Article

Chiral Nematic Structure of Cellulose Nanocrystal Suspensions and Films; Polarized Light and Atomic Force Microscopy Derek G. Gray * and Xiaoyue Mu Received: 5 October 2015 ; Accepted: 10 November 2015 ; Published: 20 November 2015 Academic Editors: Aldo R. Boccaccini, Jung Kwon (John) Oh and Juergen Stampfl Department of Chemistry, McGill University, Pulp and Paper Building, 3420 University Street, Montreal, QC H3A-2A7, Canada; [email protected] * Correspondence: [email protected]; Tel.: +1-514-398-6182

Abstract: Cellulosic liquid crystalline solutions and suspensions form chiral nematic phases that show a rich variety of optical textures in the liquid crystalline state. These ordered structures may be preserved in solid films prepared by evaporation of solvent or suspending medium. Film formation from aqueous suspensions of cellulose nanocrystals (CNC) was investigated by polarized light microscopy, optical profilometry and atomic force microscopy (AFM). An attempt is made to interpret qualitatively the observed textures in terms of the orientation of the cellulose nanocrystals in the suspensions and films, and the changes in orientation caused by the evaporative process. Mass transfer within the evaporating droplet resulted in the formation of raised rings whose magnitude depended on the degree of pinning of the receding contact line. AFM of dry films at short length scales showed a radial orientation of the CNC at the free surface of the film, along with a radial height variation with a period of approximately P/2, ascribed to the anisotropic shrinkage of the chiral nematic structure. Keywords: cellulose nanocrystals; chiral nematic order; polarized light microscopy; fingerprint texture; reflection colours; coffee-stain effect; sessile droplets; contact line pinning

1. Introduction Cellulosic liquid crystalline materials, both cellulose derivatives [1,2] and cellulose nanocrystals [3,4] form chiral nematic phases that show a rich variety of optical textures in the liquid crystalline state. The ordered structures in the liquid crystalline state may be preserved with modification in solid films prepared by simple evaporation of solvent or suspending medium. Polarized light microscopy, supplemented by scanning probe microscopy and electron microscopy provides an accessible tool to observe the structure of these ordered suspensions and films. Here, we present some observations on the chiral nematic self-organization of cellulose-based systems, with emphasis on the aqueous suspensions of cellulose nanocrystals (CNC) and of films formed from CNC by evaporation. The factors that govern phase separation, chiral nematic pitch, mechanical properties and applications of CNC and other cellulosic nanocrystals have been reviewed elsewhere, [5–7] and will not be discussed in detail here.

Materials 2015, 8, 7873–7888; doi:10.3390/ma8115427

www.mdpi.com/journal/materials

Materials 2015, 8, 7873–7888

2. Results and Discussion Materials 2015, 8, page–page 

2.1. Fluid Cellulosic Chiral Nematic Textures 2. Results and Discussion  Materials 2015, 8, page–page 

2.1.1. Fluid Phases, Fingerprint Texture

2. Results and Discussion  2.1. Fluid Cellulosic Chiral Nematic Textures 

Above some critical concentration, suspensions of cellulose nanocrystals (CNC) form chiral 2.1. Fluid Cellulosic Chiral Nematic Textures  2.1.1. Fluid Phases, Fingerprint Texture  nematic phases [3]. The liquid crystalline textures observed by polarized light microscopy depend Above  some  critical  concentration,  suspensions  cellulose  nanocrystals  (CNC)  chiral  on the orientation of the optical axis of the helicoidalof assembly relative to that of form  the microscope. 2.1.1. Fluid Phases, Fingerprint Texture  nematic phases [3]. The liquid crystalline textures observed by polarized light microscopy depend  Figure 1 shows a typical image of an aqueous CNC suspension with regions (a) of planar texture and Above  some  critical  concentration,  suspensions  of  cellulose  nanocrystals  (CNC)  form  chiral  on the orientation of the optical axis of the helicoidal assembly relative to that of the microscope. Figure 1  (b) of fingerprint texture, where the chiral nematic axis is parallel to and orthogonal to the plane of nematic phases [3]. The liquid crystalline textures observed by polarized light microscopy depend  shows a typical image of an aqueous CNC suspension with regions (a) of planar texture and (b) of  the sample respectively. on the orientation of the optical axis of the helicoidal assembly relative to that of the microscope. Figure 1  fingerprint  texture,  where the  chiral  nematic axis  is parallel  to and  orthogonal  to  the  plane of  the  shows a typical image of an aqueous CNC suspension with regions (a) of planar texture and (b) of  Asample respectively.  variety of disclinations [8] where fingerprint lines intersect are also visible. The distance where the  chiral  nematic axis  is parallel  to and  to of the  plane of  the  betweenfingerprint  the linestexture,  in fingerprint texture is P/2, where P isorthogonal  the also  pitch the chiral nematic A  variety  of  the disclinations  [8]  where  fingerprint  lines  intersect  are  visible.  The  distance  sample respectively.  between the lines in the fingerprint texture is P/2, where P is the pitch of the chiral nematic structure.  structure. The orientation of the CNC rods is shown in the idealized sketch in Figure 2. Over time, A  variety  of  disclinations  [8]  where  fingerprint  lines  intersect  are  also  visible.  The  distance  The orientation of the CNC rods is shown in the idealized sketch in Figure 2. Over time, the CNC  the CNC tend to orient parallel to the walls of the microslide, so that that the chiral nematic axis between the lines in the fingerprint texture is P/2, where P is the pitch of the chiral nematic structure.  tend  to  orient  parallel  the  walls  the  microslide,  that  that  the  chiral  nematic  axis  becomes  becomes orthogonal to the to  plane of theof sample, and theso fingerprint texture disappears. However, by The orientation of the CNC rods is shown in the idealized sketch in Figure 2. Over time, the CNC  orthogonal  to  the  plane  of  the  sample,  and  the  fingerprint  texture  disappears.  However,  by  appropriate application of ato magnetic field the CNC unidirectional textures tend  to  orient  parallel  the  walls  of  the to microslide,  so samples, that  that  the  chiral  nematic fingerprint axis  becomes  appropriate application of a magnetic field to the CNC samples, unidirectional fingerprint textures  orthogonal  to  may the  plane  of  the  sample,  and  the  fingerprint  texture  disappears.  However,  by  over extended areas be generated [9]. over extended areas may be generated [9].  appropriate application of a magnetic field to the CNC samples, unidirectional fingerprint textures  over extended areas may be generated [9]. 

    Figure  1.  Image  of  8%  aqueous  suspension  (Batch  SB‐8%,  see  experimental  section)  of  cellulose  Figure 1. Image of 8% aqueous suspension (Batch SB-8%, see experimental section) of cellulose nanocrystals in  glass  microslide.  Regions  (a)  show  planar  texture  and  (b) show  fingerprint  texture  Figure  1.  Image  of  8%  aqueous  suspension  (Batch  SB‐8%,  see  experimental  section)  of  cellulose  nanocrystals in glass microslide. Regions (a) show planar texture and (b) show fingerprint texture (see Figure 2).  nanocrystals in  glass  microslide.  Regions  (a)  show  planar  texture  and  (b) show  fingerprint  texture  (see Figure 2). (see Figure 2). 

   

Figure 2. Simplified sketch of organization of chiral nematic textures shown in Figure 1. Regions (a)  Figure 2. Simplified sketch of organization of chiral nematic textures shown in Figure 1. Regions (a)  and  (b)  correspond  to  of planar  and  fingerprint  textures,  respectively.  The  optical  axis  of  Figure 2. Simplified sketch organization of chiral nematic textures shown in Figure 1. the  Regions and  (b)  correspond  to  planar  and  fingerprint  textures,  respectively.  The  optical  axis  of  the  microscope is from top to bottom of the sketch, with the middle of the sample in focus. The light and  (a) and (b) correspond to planar and fingerprint textures, respectively. The optical axis microscope is from top to bottom of the sketch, with the middle of the sample in focus. The light and  of the dark lines under crossed polars originate from variations in birefringence due to the orientation of  microscope is from top to bottom of the sketch, with the middle of the sample in focus. The light dark lines under crossed polars originate from variations in birefringence due to the orientation of  the cellulose nanocrystals (CNCs) along and across the microscope optical axis [8].  the cellulose nanocrystals (CNCs) along and across the microscope optical axis [8].  and dark lines under crossed polars originate from variations in birefringence due to the orientation

2 of the cellulose nanocrystals (CNCs) along and across 2 the microscope optical axis [8].

7874

Materials 2015, 8, 7873–7888 Materials 2015, 8, page–page 

2.1.2.Materials 2015, 8, page–page  Flow Distortion of Fingerprint Texture 2.1.2. Flow Distortion of Fingerprint Texture  As the concentration of CNC is increased, the chiral nematic phase is first observed in a 2.1.2. Flow Distortion of Fingerprint Texture  As the concentration of CNC is increased, the chiral nematic phase is first observed in a biphasic  biphasic region, where the isotropic and chiral nematic phases co-exist [3]. On standing, a sharp 2.1.2. Flow Distortion of Fingerprint Texture  As the concentration of CNC is increased, the chiral nematic phase is first observed in a biphasic  region, where the isotropic and chiral nematic phases co‐exist [3]. On standing, a sharp boundary  boundary forms between the upper isotropic phase and the lower liquid crystalline phase [10,11]. region, where the isotropic and chiral nematic phases co‐exist [3]. On standing, a sharp boundary  As the concentration of CNC is increased, the chiral nematic phase is first observed in a biphasic  forms between the upper isotropic phase and the lower liquid crystalline phase [10,11]. However,  However, fluid samples in sealed microslides are normally rotated from the vertical position to a forms between the upper isotropic phase and the lower liquid crystalline phase [10,11]. However,  region, where the isotropic and chiral nematic phases co‐exist [3]. On standing, a sharp boundary  fluid samples in sealed microslides are normally rotated from the vertical position to a horizontal  horizontal position for viewing with a polarized light microscope, and the resultant sample flow fluid samples in sealed microslides are normally rotated from the vertical position to a horizontal  forms between the upper isotropic phase and the lower liquid crystalline phase [10,11]. However,  position  for  viewing  with  a  polarized  light  microscope,  and  the  resultant  sample  flow  causes  causes distortion of the observed texture, light  as sketched in Figure 3 and illustrated Figure 4. The position  for  viewing  with  a  polarized  microscope,  and  the  resultant  sample inflow  causes  fluid samples in sealed microslides are normally rotated from the vertical position to a horizontal  distortion of the observed texture, as sketched in Figure 3 and illustrated in Figure 4. The effect of the  effectdistortion of the observed texture, as sketched in Figure 3 and illustrated in Figure 4. The effect of the  of the gravity-driven flow at the interface leads to an apparent increase in the spacing of the position  for  viewing  with  a  polarized  light  microscope,  and  the  resultant  sample  flow  causes  gravity‐driven flow at the interface leads to an apparent increase in the spacing of the fingerprint  fingerprint lines, due to the oblique viewing angle. A similar distortion is also observed in the gravity‐driven flow at the interface leads to an apparent increase in the spacing of the fingerprint  distortion of the observed texture, as sketched in Figure 3 and illustrated in Figure 4. The effect of the  lines, due to the oblique viewing angle. A similar distortion is also observed in the bulk of the liquid bulk gravity‐driven flow at the interface leads to an apparent increase in the spacing of the fingerprint  of thelines, due to the oblique viewing angle. A similar distortion is also observed in the bulk of the liquid  liquid crystalline phase (Figure 5). crystalline phase (Figure 5).  Materials 2015, 8, page–page 

