Accepted Manuscript

Title: A Comparison of Canthaxanthine Pickering emulsions,

stabilized with cellulose nanocrystals of different origins

Authors: Sheida Hedjazi, Seyed Hadi Razavi

PII: S0141-8130(17)30245-3
DOI: http://dx.doi.org/doi:10.1016/j.ijbiomac.2017.08.030
Reference: BIOMAC 8022

To appear in: International Journal of Biological Macromolecules

Received date: 18-1-2017

Revised date: 25-5-2017

Please cite this article as: Sheida Hedjazi, Seyed Hadi Razavi, A
Comparison of Canthaxanthine Pickering emulsions, stabilized with cellulose
nanocrystals of different origins, International Journal of Biological
Macromoleculeshttp://dx.doi.org/10.1016/j.ijbiomac.2017.08.030

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.
A Comparison of Canthaxanthine Pickering emulsions,
stabilized with cellulose nanocrystals of different
origins

Short title: Canthaxanthin Pickering Emulsion

Sheida Hedjazi1, Seyed Hadi Razavi1,2*

1
Department of Food Science and Engineering, University College of Agriculture and Natural
Resources, University of Tehran, Karaj, Iran.

2
Center of Excellence for Application of Modern Technologies for Producing Functional Foods
and Drinks (FFDCE) and Bioprocess Engineering Laboratory (BPEL), University College of
Agriculture and Natural Resources, University of Tehran, Karaj, Iran.

*Corresponding author: Seyed Hadi Razavi, Email: srazavi@ut.ac.irP. O. Box: 31587-77871, Tel:
+98-26-32248804, Fax: +98-26-32249453

Research highlights

The Structure of CNCs was discussed over AFM imaging and structural changes were
reviewed by XRD and FTIR exams
CTX micro-structure droplets in pickering emulsions with different CNC content from
different source was investigated with DLS and optical microscopy
Emulsion and canthaxanthin stability to different environmental factors such as
temperature, pH, ionic concentration and light examined
Emulsions became more stable as the temperature and pH rose, and ionic concentration fall
 CCN preserved CTX from light more efficiently than BCN

Abstract

Cellulosic nanocrystals from different origins were made to stabilize the canthaxanthin (CTX) in
pickering emulsion. Nanocrystals were obtained by hydrochloric acid hydrolysis. Dynamic light
scattering (DLS) demonstrated that the length of solid particles were in the range of 112nm to
4000nm. AFM indicated the needle-like shape of the cotton cellulose nanocrystals (CCNs) and
bacterial cellulose nanocrystals (BCNs) and also illustrated the thickness of the particles to be 6
and 7 nm respectively. The crystallinity of both BCNs and CCNs was higher than the cellulose
source they were isolated (CL: 75.4%, CCNs: 86.6; BC: 79.2%, BCNs: 88.5%). Effects of different
factors such as temperature, pH and ionic concentration on stability were investigated. The results
revealed that an increase in both temperature and pH was accompanied by an improvement in
emulsion stability in all tested samples. However, increase in ionic concentration resulted in
emulsions with less stability. In all the samples, CCN emulsions had better stability than the BCN
emulsions, which was associated to smaller particle size and more coverage ability. Test of
stability to light showed that CCN emulsions can preserve CTX better than BCN against the light.

Keywords: Canthaxanthin; Cellulose nanocrystal; Pickering emulsion.

1. Introduction

Emulsion is a mixture consisting of two liquids in which one of them is immiscible in another;
therefore, a surface active polymer or a surfactant stabilizes these two phases by reducing the total
energy of the system. However, surfactants (typical stabilizers), are repeatedly adsorbed and
desorbed at the interface which lead to phase separation due to the competition between re-
adsorption and coalescence [1]. The pioneering work of Ramsden and Pickering have opened a
new window to the emulsions science, and it is now well known that solid colloidal particles can
be a good alternative for the conventional emulsion stabilizers. These particles can anchor
themselves irreversibly on the oil-water interface. These generated emulsions are called Pickering
emulsions [2, 3]. The application of surfactant in cosmetics, pharmaceutics and food industries is
undesirable,therefore, the interest in the use of Pickering emulsions have been resumed again in
recent years. These emulsions are environment friendly and ultra-stable. Additionally, the amount
of particles needed to stabilize the emulsion is much less than surfactant stabilized emulsions,
which in turn reduces the overall hazardous aspects [4].

