crossedpolarizers (Figure 4.2 A). Under reflected white light, the tactoids showed up as dark regions ona bright white background (Figure 4.2 B), which indicates a significantly lower concentration ofL-PSNPs in tactoids than in isotropic phases. Tactoids were also observed in the binary mixturesof CNCs and small-sized PSNPs (S-PSNPs, diameters of 30-57 nm, Figure 4.3) between crossedpolarizers (Figure 4.4 A) and under reflected white light (Figure 4.4 B), where they exhibited awhite color comparable to that of the surrounding isotropic phase, which indicates the existenceof S-PSNPs in tactoids. These results suggest that liquid crystalline tactoids can exclude foreignnanoparticles that are larger than a threshold size (Figure 4.5), while smaller particles can enterthese ordered microdroplets (Figure 4.6).85A BFigure 4.1. SEM images for large polystyrene nanoparticles (denoted as L-PSNPs, diameters of265-280 nm) used in this work. Scale bars, (A) 200 nm, (B) 100 nm.A BFigure 4.2. Optical microscopy images showing chiral nematic liquid crystalline tactoids formedin the presence of L-PSNPs (265 to 280 nm in diameter), which were observed between crossedpolarizers (A) or under reflected white light (B). Scale bars, 50 μm.86A BFigure 4.3. SEM images of small-size polystyrene nanoparticles (denoted as S-PSNPs, diametersof 30-57 nm) used in this work. Scale bars, 50 nm.A BFigure 4.4. Optical microscopy images showing chiral nematic liquid crystalline tactoids formedin a binary mixture of CNCs and S-PSNPs (30-57 nm in diameter), which was observed betweencrossed polarizers (A) or under reflected white light (B). Scale bars, 50 μm.87Figure 4.5. 3D model showing the exclusion effect of a liquid crystalline tactoid (mesogens aredepicted as golden brown rods) on large-sized doping nanoparticles (purple spheres).88Figure 4.6. 3D model showing that small-sized nanoparticles (depicted as red spheres) can entera liquid crystalline tactoid.894.2.2 Exclusion of Large-Sized Nanoparticles by Liquid Crystalline TactoidsFor SEM observations, nonionic precursors (compatible with the lyotropic liquid crystallinephase of CNCs) of polyacrylamide (acrylamide, crosslinker, photoinitiator) to the binary systemsof CNCs and PSNPs. After standing for a period of time (3 to 15 h), in-situ photopolymerizationwas initiated by ultraviolet-B light. A cross-linked polymer matrix rapidly formed to capture andpreserve the fluid microscopic structures. After solidification, tactoids and doping nanoparticlescould be observed at the resolution of individual mesogens by scanning electron microscopy.In a CNC/L-PSNP system, a newly formed tactoid (captured after 3 h growth) was observedas an ordered microdomain with one chiral nematic layer and an elliptical boundary (Figure 4.7).Inside this tactoid is the liquid crystalline phase, in which CNCs adopt left-handed chiral nematicorder, but no doping nanoparticles were observed (Figure 4.8). In the isotropic phase outside theboundary of this tactoid (Figure 4.9), in which the CNC mesogens are disordered, the L-PSNPscould be distinguished as either hemispherical cavities or partly stretched spheres with long tailsoutside the cross-section surface (Figure 4.10).Here the hemispherical cavities and long-tail-shaped protrusions are left by the polystyrenenanoparticles when the polyacrylamide matrix is cracked. In a typical situation, when a piece ofpolyacrylamide plastic block containing cellulose nanocrystals and polystyrene spheres is beingcracked, the polystyrene nanospheres located along the fracture become stretched (they initiallyadhere to both sides) until they break free from one of the two side. This process leaves an emptyhemispherical cavity on one surface and a protruding tail (i.e., a stretched polystyrene sphere) onthe other.The phase behavior of rod/sphere binary mixtures has been modeled by the Onsager theoryin previous literature,84 in which the phase transitions are driven by the tendency to minimize the90excluded volume of the dispersed particles. This principle explains why newly formed tactoids inCNC/L-PSNP mixtures are free from L-PSNPs (Figure 4.11), since the presence of large dopingnanoparticles in the tactoid will increase the excluded volume of the mesogens by disrupting theliquid crystalline order.Furthermore, the L-PSNPs were scarcely observed in large tactoids with five or more chiralnematic layers (Figure 4.12). These results indicate that tactoids can exclude large-sized foreignnanoparticles. This can be explained by the coalescence mechanism for the evolution of tactoids,in which multiple small tactoids that are free from foreign nanoparticles merge together to form alarger tactoid (Figure 4.13).91Figure 4.7. SEM image of a tactoid formed in a CNC/L-PSNP mixture. Scale bar, 2 μm.A BFigure 4.8. SEM images showing the inside area of a newly formed liquid crystalline tactoid in aCNC/L-PSNP binary mixture. Scale bars, (A) 1 μm, (B) 500 nm.92A DB EC FFigure 4.9. SEM images of (A) the upper left, (B,C) the left, and (D-F) the upper boundaries ofa newly emergent chiral nematic liquid crystalline tactoid in a CNC/L-PSNP mixture. Scale bars,(A,C,E) 500 nm, (B,D) 1 μm, (F) 200 nm.93A BFigure 4.10. SEM images showing L-PSNPs in disordered phases. Scale bars, 200 nm.Figure 4.11. 3D model for the emergence of a small tactoid in a CNC/L-PSNP mixture.94A BC DE FFigure 4.12. SEM micrographs of a liquid crystalline tactoid with about 10 chiral nematic layersformed in a CNC/L-PSNP mixture (A), as well as the boundary (B-D) and chiral nematic layersof this tactoid (E,F). Scale bars, (A) 5 μm, (B,C) 1 μm, (D,E) 500 nm, (F) 200 nm.95Figure 4.13. 3D model showing the coalescence of two liquid crystalline tactoids in the presenceof large-sized foreign nanoparticles.964.2.3 Coalescence of Tactoids Traps Nanoparticles in Topological DefectsAs revealed by SEM, large-sized doping nanoparticles could be trapped at the contact pointbetween two coalescing liquid crystalline tactoids (Figure 4.14 A and Figure 4.15), or inside theisotropic region surrounded by several merging tactoids (Figure 4.14 B and Figure 4.16). Due tothe higher density of tactoids than the isotropic phase, eventually numerous tactoids coalesce intoa macroscopically continuous liquid crystalline phase at the bottom of the suspension, while theisotropic regions enclosed between tactoids are transformed into topological defects (e.g., foldedor dislocated CNC chiral nematic layers) with the large-sized doping nanoparticles sealed inside(Figure 4.14 C and Figure 4.17).A B CFigure 4.14. (A) A few large-sized doping nanoparticles (depicted in purple) could be trapped atthe contact point between two coalescing tactoids. (B) A lot of large doping nanoparticles can besealed within the broad isotropic region surrounded by several merging tactoids. (C) Topologicaldefects generated during the coalescence of tactoids would eventually remain in the macroscopicliquid crystalline phases with a large number of doping nanoparticles sealed inside.97A BC DFigure 4.15. (A) Cross-sectional SEM micrograph showing the initiation of coalescence betweentwo liquid crystalline tactoids formed in a CNC/L-PSNP binary mixture. The boundaries of thesetwo tactoids can exclude the large-sized doping nanoparticles (B), but several nanoparticles weretrapped at the contact point between them (C,D), which would be sealed in the liquid crystallinephase. Scale bars, (A) 5 μm, (B,C) 1 μm, (D) 200 nm.98A DB EC FFigure 4.16. SEM images showing the contact between two tactoids in a CNC/L-PSNP mixture(A), the exclusion of L-PSNPs by tactoid boundaries (B,C), and L-PSNPs trapped in the contactregion (D-F). Scale bars, (A) 3 μm, (B,E) 500 nm, (C,D) 1 μm, (F) 200 nm.99A BC DE FFigure 4.17. Cross-sectional SEM images of a CNC/L-PSNP mixture showing large-size dopingnanoparticles trapped in the topological defects of the macroscopic liquid crystalline phase (A-Band C-F). Scale bars, (A,D) 1 μm, (B,E) 500 nm, (C) 2 μm, (F) 200 nm.1004.2.4 Existence of Small-Sized Nanoparticles in Liquid Crystalline TactoidsIn CNC/S-PSNP binary mixtures, the small-sized doping nanoparticles could be observed ashemispherical cavities and partly stretched spheres both in the isotropic phases and inside liquidcrystalline tactoids (Figure 4.18). These small doping nanoparticles already exist in the initiallyformed "baby" tactoids (Figure 4.19), which could be explained by the excluded volume theory,as small doping nanoparticles can exist in the gaps in the liquid crystal lattice of a tactoid withoutdisrupting its overall liquid crystalline order.In addition to the thermodynamic model based on the excluded volume theory, it would alsobe possible to explain the size-selective exclusion effect of tactoids using a kinetic approach. Theuniformly aligned nanorod mesogens at the boundary of a liquid crystalline tactoid may work asa microscopic filter which can size-selectively block foreign nanoparticles larger than the gaps inthe liquid crystal lattice, while particles smaller than these gaps can pass through the boundary ofthis tactoid and therefore enter the ordered phase (Figure 4.20 A).As demonstrated by Bergström and Lagerwall,51 the interaction between CNCs in the chiralnematic phase is repulsive for separation distances up to 30 nm and attractive at longer distances.According to this data, when a large foreign nanoparticle hits the boundary of a liquid crystallinetactoid from the outside (i.e., from the side of the disordered phase), the ordered lattice of liquidcrystal mesogens in front of the invading nanoparticle will be compressed in the normal directionof the contact surface (the interface between the nanoparticle and the tactoid) and stretched in thetangential directions, which results in normal repulsive forces and tangential attractive forces inthe liquid crystalline lattice (Figure 4.20 B). Therefore, the tactoid boundary repels the large-sizeinvading nanoparticles.101A DB EC FFigure 4.18. SEM images showing the existence of small-sized doping nanoparticles (diametersof 30-57 nm) as about 50 nm hemispherical cavities in tactoids formed in a CNC/S-PSNP binarymixture (A-C and D-F). Scale bars, (A) 3 μm, (B,C,F) 100 nm, (D) 2 μm, (E) 200 nm.102A BC DE FFigure 4.19. (A-E) SEM images showing hemispherical cavities created by S-PSNPs in a "baby"tactoid in a CNC/S-PSNP mixture. (F) Image showing the coexistence of S-PSNPs and CNCs inthe disordered phase. Scale bars, (A) 2 μm, (B) 200 nm, (C,D,F) 100 nm, (E) 50 nm.103A BFigure 4.20. 