crystalline phase (Figure 5).  lines, due to the oblique viewing angle. A similar distortion is also observed in the bulk of the liquid  crystalline phase (Figure 5). 

    Figure 3. Distortion of isotropic‐liquid crystalline phase boundary caused by flow of the sample fluid  Figure 3. Distortion of isotropic-liquid crystalline phase boundary caused by flow of the sample fluid   Figure 3. Distortion of isotropic‐liquid crystalline phase boundary caused by flow of the sample fluid  when placed on the microscope stage. (a) Biphasic sample in glass microslide; (b) Cross‐section of 

when placed on the microscope stage. (a) Biphasic sample in glass microslide; (b) Cross-section of when placed on the microscope stage. (a) Biphasic sample in glass microslide; (b) Cross‐section of  Figure 3. Distortion of isotropic‐liquid crystalline phase boundary caused by flow of the sample fluid  phase boundary. Viewing direction indicated by arrows.  phase boundary. Viewing direction indicated by arrows. phase boundary. Viewing direction indicated by arrows.  when placed on the microscope stage. (a) Biphasic sample in glass microslide; (b) Cross‐section of  phase boundary. Viewing direction indicated by arrows. 

   

Figure  4. Polarized light  image  of  the  isotropic‐chiral  nematic  phase  boundary  for  an  aqueous  6%    Figure  4. Polarized light  image  of  the  isotropic‐chiral  nematic  phase  boundary  for  an  aqueous  6%  CNC suspension (SB‐6%). The isotropic phase (dark under crossed polars) is on the left, the chiral  Figure 4. Polarized light image of the isotropic-chiral nematic phase boundary for an aqueous 6% CNC suspension (SB‐6%). The isotropic phase (dark under crossed polars) is on the left, the chiral  Figure  4. Polarized light  image  of  the  isotropic‐chiral  nematic  phase  boundary  for  an  aqueous  6%  nematic phase is on the right. The sample is oriented as sketched in Figure 3.  CNC suspension (SB-6%). The isotropic phase (dark under crossed polars) is on the left, the chiral nematic phase is on the right. The sample is oriented as sketched in Figure 3.  CNC suspension (SB‐6%). The isotropic phase (dark under crossed polars) is on the left, the chiral 

nematic phase is on the right. The sample is oriented as sketched in Figure 3. nematic phase is on the right. The sample is oriented as sketched in Figure 3. 

    Figure  5.  Polarized  light  microscope  image  (crossed  polars,  wave  plate)  of  6%  aqueous  CNC    of  6%  aqueous  CNC  Figure  5.  Polarized  light  microscope  image  (crossed  polars,  wave  plate)  suspension (SB‐6%), viewed between crossed polars with 530 nm red wave plate. The colors are due  suspension (SB‐6%), viewed between crossed polars with 530 nm red wave plate. The colors are due  Figure  5.  Polarized  light  microscope  image  (crossed  polars,  wave  plate)  of  6%  aqueous  CNC 

to the variations in orientation of the birefringent suspension.  Figure 5. Polarized light microscope image (crossed polars, wave plate) of 6% aqueous CNC to the variations in orientation of the birefringent suspension.  suspension (SB‐6%), viewed between crossed polars with 530 nm red wave plate. The colors are due  suspension (SB-6%), viewed between crossed polars 3 with 530 nm red wave plate. The colors are to the variations in orientation of the birefringent suspension.  3 suspension. due to the variations in orientation of the birefringent 3

7875

Materials 2015, 8, 7873–7888

As the concentration of CNC in the suspensions increases, the chiral nematic pitch should in principle decrease to values of the order of visible light, where iridescent colours should be visible. Materials 2015, 8, page–page  However, before such concentrations are reached, freezing-in of the pitch due to gelation and glass formation may occur [12]. Materials 2015, 8, page–page 

As the concentration of CNC in the suspensions increases, the chiral nematic pitch should in  principle decrease to values of the order of visible light, where iridescent colours should be visible.  2.1.3. Cellulosic Liquid Crystalline Solutions that Reflect Visible Light As the concentration of CNC in the suspensions increases, the chiral nematic pitch should in  However, before such concentrations are reached, freezing‐in of the pitch due to gelation and glass  principle decrease to values of the order of visible light, where iridescent colours should be visible.  formation may occur [12].  While CNC suspensions don’t usually achieve pitch values small enough to reflect visible However, before such concentrations are reached, freezing‐in of the pitch due to gelation and glass  light, molecularly dispersed solutions of cellulose derivatives often show iridescent colours due formation may occur [12].  2.1.3. Cellulosic Liquid Crystalline Solutions that Reflect Visible Light 

to reflection of circularly polarized light of visible wavelengths, although the colours take a long While CNC suspensions don’t usually achieve pitch values small enough to reflect visible light,  time to 2.1.3. Cellulosic Liquid Crystalline Solutions that Reflect Visible Light  form, presumably because the polymer chains take a long time to disentangle and adopt molecularly  dispersed  solutions  of  cellulose  derivatives  often  show  iridescent  colours  due  to  While CNC suspensions don’t usually achieve pitch values small enough to reflect visible light,  chiral nematic order. Thus, for the (hydroxypropyl)cellulose (HPC)-water system [1], the effect of reflection of circularly polarized light of visible wavelengths, although the colours take a long time  molecularly  dispersed  solutions  of  cellulose  derivatives  often  show  iridescent  due  to  polymerto form, presumably because the polymer chains take a long time to disentangle and adopt chiral  concentration on chiral nematic pitch is shown in Figure 6. These imagescolours  were photographed reflection of circularly polarized light of visible wavelengths, although the colours take a long time  by transmitted light between crossed polars. Thus, isotropic regions should be black, birefringent nematic order. Thus, for the (hydroxypropyl)cellulose (HPC)‐water system [1], the effect of polymer  to form, presumably because the polymer chains take a long time to disentangle and adopt chiral  on  chiral  shown  in regions Figure  6.  images  were  photographed  by  colour regions concentration  should be bright andnematic  planar pitch  chiralis nematic inThese  Figure 6a,c,d should display the nematic order. Thus, for the (hydroxypropyl)cellulose (HPC)‐water system [1], the effect of polymer  transmitted  light  between  crossed  polars.  Thus,  isotropic  regions  should  be  black,  birefringent  of circularly polarized light nematic  transmitted byshown  the sample. other hand of photographed  circularly polarized is concentration  on  chiral  pitch  is  in  Figure (The 6.  These  images  were  by  reflectedregions should be bright and planar chiral nematic regions in Figure 6a,c,d should display the colour  from the sample.) The crossed  samplespolars.  appear red isotropic  (Figure 6a), green (Figure 6b) and (Figure 6c) and transmitted  light  between  Thus,  regions  should  be  black,  birefringent  of  circularly  polarized  light  transmitted  by  the  sample.  (The  other  hand  of  circularly  polarized  is  regions should be bright and planar chiral nematic regions in Figure 6a,c,d should display the colour  blue (Figure 6d) to the naked eye, irrespective of the texture. The textures are characteristic of typical reflected from the sample.) The samples appear red (Figure 6a), green (Figure 6b) and (Figure 6c)  of  circularly  polarized  light liquid transmitted  by  the  sample.  (The  other  circularly and polarized  low molar mass chiral nematic crystals, but the samples arehand  veryof viscous, someis textures and blue (Figure 6d) to the naked eye, irrespective of the texture. The textures are characteristic of  reflected from the sample.) The samples appear red (Figure 6a), green (Figure 6b) and (Figure 6c)  require typical low molar mass chiral nematic liquid crystals, but the samples are very viscous, and some  time (weeks) to form. and blue (Figure 6d) to the naked eye, irrespective of the texture. The textures are characteristic of  textures require time (weeks) to form.  Antypical low molar mass chiral nematic liquid crystals, but the samples are very viscous, and some  example of the reflected light from a blue HPC sample illuminated by white light against a An example of the reflected light from a blue HPC sample illuminated by white light against a  black background is shown in Figure 7. In view of the discussion below, it should be noted that the textures require time (weeks) to form.  black background is shown in Figure 7. In view of the discussion below, it should be noted that the  wavelengthAn example of the reflected light from a blue HPC sample illuminated by white light against a  of reflected light, and hence the chiral nematic pitch, at the edge of the sample is smaller wavelength of reflected light, and hence the chiral nematic pitch, at the edge of the sample is smaller  black background is shown in Figure 7. In view of the discussion below, it should be noted that the  than that at the centre of the sample. In fact, the edge of the HPC sample shown in Figure 7 is clear, than that at the centre of the sample. In fact, the edge of the HPC sample shown in Figure 7 is clear,  wavelength of reflected light, and hence the chiral nematic pitch, at the edge of the sample is smaller  suggesting that the reflection wavelength has moved into the ultraviolet region as the sample dries. suggesting that the reflection wavelength has moved into the ultraviolet region as the sample dries.  than that at the centre of the sample. In fact, the edge of the HPC sample shown in Figure 7 is clear,  suggesting that the reflection wavelength has moved into the ultraviolet region as the sample dries. 