Cellulose is the most abundant source of natural polymer and macromolecule available. It is
presented in a plethora of products [5]. Treating native cellulose to harsh acid hydrolysis can
generate a super crystalline and rod-like nanoparticles suspension [6]. Different sources of
cellulose as well as different hydrolysis conditions have a considerable impact on the size
distribution and properties of the produced nanoparticles [7, 8]. Bacterial cellulose is one of the
available sources to obtain cellulose. For instance, it is presented in different bacterial products
such as Nata de coco, which is produced by fermentation of coconut water by Acetobacter xylinum
[9]. Acid hydrolysis is a miscellaneous process that consists of several steps. Sulfuric acid is
frequently used due to esterification of nanocrystals surfaces with negatively charged groups that
lead to electrostatic stabilization in aqueous solutions [10].

Carotenoids are an abundant class of pigments, consisting of highly unsaturated isoprene

derivatives [11]. These pigments are lipid-soluble compounds and cover the yellow to red spectra
of colors. They are presented in all photosynthetic organisms such as plants and algae and some
non-photosynthetic bacteria and fungi [12]. Comparing the extraction process from plant tissue to
bacterial production of carotenoids, the benefits of using bacteria are clearly evident. Bacterial
production is not seasonal or geographically dependent and the procedure is much cheaper. Also
the final product, due to less possible hazards, has more desirability over synthetic carotenoids
[13]. Current consumers demands for health promoting foods now lead the scientists to work on
added qualities of the final products. Carotenoids are used for various applications such as food
industries, pharmaceuticals, feed supplements and nutraceuticals mostly due to their antioxidant,
anti-inflammatory, antitumor and anticancer properties [14, 15]. Of the main carotenoids, -
carotene, Xantophyl, CTX and lutein are the main compounds with the most industrial interest
[11,16]

Canthaxanthin (CTX), is a ketocarotenoid with high antioxidant activity. Color shades from orange
to red makes it a desirable food colorant because of its enchanting properties. Newly discovered,
CTX production via biological resources gain ever increasing attention, which is mostly due to
possible hazards that synthetic carotenoids may cause in food and pharmaceutical industries [17].
Between different introduced biological sources, Dietzia natronolimnaea HS-1 have shown the
most promising CTX production [18].

Based on available literature and best of our knowledge there are no studies on comparison of
stabilized Canthaxanthin emulsions with cellulose nanocrystals obtained from cotton linter and
bacterial cellulose. The objective of the present paper was to generate a new surfactant free
emulsion system enriched with CTX. CNCs were prepared from cotton linter and bacterial
cellulose, hydrolyzed by sulfuric acid. The cellulose structure and, crystallinity index were
examined on CNCs. Oil/water emulsions containing bacterial CTX were successfully prepared by
sonication. Laser light scattering and optical microscopic studies were carried out to investigate
the emulsion droplet size and morphology. Finally, different environmental factors that may affect
the stability of the emulsions were investigated.

2. Materials and Methods

2.1. Materials

Cotton linter was sourced from Peter Temming AG (Glckstadt, Germany); Nata de coco was
bought at a traditional local market in San Jose (Luzon, Philippines); Canthaxanthin was purchased
from Dr. Ehrenstorfer GmbH (Augsburg, Germany); sulfuric acid was purchased from Merck.
coconut oil was purchased from a local market. The coconut oil was used for emulsion production
because of its high saturated fatty acids, which in turn reduces oxidation problems. As measured
by gas chromatography of methyl esters, the used coconut oil contained the following fatty acids
(mol %): 7.9% C8:0, 6.5% C10:0, 45.6% C12:0, 18.3% C14:0, 8.7% C16:0, 2.9% C18:0, 7.1%
C18:1 and 1.4% C18:2 .

2.2 Methods

2.2.1 Cotton linter Cellulose Nanocrystal (CCN) preparation

CCNs were prepared according to the described method by Revol et.al (1992) [19] with slight
modification. Concisely, the cotton linters were hydrolyzed in 60% wt sulfuric acid at 70C under
constant stirring for 30min. Subsequently, cold water was added to the hydrolyzed cellulose to
stop the reaction. Afterward, the suspension was washed by centrifugation at 10000g for 10 min
and the precipitate was re-dispersed in the deionized water. This procedure was repeated until a
colloidal supernatant was formed. The turbid supernatant was dialyzed using 43mm dialysis tubing
cellulose membrane with molecular weight cut-off of 12KDa (Sigma-Aldrich Co., USA) to reach
neutrality. The final dispersion was sonicated for 15min (Ultrasonic processor XL, 2020, Misonix
inc., NY, USA), and stored at 4C.