3D models illustrating (A) the size-selective permeability of an ordered-disorderedinterface to foreign nanoparticles, and (B) the exclusive force from this interface on an invadingnanoparticle with a width larger than the gaps in the liquid crystalline lattice.1044.2.5 Size-Selective Particle Permeability of Ordered-Disordered InterfacesAlthough it was difficult to directly measure the interaction forces between the boundary ofa liquid crystalline tactoid and invading nanoparticles, similar results might be obtained from thestudies of macroscopic ordered-disordered interfaces since they have the same microstructures astactoid boundaries.In a preliminary experiment, an aqueous PSNP dispersion was added on top of the isotropicphase of a phase-separated CNC suspension. The PSNPs (with a white color) gradually diffuseddownwards through the isotropic layer to the ordered-disordered interface. This interface blockedthe diffusion of L-PSNPs toward the liquid crystalline phase (Figure 4.21), as further confirmedby cross-sectional SEM observations after capturing the system in a cross-linked polymer matrix(Figure 4.22 and Figure 4.23). However, the small-size foreign nanoparticles passed through theordered-disordered interface, and thus entered the liquid crystalline phase (Figure 4.24).10 min 60 min 150 min 750 minFigure 4.21. Photographs showing that the downward diffusion of L-PSNPs was stopped by theordered-disordered interface of a phase-separated CNC dispersion.105A BC DFigure 4.22. (A) Cross-sectional SEM micrograph showing the ordered-disordered interface of aphase-separated CNC dispersion, which is the upper boundary of the macroscopically continuouschiral nematic liquid crystalline phase. (B-D) Expanded views near the interface showing that thedownward diffusion of the large-sized polystyrene nanoparticles (diameters of 265-280 nm) wasdramatically stopped by the upper boundary of the liquid crystalline phase. Scale bars, (A) 2 μm,(B) 1 μm, (C,D) 500 nm.106Figure 4.23. High-resolution cross-sectional SEM image showing the microscopic structures ofthe interface between the disordered phase (upper) and the chiral nematic liquid crystalline phase(lower) in a completely phase-separated cellulose nanocrystal dispersion, where the organizationof individual liquid crystal mesogens can be directly observed. This ordered-disordered interfacedramatically blocked the downward diffusion of the large-sized doping nanoparticles (L-PSNPs),which could be observed in the upper disordered phase as either hemispherical cavities or partlystretched nanospheres. Scale bar, 200 nm.107A BC DFigure 4.24. SEM images showing that when small-size polystyrene nanoparticles (diameters of30 to 57 nm) were added on the top of a phase-separated cellulose nanocrystal suspension, thesenanoparticles were able to pass through the isotropic-anisotropic interface by diffusion, resultingin hemispherical cavities (about 50 nm in diameter) in the chiral nematic liquid crystalline phase.Scale bars, (A) 500 nm, (B-D) 100 nm.1084.3 Size-Selective Separation of Nanoparticles with Lyotropic Liquid Crystals4.3.1 Separation of Plasmonic Gold Nanoparticles by SizeTo generalize the size-selective separation ability of tactoids, the lyotropic liquid crystals ofCNCs were homogeneously mixed with gold nanoparticles of different sizes (average diametersof 20 and 70 nm). After complete phase separation, size distributions of gold nanoparticles in theisotropic and liquid crystalline phases were measured by ultraviolet-visible (UV-Vis) absorptionspectroscopy and TEM. When using small-sized gold nanoparticles (denoted as S-AuNPs, about20 nm in diameter, absorption peak at 523 nm in water, red colored), the liquid crystalline phaseshowed a pink-red color and a plasmon resonance peak at 521 nm, which is slightly blue-shiftedas compared to the 526 nm absorption peak of the isotropic phase (Figure 4.25 and Figure 4.26).When using large-sized gold nanoparticles (denoted as L-AuNPs, average diameter of about70 nm, absorption peak at 557 nm in water, blue-purple colored), the liquid crystalline phase wascolorless, and its UV-Vis absorption spectrum showed no plasmon resonance peaks (Figure 4.25and Figure 4.27).Finally, when the mixture of large and small sized gold nanoparticles (L&S-AuNPs, with anabsorption peak at 543 nm in water, purple-red colored) was mixed with the aqueous suspensionof CNCs, after complete phase separation, the liquid crystalline phase exhibited a pink-red colorand a plasmon resonance peak at 524 nm (Figure 4.25 and Figure 4.28), indicating a significantenrichment of the small-sized gold nanoparticles in this ordered phase as the plasmon resonancepeak is very close to that of pure S-AuNPs (at 523 nm). On the other side, the absorption peak ofthe isotropic phase at 547 nm is red-shifted by 4 nm as compared to the UV-Vis spectrum of theoriginal L&S-AuNP mixture, probably due to a higher proportion of L-AuNPs. The presence of109CNCs would not change the UV-Vis absorption peak position of the gold nanoparticles as CNCshave almost no absorption in the range from 500 nm to 600 nm (Figure 4.29). The difference inthe size distribution of the doping nanoparticles in the ordered and disordered phases was furtherdirectly confirmed by transmission electron microscopy (Figure 4.30 and Figure 4.31).As demonstrated in the present experiments, liquid crystalline tactoids work as microscopicseparators that can size-selectively collect small foreign nanoparticles from the disordered phasebased on the widths of the gaps in the liquid crystal lattice. Eventually, all the tactoids will settleto the bottom of the dispersion and coalesce into a macroscopically continuous liquid crystallinephase with a significantly higher ratio of the small nanoparticles (Figure 4.32).20 nm AuNPs 20&70 nm AuNPs 70 nm AuNPsFigure 4.25. Photographs (with transmitted white light) for phase-separated CNC-AuNP binarymixtures.110Figure 4.26. UV-Vis absorption of a phase-separated CNC/S-AuNP binary mixture.Figure 4.27. UV-Vis absorption of a phase-separated CNC/L-AuNP binary mixture.111Figure 4.28. UV-Vis absorption spectra of the isotropic (Iso) and liquid crystalline (LC) phasesof a phase-separated CNC/L&S-AuNP ternary mixture.A BFigure 4.29. (A) UV-Vis absorption spectra of a phase-separated CNC dispersion, which exhibitno absorption peaks between 400 nm and 800 nm. (B) Additional UV-Vis absorption spectra ofthe small and large sized gold nanoparticles used in this study.112A BC DFigure 4.30. (A,B) TEM micrographs showing the coexistence of the large and small sized goldnanoparticles in the disordered phase of a completely phase-separated CNC/L&S-AuNP ternarymixture. (C,D) Inside the liquid crystalline phase, only the small-sized gold nanoparticles couldbe observed. Scale bars, (A,B) 100 nm, (C,D) 200 nm.113Figure 4.31. Particle size distributions of AuNPs in the isotropic (Iso) and liquid crystalline (LC)phases of a phase-separated CNC/L&S-AuNP ternary mixture, measured by TEM.Figure 4.32. A depiction showingthe size-selective collection of thesmall foreign nanoparticles (smallnanoparticles in red, large particlesin purple) by chiral nematic liquidcrystalline tactoids of CNCs. Dueto the higher density, tactoids willgradually settle to the bottom partof the dispersion and coalesce intoa macroscopic ordered phase witha significantly higher ratio of smallnanoparticles.1144.3.2 Multiple-Cycle Size-Selective Separation of Magnetic NanoparticlesThe separation medium of liquid crystals could be recycled for multiple-cycle separation byisolating the small nanoparticles from the ordered phase after equilibration; in the next separationcycle, the regenerated liquid crystalline phase is homogeneously mixed with the isotropic phase,then the dispersion phase-separates again. It was found that when small-sized superparamagneticmagnetite nanoparticles (denoted as S-MNPs, about 8.7 nm in diameter) were mixed with CNCs,they could exist in the liquid crystalline phase after phase separation, as indicated by the yellowcolor observed with both transmitted (Figure 4.33) and reflected white light (Figure 4.34). Theconcentration of S-MNPs in the liquid crystalline phase was about 30% to 33% of that inside thedisordered phase as determined by UV-Vis absorption spectroscopy (Figure 4.35 A), where theS-MNPs had been magnetically isolated from the dispersions and then redispersed in pure waterto eliminate the influence of liquid crystal mesogens.On the other hand, when large-sized superparamagnetic magnetite nanoparticles (denoted asL-MNPs, about 107 nm in diameter) were used to dope CNCs, after phase separation, the liquidcrystalline phase was colorless (Figure 4.33), with negligible absorption in the ultraviolet-visibleregion (Figure 4.35 B), indicating the absence of L-MNPs in this phase. In the case of a mixtureof large and small sized magnetite nanoparticles (denoted as L&S-MNPs), the liquid crystallinephase exhibited a yellow color, similar to