Accepted Manuscript

Rheology and Microstructure of Aqueous Suspensions of Nanocrystalline Cel-

lulose Rods

Yuan Xu, Aleks D. Atrens, Jason R. Stokes

PII: S0021-9797(17)30172-8

DOI: http://dx.doi.org/10.1016/j.jcis.2017.02.020

Reference: YJCIS 22047

To appear in: Journal of Colloid and Interface Science

Received Date: 5 December 2016

Revised Date: 9 February 2017

Accepted Date: 9 February 2017

Please cite this article as: Y. Xu, A.D. Atrens, J.R. Stokes, Rheology and Microstructure of Aqueous Suspensions

of Nanocrystalline Cellulose Rods, Journal of Colloid and Interface Science (2017), doi: http://dx.doi.org/10.1016/

j.jcis.2017.02.020

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers

we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and

review of the resulting proof before it is published in its final form. Please note that during the production process

errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

brought to you by COREView metadata, citation and similar papers at core.ac.uk

provided by University of Queensland eSpace

Rheology and Microstructure of Aqueous Suspensions of

Nanocrystalline Cellulose Rods

Yuan Xua, Aleks D Atrens

c, Jason R Stokes

a,b*

a School of Chemical Engineering, The University of Queensland, Brisbane, Australia

bAustralian Research Council Centre of Excellence in Plant Cell Walls, The University of Queensland,

Brisbane, Australia

cSchool of Mechanical and Mining Engineering, The University of Queensland, Brisbane, Australia

* Corresponding author:

Professor Jason Stokes

School of Chemical Engineering

The University of Queensland, Brisbane QLD 4072, Australia;

Tel: +61 7 336 54361; Fax: +61 7 336 54199;

Email: Jason.Stokes@uq.edu.au

Abstract

Hypothesis

Nanocrystalline cellulose (NCC) is a negatively charged rod-like colloid obtained from the hydrolysis

of plant material. It is thus expected that NCC suspensions display a rich set of phase behaviour with

salt and pH because of its anisotropic shape and electrical double layer that gives rise to liquid

crystallinity and self-assembly respectively. It should thus be possible to tune the rheological

properties of NCC suspensions for a wide variety of end-use applications.

Experiments

Rheology and structural analysis techniques are used to characterise surface-sulphated NCC

suspensions as a function of pH, salinity (NaCl) and NCC concentration. Structural techniques include

atomic force microscopy, Zeta potential, dynamic light scattering, and scanning electron microscopy.

Findings

A phase diagram is developed based on the structure-rheology measurements showing various states

of NCC that form as a function of salt and NCC concentration, which go well beyond those

previously reported. This extended range of conditions reveals regions where the suspension is a

viscous fluid and viscoelastic soft solid, as well as regions of instability that is suggested to arise

when there is sufficient salt to reduce the electrical double layer (as explained qualitatively using

DLVO theory) but insufficient NCC to form a load bearing network.

Keyword: Nanocrystalline cellulose, rheology, microstructures, ionic strength, colloidal rod.

1. Introduction

Nanocrystalline cellulose (NCC) are nano-scaled (colloidal) rod-like particles of cellulose obtained

from hydrolysis of plant mass. They are attracting increasing scientific and engineering attention since

the influential works of Revol et al. [1] on nematic self-ordering of NCC in aqueous suspension and

Favier et al. [2] on NCC’s significant reinforcement of polymer composites. The advantages of NCC

as an advanced material are their high strength, abundance, biodegradability, non-toxicity, good

thermal stability and relatively low cost [3, 4]. Its large aspect ratio and the high density of surface

hydroxyl groups enable NCC suspensions to have unique rheological properties, whereby isotropic

rod-like particle suspension behaviour occurs at low concentrations, and a chiral nematic liquid

crystalline structure is observed over a certain concentration range that is followed by gelation at

higher concentrations. In addition, colloids rods are more efficient at building low-shear viscosity,

yield stress and gel strength in suspension than spherical particles[5]. Thus NCC has potential to be

used in many application fields for rheological control, including within consumer products, drilling

fluids [6], drug delivery [7], and in artificial tissue formation [8]. In order to tune their ability to

control the rheology in a wide variety of applications, it is necessary to understand its relationship to

the underlying microstructure and how this is influenced by interparticle forces between the colloidal

rods. The specific objective of this study is to map the structure and rheology of NCC suspensions in

aqueous media as a function of pH, salt, and NCC concentration.

Recent research on the rheology of NCC suspensions has focused on its liquid crystalline or self-

ordering behaviour and its gelation properties [9, 10]. Their large aspect ratio and the presence of

surface hydroxyl groups play an important role on the interactive forces between NCC particles and

thus suspension rheology [11]. Surface hydroxyl groups render NCC colloidal rods to be negatively

charged, which electrostatically stabilises the suspension in aqueous media. They assemble into a gel

structure provided there is a sufficient salt concentration to screen surface charges so as to promote

adhesive interactions to occur [12].

The rheology of NCC suspensions in aqueous media and their gelation behaviour is of interest in this

study. NCC suspensions are considered to form liquid crystalline gels at a critical NCC concentration

and a randomly entangled gel at even higher concentration[13]. The critical concentration is affected

by size, aspect ratio and surface modifications of NCC as well as the ionic strength of the solution [14,

15]. However, the mechanism about how, and to what extent, particle interactions affect the

rheological and gelation behaviour of suspensions of NCC particles is yet to be comprehensively

studied as well as the change in linear viscoelasticity of the suspension as a function of phase

behaviour in terms of its colloidal stability and microstructure. In this work, we reveal structure-

property relationships for NCC suspensions by investigating their rheology as a function of ionic

strength, and consider the corresponding effect of NCC surface properties on the structure and

morphology of NCC network via Fourier transform infrared spectroscopy (FTIR), atomic force

microscopy (AFM), scanning electron microscopy (SEM), and Zeta potential measurement.

Monovalent sodium salt has been used to alter ionic strength of solution without causing ion mediated

cross-links between the NCC particles.

2. Materials and Methods

2.1. Materials.

NCC are obtained from the primary cell wall of plants, which is made up of semicrystalline cellulose

with randomly oriented amorphous regions; these regions have a low density and are highly

accessible to ionic species during acid hydrolysis[16, 17]. The amorphous regions of the cellulose

readily decompose under controlled temperature and concentration of inorganic acid, while the

crystalline regions are left intact. The crystalline components, which are only a few nanometres in size,

grow after being released from the cell wall during hydrolytic treatment. The literature NCC colloidal

rods to have a diameter of 4 - 30 nm and length of 50 - 500 nm [9, 18]; their size and aspect ratio are

dependent on the source of primary cellulose material and the hydrolysis conditions [19, 20].

