You are on page 1of 8

Reactive & Functional Polymers xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Reactive & Functional Polymers


journal homepage: www.elsevier.com/locate/react

Influence of mechanical treatments on the properties of cellulose


nanofibers isolated from microcrystalline cellulose
Davide Bandera a,b, Janak Sapkota c, Sébastien Josset a, Christoph Weder c, Philippe Tingaut a, Xin Gao d,
E. Johan Foster c,e,⇑, Tanja Zimmermann a,⇑
a
Swiss Federal Laboratories for Materials Science and Technology, Laboratory for Applied Wood Materials, Überlandstrasse 129, CH-8600 Dübendorf, Switzerland
b
Swiss Federal Laboratories for Materials Science and Technology, Laboratory for Bioactive Materials, Lerchenfeldstrasse 5, CH-9014 St. Gallen, Switzerland
c
Adolphe Merkle Institute, Polymer Chemistry and Materials, University of Fribourg, Rte de l’Ancienne Papeterie, CH-1723 Marly 1, Switzerland
d
Frewitt Fabrique des Machines SA, Case postale 615, 1701 Fribourg, Switzerland
e
Virginia Tech, Department of Materials Science & Engineering, 445 Old Turner Street, 213 Holden Hall, Blacksburg, VA 24061, USA

a r t i c l e i n f o a b s t r a c t

Article history: The possibility of preparing cellulose whiskers-like materials by mechanical treatment of commercially
Received 30 June 2014 available microcrystalline cellulose (MCC) was explored. High shear homogenization, grinding with a
Received in revised form 10 September supermass colloider, and hammer-milling were the processes selected to disintegrate the MCC, which
2014
yielded F-MCC, G-MCC and H-MCC, respectively. Processing aqueous dispersions with high solid content
Accepted 12 September 2014
Available online xxxx
allowed for the production of cellulose particles with greatly reduced dimensions. Morphological charac-
terization revealed that homogenization and grinding reduced the particle size more effectively than
hammer-milling, although the disintegration was incomplete in all cases. The reinforcing potential of
Keywords:
Microcrystalline cellulose
the materials was evaluated against commercially available whiskers by using the various particles as
Cellulose whiskers fillers to mechanically reinforce hydroxypropylcellulose. Nanocomposite films containing 5, 10, or
Mechanical treatments 20 wt.% of the filler were prepared and the mechanical properties were characterized. The results show
Film nanocomposites that H-MCCs are just slightly better than the original MCC, whereas F-MCC and G-MCC performed similar
Hydroxypropylcellulose to whiskers at 10 wt.% loading, despite the presence of a fraction of micrometer-sized particles. It is
therefore reasonable to envision the use of the more easily produced F-MCC and G-MCC as replacement
of cellulose whiskers in some applications.
Ó 2014 Published by Elsevier B.V.

1. Introduction Different treatments of native cellulosic materials have shown


to be efficient in producing nanometer or micrometer sized fibers
Cellulose is among the most abundant biopolymers available on [12–14]. Nanofibrillated cellulose (NFC), is obtained by disintegra-
earth [1] and can be found in plants, marine animals and bacteria tion [15] or refining [16] of cellulose fibers by mechanical
[2–6]. This biopolymer displays many desirable properties, includ- treatment, which is often combined with chemical or enzymatic
ing biodegradability, renewability, high thermal stability, high pre-treatments. This process affords water-dispersible and inter-
strength and extraordinary stiffness [7]. The chemical modification connected nanofibers with lengths in the micrometer range and
of cellulose hydroxyl groups has facilitated the expansion of its widths of usually above 10 nm. Many technologies have been
applications. It allows for the preparation of modified cellulosics employed such as high shear homogenization and grinding
possessing improved properties and the ability to interact with processes [17,18]. NFC can be produced from a wide range of cel-
several other materials of various polarity [8]. Over the last lulose-based raw materials including wood [19–22], wheat straw
decades its use has been exploited particularly in applications [23,24], bamboo [25,26], or aquatic weeds [27]. In these processes
including paper, packaging films, pharmaceuticals, food and both the amorphous and crystalline parts of cellulose are present
textiles [9–11]. in the final products [13].
Cellulose whiskers, also referred to as cellulose nanocrystals
(CNCs) in the literature, on the other hand, are individualized
⇑ Corresponding authors. Tel.: +41 (0)26 300 9281; fax. +41 (0)26 300 96 24 (E.J.
rod-like particles with aspect ratios between 10 and 100 and
Foster). Tel.: +41 (0)58 765 4115; fax. +41 (0)58 765 4007 (T. Zimmermann).
diameters from 4 to 30 nm, and are in general much shorter than
E-mail addresses: johan.foster@unifr.ch (E.J. Foster), tanja.zimmermann@empa.
ch (T. Zimmermann). micro- or nanofibrils [28]. They are usually produced by treating