2.2.2 Bacterial Cellulose Nanocrystals (BCN) preparation

BCN was obtained from Nata de coco according to the method used by Kalashnikova, et.al. (2011)
[4] with slight changes. Nata de coco cubes were grounded in a blender at high speed. The resulting
slurry was then filtered and maintained in 0.4 M NaOH at 45C for 5h. The alkali was then removed
by washing with distilled water until neutralization. In the next step, bleaching was done by mixing
the suspension with 8 g/L NaClO2 solution in sodium acetate buffer (pH= 4.7) at 45C for 4h. The
bleached cellulose was washed with distilled water until a stable pH was achieved. The BC was
then hydrolyzed with 50% (w/w) H2SO4 at 45C for 2h. The acid was removed by successive
centrifugation at 10000 g for 10 min until a turbid suspension was formed. The resulting
suspension was then homogenized and dialyzed against distilled water to remove contaminants
and was kept at 4C.

2.2.3 X-ray diffraction (XRD)

The degree of crystallinity of cotton linter, CCN, BC, and BCN was measured by X-ray diffraction
(XRD). Analysis was carried out using a D5000 (SIEMENS Ltd., Germany) operating at a power
of 35kV with a current of 30mA and Co K radiation (1.5406 A). The X-ray diffractograms were
obtained at room temperature within a 2 ranging from 5 to 80using a step size of 0.02. The
crystallinity index (CRI) of CNCs was reported using the empirical method as shown in the
following equation:

002
= 100% (1)
002

where I002 is the peak intensity of the (002) lattice diffraction, which is located at the diffraction
angle of around 2= 26.36. Iam is the intensity scattered by amorphous part of the samole and was
measured as the lowest intensity [20].

2.2.4 Attenuated total reflectance (FTIR-ATR) Spectroscopy

FTIR-ATR spectra were recorded on an Equinox 55 Fourier transform instrument from Bruker.
The ATR accessory (Specac Ltd., UK) contained a ZnSe crystal (25mm5mm2mm) at a nominal
incident angle of 45, yielding about 12 internal reflections at the sample surface to obtain spectra
for samples. All spectra were recorded at 25C. The samples were directly placed on the ATR
crystal without further modification.

2.2.5 Size distribution and Zeta potential measurement

All the experiments were done in the diluted CNCs suspension (0.1 g/L) in which the CNC
particles oriented randomly. Before every measurement the suspensions were sonicated for 30 s.
The dynamic light scattering (DLS, Zeta sizer Nano ZS, Malvern instrument) was used to assess
the size of CNCs in suspensions. This method provides indirect information about the size
distribution of CNCs. Results are reported as the average of three measurements.

The zeta potential of CNCs was measured over a range of NaCl concentration (0-100 mM) and
pH (4.5-7.8) at 25C. The electrophoretic mobility of the suspensions were measured in triplicate
with a Malvern 3000 zetasizer NanoSZ (Malvern instrument UK).

2.2.6 Atomic force microscopy

Atomic force microscopy (AFM) was used to characterize the topography and thickness of CNC
particles. The AFM imaging was carried out by a Dual scope/Raster scope C26, Scanner DS95-
50E, DME (Denmark), in non-contact mode. The CNCs samples were diluted with distilled water
to 100g ml-1, sonicated for one minute, and then, dripped onto a cleaved mica and air-dried before
microscopic imaging. The probe for this test was a silicone tip probe with tip curvature<10nm.

2.3 emulsion preparation and characterization

2.3.1 Preparation of Canthaxanthin Pickering emulsion with CNCs

Canthaxanthin was dissolved in coconut oil. As the solubility of CTX in oil at room temperature
is low, it was dissolved in hot coconut oil at a constant ratio of 1:50. According to Gharibzahedi
and et.al (2013) [10], this was the best ratio for optimum CTX solubility. Although, CTX is very
susceptible to degradation at high temperatures, so the exposure time should be restricted. The best
CTX solubility with least oxidation rate was resulted at 97c for 7.5 s under vacuum condition
[10]. The emulsions aqueous phase was at room temperature range. All experiments were done
with the final emulsions volume of 4ml using oil /aqueous phase ratio of 20/80 and different CNCs
concentrations (in aqueous phase: 0.3-0.6-0.9%). Sonication was carried out at 1.5 power level
(3mm titanium probe, 2mm below the surface, 2W.mL-1 applied energy) by alternative 3 s
sonication with 3 s standby for 21 s.

2.3.2 Canthaxanthin Pickering emulsions morphology

In order to visualize the emulsion droplets, optical micrograph was used. The Pickering emulsions
stabilized by CNCs were captured by a Carl Zeiss standard 20 optical microscope fitted with a
digital camera (Generic Digital 1" C-Mount). 20L of emulsion was added to 1mL distilled water
and vortexed then the emulsions droplet was placed on a microscope slide and viewed under 4-
100 magnification.