The nanocrystalline cellulose used in this study was sourced from Maine University Process

Development Centre (Orono, ME) as 11.9 wt% NCC aqueous suspension with 0.9 wt% sulphur on

dry NCC sodium form which is equivalent to 4.62 SO3- per 100 anhydroglucose units [12]. The NCC

suspension was made by re-dispersing freeze-dried powder from hydrolysis of cellulose by applying

ultrasonic treatment. Sodium chloride from Merck KGaA (Darmstadt Germany) is used to investigate

the effect of salt. Hydrochloric acid and sodium hydroxide (from Ajax Finechem, NSW Australia)

were used to adjust pH values of suspensions. Diluted samples were produced using deionized water

produced by reverse osmosis water system, which has a resistivity of 18.2 MΩ•cm (Sartorius Stedim).

2.2. Preparation of NCC suspensions.

The 11.9 wt% NCC suspension was diluted to 0.1, 1, 2, 3, 4, 5, 7, and 9 wt% by addition of deionized

water. No ultrasonic treatment was further applied to sample i.e. particle size and aggregation pattern

was unchanged during sample preparation. The resultant suspensions were at pH of 6. For studies on

NCC suspension at different pH, incremental addition of 10 µL of 1 M NaOH or HCl solution were

used to achieve required pH values. The volume change due to the addition of acid or base is less than

0.17 vol%, which was considered negligible. For studies on the influence of ionic strength, solid

NaCl was added to NCC suspensions to levels required. For samples at very dilute NaCl (<0.01M),

salinity was adjusted by pipetting pre-prepared 0. 1M NaCl solution into corresponding NCC

suspensions.

2.3. Rheological Measurement.

An AR1500 rheometer (TA Instrument Inc., New Castle, DE) equipped with concentric cylinder

geometry was used to measure samples at room temperature. The inner and outer radii for the

concentric cylinder were 14 and 15 mm respectively. For fluid-like systems in this geometry, we

checked whether measurements were affected by the potential for ‘slip’ at smooth surfaces by

comparing measurements between smooth and rough surfaces; negligible differences are observed

with a result included in Supplementary Information. In contrast, a vane geometry was used for thick

and gel-like samples to avoid the effects of slip and structure modification during sample loading [21]

in the concentric cylinder; this geometry provides an accurate means in which to determine the

apparent yield stress and low-shear viscosity [21, 22]. The inner (Rv) and outer RC radii of the vane

tool was 11 and 22.5 mm respectively, and height (H) is 25 mm. Since the radius-ratio between cup

and vane is about 2 (RC/RV), the equations for Couette flow do not necessarily apply. The shear stress

(σ) and apparent shear rate ( ) is obtained from torque (M) and vane rotation rate in s-1

(Ω) using the

following equations [21, 22]:

(1)

(2)

Steady-state shear flow curves were measured in the shear range from 0.01 to 200 s-1 using shear rate

sweep tests with 25 s per data point. Unless stated otherwise, all experiments are at 25 °C and pH 6.

Measurement of the storage modulus (G’) and loss modulus (G’’) were conducted at frequency range

100 to 0.2 rad/s at controlled shear stress of 1 Pa (within the linear viscoelastic regime), for 0.01, 0.1,

1, 2, 3, and 5 wt% NCC suspensions with 0.1 M NaCl. The yield point for samples with 0.1 M salt

were measured using a shear stress sweep over the over the range of 0.001 to 200 Pa. The yielding

point was taken as the value of shear stress at the point on viscosity vs. shear stress plot where the

viscosity value started deviating from the tangent line of the linear region [21].

2.4. FT-IR spectra.

PIKE GladiATR FT-IR spectrometer (PerkinElmer, Waltham, Massachusetts) was used for

characterising the surface chemistry of NCC samples. Samples for analysis were from 5wt% water

suspensions form with pH 1.5, 6 and 12.

2.5. ζ Potential Measurement.

ZetaPals zeta potential analyser (Brookhaven Instrument, US) was used to measure the electrophoretic

mobility of NCC particles in suspension at different salt concentration. Zeta (ζ) potentials were

obtained from converting mobility values via Smoluchowski equation. Measurement of ζ potential for

the high-salinity [> 10mM] NCC suspensions was difficult due to rapid agglomeration. Results were

obtained by careful experimental technique as follows. The measurement chute was initially filled

with only saline solution, to which a drop of dilute NCC suspension was added. The first two data

values from instrument were recorded. The same procedure was repeated ten times and twenty data

values in total were obtained; the reported values are the mean of these measurements. Due to rapid

agglomeration at higher salinities, ζ potential was only measurable for concentrations up to 0.07 M

NaCl.

2.6. Dynamic Light Scattering (DLS).

DLS measurements were made using BI-200SM light scattering instrument (Brookhaven Instrument,

US). The detection was made at a scattering angle of 173°. The intensity size distributions f (Rh) were

obtained using CONTIN algorithm to analyse correlation function in the instrument software where

the hydrodynamic radius (Rh) is calculated using Stokes-Einstein relation;

(3)

where kb is Boltzmann constant, T is temperature, η0 is the viscosity of solvent, and D is translational

diffusion constant. Samples with salt water are prepared in the same way as that for ζ potential

measurements.

2.7. Scanning Electron Microscope.

The structure of 1 wt% and 5 wt% NCC suspension with salt concentration 0, 0.01 and 0.1 M was

investigated. The suspensions were freeze-dried at -68 °C. The resultant samples were foam-like

structures, and NCC suspensions at higher concentrations of salt were denser and stronger while NCC

with lower concentrations of salt was softer. Dried samples were ground to fine powder and putter-

coated with platinum and imaged using a Hitachi S-4800 ultra-high definition scanning electron

microscope (5 kV).

2.8. Atomic Force Microscope.

Samples for imaging using an AFM were produced by placing a drop of 0.05 wt% NCC aqueous

suspension onto a freshly cleaved mica surface and allowing subsequent evaporation at room

temperature. For pictures with isolated NCC particles, the method from Lahiji et al. [23] is used. This

involved placing 2 or 3 drops of 1 wt% NCC suspension onto a freshly cleaved mica plate in order to

gain better adsorption, followed by rinsing using excess deionised water before being dried and blown

by N2 gas. Samples were imaged using WI Tec Alpha 300 AFM machine.

3. Results and Discussion

3.1. Rheology of NCC Suspensions in water.

Figure 1 (a) and (b) depict the steady state flow curves of NCC dispersed in deionized water in a

concentration range from 1 to 11.9 wt%. NCC suspensions are near-Newtonian up to 3 wt%; the flow

curves show minimal shear thinning (Figure 1a, b) and a power-law response between stress and

strain rate with exponents of between 0.9 and 1 (Table 1). For samples with 4 to 7 wt%, at low-shear

rates (<10 s-1) the suspensions exhibit slight shear thinning (i.e. not quite a zero-shear viscosity

plateau) but at higher shear rates there is significant shear thinning with a power law exponent

decreasing for each sample from 0.74 to 0.29 respectively. The shape of the curve is reminiscent of

samples undergoing slip (e.g. [21]), but an identical curve is observed when comparing rough and

smooth surfaces to indicate that slip does not to play a role here, as shown in supplementary material.