http://dx.doi.org/10.1016/j.reactfunctpolym.2014.09.009
1381-5148/Ó 2014 Published by Elsevier B.V.

Please cite this article in press as: D. Bandera et al., React. Funct. Polym. (2014), http://dx.doi.org/10.1016/j.reactfunctpolym.2014.09.009
2 D. Bandera et al. / Reactive & Functional Polymers xxx (2014) xxx–xxx

the cellulose source with strong acids such as sulfuric [29], (0.99% sulfur on dry CNC). For all experiments deionized water
hydrochloric [30], phosphoric [31] or hydrobromic acid [32]. was used.
Combinations of acidic treatment and high-pressure homogeniza-
tion have also been reported [33,34]. Acid treatments remove the 2.2. Preparation of microfluidized MCC (F-MCC)
amorphous parts of cellulose leaving behind the highly crystalline
regions. Cotton [29], softwood [32], hardwood [35], sisal [36], bac- F-MCC was produced from a dispersion prepared with 0.3 kg of
terial cellulose [37], tunicates [38] and microcrystalline cellulose MCC in 1.5 kg of deionized water (20 wt.%), then stirred for 14 h at
(MCC) [39–41] have been frequently reported as starting materials room temperature. The dispersion was then fibrillated with a high
for the production of cellulose whiskers. pressure according to the setup previously described [17]. In par-
Increased attention has been given over the last decades to the ticular the 20% dispersion was pumped at 400 kPa through: (a)
study of polymers nanocomposites, containing mechanically two consecutive chambers of 400 and 200 lm (H230Z400lm and
strong fillers, such as clays, carbon nanotubes and silica, which H30Z200lm) for 3 passes; (b) two consecutive chambers of 200
are capable to improve the mechanical properties of neat polymers and 75 lm (H30Z200lm and H20Y75lm) for 7 passes. A defined
[42]. In a context of environmental preservation and optimization amount of water was added after each pass in order to decrease
of resources, the use of biobased cellulose nanofibers as reinforcing the viscosity of the suspension. After the first pass 400 mL of deion-
agents in composite materials has been widely reported [12]. ized water were added and then 150 mL after each pass. At the end
Elastic moduli of 150 GPa and 114 GPa have already been esti- of the mechanical treatment, the dispersion was stored at 4 °C and
mated for cellulose whiskers and the crystalline parts of nanofibr- the solid content determined by drying a known amount of disper-
illated cellulose, respectively [43,44]. Both types of nanofibers have sion into an oven at 105 °C until stable mass. The final solid content
been successfully employed with several polymers providing a was 10.9 wt.%.
valuable, but also renewable alternative to conventional fillers.
Some examples of polymer composites containing different types
2.3. Preparation of grinded MCC (G-MCC)
of whiskers or nanofibrillated cellulose, which showed improved
properties, are poly(styrene-co-butyl acrylate) [45], poly(caprolac-
G-MCC was produced starting from a dispersion prepared with
tone) [36,46], poly(vinyl acetate) [47] and hydroxypropylcellulose
1.3 kg of MCC in 8 kg of deionized water (16.3%), then stirred for
[17,48,49].
16 h at room temperature. The dispersion was then grinded follow-
Nevertheless, the production of cellulose whiskers by acid
ing previous processing conditions [18] with a Supermass Colloider
hydrolysis requires tedious neutralization steps that have so far
(MKZA10-20 J CE) from Masuko Sangyo Co. Ltd. (Kawaguchi/Saitam-
partially hindered the mass-scale production of these nanofibers.
a, Japan) equipped with non-porous grinding stones made of SiC/
Accordingly, in this paper we present the impact of different
Al2O3 particles in resin. The grinding stone rotated at ca.