2.3.3 Droplet diameters measurement

Droplet diameter was measured using ImageJ software and the results were compared to the
droplet size distribution outcome of the laser light diffraction method employing Malvern 2000
granulometer apparatus equipped with a HeNe laser (Malvern Instruments, U.K.) with
Fraunhofer diffraction. The measurements were carried out in triplicate and the diameters were
demonstrated as surface mean diameter D(2,3) (the Sauter diameter).

2.3.4 Zeta potential measurement of emulsions

CNCs stabilized emulsions were examined for their zeta potential using a particle electrophoresis
instrument (Zetasizer Nano ZS, Malvern instruments Ltd., Malvern, and Worcestershire, UK). To
steer clear of multiple scattering effects, the samples were diluted to a concentration of roughly
0.05% w/w using deionized water. The results were reported as the average and standard deviation
of measurements made on three freshly prepared samples, with three readings per samples.

2.3.5 Emulsion stability to creaming

Test of emulsion stability was carried out by centrifugation (10min at 4000g) to speed up the
creaming process. As a result of different densities, centrifugation led to a concentrated emulsion
under close packing conditions. The thickness of the creaming layer was measured with a digital
caliper and the results were converted to volume [4]. Furthermore, different factors that may have
an impact on emulsions stability were also investigated. As previously mentioned, Canthaxanthin
was dissolved into the warm coconut oil at a ratio of 1:50. The ratio of oil/water was 20:80 in the
final emulsions. CNCs concentrations were 0.3, 0.6%, 0.9% w/w. These emulsions were then
examined in different pH values, temperatures and NaCl concentrations. In order to evaluate the
temperature stability, emulsions were diluted and then incubated in a water bath for 30min at
30,40,50,60 and 70C. The test of pH stability was carried out by adjusting the pH with 0.1M
NaOH and 0.1M HCl to the final pH of 4.5, 5.6, 6.3, 7.1, and 8. Finally the emulsions were diluted
using NaCl solution to the endmost concentration of 0-100mM, so that the ionic strength of the
emulsions could be investigated. After each treatment, the treated emulsions were centrifuged at
4000g for 10min, and the creaming layer was measured using a digital caliper and then converted
to volume.

2.3.6 Effect of light on Canthaxanthin stability

CTX stability to light was investigated in two CTX conditions: 1) CTX in coconut oil and 2) CTX
in pickering emulsions. Samples were incubated in a 30C under light treatment for 30 days. The
lamp intensity was 2000 lx. White cool and fluorescent lamps were used 20 cm away from the
samples. Changes in CTX content were measured using spectrophotometric method in day 1, 7,
14 and 30 using the following equation:

CTX%= ( ) 100. (2)

where initial and final CTX content were representing the average of three readings for each
sample.

3. Results and Discussion

3.1. Cellulose nano-crystals preparation and characterization

CNCs were obtained from cotton linter and bacterial cellulose with acid treatment. During acid
hydrolysis, hydrogen atoms from sulfuric acid protonate the glucosiding oxygen or cyclic oxygen
and in order to accelerate the splitting of glucosidic bonds, water was added [21]. In native
cellulose, amorphous regions are more attainable to acid and therefore can be hydrolyzed easier
than crystalline regions [22]. As the reaction time expands the hydrolysis of amorphous regions
intensify. The reaction condition leads to a negatively charged, homogenous and stable suspension.
The negative charge of CNC is due to sulfate ester groups that attach to glucose units during acid
hydrolysis. Dynamic light scattering showed the average size of CNCs was 112.6nm which means
the CNCs were produce after acid treatment. Fig. 1 illustrates the size distribution of the CNC
samples.

The atomic force microscopy (AFM) of CNCs is reported in Fig. 4. AFM images shows the rod-
like shape of CNCs, in addition, it was used for measuring the thickness of the CNC particles.
AFM showed that even after drying of CNCs for imaging, the shape of CNCs did not alter which
indicate their strong structure and elastic nature [23]. As it is shown in Fig. 4, CCNs represented
smaller nanocrystals compared to BCNs.