  Figure  6.  (Hydroxypropyl)cellulose  (HPC)/water  polarized  light  microscope,  crossed    polars,  polars, Figure 6. (Hydroxypropyl)cellulose (HPC)/water polarized light microscope, crossed transmission images (a) ~45% HPC, planar and focal conic textures; (b) 55% HPC, focal conic texture;  Figure images 6.  (Hydroxypropyl)cellulose  (HPC)/water  polarized  light  (b) microscope,  crossed  polars, texture; transmission (a) ~45% HPC, planar and focal conic textures; 55% HPC, focal conic (c) 55% HPC, oily streak texture; (d) ~65% HPC, planar and focal conic textures. The colours of the  transmission images (a) ~45% HPC, planar and focal conic textures; (b) 55% HPC, focal conic texture;  (c) 55% planar regions between crossed polars correspond to the reflection colours of the samples.  HPC, oily streak texture; (d) ~65% HPC, planar and focal conic textures. The colours of the (c) 55% HPC, oily streak texture; (d) ~65% HPC, planar and focal conic textures. The colours of the  planar regions between crossed polars correspond to the reflection colours of the samples. planar regions between crossed polars correspond to the reflection colours of the samples. 

  Figure 7. Reflection color from a thin 65 wt % solution of HPC in water, between a microscope slide    and  a  22  mm  square  cover  glass.  The  sample  is  illuminated  by  white  light  against  a  black  Figure 7. Reflection color from a thin 65 wt % solution of HPC in water, between a microscope slide  Figure background. Evaporation of water leads to a higher concentration and shorter reflection wavelength at  7. Reflection color from a thin 65 wt % solution of HPC in water, between a microscope and  a  22  mm  square  cover  glass.  The  sample  is  illuminated  by  white  light  against  a  black  edge of sample.  slide and a 22 mm square cover glass. The sample is illuminated by white light against a black background. Evaporation of water leads to a higher concentration and shorter reflection wavelength at  4 concentration and shorter reflection wavelength background. Evaporation of water leads to a higher edge of sample. 

at edge of sample.

4

7876

Materials 2015, 8, 7873–7888 Materials 2015, 8, page–page 

2.2. Solid Cellulosic Films with Chiral Nematic Properties  2.2. Solid Cellulosic Films with Chiral Nematic Properties Cellulose and cellulose derivatives have long been exploited for their film‐forming properties.  Cellulose and cellulose derivatives have long been exploited for their film-forming properties. Examples include  cellophane, the the first first transparent transparent food  wrap, and and cellulose cellulose acetate,  whose use as  Examples include cellophane, food wrap, acetate, whose use as base for photographic films has been replaced by its use as protective and passive optical elements in  base for photographic films has been replaced by its use as protective and passive optical elements liquid  combined  with with their their  in liquidcrystal  crystalflat‐screen  flat-screendisplays.  displays.If Ifthese  thesefilm‐forming  film-formingproperties  properties could  could be  be combined tendency to form chiral nematic solutions and suspensions, the unique optical properties of chiral  tendency to form chiral nematic solutions and suspensions, the unique optical properties of chiral nematic  liquid  crystals would would become become available  in a cheap  easily handled  solid film, film, based based on on an an  nematic liquid crystals available in a cheap easily handled solid essentially renewable  resource. To To date,  the cellulose cellulose derivatives derivatives have  proved difficult, in  general  essentially renewable resource. date, the have proved difficult, in general forming viscous viscous solutions solutions and and solids solids that that are are slow slow toto develop develop chiral chiral nematic nematic organization. organization. CNC CNC  forming suspensions  in  water  are  less  viscous,  and  chiral  nematic  suspensions  and  films  are  more  readily  suspensions in water are less viscous, and chiral nematic suspensions and films are more readily prepared.  However,  films  remains  challenging.  In  this  prepared. However, control  controlof  ofthe  theorientation  orientationand  andthe  theCNC  CNCin indry  dry films remains challenging. In work,  we  focus  on  CNC  films  produced  by  simple  evaporation  of  water  from  a  sessile  droplet  of  this work, we focus on CNC films produced by simple evaporation of water from a sessile droplet of CNC  suspension  on  a  flat  surface.  Hoeger  et  al.  [13]  showed  by  means  of  a  convective‐shear  CNC suspension on a flat surface. Hoeger et al. [13] showed by means of a convective-shear assembly assembly  that capillary shear  and  capillary  forces alignment generated  alignment  parallel  and  normal  to  the  withdrawal  that shear and forces generated parallel and normal to the withdrawal direction direction of a deposition plate, respectively. Both shear and capillary forces might be expected to act  of a deposition plate, respectively. Both shear and capillary forces might be expected to act on the on the orientation of CNC deposited at the edge of an evaporating droplet. It is of interest to test  orientation of CNC deposited at the edge of an evaporating droplet. It is of interest to test which which effect predominates.  effect predominates. 2.2.1. Evaporation of Water from Edge of Covered Chiral Nematic CNC Suspensions 2.2.1. Evaporation of Water from Edge of Covered Chiral Nematic CNC Suspensions  For samples sandwiched between microscope slide and cover glass, evaporation can only occur For samples sandwiched between microscope slide and cover glass, evaporation can only occur  from the edge of the sample, resulting in a concentration gradient decreasing from the edge to the from the edge of the sample, resulting in a concentration gradient decreasing from the edge to the  centre the sample. This gradient willwill  influence the orientation of theof  nanocrystals at the point where centre ofof  the  sample.  This  gradient  influence  the  orientation  the  nanocrystals  at  the  point  they are frozen into the solid. Other factors will also affect the orientation and texture of the film where they are frozen into the solid. Other factors will also affect the orientation and texture of the  produced on drying (Figure(Figure  8). When coverthe  glass on top of aon  droplet microscope film  produced  on  drying  8). placing When  the placing  cover  glass  top  of ona the droplet  on  the  slide (Figureslide  8a), (Figure  the droplet is forced to is  spread forms as  a thin layera  of uniform thickness microscope  8a),  the  droplet  forced asto itspread  it  forms  thin  layer  of  uniform  (Figure 8b).(Figure  This process generates both radial and transverse shear forces on shear  the anisotropic thickness  8b).  This  process  generates  both  radial  and  transverse  forces  on and the  viscoelastic liquid crystalline fluid. Low viscosity samples will be drawn to the edge of the cover anisotropic and viscoelastic liquid crystalline fluid. Low viscosity samples will be drawn to the edge  of  the  glass  by  capillary  forces.  If  the  contact  and  the  slide  and  glass bycover  capillary forces. If the contact lines between thelines  fluidbetween  and the the  slidefluid  and coverglass are not coverglass are not pinned, then the droplet radius may shrink during evaporation (Figure 8c).  pinned, then the droplet radius may shrink during evaporation (Figure 8c).

  Figure 8. Sketch of sample preparation between microscope slide and cover glass. (a) Sessile droplet  Figure 8. Sketch of sample preparation between microscope slide and cover glass. (a) Sessile droplet with  finite  contact  angle  on  microscope  slide;  (b)  Droplet  spreads  under  cover  glass;  (c)  Droplet  with finite contact angle on microscope slide; (b) Droplet spreads under cover glass; (c) Droplet shrinks due to edge evaporation.  shrinks due to edge evaporation.

Complex textures are observed for dry films of 8% CNC suspensions allowed to dry between  Complex textures are observed for dry films of 8% CNC suspensions allowed to dry between microscope slide and cover glass (Figure 9). For this sample and CNC concentration, the diameter of  microscope slide and glass (Figure 9). initial  For thisdroplet  samplebetween  and CNC concentration, diameter of the  dry  film  was  the cover same  as  that  of  the  slide  and  cover the glass.  Thus  the  the dry film was the same as that of the initial droplet between slide and cover glass. Thus the contact contact  lines  between  the  drying  droplet  and  the  glass  surfaces  must  have  been  pinned,  and   the evaporation of water from the edge of the sample resulted in a decrease in thickness only; the  situation sketched in Figure 8c did not apply here. The graininess of the texture was smallest close to  7877 5

Materials 2015, 8, 7873–7888

lines between the drying droplet and the glass surfaces must have been pinned, and the evaporation of water from the edge of the sample resulted in a decrease in thickness only; the situation sketched in Materials 2015, 8, page–page  Figure 8c did not apply here. The graininess of the texture was smallest close to the edge of the sample the edge of the sample (bottom right of the image), increasing towards the centre of the cover glass.  (bottom right of the image), increasing towards the centre of the cover glass. What appear to be Materials 2015, 8, page–page  What appear to be residual areas of distorted fingerprint texture are evident away from the sample  residual areas of distorted fingerprint texture are evident away from the sample centre. This suggests centre.  This  suggests  that  the  relatively  rapid  drying  at  the  edge  of  the  sample  generates  a  the edge of the sample (bottom right of the image), increasing towards the centre of the cover glass.  that the relatively rapid drying at the edge of the sample generates a polydomain structure, but as the polydomain  structure,  but  as  the  process  proceeds,  the  domains  enlarge  until,  at  the  centre,  the  What appear to be residual areas of distorted fingerprint texture are evident away from the sample  process proceeds, the domains enlarge until, at the centre, the sample assumes an essentially planar sample assumes an essentially planar texture, oriented with the chiral nematic axis orthogonal to the  centre.  This  suggests  that  the  relatively  rapid  drying  at  the  edge  of  the  sample  generates  a  texture, oriented with the chiral nematic axis orthogonal to the microscope slide and cover glass. microscope slide and cover glass.  polydomain  structure,  but  as  the  process  proceeds,  the  domains  enlarge  until,  at  the  centre,  the  sample assumes an essentially planar texture, oriented with the chiral nematic axis orthogonal to the  microscope slide and cover glass. 