mechanical treatments on commercially available MCC to produce
1500 rpm with a motor-load of 15 kW. The dispersion was passed
cellulose nanomaterials, as an alternative to the classical acidic
12 times through the grinder. 1 L of deionized water was added
process commonly reported for production of crystal-like material.
after each pass from pass 4 to pass 10 in order to decrease the vis-
We then evaluated the ability of the produced materials to rein-
cosity of the suspension (7 L of water total). After this mechanical
force hydroxypropylcellulose (HPC), a water soluble polymer
treatment the dispersion was stored at 4 °C and the solid content
matrix. Since our nanomaterials displayed stable suspensions in
determined by drying a known amount of dispersion into an oven
water, an improved dispersibility within the polymeric matrix
at 105 °C until stable mass. The final solid content was 8.7 wt.%.
has been expected. In particular, we compared the performance
of the materials we produced with regards to commercially avail-
able nanocrystals and possibly foresee the replacement of such 2.4. Preparation of dry Hammer-milled MCC (H-MCC)
products in niche applications. A high shear homogenization pro-
cess involving a microfluidizer [48,49] and a grinding process using The dry milling of the MCC was performed using a Hammer
a Masuko Sangyo supermass colloider [18] have been evaluated in Witt-LAB (Frewitt SA, Fribourg, Switzerland) at hammer side using
this study, together with a dry hammer-milling treatment. All a screen ‘Chevron’ 0.2 mm at 7000 rpm. 7 consecutive cycles were
these methods have potential to be used in scale-up productions. operated in order to obtain finer particles (1 kg of MCC was milled
The obtained materials were characterized in terms of morphology in 42 s cycles).
and crystallinity. The composite films were prepared by a solvent
casting approach. The morphological and mechanical properties 2.5. Preparation of HPC composite films containing xMCC (x = F, G or
of the HPC composites were systematically evaluated using SEM, H) or commercial cellulose whiskers (W)
AFM, optical microscopy, sieve analysis and tensile test experi-
ments, respectively. The material produced by high-shear homog- Aqueous dispersions with a solid content of 3 wt.% HPC or
enization, grinding and hammer milled processes were 3 wt.% HPC and 5, 10 or 20 wt.% xMCC or whiskers (with respect
denominated F-MCC, G-MCC and H-MCC, respectively. Commer- to the 3 wt.% of polymer) were prepared by mixing a 7% HPC solu-
cially available whiskers were labelled W. tion with the desired amount of xMCC or whiskers and deionized
water. F-MCC and G-MCC were used as aqueous dispersions as
obtained from mechanical treatment, whereas H-MCC and whisker
2. Experimental dispersions were prepared by ultrasonicating for 2 min (10 s ON-
10 s OFF cycles at 10% power) a 3.8 wt.% and a 0.8 wt.% suspension,
2.1. Materials respectively. The prepared mixtures were all treated with an Ika
ultraturrax system for 1 min at 8000 rpm, then heated and mag-
Microcrystalline cellulose Avicel Lattice NT 100 was purchased netically stirred for 30 min at 40 °C, left cooling to room tempera-
from FMC Biopolymer and hydroxypropylcellulose (HPC) with a ture and degassed under vacuum for 5 min. 55 g of each dispersion
weight-average molecular weight MW = 100,000 g mol 1 was pur- were poured into a silicon mold (15  18 cm) and the water was
chased from Sigma Aldrich. Commercially available cellulose left evaporating in a climatized room at 20 °C and 60% RH for
nanocrystals (herein referred to as whiskers, W) were purchased 5 days. The films were prepared in duplicate and had thicknesses
from the University of Maine Process Development Center from 50 to 70 lm.