when S-MNPs were used (Figure 4.33).Furthermore, recycling of the separation medium (the lyotropic liquid crystalline mesogens)was investigated using a CNC/L&S-MNP ternary mixture (Figure 4.36). After phase separation,the liquid crystalline phase was extracted into a separate vial, then the MNPs were magneticallyisolated from the liquid; the regenerated cellulose nanocrystals were reused in the next separationcycle to keep separating small-sized MNPs from large ones. The concentration of S-MNPs in the115liquid crystalline phase (also in the whole system) decreases after each separation cycle, resultingin a lighter brown color (Figure 4.37). As revealed by TEM observations, the initial L&S-MNPmixture contains about 98.5% S-MNPs and 1.5% L-MNPs by number (Figure 4.36 A). Either inthe 1st or in the 9th separation cycles, the liquid crystalline phase contains almost only S-MNPs(Figure 4.36 B,C), confirming the size-selectivity. Moreover, after the 9th separation cycle, thedisordered phase contains about 38% L-MNPs and 62% S-MNPs by number (Figure 4.36 D andFigure 4.38), which indicates L-MNPs are dominant in mass as they are about 1000 times largerthan S-MNPs by volume. These results suggest that it should be possible to completely separatesmall-sized nanoparticles from large ones by multiple-cycle separations.S-MNPs L&S-MNPs L-MNPsFigure 4.33. Photographs showing phase-separated dispersions of cellulose nanocrystals mixedwith 8.7 nm (left), 8.7 nm & 107 nm (middle), and 107 nm (right) superparamagnetic magnetitenanoparticles, taken with transmitted white light.116S-MNPs L&S-MNPs L-MNPsFigure 4.34. Photographs showing phase-separated dispersions of cellulose nanocrystals mixedwith 8.7 nm (left), 8.7 nm & 107 nm (middle), and 107 nm (right) superparamagnetic magnetitenanoparticles. The photos were taken with reflected white light.A BFigure 4.35. UV-Vis absorption spectra for phase-separated CNC/S-MNP (A) and CNC/L-MNP(B) binary mixtures.117A BC DFigure 4.36. TEM images showing the changes in the size distribution of magnetic nanoparticlesduring the multiple-cycle size-selective separation process, which was conducted with a ternarymixture of S-MNPs, L-MNPs, and CNCs. (A) The initial mixture of S-MNPs and L-MNPs usedin the multicycle separation. (B) MNPs in the liquid crystalline phase of the 1st separation cycle.(C) MNPs isolated from the liquid crystalline phase of the 9th separation cycle. (D) MNPs in theisotropic phase of the 9th separation cycle. All the images were taken at the same magnification.Scale bars, (A-D) 100 nm.1181st 3rd 5th 7th 9thFigure 4.37. Photographs showing the phase separations of a CNC/L&S-MNP ternary mixture inthe 1st, 3rd, 5th, 7th, and 9th separation cycles. The photographs were taken with reflected whitelight.Figure 4.38. The size distributions of MNPs in the initial mixture of L-MNPs and S-MNPs (redcolor), in the liquid crystalline phase of the 1st (green) and the 9th (cyan) separation cycles, andin the isotropic phase of the 9th separation cycle (blue).1194.4 ConclusionsIn this study, the structures and evolution of tactoids in the presence of doping nanoparticleswere investigated by optical and electron microscopy, where the size-selective exclusion effectsof liquid crystalline tactoids on foreign nanoparticles were observed. Nanoparticles smaller thanthe gaps in the liquid crystal lattice are able to enter tactoids, while larger particles are excludedfrom these ordered microdroplets. Using tactoid-mediated phase separation process of lyotropicliquid crystals, polydisperse gold nanoparticles were separated by size. Afterwards, the recyclingof liquid crystals for multicycle separation was demonstrated with superparamagnetic magnetitenanoparticles. Interestingly, in each case, the smaller nanoparticles are partitioned into the lowerphase, while the larger, heavier particles remain in the top, isotropic phase. Thus, this separationprocess proceeds counter to gravity. These new insights into the structures and behavior of CNCtactoids may provide an explanation for the enrichment of nanoparticles in the isotropic phase aswell as the aggregation of nanoparticles in the topological defects of liquid crystals.85 Moreover,the size-selective uptake of nanoparticles by liquid crystalline tactoids may be the basis of a newseparation method for nanomaterials.1204.5 Experimental Methods4.5.1 Photopolymerization of Liquid Crystals with Doping NanoparticlesIn a representative experiment, acrylamide (1.0 g), N,N′-methylenebisacrylamide (100 mg),and 2-hydroxy-4′-(2-hydroxyethoxy)-2-methylpropiophenone (5 mg) were added to an aqueousdispersion of cellulose nanocrystals (volume of 9.0 mL, 4 wt.%), then an aqueous suspension ofeither L-PSNPs or S-PSNPs (1.0 mL, 10 wt.%) was added, and the mixture was homogenized bysonication in cold water for 10 minutes. After standing in the dark for a prescribed period of time(3 to 15 hours), the system was irradiated with ultraviolet light (wavelength of 300 nm, power of8 W) for 10 to 20 minutes to form a crosslinked polyacrylamide matrix, which was further driedby heating in air at 60 °C for 12 hours. Samples for cross-sectional scanning electron microscopyobservations can be obtained by cracking the completely dried polymer matrix into small pieceswith a hammer. Samples for optical microscopy were prepared by homogeneously mixing CNCand PSNP dispersions in a 9:1 volume ratio, then the binary mixture was observed between twoparallel glass slides that are separated by 0.087 mm thick spacers.4.5.2 Diffusion of Nanoparticles through Ordered-Disordered InterfacesAn aqueous dispersion of cellulose nanocrystals (2.0 mL, 4 wt.%) was allowed to stand in asealed vial (inner diameter of 12.60 mm) until complete phase separation, afterwards, an aqueoussuspension of either L-PSNPs or S-PSNPs (volume of 1.0 mL) was added on top of the isotropicphase to form a metastable system consisting of three different liquid layers (from top to bottom:a white opalescent PSNP layer, a clear isotropic layer, and a birefringent liquid crystalline layer).Then PSNPs in the top layer gradually diffused downwards during the following 12 hours, where121the process was photographed at different time intervals. For SEM observations, polyacrylamideprecursors (ratio as described above) were added to the CNC and PSNP mixtures before startingthe experiments, after the diffusion of the PSNPs had reached equilibrium (here the existence ofnon-ionic polymer precursors does not change the diffusion of either L-PSNPs or S-PSNPs), thesystem was irradiated with ultraviolet light to rapidly form a crosslinked polymer matrix, whichwas then taken out from the vial and dried for scanning electron microscopy analysis.4.5.3 Size-Selective Separation of Gold Nanoparticles by Liquid CrystalsAn aqueous dispersion containing CNCs (4.0 wt.%) and gold nanoparticles (0.05 wt.%) washomogenized by sonication, then allowed to stand in a separatory funnel until macroscopic phaseseparation, where the isotropic and liquid crystalline phases were separately collected in differentcontainers. A separated phase could be diluted to 5 times its volume with pure water for UV-Visabsorption spectroscopy, or to 50 times for TEM observations.4.5.4 Recycling of Liquid Crystalline Phases for Multicycle SeparationsAn aqueous dispersion of CNCs (9.0 mL, 4 wt.%, pH = 6.4) was combined with dispersionsof citrate-stabilized L-MNPs (0.2 mL, 2 wt.%) and S-MNPs (0.1 mL, 5 wt.%), this mixture wassealed in a vial (inner diameter of 16.50 mm) and thoroughly homogenized by sonication. Aftercomplete phase separation, the liquid crystalline phase was extracted from the bottom of the vialusing a syringe equipped with a long needle, and transferred into a clean vial. The paramagneticmagnetite nanoparticles in this phase (mainly S-MNPs) were magnetically isolated, and then theregenerated cellulose nanocrystals were recombined with the isotropic phase in the initial vial tostart the next separation cycle. The isolated MNPs could be redispersed in pure water to 5 times122the original volume for UV-Vis absorption spectroscopy measurements, or to 50 times for TEMobservations.4.5.5 MaterialsCellulose nanocrystal suspensions (4.2 wt.% in water, pH = 2.1, conductivity = 2.1 mS/cm)were provided by CelluForce, which were obtained by hydrolysis of wood pulp in sulfuric acid.The large and small sized polystyrene nanoparticles were synthesized according to literaturemethods using a mixture of methacrylic acid and styrene (1:9 v/v) as monomers.86,87 The PSNPswere purified by dialysis in pure water to remove electrolytes, and prepared as 10 wt.