Shear-Induced Alignment of Anisotropic Nanoparticles in a Single

Dec 4, 2017 - Department of Chemical Engineering and Applied Chemistry, University of Toronto, 200 College Street, Toronto, Ontario M5S 3E5, Canada. â...
0 downloads 9 Views 1MB Size
Subscriber access provided by READING UNIV

Article

Studies of Shear-induced Alignment of Anisotropic Nanoparticles in a Single-Droplet Oscillatory Microfluidic Platform Moien Alizadehgiashi, Amir Khabibullin, Yunfeng Li, Elisabeth Prince, Milad Abolhasani, and Eugenia Kumacheva Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.7b03648 • Publication Date (Web): 04 Dec 2017 Downloaded from http://pubs.acs.org on December 9, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Studies of Shear-Induced Alignment of Anisotropic Nanoparticles in a Single-Droplet Oscillatory Microfluidic Platform Moien Alizadehgiashi†, Amir Khabibullin†, Yunfeng Li†, Elisabeth Prince†, Milad Abolhasani‡*, Eugenia Kumacheva†, §,∥ * †

Department of Chemistry, University of Toronto, 80 Saint George Street, Toronto, Ontario, M5S 3H6, Canada.



Department of Chemical Engineering, North Carolina State University, 911 Partners Way, Raleigh, North Carolina, 27695-7905, USA.

§

Department of Chemical Engineering and Applied Chemistry, University of Toronto, 200 College Street, Toronto, Ontario, M5S 3E5, Canada.

∥Institute

of Biomaterials and Biomedical Engineering, University of Toronto, 4 Taddle Creek Road, Toronto, Ontario, M5S 3G9, Canada

Keywords. Nanoparticles, cellulose nanocrystals, microfluidics, shear-induced alignment, anisotropy, birefringence.

ACS Paragon Plus Environment

1

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 30

Abstract

Flow-induced alignment of shape-anisotropic colloidal particles is of great importance in fundamental research and in the fabrication of structurally anisotropic materials, however rheooptical studies of shear-induced particle orientation are time- and labor-intensive and require complicated experimental setups. We report a single-droplet oscillatory microfluidic strategy integrated with in-line polarized light imaging as a strategy for studies of shear-induced alignment of rod-shape nanoparticles. Using an oscillating droplet of an aqueous isotropic suspension of cellulose nanocrystals (CNCs), we explore the effect of the shear rate and suspension viscosity on the flow-induced CNC alignment and subsequent relaxation to the isotropic state. The proposed microfluidic strategy enables high-throughput studies of shearinduced orientations in structured liquid under precisely controlled experimental conditions. The results of such studies can be used in the development of structure-anisotropic materials.

Introduction Liquid crystalline polymers1,2,3, DNA molecules4, and high-aspect ratio colloidal particles such as rod-like fd virus5 and wormlike micelles6-8, exhibit shear-mediated alignment in the direction of flow. This effect has great implications for fundamental understanding of flow of molecules and colloidal particles, as well as their applications for producing fibers9 and sheets10, threedimensional printing of high-strength structure-anisotropic materials11,12 and the fabrication of strain sensors.13 Shear-induced alignment of colloidal particles occurs when the shear rate exceeds the characteristic relaxation time of their suspensions. Understanding how the alignment and the

ACS Paragon Plus Environment

2

Page 3 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

relaxation time depend on the size and shape of flow-oriented particles, as well as on the viscosity of the liquid and operation parameters is fundamentally and practically important. Microfluidics (MFs) offers a unique capability to study the interplay of confinement and shearinduced alignment of colloids14,15, however, most of MF experiments have been limited to studies of flow of nematic liquid crystals.16-19 One of the limitations of the continuous MF flow in studies of shear-induced nanoparticle orientation stems from the time-to-distance transformation, that is, the correlation between the time and position of a particular event. This relationship may result in a very significant increase in dimensions of MF devices used for studies of slow shear-induced orientational transitions and their relaxation in structured fluids.5, 20-22

Recently, this limitation has been addressed with the introduction of the oscillatory MF

platform23 that utilized a periodic change in the direction of motion of a liquid segment placed in a tube or a microchannel.24-27 Importantly, the assembly of the tube-based oscillatory MF platform was cost- and labor-efficient, due to the elimination of the time-consuming and expensive microfabrication process.28,29 The MF oscillatory platform has been used for studies of shear-induced elongation of DNA molecules30, reactive etching of inorganic nanoparticles25, optimization of Pd-catalyzed reaction27, studies of growth dynamics of isogenic cell populations23,31, and screening of liquid-liquid phase-separation and extraction processes.24,32 In these studies, the role of oscillatory flow was limited to the enhanced mixing and increased residence times of the liquid segment. Flow-induced alignment of shape-anisotropic nanoparticles has not been reported. Here, we report the results of experimental studies and simulations of flow-induced alignment of rod-like cellulose nanocrystals (CNCs) in their aqueous suspensions. Over the past decade, CNCs have attracted strong interest of the materials science community due to their high