Please cite this article in press as: D. Bandera et al., React. Funct. Polym. (2014), http://dx.doi.org/10.1016/j.reactfunctpolym.2014.09.009
D. Bandera et al. / Reactive & Functional Polymers xxx (2014) xxx–xxx 3

2.6. Scanning Electron Microscopy (SEM) between elongations from 0.05% and to 0.25%. The results pre-
sented are the mean values and the standard deviation of at least
SEM images were taken from samples of diluted 0.05 wt.% sus- 16–20 measurements.
pensions of which 2–3 drops were placed on a freshly cleaved mica
surface sample holder. All samples were sputter-coated directly 3. Results and discussion
with a platinum layer of about 7 nm (BAL-TEC MED 020 Modular
High Vacuum Coating Systems, BAL-TEC AG, Liechtenstein) in Ar The starting material selected for this comparative study was
as a carrier gas at 5  10 2 mbar. SEM images were finally recorded commercially available MCC because of its availability and low cost
with a FEI Nova NanoSEM 230 instrument (FEI, Hillsboro, Oregon, (64 $/kg vs. 1000 $/kg). The morphology of this crystalline material
USA) at an accelerating voltage of 5 kV and a working distance of has been evaluated by scanning electron microscopy (SEM). The
5 mm. SEM images reveal large particles with typical sizes below
120 lm (Fig. 1). This result has also been confirmed by sieve anal-
2.7. Atomic Force Microscope Imaging (AFM) ysis which yielded an average value of 108 lm (see Fig. S1).

AFM measurements were performed using atomic force micro- 3.1. Preparation of cellulose nanomaterials from MCC
scope (JPK instruments AG, Berlin, Germany) with a Nanowizard II
head and a SPM Controller III. A dilute drop of aqueous suspension A first product was obtained by mechanically treating MCC with
(0.01–0.1 mg ml 1) was dispersed on the freshly cleaved musco- the high shear homogenization process. An unusual high cellulose
vite mica disc surface (SPI Supplies, West Chester, Pennsylvania, concentration of 20 wt.% in water has been envisaged in our study.
USA, 9.9 mm diameter) mounted on the optical glass slide and This process is usually carried out at concentrations 10 times or
allowed to dry at 70 °C for an hour prior to analysis. AFM images more lower, both for the production of fibrils [17] and whiskers
were recorded in air with a silicon nitride cantilever tip (NANO [51]. The reason for this choice was that a higher concentration
WORLD, TESPA-50) in tapping mode with a spring constant of increases the viscosity of the dispersion, leading to higher shear
42 N/m at the resonance frequency of 320 kHz and a scan rate of stress on the MCC particles and the probability of friction forces
0.7 line s 1. Lengths and width reported in our study are the aver- between crystals. This was assumed to render the process more
age of 70 analyzed objects. efficient towards size reduction, but also to increase the output

2.8. Optical microscopy

Optical microscopy analyses were performed using Olympus


BX51 (Olympus Corporation) on samples of 20–30 lm thick films.
Each film was mounted on a glass microscopy slide and the images
were recorded with an objective lens of 50x magnification
equipped with DP72 digital camera.

2.9. Sieving

Sieve analysis was performed using Retsch Analytical Sieve


Shaker AS 200 Control for 30 g of each samples with a vibration
rate of 2 min and amplitude of 2 mm g 1. Vulkollan cubes were
added as sieving aid.

2.10. X-ray diffraction

The XRD measurements were performed with a Bruker D8


Advanced Diffractometer (Cu Ka radiation, k = 1.54184 Å), tube
parameters: V = 40 kV, I = 40 mA, using a parallel incident beam
prepared by a Göbel X-ray mirror. The patterns were recorded with
Fig. 1. Scanning electron microscopy image of commercially available MCC.
a 1D position sensitive detector (VÅNTEC). The samples were
investigated in h/2h mode using a rotating disk-shaped sample
holder (rotated at 15 rpm) on which the sample was spread to form
a flat surface. The scans were recorded in the angular range from
5 < 2h < 90° with a step width of D2h = 0.015° at a speed of
5° min 1, respectively. The data have been analyzed using Micro-
soft excel according to the peak height method [50].