The slight shear thinning at low shear rates may thus indicate some weak associations between the

rods, and it has been suggested previously to be associated with variations on the disclination lines of

the polydomain texture in the liquid crystal[24]; the origin for this initial shear thinning region is still

currently in debate [13, 25, 26]. At higher concentrations, the samples exhibit shear thinning power

law behaviour at all measured shear rates, with the exponent approaching zero for 11.8% NCC to

indicate that the response is characteristic of a yield stress fluid. This phenomenon is attributed to the

formation of closely packed structure, which inhibits the particle motion due to an excessive

concentration[13, 27]. Some literature suggests that this non-linear power-law behaviour i.e. a plateau

region in a flow curve is typically expected for lyotropic liquid crystal phase formed in rods

suspensions where rods can align themselves in direction of flow at a high shear rate [13, 28]

3.2. Rheology of NCC Samples as a function of pH and Salinity.

The effect of pH and salinity on NCC suspensions was studied for the 5 wt% NCC suspension. The

effect of pH on the rheology of this NCC suspension is shown in Figure 2(a), which shows the

viscosity at shear rates of 0.1, 1 or 10 s-1

. A peak in viscosity is observed at pH 6. The viscosity

decreases steadily as pH values are lowered to below 2 or increased to above 12, beyond which

viscosity at each shear rate increases dramatically.

Salinity has a significant effect on the rheology of NCC suspensions, as shown in Figure 2(b). The

viscosities at the three shear rates shown decrease with addition of NaCl up to a concentration of 0.01

M. The viscosity values pass through a minimum and increases slightly with further increases in NaCl

until a dramatic increase occurs at a concentration of ca. 0.1 M. At this concentration, the viscosity

increases by two orders of magnitude and the suspensions visually transitions from being transparent

to opaque, as shown in Figure 2 (d). An additional experiment (Figure ii (a) in supplementary

material) showed no change in rheology for samples containing 0.1 M NaCl concentration when the

pH was varied between 1 and 13. Figure ii (b) in supplementary material shows that the pH of

solution did not cause additional changes to the ζ potential of NCC particle suspended in 0.01M NaCl

besides the increasing ionic strength due to added acid or base. It suggests that pH and salinity

influence the NCC suspension via mechanisms controlled by the ionic strength of the suspension.

NCC gels form due to high ionic strength and these show significant shear thinning as evident in

Figure 2b. The low viscosity at high shear rate suggests that the contacts between rods are relatively

weak and can be perturbed by external deformation.

Figure 3 presents the yield stress (σy) and linear viscoelastic moduli (G’ and G’’ at 6.28 rad/s) for

suspensions at 0.1 M NaCl as a function of NCC concentration. The yield stress is the minimum shear

stress required to disrupt the gel network structure to enable the system to flow. The cross-over

between G’ and G’’ occurs at a concentration of 0.83 wt% NCC, which is one indicator for the

transition point of the suspension between liquid and solid-like (gel) rheology. G’, G”, and σy each

show a power law dependency on NCC concentration (solid lines in Figure 3) with an exponent of

around 3; this exponent is commonly observed for percolating systems and colloidal gels [5, 29].

The rheological behaviour of NCC suspension in various salt concentrations can be divided into three

regions, and the transitions between them can be generally explained by DLVO theory as follows:

(1) At low ionic strength (below 10 mM of NaCl), external ions compress the surface double layer of

particles, which therefore reduces the effective volume fraction and thus viscosity. At this stage, the

original liquid crystalline gel structure is weaker but generally retained.

(2) When salt concentration is increased to about 10 mM, the inter-particle repulsion is reduced by

sufficiently high ionic strength through the electro-screening effect. This allows for potential contact

to occur between the colloidal rods and thus formation of large agglomerations[15].

(3) At NaCl concentrations at and above 0.1 M, the NCC rods agglomerate into a network gel

structure. Development of the network structure may involve a growth of aggregates into fractals due

to attractive surface forces, and subsequent percolation as the fractals inter-link. The surface forces

are attributed to van der Waals attraction between NCC rods and sufficient particle proximity for

hydrogen bond formation to occur between them. As a consequence, the suspension becomes opaque

due to the transition from an ordered-to-disordered structure, and consequently, the viscosity and gel

strength increase with increasing salt concentration.

3.3. Morphology and Colloidal Properties of NCC.

The following techniques are used to explore the dependence of the rheology on salinity, which

controls the surface properties of the NCC particles and the suspension microstructure: FTIR, to

show the surface functional group of NCC particles; ζ potential measurements to show the

electrostatic stability of suspension; and DLS, to show hydrodynamic particle size distribution and

degree of agglomeration.

3.3.1. Surface Chemistry and Charge on NCC

Figure 4 shows FTIR results for the surface functional groups on NCC under different solution

chemistry. Characteristic spectral peaks are observed at 3400 cm-1

and 2800 cm-1

that are due to

stretching of OH- and CH- bonds respectively[30]. Absorption at around 1640 cm-1 is ascribed to

bound water within NCC. Peaks at around 1371 cm-1

and 897 cm-1

originate from the O-H and C-H

vibration; while 1030 cm-1

and 665cm-1 are from stretching of C-O and C-OH bonds[31]. Peaks at

1200 to 1240 cm-1

signal the presence of sulphate groups on the surface of NCC particles[32]. NCC

suspensions at different pH values show almost identical IR spectra, indicating that no chemical

change occurs at the NCC particle surface due to high or low pH and ionic strength.

ζ potential measurements are shown in Figure 5. Colloidal suspensions are predicted to be

electrostatically stable if the colloidal particles have an absolute ζ potential above 30 mV[5]. This

value of ζ potential is obtained at between 5 and 10 mM NaCl. Figure 5 also includes the

corresponding suspension viscosity, whereby a minimum in viscosity is observed at a concentration of

10 mM NaCl. The decrease in viscosity with salinity is likely to arise from compression of surface

double layer that reduces the specific volume occupied by the NCC rods and thus the effective

volume fraction. As salt levels are increased, electrostatic repulsion is reduced and attractive particle

interactions dominate to promote aggregation. Above 10 mM, ζ potential is approximately constant

with increasing NaCl concentration up to the maximum tested concentration of 70 mM. This

independence of ζ potential with salt level suggests there is no sodium ions adsorbed onto the surface

of NCC particles above 10 mM. However, the apparent viscosity values gradually increase with salt

concentration in this regime. We attribute this to dynamic aspects of the gel formation process; that is,

an interconnected network structure forms slowly at salinities slightly above the critical point, but

more rapidly the greater the salinity. At salt concentrations above 70 mM, we visually observed that

large NCC aggregates form instantaneously.