ACS Paragon Plus Environment

3

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 30

mechanical strength33,34, surface chemistry amenable to chemical or physical modification35 and the formation of cholesteric liquid crystalline phases in aqueous CNC suspensions.36 Shearinduced orientation of CNCs in fibers produced by spinning37-39 and hydrodynamic flow focusing40 greatly influenced fiber performance. Currently, MF studies of CNC suspensions have been limited to the generation of cholesteric droplets,41,42 however MFs can provide enhanced understanding of flow-induced structural transitions in CNC suspensions, which can be used for the fabrication of new high-performance materials with anisotropic structures and properties.43,44 In the present work, we developed a single-droplet MF oscillatory platform integrated with polarized optical microscopy for in-situ studies of shear-induced alignment of CNCs in their aqueous shear-thinning suspensions. The alignment and relaxation of CNCs at varying shear rates in suspension with different viscosities was studied in capillaries with a circular or square cross-section. The ease of the assembly of the single-droplet oscillatory platform from the commercially available components and the ability to integrate it with fluorescence and polarized optical microscopy make the MF approach an efficient tool for studies of shear-induced behavior of shape-anisotropic colloidal particles.

ACS Paragon Plus Environment

4

Page 5 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 1. (a) Schematics of oscillatory microfluidic platform. (b) Top: A segment of a quiescent isotropic CNC suspension confined in a capillary with square cross-section. (I) mid-plane top view and (II) cross-section of the suspension with a random CNC orientation. (c) Top: flowinduced CNC alignment in the suspension segment subjected to oscillatory motion. (I) midplane top view and (II) cross-section of the suspension with a CNC alignment in the direction of flow. In (I) the CNC alignment increases from the center to the walls. (d) Schematic of shear rate profile in the capillary: (I) shear rate distribution in the cross-section of the capillary; (II) midplane shear rate distribution (a view from the top of the capillary). Experimental Materials A 12 wt% aqueous suspension of CNCs was supplied by the USDA Forest Product Laboratory (USDA FPL, Madison, WI, USA). Trimethoxy(octadecyl) silane (TOS), denatured ethanol (reagent grade), HCl and NaOH were purchased from Sigma-Aldrich. Fluorescein-5’-

ACS Paragon Plus Environment

5

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 30

isothiocyanate (FITC) (90%) was obtained from ACROS Organics. Glycerol was purchased from ACP Chemicals Inc. (Montreal, QC, Canada). All reagents, unless specified, were used as received. Milli-Q grade deionized water was used to dilute CNC suspensions. Two glass tubes, one with a circular cross-section (with an inner diameter of 1.6 mm and wall thickness of 1.4 mm); and the other one with a square cross-section (inner dimensions of 0.8 mm and wall thickness of 0.14 mm) were supplied by Friedrich & Dimmock Inc., Millville, NJ, USA. The capillary with a circular cross-section was connected to the syringe pumps and pressurized vessel using PEEK fittings and an adapter (Upchurch Scientific, Oak Harbor, WA, USA). The capillary with a square cross-section was glued with epoxy glue at both ends to the PTFE tubing (McMaster-Carr, USA). Silanization of a glass capillary The capillaries were washed with 10 wt% HCl solution in deionized water, rinsed with acetone, transferred for 24 h into a mixture of 1 M solution of NaOH in 80 vol% IPA and 20 vol% deionized water, rinsed with deionized water, ethanol and acetone, and dried in a vacuum oven at 80 oC for 24 h. After placing the capillaries in a plasma chamber at 300 mTorr for 20 min, they were filled for 8 h at 50 oC with a 5 wt% solution of TOS in ethanol. The solution was then removed and the capillaries were placed in a vacuum oven at 80 oC for 2 h. The wetting angle of the 3 wt% CNC suspension on the silanized surface of the capillary was 85±5˚ (Supporting Information, Figure S1). Labeling CNCs with a fluorescent dye Using a method reported elsewhere46, FITC was covalently attached to the CNC surface in a