2.11. Tensile testing

For each nanocomposite film composition 10 to 12 dog-bone


shaped specimens were cut of (length 5 cm, surface 330 mm2,
width of the central part 4 mm). They were allowed to equilibrate
for 48 h at 23 °C and 50% RH before measurement. The mechanical
properties in tension were determined using a Zwick Z010/TH2A
(Zwick GmbH & Co. KG, Ulm, Germany): the pre-load applied
was 0.2 N, the loading speed 10 mm min 1 on a 100 N loading cell.
The tensile modulus (E) was determined using the strain curve Fig. 2. X-ray diffractograms plot of MCC, G-MCC, F-MCC, H-MCC and W.

Please cite this article in press as: D. Bandera et al., React. Funct. Polym. (2014), http://dx.doi.org/10.1016/j.reactfunctpolym.2014.09.009
4 D. Bandera et al. / Reactive & Functional Polymers xxx (2014) xxx–xxx

of the product. A significant increase in the suspension viscosity Hammer-milling of MCC was performed for one kg of starting
was noticed after only 3 passes. Remediation to this issue, which MCC in the dry state affording powder particles of reduced diame-
made the process more difficult due to the lower flow rate, con- ter in every consecutive step of 7 cycles. No heating of the product
sisted in the addition of a defined volume of water after each pass was observed during milling and hence no cooling medium was
and proceeding with the homogenization treatment for 10 passes. used. The final average particle size was determined to be 59 lm
The final product (F-MCC) had a concentration of 10.9 wt.%. according to sieve analysis (Fig. S2).
The grinding treatment of MCC with the Masuko Sangyo super-
mass colloider [18] was performed starting with a 16.3 wt.% cellu-
3.2. Crystallinity analysis
lose suspension in water, which was gradually introduced into the
equipment and the material was mechanically grinded through
The X-ray diffraction patterns of MCC and the products
two rotating disks having an estimated clearance of 100 lm. Dur-
obtained by the different mechanical treatments are presented in
ing this treatment, the problematic increase of the suspension vis-
Fig. 2. The crystallinity index (CI) of each material is reported in
cosity was also addressed by adding a defined volume of water
Table 1. A CI of 91% has been measured for the original MCC, which
after each pass. The final dispersion (G-MCC) had a concentration
is in agreement with other studies reported in the literature [41].
of 8.7 wt.%.
Similar CIs have been calculated for all treated materials, which
suggests that the crystal structure of MCC was not affected by
Table 1
Summary of crystallinity and size measurements for the processed materials. the strong mechanical treatments used in our study. Commercially
available cellulose whiskers displayed also a CI similar to the origi-
Material CIa Length Width
nal MCC and the other materials produced.
MCC 84 – 108 lmc
F-MCC 90 416 ± 105 nmb 8 ± 2 nmb
G-MCC 82 270 ± 80 nmb 38 ± 9 nmb 3.3. Morphological characterization of mechanically treated MCC
H-MCC 82 – 59 lmc
W 79 343 ± 93 nmb 51 ± 14 nmb
The morphology of the mechanically treated MCC has been
a
Crystallinity index determined by XRD. evaluated by SEM. A substantial size reduction can be observed
b
Determined by AFM. for F-MCC (Fig. 3) after the homogenization treatment as compared
c
Determined by sieve analysis as diameter. with the starting material, but the material was not homogeneous.

Fig. 3. Scanning electron microscopy images of (a) F-MCC deposited from a 0.05 wt.% aqueous dispersion and (b) a magnification of the area marked with the white square.

Fig. 4. Scanning electron microscopy images of (a) G-MCC and (b) H-MCC deposited from a 0.05 wt.% aqueous dispersion.