  Figure  9.  Texture  (polarized  light  microscope,  crossed  polars,  530  nm  red  wave  plate)  of  8  wt  %  

Figure 9. Texture (polarized light microscope, crossed polars, 530 nm red wave plate) of 8 wt % (Batch SB‐8%) CNC suspension on drying between microscope slide and cover glass.    (Batch SB-8%) CNC suspension on drying between microscope slide and cover glass. Figure  9.  Texture  (polarized  light  microscope,  crossed  polars,  530  nm  red  wave  plate)  of  8  wt  %   (Batch SB‐8%) CNC suspension on drying between microscope slide and cover glass. 

2.2.2. Evaporation of Water from Uncovered Droplets of Chiral Nematic CNC Suspensions 

2.2.2. Evaporation of Water from Uncovered Droplets of Chiral Nematic CNC Suspensions

The edge‐wise evaporation of water from a sample of CNC suspension between glass slide and 

2.2.2. Evaporation of Water from Uncovered Droplets of Chiral Nematic CNC Suspensions  cover glass is slow and impractical on a larger scale. What happens if the droplet is left uncovered?  The edge-wise evaporation of water from a sample of CNC suspension between glass slide and Aqueous droplets containing low concentrations of rod‐like CNC have been reported to display a  cover glass is slow and impractical on a larger scale. What happens if the droplet is left uncovered? The edge‐wise evaporation of water from a sample of CNC suspension between glass slide and  nematic‐like arrangement at the edge of the ring [14]. The orientation of CNC in the residue deposited  cover glass is slow and impractical on a larger scale. What happens if the droplet is left uncovered?  Aqueous droplets containing low concentrations of rod-like CNC have been reported to display a on  a  glass  slide  by  the  evaporation  of  a  dilute  suspension  is  readily  observed  by  polarized  light  Aqueous droplets containing low concentrations of rod‐like CNC have been reported to display a  nematic-like arrangement at the edge of the ring [14]. The orientation of CNC in the residue deposited microscopy  with  530  nm  red  plate  (Figure  10).  The  530  nm  red  plate  generates  a  red  colour  for  nematic‐like arrangement at the edge of the ring [14]. The orientation of CNC in the residue deposited  on a glass slide by the evaporation of a dilute suspension is readily observed by polarized light isotropic material and blue and yellow colours for birefringent material oriented as indicated in the  on  a  glass  slide  by nm the red evaporation  of  a  dilute  suspension  is  readily  observed  by  polarized  light  for microscopy with 530 platebecause  (Figure The 530 nm redleft  plate generates a red colour sketches.  The  colours  are  red  faint,  of 10). the  suspension  a generates  thin  film  on  the  glass.  The  microscopy  with  530  nm  plate  (Figure  10). dilute  The  530  nm  red  plate  a  red  colour  for  isotropic material and blue and yellow colours for birefringent material oriented as indicated quadrants  occupied  by  the  blue  and  yellow  colours  indicate  that  the  orientation  direction  of  the  in isotropic material and blue and yellow colours for birefringent material oriented as indicated in the  the sketches. The colours are faint, because of the dilute suspension left a thin film on the glass. nanocrystals deposited by the evaporating droplet is radial (parallel to the edge of the droplet).  sketches.  The  colours  are  faint,  because  of  the  dilute  suspension  left  a  thin  film  on  the  glass.  The  The quadrants occupiedby  bythe  theblue  blueand  andyellow  yellowcolours  colours indicate orientation direction of the quadrants  occupied  indicate  that that the the orientation  direction  of  the  nanocrystals deposited by the evaporating droplet is radial (parallel to the edge of the droplet). nanocrystals deposited by the evaporating droplet is radial (parallel to the edge of the droplet). 

  Figure  10.  Polarized  light  image  (crossed  polars,  530  nm  red  wave  plate)  of  dry  film  left  on  glass  microscope  slide  by  evaporating  droplet  of  dilute  (initially  0.52  wt  %)  aqueous    CNC  suspension.   Two  segments  of  the  disc‐shaped  residue  are  shown.  The  blue  and  yellow  colours  observed  with  the   Figure  10.  Polarized  light  image  (crossed  polars,  530  nm  red  wave  plate)  of  dry  film  left  on  glass  Figure 10. Polarized light image (crossed polars, 530 nm red wave plate) of dry film left on glass 530 nm red wave plate [15] indicate that the birefringent CNC are oriented radially, as indicated in  microscope  slide  by  evaporating  droplet  of  dilute  (initially  0.52  wt  %)  aqueous  CNC  suspension.   microscope slide by evaporating droplet of dilute (initially 0.52 wt %) aqueous CNC suspension. the lower sketches.  Two  segments  of  the  disc‐shaped  residue  are  shown.  The  blue  and  yellow  colours  observed  with  the   Two segments of the disc-shaped residue are shown. The blue and yellow colours observed with 530 nm red wave plate [15] indicate that the birefringent CNC are oriented radially, as indicated in  Evaporation of suspensions above the critical concentration for liquid crystal formation showed  the 530 nm red wave plate [15] indicate that the birefringent CNC are oriented radially, as indicated the lower sketches. 

a more dramatic effect. Preliminary observations of the evaporation of droplets starting from CNC  in the lower sketches. concentrations high enough to form chiral nematic suspensions showed the formation of films with  Evaporation of suspensions above the critical concentration for liquid crystal formation showed  6 a more dramatic effect. Preliminary observations of the evaporation of droplets starting from CNC  concentrations high enough to form chiral nematic suspensions showed the formation of films with 

7878 6

Materials 2015, 8, page–page 

a remarkable 3‐Dimensional iridescent pattern [16]. The contact line at the edge of the droplet was  Materials 2015, 8, 7873–7888 pinned, and the CNC in colloidal suspension were transported during evaporation to the outer edge  of  the  droplet  by  the  so‐called  “coffee‐stain”  effect.  Here  we for examine  the formation evaporation  process  in  Evaporation of suspensions above the critical concentration liquid crystal showed a more dramatic effect. Preliminary observations of the evaporation of droplets starting from CNC more detail, focusing on the textures observed by polarized light microscopy and on the topology of  concentrations high enough to form chiral nematic suspensions showed the formation of films with the films observed by profilometry and atomic force microscopy.  a remarkable 3-Dimensional iridescent pattern [16]. The contact line at the edge of the droplet was The drying process initially causes a reduction in chiral nematic pitch, if for no other reason that  pinned, and the CNC in colloidal suspension were transported during evaporation to the outer edge the concentration of chiral centres increases with increasing concentration. Thus, there are no reports  of the droplet by the so-called “coffee-stain” effect. Here we examine the evaporation process in more of  the  reflection  of  circularly  polarized  light  from  fluid  CNC  suspensions,  but  iridescence  detail, focusing on the textures observed by polarized light microscopy and on the topology of the is  often  films observed by profilometry and atomic force microscopy. observed for films made from such suspensions [17–21].  The drying process initially causes a reduction in chiral nematic if for other reason An  example  of  the  texture  of  an  iridescent  film  is  shown  in pitch, Figure  11. noThe  reflection  band  that the concentration of chiral centres increases with increasing concentration. Thus, there are no colours range from blue to red, as indicated on the insert, although the colours towards the edge of  reports of the reflection of circularly polarized light from fluid CNC suspensions, but iridescence is the droplet are less intense due to the thickening and misalignment of the sample due to the coffee‐stain  often observed for films made from such suspensions [17–21]. effect [16]. The dark lines are disclination lines at the edge of regions of essentially planar orientation.  An example of the texture of an iridescent film is shown in Figure 11. The reflection band colours No  fingerprint  lines  present,  because  chiral although nematic  pitch  for towards the  sample  is  in  the  visible  range from blueare  to red, as indicated on the  the insert, the colours the edge of the droplet are less intense due to the thickening and misalignment of the sample due to the coffee-stain region. Note that a wave plate was not used to generate the main image in Figure 11; the colours  effect [16]. The dark lines are disclination lines at the edge of regions of essentially planar orientation. observed in transmission match those of the reflection band, as explained for the much more intense  No fingerprint lines are present, because the chiral nematic pitch for the sample is in the visible region. colours shown by the (hydroxypropyl)cellulose/water system (Figure 6).  Note that a wave plate was not used to generate the main image in Figure 11; the colours observed The in freely  dried  droplets  another  difference  in  that  the  transmission match those ofshow  the reflection band, as explainedfrom  for theedge  much evaporation,  more intense colours progression of colours from edge to centre is reversed. For covered droplets, evaporation from the  shown by the (hydroxypropyl)cellulose/water system (Figure 6). The freely dried droplets show another difference from edge evaporation, in that the progression edge leads to a higher concentration and hence a shorter pitch at the edge, as shown for example in  of colours from edge to centre is reversed. For covered droplets, evaporation from the edge leads to Figure 7. For freely dried droplets, as mentioned above, the longer wavelength reflections are near  a higher concentration and hence a shorter pitch at the edge, as shown for example in Figure 7. the edge. A similar trend has been observed by Lagerwall et al. [4] and Dumali et al. [22]. However, for  For freely dried droplets, as mentioned above, the longer wavelength reflections are near the edge. vacuum‐dried CNC films, the colours were blue at edge and red at centre [20]; evidently the rate of  A similar trend has been observed by Lagerwall et al. [4] and Dumali et al. [22]. However, for drying influences the colour sequence.  vacuum-dried CNC films, the colours were blue at edge and red at centre [20]; evidently the rate of drying influences the colour sequence. Even where iridescent reflection is not observed, the pitch values, indicated by fingerprint line  Even where iridescent reflection is not observed, the pitch values, indicated by fingerprint line spacings (P/2) are much reduced on drying (Figure 12). In solid films, the line spacing is only visible  spacings (P/2) are much reduced on drying (Figure 12). In solid films, the line spacing is only visible over small areas of the film, and the spacing often varies with location because the chiral nematic  over small areas of the film, and the spacing often varies with location because the chiral nematic axis axis  is  not  in inthe  of the the  film  so  that  the  apparent  line isspacing  is P/2 larger  due  to  the  is not theplane  plane of film so that the apparent line spacing larger than due than  to the P/2  oblique oblique viewing angle.  viewing angle.