% aqueousdispersions. The gold nanoparticles were synthesized by reducing HAuCl4 with sodium citrate athigh temperature following a literature method.88 Citrate-stabilized magnetite nanoparticles weresynthesized using literature methods.89,90 Both the small and large sized magnetite nanoparticlesare superparamagnetic. The large-sized magnetite nanoparticles are composed of numerous smallmagnetite nanocrystals with widths of about 5 to 10 nm, which endowed the large nanoparticleswith superparamagnetism.90Polyacrylamide precursors are the same as those used in the previous experiments. Styrene(Aldrich, 99%), methacrylic acid (Aldrich, 99%), potassium persulfate (Fisher, 99.9%), sodiumdodecyl sulfate (Aldrich, 99%), iron (II) chloride tetrahydrate (Aldrich, 98%), iron (III) chloridehexahydrate (Aldrich, 99%), tetrachloroauric (III) acid (Aldrich, 99.9%), sodium citrate tribasicdihydrate (Aldrich, 99%), and citric acid anhydrous (Fisher, 99.8%) were used as received.1234.5.6 CharacterizationThe UV-Vis spectra were recorded on a Varian Cary 5000 UV-Vis-NIR Spectrophotometer.Other characterization methods are the same as those used in the previous chapters.124Chapter 5: Movement and Orientation Control of Liquid Crystalline Tactoidsin Competitive Acceleration Fields: Phase Separations Beyond Gravity5.1 IntroductionIn many lyotropic systems, the phase transitions are mediated by liquid crystalline tactoids,where they spontaneously emerge from disordered phases, grow larger by coalescence (or othermechanisms), and eventually merge into macroscopically continuous ordered phases.16 From theperspective of tactoids, there are two primary driving forces for phase separation processes: theordered-disordered interfacial tension and gravity. The interfacial tension tends to minimize theoverall interfacial energy of the boundary between ordered and disordered phases, promoting thecoalescence of discrete tactoids into continuous macroscopic liquid crystalline regions. Gravity,the other driving force, leads to the unidirectional movement of all tactoids to the bottom of thedispersion as tactoids have a higher density than isotropic phases due to the ordered arrangementof mesogens. Therefore, phase separation generally results in the formation of a disordered phaseabove a macroscopic liquid crystalline phase.Although controlling the orientation of tactoids and macroscopic liquid crystalline phases inlyotropic systems could be important for the development of new materials and devices, tactoidsare difficult to manipulate as they are similar in chemical composition and physical properties todisordered phases. In this study, a new approach is demonstrated, which allows control over themovement and orientation of discrete tactoids and the resulting macroscopic ordered phases. Thekey to this approach is the discovery that disordered phases could be endowed with significantlyhigher volume magnetic susceptibility (χm) than liquid crystalline tactoids based on the exclusion125effects of tactoids on paramagnetic magnetite doping nanoparticles (Figure 5.1). When applyingan external gradient magnetic field, the disordered phases experience unidirectional acceleration(driven by magnetic body forces) parallel to the magnetic field and move to high magnetic fieldregions; meanwhile, since liquid crystalline tactoids are discrete microdroplets with much lowervolume magnetic susceptibility (due to the exclusion effects of tactoids on doping nanoparticles),they will be accelerated by magnetic buoyancy forces from the surrounding continuous isotropicphases and move in the opposite direction to low magnetic field regions. Moreover, the tactoidsare simultaneously oriented with the nematic layers parallel to the gradient magnetic field, whichis driven by the shear forces from the isotropic phase.52 As tactoids are only slightly denser thanisotropic phases,53 even weak gradient magnetic fields of several hundred Gauss/cm can overridethe gravitational effects, enabling control of the phase separation configuration (i.e., the relativepositions of the ordered and disordered phases) and rate, as well as the orientation of the directorfields in both discrete tactoids and continuous liquid crystalline phases, by changing the directionand strength of the gradient magnetic field.As shown in the following experiments, the phase separation of a lyotropic system generallyresults in the formation of a disordered phase above a liquid crystalline phase when gravity is theonly macroscopic driving force for the movement of tactoids. However, in the presence of otheracceleration fields, the effects of gravity to control the direction and rate of phase separation maybe overridden. In the case of superparamagnetic nanoparticles that are excluded from the tactoidsand thus preferentially concentrated in the isotropic region, the isotropic phase could be attractedto the bottom by a gradient magnetic field despite its lower density, therefore reversing the phaseseparation observed under normal conditions.126Figure 5.1. 3D Model showing the exclusion effects of liquid crystalline tactoids (mesogens aredepicted as golden-brown rods) on superparamagnetic magnetite doping nanoparticles (depictedas black spheres), thus the disordered phase has significantly higher magnetic susceptibility thanliquid crystalline tactoids. However, the tactoids have a higher density than isotropic phases dueto the ordered arrangements of mesogens.1275.2 Magnetic Buoyancy Forces on Tactoids with Lower Magnetic SusceptibilityRegarding lyotropic liquid crystals that have been doped with paramagnetic nanoparticles asferrofluids, the magnetic body force per unit volume can be calculated based on the Kelvin law:91dFm = μ0 (M∙∇)H dV = μ0 χm (H∙∇)H dVHere Fm is the Kelvin magnetic body force, μ0 is the vacuum permeability (N/A2), M is themagnetization,∇ is the Nabla operator, χm is the volume magnetic susceptibility of the material,H is the magnetic field (A/m), and dV is the volume element.A discrete liquid crystalline tactoid in a continuous isotropic phase mainly experiences fourkinds of external driving forces that can affect its movement direction and velocity (here despitethe viscous resistance from the isotropic phase since it is not a primary driving force), which arethe weight (Fg) of the tactoid, the gravitational buoyancy force (Fbg) and the magnetic buoyancyforce (Fbm) both exerted by the continuous disordered phase surrounding this tactoid (for an idealcase where there is only one liquid crystalline tactoid in an infinite continuous disordered phase,the continuous phase could be considered as a quasi-static phase that does not move), as well asthe magnetic force (Fm) from the external gradient magnetic field (Figure 5.2):dFg = ρ(LC) g dVLCdFbg = − ρ(Iso) g dVLCdFm = μ0 χm(LC) (H∙∇)H dVLCdFbm = − μ0 χm(Iso) (H∙∇)H dVLCWhere ρ(LC) and ρ(Iso) are the densities of the liquid crystalline (LC) tactoid and isotropic (Iso)phases; g is the standard gravitational acceleration; χm(LC) is the volume magnetic susceptibility ofthe liquid crystalline tactoid, χm(Iso) represents the volume magnetic susceptibility of the isotropic128phase; and dVLC is the volume element of the tactoid. Generally, there are ρ(LC) > ρ(Iso) due to theordered arrangements of mesogens in tactoids and χm(Iso) >> χm(LC) caused by the exclusion effectsof tactoids on paramagnetic doping nanoparticles (i.e., the nanoparticle concentration in tactoidsis lower than that in the isotropic phases).The total force on a liquid crystalline tactoid FLC = Fg + Fbg + Fm + Fbm is the vector sum ofthe four external forces affecting its movement direction:dFLC = [ρ(LC) − ρ(Iso)] g dVLC − μ0 [χm(Iso) − χm(LC)] (H∙∇)H dVLCor, in another form:dFLC = [ρ(LC) − ρ(Iso)] g dVLC + μ0 [χm(Iso) − χm(LC)] [− (H∙∇)H] dVLCHerein, the vector [− (H∙∇)H] is in the opposite direction of the magnetic body force, awayfrom the magnetic pole to low magnetic field regions. This equation reveals that in the presenceof superparamagnetic doping nanoparticles, the movement direction of liquid crystalline tactoidsis determined by the competition between the acceleration forces resulting from gravitational andmagnetic fields:ΣFgrav = [ρ(LC) − ρ(Iso)] g VtactoidΣFmag = μ0 [χm(Iso) − χm(LC)] [− (H∙∇)H] VtactoidNote that the direction and magnitude of ΣFmag could be controlled by changing the gradientmagnetic field and the doping concentration of paramagnetic nanoparticles, but ΣFgrav is a nearlyconstant vector always in the vertical direction as [ρ(LC) − ρ(Iso)] will not be significantly affectedby trace amounts of doping nanoparticles.129Figure 5.2. When a paramagnetic-nanoparticle-doped lyotropic system is subjected to a gradientmagnetic field (H), the movement direction and velocity of a discrete liquid crystalline tactoid isdetermined by the vector sum of four external forces: the weight (Fg) in the vertically downwarddirection, the magnetic body force (Fm) along the gradient magnetic field to the high field region,the magnetic buoyancy force (Fbm) along the gradient magnetic field to the low field region (thisforce is exerted by the continuous isotropic phase surrounding the tactoid), and the gravitationalbuoyancy force (Fbg) in the vertically upward direction exerted by the continuous isotropic phaseas well. The unidirectional movement of tactoids (which is mainly driven by Fbm) results in shearforces, by which the chiral nematic layers are oriented parallel to the magnetic field.1305.3 Unidirectional Phase Separations: Acceleration by Gradient Magnetic Fields5.3.1 Lyotropic Liquid Crystals Doped with Paramagnetic NanoparticlesExperiments in this study were conducted with the chiral nematic liquid crystals formed bycellulose nanocrystals in aqueous dispersions, where tactoids appear as ellipsoidal microdropletswith periodically spaced birefringent layers. Citrate-coated superparamagnetic magnetite (Fe3O4)nanoparticles (MNPs) with diameters of 15-20 nm were used to dope the lyotropic liquid crystalsof CNCs (Figure 5.3), forming homogeneous aqueous dispersions of CNCs (4.0 wt.