Please cite this article in press as: D. Bandera et al., React. Funct. Polym. (2014), http://dx.doi.org/10.1016/j.reactfunctpolym.2014.09.009
D. Bandera et al. / Reactive & Functional Polymers xxx (2014) xxx–xxx 5

The disintegrated F-MCC displayed cellulosic nanocrystals together The dry hammering process yielded an H-MCC material with
with larger particles in the micrometer range although our high reduced size compared to MCC but still in the micrometer length
concentration process seems more efficient than the previously range. From SEM images of MCC (Fig. 1) and H-MCC in Fig. 4b
reported one where the extent of size reduction seemed less effec- and Table 1, the width of the latter appears to be reduced but
tive [51]. not as much as after homogenization or grinding, and larger
The grinded material (G-MCC) presented a structure similar to objects are still present. This was confirmed by sieve analysis of
F-MCC with the presence of nanometer and micrometer sized par- the powders which resulted in the confirmation of size reduction
ticles (Fig. 4a), although in this case the smaller sized particles from an average of 108 lm for MCC to 59 lm for H-MCC (Fig. S1
seemed to be more abundant. and Fig. S2 in the SI).

500 nm 500 nm

500 nm 500 nm

500 nm 500 nm

1
Fig. 5. Height (left) and phase (right) atomic force microscopy images of (a) F-MCC, (b) G-MCC, and (c) commercial whiskers deposited from a 0.01–0.1 mg ml aqueous
dispersion.

Please cite this article in press as: D. Bandera et al., React. Funct. Polym. (2014), http://dx.doi.org/10.1016/j.reactfunctpolym.2014.09.009
6 D. Bandera et al. / Reactive & Functional Polymers xxx (2014) xxx–xxx

The dimensions of smaller crystals found in F-MCC, G-MCC and content of 5, 10 and 20 wt.% were produced by solvent casting.
commercially available whiskers were further accessed by Atomic Their properties were compared with HPC films reinforced with
Force Microscopy (AFM, Fig. 5). The average length and width mea- the same content of MCC and commercially available cellulose
sured for these materials are reported in Table 1. It is immediately whiskers (W).
noticeable that F-MCC contained nanometer-sized objects with a The morphology of the solvent-cast films has been analyzed
length of 416 ± 105 nm and a width of 8 ± 2 nm, similar to cellulose with optical microscopy (Fig. 6). The neat HPC film was transparent
whiskers already reported in the literature from the same source. as expected, except for small bubbles in the films, which could
The analysis of G-MCC, confirmed the presence of cellulose whis- have been formed during the film preparation by solution casting.
kers, which display lengths of 270 ± 80 nm and width of The nanocomposite films containing 10 wt.% of the various fillers
38 ± 9 nm, in agreement with previous observations of materials show well embedded filler particles within the polymer matrix
produced by sonication of MCC [40]. Commercially available whis- but the dispersion state was difficult to assess. Larger filler parti-
kers display lengths of 343 ± 93 nm and width of 51 ± 14 nm. cles are clearly distinguishable in the optical images for materials
From this morphological analysis, it appears that the homogeni- with larger particles such as MCC, H-MCC. The observed particles
zation process produced thinner fibers compared to the grinding, are smaller for F-MCC, G-MCC containing films and for W films
whereas lengthwise the two are similar. The hammer-milling, on they are not distinguishable because of the nanoscale dimensions
the other end, reduced the size of MCC but not to the same extent of this filler.
achieved for F-MCC and G-MCC. In general, it seems possible to tai- The mechanical properties of the composite films have been
lor the size reduction of MCC, based on the type of mechanical evaluated by tensile tests. The evolution of the tensile modulus
treatment. (E), tensile strength (r) and strain at break (e) as a function of filler
content is shown for the various materials in Fig. 7 (see also
3.4. Evaluation of the reinforcing potential of F-MCC, G-MCC and Table S1, S2 and S3 of the SI for detailed values).
H-MCC in HPC nanocomposites As it is well known that increased interfacial interaction
between the filler and the polymer matrix results in better
Although neither of the processes employed here completely mechanical performance [52–54], we expected this to favor the
disintegrates MCC to CNCs/whiskers, we speculate that the signif- interaction of our filler with HPC, because both materials are
icant content of CNCs/whiskers would render the products suitable hydrophilic and capable of hydrogen bonding.
as fillers. Hence, the reinforcing potential of F-MCC, G-MCC and H- Neat HPC films displayed a tensile modulus of 561 ± 30 MPa, a
MCC has been evaluated by preparing nanocomposites with tensile strength of 15.5 ± 0.7 MPa and a strain at break of
hydroxypropylcellulose (HPC). Thin films with cellulose nanofiller 29.1 ± 4.1%. The introduction of MCC provoked a negligible