ACS Paragon Plus Environment

6

Page 7 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

0.1 M NaOH solution for 5 days at r. t. in the dark. The CNCs were separated from free (nonattached) FITC molecules by five cycles of centrifugation (7500 g for 10 min), washing with 0.1 M NaOH solution and dialyzed for two days against deionized water using a cellulose membrane (SIGMA, 33 mm avg. flat width and molecular weight cut-off of 14000 Da) in the dark. The color of the suspension of FITC-labeled CNCs changed from white to yellow. The suspension exhibited fluorescence emission at 530 nm when excited at 490 nm (Supporting Information, Figure S2 and S3). Experimental setup Figure 1a shows schematically the MF oscillatory platform used in the present work. The platform included a 20 cm-long glass capillary with a square or a circular cross-section (Supporting Information, Figure S4). A signal from two photoresistor-based liquid sensors (TT Electronics, OPTEK OCB 350 Series, Woking, Surrey, England) that were set 20 cm apart was received using a data acquisition box (National Instrument DAQ USB-6008). A computercontrolled syringe pump (PHD Ultra, Harvard Apparatus, Holliston, MA, USA) loaded with a gas-tight glass syringe (10 mL, SGE Analytics) was used to apply the alternating pressure gradient to the capillary. To reduce the effect of gas compressibility on the response time of the system to the push/pull motion, the capillary was pressurized with a nitrogen gas using a vessel connected to the capillary outlet (nitrogen cylinder, Grade 4.8, Linde BOC, Canada). A 5 cmlong plug of the 3 wt% aqueous CNC suspension was introduced in the capillary using another syringe pump (PHD2000, Harvard Apparatus, Holliston, MA, USA). Owing to the alternating pressure gradient in the capillary, the liquid segment underwent motion with a periodically changing direction (oscillatory motion) at the average velocity, v, in the range from 5 to 158

ACS Paragon Plus Environment

7

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 30

mm/s. The half-oscillation time of the plug varied from 1 to 30 s for ν of 158 to 5 mm/s, respectively, the oscillation amplitude was maintained at 15 cm. A capillary containing a plug of the CNC suspension was positioned on the microscope stage. The polarized light images (PLI) were recorded using a camera (Nikon D7200) connected to the binocular port of the polarized optical microscope (Olympus BX 51). The flow-induced birefringence (BF) was measured when the moving plug was in the middle of the capillary. The fluorescence images of the segment of the aqueous suspension of FITC-labelled CNCs were acquired using an inverted fluorescent microscope (Nikon Eclipse-Ti) equipped with Mira Light Engine (Lumencor, Beaverton, OR, USA) using the FITC channel. Fluorescence intensity profiles were measured using a NIS-elements AR Analysis software. Birefringence Measurement A monochromatic light with FITC filter cube (U-N41001 HQ:F, Excitation 480±23 nm and Emission 535±25 nm, Chroma Tech Corp, VT, USA) was used to illuminate the sample. Using a cooled monochrome CCD camera (Evolution VF, Media Cybernetics, MD, USA), the PLIs of the suspension plug were acquired and recorded in a gray-scale, flat binary raw unprocessed format. Light intensity at different distances from the middle of the capillary was evaluated using ImageJ software. Viscosity measurements The viscosity of the CNC suspensions was measured using a Carri-Med CSL2 Rheometer (TA Instruments, New Castle, DE, USA) with a parallel-plate geometry at 60 µm gap. The measurements were conducted at 25 °C at the shear rates in the range from 1 to 1300 s-1.

ACS Paragon Plus Environment

8

Page 9 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Results In half a cycle of the oscillatory motion of the liquid segment of the CNC suspension, the velocity of the liquid increases from zero to a maximum positive value almost instantly27 and remains invariant, until it reaches the sensor. At this point, the velocity instantly reduces to zero. Then, the plug changes the direction of motion, and the velocity instantly reaches its maximum negative value and remains constant, until it reaches the original position, i.e. the second sensor. Subsequently, the cycle is repeated. Thus, the shear oscillations are comprised of square pulses moving the suspension segment forward and backwards. Figure 1b and c illustrate the alignment of CNCs in a liquid segment undergoing oscillatory motion in a capillary with a square cross-section. In the quiescent suspension, the CNCs acquire random orientation. Under applied shear, the CNCs align in the direction of flow and relax to the isotropic state when velocity reduces to zero. Figure 1d illustrates qualitatively the shear stress profile in the capillary, when a pressuredriven flow is applied.47, 48 The velocity is at its maximum, and the shear stress is zero in the middle of the capillary. At the walls of the capillary and away from the corners, the shear stress is at its maximum and the velocity is zero (with the no-slip boundary conditions in place). We conducted 3D simulations (COMSOL Multiphysics 4.4, Burlington, MA, USA) to compute the shear rate, γ , distribution in the oscillating plug of the CNC suspension. These results were used to relate the intensity of BF (and thus CNC alignment) to the local ( ) in the capillary. We assumed that the effect of inertial forces was negligible, the density of the

ACS Paragon Plus Environment

9

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 30



suspension is 1000 , and viscosity is a function of  (see Supporting Information). The simulations were conducted for v in the range from 5 to 150 mm/s.