  Figure  11. Polarized light image (transmission, between crossed polars) of film cast from 5.2 wt %  Figure 11. Polarized light image (transmission, between crossed polars) of film cast from 5.2 wt % (SA-5.2%) aqueous CNC suspension on glass. The insert shows a reflection image of a similar droplet (SA‐5.2%) aqueous CNC suspension on glass. The insert shows a reflection image of a similar droplet  dried on a black acrylic surface. The observed iridescent region is indicated by the arrow on the insert. dried on a black acrylic surface. The observed iridescent region is indicated by the arrow on the insert. 

7879

  Figure  11. Polarized light image (transmission, between crossed polars) of film cast from 5.2 wt %  (SA‐5.2%) aqueous CNC suspension on glass. The insert shows a reflection image of a similar droplet  Materials 2015, 8, 7873–7888 dried on a black acrylic surface. The observed iridescent region is indicated by the arrow on the insert. 

  Figure 12. 12. Fingerprint Fingerprint  textures,  polarized  light  microscope,  crossed  polars.  (a) wt 5.2  %  aqueous  Figure textures, polarized light microscope, crossed polars. (a) 5.2 % wt  aqueous CNC CNC suspension (Batch SA‐5.2%); (b) Film cast from same suspension.  suspension (Batch SA-5.2%); (b) Film cast from same suspension.

7 2.2.3.Materials 2015, 8, page–page  Evaporation of Water from Droplets of Chiral Nematic Suspensions: Stick-Slip Textures Materials 2015, 8, page–page 

With a new batch of CNC, some novel textures were observed with the polarized light 2.2.3. Evaporation of Water from Droplets of Chiral Nematic Suspensions: Stick‐Slip Textures  microscope. A distinct banded structure was observed near the edge of the film (Figure 13). At 2.2.3. Evaporation of Water from Droplets of Chiral Nematic Suspensions: Stick‐Slip Textures  With  a  new  batch  of  CNC,  some  novel  textures  were  observed  with  the  polarized  light  the very edge of the film (bottom right, next to the red area indicating the isotropic glass support) With  a  new  batch  of  CNC,  some  novel  textures  were  observed  with  the  polarized  light  microscope. A distinct banded structure was observed near the edge of the film (Figure 13). At the  a thinvery edge of the film (bottom right, next to the red area indicating the isotropic glass support) a thin  blue band shows that the nanocrystals in this region are oriented radially. Moving towards microscope. A distinct banded structure was observed near the edge of the film (Figure 13). At the  the centre of the sample distinct radial bands with more complex textures. Atthe  higher very edge of the film (bottom right, next to the red area indicating the isotropic glass support) a thin  blue  band  shows  that are the several nanocrystals  in  this  region  are  oriented  radially.  Moving  towards  magnification, these radial bands, of the order of 100 µm wide, are themselves seen to be made blue  band  shows  that are  the several  nanocrystals  in  radial  this  region  are  oriented  towards  the  up centre  of  the  sample  distinct  bands  with  more radially.  complex Moving  textures.  At  higher  centre  of  the  sample  are  several  complex  At  higher  magnification, these radial bands, of the order of 100 μm wide, are themselves seen to be made up of  of more or less radially aligned wavydistinct  featuresradial  aboutbands  2 µmwith  widemore  extending fortextures.  up to several hundred magnification, these radial bands, of the order of 100 μm wide, are themselves seen to be made up of  more or less radially aligned wavy features about 2 μm wide extending for up to several hundred  micron (Figure 14). Their appearance suggests a distorted fingerprint pattern. Towards the centre micron (Figure 14). Their appearance suggests a distorted fingerprint pattern. Towards the centre of  of themore or less radially aligned wavy features about 2 μm wide extending for up to several hundred  samples, the texture becomes more uniform, the orientation of the features becomes more micron (Figure 14). Their appearance suggests a distorted fingerprint pattern. Towards the centre of  the  samples,  the  texture  becomes  more  uniform,  the  orientation  of becomes the  features  more  random, and the spacing becomes larger. The radial banded structure morebecomes  apparent at some the  samples,  texture  becomes  larger.  more  uniform,  the  orientation  of  the  features  becomes  more  random,  and  the  spacing  The  radial  banded  structure  becomes  more  apparent  at  locations on the film, where intense dark lines appear between the rings (Figure 15). The complex random,  and  the  spacing  becomes  radial  apparent  at  some  locations  on  the  film,  where larger.  intense The  dark  lines banded  appear structure  between becomes  the  rings more  (Figure  15).  The  nature of these images the  generated by the stick-slip evaporation process the  shows that the orientation of some  locations  film,  where  intense  dark  lines  appear evaporation  between  rings  (Figure  complex  nature on  of  these  images  generated  by  the  stick‐slip  process  shows  15).  that The  the  the nanocrystals is far from a simple chiral nematic planar or fingerprint texture. complex  nature  of  these  images  generated  by  the  stick‐slip  evaporation  process  shows  that  the  orientation of the nanocrystals is far from a simple chiral nematic planar or fingerprint texture.  orientation of the nanocrystals is far from a simple chiral nematic planar or fingerprint texture. 

    Figure  13. Polarized light image (crossed polars, 530 nm red wave plate) of film cast from 6 wt % 

Figure 13. Polarized light image (crossed polars, 530 nm red wave plate) of film cast from 6 wt % CNC Figure  13. Polarized light image (crossed polars, 530 nm red wave plate) of film cast from 6 wt %  CNC suspension (SB‐6%) on glass.  suspension (SB-6%) on glass. CNC suspension (SB‐6%) on glass. 

    Figure  14. Polarized light image (crossed polars, 530 nm red wave plate) of film cast from 6 wt %  Figure  14. Polarized light image (crossed polars, 530 nm red wave plate) of film cast from 6 wt %  CNC suspension (SB‐6%) on glass showing banded texture at higher magnification.  Figure 14. Polarized light image (crossed polars, 530 nm red wave plate) of film cast from 6 wt % CNC CNC suspension (SB‐6%) on glass showing banded texture at higher magnification. 

suspension (SB-6%) on glass showing banded texture at higher magnification.

7880

  Figure  Materials 2015, 14. Polarized light image (crossed polars, 530 nm red wave plate) of film cast from 6 wt %  8, 7873–7888 CNC suspension (SB‐6%) on glass showing banded texture at higher magnification. 

  Figure  15. Polarized light image (crossed polars, 530 nm red wave plate) of film cast from 6 wt %  Figure 15. Polarized light image (crossed polars, 530 nm red wave plate) of film cast from 6 wt % CNC CNC suspension (SB‐6%) on glass showing region with cracks between bands.  suspension (SB-6%) on glass showing region with cracks between bands. Materials 2015, 8, page–page 

8

2.2.4. Profilometry of Film Topography 2.2.4. Profilometry of Film Topography  To interpret these images, profilometer scansscans  and atomic force microscopy (AFM) images To try try and and  interpret  these  images,  profilometer  and  atomic  force  microscopy  (AFM)  of the surface of the dried droplets were obtained. As a first observation by means of optical images of the surface of the dried droplets were obtained. As a first observation by means of optical  profilometry, there appear to be concentric rings of material, with raised edges. However, the rings profilometry, there appear to be concentric rings of material, with raised edges. However, the rings  appeared to be discontinuous, and the sample signal was very noisy. It seemed that the oriented and appeared to be discontinuous, and the sample signal was very noisy. It seemed that the oriented and  birefringent nature of the CNC film was interfering with the optical interferometric measurements birefringent nature of the CNC film was interfering with the optical interferometric measurements  used instrument. ToTo  avoid this this  problem, samples were coated with gold to provide optically used by by the the  instrument.  avoid  problem,  samples  were  coated  with  gold  to  an provide  an  homogeneous surface, and the samples were scanned with the optical profilometer (Figure 16). In this optically homogeneous surface, and the samples were scanned with the optical profilometer (Figure 16).  case a clearer topography emerges, with rings of increasing thickness towards the centre of the film. In this case a clearer topography emerges, with rings of increasing thickness towards the centre of  The edges of the rings are often raised, and clearly the films contain radial cracks (coloured blue) that the film. The edges of the rings are often raised, and clearly the films contain radial cracks (coloured  show the level of the underlying glass surface. blue) that show the level of the underlying glass surface 

  Figure 16. Optical profilometry of gold‐coated CNC film. Perspective view showing increasing step  Figure 16. Optical profilometry of gold-coated CNC film. Perspective view showing increasing step heights towards the centre of the film. Blue regions indicate cracks in the film.  heights towards the centre of the film. Blue regions indicate cracks in the film.

The heights of steps and cracks are shown in the X and Y profiles across the film (Figure 17) are  The heights of steps and cracks are shown in the X and Y profiles across the film (Figure 17) are of the order of 2 μm, which is much smaller than the >50 μm height difference previously observed  of the order of 2 µm, which is much smaller than the >50 µm height difference previously observed at at  the  edge  of  droplets  with  pinned  contact  lines  [16].  However,  the  profilometry  shows  that  the  the edge of droplets with pinned contact lines [16]. However, the profilometry shows that the outer outer edge of each step is also raised, suggesting that a minor “coffee‐stain effect” is also occurring  edge of each step is also raised, suggesting that a minor “coffee-stain effect” is also occurring during during  evaporation.  Thus  a  multiple  “stick‐slip”  situation  is  occurring  in  this  case,  as  sketched  in  evaporation. Thus a multiple “stick-slip” situation is occurring in this case, as sketched in Figure 18. Figure  18.  The  radial  nature  of  the  cracks  observed  in  Figures  15–17  indicates  that  there  exists  an  The radial nature of the cracks observed in Figures 15–17 indicates that there exists an anisotropy in anisotropy  in  mechanical  properties  of  the  dry  films  that  must  reflect  the  local  orientation  of  the  mechanical properties of the dry films that must reflect the local orientation of the CNC. As with any CNC. As with any fibrous solid, the strong direction of films would be expected to be aligned along  fibrous solid, the strong direction of films would be expected to be aligned along the direction of the the direction of the rod‐like CNC, with the weakest direction normal to the CNC orientation. The  rod-like CNC, with the weakest direction normal to the CNC orientation. The radial alignment of the radial alignment of the CNC thus favors radial cracking during drying.  CNC thus favors radial cracking during drying.