a b

c d

e f

Fig. 6. Optical micrographs of selected nanocomposite films containing hydroxypropylcellulose (HPC) and 10 wt.% of the nanocellulose filler: (a) commercial MCC, (b) F-MCC,
(c) G-MCC, (d) H-MCC, (e) whiskers and (f) Neat HPC (scale bars = 0.5 lm).

Please cite this article in press as: D. Bandera et al., React. Funct. Polym. (2014), http://dx.doi.org/10.1016/j.reactfunctpolym.2014.09.009
D. Bandera et al. / Reactive & Functional Polymers xxx (2014) xxx–xxx 7

Furthermore, differences in reinforcing potential have been


noticed between these fillers. A progressive increase in r has been
observed for HPC reinforced with G-MCC and F-MCC with increas-
ing filler loadings, while no evolution was seen for H-MCC. These
differences have been attributed to the higher aspect ratios of
G-MCC and F-MCC compared to H-MCC, as well as the possibility
of nanoparticles percolation/interaction within the polymeric
matrix. These results are in agreement with the particles morpho-
logical differences discussed previously and the confirmed
presence of nanometer scale objects in G-MCC and F-MCC.
Interestingly, a similar reinforcing potential has been observed
for G-MCC and F-MCC, with maximum r values reaching
24.5 ± 2.5 MPa and 25.4 ± 2.4 MPa for 20 wt.% of fillers, respec-
tively, in line with other works where maximum improvement
was observed with 20 wt.% of filler [17,48]. This observation sug-
gests that although G-MCC has a significantly larger diameter
and their length is comparable to F-MCC, in the end this difference
does not affect the reinforcing power displayed, which is analo-
gous. According to percolation theory, the threshold for these
types of filler can be calculated with the simplified formula
mc  [0.7/(L/d)] 100 where L and d are the length and the width
of the particles, respectively (considering only the fraction of mate-
rial having whisker-like size) [28]. It turns out that for F-MCC and
G-MCC the threshold would be 1.3% and 9.8% respectively. These
values fit well with our results for which we observe a more
modest increase of E for F-MCC and G-MCC with respect to the fil-
ler loading. For 5 wt.% our results agree with a previous report [49].
HPC composites reinforced with W displayed higher E and
r than any other materials for 5 wt.% and 20 wt.% filler content,
but similar to F-MCC and G-MCC at 10 wt.%. These results highlight
the potential of F-MCC and G-MCC in composite applications and
represent a viable alternative to cellulose whiskers prepared with
strong acidic conditions.
For all composites in this study, a typical decrease of e has been
noted with the filler content, as already reported for composite
materials reinforced with cellulose particles.