%, pH = 6.4)and MNPs (typically 140.5 ppm by weight).Without external magnetic fields, the CNC-MNP mixture phase separates into a birefringentliquid crystalline phase below a clear disordered phase with the relative positions determined bygravity (Figure 5.4), which is identical to pure CNCs (Figure 5.5).When a static gradient magnetic field (about 1050 Gauss/cm) is applied to the binary systemof CNCs and MNPs, phase separation occurs along the direction of the magnetic field, where themacroscopic isotropic phase forms in the high-magnetic-field region with the ordered-disorderedinterface nearly perpendicular to the magnetic field (Figure 5.6); in a vertically oriented gradientmagnetic field, the isotropic phase could even form below the liquid crystalline phase which hasa higher density, indicating that ΣFmag > ΣFgrav and therefore ΣFmag is the main driving force thatdetermines the configuration of phase separation (Figure 5.7 and Figure 5.8).131A BC DFigure 5.3. TEM micrographs of the superparamagnetic magnetite (Fe3O4) nanoparticles used inthis study. Scale bars, (A,B) 30 nm, (C,D) 20 nm.1320 minutes 10 minutes 60 minutes 90 minutes2 hours 3 hours 4 hours 5 hours6 hours 7 hours 8 hours 9 hours10 hours 11 hours 12 hoursFigure 5.4. When only affected by the gravity, a CNC-MNP binary mixture phase-separates intoa liquid crystalline phase (exhibiting higher brightness between two crossed polarizers) below adisordered phase over several hours. The CNC-MNP binary mixture has a volume of 1.0 mL andthe phase separation process was observed between two crossed polarizers (in the horizontal andvertical directions) with transmitted white light.133A BFigure 5.5. Photographs of phase-separated CNC dispersions (colorless) and CNC-MNP binarymixtures (yellow colored). A grid background was used to distinguish the isotropic phase (whichis clear) from the liquid crystalline phase (which is birefringent) in (B).A BFigure 5.6. In a horizontally oriented gradient magnetic field (1050 Gauss/cm) from a permanentmagnet placed beside the vial, the CNC-MNP binary mixture phase-separated into a macroscopicliquid crystalline phase (with a lighter yellow color) in the low-magnetic-field region as well as adisordered phase in the high-magnetic-field region (A), while the pure CNC suspension was notsignificantly affected (B).1340 minutes 5 minutes 10 minutes 20 minutes30 minutes 40 minutes 60 minutes 90 minutes110 minutes 120 minutes 150 minutes 180 minutes210 minutes 240 minutes 5 hours 6 hours7 hours 8 hours 9 hours 10 hoursFigure 5.7. In a vertical gradient magnetic field (intensity about 1050 Gauss/cm) from a magnetunder the vial, phase separation of the CNC-MNP binary mixture results in a macroscopic liquidcrystalline phase (with a higher brightness between crossed polarizers) above a disordered phase.This process is much faster than in the absence of the gradient magnetic field.135A BFigure 5.8. In a vertical gradient magnetic field from a permanent magnet placed under the vial,the CNC-MNP binary mixture phase-separated into a macroscopic liquid crystalline phase (witha lighter yellow color) above a disordered phase (A). However, the pure CNC dispersion was notsignificantly influenced by the magnetic field (B).1365.3.2 Orientation of Liquid Crystalline Tactoids in Gradient Magnetic FieldsTo directly observe the microstructures of liquid crystalline tactoids and ordered-disorderedinterfaces by scanning electron microscopy, nonionic precursors of polyacrylamide (acrylamide,crosslinker, and photoinitiator) were added into the dispersions, after the magnetic-field-inducedphase separation, photopolymerization was initiated by ultraviolet light (wavelength of 300 nm),rapidly forming crosslinked polymer networks to solidify these fluid systems (Figure 5.9).If the phase separation of a CNC-MNP binary mixture is only driven by gravity, most chiralnematic layers in the continuous ordered phase will be horizontally oriented, whereas the discretetactoids in the isotropic phase have random orientations, as revealed by both POM (Figure 5.10)and SEM micrographs (Figure 5.11). These structures are similar to those observed in pure CNCdispersions (Figure 5.12 and Figure 5.13).As previously demonstrated, if the paramagnetic-nanoparticle-doped lyotropic liquid crystalphase-separates in a gradient magnetic field (about 1050 Gauss/cm), the macroscopic continuousordered phase will form in the lower magnetic field region with the ordered-disordered interfaceperpendicular to this gradient magnetic field, or, more precisely, perpendicular to the vector sumof (ΣFmag + ΣFgrav) (Figure 5.14 and Figure 5.15). Besides, during the phase separation process,discrete liquid crystalline tactoids will be unidirectionally oriented with the chiral nematic layersparallel to the gradient magnetic field, as revealed by cross-sectional POM (Figure 5.16).Here the alignment may be caused by the shear forces exerted on tactoids by the continuousdisordered phase during the unidirectional movement of tactoids driven by magnetic accelerationforces.52 In our work, the chiral nematic layers of tactoids are oriented parallel to a weak gradientmagnetic field (maximum 1600 Gauss, i.e., 0.16 Tesla) during the phase separation process, butchanging the direction of the magnetic field (by moving the magnet to a different side of the vial)137after completion of the phase separation will not lead to reorientation of the chiral nematic layersparallel to the new magnetic field direction, although the positions of the disordered and orderedphases will follow the magnetic field. This mechanism is different from the magnetic alignmentof liquid crystals reported by Gray and others, where the chiral nematic layers of tactoids can bedirectly aligned perpendicular to a strong magnetic field (about 7 Tesla) due to the paramagneticor diamagnetic anisotropy of the mesogen nanocrystals themselves.92−96A BFigure 5.9. Photographs of phase-separated CNC-MNP binary mixtures captured in cross-linkedpolyacrylamide networks by in-situ photopolymerization. The phase separation of these systemsoccurred in (A) horizontally or (B) vertically oriented gradient magnetic fields.138A BC DE FFigure 5.10. POM images of a CNC-MNP dispersion (no magnets). Scale bars, 50 μm.139A BC DFigure 5.11. Cross-sectional SEM micrographs for a phase-separated CNC-MNP dispersion (nomagnetic fields). Chiral nematic layers in the macroscopically continuous ordered phase adopt ahorizontal orientation, while discrete liquid crystalline tactoids in the continuous isotropic phaseare nearly randomly oriented. Scale bars, (A) 10 μm, (B) 1 μm, (C) 200 nm, (D) 100 nm.140Figure 5.12. 3D model of a CNC-MNP binary mixture (no magnetic fields).A BFigure 5.13. SEM images of a pure CNC dispersion. Scale bars, (A) 40 μm, (B) 20 μm.141A BC DE FFigure 5.14. SEM images showing the ordered-disordered interfaces of CNC-MNP suspensionsthat phase-separated in a horizontal gradient magnetic field. The interfaces are tilted. Scale bars,(A) 10 μm, (B) 2 μm, (C,F) 1 μm, (D) 500 nm, (E) 3 μm.142Figure 5.15. 3D model showing the unidirectional orientation of chiral nematic layers in both thecontinuous ordered phase and discrete liquid crystalline tactoids of a phase-separated CNC-MNPbinary mixture. The phase separation of this system occurred in a horizontally oriented gradientmagnetic field, which resulted in a nearly vertical ordered-disordered interface.143A BC DFigure 5.16. POM micrographs showing the unidirectional orientation of chiral nematic layers ofliquid crystalline tactoids in a CNC-MNP suspension subjected to an external gradient magneticfield (in the direction from left to right with respect to these images). The system was captured ina crosslinked polyacrylamide matrix by photopolymerization during the phase separation process(after a standing time of 60 minutes). The orientation of chiral nematic layers in these tactoids isparallel to the external gradient magnetic field. From (A,B) to (C) and (D), the imaging area wasincreasingly further from the high-magnetic-field region at the magnetic pole. These micrographsindicate that tactoids were driven by magnetic buoyancy forces (from the continuous disorderedphase that has a higher magnetic susceptibility) and move to lower magnetic field regions, wherethey aggregate and coalesce into continuous liquid crystalline phases. Scale bars, 50 μm.1445.3.3 Vertically Aligned Chiral Nematic Layers: Reversed Phase SeparationPhase separation of a CNC-MNP binary mixture in a vertical gradient magnetic field (froma magnet placed under the dispersion) leads to the vertical orientation of all chiral nematic layersin both the discrete liquid crystalline tactoids and macroscopically continuous ordered phases, asrevealed by POM (Figure 5.17) and SEM micrographs (Figure 5.18). Moreover, high-resolutionSEM images confirm that all the pseudo nematic layers in the chiral nematic phase are aligned inthe vertical direction (Figure 5.19). This alignment is strong enough to overcome the anchoringenergy near the boundary of the liquid crystalline phase, as the vertical chiral nematic layers areperpendicular to both the horizontal ordered-disordered interface at the bottom and the horizontalliquid-air interface at the top (Figure 5.20 to Figure 5.25).Based on SEM observations of the biphasic regions (Figure 5.26), discrete liquid crystallinetactoids in the isotropic phase are also vertically oriented with their chiral nematic layers alignedparallel to the gradient magnetic field. As previously discussed, this alignment might result fromthe shear forces exerted on tactoids by the continuous disordered phase during the unidirectionalmovement of tactoids, which is mainly driven by the magnetic acceleration forces (ΣFmag).These results reveal that in the presence of a gradient magnetic field, the phase separation ofa paramagnetic-nanoparticle-doped liquid crystal may result in the reversed phase configuration,where the macroscopic ordered phase forms above the disordered phase, despite the fact that theordered phase has a higher density due to the more efficient packing of the mesogens. This studyalso suggests a new approach to control the orientation of ordered microstructures in soft mattersystems, which may help develop new optical devices.145Figure 5.17. POM micrograph of a CNC-MNP binary mixture that phase-separated in a verticalgradient magnetic field, where the chiral nematic layers in both the continuous ordered phase anddiscrete liquid crystalline tactoids are vertically oriented. Scale