JOURNAL OF

POLYMER SCIENCE WWW.POLYMERPHYSICS.ORG REVIEW

Cellulose Nanocrystals and Related Nanocomposites: Review of some

Properties and Challenges

Marcos Mariano,1,2 Nadia El Kissi,2 Alain Dufresne1

1
Grenoble Institute of Technology (Grenoble INP)—The International School of Paper, Print Media and Biomaterials (Pagora),
CS10065, 38402 Saint Martin d’Hères Cedex, France
2
Laboratoire Rheologie et Proce
de
s, Grenoble INP-CNRS-UJF, UMR 5520, BP 53, 38041 Grenoble Cedex 9, France
Correspondence to: A. Dufresne (E - mail: [email protected])

Received 25 February 2014; accepted 2 April 2014; published online 19 April 2014
DOI: 10.1002/polb.23490

ABSTRACT: Cellulosic nanoparticles with high Young’s modu- chemically assisted deconstructing strategy, and owing to their
lus, crystallinity, specific surface area, and aspect ratio can be highly reactive surface ensuing nanomaterials can be chemically
found in the natural structure of plant fibers. Indeed, lignocel- modified to tailor their properties for a wide range of applica-
lulosic fibers consist of semicrystalline cellulose nanofibrils tions. This review is limited to cellulose chemically extracted
embedded in an amorphous matrix mainly composed of lignin nanocrystals and aims to provide an overview about several
and hemicelluloses. These nanostructures give the mechanical aspects that involve this material, including sources, properties,
strength to higher plant cells, and are biodegradable, renew- challenges, and perspectives. V C 2014 Wiley Periodicals, Inc.
able, resistant, and widely available to produce nanocompo- J. Polym. Sci., Part B: Polym. Phys. 2014, 52, 791–806
sites with low density, and improved and controlled
mechanical, optical, and barrier properties. Nanoparticles can KEYWORDS: barrier; biofibers; cellulose; mechanical properties;
be extracted from cellulose using a top-down mechanically or nanocomposites; nanoparticles; polysaccharides

INTRODUCTION The prefix nano has become common jargon increase. Consequently, the nanomaterial has a strong tend-
in modern society; this term crossed the lines of the scien- ency for agglomeration. This issue is particularly significant
tific community and is now present in markets, drugstores, for cellulose, which displays a highly reactive surface with
computers, smartphones, and so forth. In 2009, the world- plenty of hydroxyl groups. Figure 1 shows the evolution of
wide market for products incorporating nanotechnology the specific surface area for rod-like particles with a density
reached about USD 254 billion and this number can double of 1.5 g cm23 (density of crystalline cellulose) as a function
each 3 years until 2020, when this value could be around of their diameter.5 It can be seen that Asp is increasing very
USD 3 trillion.1 The nano scale refers to values of the order sharply when the diameter drops down to about 20 nm.
of 1029 meters. By definition materials having at least one This diameter range corresponds to the size of cellulose
linear dimension in the range from approximately 1 nm to nanocrystals.
100 nm are called nanomaterials. This definition makes a
differentiation with nanoparticles, which are nano-objects A nanocomposite is a multiphase solid material where one
with all three external dimensions in the nanoscale.2 A of the phases has one, two, or three dimensions in the nano-
review of recent trends in the dispersion, purification, and meter scale, that is of less than 100 nm. Nanocomposites
assembly of colloidal nanoparticles highlighted a number of show unique properties, because of the nanometric size
growing analogies with ideas borrowed from polymer effect, compared to conventional composite even at low filler
science.3 content. Indeed, nanofillers have strong reinforcing effects
and studies have also shown their positive impact in barrier
As the size of a particle is decreasing down to the nanometer properties. However, for decades studies have been con-
scale important changes occur. For example, a spherical par- ducted with non-renewable inorganic fillers. Increasing envi-
ticle 30 nm in diameter has about 5% of its atoms on its ronmental concerns have led to investigating the potential
surface; with 10 nm this number changes to almost 15%; uses of renewable resources for such application.
and a very small particle 3 nm in diameter can have about
50% of its atoms on the surface.4 As a consequence, both Since the first investigation on cellulose-based nanocompo-
the specific surface area (Asp) and total surface energy sites and the reinforcing effect of cellulose nanocrystals,6

C 2014 Wiley Periodicals, Inc.

V

WWW.MATERIALSVIEWS.COM JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806 791
 JOURNAL OF
REVIEW WWW.POLYMERPHYSICS.ORG POLYMER SCIENCE

Marcos Mariano received the bachelor degree in chemistry and master degree in physical-
chemistry from Federal University of Santa Catarina. Currently he is a PhD student at Gre-
noble Institute of Technology (INPG) and his research focuses on cellulose nanomaterials
and nanocomposites.

Nadia El Kissi received her PhD in Mechanical Engineering, in 1989 from Grenoble Insti-
tute of Technology, France. She is senior scientist at the Centre National de la Recherche
Scientifique (CNRS). Her research expertise concerns rheology and process engineering of
complex model, industrial, and biological fluids. All scales, from the nano to the macro-
scopic scale, and all fields, from the elaboration to the processing, are considered.

Alain Dufresne has a background in solid-state physics. He received his PhD in 1991 from
the Department of Electronic at the Toulouse National Institute of Applied Science. He is
professor at the International School of Paper, Print Media, and Biomaterials (Pagora) in
Grenoble Institute of Technology. His main research interests concern the processing and
characterization of polymer nanocomposites reinforced with nanoparticles extracted from
renewable resources.

almost 20 years ago, the number of publications has grown and 2004, respectively.7 This shows the high interest in this
considerably (as nanotechnology) due to both increasing area, and there is a strong tendency for a bigger expansion
environmental concerns and the huge range of potential since a number of manufacturing facilities have been or are
applications. The number of scientific articles involving key- being built that will increase production to upward of multi-
words like “nanocellulose” (including mechanically extracted ple tons per day. Production has increased by 1000% in two
cellulose nanoparticles or nanofibrillated cellulose—NFC) or years and will likely increase by a further 500% at least by
“cellulose nanocrystals” or “cellulose (nano)whiskers” has 2017, leading to an increase in perspective production in the
expanded rapidly as shown in Figure 2. It was reported that range of 1000% in the next two years.8
the number of scientific publications and patents related to
nanocellulose have begun to increase markedly after 2003

FIGURE 1 Evolution of the specific surface area of rod-like FIGURE 2 Number of publications related with nanocellulose
nanoparticles as a function of their diameter, assuming a den- terms during the last decade (Data from main scientific
sity of 1.5 g cm23. (Reproduced from ref. 5). databases).

792 JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806
 JOURNAL OF
POLYMER SCIENCE WWW.POLYMERPHYSICS.ORG REVIEW

FIGURE 3 Scanning electron micrograph of a sisal fiber and its organization.

HIERARCHICAL STRUCTURE OF PLANT FIBERS fibers using chemical treatments consisting of alkali extrac-
Plant (or lignocellulosic) fibers are cellular, complex hier- tion and bleaching. The alkali extraction treatment (usually
archical biocomposites designed by nature. They are basi- 2% NaOH at 80  C) allows removal of soluble polysaccha-
cally semicrystalline cellulose microfibril-reinforced rides. The subsequent bleaching treatment removes most of
amorphous matrices made of hemicellulose, lignin, waxes, the residual phenolic molecules like lignin or polyphenols.
extractive, and trace elements (Fig. 3). They consist of bun- Cellulose is the most important polymer on Earth. It is com-
dles of elongated honeycomb cells cemented together by an posed of D-glucopyranose (C6H11O5) units linked through
intracellular ligneous material. In each cell, the inner lumen b(1,4) links, forming cellobiose, the repeating unit of cellu-
is surrounded by multiple concentric layers in one primary lose (Fig. 4). Even if it is difficult to determine experimen-
and three secondary cell walls sticked together in a multiple- tally the degree of polymerization of native cellulose, it is
layer-composite, in which cellulose microfibrils confer the reported17 to be close to 10,000, corresponding to a molecu-
mechanical properties to the plant cells. This dynamic hier- lar weight around 3.2 3 106 g mol21. The biosynthesis of
archical structure leads to the unique strength and high celluloses occurs by a process where the monomeric unit
mechanical properties of different species of plants, in partic- and the microfibril construction are closely related.
ular their unusual ability to provide high mechanical
strength and high strength-to-weight ratio while allowing for The presence of many hydroxyl groups along the cellulose
flexibility to counter large dimensional changes due to swel- chain results in the formation of a network of intra- and
ling and shrinking. Such microfibrils typically have a diame- intermolecular bonds.18 In addition, a network of Van der
ter of about 2–20 nm and are made up of 30–100 cellulose Waals connections is established between the chain layers.
molecules in extended chain conformation. These linked networks, and the stiffness of the glycosidic
bond, allow the construction of ordered crystalline structures
The structure and chemical composition of the fiber can vary 2–20 nm in width and up to a few microns for the length.
drastically depending on the plant species, age of the plant, These crystalline domains are interspersed with amorphous
part of the plant it was extracted, growth area, and climate.9 regions in the fibril structure (Fig. 5). For some plant spe-
It causes a great variation in the characteristics of the fibers, cies, these amorphous regions can represent up to 50% of
and brings difficulties in establishing a standard of quality.10 the structure, while in bacterial cellulose and cellulose
extracted from some algae the crystalline domains corre-
CELLULOSE FIBERS sponds to almost 100% of the fibril. These crystalline
domains are almost defect-free.
Many authors suggest methodologies to isolate cellulose
from their natural matrix with the removal of lignin, hemi-
celluloses, and resins. It generally involves basic and/or oxi-
dative treatments11–13 that can affect the morphology of
cellulose, or acidic treatment with peracids.14–16 This step is
very important because it can change the crystalline organi-
zation and polymorphism of cellulose, as well as chemical
and thermal stability of the material. Classically, the matrix
material can be removed to extract cellulose from plant FIGURE 4 Chemical structure of cellulose.