The X-ray diffraction profiles of cotton linter, Bacterial cellulose, CCN, and BCN are presented
in Fig. 2. They showed a sharp peak at 2=26.36 and three weak peaks at 2=17.29, 19.05,
24.32. The peaks observed at 2=17.29and 19.05 are corresponded to (101), (10) planes of
cellulose I; which is the most abundant form of cellulose consisting of repeating -(14)-D-
glucopyranose in the form of parallel glucan chains. Diagnostic peak at 2=24.32 is attributed to
(110) plane of cellulose II. As it was expected, the magnitude of these peaks increased as the
amorphous parts were removed after acid treatment. The sharpest peak, occurring at 2=26.36
represents the (002), lattice plane of cellulose I [24-27]. It can be observed that the peaks arise
from BC and BCN were approximately in the same angles as CL and CCN. Based on the XRD
findings and the results of Li, Wang, Liu, (2011) [28], it is suggested that the crystalline structure
of the cellulose was preserved during acid treatment. The crystallinity index (CrI) of cotton linter
and CCN were calculated by equation 1 and results were 75.4% and 86.6% respectively. Also CrI
for bacterial cellulose and BCN were 79.1% and 91.6% respectively.

Fig. 3 shows the FTIR-ATR spectra of cotton linter, CCN, bacterial cellulose, and BCN. The peak
at 3328-3400 is corresponded to alcoholic hydroxyl groups. The results revealed that after acid
treatment the amount of hydroxyl groups increased. ATR spectra of CCN displayed a signal of C-
H stretching at 2900 cm-1. The peak at 1426 cm-1 is attributed to H-C-H and O-C-H in-plane
bending vibration. The peak observed at 1051 cm-1 is associated with -C-O stretching. C-O-C
stretch of the ether linkage (1,4--D-glucoside) of cellulose. The wavenumber at 1159 cm-1
indicated the presence of cellulose crystals. The motion of C-5 and C-6 atoms led to deformation
modes and stretching vibration of C-O-C, C-C-O and C-C-H at 885 cm-1. Absorbance at 665cm-1
is refer to the C-O-H out of plane bending mode, which also proved the cellulosic structure of the
product [29, 30]. The signals at 1426, 1157, 1106 and 885 cm-1 showed that CNCs were primarily
in the form of cellulose I. The results exhibited that the cellulose structure did not change
drastically, and no mercerizing occurred [31, 32]

3.2. Emulsion preparation

The aqueous phase of all tested emulsions consisted of CNCs dispersed in distilled water at the
required concentration without additional treatment. CTX was dissolved into the warm coconut
oil and was added afterward to the watery phase. The sonication of biphasic system resulted in an
ultra-stable emulsion. Fig. 5 shows the optical microscopy images of the emulsion at 0.3%, 0.6%
and 0.9% of CCN and BCN concentrations. Emulsions with 0.3 % CNCs content created bigger
droplet compared to emulsions with 0.6% and 0.9% CNCs. One possible explanation for this
occurrence may be the fact that as the CNC content decrease, there will be less particles to cover
the droplets, so that some parts of the droplets may stay overt and as a result, the uncovered parts
of the droplets will unite and generate a bigger droplet. According data in Table 1 about size
distribution, emulsions with 0.6% CCNs had surface mean diameter of 574nm whilst those with
0.6% BCN had surface mean diameter of 1.2. In all the samples, as the CNC concentration
increased the droplet diameter of CTX pickering emulsions decreased significantly (p < 0.05, Fig.
4 and Table. 1). Emulsions with 0.6% CNCs concentration exhibited good stability and therefore
were used for further experiments.

3.3. Emulsion stability to creaming

Stable emulsion is considered an emulsion that can be resistant to physical changes over a period
of time. An emulsion is called stable as long as no coalescence occurs, meaning that the size and
size distribution do not change. One of the most common method to evaluate the emulsions
stability is centrifugation. Two factors were under investigation after each centrifugations, 1)
droplets size distribution 2) emulsions volume.

The effect of temperature on the properties of CNCs on the droplets interfaces was investigated at
various temperatures (30 to 70C). The Results indicated that for both CCN and BCN, as the
temperature increased, the emulsions stability against creaming improved. It can be explained by
the fact that heating caused some irreversible structural changes. The kinetic energy of the particles
increased so that their movement enhanced and as a result, they rearranged themselves on the
interface. Furthermore, the diffuse-double layer thickness and inner particle distances are reduced,
which led to increase in forces between CNCs [33, 34]. Consequently, the emulsion stability to
creaming is increased. As the concentration increased (from 0.3 to 0.9) the volume of emulsions
after centrifugation and temperature treatment, increased. Fig. 6 illustrates the emulsions after
centrifugation under different temperatures. Increase in temperature had no notable impact on the
droplets size. Comparatively, creaming stability of emulsions with BCN was lower than CCN
emulsions, which may be caused by poor droplets coverage.