bar, 50 μm.A BFigure 5.18. SEM images showing the vertical orientation of chiral nematic layers in continuousordered phases and discrete liquid crystalline tactoids. Scale bars, (A) 15 μm, (B) 10 μm.146A BC DFigure 5.19. High-resolution SEM micrographs revealing the ordered arrangements of cellulosenanocrystal mesogens in the continuous liquid crystalline phase of a CNC-MNP suspension thatphase-separated in a vertical gradient magnetic field. The pseudo nematic layers formed by theseCNC mesogens are in the vertical direction, which are parallel to the external gradient magneticfield. Scale bars, (A) 300 nm, (B,C) 200 nm, (D) 100 nm.147A BC DFigure 5.20. Cross-sectional POM micrographs showing the continuous liquid crystalline phaseof a CNC-MNP binary mixture, which phase-separated in a vertical gradient magnetic field. Thechiral nematic layers are aligned in a vertical orientation throughout this liquid crystalline phase,from the liquid-air interface at the top of the dispersion (A) to the ordered-disordered interface atthe bottom of the anisotropic phase (B-D). Scale bars, 50 μm.148A BC DE FFigure 5.21. SEM images showing the horizontal ordered-disordered interface and vertical chiralnematic layers in a CNC-MNP binary system that phase-separated in a vertical gradient magneticfield. Scale bars, (A) 5 μm, (B) 3 μm, (C) 2 μm, (D) 1 μm, (E) 500 nm, (F) 200 nm.149A BC DE FFigure 5.22. SEM images showing the horizontal ordered-disordered interface and vertical chiralnematic layers in a CNC-MNP binary system that phase-separated in a vertical gradient magneticfield. Scale bars, (A) 30 μm, (B) 10 μm, (C) 3 μm, (D) 1 μm, (E) 300 nm, (F) 150 nm.150A BC DFigure 5.23. Additional SEM images showing the vertical orientation of chiral nematic layers ina phase-separated CNC-MNP binary mixture. These chiral nematic layers are aligned parallel tothe external gradient magnetic field, perpendicular to the horizontal ordered-disordered interface.In this case, the cross-sectional appearance of chiral nematic structures is also determined by thefracture angle when the cross-section surface was created from the CNC-MNP-PAAm compositepolymer matrix. Scale bars, (A) 5 μm, (B) 2 μm, (C) 1 μm, (D) 400 nm.151Figure 5.24. 3D model showing the cross-sectional structure of a CNC-MNP binary mixture thatphase-separated in a vertical gradient magnetic field from a magnet placed under the suspension.In this system, chiral nematic layers in both the continuous ordered phase (the upper region) anddiscrete liquid crystalline tactoids (in the lower region) are vertically oriented, while the interfacebetween the ordered and disordered phases is in the horizontal direction.152A BFigure 5.25. SEM images showing the liquid-air interface at the top of a CNC-MNP suspensionthat phase-separated in a vertical gradient magnetic field. The chiral nematic layers are verticallyoriented near this interface. Scale bars, (A) 100 μm, (B) 30 μm.A BFigure 5.26. SEM micrographs showing the biphasic region in a CNC-MNP binary mixture. Thesystem was captured by photopolymerization while the phase separation process was ongoing ina vertical gradient magnetic field from a magnet placed under the dispersion. The chiral nematiclayers in both the continuous ordered phase and discrete liquid crystalline tactoids are verticallyoriented. Scale bars, (A) 20 μm, (B) 10 μm.1535.4 Competitions between Gravitational and Magnetic Acceleration ForcesWhen a magnet is placed under a CNC-MNP dispersion, the phase separation configurationand rate will be determined by the competition between ΣFmag and ΣFgrav since these two vectorsare in opposite directions (i.e., ΣFmag vertically upwards, ΣFgrav vertically downwards):(1) |ΣFmag| < |ΣFgrav| leads to "Normal" phase separations, where the liquid crystalline phaseforms below the isotropic phase;(2) |ΣFmag| > |ΣFgrav| leads to "Reverse" phase separations, where the liquid crystalline phaseforms above the isotropic phase;(3) |ΣFmag + ΣFgrav| determines the phase separation rate, i.e., for either the "Normal" or the"Reverse" phase separation, a higher absolute value of the vector sum of ΣFmag and ΣFgrav meansa faster phase separation rate, or a shorter phase separation time.For experimental verification, influences of the doping concentration of superparamagneticFe3O4 nanoparticles (maximum 140.5 ppm by weight) and strength of the gradient magnetic field(0, 600, 1050, and 3050 Gauss/cm) on the phase separation rates of CNC-MNP binary mixtures(volume of 1.0 mL) were investigated, where the experimental results agree with the theoreticalpredictions above: increasing either the gradient magnetic field (raising the |H|) or the total Fe3O4concentration (raising [χm(Iso) − χm(LC)]) will increase ΣFmag and therefore slow down a "Normal"or accelerate a "Reverse" phase separation process (Table 5.1 and Figure 5.27). And moreover,the phase separation time reaches the maximum value when the inversion between the "Normal"and "Reverse" configurations occurs, where the magnitude of the vector sum of (ΣFmag + ΣFgrav)approaches zero.As verified by ultraviolet-visible absorption spectroscopy (at 300 nm wavelength), there is apositive correlation between [χm(Iso) − χm(LC)] (the magnetic susceptibility difference) and the total154concentration of paramagnetic Fe3O4 nanoparticles, where the distribution ratio of Fe3O4 dopingnanoparticles in the isotropic and liquid crystalline phases ([Fe3O4]Iso/[Fe3O4]LC) increased fromaround 2.40 to 4.41 as the total Fe3O4 concentration increased from 11.7 to 140.5 ppm by weight(Table 5.2; Figure 5.28 and Figure 5.29). These results should be applicable to liquid crystallinetactoids as the coalescence of discrete tactoids leads to the formation of macroscopic continuousordered phases.Table 5.1. Phase separation times of CNC-Fe3O4 binary mixtures.Conc. of Fe3O4NPs (wt. ppm) Zero Gauss/cm 600 Gauss/cm 1050 Gauss/cm 3050 Gauss/cm0 6 h 06 min N 6 h 09 min N 5 h 57 min N 6 h 08 min N11.7 5 h 54 min N 6 h 20 min N 7 h 53 min N 9 h 46 min R23.4 5 h 59 min N 6 h 32 min N 24 h 08 min N 7 h 57 min R35.1 5 h 52 min N 6 h 47 min N 13 h 47 min R 3 h 03 min R46.8 5 h 49 min N 7 h 04 min N 7 h 57 min R 1 h 52 min R93.7 5 h 27 min N 14 h 03 min R 2 h 54 min R 55 min R140.5 5 h 48 min N 4 h 13 min R 2 h 21 min R 49 min RNotes: The CNC-MNP dispersions were placed in glass vials (inner diameter of 12.60 mm, outerdiameter of 15.00 mm); the completion of phase separation was determined by the formation of aclear and smooth boundary between two phases (the errors are about +/− 10 minutes); "N" meansnormal phase separation, where the isotropic phase forms above the liquid crystalline phase; "R"represents reverse phase separation, where the isotropic phase forms below the liquid crystallinephase.155Figure 5.27. Phase separation times of CNC-MNP dispersions (volume of 1.0 mL) with differentdoping concentrations of Fe3O4 nanoparticles (0 to 140.5 ppm by weight) and in vertical gradientmagnetic fields with different strengths (0, 600, 1050, and 3050 Gauss/cm). The striped columnsrepresent reversed phase separations, where the liquid crystalline phase forms above the isotropicphase. (Errors in the determination of phase separation times are +/− 10 minutes).156Table 5.2. Distribution ratio of Fe3O4 NPs in isotropic and liquid crystalline phases.Conc. of Fe3O4 NPs(ppm by weight)Abs. of Fe3O4 NPs(Isotropic Phase)Abs. of Fe3O4 NPs(Ordered Phase) [Fe3O4]Iso/[Fe3O4]LC11.7 0.120 0.050 2.4023.4 0.200 0.073 2.7435.1 0.260 0.085 3.0646.8 0.325 0.094 3.4693.7 0.622 0.153 4.07140.5 0.854 0.193 4.41Notes: The distribution ratios of Fe3O4 nanoparticles were determined by UV-Vis absorbance atthe wavelength of 300 nm, where the absorbance of cellulose nanocrystals has been subtracted.Figure 5.28. UV-Vis absorption spectra of the isotropic phase (red curve) and liquid crystallinephase (blue curve) of a phase-separated pure CNC dispersion. The two phases were isolated anddiluted by water to 10 times their original volumes.157[Fe3O4]Total = 11.7 wt. ppm [Fe3O4]Total = 46.8 wt. ppm[Fe3O4]Total = 23.4 wt. ppm [Fe3O4]Total = 93.7 wt. ppm[Fe3O4]Total = 35.1 wt. ppm [Fe3O4]Total = 140.5 wt. ppmFigure 5.29. Absorbance of MNPs in isotropic (red) and liquid crystalline phases (blue).1585.5 ConclusionsIn this study, by doping lyotropic liquid crystals with paramagnetic magnetite nanoparticles,the disordered phases were endowed with significantly higher magnetic susceptibility than liquidcrystalline tactoids since tactoids have an exclusion effect on paramagnetic doping nanoparticles.Due to the lower magnetic susceptibility, the movement direction and orientation of tactoids aredetermined by the competition between gravitational and magnetic acceleration fields, where thetactoids are mainly driven by magnetic buoyancy forces. This method enables control of the rateand configuration of phase separation, as well as the orientation of the director fields in both thediscrete tactoids and macroscopic continuous ordered phases. These results may help understandthe physical properties of liquid crystalline tactoids and other soft matter systems, and might alsoprovide new methods to precisely control self-assembly processes for developing new functionalmaterials.1595.6 Experimental Methods5.6.1 Synthesis of Superparamagnetic Magnetite (Fe3O4) NanoparticlesCitrate-stabilized superparamagnetic magnetite (Fe3O4) nanoparticles were synthesized by acoprecipitation method.89 In a typical synthesis, iron (II) chloride tetrahydrate (0.994 g, 5 mmol)and iron (III) chloride hexahydrate (2.703 g, 10 mmol) were thoroughly dissolved in pure water(50.0 mL). The system was stirred (at 1150 RPM) and deoxygenated by bubbling argon throughthe solution. An ammonium hydroxide solution (28-30 wt.