WWW.MATERIALSVIEWS.COM JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806 793
 JOURNAL OF
REVIEW WWW.POLYMERPHYSICS.ORG POLYMER SCIENCE

biomimics nature. The goal of scientists is to mimic nature,

to exploit natural biocomposites, in order to develop novel
materials that can be suitable to our needs without being
harmful to the environment.

CELLULOSE NANOCRYSTALS

Extraction of Cellulose Nanocrystals

The extraction of cellulose nanocrystals from cellulosic fibers
usually consists of an acid-induced destructuring process,

FIGURE 5 Schematic organization of crystalline and amor- TABLE 1 Longitudinal Modulus (EL) of Crystalline Cellulose
phous domains in cellulose fiber.
Method Material EL (GPa) Ref.

POTENTIAL REINFORCEMENT OF CELLULOSE Calculation Cellulose I 77–121 19

56 20
The reinforcing ability of cellulose is primarily ascribed to its
semicrystalline nature and the extended chain conformation 172.9a 21
b
of cellulose molecules in the crystalline regions. This particu- 70.8 22
lar conformation results from the strong intermolecular 136 6 6 23
hydrogen bonding system arising from the high density of 167.5 24
hydroxyl groups borne by the cellulose molecule. The elastic 134–135 25
properties of the cellulose crystal have been investigated
Cellulose Ia 127.8 26
since the mid-1930s either by theoretical evaluations or by
136–155a 27
experimental measurements (X-ray diffraction, Raman spec-
troscopy, and atomic force microscopy (AFM)). Some values 114–117b
of the longitudinal modulus of crystalline cellulose reported Cellulose Ib 115.2 26
in the literature are collected in Table 1. Few studies 124–155 28
reported the transverse modulus that was obviously found 116–149a 27
to be much lower than the longitudinal modulus. As shown 124–127b
in Table 1 a broad range of values have been reported that 156 29
obviously depend on the material under study. a
85.2–126.0 / 30
For cellulose I (native cellulose), which is a mixture of two pol- 37.6–63.3b
ymorphs, cellulose Ia and cellulose Ib, the average value is 139.5 31
around 137 GPa when considering intramolecular hydrogen 206 32
bonding, and 92 GPa without intramolecular hydrogen bonding, Cellulose II 89 6 4 23
evidence of the important role of intramolecular bonding on 162.1 24
the determination of the crystallite modulus and chain defor-
83 25
mation mechanism. For cellulose II, the average value is around
155 33
113 GPa. The lower value observed for form II is ascribed to
lower intramolecular hydrogen bonds, showing again the 109–166a 27
b
importance of this parameter, whereas the intermolecular 101–106
hydrogen bonds were found to play a minor role. X-ray diffraction Cellulose I 134 34
122–135 35
These impressive mechanical properties make cellulose crys-
138 36
tal an ideal candidate for the processing of reinforced poly-
Cellulose II 70–90 37
mer composites. Its Young’s modulus, with a density for
crystalline cellulose around 1.5–1.6 g cm23, is much higher 106–112 35
than for glass fibers, around 70 GPa with a density around 88 36
2.6 g cm23, which are classically used in composite applica- Raman Cellulose I 57–105 38
tions. It is similar to Kevlar (60–125 GPa, density around Cellulose Ib 143 39
1.45 g cm23) and potentially stronger than steel (200–220 Inelastic X-ray Cellulose I 220 40
GPa, density around 8 g cm23). Indeed, the specific Young’s scattering
modulus, which is the ratio between the Young’s modulus AFM Cellulose Ib 145–150 41
and the density of cellulose crystal is around 85 J g21
whereas it is around 25 J g21 for steel.42 Using these nano- a
With, and
b
particles in a synthetic or natural polymeric matrix therefore without intramolecular hydrogen bonding.

794 JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806
 JOURNAL OF
POLYMER SCIENCE WWW.POLYMERPHYSICS.ORG REVIEW

trostatic layer covering the nanocrystals and promotes their

dispersion in water. The high stability of sulfuric acid-
hydrolyzed cellulose nanocrystals results, therefore, from an
electrostatic repulsion between individual nanoparticles.
However, it compromises the thermal stability of the nano-
particles.45 To increase the thermal stability of H2SO4-pre-
pared nanocrystals, neutralization of the nanoparticles by
sodium hydroxide (NaOH) can be carried out.46

The nomenclature used to designate these acid-hydrolyzed

crystalline cellulose nanoparticles has been unclear for a
long time and different descriptors including cellulose
whiskers or nanowhiskers, nanocrystalline cellulose, or even
cellulose microcrystals have been used in the literature.
However, recently TAPPI (Technical Association of the Pulp
and Paper Industry) has proposed to standardize the termi-
FIGURE 6 Standard terms for cellulose nanomaterials (TAPPI
nology (Standard Terms and Their Definition for Cellulose
WI3021).
Nanomaterial WI 3021). Nomenclature, abbreviation, and
dimensions applicable to each sub-group are reported in Fig-
during which the heterogeneous acid hydrolysis involves the ure 6. Cellulose nanocrystal was the TAPPI standard
diffusion of acid molecules into cellulose fibers, followed by recommendation.
cleavage of glycosidic bonds. This acid hydrolysis step is fol-
lowed by centrifugation, dialysis, and ultrasonication. Differ- Morphology of Cellulose Nanocrystals
ent strong acids have been shown to successfully degrade The ensuing nanoparticles occur as high aspect ratio rod-like
non-crystalline (amorphous) regions of cellulose fibers to nanocrystals. Their geometrical dimensions depend on the
release crystalline cellulosic nanoparticles, such as sulfuric, origin of the cellulosic substrate and hydrolysis conditions.
hydrochloric, phosphoric, hydrobromic, nitric acids, and a For example, the effect of acid hydrolysis conditions on the
mixture composed of hydrochloric and organic acids.43 dimensions and crystallinity of cellulose nanocrystals iso-
Nevertheless, the hydrolysis treatment with sulfuric acid to lated from sisal fibers was investigated.47 This study showed
prepare cellulose nanocrystals has been extensively investi- the influence of temperature (45 and 60  C) and duration of
gated and appears to be the most effective method. One of hydrolysis (30, 60, and 75 min). Higher crystallinity was
the main reasons for using sulfuric acid as a hydrolyzing reported for materials hydrolyzed at the higher temperature
agent is that, if nanocrystals are prepared using hydrochloric for lower times. Figure 7 shows cellulose nanocrystals
acid, their ability to disperse in solvents is limited and the obtained from two different cellulosic sources. Each rod can
suspension is unstable, tending to flocculate. Indeed, during be considered as a cellulose crystal with no apparent defect.
hydrolysis, sulfuric acid reacts with the surface hydroxyl They usually present a relatively broad distribution in length
groups via an esterification process allowing the grafting of because of the diffusion-controlled nature of the acid hydro-
anionic sulfate ester groups (AOSO32).44 These sulfate lysis. The average length is generally of the order of a few
groups are randomly distributed on the surface of the cellu- hundred nanometers and the width is of the order of a few
losic nanoparticle. The presence of these negatively charged nanometers.5 An important parameter for CNCs is the aspect
sulfate ester groups induces the formation of a negative elec- ratio, which is defined as the ratio of the length to the width.

FIGURE 7 Transmission electron micrographs from a dilute suspension of (a) tunicin, and (b) ramie cellulose nanocrystals. (Repro-
duced from ref. 49, with permission from [American Chemical Society], and ref. 46, with permission from [Royal Society of
Chemistry]).