3.3.2. Morphology and Aggregation of NCC

The rod-like nature of NCC particles was evaluated using AFM, as shown in Figure 6. An ordered

structure is shown in figure 6 (a) for NCC suspension when “dropped” onto a mica plate. Figure 6 (b)

shows an example of an isolated NCC rod, obtained following methods described in experimental

section (Figure iii in suplementary material includes an AFM image of the mica plate containing

numerous isolated NCC rods). From measurements performed on twenty individual NCC particles,

the average (sd = standard deviation) length and diameter is 210 (sd = 10) and 15 (sd = 4) nm

respectively, and the average aspect ratio is 14 (range of 10-20).

Figure 7 shows the distribution of NCC particle sizes at different NaCl concentrations, measured

using DLS. Note, this does not measure the actual particle size, but rather their equivalent

hydrodynamic radius; for isolated (non-aggregating) rod-like particles with large aspect ratio, this is

comparable to its length although not a measure of it [5]. While DLS does not provide an accurate

measure of the size and morphology of the NCC rods in the same manner that is possible with AFM,

we use it to compare the effect of solution variables on the degree of aggregation.

The DLS measurement indicate that for the NCC suspension at 0 mM NaCl, there are two

characteristic peaks present, which correspond to particles with a hydrodynamic radius of mostly ca.

100 nm and a small portion of particles with diameter hydrodynamic of ca. 10 nm. The small peak

may be remnants from the hydrolysis process [33, 34]. Both characteristic peaks are present when

NaCl is increased to 10 mM, but the measured hydrodynamic radius is greater which indicates the

formation of aggregates. Further increase in NaCl causes the smaller peak to disappear and the larger

peak to increase further; the peak hydrodynamic particle radius at both 50 mM and 70 mM NaCl is

observed to be similar at a value of 200 nm. We note that it is around 10 mM that destabilisation and

aggregation is predicted to occur from ζ potential measurements in Figure 6 due to a weakening of

repulsive electrostatic forces, which is thus supported by DLS measurements.

3.4. Microscopic Analysis by Scanning Electron Microscope

Microscopic analysis was performed using SEM to determine how the microstructure of NCC

suspensions alters as a function of NaCl concentration, to support proposed mechanisms for observed

changes in rheology, surface potential and particle size (aggregation). SEM results for 5 wt% NCC

suspension in 0, 0.01 and 0.1 M NaCl, and 1wt% NCC suspension in 0.01 and 0.1 M NaCl are shown

in Figure 8 and Figure 9 respectively. Note, the size bar in the left hand images in Figure 8 and 9 is 3

microns, which is ca. 15 times greater than the length of the isolated NCC rod measured on the AFM.

Characteristic changes in microstructure are observed upon varying NCC and NaCl concentrations.

Interpretation of SEM images of the type observed are well described in the literature for various

suspensions (e.g. [35-37]). As noted in [35], the apparent structures observed in SEM are affected by

the nucleation and growth rate of ice crystals which are ultimately controlled by the freezing rate and

sub-cooling temperature[35]. Figure 8 (a) shows a layered porous lamellar structure for 5 wt% NCC

in 0 mM NaCl; layered structures are typically seen as the result of ice crystal growth during SEM

sample preparation [35]. The structure is indicative of NCC rods aligning in the direction of ice

growth, and there is no indication of fractals (percolation or aggregation) in this state. Figure 10

provides a schematic of the effect of freezing on an anisotropic suspension of rods, based on the

qualitative description proposed for a similar system [38]. It is seen in Figure 10 that the free

colloidal particles can align with the ice crystal growth whilst the fractal and gelled structure resit

such alignment, which therefore results in different structural morphologies in the SEM images.

When the salt concentration is increased to 10 mM NaCl for the 5 wt% NCC suspension, the structure

is completely altered as shown in Figure 8 (b, c). Two structures were observed in SEM images;

figure 8b shows a ‘collapsed’ porous lamellar structure, whereby the original network structure is

disrupted by a small amount of ions. The measured surface in this case was rougher, thicker with

broken pores still seen in the image. As this salt level corresponds to the point of destabilisation, the

images suggest that the aggregation of NCC rods form irregularly shaped particles prevents the

formation of the smooth layered structure along the ice growth direction. We also observe large,

spherical clusters together with collapsed lamella, as shown in figure 8c. The second schematic in

Figure 10 illustrates this process.

When NaCl concentration is increased to 100 mM NaCl for 5 wt% NCC suspension, which also

corresponds to when a very high viscosity and gel strength is observed, the SEM images display a

thick and rough layered structure consisting of large fibrils (Figure 8, d). At a higher magnification,

figure 8e shows a fractured section to comprise of densely aggregated rods. This observation is in

agreement with our proposed gelling mechanism i.e. NCC particles aggregate into larger and longer

fibrils and then these fibrils form entanglements to create a gelled network structure. We observed

that this particular suspension is a strong gel (G’ >> G’’), and thus we suggest that such a structure

resists ice crystal growth along the preferred direction in its crystallography as shown in the third

schematic of Figure 10.

Figure 9 shows the SEM images for suspensions at a lower NCC concentration, 1 wt% NCC. At 10

mM, figures 9a shows the presence of clusters, which is particularly noticeable in the less-magnified

image in figure 8b. These clusters are smaller in size than that at 5 wt% NCC sample, and no network

or lamella structure is observed. This indicates that the sample contained too few particles to form an

interconnected network structure (required for a gel), which is in agreement with the rheological

measurement that showed liquid behaviour. At the higher salt concentration of 100 mM, dual phases

were observed in the suspensions at 1wt% NCC in Figure 9(c, d). Fractured, non-aligned and rough

surface structure are observed in figure 9d, indicating small irregularly shaped aggregates and

potentially interconnections between aggregates. However, the layer is much less thick and dense than

that observed for the 5wt% NCC suspension in Figure 8 (d,e), indicating that the interactions are

weaker and less numerous. Also, large clusters are present as observed in Figure 9c. This is consistent

with our rheology measurement (Figure 3), whereby 1 wt% NCC at 0.1 M NaCl condition

corresponds to the transition point between liquid-like and gel-like rheology.

3.5 Phase diagram

The salt (NaCl)-concentration phase diagram of NCC suspension is summarised in Figure 11, based

on our rheological and structural characterisation. At this stage, the lines are indicative only of where

phase transitions are likely to occur based on our current measurements.