% in water, 6.0 mL) was added to thesystem using a syringe. After 30 minutes, an aqueous solution of citric acid (1.50 g of citric acidanhydrous dissolved in 2.0 mL of pure water) was introduced into the reaction mixture, then thesystem was stirred at room temperature for another 120 minutes. The Fe3O4 nanoparticles couldbe magnetically isolated then redispersed in pure water for several cycles to remove electrolytes,and eventually a stable homogeneous dispersion of Fe3O4 nanoparticles (about 1.41% by weight)was obtained.5.6.2 Solidification of CNC-Fe3O4 Mixtures by In-Situ PhotopolymerizationIn a standard experiment, an aqueous dispersion of superparamagnetic Fe3O4 nanoparticles(1.41 wt.%, 50 μL) was added to a cellulose nanocrystal suspension (4.0 wt.%, pH 6.4, 5.0 mL),then the system was mixed with acrylamide (500 mg, monomer), N,N′-methylenebisacrylamide(50 mg, crosslinker), and 2-hydroxy-4′-(2-hydroxyethoxy)-2-methylpropiophenone (2.5 mg, thephotoinitiator). The CNC-Fe3O4 dispersion was allowed to stand in the dark (with or without anexternal gradient magnetic field) for about 12 hours, during which the phase separation occurred.Afterwards, UV-B irradiation (wavelength of 300 nm, 8 W) was applied to the system to initiate160the photopolymerization process, rapidly forming a crosslinked polyacrylamide (PAAm) matrixto capture liquid crystalline tactoids and other microstructures. After complete removal of water(by heating at 65 °C in air), the solid-state polyacrylamide matrix could be cracked to give freshcross-sections for SEM observations.5.6.3 MaterialsThe cellulose nanocrystal dispersions (4.0 wt.%, pH = 6.4, conductivity = 395 μS/cm) wereprovided by FPInnovations. Other chemicals are the same as those used in previous chapters.5.6.4 CharacterizationThree different permanent magnets were used in the performance of magnetic-field-inducedphase-separation experiments. The strengths of the gradient magnetic fields were measured witha Bell Model 620 Hall Effect Gaussmeter. For each of these permanent magnets, the strengths ofits magnetic field at the surface of the magnetic pole and at 1.0 centimeter away from the surface(on the central axis) were measured, which are 1600 Gauss / 550 Gauss, 1000 Gauss / 400 Gauss,and 3600 Gauss / 550 Gauss for the three magnets, respectively. Other characterization methodsare the same as those used in previous chapters.161Chapter 6: Conclusions and Future Perspectives6.1 ConclusionsThe improvements in observational techniques or methods, which can provide higher spatialand temporal resolutions, are the material basis of a better understanding of physical phenomenaand therefore new theoretical models. For condensed matter physics, electron microscopy is oneof the most important characterization methods, as it can directly reveal the microstructures of asystem at the resolution of individual particles.An attractive topic in condensed matter physics is the spontaneous emergence and evolutionof ordered structures in an initially disordered system, for example, the formation of anisotropicphases in isotropic dispersions. In lyotropic liquid crystals, this process is generally mediated bytactoids, which are discrete ordered microdroplets in continuous disordered phases. However, thestructural studies of tactoids were difficult as these soft fluid droplets cannot be directly observedby electron microscopy.In Chapter 2 of this thesis, a new method was developed to enable the direct observation ofliquid crystalline tactoids by electron microscopy, where these metastable ordered microdropletscould be rapidly captured in a crosslinked polyacrylamide matrix by in-situ photopolymerization.This method provides an approach to the solidification of soft matter systems and preservation oflyotropic liquid crystalline structures while removing the solvents, enabling electron microscopyobservations of tactoids with the resolution of individual liquid crystal mesogens. The emergenceof tactoids in initially disordered phases, the coalescence of multiple tactoids and the generationof topological defects, as well as the sedimentation of tactoids and the formation of macroscopiccontinuous liquid crystalline phases were systemically investigated, which provides new insights162into the structure and evolution of tactoids, topological defects and other microscopic metastabletransition states in lyotropic liquid crystals.In Chapter 3 of this thesis, the in-situ photopolymerization of lyotropic liquid crystals wasextended to an inverse emulsion system. The structural transformation of chiral nematic tactoidsin spherical confinement was captured at different evolution stages and examined by both opticaland cross-sectional electron microscopy. Furthermore, solid-state polymer and mesoporous silicamicrospheres with chiral nematic ordering were obtained, which may have potential applicationsin optics, sensing and chiral separation.Chapter 4 of this dissertation focuses on the behavior of tactoids in the presence of foreignnanoparticles. Based on the in-situ photopolymerization method, liquid crystalline tactoids wereobserved to have size-selective exclusion effects on doping nanoparticles. This research directlyrevealed the initial emergence of self-assembly in binary mixtures of nanorods and nanospheresby electron microscopy, and investigated the interactions between ordered-disordered interfacesand foreign nanoparticles. The observations in these experiments provide new information aboutthe metastable transition states in multicomponent colloidal systems, and could help explain theaggregation of doping nanoparticles in topological defects of lyotropic liquid crystals. Moreover,a size-selective separation method for nanoparticles was successfully developed according to theexclusion effects of tactoids, which may have applications in the purification of nanomaterials.In Chapter 5 of this thesis, the influence of gradient magnetic fields on the phase separationof paramagnetic-nanoparticle-doped lyotropic liquid crystals was investigated. This research wasbased on the exclusion effects of tactoids on doping nanoparticles, by which the liquid crystallinetactoids could be endowed with significantly lower volume magnetic susceptibility than isotropicphases. This difference in magnetic properties facilitates the manipulation of tactoids by gradient163magnetic fields, where the movement and orientation of liquid crystalline tactoids are determinedby the competition between gravitational and magnetic acceleration forces, which allows controlover the phase separation rate and configuration, as well as the orientation of the director field inboth discrete tactoids and continuous ordered phases. These studies provide new insights into theevolution of tactoids and phase separation processes of lyotropic liquid crystals, and may help todevelop new functional materials with ordered microstructures.In summary, this thesis investigates the microscopic structures of liquid crystalline tactoids,which are representative of the metastable transition states of lyotropic systems. Microstructuresof transition states are difficult to observe since they are always moving and changing, especiallywhen these structures are in the nanoscale or atomic scale, where electron microscopy or atomicforce microscopy is required to achieve sufficiently high resolution. This thesis partly addressesthe above issue by an in-situ photopolymerization method, where the metastable transition statesof lyotropic liquid crystals can be rapidly captured and solidified in a crosslinked polymer matrixfor electron microscopy observations. Moreover, these studies also provide new insights into theearly stages of self-assembly processes, e.g., the emergence of tactoids from an initially isotropicphase. This phenomenon may be of significance since it might share some common features withthe origin of highly ordered hierarchical structures in plants and animals, such as those observedin Pollia fruits,97 butterflies,98 and beetles.991646.2 Future Perspectives6.2.1 Liquid Crystalline Tactoids in Other Soft Matter SystemsIn this dissertation, the direct electron microscopy observation of liquid crystalline tactoidswas successfully achieved; however, the lyotropic liquid crystal formed by cellulose nanocrystalsis not a perfect model for the study of self-assembly structures or processes. Although individualcellulose nanocrystals could be directly observed by scanning electron microscopy, these organicnanoparticles are not very stable under an electron beam. Furthermore, the chirality of this liquidcrystal makes the structures more complicated, especially near the ordered-disordered interfacesand topological defects. To avoid these problems, nematic or columnar liquid crystals formed byinorganic (e.g., metal hydroxides or metal phosphates) nanoparticles would be better candidatesfor the structural studies of tactoids and topological defects.6.2.2 Size-Selective Separation of Molecules by Liquid Crystalline TactoidsAnother hypothesis worth further investigation is the size-selective separation of proteins orother organic macromolecules during the tactoid-mediated phase separation process of molecularlyotropic liquid crystals, for example, the nematic phase formed by Sunset Yellow FCF.100 Thesestudies would provide valuable information about the interactions between mesogens in tactoidsand macroscopic ordered phases, and may help to develop new separation techniques for organiccompounds, atom clusters or biomacromolecules.As the smallest units of ordered phases and early stages of self-assembly, liquid crystallinetactoids play an important role in the phase transitions of lyotropic systems. The investigation oftactoids can provide a better understanding of the formation and properties of liquid crystals, and165promote the development of functional materials with ordered microstructures. Furthermore, thein-situ photopolymerization method may be used to capture and solidify the metastable transitionstates in other soft matter systems, such as foams, colloids, and some biological materials, whichwould enable the direct observation of these fluid microstructures by electron microscopy.First Draft Completed on 13th July 2018 Friday by PEI-XI WANG初稿完成于二零一八年七月十三日夜 王佩玺166Bibliography(1) Reinitzer, F. Monatsh. Chem. 1888, 9, 421-441.