WWW.MATERIALSVIEWS.COM JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806 795
 JOURNAL OF
REVIEW WWW.POLYMERPHYSICS.ORG POLYMER SCIENCE

It varies roughly between 10 for cotton48 and 67 for tuni- exchange procedure is another option to suspend the nano-
cin49 or capim dourado (golden grass).50 particles in the proper liquid medium for further surface
chemical modification, or mixing with a polymer solution or
PROCESSING OF NANOCOMPOSITES monomer for subsequent in situ polymerization.57 With this
method the aqueous suspension is progressively solvent
Among other possible applications, cellulose nanocrystals
exchanged in liquids of decreasing polarity by several succes-
have been extensively used as “nano-additive” for polymer
sive centrifugation and redispersion operations, using sonica-
nanocomposites to improve properties such as mechanical or
tion after each solvent exchange step to avoid aggregation.
barrier properties, or swelling behavior. The introduction of
Acetone is routinely used for the first solvent exchange step.
cellulose wood fibers in nanocomposite materials has been
identified as one of the four biggest discoveries since 2000 Stable cellulose nanocrystal dispersions in apolar or low
in the Nanotechnology Research Directions for Societal Needs polarity solvent can be obtained by physically coating their
in 2020 with the mass use in nanotechnology of renewable surface with a surfactant,58,59 or covalently grafting apolar
and earth-abundant raw materials as the “Holy Grail” to moieties onto the surface. Both methods allow tuning the
attain and barriers to overcome by 2020.49 surface and decreasing the surface energy of the nanopar-
ticle to fit with the one of the liquids. Surface modification is
However, the homogeneous dispersion of cellulose nanocrys-
a way to overcome limitations of cellulose nanocrystals like
tals within a continuous polymeric matrix is the key step to
poor interaction with hydrophobic matrix, improve the ther-
benefit from their outstanding properties. It is fixed during
mal resistance, and create more stable suspensions in apolar
the processing of the material. These nanoparticles have a
liquids. The surface chemical modification of cellulose nano-
strong tendency for self-association because of the omnipre-
crystals obviously involves the ample surface hydroxyl
sence of interacting surface hydroxyl groups. This property,
groups resulting from their nanoscale dimensions and ensu-
which is the basis of the strength of paper sheets, is a desir-
ing high surface area (see Fig. 1). However, experimental
able feature for the formation of load-bearing percolating
chemical grafting conditions should avoid swelling media
architectures within the host polymer matrix. However, these
and peeling effect of surface-grafted chains otherwise induc-
inter-particle interactions can cause aggregation during the
ing their dissolution in the reaction medium. Therefore, the
preparation of the nanocomposite, thus inducing the loss of
chemical grafting process has to be mild enough in order to
the nanoscale, and limiting the potential of mechanical rein-
preserve the integrity of the nanoparticle. Some examples of
forcement. This aggregation phenomenon is magnified when
surface chemical modification of cellulose nanocrystals and
the specific surface area increases and then when the size of
the objective of this modification are reported in Table 2.
the particle decreases. Different strategies, briefly described
below, have been reported in the literature to homogene- Long chain surface chemical modification of cellulose nano-
ously mix cellulose nanoparticles with a polymeric matrix. crystals can also be performed and examples are given in
More detailed information can be found elsewhere.5,52 Table 3. This strategy consists in grafting polymeric chains
bearing a reactive end group and a long compatibilizing tail.
Processing from Aqueous Medium It not only decreases the surface energy of cellulose nano-
Cellulose nanocrystals form stable colloidal dispersions in crystals but the grafted chains may also interdiffuse upon
water. It is therefore the most suitable processing medium. heating to form the polymer matrix phase. Moreover, possi-
Both water-soluble polymers and polymer aqueous disper- ble chain entanglements as well as co-crystallization between
sions (latex) can be used. After mixing the nanoparticle dis- covalently linked chains at the surface of the nanoparticle
persion with the polymer solution/dispersion, a solid with those from the matrix should further improve the prop-
nanocomposite film can be obtained by simple casting and erties of the final composite.
water evaporation. This mode of processing allows preserv-
ing the individual state of the nanoparticles in water and has The chemically modified nanoparticles can be dispersed in
been therefore extensively used in the literature. A copoly- organic liquids of low polarity and mixed with polymer solu-
mer of styrene and butyl acrylate (poly(S-co-BuA)) in latex tion or eventually directly added in the polymer melt after
form and cellulose nanocrystals extracted from tunicin were drying. However, two conflicting effects arise from this proce-
used in the pioneering work.6,53 dure. On the one hand, it allows improving the dispersion of
the modified nanoparticles in the continuous apolar medium
Processing from Non-Aqueous Liquid Medium which is beneficial to optimize the mechanical properties of
Other polar liquid media can also be used for dispersing cel- the ensuing nanocomposite. On the other hand, it restricts the
lulose nanocrystals and processing medium, such as dime- interactions between nanoparticles through hydrogen-bonding
thylformamide (DMF),54 dimethyl sulfoxide (DMSO), N- which is the basis of the outstanding mechanical properties of
methyl pyrrolidine (NMP), formic acid, and m-cresol.55 Cast- polysaccharide nanocrystal-based nanocomposites.
ing from a mixture of solvents can also be used to prepare
cellulose nanocrystal-reinforced nanocomposites. By this Melt Processing
method, the aqueous dispersion of nanoparticles is mixed Most processing techniques used to prepare polymer nano-
with a polymer solution involving a solvent miscible with composites reinforced with cellulose nanocrystals are mainly
water, for example tetrahydrofuran (THF).56 Solvent restricted to wet processing methods such as casting/

796 JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806
 JOURNAL OF
POLYMER SCIENCE WWW.POLYMERPHYSICS.ORG REVIEW

TABLE 2 Common Surface Chemical Modifications of Cellulose Nanocrystals

Grafting Agent Objective Ref.

Esterification, Acetic anhydride and acetic acid Ultrastructural aspect of acetylation 60

acetylation
and acylation
Iso-octadecenyl succinic anhydride and Dispersion in solvents of low polarity 61
n-tetradecenyl succinic anhydride
Acetic anhydride Redispersion in water 62
Organic acid chloride aliphatic chain Dispersion in PE 10
Palmitoyl acid Hydrophobization 63
Acetic and butyric acids Dispersion in ethyl acetate and toluene 64,65
Vinyl acetate Dispersion in THF 66
Acetic anhydride Dispersion in chloroform 67
Succinic anhydride Adsorption of heavy metal ions 68
Alkenyl succinic anhydride Dispersion in PP and PLA 69
Silylation Hexamethyldisilazane Dispersion in acetone and acetate butyrate 70,71
Isopropyl, n-butyl, n-octyl, and Dispersion in THF 72
n-dodecyldimethylchlorosilane
n-dodecyldimethyl-chlorosilane Dispersion in THF and chloroform, 73
compatibilization with PLA
Aminopropyltriethoxysilane, Dispersion in PLA 74
n-propyltrimethoxysilane,
methacryloxypropyltrimethoxysilane,
acryloxypropyltrimethoxysilane
Isocyanate propyl triethoxysilane Hydrophobization 75
Carbamination n-Octadecylisocyanate Dispersion in PCL 57,76–78

evaporation, which has been extensively reported in the lit- crystals hinders their self-aggregation during melt process-
erature. The main advantage of this strategy relies in the fact ing. For example, organic acid chlorides-,10 poly(e-
that it preserves the dispersion state of the nanoparticles in caprolactone) (PCL)-,46 and polylactic acid (PLA)-grafted
the liquid. However, it limits the number of polymer matrices cellulose nanocrystals89 have been successfully extruded
that can be used in association with cellulose nanocrystals. with low density polyethylene (LDPE), PCL, and PLA,
Moreover, this procedure is both non-industrial and non- respectively. Homogeneous dispersion of the nanoparticles
economic. It should be used for niche applications. A number in the polymeric matrix was obtained. In the former work,
of organizations have announced nanocellulose demonstra- the surface of cellulose nanocrystals was modified with
tion plants.5 Therefore, more industrial nanocomposite proc- organic acid chloride aliphatic chains of different lengths,
essing techniques should be developed. Melt-compounding namely hexanoyl chloride, lauroyl chloride, and stearoyl
methods, such as extrusion or injection molding, are com- chloride, and it was found that the longer acid chains gave
monly used to process thermoplastic polymers. They are the best results. The extrusion temperature was 160  C and
“green” (solvent-free), and industrially and economically via- no sign of carbonization, a typical indication of cellulose
ble. However, these conventional processing techniques are thermal degradation, was observed (Fig. 8). However, this
infrequently employed for the preparation of cellulose nano- strategy involves an additional functionalization step which
crystal reinforced polymer nanocomposites. This is ascribed is time-consuming and can be hardly compatible with
to inherent incompatibility and thermal stability issues. The industrial application.
hydrophilic nature of cellulose causes irreversible agglomera-
tion during drying and aggregation in non-polar matrices Coating of cellulose nanocrystals with high molecular weight
because of the formation of additional hydrogen bonds polyoxyethylene (PEO)81,96 or polyamide 6 (PA6)97 prior to
between nanoparticles. Moreover, the thermal stability of cel- extrusion also showed encouraging results. Coating of the
lulose nanocrystals is relatively low, mainly when sulfuric nanoparticles was performed in the polymer solution using
acid is used for hydrolysis. Few solutions have been pro- water and formic acid for PEO and PA6, respectively. Dry
posed to address this challenge. A glance at the literature polymer-coated cellulose nanocrystals were obtained by
provides different strategies.52 freeze-drying or precipitation in water, respectively, and
extrusion was realized with polystyrene (PS),81 LDPE,96 and
Among them, a limited number hinted at encouraging PA6.97 Both improved dispersibility and thermal stability
results. Functionalization of the surface of cellulose nano- were observed. Grafting of polyethylene glycol (PEG) on

WWW.MATERIALSVIEWS.COM JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806 797
 JOURNAL OF
REVIEW WWW.POLYMERPHYSICS.ORG POLYMER SCIENCE