The diagram shows that as the concentration of NCC is increased (lowest NaCl level), the suspension

transitions from being a viscous fluid, where NCC behave as isotropic colloidal rods, to an anisotropic

liquid with an ordered structure (i.e. ‘anisotropic suspension’), where the approximated boundary is

indicated by line E in Figure 11. Further increase in NCC results in the transition (line F) to a

structure that behaves as a viscoelastic soft solid (G’>G’’). These transition boundaries are

anticipated to be quantifiable in future work by considering the specific volume (and thus the effective

phase volume) of the NCC rods in different solution environments.

At a constant NCC concentration (1 and 5wt% NCC in this study), as salt concentration is increased,

the electrical double layer on NCC rods is reduced that leads to aggregation and an unstable regime

(labelled ‘destabilised suspension’). Suspensions in this region show a lower viscosity than without

salt at the same NCC concentration, which is attributed to phase separation that results in the

formation of micron-sized aggregates. However, at larger NaCl concentration, the NCC rods come

into close contact to form an interconnected network structure that has rheological characteristics of a

viscoelastic soft solid (i.e. G’> G’’ over an appreciable frequency range and an apparent yield stress).

Line A is the estimated boundary for gelation (liquid-solid transition) based on our experimental

observations. A goal of future work is to predict this phase boundary using computational models [39]

of the percolation process and/or more detailed rheological characterisation[40]. We consider that

line B corresponds to the point at which clusters of rods begin to form due to attractive interactions,

but interconnections between rods is not extensive enough to form a percolated network structure.

Therefore the region bounded by line A and B has the potential for microphase separation [41, 42]

and is in a transient state between liquid and solid. Macroscopic phase separation is not observed due

to the density of cellulose being close to that of the solvent, water.

The viscoelastic soft solid phases at either high NaCl or high NCC concentrations (above line F) are

characterised by G’> G’’ over an appreciable frequency range, as well as an apparent yield stress and

shear thinning under non-linear shear conditions. Line C and D separate the solid states of NCC

suspension formed at low and high ionic strengths. We believe that there can be structural difference

between these two solids because the particle interaction is repulsive below line D and is attractive

above it according to ζ potential results shown here and DLVO theory [43, 44]. This is a fascinating

area in which to pursue future work to quantify the convoluted structural states arising from a

combination of interactive forces between colloidal rods and crowding.

The phase diagram is based on the experimental results of the NCC we used, but the respective

positions of the proposed boundaries are anticipated to depend on the particular source of NCC used,

including their size, aspect ratio, and surface functional groups. In comparison to the monovalent

sodium salt used in this work, which is relatively well understood in terms of electrostatic screening

effect common in many charged colloidal systems, we should note that the structure-property

relationships for multivalent ions may be quite different due to the potential for ion-mediated cross-

links between NCCs [10].

4. Conclusions

This study builds on previously reported literature on the shear rheology, colloidal stability and

gelation behaviour of NCC suspensions[10, 11, 13, 45, 46], by significantly expanding the range of

examination in terms of concentration of NCC and NaCl, as well as pH. A key finding from the work

is a phase diagram (figure 11), which specifically utilises our rheological and structural

characterisation to provide insight into the state of the suspension. The concentration of NCC and

NaCl in the phase diagram extends to more than double that of previous works [10], and thus we show

many additional phases that arise due to both NCC’s anisotropy and its surface charge. Our phase

diagram demonstrates clear domains where suspensions are a stable viscoelastic solid and viscous

liquid, whilst also showing a region in NCC and NaCl concentration where it was unstable between

these two extremes. The region of instability is triggered by phase separation due to reduced

repulsion between particles, as observed in other similar cellulose and non-spherical colloidal systems

[47-49] although it has not been previously reported experimentally for other colloidal rods systems

that we are aware. In this unstable region, we hypothesise that salt levels are sufficient to decrease the

double layer and thus decrease repulsive forces between NCC rods, but there is insufficient NCC to

form an interconnected network structure.

It is anticipated that the rich set of phenomena observed in this NCC system as a suspension of

colloidal rods will inspire fundamental investigations into their dynamics under a range of conditions.

This includes their dynamics as colloidal glasses and gels, and development of specific rheology-

microstructural relationships in each phase domain and utilisation of DVLO theory with

computational studies to predict the boundaries between them.

The results from this work also provide a basis for controlling the rheology of this NCC in a variety of

current and emerging applications, whilst also establishing a basis for comparison in future work that

examines the effect on the structure-rheology phase diagram by altering the properties of the NCC.

NCC properties of high interest include: (i) surface functionalisation and/or interactions with

surfactants [12, 50]; (ii) examining NCC from a variety of feedstocks and processing methodologies

to vary aspect ratio, crystallinity, hemicellulose content and surface charge, with an example being

recently reported high aspect ratio cellulose nanofibers that are obtained from spinifex grasses[51, 52].

Current and emerging applications for NCC where its rheology-structure is relevant include its use in

fast moving consumer goods such as food, personal care, pharmaceutical and coating products [53-55],

as well as during processing towards the application of NCC as unique fillers in advanced materials

and biomaterials [56], whilst an emerging application is in developing novel nanostructured materials

using so-called freeze-casting process [35, 38]. In addition, our future work seeks to build on the

foundation set in this paper and examine the use of NCC within drilling fluids [6] for accessing

underground energy resources (oil, gas, geothermal), where precise rheology control and thermal

stability is needed under variable environmental conditions such as salt/pH as well as high pressure

and temperatures.

Acknowledgements

Yuan Xu thanks the University of Queensland (UQ) for an International (UQI) Scholarship. This

work was performed in part at the Queensland node of the Australian National Fabrication Facility

(ANFF-Q), a company established under the National Collaborative Research Infrastructure Strategy

to provide nano and micro-fabrication facilities for Australia’s researchers. We thank: Kinnari Shelat

from ANFF-Q for her assistance in making AFM measurements; Xiaotong Li from School of

Chemistry and Molecular Bioscience (UQ) for assistance in measuring infra-red spectra; Gleb

Yakubov from School of Chemical Engineering (UQ) for the helpful discussion about AFM, cellulose

and colloid science; and Jian Xu from Department of Chemistry and Chemical Engineering,

Shandong University, China for the advice on experiment design and data interpretation. The research

is partially supported by the Australian Research Council Centre of Excellence in Plant Cell Walls

(CE110001007).

References

1. Revol, J.F., et al., Helicoidal self-ordering of cellulose microfibrils in aqueous suspension. International Journal of Biological Macromolecules, 1992. 14(3): p. 170-172.

2. Favier, V., H. Chanzy, and J.Y. Cavaille, Polymer Nanocomposites Reinforced by Cellulose Whiskers. Macromolecules, 1995. 28(18): p. 6365-6367.