Emulsions stabilization under different pH and ionic concentrations were tested by measuring the
zeta potential of the droplets. Zeta potential of emulsions (in 0.6% w/w CCNs and BCNs
concentration) at tested of NaCl concentrations, were following the same pattern as CCNs and
BCNs (Fig. 7). The results represented that as the salt concentration increased; since the repulsive
forces are weakened by the increase in electrolyte concentration; the tendency of CNC particles to
aggregate to each other enhanced. At the final NaCl concentration, the CNCs partially aggregated
and zeta potential of the dispersion fell into the low charged surface range of -25 mV and below.
Absolute value of zeta potential for CNC remains higher than emulsions in all samples. Low charge
of the droplets can be explained by of the relatively low surface load of CNCs. Another reason
might be due to the possible presence of ionic impurities in the samples and their adsorption to the
surface of the droplets, which could leads to a change in electrical charge [35]. as the ionic
concentration is elevated (from 0 to 100 mM) the zeta potential is reduced from -58 to -22 mV,
which may attributed to the electrostatic screening of the negatively charged -COO- groups by the
counter ions of Na+.

Fig. 7 shows the trends of zeta potential of 0.6% w/w CNCs stabilized emulsions and CNCs
dispersions in the different pH environments. Results indicated similarities in the zeta potential
trends of emulsions and CNCs dispersions, and both of the samples showed that as the pH elevated
the zeta potential increased. The values of zeta potential at all evaluated of pH range remained
negative, which is due to negatively charged groups of -C-O-SO3- on CNC particles. Furthermore,
CNCs did not have any cationic groups, so they did not charged positive at any pH level. Following
an increase in pH, the magnitude of the negative charge on the droplets increased from -30.1 to -
50.7mV, this occurrence may arise from deprotonation of some of the protonated carboxyl groups
(COOH COO- +H) [36]. It is evident that an increase in surface charge leads to an improvement
in emulsions stability. This trend is due to the increase in repulsive forces between the droplets
that reduce the risk of flocculation and coalescence. An increase in pH value was accompanied by
a decrease in charge from -29.6 to -59.2 mV. Consequently, the stability of final emulsions was
improved [37]. In all the samples, CCN emulsions showed better stabilization properties than that
observed for BCN emulsions, which is because of smaller particle size. The size of colloidal
particles is one of the most important factors affecting emulsions stability by controlling the
ability of the particles to reside at the interface [38]. Another reason for poor stability of BCN
emulsions might be the low droplet coverage. Kalashnikova, et. al., 2013 [1] indicated that, the
coverage ratio of the BCN was 44% when CNC coverage was above 90%. BCN will produce a
porous net around the droplet and cause possible leakage of the oil from the inside. Additionally,
since the particle size is larger they produce bridges between droplets and bridging can cause
flocculation of the droplets which later will lead to instability [4]. Fig. 8 represent a schematic
CCN and BCN particles placement on the droplets surface.

The droplet size of the emulsions after ionic treatment started to increase in a period of 72 h (data
not presented). The reason for this event can be expressed as the reduction of surface charges and
repulsive forces after treatment, which increased the risk of flocculation and coalescence. The
effect of pH on droplet size were not noticeable.

3.4 Effect of light on Canthaxanthin stability

Fig. 9 shows the light stability of CTX in oil and encapsulated CTX (CNC pickering emulsion) at
30C in light and dark conditions over a month. Results were represented as a percentage, assuming
that at zero time the CTX concentration was 100% in both blank and encapsulated samples.
According to Fig. 9, CTX was consumed much more in the light than in the dark. The light had a
significant effect (p < 0.05) on CTX content at 30C which is in agreement with Palozza,
Muzzalupo, Trombino, Valdannini, and Picci (2006) [39] who worked on the - carotenes
stabilized in niosomes at room temperature in light. They declared that light consumes more
carotenoids due to the presence of double bonds in their chains which are sensitive to
environmental factors such as light. However, in our work it was found that the use of CCN as a
stabilizer increased the stability of the CTX to light more than BCN. The mechanism responsible
for this occurrence may be the fact that CCN particles make a compact shell-like barrier around
CTX droplets, which partly preserves the CTX from light. However, BCN showed less stability,
which may be related to its poor coverage.