7881

during  evaporation.  Thus  a  multiple  “stick‐slip”  situation  is  occurring  in  this  case,  as  sketched  in  Figure  18.  The  radial  nature  of  the  cracks  observed  in  Figures  15–17  indicates  that  there  exists  an  anisotropy  in  mechanical  properties  of  the  dry  films  that  must  reflect  the  local  orientation  of  the  CNC. As with any fibrous solid, the strong direction of films would be expected to be aligned along  the direction of the rod‐like CNC, with the weakest direction normal to the CNC orientation. The  Materials 2015, 8, 7873–7888 radial alignment of the CNC thus favors radial cracking during drying. 

  Figure  17.  Optical  profilometry  of  gold‐coated  CNC  film.  (a)  Topographic  image,  showing  ridges  Figure 17. Optical profilometry of gold-coated CNC film. (a) Topographic image, showing ridges (yellow) and cracks (black); (b) Corresponding X profile; (c) Y profile across image.   (yellow) and cracks (black); (b) Corresponding X profile; (c) Y profile across image. Materials 2015, 8, page–page 

9

  Figure 18. Sketch of droplet evaporation under (a) pinned and (b) stick‐slip conditions; (c) Multiple  Figure 18. Sketch of droplet evaporation under (a) pinned and (b) stick-slip conditions; (c) Multiple coffee‐stain deposition stages are shown, with rings of increasing thickness and raised edges.  coffee-stain deposition stages are shown, with rings of increasing thickness and raised edges.

The contrast between the relatively high ring of CNC deposited at the edge of the evaporating  The contrast between the relatively high ring of CNC deposited at the edge of the evaporating droplet for a pinned contact line [16] and the much more diffuse effect noted above suggests that the  droplet for a pinned contact line [16] and the much more diffuse effect noted above suggests that tendency  for  contact  line  pinning  is  controlling  the  texture.  Accordingly,  we  tried  to  decrease  the  the tendency for contact line pinning is controlling the texture. Accordingly, we tried to decrease pinning  tendency  by  modifying  the  suspension  with  a  suitable  additive.  Small  amounts  of  a  the pinning tendency by modifying the suspension with a suitable additive. Small amounts of a water‐soluble  compound,  propane‐1,2‐diol  (propylene  glycol,  PG),  were  added  to  the  evaporating  water-soluble compound, propane-1,2-diol (propylene glycol, PG), were added to the evaporating droplet.  PG  has  a  lower  surface  tension  and  vapour  pressure  than  water,  and  so  should  become  droplet. PG has a lower surface tension and vapour pressure than water, and so should become enriched near the contact line during evaporation. The presence of the PG rich fluid at the contact  enriched near the contact line during evaporation. The presence of the PG rich fluid at the contact line line inhibits contact line pinning during evaporation. Figure 19 shows a polarized light image of the  inhibits contact line pinning during evaporation. Figure 19 shows a polarized light image of the film film prepared with PG; it shows none of the banding shown by PG‐free suspensions (Figures 16–18),  prepared with PG; it shows none of the banding shown by PG-free suspensions (Figures 16–18), and and while dark lines attributed to cracking do occur, they are much closer together than in the case  while dark lines attributed to cracking do occur, they are much closer together than in the case films films prepared from PG‐free suspensions. Thus the presence of the PG, presumably as a layer on the  prepared from PG-free suspensions. Thus the presence of the PG, presumably as a layer on the surface surface of the glass, decreases the tendency for contact line pinning. Recently, dramatic effects on  of the glass, decreases the tendency for contact line pinning. Recently, dramatic effects on contact contact line pinning and droplet mobility have been observed for water‐PG droplets on high‐energy  line pinning and droplet mobility have been observed for water-PG droplets on high-energy glass glass  surfaces  [23].  While  the  glass  surfaces  that  we  used  had  not  been  subjected  to  the  rigorous  surfaces [23]. While the glass surfaces that we used had not been subjected to the rigorous procedure procedure used by Cira et al., nevertheless addition of low volatility low surface tension components  used by Cira et al., nevertheless addition of low volatility low surface tension components such as such  as  PG  to  CNC  suspensions  would  be  expected  to  alter  contact  line  pinning  and  hence  the  PG to CNC suspensions would be expected to alter contact line pinning and hence the orientational orientational order of the resulting film.  order of the resulting film.

7882

surface of the glass, decreases the tendency for contact line pinning. Recently, dramatic effects on  contact line pinning and droplet mobility have been observed for water‐PG droplets on high‐energy  glass  surfaces  [23].  While  the  glass  surfaces  that  we  used  had  not  been  subjected  to  the  rigorous  procedure used by Cira et al., nevertheless addition of low volatility low surface tension components  such  as 2015, PG 8,to  CNC  suspensions  would  be  expected  to  alter  contact  line  pinning  and  hence  the  Materials 7873–7888 orientational order of the resulting film. 

  Figure 19. Polarized light microscope (crossed polars, 530 nm wave plate) of film cast from 8 wt %  Figure 19. Polarized light microscope (crossed polars, 530 nm wave plate) of film cast from 8 wt % CNC suspension (SB‐8%) containing 2 wt % added propane‐1,2‐glycol (PG).  CNC suspension (SB-8%) containing 2 wt % added propane-1,2-glycol (PG).

2.3. Atomic Force Microscopy of CNC Orientation and Topography in Dry Films  2.3. Atomic Force Microscopy of CNC Orientation and Topography in Dry Films The  profilometric  measurements  reported  above  demonstrate  the  large  scale  deviations  from  The profilometric measurements reported above demonstrate the large scale deviations from planarity  generated  by  colloidal  forces  during  evaporation  from  anisotropic  suspensions  of  CNC.   planarity generated by colloidal forces during evaporation from anisotropic suspensions of CNC. At  much  higher  magnifications,  the  orientations  of  individual  nanocrystals  in  solid  films  may  be  At much higher magnifications, the orientations of individual nanocrystals in solid films may be observed.  For  films  cast  from  dilute  CNC  suspensions,  below  the  critical  concentration  for  liquid  observed. For films cast from dilute CNC suspensions, below the critical concentration for liquid crystal formation, the orientation of the nanocrystals is essential random (Figure 20), despite the fact  crystal formation, the orientation of the nanocrystals is essential random (Figure 20), despite the that  the  suspension  must  pass  through  the  concentration  range  where  chiral  nematic  phase  is  fact that the suspension must pass through the concentration range where chiral nematic phase is expected.  The  ordering  phenomenon  is  presumably  slower  than  evaporation,  or  alternatively  expected. The ordering phenomenon is presumably slower than evaporation, or alternatively gelation gelation occurs below the critical concentration for liquid crystal formation for this sample.  occurs below the critical concentration for liquid crystal formation for this sample. Materials 2015, 8, page–page  10

  Figure  20. 20.  Atomic  of  Figure Atomic force  force microscope  microscope (AFM)  (AFM) height  height image  image of  of surface  surface of  of film  film cast  cast from  from droplet  droplet of isotropic (0.0032 wt %) CNC suspension.  isotropic (0.0032 wt %) CNC suspension.

The surface orientation of CNC in films cast from suspensions above the critical concentration  The surface orientation of CNC in films cast from suspensions above the critical concentration for for  chiral  nematic  phase  formation is  quite  different.  Atomic force  microscope  (AFM)  images  of a  chiral nematic phase formation is quite different. Atomic force microscope (AFM) images of a typical typical  area  of  a  film  cast  from  an  8  wt  %  CNC  suspension  on  a  mica  substrate  are  shown  in   area of a film cast from an 8 wt % CNC suspension on a mica substrate are shown in Figure 21. The Figure 21.The amplitude mode image (a) shows that the CNC are fairly uniformly oriented parallel  amplitude mode image (a) shows that the CNC are fairly uniformly oriented parallel to the edge to the edge of the sample. The height mode image (b) shows that the surface is uneven, with ridges  of the sample. The height mode image (b) shows that the surface is uneven, with ridges of the of the order of 10 nm high and around 2–3 μm apart, also oriented parallel to the edge of the sample.  order of 10 nm high and around 2–3 µm apart, also oriented parallel to the edge of the sample. A A similar periodic relief at the surface of a chiral nematic oligomer has been observed by AFM, and  similar periodic relief at the surface of a chiral nematic oligomer has been observed by AFM, and interpreted as relating to the half‐helical pitch [24]. Out‐of‐plane AFM deformations that correspond  interpreted as relating to the half-helical pitch [24]. Out-of-plane AFM deformations that correspond to  optical  observations  on  dry  CNC  films  have  been  observed  for  parabolic  focal  conic  and  to optical observations on dry CNC films have been observed for parabolic focal conic and fingerprint fingerprint  textures  [25].  The  spacing  of  the  ridges  observed  by  AFM  is  of  the  same  order  of  textures [25]. The spacing of the ridges observed by AFM is of the same order of magnitude as the magnitude  as  the  line  spacings  of  the  fingerprint  pattern  observed  by  optical  microscopy.  It  is  line spacings of the fingerprint pattern observed by optical microscopy. It is reasonable to assume reasonable to assume that the anisotropic orientation of the CNC in the film is the cause.  that the anisotropic orientation of the CNC in the film is the cause. A possible explanation is sketched in Figure 22. The orientation of CNC in successive adjacent  A possible explanation is sketched in Figure 22. The orientation of CNC in successive adjacent nematic‐like layers of a chiral nematic structure close to an open surface is shown in the upper series  nematic-like layers of a chiral nematic structure close to an open surface is shown in the upper series of cross‐sections. The chiral nematic axis is taken to be perpendicular to the page, and the sketches  cover one half of the chiral nematic pitch. Assume that the concentration of CNC is sufficient to form  a gel. The gel properties will depend on the orientation of the CNC; in particular, gels will shrink  7883 more in directions  perpendicular  to  the  CNC  and less  in  directions  parallel  to  the  CNC.  Thus  the  reduction  in  volume  will  be  anisotropic,  resulting  in  a  periodicity  of  P/2  in  height  along  the  helicoidal  axis,  where  P  is  the  pitch  in  the  final  dry  film.  This  type  of  anisotropic  swelling  and 