3. Azizi Samir, M.A., F. Alloin, and A. Dufresne, Review of recent research into cellulosic whiskers, their properties and their application in nanocomposite field. Biomacromolecules, 2005. 6(2): p. 612-26.

4. Eichhorn, S.J., Cellulose nanowhiskers: promising materials for advanced applications. Soft Matter, 2011. 7(2): p. 303-315.

5. Mewis, J. and N.J. Wagner, Colloidal Suspension Rheology. 2012, New York: Cambridge University Press.

6. Li, M.C., et al., Cellulose nanoparticles as modifiers for rheology and fluid loss in bentonite water-based fluids. ACS Appl Mater Interfaces, 2015. 7(8): p. 5006-16.

7. Wallace, D., Collagen gel systems for sustained delivery and tissue engineering. Advanced Drug Delivery Reviews, 2003. 55(12): p. 1631-1649.

8. Klemm, D., et al., Nanocelluloses as Innovative Polymers in Research and Application. 2006. 205: p. 49-96.

9. Habibi, Y., L.A. Lucia, and O.J. Rojas, Cellulose nanocrystals: chemistry, self-assembly, and applications. Chem Rev, 2010. 110(6): p. 3479-500.

10. Chau, M., et al., Ion-Mediated Gelation of Aqueous Suspensions of Cellulose Nanocrystals. Biomacromolecules, 2015. 16(8): p. 2455-62.

11. Shafiei-Sabet, S., W.Y. Hamad, and S.G. Hatzikiriakos, Rheology of nanocrystalline cellulose aqueous suspensions. Langmuir, 2012. 28(49): p. 17124-33.

12. Eyley, S. and W. Thielemans, Surface modification of cellulose nanocrystals. Nanoscale, 2014. 6(14): p. 7764-79.

13. Ureña-Benavides, E.E., et al., Rheology and Phase Behavior of Lyotropic Cellulose Nanocrystal Suspensions. Macromolecules, 2011. 44(22): p. 8990-8998.

14. Hirai, A., et al., Phase separation behavior in aqueous suspensions of bacterial cellulose nanocrystals prepared by sulfuric acid treatment. Langmuir, 2009. 25(1): p. 497-502.

15. Shafiei-Sabet, S., W.Y. Hamad, and S.G. Hatzikiriakos, Ionic strength effects on the microstructure and shear rheology of cellulose nanocrystal suspensions. Cellulose, 2014. 21(5): p. 3347-3359.

16. de Souza Lima, M.M. and R. Borsali, Rodlike Cellulose Microcrystals: Structure, Properties, and Applications. Macromolecular Rapid Communications, 2004. 25(7): p. 771-787.

17. Thielemans, W., C.R. Warbey, and D.A. Walsh, Permselective nanostructured membranes based on cellulose nanowhiskers. Green Chemistry, 2009. 11(4): p. 531.

18. Kontturi, E. and T. Vuorinen, Indirect evidence of supramolecular changes within cellulose microfibrils of chemical pulp fibers upon drying. Cellulose, 2008. 16(1): p. 65-74.

19. Kim, J.-H., et al., Review of nanocellulose for sustainable future materials. International Journal of Precision Engineering and Manufacturing-Green Technology, 2015. 2(2): p. 197-213.

20. Cai, J., et al., Science of Cellulose Materials. 2015, Beijing: Chemical Industry Press. 21. Stokes, J.R. and J.H. Telford, Measuring the yield behaviour of structured fluids. Journal of

Non-Newtonian Fluid Mechanics, 2004. 124(1-3): p. 137-146. 22. Dzuy, N.Q. and D.V. Boger, Direct Yield Stress Measurement with the Vane Method. Journal

of Rheology, 1985. 29(3): p. 335-347. 23. Lahiji, R.R., et al., Atomic force microscopy characterization of cellulose nanocrystals.

Langmuir, 2010. 26(6): p. 4480-8.

24. Walker, L. and N. Wagner, Rheology of region I flow in a lyotropic liquid‐crystal polymer: The effects of defect texture. Journal of Rheology, 1994. 38(5): p. 1525-1547.

25. Burghardt, W.R., Molecular orientation and rheology in sheared lyotropic liquid crystalline polymers. Macromolecular Chemistry and Physics, 1998. 199(4): p. 471-488.

26. Hongladarom, K. and W.R. Burghardt, Molecular orientation, "Region I" shear thinning and the cholesteric phase in aqueous hydroxypropylcellulose under shear. Rheologica Acta, 1998. 37(1): p. 46-53.

27. Liu, D., et al., Structure and rheology of nanocrystalline cellulose. Carbohydrate Polymers, 2011. 84(1): p. 316-322.

28. Shigharu, O. and A. Tadahiro, Rheology and Rheo-Optics of Polymer Liquid Crystals, in Proceeding of the eighth international congress on rheology. 1980, Plenum Press: New York. p. 127-147.

29. van der Aerschot, E. and J. Mewis, Equilibrium properties of reversibly flocculated dispersions. Colloids and Surfaces, 1992. 69(1): p. 15-22.

30. Tang, Y., et al., Preparation and characterization of nanocrystalline cellulose via low-intensity ultrasonic-assisted sulfuric acid hydrolysis. Cellulose, 2013. 21(1): p. 335-346.

31. Li, J., et al., Homogeneous isolation of nanocellulose from sugarcane bagasse by high pressure homogenization. Carbohydr Polym, 2012. 90(4): p. 1609-13.

32. Zeng, X., et al., Characterization of sodium cellulose sulphate/poly-dimethyl-diallyl-ammonium chloride biological capsules for immobilized cultivation of microalgae. Journal of Chemical Technology & Biotechnology, 2013. 88(4): p. 599-605.

33. Lu, A., et al., Investigation on metastable solution of cellulose dissolved in NaOH/urea aqueous system at low temperature. J Phys Chem B, 2011. 115(44): p. 12801-8.

34. Cai, J., et al., Dynamic Self-Assembly Induced Rapid Dissolution of Cellulose at Low Temperatures. Macromolecules, 2008. 41(23): p. 9345-9351.

35. Deville, S., et al., Freezing as a path to build complex composites. Science, 2006. 311(5760): p. 515-8.

36. Deville, S., E. Saiz, and A.P. Tomsia, Freeze casting of hydroxyapatite scaffolds for bone tissue engineering. Biomaterials, 2006. 27(32): p. 5480-9.

37. Deville, S., E. Saiz, and A.P. Tomsia, Ice-templated porous alumina structures. Acta Materialia, 2007. 55(6): p. 1965-1974.

38. Dash, R., Y. Li, and A.J. Ragauskas, Cellulose nanowhisker foams by freeze casting. Carbohydrate Polymers, 2012. 88(2): p. 789-792.