Figure 2. Simulated shear rate profile for the liquid segment of 3 wt% CNC suspension oscillating at v =100 mm/s in the capillary with a square cross-section with dimensions of 0.8 mm. (a) Top-view middle plane shear rate distribution; (b) cross-section view of the shear rate distribution in the middle of the segment. Figure 2 shows a representative simulated  field for the segment of the CNC suspension oscillating in the capillary with a square cross-section at v=100 mm/s. (A similar result is shown in Figure S5, Supporting Information for the CNC plug oscillating in the capillary with a circular cross-section). For the pressure-driven flow, the shear profile was linear, with a zero shear in the middle of the channel and the maximum shear next to walls. Figure 2a shows the simulated distribution of  in the middle plane of the capillary (top-view). At the walls of the capillary, the maximum value of  was 1100 s-1, while in the 40 µm-wide region (that is, +/-20 µm from the capillary center) the average value of  was 100 s-1. Figure 2b shows the  profile under the same conditions under cross-section view. The maximum ( ) is at the walls, except the corners, and the minimum ( ) is in the center of the capillary cross-section.

ACS Paragon Plus Environment

10

Page 11 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Shear-Induced birefringence Prior to BF measurements, we ensured that no shear-induced change in the local CNC concentration within the plug takes place. The CNCs were covalently labelled with FITC, and the distribution of fluorescence intensity was studied for the central plane of the liquid segment. No change in the distribution of fluorescence intensity was observed within the segment moving in both types of the capillaries at v up to 158 mm/s (Supporting Information, Figures S3). Thus, we conclude that in the oscillatory motion experiments the CNCs were uniformly distributed within the suspension segment, and no CNC loss from the plug occurred due to precipitation. First, we examined shear-induced BF under white light with the polarizer set parallel to the flow direction and the analyzer placed at 90˚ with respect to the flow direction(see Supporting Information, Figure S6 for the effect of the direction of cross-polarizer with respect to the direction of flow). Under static conditions, the BF image of the 3 wt% CNC suspension was black, indicating random CNC orientation (Figure 3a). When a suspension segment was subjected to oscillatory motion, an emerging bright BF pattern was a signature of CNC alignment in the direction of flow. Figure 3b-f displays top-view BF patterns observed in the middle plane of the segment moving at various v. When the segment reached the limiting position on either side of the capillary and stopped before changing the direction of motion, the BF disappeared. At v>0, two symmetric bright regions appeared in the BF pattern of the liquid segment, which were separated by the middle black region. The bright regions were white at v=2.5 mm/s (Figure 3b), however, further increase in v resulted in the change of the number and the sequence of colors across the liquid segment. For example, at v=52 mm/s, the bright regions contained white,

ACS Paragon Plus Environment

11

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 30

yellow, orange and brown stripes (Figure 3c), while at v=158 mm/s a sequence of white, yellow, red, purple, blue and turquoise colors was observed (Figure 3f).

Figure 3. Polarized optical microscopy top-view middle plane images of the CNC suspension segment subjected to the oscillatory motion at velocity, v, of (a) 0, (b) 2.5, (c) 52, (d) 105, (e) 131, and (f) 158 mm/s in the glass capillary with a square cross-section. Scale bar is 200 µm. The polarizer-analyzer positions and flow direction for all the experiments are shown in the insets of Figure 3a. The images were taken at the moment when the suspension segment was in the center of the capillary. The yellow dashed lines show the wall boundary.