4. Conclusion

Cellulose nanocrystals from different origins were produced by sulfuric acid hydrolysis. Stable
dispersion of nano-particles with a charged surface due to carboxyl and sulfate ester groups were
generated. This process produced fine particles that were likely to stabilize an oil in water pickering
emulsion irreversibly. The degree of crystallinity of cellulose increased after acid hydrolysis which
was caused by degradation of the amorphous regions of cellulose. Canthaxanthin pickering
emulsions were prepared by varying amounts of nanocrystals introduced at high sonication power
for 21s. The creaming stability of final emulsions and size distribution was satisfactory.
Additionally, results indicated that the stability of the emulsions at low pH or high ionic
concentration decreased because of the electrostatic screening effect. Furthermore, the emulsion
stability increased as the temperature increased, which was probably because of the reduction in
diffuse-double layer thickness which led to elevation of the forces between CNCs. However, CCN
showed better stability over BCN emulsions. Large particle size, poor coverage and bridging effect
were responsible for this phenomenon. Therefore, CNCs can be introduce as an useful stabilizer
for canthaxanthin and considering its biodegradability and no use of synthetic emulsifier in the
formulation, it can be a promising stabilizer for exquisite industries that deal with peoples health
such as food industries, pharmacology and cosmetics.

Acknowledgment

This material is based upon work supported financially by the University of Tehran and Iranian
center of excellence for application of modern technologies for producing functional foods and
drinks and Iran Nanotechnology Initiative Council (INIC).
References

[1] I. Kalashnikova, H. Bizot, P. Bertoncini, B. Cathala, & I. Capron. Soft Matter 93 (2013) 952959.

[2] S. U. Pickering. J. Chem. Soc. Trans. 90 (1907) 307-314.

[3] W. Ramsden, Proc. R. Soc. Lond. 72 (1903) 156-164.

[4] I. Kalashnikova, H. Bizot, B. Cathala, I. Capron, Biomacromolcules, 13 (2011) 267275.

[5] C. Salas, T. Nypel, C. Rodriguez-Abreu, C. Carrillo, O. J. Rojas, Curr. Opin. Colloid Interface Sci.
195(2014) 383396.

[6] B.G. Rnby, Discuss. Faraday Soc., 11 (1951) 158-164.

[7] M.M. De Souza Lima, R. Borsali., Macromol. Rapid Commun. , 257 (2004) 771787.

[8] S.B. Candanedo, M. Roman, D.G. Gray, Biomacromolecules, 6 (2) (2005) 10481054

[9] I. Kalashnikova, H. Bizot, B. Cathala, I. Capron, Langmuir, 27(12) (2011) 74717479.

[10] M.A.S.A. Samir, F. Alloin, A. Dufresne. Biomacromolecules, 6 (2) (2005) 612626.

[11] G.K. Chandi, B.S. Gill, Int. J. Food. Prop. 14 (2011) 503513.

[12] S.M.T Gharibzahedi, S.H. Razavi, S.M. Mousavi. Carbohyd. Polym. 96 (2013a), 2130

[13] S.M.T Gharibzahedi, S.H. Razavi, S.M. Mousavi. Carbohyd. Polym. 101 (2014) 11471153.

[14] C. Schmidt-Dannert, Curr. Opin. Biotechnol. 11 (2000) 255261.

[15] S.M.T Gharibzahedi, S.H. Razavi, S.M. Mousavi. Eng. Life Sci. 13 (2013b) 408417

[16] J.A. del Campo, M. Garcia-Gonzalez, M.G. Guerrero, Appl. Microbiol. Biotechnol. , 74 (2007) 1163
1174.

[17]M. Hojjati, S.H. Razavi, K. Rezaei, K. Gilani, Food Sci. Biotechnol. 20 (2011) 6369

[18] F. Khodaiyan, S.H. Razavi, S.M. Mousavi, Biochem. Eng. J., 40 (2008) 415422

[19]J. Revol, H. Bradford, J. Giasson, R.H. Marchessault, D.G. Gray. Int. J. Biol. Macromolec. (1992)13.

[20]L.G.J.M.A. Segal, J.J. Creely, A.E. Martin, C.M. Conrad. Text. Res. J., 29(10) (1959) 786794.

[21] P. Lu, Y. Hsieh, Carbohyd. Polym. 82 (2010) 329336.

[22]D.N.S. Hon, N. Shiraishi,. Wood and cellulosic chemistry, revised, and expanded. CRC Press (2000)

[23] S.J. Eichhorn, et al. J. Mater. Sci. 45 (2010) 1-33.

[24] I. Besbes, S.Alila, S. Boufi. Carbohyd. Polym. 84 (2011) 975983.

[25] S.Park, J.O. Baker, M.G. Himmel, P.A. Parilla, D.K. Johnson., Biotechnol Biofuels 3 (2010) 1-10.