(2) Lehmann, O. Z. Phys. Chem. 1889, 4, 462-472.(3) Oseen, C. W. Trans. Faraday Soc. 1933, 29, 883-899.(4) Sonin, A. S. J. Mater. Chem. 1998, 8, 2557-2574.(5) Schadt, M.; Helfrich, W. Appl. Phys. Lett. 1971, 18, 127-128.(6) Smith, C. R.; Sabatino, D. R.; Praisner, T. J. Exp. Fluids. 2001, 30, 190-201.(7) Beck, J. S.; Vartuli, J. C.; Roth, W. J.; Leonowicz, M. E.; Kresge, C. T.; Schmitt, K. D.;Chu, C. T. W.; Olson, D. H.; Sheppard, E. W. J. Am. Chem. Soc. 1992, 114, 10834-10843.(8) Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck, J. S. Nature 1992, 359,710-712.(9) Yang, P.; Zhao, D.; Margolese, D. I.; Chmelka, B. F.; Stucky, G. D. Nature 1998, 396, 152-155.(10) Zhao, D.; Feng, J.; Huo, Q.; Melosh, N.; Fredrickson, G. H.; Chmelka, B. F.; Stucky, G. D.Science 1998, 279, 548-552.(11) Zocher, H. Z. Anorg. Allg. Chem. 1925, 147, 91-110.(12) Bawden, F. C.; Pirie, N. W.; Bernal, J. D.; Fankuchen, I. Nature 1936, 138, 1051-1052.(13) Coper, K.; Freundlich, H. Trans. Faraday Soc. 1937, 33, 348-350.(14) Robinson, C. Trans. Faraday Soc. 1956, 52, 571-592.(15) Revol, J. F.; Bradford, H.; Giasson, J.; Marchessault, R. H.; Gray, D. G. Int. J. Biol.Macromol. 1992, 14, 170-172.(16) Wang, P.-X.; MacLachlan, M. J. Phil. Trans. R. Soc. A 2018, 376, 20170042.167(17) Langmuir, I. J. Chem. Phys. 1938, 6, 873-896.(18) Kallmann, H.; Willstätter, M. Naturwissenschaften 1932, 20, 952-953.(19) Verwey, E. J. W. J. Phys. Chem. 1947, 51, 631-636.(20) McBride, M. B. Clays Clay Miner. 1997, 45, 598-608.(21) Hirai, A.; Inui, O.; Horii, F.; Tsuji, M. Langmuir 2009, 25, 497-502.(22) Lekkerkerker, H. N. W.; Coulon, P.; Haegen, R. V. D.; Deblieck, R. J. Chem. Phys. 1984,80, 3427-3433.(23) Onsager, L. Ann. N. Y. Acad. Sci. 1949, 51, 627-659.(24) Kaznacheev, A. V.; Bogdanov, M. M.; Taraskin, S. A. J. Exp. Theor. Phys. 2002, 95, 57-63.(25) Kaznacheev, A. V.; Bogdanov, M. M.; Sonin, A. S. J. Exp. Theor. Phys. 2003, 97, 1159-1167.(26) Prinsen, P.; van der Schoot, P. Phys. Rev. E 2003, 68, 021701.(27) Chen, W.; Gray, D. G. Langmuir 2002, 18, 633-637.(28) Zocher, H.; Jacobsohn, K. Kolloid Z. 1927, 41, 220-222.(29) Zocher, H.; Török, C. Kolloid Z. 1960, 170, 140-144.(30) Bawden, F. C.; Pirie, N. W.; Bernal, J. D.; Fankuchen, I. Nature 1936, 138, 1051-1052.(31) Dogic, Z.; Fraden, S. Phil. Trans. R. Soc. A 2001, 359, 997-1015.(32) Marchessault, R. H.; Morehead, F. F.; Walter, N. M. Nature 1959, 184, 632-633.(33) Habibi, Y.; Lucia, L. A.; Rojas, O. J. Chem. Rev. 2010, 110, 3479-3500.(34) Shopsowitz, K. E.; Qi, H.; Hamad, W. Y.; MacLachlan, M. J. Nature 2010, 468, 422.(35) Shopsowitz, K. E.; Hamad, W. Y.; MacLachlan, M. J. Angew. Chem. Int. Ed. 2011, 50,10991-10995.168(36) Shopsowitz, K. E.; Hamad, W. Y.; MacLachlan, M. J. J. Am. Chem. Soc. 2012, 134, 867-870.(37) Shopsowitz, K. E.; Stahl, A.; Hamad, W. Y.; MacLachlan, M. J. Angew. Chem. Int. Ed.2012, 51, 6886-6890.(38) Kelly, J. A.; Shukaliak, A. M.; Cheung, C. C. Y.; Shopsowitz, K. E.; Hamad, W. Y.;MacLachlan, M. J. Angew. Chem. Int. Ed. 2013, 52, 8912-8916.(39) Khan, M. K.; Giese, M.; Yu, M.; Kelly, J. A.; Hamad, W. Y.; MacLachlan, M. J. Angew.Chem. Int. Ed. 2013, 52, 8921-8924.(40) Nguyen, T.-D.; Hamad, W. Y.; MacLachlan, M. J. Adv. Funct. Mater. 2014, 24, 777-783.(41) Kelly, J. A.; Giese, M.; Shopsowitz, K. E.; Hamad, W. Y.; MacLachlan, M. J. Acc. Chem.Res. 2014, 47, 1088-1096.(42) Giese, M.; Blusch, L. K.; Khan, M. K.; MacLachlan, M. J. Angew. Chem. Int. Ed. 2015, 54,2888-2910.(43) Xu, J.; Nguyen, T.-D.; Xie, K.; Hamad, W. Y.; MacLachlan, M. J. Nanoscale 2015, 7,13215-13223.(44) Wang, P.-X.; Zamarion, V. M.; Hamad, W. Y.; MacLachlan, M. J. Dalton Trans. 2015, 44,14724-14731.(45) Zamarion, V. M.; Khan, M. K.; Schlesinger, M.; Bsoul, A.; Walus, K.; Hamad, W. Y.;MacLachlan, M. J. Chem. Commun. 2016, 52, 7810-7813.(46) Terpstra, A. S.; Hamad, W. Y.; MacLachlan, M. J. Adv. Funct. Mater. 2017, 27, 1703346.(47) Orts, W. J.; Godbout, L.; Marchessault, R. H.; Revol, J.-F. Macromolecules 1998, 31, 5717-5725.169(48) Araki, J.; Kuga, S. Langmuir 2001, 17, 4493-4496.(49) Hanley, S. J.; Revol, J.-F.; Godbout, L.; Gray, D. G. Cellulose 1997, 4, 209.(50) Elazzouzi-Hafraoui, S.; Nishiyama, Y.; Putaux, J.-L.; Heux, L.; Dubreuil, F.; Rochas, C.Biomacromolecules 2008, 9, 57-65.(51) Schütz, C.; Agthe, M.; Fall, A. B.; Gordeyeva, K.; Guccini, V.; Salajková, M.; Plivelic, T.S.; Lagerwall, J. P. F.; Salazar-Alvarez, G.; Bergström, L. Langmuir 2015, 31, 6507-6513.(52) Park, J. H.; Noh, J.; Schütz, C.; Salazar-Alvarez, G.; Scalia, G.; Bergström, L.; Lagerwall, J.P. ChemPhysChem 2014, 15, 1477-1484.(53) Wang, P.-X.; Hamad, W. Y.; MacLachlan, M. J. Nat. Commun. 2016, 7, 11515.(54) Wang, P.-X.; Hamad, W. Y.; MacLachlan, M. J. Angew. Chem. Int. Ed. 2016, 55, 12460-12464.(55) Wang, P.-X.; Hamad, W. Y.; MacLachlan, M. J. Angew. Chem. Int. Ed. 2018, 57, 3360-3365.(56) Zeppieri, S.; Rodríguez, J.; López de Ramos, A. L. J. Chem. Eng. Data 2001, 46, 1086-1088.(57) Erdmann, J. H.; Žumer, S.; Doane, J. W. Phys. Rev. Lett. 1990, 64, 1907-1910.(58) Kralj, S.; Žumer, S.; Allender, D. W. Phys. Rev. A 1991, 43, 2943-2952.(59) Humar, M.; Muševič, I. Opt. Express 2010, 18, 26995-27003.(60) Humar, M.; Muševič, I. Opt. Express 2011, 19, 19836-19844.(61) Iwai, Y.; Kaji, H.; Uchida, Y.; Nishiyama, N. Mol. Cryst. Liq. Cryst. 2015, 615, 9-13.(62) Xu, F.; Crooker, P. P. Phys. Rev. E 1997, 56, 6853-6860.(63) Seč, D.; Porenta, T.; Ravnik, M.; Žumer, S. Soft Matter 2012, 8, 11982-11988.(64) Seč, D.; Čopar, S.; Žumer, S. Nat. Commun. 2014, 5, 3057.170(65) Orlova, T.; Aßhoff, S. J.; Yamaguchi, T.; Katsonis, N.; Brasselet, E. Nat. Commun. 2015, 6,7603.(66) Granick, S. Science 1991, 253, 1374-1379.(67) Yang, H.; Yan, R.; Chen, H.; Lee, D. H.; Zheng, C. Fuel 2007, 86, 1781-1788.(68) Kaşgöz, H.; Özgümüş, S.; Orbay, M. Polymer 2003, 44, 1785-1793.(69) Aguiar, H.; Serra, J.; González, P.; León, B. J. Non-Cryst. Solids 2009, 355, 475-480.(70) Li, Y.; Suen, J. J.-Y.; Prince, E.; Larin, E. M.; Klinkova, A.; Thérien-Aubin, H.; Zhu, S.;Yang, B.; Helmy, A. S.; Lavrentovich, O. D.; Kumacheva, E. Nat. Commun. 2016, 7, 12520.(71) Parker, R. M.; Frka-Petesic, B.; Guidetti, G.; Kamita, G.; Consani, G.; Abell, C.; Vignolini,S. ACS Nano 2016, 10, 8443-8449.(72) Schwartz, M.; Lenzini, G.; Geng, Y.; Rønne, P. B.; Ryan, P. Y. A.; Lagerwall, J. P. F. Adv.Mater. 2018, 30, 1707382.(73) Bell, A. T. Science 2003, 299, 1688-1691.(74) Giljohann, D. A.; Seferos, D. S.; Daniel, W. L.; Massich, M. D.; Patel, P. C.; Mirkin, C. A.Angew. Chem. Int. Ed. 2010, 49, 3280-3294.(75) You, H.; Yang, S.; Ding, B.; Yang, H. Chem. Soc. Rev. 2013, 42, 2880-2904.(76) Saha, K.; Agasti, S. S.; Kim, C.; Li, X.; Rotello, V. M. Chem. Rev. 2012, 112, 2739-2779.(77) Sharma, V.; Park, K.; Srinivasarao, M. Proc. Natl. Acad. Sci. USA 2009, 106, 4981-4985.(78) Sweeney, S. F.; Woehrle, G. H.; Hutchison, J. E. J. Am. Chem. Soc. 2006, 128, 3190-3197.(79) Hanauer, M.; Pierrat, S.; Zins, I.; Lotz, A.; Sönnichsen, C. Nano Letters 2007, 7, 2881-2885.(80) Wei, G.-T.; Liu, F.-K. J. Chromatogr. A 1999, 836, 253-260.(81) Giddings, J. Science 1993, 260, 1456-1465.171(82) Robertson, J. D.; Rizzello, L.; Avila-Olias, M.; Gaitzsch, J.; Contini, C.; Magoń, M. S.;Renshaw, S. A.; Battaglia, G. Sci. Rep. 2016, 6, 27494.(83) Schütz, C.; Agthe, M.; Fall, A. B.; Gordeyeva, K.; Guccini, V.; Salajková, M.; Plivelic, T.S.; Lagerwall, J. P. F.; Salazar-Alvarez, G.; Bergström, L. Langmuir 2015, 31, 6507-6513.(84) Adams, M.; Dogic, Z.; Keller, S. L.; Fraden, S. Nature 1998, 393, 349-352.(85) Mondiot, F.; Botet, R.; Snabre, P.; Mondain-Monval, O.; Loudet, J.-C. Proc. Natl. Acad. Sci.USA 2014, 111, 5831-5836.(86) Chen, X.; Cui, Z.; Chen, Z.; Zhang, K.; Lu, G.; Zhang, G.; Yang, B. Polymer 2002, 43,4147-4152.(87) Ding, X.; Zhao, J.; Liu, Y.; Zhang, H.; Wang, Z. Mater. Lett. 2004, 58, 3126-3130.(88) Frens, G. Nature 1973, 241, 20-22.(89) Sahoo, Y.; Goodarzi, A.; Swihart, M. T.; Ohulchanskyy, T. Y.; Kaur, N.; Furlani, E. P.;Prasad, P. N. J. Phys. Chem. B 2005, 109, 3879-3885.(90) Liu, J.; Sun, Z.; Deng, Y.; Zou, Y.; Li, C.; Guo, X.; Xiong, L.; Gao, Y.; Li, F.; Zhao, D.Angew. Chem. Int. Ed. 2009, 48, 5875-5879.(91) Auernhammer, G. K.; Brand, H. R. Eur. Phys. J. B 2000, 16, 157-168.(92) Revol, J.-F.; Godbout, L.; Dong, X.-M.; Gray, D. G.; Chanzy, H.; Maret, G. Liq. Cryst.1994, 16, 127-134.(93) Dong, X. M.; Gray, D. G. Langmuir 1997, 13, 3029-3034.(94) De France, K. J.; Yager, K. G.; Hoare, T.; Cranston, E. D. Langmuir 2016, 32, 7564-7571.(95) Frka-Petesic, B.; Guidetti, G.; Kamita, G.; Vignolini, S. Adv. Mater. 2017, 29, 1701469.172(96) Nakayama, M.; Kajiyama, S.; Kumamoto, A.; Nishimura, T.; Ikuhara, Y.; Yamato, M.;Kato, T. Nat. Commun. 2018, 9, 568.(97) Vignolini, S.; Rudall, P. J.; Rowland, A. V.; Reed, A.; Moyroud, E.; Faden, R. B.;Baumberg, J. J.; Glover, B. J.; Steiner, U. Proc. Natl. Acad. Sci. USA 2012, 109, 15712-15715.(98) Ghiradella, H.; Aneshansley, D.; Eisner, T.; Silberglied, R. E.; Hinton, H. E. Science 1972,178, 1214-1217.(99) Bouligand, Y. Tissue Cell 1972, 4, 189-217.(100) Lydon, J. J. Mater. Chem. 2010, 20, 10071-10099.

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0372765/manifest

Comment

Related Items