TABLE 3 Long Chain Surface Chemical Modifications of Cellu- BuA)) in latex form and cellulose nanocrystals.6,53 In the
lose Nanocrystals rubbery state of the thermoplastic matrix, the modulus of
the composite with a loading level as low as 6 wt %, that is
Grafted Chain Objective Ref. around 4 vol %, was more than two order of magnitude
PEG/PEO Dispersion in water and 79
higher than the one of the unfilled matrix. Moreover, the
non-aqueous solvents introduction of 6 wt % or more cellulosic nanocrystals pro-
vides an outstanding thermal stability of the matrix modulus
Steric stabilization in water 80
up to the temperature at which cellulose starts to degrade
Dispersion in PS 81
(around 500 K).
PEO-co-PPO Thermosensitive properties 82
PS Liquid crystalline behavior in DMF 83 The prediction of this reinforcing effect using the theoretical
Pollutant removal from water 84 model of Halpin–Kardos (mean field approach),98 which has
Dispersion in organic solvents 85
been extensively used to predict the elastic shear modulus of
short-fiber composites and semicrystalline polymers, failed
PCL Dispersion in dichloromethane 46,86
to describe the experimental data. On the contrary, a good
and PCL
agreement between experimental and predicted data was
Dispersion in dichloromethane 87
reported when using the series–parallel model of Takayanagi
and PLA
modified to include a percolation approach.6,53 It was con-
Dispersion in PCL 88 cluded that all the stiffness of the nanocomposite materials
89 was due to infinite aggregates of cellulose nanocrystals.
PMMAZO Liquid crystalline behavior in 90 Above the percolation threshold, the cellulosic nanoparticles
chlorobenzene can connect and form a three-dimensional continuous path-
PDMAEMA Liquid crystalline behavior in water 91 way through the nanocomposite film. The formation of this
PGMA Dispersion in PLA 92 cellulose network was supposed to result from strong inter-
PAA Polymer brush architecture 93 actions between nanocrystals, like hydrogen bonds. This
mechanical percolation effect explained both the high rein-
WPU Nanocomposite processing 94
forcing effect and the thermal stabilization of the composite
PLA Dispersion in PLA 95
modulus up to 500 K for cellulose nanocrystal reinforced
poly(S-co-BuA) films.
cellulose nanocrystals prior to coating with PEO was found
The formation of this rigid network was assumed to be gov-
to further improve the extrusion process at 200  C of PS-
erned by a percolation mechanism and the most important
based nanocomposites as shown in Figure 9.81
parameters are the aspect ratio of the rod-like nanoparticles,
which is the ruler of the percolation threshold, and the mod-
PROPERTIES OF NANOCOMPOSITES ulus of the percolating nanoparticle network, which is obvi-
ously different from the one of individual nanocrystals. This
Mechanical Properties modulus can be assumed to be similar, in principle, to the
The mechanical properties of cellulose nanocrystals are their one of a paper sheet for which the hydrogen bonding forces
most important and most investigated asset. Outstanding provide the basis of its stiffness. However, because of the
mechanical properties can be obtained by blending cellulose nanoscale effect stronger interactions are expected. This
nanocrystals and a polymer matrix even at low filler load- modulus can be experimentally determined from tensile tests
ing.5,6 Such properties originate from the high stiffness of performed on films prepared by water evaporation of a sus-
crystalline cellulose that provides the strength to higher pension of nanocrystals. Values ranging from about 1 to 15
plants, the nanoscale dimensions, and high aspect ratio of GPa have been reported and this value was found to be
the nanoparticles, and the high reactivity of cellulose. How- roughly correlated to the aspect ratio of the nanocrystal as
ever, to improve the mechanical properties of the host mate- shown in Figure 10.99
rial and take advantage of this property special care needs
to be paid to the processing of the nanocomposite. Therefore, any parameter that affects the formation of the
percolating nanocrystal network or interferes with it is a
The highest mechanical reinforcement effect of cellulose critical issue that changes dramatically the mechanical per-
nanocrystals is observed when using unmodified nanopar- formances of the composite.100 Three main parameters were
ticles bearing hydroxyl groups and for materials processed identified to affect the mechanical properties of such materi-
by casting/evaporation from a liquid medium in which they als, viz. the morphology and dimensions of the nanoparticles,
are homogeneously dispersed. It therefore mainly limits the the processing method, and the microstructure of the matrix
processing to water or polar liquid medium in which the cel- and matrix/filler interactions. As shown in Figure 10, higher
lulose nanocrystals can be dispersed without any surface aspect ratio nanocrystals provide higher reinforcement.
functionalization. In the pioneering work on cellulose
nanocrystal-reinforced materials, nanocomposite films were Regarding the processing method, slow wet processes (cast-
obtained by casting and evaporating a mixture of (poly(S-co- ing/evaporation) involving pristine cellulose nanocrystals

798 JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806
 JOURNAL OF
POLYMER SCIENCE WWW.POLYMERPHYSICS.ORG REVIEW

reinforcing effect. The resulting structure (after the coales-

cence of latex particles and/or interdiffusion of polymeric
chains) is completely relaxed and direct contacts between
pristine nanocrystals are then created.

To broaden the range of polymeric matrices that can be used

in association with cellulose nanocrystals, homogeneous dis-
persion of the nanoparticles in any liquid medium, regard-
less its polarity, is desirable. To promote this dispersion it is
necessary to lower the surface energy of the nanoparticles
to fit with that of the liquid or polymer melt. This is gener-
ally achieved by coating their surface with a surfactant or
replacing part of the surface hydroxyl groups by less polar
groups (surface chemical modification or physical coating).
However, even if the dispersion of the nanoparticles in the
liquid medium (in which the polymeric matrix is generally
FIGURE 8 Photographs of extruded (a) neat LDPE film and solubilized) is improved, hindered interactions between
nanocomposite films reinforced with 10 wt % (b) unmodified, physically or chemically modified nanocrystals limit the for-
and (c) stearoyl acid chloride-grafted cellulose nanocrystals. mation of a strong percolating nanoparticle network and
(Reproduced from ref. 10, with permission from [Elsevier]). then the reinforcing effect.

An abundant and flourishing literature is devoted to the

were reported to give the highest mechanical performance mechanical properties of cellulose nanocrystal-reinforced
materials compared to other processing techniques. Indeed, polymer nanocomposites. In most cases a reinforcing effect
during liquid evaporation strong interactions between nano- of the nanoparticles is claimed even if their potential is not
particles can settle and promote the formation of a strong always fully exploited, that is if the formation of a three-
percolating network. During slow water (or polar liquid) dimensional nanocrystal network within the polymeric
evaporation, because of Brownian motions in the suspension matrix is not achieved.
or solution (whose viscosity remains low, up to the end of
the process when the latex particle or polymer concentration This percolation phenomenon of cellulose nanocrystals
becomes very high), the rearrangement of the nanoparticles within a polymeric matrix has been used to prepare biomi-
is possible. They have time enough to interact and connect metic, stimuli-responsive nanocomposites using a low-
to form a percolating network which is the basis of their modulus matrix and cellulose nanocrystals as high-modulus

FIGURE 9 Photographs of extruded neat PS film and nanocomposite films reinforced with pristine (CN-PS), PEO-coated (CN/PEO-
PS), PEG-grafted (CN-g-PEG-PS), and both PEG-grafted and PEO-coated (CN-g-PEG/PEO-PS) cellulose nanocrystals. (Reproduced
from ref. 81, with permission from [American Chemical Society]).

WWW.MATERIALSVIEWS.COM JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806 799
 JOURNAL OF
REVIEW WWW.POLYMERPHYSICS.ORG POLYMER SCIENCE

explained on the basis of a helicoidal arrangement of birefrin-

gent layers, as is the case for cellulose nanocrystals in a chiral
nematic liquid crystal.110 The reflected wavelength becomes
shorter at oblique viewing angles, giving rise to visible irides-
cence colors when the pitch of the helix is on the order of the
wavelengths of visible light (between 400 and 700 nm).The
reflection color of these films can be controlled by simply
adding a little salt to the deionized suspension. Intensive
works have been reported on this behavior by Gray and
coworkers at McGill University.111,112

Ultrasound treatment was also found to increase the chiral

nematic pitch in suspension and red-shift the reflection
wavelength of cellulose nanocrystal films as the applied
energy increased.113 Figure 12 shows solid films cast from
aliquots of 2.8 wt % cellulose nanocrystal suspensions pre-
FIGURE 10 Evolution of the Young’s modulus of cellulose
pared by sulfuric acid hydrolysis from bleached softwood
nanocrystal films determined from tensile tests as a function of
kraft pulp and sonicated with increasing (left to right)
the constituting nanocrystals. (Reproduced from ref. 99, with
energy inputs. The films exhibit reflected iridescence with
permission from [Elsevier]).
colors ranging from blue-violet to red. By combining sonica-
tion and electrolyte addition the reflective properties of the
filler. The mechanical properties of these materials can be
film can be predictably tuned. The effects of sonicating a cel-
selectively and reversibly controlled through the formation lulose nanocrystal suspension were shown to be cumulative
and decoupling of the three-dimensional network of well- and permanent. Moreover, suspensions sonicated with differ-
individualized nanocrystals in response to specific chemical ent energy inputs can be mixed to prepare films having a
or thermal stimuli.101–106 reflection band intermediate between those obtained from
the individual suspensions. It was suggested that the
Optical Properties
ultrasound-induced red-shift is electrostatic in nature.
Birefringence of Nanocrystal Suspensions
Birefringence is the optical property of a material having a
refractive index that depends on the polarization and propaga-
tion direction of light. The birefringence of a cellulose nanocrys-
tal suspension can be observed using two linear polarizers at
90 . It gives a quick indication of the successful preparation of
well-dispersed nanocrystals obtained by sulfuric acid hydrolysis.
Figure 11 shows photographs of tunicin nanocrystal dispersions
observed in polarized light and formation of birefringent
domains.54 It has been first observed in 1959.107 This birefrin-
gence results from two origins: (i) a structural form anisotropy
of cellulose (anisotropic refractive index, Dn 0.05), and (ii) a
flow anisotropy resulting from alignment of the nanorods under
flow generally operated before observation.