The flow-induced BF patterns of the plug of the CNC suspension oscillating at 0≤v≤158 mm/s in the cylindrical capillary (Supporting Information, Figs. S7 and S8) were qualitatively similar to that observed in the capillary with a square cross-section capillary. To quantify the shear-induced BF of the CNC suspension, we performed transmission mode POM experiments using monochromatic light at λ=500 nm, with cross-polarizers oriented at 45o with respect to the flow direction49. Figure 4, top row, left-to-right shows representative PLI images of the plug subjected to oscillations at increasing v. The dashed yellow lines show

ACS Paragon Plus Environment

12

Page 13 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

the capillary walls and the central dashed line shows the middle plane of the capillary. With v increasing from 5 to 50 mm/s, a bright region in the middle of the capillary gradually became dark, and two darker zones adjacent to the capillary walls became bright. Further increase in v to 130 mm/s resulted in the opposite effect: the central region appeared bright and the regions next to the walls became dark. This effect was caused by the periodically changing light intensity,49 as it changed from the first-order maximum at v=5 mm/s, the first-order minimum at v=26 mm/s, and the second-order maximum at v=130 mm/s, all at the center line of the plug (Supporting Information, Figure S10). The change in BF across the oscillating liquid plug was characterized as

=



∆ ,

(1)

where is the light wave retardation, λ is the wavelength of the monochromatic light (λ=500 nm), d is the plug thickness or the path of light (d=800 µm), and ∆n is the difference in ordinary  and extraordinary refractive indices  , ∆=  −  , which determines BF. The retardation wavelength was determined as 

= 2 ×   ,  

(2)

where ! is the intensity of transmitted light, and !" is the intensity of light passing through of a parallel polarizer-analyzer system under quiescent conditions.50 Since the distribution of light intensity is a periodic function,49,51 for #$ ≥ 2& (when intensity of light reached the first-order maximum), the calculated value of δ was corrected (Supporting Information, Figure S10). Figure 4, middle row shows the variation in ∆ plotted as a function of the distance from the center line of the plug that moves at different v. The general trend observed for all the values of v was the increase in ∆ from the center of the liquid plug to its

ACS Paragon Plus Environment

13

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 30

periphery (that is, the capillary walls), consistent with the increase in the shear rate determined in numerical simulations (Figure 4, bottom row and Figure 2).

Figure 4. Shear and birefringence distribution in the square cross-section capillary at (a) v=5 mm/s, (b) v=26 mm/s, and (c) v=130 mm/s. Yellow dashed lines show the boundary of the channel. The black dashed line shows the middle of the channel. Inset of panel a, first row shows the orientation of cross-polarizer for all of the intensity based experiments.

In the analysis, we assumed that the PLI of the plug represents the BF (=∆) induced in the middle-plane of the plug. The BF became stronger with increasing value of v. More specifically, for 5≤v≤130 mm/s, the BF value increased from 4.3×10-4 to 10.3×10-4 in the middle of the plug and from 6.1×10-4 to 12.4×10-4 at d=200 µm. The corresponding shear

ACS Paragon Plus Environment

14

Page 15 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

rates increased with v from 4 to 150 s-1 in the middle of the plug, and from 11 to 790 s-1 at d=200 µm. In the entire range of v values studied, the value of ∆ monotonically increased from the middle of the plug (at d=0) to d=200 µm.

Effect of viscosity on shear-induced CNC alignment and relaxation To explore the effect of viscosity of the suspension on CNC alignment, we prepared 3 wt% CNC suspensions in water/glycerol mixtures with glycerol volume fractions of 0, 0.5 and 0.75. The steady-state viscosity, µ" , of these CNC suspensions was determined to be 4.5, 26, and 96 cP respectively. First, we examined the variation in viscosity of the CNC suspensions with the shear rate,  in a parallel-plate rheometer (Figure 5a). All three suspensions showed a shear thinning behavior, with a stronger viscosity reduction for suspensions with a higher volume fraction of glycerol. For  ≥400 s-1, no further significant change in viscosity was observed for all suspensions. The relative viscosity of the CNC suspensions (determined by dividing viscosity at each

 by the corresponding solvent viscosity '( ) was significantly higher for the CNC

suspension with the glycerol volume fraction of 0.75, in comparison with other suspensions (Fig 5a). This effect was caused by the desulfation of CNC surface sulphate groups at high glycerol content, resulting in the reduced electrostatic repulsion between individual CNCs and their association due to hydrogen bonding between the surface hydroxyl groups.52 In the next step, we examined the evolution of BF in the oscillating segment of CNC suspensions with different '" . We note that the value of  changes across the plug, increasing with the distance d from its middle plane, as shown in Figure 5b. For example, at v=80 mm/s, a 20-fold increase in  occurs for d=200 µm, in comparison with that at d=0. Since a different degree of shear thinning could occur across the segment of the CNC suspension, we examined

ACS Paragon Plus Environment

15

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 30

the evolution of BF in the CNC suspensions with different '( at two different locations of the plug, that is, at d=0 and d=200 µm. Figure 5c shows the development of ∆ at d=0 µm and