interpreted as relating to the half‐helical pitch [24]. Out‐of‐plane AFM deformations that correspond  to  optical  observations  on  dry  CNC  films  have  been  observed  for  parabolic  focal  conic  and  fingerprint  textures  [25].  The  spacing  of  the  ridges  observed  by  AFM  is  of  the  same  order  of  magnitude  as  the  line  spacings  of  the  fingerprint  pattern  observed  by  optical  microscopy.  It  is  reasonable to assume that the anisotropic orientation of the CNC in the film is the cause.  Materials 2015, 8, 7873–7888 A possible explanation is sketched in Figure 22. The orientation of CNC in successive adjacent  nematic‐like layers of a chiral nematic structure close to an open surface is shown in the upper series  of cross-sections. The chiral nematic axis is taken to be perpendicular to the page, and the sketches of cross‐sections. The chiral nematic axis is taken to be perpendicular to the page, and the sketches  cover one half of the chiral nematic pitch. Assume that the concentration of CNC is sufficient to form cover one half of the chiral nematic pitch. Assume that the concentration of CNC is sufficient to form  aa gel. The gel properties will depend on the orientation of the CNC; in particular, gels will shrink  gel. The gel properties will depend on the orientation of the CNC; in particular, gels will shrink more in directions perpendicular less in more in directions  perpendicular to to  the the  CNC CNC  and and less  in  directions directions  parallel parallel  to to  the the  CNC. CNC.  Thus Thus  the the  reduction in volume will be anisotropic, resulting in a periodicity of P/2 in height along the helicoidal reduction  in  volume  will  be  anisotropic,  resulting  in  a  periodicity  of  P/2  in  height  along  the  axis, whereaxis,  P is where  the pitch inthe  thepitch  final in  drythe  film. This ofThis  anisotropic andswelling  shrinkage is helicoidal  P  is  final  dry type film.  type  of swelling anisotropic  and  well-known on a larger length scale; the orientation of the cellulose microfibrils along the secondary shrinkage is well‐known on a larger length scale; the orientation of the cellulose microfibrils along  wall of wood pulp fibres causes them to shrink on drying much more laterally than longitudinally, the secondary wall of wood pulp fibres causes them to shrink on drying much more laterally than  with important consequences for the bonding and properties of paper sheet [26]. longitudinally, with important consequences for the bonding and properties of paper sheet [26]. 

  Figure 21. AFM (a) amplitude and (b) height images of an area of film cast from droplet of 8 wt %  Figure 21. AFM (a) amplitude and (b) height images of an area of film cast from droplet of 8 wt % aqueous  CNC  suspension  (SB‐8%)  on  a  mica  substrate.  The  white  arrow  (a)  indicates  the  aqueous CNC suspension (SB-8%) on a mica substrate. The white arrow (a) indicates the approximate approximate  orientation  of  the  nanocrystals;  the  black  arrows  (b)  indicate  the  approximate  orientation of the nanocrystals; the black arrows (b) indicate the approximate orientation and spacing orientation and spacing of ridged areas of the sample. The height colour scale in (b) is from −19 to +19 nm.  of ridged areas of the sample. The height colour scale in (b) is from ´19 to +19 nm. Materials 2015, 8, page–page 

11

  Figure 22. 22.  (a) The  sketches show  the  orientation orientation  of of CNC CNC in  sequential sections sections of of aa chiral chiral nematic nematic  Figure (a) The sketches show the in sequential helicoid with axis normal to the plane of the page. (In other words, the orientations shown side by  helicoid with axis normal to the plane of the page. (In other words, the orientations shown side side should be envisaged as being one behind the other, with the sketch on the left being at the front.)  by side should be envisaged as being one behind the other, with the sketch on the left being at the The CNC concentration is assumed high enough for gel formation, and the free surface is at the top  front.) The CNC concentration is assumed high enough for gel formation, and the free surface is at of the sketches. On evaporation (b) the tendency of the CNC gel to shrink to a greater extent parallel  the top of the sketches. On evaporation (b) the tendency of the CNC gel to shrink to a greater extent to the CNC orientation may lead to a surface ripple with a periodicity of P/2 along the chiral nematic axis.   parallel to the CNC orientation may lead to a surface ripple with a periodicity of P/2 along the chiral nematic axis.

This  model  is  clearly  oversimplified;  the  drying  process  will  generate  in‐plane  as  well  as  out‐of‐plane shrinkage and the chiral nematic axis may well be at an angle to the evaporating surface.  This model is clearly oversimplified; the drying process will generate in-plane as well as However, where clear fingerprint patterns are detected optically near a free surface, a corresponding  out-of-plane shrinkage and the chiral nematic axis may well be at an angle to the evaporating surface. surface  topology  may  be  expected.  The  effect  resembles  the  theoretical  predictions  of  nanoscale  However, where clear fingerprint patterns are detected optically near a free surface, a corresponding undulations and disclination lines for appropriately anchored chiral nematic liquid crystals, driven  surface topology may be expected. The effect resembles the theoretical predictions of nanoscale by  director  surface  gradients  and  elastic  constants  [27].  However,  the  surface  distortion  for  CNC  undulations and disclination lines for appropriately anchored chiral nematic liquid crystals, driven films is driven by the anisotropic shrinkage of a chiral nematic gel.  by director surface gradients and elastic constants [27]. However, the surface distortion for CNC films is driven by the anisotropic shrinkage of a chiral nematic gel. 3. Experimental Section  3.1. CNC Preparation and Purification  All  the  CNC  suspensions  were  prepared 7884 as  described  previously  [28].  Briefly,  40  g  of  dried  Whatman cotton powder was reacted with a preheated sulfuric acid solution (64% w/w, 700 mL) at  45  °C  for 45 min, after  which  the reaction  was  quenched  by  5  L of  deionized water (18.2 MΩ∙cm,  Millipore Milli‐Q Purification System, EMD Millipore, Billerica, MA, USA) and left to settle for 2 h. 

Materials 2015, 8, 7873–7888

3. Experimental Section 3.1. CNC Preparation and Purification All the CNC suspensions were prepared as described previously [28]. Briefly, 40 g of dried Whatman cotton powder was reacted with a preheated sulfuric acid solution (64% w/w, 700 mL) at 45 ˝ C for 45 min, after which the reaction was quenched by 5 L of deionized water (18.2 MΩ¨ cm, Millipore Milli-Q Purification System, EMD Millipore, Billerica, MA, USA) and left to settle for 2 h. The suspension was repeatedly concentrated and washed by centrifugation (6000 rpm, 10 min cycles) until the supernatant appeared turbid, the suspension was then extensively dialysized against deionized water until the effluent remained at neutral pH (Spectra/Por 4, Molecular weight cut-off 12,000–14,000 Daltons). Following the removal of excess acid, the suspensions were dispersed by sonication using a Vibracel sonicator (five 7-min cycles at 60% energy output, Vibracel Sonics and Materials, Newtown, CT, USA). An ice bath was used to prevent overheating. Two batches of CNC, labelled SA and SB were prepared. The main difference was in the ion exchange treatment used to remove excess ion content from the suspensions. Batch SA was stirred over Dowex Marathon MR-3 (Dow Chemical Co., Midland, MI, USA) for 24 h. Batch SB was also stirred over Dowex Marathon MR-3 for 24 h, but then the mixed bed resin was filtered off and replaced by Dowex Marathon C, a strong acid cation exchange resin that had previously been rinsed repeated with ethanol to remove colour. The second resin treatment was to ensure that the resin was in acid form. The suspensions were next filtered through glass microfibre filters (0.45 µm pore size) to remove any particulate impurities introduced by the sonicater tip, and concentrated by means of a Labconco RapidVap vacuum evaporation system. The concentration of the suspensions was determined gravimetrically by drying small volumes in an oven at 105 ˝ C. Sample SA was used as a 5.2 wt % suspension, and sample SB was concentrated to give 6 wt % (SB-6%) and 8 wt % (SB-8%) suspensions. The CNC samples were stored at ´4 ˝ C for at least 2 weeks until the suspensions had separated into isotropic and anisotropic phases. Small amounts of each sample were sealed in glass micro-slides (VitroCom, Mountain Lakes, NJ, USA) and kept standing vertically for 2–3 days before imaging by polarized light microscopy. 3.2. Polarized Light Microscopy Photographs of the liquid crystalline suspensions and the dried films prepared from CNC were taken with a Nikon Eclipse LV100POL polarized light microscope and DS-Fi1 camera (Nikon, Tokyo, Japan). A 530 nm red wave plate was inserted between crossed polars to indicate CNC orientation [15]. Some images of chiral nematic cellulose derivatives were prepared as described in the cited references. 3.3. Fabrication of CNC Films CNC suspensions were drop cast onto glass (for transmitted light microscopy and optical profilometry), black acrylic (for reflected light imaging) or mica surfaces (for atomic force microscopy). The suspensions were allowed to evaporate undisturbed at ambient conditions (room temperature 20–25 ˝ C, relative humidity 20%–60%) for 12 h to give disc-shaped solid films on the glass, acrylic or mica surfaces. 3.4. Optical Profilometry The topography of mm-scale areas of the dry CNC films was investigated by optical profilometry. CNC films deposited on glass substrates were scanned with an optical profilometer (Wyko NT8000, Veeco Instruments Inc., Plainview, NY, USA). The initial results were noisy, but the quality of the topographic images was much improved by coating the surface of the CNC film with