Above a critical concentration of nanoparticle, chiral nematic

liquid crystalline order forms in aqueous suspensions of cellu-
lose nanocrystals.108 The observed chiral nematic order can
be enhanced by placing the suspension in a magnetic field,109
as cellulose has a negative diamagnetic susceptibility, meaning
that the rods tend to align perpendicular to the magnetic
field. The ratio of this anisotropic phase increases as the con-
centration increases, that is as the water evaporates. Under
proper conditions, the suspensions can be slowly evaporated
to produce solid semi-translucent films with optical properties
characteristic of chiral nematic liquid crystals that retain the
self-assembled chiral nematic liquid crystalline order formed FIGURE 11 Photographs of suspensions of tunicin nanocrystals
in the suspension. These films exhibit iridescence by reflecting (0.5 wt %) in (a) water, and (b) dimethylformamide (DMF)
left-handed circularly polarized light in a narrow wavelength observed between cross nicols, showing the formation of bire-
band determined by the chiral nematic pitch and the refrac- fringent domains. (Reproduced from ref. 54, with permission
tive index of the film. This phenomenon of reflectance was from [American Chemical Society]).

800 JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806
 JOURNAL OF
POLYMER SCIENCE WWW.POLYMERPHYSICS.ORG REVIEW

FIGURE 12 Cellulose nanocrystal films produced from suspensions treated with increasing applied ultrasonic energy (0, 250, 700,
1800, and 7200 J g21 of cellulose nanocrystal) from left to right. Viewing is normal to the film surface under diffuse lighting. Scale
marker 1 cm. (Reproduced from ref. 113, with permission from [American Chemical Society]).

Optical Properties of Nanocomposites n-octadecylisocyanate-grafted cellulose nanocrystals were

Generally, in addition to the mechanical reinforcing effect used as the reinforcing phase. Visual examination of the
provided by the inclusion of cellulose nanocrystals in a poly- films indicated that good filler dispersion was achieved for
mer matrix, the optical transparency of the polymeric matrix the lowest cellulose nanocrystal content (2.5 wt %) and the
is another aspect which is often aimed to be preserved. The transparency was considered like that of the neat polymeric
optical properties of cellulose nanocrystal-based nanocompo- matrix. The films reinforced with chemically modified nano-
site films can be investigated by determining the regular crystals [Fig. 13(b)] was always more transparent than for
light transmittance with a UV–visible spectrometer. Measure- unmodified nanoparticles [Fig. 13(a)]. At higher filler load-
ments are performed in the wavelength range 200–1000 nm. ings, aggregation of the unmodified nanoparticles in the
The regular light transmittance at 600 nm wavelength, which matrix can be observed giving rise to a granular structure. In
is in the middle of the visible wavelength range, is generally the case of modified nanoparticles, films became opaque but
reported. no aggregates were observed.

In general, 40 nm is the upper limit for nanoparticle diame- However, surprisingly, it was shown that the use of smaller
ter to avoid intensity loss of transmitted light due to Ray- diameter cellulose nanocrystals was more detrimental for
leigh scattering.114 Given the mismatch in the refractive the transparency of the polymeric matrix.117 This effect was
index of cellulose (1.58)115 and the one of most host matri- attributed to the percolation effect of the nanofiller. Indeed,
ces, the critical factor in the transparency of cellulose thin and high aspect ratio nanoparticles can more easily
nanofiller-based nanocomposites is the width of the nano- form a continuous network and in turn this cellulosic skele-
crystal or more specifically the effective cross-sectional area ton impacted negatively the transparency of the matrix by
of a scattering and its dispersion level within the host poly- promoting light scattering. It was then shown that mechani-
mer matrix. High transmittance values are therefore indica- cal and optical properties were two antinomic properties.
tive of a good transparency and dispersion of the cellulose
nanocrystals within the nanocomposite film. Barrier Properties
Nanofillers have strong reinforcing effects but some investi-
For example, Figure 13 shows photographs of nanocomposite gations have also shown their positive impact for barrier
films obtained by casting/evaporating a mixture of cellulose packaging materials. For decades, studies have been con-
nanocrystals and PLA in chloroform.116 Both unmodified and ducted with non-renewable inorganic fillers and petroleum-

FIGURE 13 Photographs of (a) neat PLA film and nanocomposites reinforced with (b) 2.5, (c) 7.5, and (d) 15 wt % unmodified or
(e) 2.5, (f) 7.5, and (g) 15 wt % chemically modified cellulose nanocrystals. (Reproduced from ref. 116, with permission from
[Elsevier]).

WWW.MATERIALSVIEWS.COM JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806 801
 JOURNAL OF
REVIEW WWW.POLYMERPHYSICS.ORG POLYMER SCIENCE

based matrices. Increasing environmental concerns have led

to developing new flexible barrier bio-based packaging and
investigating the potential uses of renewable resources for
such application.

The increasing interest in the barrier properties of cellulosic

nanoparticles is due to the increased tortuosity they pro-
vided. Indeed, because of their small size, the surface-to-
volume ratio of the nanoparticles is significantly greater than
for microparticles. Moreover, the low permeability of cellu-
lose can be enhanced by the highly crystalline nature of cel-
lulose nanocrystals and its ability to form a dense
percolating network. Provided that strong particle–polymer
molecular interactions exist, the smaller particles have a
greater ability to bond to the surrounding polymer material,
thereby reducing the chain segmental mobility and thus the FIGURE 14 Water uptake at equilibrium for glycerol plasticized
penetrant diffusivity. Barrier properties are commonly in waxy maize starch films conditioned at 25  C and 98% RH as a
connection with particular permeable objects, which include function of tunicin nanocrystal content. (Reproduced from ref.
common gases, water vapor, liquid, and organic substance, 49, with permission from [American Chemical Society]).
and so forth. In polymers they are necessarily associated to
their inherent ability to permit the exchange, to a higher or is inhibited, while positively-charged species are adsorbed by
lower extent, of low molecular weight substances through the film.120 Cellulose nanocrystals were also used to prepare
mass transport processes like permeation. The whole trans- ultrafine membranes for water purification.121
mitting process includes adsorption, dissolution, diffusion,
and desorption. Gases or water vapor enter the surface of
the material from higher density side. After diffusing inside Swelling and Sorption of Nanocomposites
the material, it desorbs on the low-density side. The permea- The interaction of polymeric materials with solvents can be
tion of low-molecular weight chemical species usually takes a problem from a technological point of view because the
place through the polymer amorphous phase and the crystal- dimensions and physical properties of the material may
line phase is considered impenetrable. change with the penetration of solvent molecules into the
specimen.

Barrier Properties of Nanocrystal films The water resistance of nanocomposite films reinforced with
The swelling and barrier properties of neat cellulose nano- cellulose nanocrystals depends on the hydrophilicity of the
crystal films have been much less investigated than for matrix. For hydrophilic matrices such as plasticized
mechanically extracted nanocellulose (NFC). The swelling of starch,49,122–129 soy protein isolate (SPI),130 glycerin plasticized
cellulose nanocrystal films was reported to be higher than carboxymethyl cellulose,131 carrageenan,132 and chitosan133
for NFC films.118,119 It was ascribed to water penetration both the water uptake and diffusion coefficient of water were
between the nanocrystals, causing their separation and found to decrease upon adding cellulose nanocrystals. For
allowing for a further imbibition into the film. Moreover, the example, Figure 14 shows the evolution of the water uptake of
sulfuric acid-prepared nanocrystals contain sulfate groups glycerol plasticized starch conditioned at 98% relative humid-
aiding in creating an osmotic pressure separating the crys- ity (RH) as a function of cellulose nanocrystal content.49 It
tals in the film and exposure to liquid water.118 The water decreased nonlinearly from about 60% for the unfilled matrix
diffusion coefficient was also higher for nanocrystal films down to 40% for nanocomposite films reinforced with 25 wt
compared to NFC films.119 It was attributed to the presence % tunicin nanocrystals. This behavior was ascribed to the exis-
of residual lignin, extractive substances, and fatty acids at tence of strong hydrogen bonding interactions between nano-
the surface of NFC. Nanocrystal films also showed signifi- particles and between the polar matrix and nanoparticles.
cantly higher permeability to gases and diffusion coefficients However, in some cases higher nanocrystal contents resulted
than NFC films. It was suspected that gas molecules pene- in agglomeration and increase of the water uptake.132
trate much more slowly in the NFC films because of a longer
diffusion path. In addition to a different surface chemistry, it For hydrophobic matrices, the opposite trend is obviously
was supposed that entanglements of the fibrils and lower observed as expected. This behavior was reported for polyvi-
porosity of the NFC films represented barrier domains nyl acetate (PVAc),104,134 poly(methyl methacrylate)
increasing the tortuosity of the diffusion pathway. (PMMA),105 and natural rubber (NR)135,136 reinforced with
cellulose nanocrystals. Diffusion of water molecules toward
Because of surface negatively-charged surface groups resulting the hydrophilic nanoparticles and localization in the weak
from the sulfuric acid hydrolysis step, cellulose nanocrystal interfacial zones were set out as the main explanation for this
membranes present permselective properties. The transfer of phenomenon. Therefore, possible formation of a continuous
negatively-charged species through the nanocrystal membrane percolation cellulose nanocrystal network and then aspect