[26] F. Jiang, A.R. Esker, M. Roman,. Langmuir, 26(23) (2010) 1791917925

[27] Q. Li, , S. Renneckar,. Biomacromolecules, 12(3) (2011) 650659

[28] W. Li, R. Wang, S. Liu. BioResources. 6(4) (2011) 42714281.

[29] C.H. Chan, S. Zakaria, I. Ahmad, A. Dufresne. BioResources , (2013) 785794 .

[30] S.K. Bajpai, V.Pathak, N. Chand, B. Soni,.J. Macromol. Sci. A. (2013)

[31] F. Jiang, S. Han, Y.L. Hsieh. RSC Advances. 3 (2013), 12366-12375

[32] A.C. Leung, S. Hrapovic, E. Lam, Y.Liu, K.B. Male, K.A. Mahmoud, et al. Small, 7(3) (2011) 302
305

[33] M.V. Tzoumaki, T. Moschakis , C.G. Biliaderis.. Biomacromolecules, 11 (1)(2010) 175181

[34] C. Wen, Q.Yuan, H. Liang, F. Vriesekoop. Carbohyd. Polym. 112 (2014) 695700 .
[35] J. Surh, E.A. Decker, D.J. Mcclements. Properties and stability of oil-in-water emulsions stabilized by
fish gelatin. Food Hydrocoll. 20 (2006) 596606.

[36] T. Winuprasith, M. Suphantharika. Food Hydrocoll. 32(2) (2013) 383394.

[37] E. Dickinson. Food Hydrocoll. 23(6) (2009) 14731482.

[38] D.E. Tambe, M.M. Sharma. Adv. Colloid Interface Sci. 52 (1994) 1-63.

[39] P. Palozza, R. Muzzalupo, S.Trombino, A. Valdannini, N. Picci, (2006). Chem. Phys. Lipids. 139
(2006) 3242
Figure captions

Figure 1. Size distribution of CCN and BCN by volume and their suspensions (0.9 wt%) after
sulfuric acid treatment.

Figure 2. X-ray diffractorgrams of cotton linter (CL) cellulose nanocrystals (CCN), bacterial
cellulose (BC) and bacterial cellulose nanocrystal (BCN).

Figure 3. FTIR-ATR spectra of cotton linter (CL) cellulose nanocrystals (CCN), bacterial cellulose
(BC) and bacterial cellulose nanocrystal (BCN).

Figure 4. AFM images of A: bacterial cellulose nanocrystals, B: cellulose nanocrystals from cotton
linter (height image of surface topography).

Figure 5. Droplet size dependence on CNCs concentration with 20/80 oil/water ratio. A: 0.9%
CCN, B: 0.6% CCN, C: 0.3% CCN, D: 0.9% BCN, E: 0.6% BCN, F: 0.3% BCN

Figure 6. Effect of temperature on the emulsions stability with (A) 0.3% w/w CNCs, (B) 0.6%
w/w CNCs and (C) 0.9% w/w CNCs, (D) 0.3% w/w BCN, (E) 0.6% w/w BCN and (F) 0.9% w/w
BCN.

Figure 7. (A) Effect of ionic concentration (NaCl) and (B) pH on CNCs suspensions and CTX
pickering emulsions zeta potential.

Figure 8. Schematic presentation of CCN and BCN particles placement on the emulsion droplets.

Figure 9. CNC content in emulsion or in the oil evaluated for various periods of time at
temperature of 30C under light treatment, or in the dark. The values were the mean amount of
three different experiments. The treatment/time interaction was significant (p < 0.05). Values not
sharing the same superscript were significantly different (A: p < 0.01, B: p < 0.01) (tukeys test).

Below: image taken from emulsions before and after 30 days of light treatment A: new emulsion
at day 0, B: CCN emulsion C: BCN emulsion.
 25.00

20.00
CNC
Channel%

15.00

10.00

5.00

0.00
0 200 400 600 800 1000
Size (d.nm)

Fig. 1.
Fig. 2.
Fig. 3.
 A B

Fig. 4.
Fig. 5.
Fig. 6.
Fig. 7.
Fig. 8
Fig. 9.
Table caption

Table 1. Average diameters of CTX pickering emulsions at different CCN and BCN
concentrations. Results showed the mean amount of triplicate exams.

CNC concentration (% w/w) BCN-emulsion Droplet CCN-emulsion Droplet

surface mean diameter () surface mean diameter ()
0.3 3.10.4a 0.5040.012a
0.6 2.30.3a 0.4310.019a
0.9 1.10.3b 0.3500.011a
Means with different subscripts differ significantly (p < 0.05).