39. Coniglio, A., et al., Percolation, gelation and dynamical behaviour in colloids. Journal of Physics: Condensed Matter, 2004. 16(42): p. S4831-S4839.

40. Winter, H.H. and M. Mours, Rheology of Polymers Near Liquid-Solid Transitions. 1997. 134: p. 165-234.

41. Zaccarelli, E., Colloidal gels: equilibrium and non-equilibrium routes. Journal of Physics: Condensed Matter, 2007. 19(32): p. 323101.

42. Charbonneau, P. and D.R. Reichman, Phase behavior and far-from-equilibrium gelation in charged attractive colloids. Phys Rev E Stat Nonlin Soft Matter Phys, 2007. 75(5 Pt 1): p. 050401.

43. Saunders, J.M., et al., A Small-Angle X-ray Scattering Study of the Structure of Aqueous Laponite Dispersions. The Journal of Physical Chemistry B, 1999. 103(43): p. 9211-9218.

44. Secor, R.B. and C.J. Radke, Spillover of the diffuse double layer on montmorillonite particles. Journal of Colloid and Interface Science, 1985. 103(1): p. 237-244.

45. Lu, A., et al., Unique viscoelastic behaviors of colloidal nanocrystalline cellulose aqueous suspensions. Cellulose, 2014. 21(3): p. 1239-1250.

46. Prathapan, R., et al., Modulating the zeta potential of cellulose nanocrystals using salts and surfactants. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 2016. 509: p. 11-18.

47. Appaw, C., et al., Phase separation and heat-induced gelation characteristics of cellulose acetate in a mixed solvent system. Cellulose, 2010. 17(3): p. 533-538.

48. Takahashi, M., M. Shimazaki, and J. Yamamoto, Thermoreversible gelation and phase separation in aqueous methyl cellulose solutions. Journal of Polymer Science Part B: Polymer Physics, 2001. 39(1): p. 91-100.

49. Cummins, H.Z., Liquid, glass, gel: The phases of colloidal Laponite. Journal of Non-Crystalline Solids, 2007. 353(41-43): p. 3891-3905.

50. Hu, Z., et al., Tuning cellulose nanocrystal gelation with polysaccharides and surfactants. Langmuir, 2014. 30(10): p. 2684-92.

51. Raj, P., et al., Gel point as a measure of cellulose nanofibre quality and feedstock development with mechanical energy. Cellulose, 2016. 23(5): p. 3051-3064.

52. Amiralian, N., et al., Easily deconstructed, high aspect ratio cellulose nanofibres from Triodia pungens; an abundant grass of Australia's arid zone. RSC Adv., 2015. 5(41): p. 32124-32132.

53. Gómez H, C., et al., Vegetable nanocellulose in food science: A review. Food Hydrocolloids, 2016. 57: p. 178-186.

54. Qing, W., et al., The modified nanocrystalline cellulose for hydrophobic drug delivery. Applied Surface Science, 2016. 366: p. 404-409.

55. Li, F., et al., Multi-functional coating of cellulose nanocrystals for flexible packaging applications. Cellulose, 2013. 20(5): p. 2491-2504.

56. Yu, H.-y., Z.-y. Qin, and Z. Zhou, Cellulose nanocrystals as green fillers to improve crystallization and hydrophilic property of poly(3-hydroxybutyrate-co-3-hydroxyvalerate). Progress in Natural Science: Materials International, 2011. 21(6): p. 478-484.

Table 1 A list of the Power law ( n) parameters obtained from fitting regions of the flow

curves presented in figure 1.

NCC wt% 1 2 3 4 5 7 9 11.9

K (Pasn) 0.002 0.0052 0.011 0.043 0.13 2.2 6.7 51.5

n 1 0.96 0.91 0.74 0.64 0.29 0.22 0.085

List of Figures

Figure 1. Shear rheology of 1 to 11.9 wt% NCC dispersed in water, showing (a) apparent viscosity as

a function of shear rate and (b) shear stress against shear rate. Solid lines are power law fits to regions

of the flow curves (see Table 1).

Figure 2. Steady-state viscosity at 3 shear rates for 5 wt% NCC suspension as a function of (a) pH, (b)

salt concentration, and (c) as a function of NCC wt% at 0.1 M NaCl. (d) Shows the relative

translucency of suspensions containing 5 wt% NCC at 0, 0.01 and 0.1 M of NaCl respectively.

Figure 3. Apparent yield stress (σy), storage (G’) and loss (G’’) modulus at frequency of 6.28 rad/s as

a function of NCC concentration (c, wt%) at 0.1 M NaCl for the solid-like suspensions (G’ > G’’).

Solid lines are power law fits to the data, according to: G’=8.6 c2.7

; G’’= 1.6 c2 ; and σy=0.5 c

3.1 .

Figure 4. IR spectra for NCC at different pH values. (Vertical axis was not scaled)

Figure 5. Zeta potential of NCC particles (dilute concentrations) and shear viscosity at 1 s-1

for 5 wt%

NCC suspensions as a function of NaCl concentration. Error bars are standards deviations. Dashed

lines are included to guide the eye. The vertical grey line corresponds to the NaCl concentration at

which the particle potential is ca. -30 mV, above which the suspension is no longer electrostatically

stable due to the electrostatic screening effect.

Figure 6.AFM images for NCC on mica substrate, showing: (a) suspension of NCC colloidal rods; (b)

an isolated NCC colloidal rod with a measure of its length.

Figure 7. Measurement of particle size distribution of dilute NCC suspensions for varying NaCl

concentrations. The vertical axis is the distribution function based on intensity of scattered light and is

calculated from CONTIN analysis; and the horizontal axis is hydrodynamic radius for an equivalent

sphere.

Figure 8. SEM images of 5 wt% NCC suspensions in: (a) water, (b,c) 0.01 M NaCl, and (d, e) 0.1 M

NaCl.

Figure 9. A representative sample of SEM images for 1 wt% NCC suspensions in: (a, b) 0.01 M NaCl,

and (c, d) 0.1M NaCl.

Figure 10. Schematic of NCC rods (short lines) before and after freezing (short dashed arrows

showing the direction of ice crystal growth) arising from SEM procedures.

Figure 11. A schematic of the phase behaviour of NCC suspension with different concentration of

NaCl concentration in water based on the microstructure and rheological measurements performed in

this work (circles). Although dashed lines are used to indicate the boundaries between phases, further

experiments are still required to substantiate these as well as the properties of each phase. A to F are

labels of boundary lines. Note, the axes are not scaled.