802 JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806
 JOURNAL OF
POLYMER SCIENCE WWW.POLYMERPHYSICS.ORG REVIEW

ratio of the nanoparticles are important parameters with two A reduction of OP value was observed when adding up to 5
regimes, below and above the percolation threshold. Above wt % nanocrystals, with an optimal value around 2–3 wt %
this critical value the diffusion of water molecules is facili- filler loading. As for WVP experiments, this effect was attrib-
tated. When using a solvent of the polymeric matrix as swel- uted to the nanocrystal-induced crystallinity of the PLA
ling medium instead of water, the addition of cellulose matrix. This experimental permeability drop was higher than
nanocrystals decreases the swelling of the material.135,137 the one predicted by Nielsen and Fricke models even if
experimental data were corrected for crystallinity alterations.
Water Vapor Transfer and Permeability of Because experiments were conducted at 80% RH, it was
Nanocomposites hypothesized that sorbed moisture was thought to fill in the
The evolution of the water vapor transmission rate (WVTR) existing free volume reducing therefore the gas permeability.
or water vapor permeability (WVP) when adding cellulose Decreased oxygen transmission rate (OTR) and carbon diox-
nanocrystals also strongly depends on the hydrophilic char- ide transmission rate of PCL was observed with 5 wt % cel-
acter of the matrix. Decreased WVTR or WVP values were lulose nanocrystal incorporation and further improvement
reported for hydrophilic matrices such as polyvinyl alcohol was reported upon gamma irradiation treatment.144
(PVA),138 alginate-acerola,139 mango puree,140 glycerol plasti-
cized chitosan,141 and methylcellulose (MC)142 when rein- The barrier properties to oxygen and carbon dioxide of ter-
forced with cellulose nanocrystals. This behavior was nary systems consisting of NR reinforced with cellulose
attributed to the physical barrier role provided by nanocrys- nanocrystals and natural montmorillonite (MMT) were also
tals creating a tortuous path for the permeating moisture. investigated.146 The total filler content was kept constant
and equal to 5 wt % but changing the weight content of
For hydrophobic matrices, opposed tendencies have been each filler. Improved gas barrier properties were reported
reported. For NR reinforced with cellulose nanocrystals, an and explained by a tortuosity effect. For binary systems, a
increase of WVP was observed when adding the nanoparticles more efficient barrier effect was observed for MMT com-
up to a concentration of 7.5 wt % followed by a decrease.136 pared to cellulose nanocrystals. The tortuosity values calcu-
The former increase was ascribed to the hydrophilic nature of lated from permeability and diffusion coefficients indicated
cellulose and the decrease for higher cellulose nanocrystal that the simultaneous use of both fillers greatly slowed
contents was associated with the formation of a percolating down the gas diffusion leading to a synergistic effect. This
nanoparticle network. Reduction of the WVP value of PLA,132 phenomenon was ascribed to the formation of MMT-cellulose
xylan,143 and PCL144 was also reported when adding cellulose nanocrystal subassembly, the most efficient effect being
nanocrystals. For PCL-based films larger diffusion coefficients observed for MMT contents higher than 2.5 wt %.
reported for nanocomposites showed an increase of water dif-
fusivity due to hydrophilic nanocrystals whereas a reduction CONCLUSIONS
of water diffusivity was found when nanocrystals were
surface-modified owing to the reduction in polymer chain There has been an explosion of interest in the use of bio-
mobility and the water accumulation at the nanofiller–PCL mass as a source of renewable energy and materials. Despite
interfaces.77 Similar results were reported for cellulose nano- being the most available natural polymer on earth, it is only
crystal reinforced PLA.116 For PLA-based systems, quite recently that cellulose has gained prominence as a
nanocrystal-induced crystallization of the matrix was sus- nanostructured material, in the form of nanocellulose. This
pected to contribute to the reduction of water transmis- term covers the range of materials derived from cellulose
sion.132 Moreover, the reduction in water transmission of with at least one dimension in the nanometer range. It
xylan films was found to be more significant with sulfuric mainly consists of chemically (cellulose nanocrystals) or
acid-hydrolyzed than with hydrochloric acid-hydrolyzed cellu- mechanically extracted nanoparticles (nanofibrillated).
lose nanocrystals.143 Further decrease of WVP was observed Nanocellulose-based materials are carbon-neutral, sustain-
with 10 kGy gamma irradiation treatment.144 A continuous able, recyclable, and non-toxic. They thus have the potential
increase of the WVTR of nanocomposite films made of poly- to be truly green nanomaterials, with many useful and unex-
sulfone and cellulose nanocrystals was observed when pected properties. What is not to love?
increasing the nanoparticle content.145 It was associated with Many opportunities and potential applications have recently
the formation of a percolating cellulose network within the emerged. Among them novel functionalities added, including
polymeric matrix acting as a channel for the diffusion of as hydrogels or drug-delivery systems,147 as antimicrobial
water through the film. However, as cellulose crystals are agents (with silver nanoparticles, too),148 as smart materials,
rather impermeable to water, the formation of a weak inter- including co-polymerized with polyaniline for electroconduc-
phase in which diffusion is facilitated compared to the bulk tivity,149 and as edible films.139–141
matrix was suggested. This was supported by the low compat-
ibility between cellulose and the polysulfone matrix. An important feature of cellulose nanocrystals is their high
stiffness and elongated morphology. However, the properties
Gas Permeability of Nanocomposites of cellulose nanocrystal-reinforced polymer nanocomposites
The oxygen permeability (OP) of cellulose nanocrystal rein- are mainly governed by the possibility of strong inter-
forced PLA films was characterized at 24  C and 80% RH.132 particle interactions through hydrogen bonding, which are

WWW.MATERIALSVIEWS.COM JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806 803
 JOURNAL OF
REVIEW WWW.POLYMERPHYSICS.ORG POLYMER SCIENCE