7885

Materials 2015, 8, 7873–7888

a 20 nm thick layer of gold by means of a NexDep Ebeam Evaporator (Angstrom Engineering Inc., Kitchener, ON, Canada). 3.5. Atomic Force Microscopy The topography of the CNC film surfaces at much smaller length scales was investigated by AFM. To prepare very dilute CNC samples on mica for AFM imaging of individual nanocrystals, a drop of poly-L-lysine (0.01 wt % in water) was placed onto a freshly cleaved mica surface (Electron Microscope Supplies, Hatfield, PA, USA), rinsed with water, followed by a drop of dilute CNC suspension (~0.005 wt %). The samples were left to dry in Petri dishes under ambient conditions. For images of CNC orientation in films cast from chiral nematic suspensions, a droplet of 8 wt % suspension was allowed to evaporate on the mica surface. A Multi-Mode-3 AFM (Digital Instruments, Bruker, Billerica, MA, USA) was used in tapping mode to image the samples, with probes obtained from NanoScience (Si, N-type, 0.01–0.025 Ω/cm, 200–400 kHz resonance frequency, with nominal tip radius of 10 nm). 4. Conclusions Simple evaporation of a droplet of aqueous CNC suspension leads to a surprising range of self-organizing phenomena: 1. Starting with a dilute suspension of CNC, well below the critical concentration for liquid crystal formation, the CNC tend to align radially at the edge of the droplet. 2. For more concentrated suspensions that are initially above the critical concentration, some radial initial orientation of CNCs at the edge of the film is observed. This is in line with recent observations of the orientation of CNCs in shear flow, where the acid-form CNCs self-oriented across the flow direction [29]. 3. Drying droplets of concentrated CNC display a topography governed by evaporation-driven mass transfer (the “coffee-stain effect”). 4. The magnitude of the coffee-stain effect is greatest for strongly pinned droplets, where the height near the edge of the dry film is several times the height in the middle of the film [16]. 5. When pinning of the contact line is weaker, a “stick-slip” situation is observed, with layers of increasing thickness are laid down by the receding contact line. The layers give a complex optical texture; radial cracking of the drying film was sometimes observed. 6. The surface topography of dry films at short length scales showed a radial orientation of the CNC at the free surface of the film, along with a radial height variation with a period of approximately P/2, ascribed to the anisotropic shrinkage of the chiral nematic structure. 7. The rich liquid crystalline textures displayed by the evaporation of water from CNC suspensions depend on the interplay of concentration gradients that influence chiral nematic pitch, director orientation and gelation with the effects of surface orientation and the elastic constants normally associated with liquid crystalline phases. The qualitative results here are very preliminary, but hopefully indicate some of the variables involved in the formation of chiral nematic solids from cellulose nanocrystal suspensions. Acknowledgments: Funding from the Natural Sciences and Engineering Research Council Canada, and support from Fonds Québecois de la Recherche sur la Nature et les Technologies through the Center for Self-Assembled Chemical Structures are gratefully acknowledged. We thank M. Ramkaran for assistance with the AFM imaging. Author Contributions: X.M. prepared the CNC suspensions and films, took the photomicrographs and contributed to the manuscript. D.G.G. supervised the work and wrote the manuscript. Conflicts of Interest: The authors declare no conflict of interest.

7886

Materials 2015, 8, 7873–7888

References 1. 2. 3. 4.

5. 6. 7. 8. 9.

10. 11. 12. 13.

14. 15. 16. 17. 18. 19. 20. 21. 22.

23.

Werbowyj, R.S.; Gray, D.G. Liquid crystalline structure in aqueous hydroxypropyl cellulose solutions. Mol. Cryst. Liq. Cryst. Lett. 1976, 34, 97–103. [CrossRef] Nishio, Y.; Sato, J.; Sugimura, K. Liquid crystals of cellulosics: Fascinating ordered structures for the design of functional material systems. Adv. Polym. Sci. 2015. [CrossRef] Revol, J.F.; Bradford, H.; Giasson, J.; Marchessault, R.H.; Gray, D.G. Helicoidal self-ordering of cellulose microfibrils in aqueous suspension. Int. J. Biol. Macromol. 1992, 14, 170–172. [CrossRef] Lagerwall, J.P.F.; Schütz, C.; Salajkova, M.; Noh, J.H.; Park, J.H.; Scalia, G.; Bergström, L. Cellulose nanocrystal-based materials: From Liquid Crystal Self-Assembly and Glass Formation to Multifunctional Thin Films. Available online: http://www.nature.com/am/journal/v6/n1/abs/am201369a.html (accessed on 10 January 2014). Habibi, Y.; Lucia, L.A.; Rojas, O.J. Cellulose nanocrystals: Chemistry, self-assembly and applications. Chem. Rev. 2010, 110, 3479–3500. [CrossRef] [PubMed] Klemm, D.; Kramer, F.; Moritz, S.; Lindström, T.; Ankerfors, M.; Gray, D.; Dorris, A. Nanocelluloses: A new family of nature-based materials. Angew. Chem. Int. Ed. 2011, 50, 5438–5466. [CrossRef] [PubMed] Moon, R.J.; Martini, A.; Nairn, J.; Simonsen, J.; Youngblood, J. Cellulose nanomaterials review: Structure, properties and nanocomposites. Chem. Soc. Rev. 2011, 40, 3941–3994. [CrossRef] [PubMed] Chandrasekhar, S. Liquid Crystals, 2nd ed.; Cambridge University Press: Cambridge, UK, 1992; pp. 213–299. Revol, J.F.; Godbout, L.; Dong, X.M.; Gray, D.G.; Chanzy, H.; Maret, G. Chiral nematic suspensions of cellulose crystallites; phase separation and magnetic field orientation. Liq. Cryst. 1994, 16, 127–134. [CrossRef] Dong, X.M.; Kimura, T.; Revol, J.F.; Gray, D.G. Effects of ionic strength on the phase separation of suspensions of cellulose crystallites. Langmuir 1996, 12, 2076–2082. [CrossRef] Chen, W.; Gray, D.G. Interfacial tension between isotropic and anisotropic phases of a suspension of rod-like particles. Langmuir 2002, 18, 633–637. [CrossRef] Mu, X.; Gray, D.G. Formation of chiral nematic films from cellulose nanocrystal suspensions is a two-stage process. Langmuir 2014, 30, 9256–9260. [CrossRef] [PubMed] Hoeger, I.; Rojas, O.J.; Efimenko, K.; Velev, O.D.; Kelley, S.S. Ultrathin film coatings of aligned cellulose nanocrystals from a convective-shear assembly system and their surface mechanical properties. Soft Matter 2011, 7, 1957–1967. [CrossRef] Uetani, K.; Yano, H. Self-organizing capacity of nanocelluloses via droplet evaporation. Soft Matter 2013, 9, 3396–3401. [CrossRef] Introduction to Polarized Light Microscopy. Available online: https://www.microscopyu.com/articles/ polarized/polarizedintro.html (accessed on 2 November 2015). Mu, X.; Gray, D.G. Droplets of cellulose nanocrystal suspensions on drying give iridescent 3-D “coffee-stain” rings. Cellulose 2015, 22, 1103–1107. [CrossRef] Revol, J.-F.; Godbout, L.; Gray, D.G. Solid films of cellulose with chiral nematic order and optically variable properties. J. Pulp Paper Sci. 1998, 24, 146–149. Beck, S.; Bouchard, J.; Berry, R. Controlling the reflection wavelength of iridescent solid films of nanocrystalline cellulose. Biomacromolecules 2011, 12, 167–172. [CrossRef] [PubMed] Beck, S.; Bouchard, J.; Chauve, G.; Berry, R. Controlled production of patterns in iridescent solid films. Cellulose 2013, 20, 1401–1411. [CrossRef] Tang, H.; Guo, B.; Jiang, H.; Xue, L.; Li, B.; Cao, X.; Zhang, Q.; Li, P. Fabrication and characterization of nanocrystalline cellulose films prepared under vacuum conditions. Cellulose 2013, 20, 2667–2674. [CrossRef] Dumanli, A.G.; van der Kooij, H.M.; Kamita, G.; Reisner, E.; Baumberg, J.J.; Steiner, U.; Vignolini, S. Digital color in cellulose nanocrystal films. ACS Appl. Mater. Interfaces 2014, 6, 12302–12306. [CrossRef] [PubMed] Dumanli, A.G.; Kamita, G.; Landman, J.; van der Kooij, H.M.; Glover, B.J.; Baumberg, J.J.; Steiner, U.; Vignolini, S. Controlled, bio-inspired self-assembly of cellulose-based chiral reflectors. Adv. Opt. Mater. 2014, 2, 646–650. [CrossRef] [PubMed] Cira, N.J.; Benusiglio, A.; Prakash, M. Vapour-mediated sensing and motility in two-component droplets. Nature 2015, 519, 446–450. [CrossRef] [PubMed]

7887

Materials 2015, 8, 7873–7888

24. 25. 26. 27. 28. 29.

Meister, R.; Dumoulin, H.; Halle, M.A.; Pieranski, P. The anchoring of a cholesteric liquid crystal at the free surface. J. Phys. II 1996, 6, 827–844. [CrossRef] Roman, M.; Gray, D.G. Parabolic focal conics in self-assembled solid films of cellulose nanocrystals. Langmuir 2005, 21, 5555–5561. [CrossRef] [PubMed] Page, D.H.; Tydeman, P.A. Consolidation of the Paper Web; Bolam, F.M., Ed.; British Paper and Board Makers Association: London, UK, 1965; pp. 371–396. Rofouie, P.; Pasini, D.; Rey, A.D. Nano-scale surface wrinkling in chiral liquid crystals and plant-based plywoods. Soft Matter 2015, 11, 1127–1139. [CrossRef] [PubMed] Kloser, E.; Gray, D.G. Surface PEG-grafting of cellulose nanocrystals in aqueous media. Langmuir 2010, 26, 13450–13456. [CrossRef] [PubMed] Tatsumi, M.; Teramoto, Y.; Nishio, Y. Different orientation patterns of cellulose nanocrystal films prepared from aqueous suspensions by shearing under evaporation. Cellulose 2015, 22, 2983–2992. [CrossRef] © 2015 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons by Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).

7888