(a)

Shear Rate (s-1

)

10-2 10-1 100 101 102 103 104

Vis

co

sity (

Pa.s

)

0.001

0.01

0.1

1

10

100

1000

10000water

1 wt% NCC

2 wt% NCC

3 wt% NCC

4 wt% NCC

5 wt% NCC

7 wt% NCC

9 wt% NCC

11.9 wt% NCC

(b)

Shear Rate (s-1

)

10-2 10-1 100 101 102 103 104

She

ar

Str

ess (

Pa)

0.001

0.01

0.1

1

10

100

Figure 1. Shear rheology of 1 to 11.9 wt% NCC dispersed in water, showing (a) apparent viscosity as

a function of shear rate and (b) shear stress against shear rate. Solid lines are power law fits to regions

of the flow curves (see Table 1).

NaCl concentration M

0.001 0.01 0.1 1 10

Vis

co

sity (

Pa.s

)

0.1

1

10

100

1000

wt% of NCC

0 1 2 3 4 5

Vis

co

sity (

Pa.s

)

0.1

1

10

100

1000

a

(c)(d)

pH

0 2 4 6 8 10 12 14

Vis

co

sity (

Pa.s

)

0.1

1

10

100

1000

0.1 s-1

1 s-1

10 s-1

(a) (b)

Figure 2. Steady-state viscosity at 3 shear rates for 5 wt% NCC suspension as a function of (a) pH, (b)

salt concentration, and (c) as a function of NCC wt% at 0.1 M NaCl. (d) shows the relative

translucency of suspensions containing 5 wt% NCC at 0, 0.01 and 0.1 M of NaCl respectively.

wt% of NCC

0 1 2 3 4 5 6

G', G

'' o

r y(

Pa)

0.1

1

10

100

1000

G'

G''

y

Figure 3. Apparent yield stress (σy), storage (G’) and loss (G’’) modulus at frequency of 6.28 rad/s as

a function of NCC concentration (c, wt%) at 0.1 M NaCl for the solid-like suspensions (G’ > G’’).

Solid lines are power law fits to the data, according to: G’=8.6 c2.7

; G’’= 1.6 c2 ; and σy=0.5 c

3.1 .

Figure 4. IR spectra for NCC at different pH values. (Vertical axis was not scaled)

NaCl Concentration (M)

0.00 0.02 0.04 0.06 0.08 0.10

Ze

ta P

ote

ntia

l (m

V)

-50

-40

-30

-20

-10

0

Vis

co

sity

(Pa

.s)

0.01

0.1

1

10

100

Viscosity

Zeta Potential

NaCl at = -30 mV

Figure 5. Zeta potential of NCC particles (dilute concentrations) and shear viscosity at 1 s-1

for 5 wt%

NCC suspensions as a function of NaCl concentration. Error bars are standards deviations. Dashed

lines are included to guide the eye. The vertical grey line corresponds to the NaCl concentration at

which the particle potential is ca. -30 mV, above which the suspension is no longer electrostatically

stable due to the electrostatic screening effect.

Figure 6.AFM images for NCC on mica substrate, showing: (a) suspension of NCC colloidal rods;

(b) an isolated NCC colloidal rod with a measure of its length.

Rh, apparent

(nm)

1 10 100 1000

f (R

h)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0 M NaCl

0.01 M NaCl

0.05 M NaCl

0.07 M NaCl

Figure 7. Measurement of particle size distribution of dilute NCC suspensions for varying NaCl

concentrations. The vertical axis is the distribution function based on intensity of scattered light and is

calculated from CONTIN analysis; and the horizontal axis is hydrodynamic radius for an equivalent

sphere.

Figure 8. A representative sample of SEM images for 5 wt% NCC suspensions in: (a) water, (b,c)

0.01 M NaCl, and (d, e) 0.1 M NaCl.

Figure 9. A representative sample of SEM images for 1 wt% NCC suspensions in: (a, b) 0.01 M NaCl,

and (c, d) 0.1M NaCl.

Figure 10. Schematic of NCC rods (short lines) before and after freezing (short dashed arrows

showing the direction of ice crystal growth) arising from SEM procedures.

Figure 11. A schematic of the phase behaviour of NCC suspension with different concentration of

NaCl concentration in water based on the microstructure and rheological measurements performed in

this work (circles). Although dashed lines are used to indicate the boundaries between phases, further

experiments are still required to substantiate these as well as the properties of each phase. A to F are

labels of boundary lines. Note, the axes are not scaled.

Rheology and Microstructure of Aqueous Suspensions of

Nanocrystalline Cellulose Rods

Supplementary material

List of Figures

Figure i – Comparison between rough (emery paper) and smooth parallel plates (diameter = 50 mm) at

1 mm gap for 7 wt% NCC, in comparison to measurements on smooth concentric cylinders in Figure

1. There is negligible difference between the 3 measurements under the conditions tested.

Figure ii – (a) pH effects on viscosity of 5wt% NCC suspension with 0.1 M NaCl and zeta potential of

NCC particle in 0.01M NaCl. (b) pH effects on zeta potential of NCC in 0.01M NaCl plotted as

equivalent NaCl concentration. Dotted line is the zeta potential change due to increasing salinity at

constant pH, and is for reference.

Figure iii – The AFM picture of mica plate with isolated NCC rods. Note that the length scale of these

images is considerably greater than those in figure 6.

Figure i – Comparison between rough (emery paper) and smooth parallel plates (diameter =

50 mm) at 1 mm gap for 7 wt% NCC, in comparison to measurements on smooth concentric cylinders in Figure 1. There is negligible difference between the 3 measurements under the

conditions tested.

0.001

0.01

0.1

1

10

0.001 0.01 0.1 1 10 100 1000 10000

Vis

cosi

ty P

a.s

Shear rate s-1

Data in Figure 1

1.0 mm rough

1.0 mm smooth

Figure ii – (a) pH effects on viscosity of 5wt% NCC suspension with 0.1 M NaCl and zeta potential

of NCC particle in 0.01M NaCl. (b) pH effects on zeta potential of NCC in 0.01M NaCl plotted as

equivalent NaCl concentration. Dotted line is the zeta potential change due to increasing salinity at

constant pH, and is for reference.

-45

-40

-35

-30

-25

-20

-15

-10

-5

0

1

10

100

1000

0 5 10 15

Zeta

Po

ten

tial

mV

Vis

cosi

ty P

a.s

pH

0.1 s-1

1 s-1

10 s-1

Zeta potential

-45

-40

-35

-30

-25

-20

-15

-10

-5

0

0.001 0.01 0.1

Zeta

po

ten

tial

mV

Equivalent NaCl concentration M

0.01M NaCl with varied pH

pH=6.8 with varied salinity

a

b

Figure iii – The AFM picture of mica plate with isolated NCC rods. Note that the length scale of

these images is considerably greater than those in figure 6.

20

15

10

5

0

µm

20151050

µm

-4

-2

0

2

4

nm

5

4

3

2

1

0

µm

543210

µm

-4

-2

0

2

4

nm

a

b

Graphical abstract