exacerbated by the nano-scale effect resulting from their 18 T. Kondo, In Polysaccharides: Structural Diversity and Func-
dimensions. The challenge consists therefore in promoting tional Versatility; S. Dimitriu, Ed.; Marcel Dekker Inc.: New
the homogeneous dispersion of the cellulosic nanoparticles York, 1998; Chapter 3, pp 131–172–172.
and avoiding agglomeration during processing, thus requir- 19 K. Meyer, W. Lotmar, Helv. Chim. Acta 1936, 19, 68–86.
ing eventually favorable filler/matrix interactions, and at the 20 L. R. G. Treloar, Polymer 1960, 1, 290–303.
same time promoting filler/filler interactions to allow the 21 K. Tashiro, M. Kobayashi, Polym. Bull. 1985, 14, 213–218.
beneficial formation of a percolating network of nanopar- 22 M. A. Jaswon, P. P. Gillis, R.E. Mark, Proc. Roy. Soc. 1968,
ticles. These two requirements are simply conflicting. 306, 389–412.
23 L. M. J. Kroon-Batenburg, J. Kroon, M. G. Northolt, Polym.
After intensive research, several initiatives have emerged in Commun. 1986, 27, 290–292.
the perspective of producing nanocellulose on a large scale. 24 K. Tashiro, M. Kobayashi, Polymer 1991, 32, 1516–1526.
However, several challenges remain, for example sustainabil- 25 S. Reiling, J. Brickmann, Macromol. Theory Simul. 1995, 4,
ity of supply, melt extrusion of cellulosic nanoparticles- 725–743.
reinforced nanocomposites, and improvement of nanocompo- 26 S. Neyertz, A. Pizzi, A. Merlin, B. Maigret, D. Brown, X.
site properties in moist atmosphere among others. Deglise, J. Appl. Polym. Sci. 2000, 78, 1939–1946.
27 S. J. Eichhorn, G. R. Davies, Cellulose 2006, 13, 291–307.
ACKNOWLEDGMENTS 28 F. Tanaka, T. Iwata, Cellulose 2006, 13, 509–517.
29 M. Bergenstråhle, L. A. Berglund, K. Mazeau, J. Phys. Chem.
The authors gratefully acknowledge “Ci^encia Sem Fronteiras”
B 2007, 111, 9138–9145.
Program for the financial support (PhD fellowship of M. M.).
 n, G. P. Johnson, A. D. French, Cellulose 2011,
30 M. S. Cintro
18, 505–516.
31 X. Wu, R. Moon, A. Martini, Cellulose 2013, 20, 43–55.
REFERENCES AND NOTES
32 F. L. Dri, L. G. Hector Jr., R. J. Moon, P. D. Zavattieri, Cellu-
1 C. M. Roco, C. A. Mirkin, M. C. Hersam, J. Nanopart. Res. lose 2013, 20, 2703–2718.
2011, 13, 897–919. 33 J. Ganster, J. Blackwell, J. Mol. Model. 1996, 2, 278–285.
2 Terminology for Nanomaterials, British Standards, PAS 34 I. Sakurada, Y.Nukushina, T. Ito, J. Polym. Sci. 1962, 57,
136:2007; BSI, London, UK, 2007. 651–660.
3 J. B. Miller, E. K. Hobbie, J. Polym. Sci. Part B: Polym. Phys. 35 M. Matsuo, C. Sawatari, Y. Iwai, F. Ozaki, Macromolecules
2013, 51, 1195–1208. 1990, 23, 3266–3275.
4 K. J. Klabunde, J. Stark, O. Koper, C. Mohs, D. G. Park, S. 36 T. Nishino, K. Takano, K. Nakamae, J. Polym. Sci. Part B:
Decker, Y. Jiang, I. Lagadic, D. Zhang, J. Phys. Chem. 1996, Polym. Phys. 1995, 33, 1647–1651.
100, 12142–12153. 37 J. Mann, L. Roldan-Gonzales, Polymer 1962, 3, 549–553.
5 A. Dufresne, Nanocellulose: From Nature to High Perform- 38 R. Rusli, S. J. Eichhorn, Appl. Phys. Lett. 2008, 93, 033111-1-
ance Tailored Materials; Walter de Gruyter GmbH: Berlin/Bos- 3.
ton, 2012. 
39 A. Sturcova, G. R. Davies, S. J. Eichhorn, Biomacromole-
6 V. Favier, G. R. Canova, J. Y. Cavaille  , H. Chanzy, A. cules 2005, 6, 1055–1061.
Dufresne, C. Gauthier, Polym. Adv. Technol. 1995, 6, 351–355.
€ ller, Macromole-
40 I. Diddens, B. Murphy, M. Krisch, M. Mu
7 D. H. Milanez, R. M. Do Amaral, Mater. Res, 2013, 16, 635– cules 2008, 41, 9755–9759.
641.
41 S. Iwamoto, W. Kai, A. Isogai, T. Iwata, Biomacromolecules
8 The Global Market for Nanocellulose. Future markets Inc, 2009, 10, 2571–2576.
2013. 42 A. Dufresne, Mater. Today 2013, 16, 220–227.
9 T. Nishino, In Green Composites; C. Baillie, Ed.; CRC Press, 43 Y. Habibi, L. A. Lucia, O. J. Rojas, Chem. Rev. 2010, 110,
Boca Raton, Boston, New York, Washington DC, 2004; Chapter 3479–3500.
4, pp 61–79–79.
44 B. G. Rånby, Acta Chem. Scand. 1949, 3, 649–650.
10 A. J. de Menezes, G. Siqueira, A. A. S. Curvelo, A. Dufresne,
45 M. Roman, W. T. Winter, Biomacromolecules 2004, 5 1671–
Polymer 2009, 50, 4552–4563.
1677.
11 C. Mancera, F. Ferrando, J. Salvado , N. E. El Mansouri, Bio-
46 Y. Habibi, A. L. Goffin, N. Schiltz, E. Duquesne, P. Dubois,
mass Bioenerg. 2011, 35, 2072–2079.
A. Dufresne, J. Mat. Chem. 2008, 18, 5002–5010.
12 C. Crestini, M. Crucianelli, M. Orlandi, R. Saladino, Catal. ^a, A. de
47 K. B. R. Teodoro, E. M. Teixeira, A. C. Corre
Today 2010, 156, 8–22.
Campos, J. M. Marconcini, L. H. C. Mattoso, Polımeros 2011,
13 S. Qian, H. Mao, K. Sheng, J. Lu, Y. Luo, C. Hou, J. Appl. 21, 280–285.
Polym. Sci. 2013, 130, 1667–1674. 48 T. Ebeling, M. Paillet, R. Borsali, O. Diat, A. Dufresne, J. Y.
14 K. Kamide, Cellulose and Cellulose Derivatives; Elsevier , H. Chanzy, Langmuir 1999, 15, 6123–6126.
Cavaille
B.V., Amsterdam, The Netherlands, 2005. 49 M. N. Anglès, A. Dufresne, Macromolecules 2000, 33, 8344–
15 H. U. Suess, Pulp Bleaching Today; de Gruyter, GmbH & 8353.
Co. KG: Berlin/New York, 2010. 50 G. Siqueira, H. Abdillahi, J. Bras, A. Dufresne, Cellulose
16 L. B. Brasileiro, J. L. Colodette, D. P. Veloso, Quim. Nova 2010, 17, 289–298.
2001, 24, 819–829. 51 M. C. Roco, C. A. Mirkin, M. C. Hersam, Nanotechnology
17 M. J. John, S. Thomas, Carbohydr. Polym. 2008, 71, 343– Research Directions for Society Needs in 2020; WTEC Panel
364. Report, Springer, Boston and Berlin, 2010.

804 JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806
 JOURNAL OF
POLYMER SCIENCE WWW.POLYMERPHYSICS.ORG REVIEW

52 A. Dufresne, Int. Polym. Proc. 2012, 27, 557–564. 84 G. Morandi, L. Heath, W. Thielemans, Langmuir 2009, 25,
, Macromolecules 1995,
53 V. Favier, H. Chanzy, J. Y. Cavaille 8280–8286.
28, 6365–6367. € m, A. Hult, K. L.
85 S. Harrisson, G. L. Drisko, E. Malmstro
54 M. A. S. Azizi Samir, F. Alloin, J. Y. Sanchez, N. El Kissi, A. Wooley, Biomacromolecules 2011, 12, 1214–1223.
Dufresne, Macromolecules 2004, 37, 1386–1393. 86 Y. Habibi, A. Dufresne, Biomacromolecules 2008, 9, 1974–
55 O. Van den Berg, J. R. Capadona, C. Weder, Biomacromole- 1980.
cules 2007, 8, 1353–1357. 87 N. Lin, G. Chen, J. Huang, A. Dufresne, P. R. Chang, J. Appl.
56 M. Schroers, A. Kokil, C. Weder, J. Appl. Polym. Sci. 2004, Polym. Sci. 2009, 113, 3417–3425.
93, 2883–2888. 88 J. O. Zoppe, M. S. Peresin, Y. Habibi, R. A. Venditti, O. J.
57 G. Siqueira, J. Bras, A. Dufresne, Langmuir 2009, 26, 402– Rojas, ACS Appl. Mater. Interfaces 2009, 1, 1996–2004.
411. 89 A. L. Goffin, J. M. Raquez, E. Duquesne, G. Siqueira, Y.
58 L. Heux, G. Chauve, C. Bonini, Langmuir 2000, 16, 8210– Habibi, A. Dufresne, Ph. Dubois, Polymer 2011, 52, 1532–1538.
8212. 90 Q. Xu, J. Yi, X. Zhang, H. Zhang, Eur. Polym. J. 2008, 44,
59 I. Kvien, B. S. Tanem, K. Oksman, Biomacromolecules 2005, 2830–2837.
6, 3160–3165. 91 J. Yi, Q. Xu, X. Zhang, H. Zhang, Cellulose 2009, 16, 989–
60 J. F. Sassi, H. Chanzy, Cellulose 1995, 2, 111–127. 997.
61 H. Yuan, Y. Nishiyama, M. Wada, S. Kuga, Biomacromole- 92 M. Martinez-Sanz, M. A. Abdelwahab, A. Lopez-Rubio, J. M.
cules 2006, 7, 696–700. Lagaron, E. Chiellini, T. G. Williams, D. F. Wood, W. J. Orts, S.
H. Imam, Eur. Polym. J. 2013, 49, 2062–2072.
62 N. Wang, E. Ding, R. Cheng, Front. Chem. Eng. China 2007,
1, 228–232. 93 J. Majoinen, A. Walther, J. R. McKee, E. Kontturi, V. Aseyev,
J. M. Malho, J. Ruokolainen, O. Ikkala, Biomacromolecules
63 S. Berlioz, S. Molina-Boisseau, Y. Nishiyama, L. Heux, Bio-
2011, 12, 2997–3006.
macromolecules 2009, 10, 2144–2151.
94 X. Cao, Y. Habibi, L. A. Lucia, J. Mater. Chem. 2009, 19,
64 B. Braun, J. R. Dorgan, Biomacromolecules 2009, 10, 334–
7137–7145.
341.
95 A. L. Goffin, J. M. Raquez, E. Duquesne, G. Siqueira, Y.
65 M. J. Sobkowicz, B. Braun, J. R. Dorgan, Green Chem. 2009,
Habibi, A. Dufresne, Ph. Dubois, Biomacromolecules 2011, 12,
11, 680–682.
2456–2465.
66 N. S. Çetin, P. Tingaut, N. Ozmen, N. Henry, D. Harper, M. 96 K. Ben Azouz, E. C. Ramires, W. Van den Fonteyne, N. El
Dadmun, G. Sebe, Macromol. Biosci. 2009, 9, 997–1003. Kissi, A. Dufresne, ACS Macro Lett. 2012, 1, 236–240.
67 N. Lin, J. Huang, P. R. Chang, J. Feng, J. Yu, Carbohydr. 97 A. C. Corre ^ a, E. M. Teixeira, V. B. Carmona, K. B. R.
Polym. 2011, 83, 1834–1842. Teodoro, C. Ribeiro, L. H. C. Mattoso, J. M. Marconcini, Cellu-
68 X. Yu, S. Tong, M. Ge, L. Wu, J. Zuo, C. Cao, W. Song, J. lose 2014, 21, 311–322.
Environ. Sci. 2 013, 25, 933–943. 98 J. C. Halpin, J. L. Kardos, J. Appl. Phys. 1972, 43, 2235–
69 V. Khoshkava, M. R. Kamal, Biomacromolecules 2013, 14, 2241.
3155–3163. 99 J. Bras, D. Viet, C. Bruzzese, A. Dufresne, Carbohydr. Polym.
70 M. Grunert, W. T. Winter, J. Polym. Environ. 2002, 10, 27– 2011, 84, 211–215.
30. 100 A. Dufresne. J. Nanosci. Nanotechnol. 2006, 6, 322–330.
71 M. Grunert, W. T. Winter, Polym. Mater. Sci. Eng. 2002, 86, 101 J. R. Capadona, K. Shanmuganathan, D. J. Tyler, S. J.
367–368. Rowan, C. Weder, Science 2008, 319, 1370–1374.
72 C. Gousse , H. Chanzy, G. Excoffier, L. Soubeyrand, E.
102 J. Mendez, P. K. Annamalai, S. J. Eichhorn, R. Rusli, S. J.
Fleury, Polymer 2002, 43, 2645–2651. Rowan, E. J. Foster, C. Weder, Macromolecules 2011, 44, 6827–
73 A. Pei, Q. Zhou, L. Berglund, Compos. Sci. Technol. 2010, 6835.
70, 815–821. 103 K. Shanmuganathan, J. R. Capadona, S. J. Rowan, C.
74 J. M. Raquez, Y. Murena, A. L. Goffin, Y. Habibi, B. Ruelle, Weder, Prog. Polym. Sci. 2010, 35, 212–222.
F. DeBuyl, P. Dubois, Compos. Sci. Technol. 2012, 72, 544–549. 104 K. Shanmuganathan, J. R. Capadona, S. J. Rowan, C.
75 M. de Oliveira Taipina, M. M. Favaro Ferrarezi, I. V. Pagotto Weder, J. Mater. Chem. 2010, 20, 180–186.
Yoshida, M. do Carmo Gonçalves, Cellulose 2013, 20, 217–226. 105 K. Shanmuganathan, J. R. Capadona, S. J. Rowan, C.
76 G. Siqueira, J. Bras, A. Dufresne, Biomacromolecules 2009, Weder, ACS Appl. Mater. Interfaces 2010, 2, 165–174.
10, 425–432. 106 K. L. Dagnon, K. Shanmuganathan, C. Weder, S. J. Rowan,
77 N. Follain, S. Belbekhouche, J. Bras, G. Siqueira, S. Marais, Macromolecules 2012, 45, 4707–4715.
A. Dufresne, J. Membr. Sci. 2013, 427, 218–229. 107 R. H. Marchessault, F. F. Morehead, N. M. Walter, Nature
78 G. Siqueira, J. Bras, N. Follain, S. Belbekhouche, S. Marais, 1959, 184, 632–633.
A. Dufresne, Carbohydr. Polym. 2013, 91, 711–717. 108 J. F. Revol, H. Bradford, J. Giasson, R. H. Marchessault, D.
79 J. Araki, M. Wada, S. Kuga, Langmuir 2001, 17, 21–27. G. Gray, Int. J. Biol. Macromol. 1992, 14, 170–172.
80 E. Kloser, D. G. Gray, Langmuir 2010, 26, 13450–13456. 109 J. F. Revol, L. Godbout, X. M. Dong, D. G. Gray, H. Chanzy,
81 N. Lin, A. Dufresne, Macromolecules 2013, 46, 5570–5583. G. Maret, Liq. Cryst. 1994, 16, 127–134.
82 F. Azzam, L. Heux, J. L. Putaux, B. Jean, Biomacromolecules 110 H. I. de Vries, Acta Crystallogr. 1951, 4, 219–226.
2010, 11, 3652–3659. 111 X. M. Dong, T. Kimura, J. F. Revol, D. G. Gray, Langmuir
83 J. Yi, Q. Xu, X. Zhang, H. Zhang, Polymer 2008, 49, 4406– 1996, 12, 2076–2082.
4412. 112 X. M. Dong, D. G. Gray, Langmuir 1997, 13, 2404–2409.

WWW.MATERIALSVIEWS.COM JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806 805
 JOURNAL OF
REVIEW WWW.POLYMERPHYSICS.ORG POLYMER SCIENCE

113 S. Beck, J. Bouchard, R. Berry, Biomacromolecules 2011, 132 M. D. Sanchez-Garcia, J. M. Laragon, Cellulose 2010, 17,
12, 167–172. 987–1004.
114 H. Althues, J. Henle, S. Kaskel, Chem. Soc. Rev. 2007, 36, 133 J. P. de Mesquita, C. L. Donnici, I. F. Teixeira, F. V. Pereira,
1454–1465. Carbohydr. Polym. 2012, 90, 210–217.
115 V. Landry, A. Alemdar, P. Blanchet, Forest Prod. J. 2011, 134 N. L. Garcia de Rodriguez, W. Thielemans, A. Dufresne,
61, 104–112. Cellulose 2006, 13, 261–270.
116 E. Espino-Pe rez, J. Bras, V. Ducruet, A. Guinault, A.
135 A. Bendahou, H.Kaddami, A.Dufresne, Eur. Polym. J. 2010,
Dufresne, S. Domenek, Eur. Polym. J. 2013, 49, 3144–3154. 46, 609–620.
117 S. Boufi, H. Kaddami, A. Dufresne, Macromol. Mater. Eng.,
136 J. Bras, M. L. Hassan, C. Bruzesse, E. A. Hassan, N. A. El-
in press. Article first published online: 7 Oct 2013, DOI:
Wakil, A. Dufresne, Ind. Crops Prod. 2010, 32, 627–633.
10.1002/mame.201300232.
, G. R. Canova, R. Dendievel, B.
137 L. Chazeau, J. Y. Cavaille
118 C. Aulin, S. Ahola, P. Josefsson, T. Nishino, Y. Hirose, M.
€
Osterberg, L. Wågberg, Langmuir 2009, 25, 7675–7685. Boutherin, J. Appl. Polym. Sci. 1999, 71, 1797–1808.
119 S. Belbekhouche, J. Bras, G. Siqueira, C. Chappey, L. 138 S. A. Paralikar, J. Simonsen, J. Lombardi, J. Membr. Sci.
Lebrun, B. Khelifi, S. Marais, A. Dufresne, Carbohydr. Polym. 2008, 320, 248–258.
2011, 83, 1740–1748. 139 H. M. C. Azeredo, K. W. E. Miranda, M. F. Rosa, D. M.
120 W. Thielemans, C. R. Warbey, D. A. Walsh, Green Chem. Nascimento, M. R. de Moura, Food Sci. Technol. 2012, 46, 294–
2009, 11, 531–537. 297.
121 H. Ma, C. Burger, B. S. Hsiao, B. Chu, Biomacromolecules 140 H. M. C. Azeredo, L. H. C. Mattoso, D. Wood, T. G.
2011, 12, 970–976. Williams, R. J. Avena-Bustillos, T. H. McHugh, J. Food Sci.
122 A. P. Mathew, A. Dufresne, Biomacromolecules 2002, 3, 2009, 74, N31–N35.
609–617. 141 H. M. C. Azeredo, L. H. C. Mattoso, R. J. Avena-Bustillos,
123 Y. Lu, L. Weng, X. Cao, Macromol. Biosci. 2005, 5, 1101– G. C. Filho, M. L. Munford, D. Wood, T. H. McHugh, J. Food
1107. Sci. 2010, 75, N1–N7.
124 Y. Lu, L. Weng, X. Cao, Carbohydr. Polym. 2006, 63, 198– 142 R. A. Khan, S. Salmieri, D. Dussault, J. Uribe-Calderon, M.
204. R. Kamal, A. Safrany, M. Lacroix, J. Agric. Food Chem. 2010,
125 X. Cao, Y. Chen, P. R. Chang, M. Stumborg, M. A. 58, 7878–7885.
Huneault, J. Appl. Polym. Sci. 2008, 109, 3804–3810. 143 A. Saxena, T. J. Elder, A. J. Ragauskas, Carbohydr. Polym.
126 X. Cao, Y. Chen, P. R. Chang, A. D. Muir, G. Falk, eXPRESS 2011, 84, 1371–1377.
Polym. Lett. 2008, 2, 502–510. 144 R. A. Khan, S. beck, D. Dussault, S. Salmieri, J. Bouchard,
127 Y. Chen, C. Liu, P. R. Chang, X. Cao, D. P. Anderson, Car- M. Lacroix, J. Appl. Polym. Sci. 2013, 129, 3038–3046.
bohydr. Polym. 2009, 76, 607–615.
145 S. Noorani, J. Simonsen, S. Atre, Cellulose 2007, 14, 577–
128 D. Liu, T. Zhong, P. R. Chang, K. Li, Q. Wu, Bioresource 584.
Technol. 2010, 101, 2529–2536.
146 A. Bendahou, H. Kaddami, E. Espuche, F. Gouanve  , A.
129 E. M. Teixeira, D. Pasquini, A. A. S. Curvelo, E. Corradini,
Dufresne, Macromol. Mater. Eng. 2011, 296, 760–769.
M. N. Belgacem, A. Dufresne, Carbohydr. Polym. 2009, 78, 422–
431. 147 N. Lin, A. Dufresne, Biomacromolecules 2013, 14, 871–880.
130 Y. Wang, X. Cao, L. Zhang, Macromol. Biosci. 2006, 6, 524– € ller,
148 S. Berndt, F. Wesarg, C. Wiegand, D. Kralisch, F. A. Mu
531. Cellulose 2013, 20, 771–783.
131 Y. J. Choi, J. Simonsen, J. Nanosci. Nanotechnol. 2006, 6, 149 L. H. Mattoso, E. S. Medeiros, D. A. Baker, J. Avloni, D. F.
633–639. Wood, W. J. Orts, J. Nanosci. Nanotechnol. 2009, 9, 2917–2922.

806 JOURNAL OF POLYMER SCIENCE, PART B: POLYMER PHYSICS 2014, 52, 791–806