4. Conclusions

Mechanical treatments of MCC offer the possibility to reduce


the particle size of this highly crystalline cellulosic material. All
methods used here potentially allow for easy and high-throughput
production of materials with a method-dependent fraction and
morphology of nanoparticles. Hammer-milling affords H-MCC,
which is mainly composed of low-aspect-ratio particles. Homoge-
nization produces F-MCC with long and thin fibers, whereas grind-
ing affords G-MCC with particles that feature a similar length but a
Fig. 7. Graphs show: (a) the evolution of the tensile modulus (E), (b) tensile
larger diameter. Despite this size difference, F-MCC and G-MCC
strength (r) and (c) strain at break (e) for HPC nanocomposites reinforced with G-
MCC, F-MCC, H-MCC and W as function of the filler content. display a similar reinforcing behavior in HPC films for a range of fil-
ler loadings. Well-defined whiskers, which were used as reference,
have a higher reinforcing capability than any of the MCC-based
increase of E but a sharp decrease in tensile strength, which might
materials at low and high filler content, but G-MCC and F-MCC
be associated with an inhomogeneous dispersion of the filler in the
reinforced composites match the properties of whisker-reinforced
composite materials or the formation of a weak MCC-HPC
HPC at a filler content of 10 wt.%. Thus, even though the mechani-
interface.
cal treatments afford cellulosic materials that are comprised of
A gradual increase of E has been noted upon the increase of G-
mixtures of nano- and microparticles, this distribution does, at
MCC, F-MCC and H-MCC contents in the HPC matrix, as expected,
least not in the HPC composites studied here, significantly limit
which reflected a gradual increase of the composites stiffness.
the reinforcing capability in comparison to well-defined whiskers.
Hence, the cellulosic particles display a higher stiffness than the
On the other hand, for use in optical applications a better disinte-
soft HPC polymeric material. Notably, E in F-MCC and G-MCC nano-
gration may be required.
composites show similar behavior and the reinforcing effect at 10%
is indistinguishable from that of the whisker nanocomposites.
Lower E values were noted for composites reinforced with Acknowledgments
H-MCC and levelling off effect has been observed above 10 wt.%
of filler content, which was possibly due to the larger particles as We thank Dr. Arndt Remhof for helping with XRD data acquisi-
well as a filler aggregation, respectively [36,55]. tion and Ms. Esther Strub for tensile testing, both at the Swiss

Please cite this article in press as: D. Bandera et al., React. Funct. Polym. (2014), http://dx.doi.org/10.1016/j.reactfunctpolym.2014.09.009
8 D. Bandera et al. / Reactive & Functional Polymers xxx (2014) xxx–xxx

Federal Laboratories for Materials Science and technology. The [21] M. Henriksson, G. Henriksson, L.A. Berglund, T. Lindström, Eur. Polym. J. 43
(2007) 3434–3441.
authors gratefully acknowledge financial support from the Swiss
[22] T. Saito, S. Kimura, Y. Nishiyama, A. Isogai, Biomacromolecules 8 (2007) 2485–
National Science Foundation (NRP66: Resource Wood, Nr. 2491.
406640_136911/1) and the Adolphe Merkle Foundation. [23] A. Kaushik, M. Singh, Carbohyd. Res. 346 (2011) 76–85.
[24] M.S.M. Hrabalova, R. Wimmer, A. Gregorova, T. Zimmermann, N. Mudliger,
Bioresources 6 (2011) 1631–1647.
Appendix A. Supplementary material [25] W. Chen, H. Yu, Y. Liu, Y. Hai, M. Zhang, P. Chen, Cellulose 18 (2011) 433–442.
[26] K. Abe, H. Yano, Cellulose 17 (2010) 271–277.
[27] G. Nyström, A. Razaq, M. Strømme, L. Nyholm, A. Mihranyan, Nano Lett. 9
Supplementary data associated with this article can be found, in (2009) 3635–3639.
the online version, at http://dx.doi.org/10.1016/j.reactfunctpolym. [28] M.A.S. Azizi Samir, F. Alloin, A. Dufresne, Biomacromolecules 6 (2005) 612–
2014.09.009. 626.
[29] X. Dong, J.-F. Revol, D. Gray, Cellulose 5 (1998) 19–32.
[30] J. Araki, M. Wada, S. Kuga, T. Okano, Coll. Surf. A 142 (1998) 75–82.
References [31] S. Camarero Espinosa, T. Kuhnt, E.J. Foster, C. Weder, Biomacromolecules 14
(2013) 1223–1230.
[1] D. Klemm, B. Heublein, H.-P. Fink, A. Bohn, Angew. Chem. Int. Ed. 44 (2005) [32] I. Filpponen, D.S. Argyropoulos, Biomacromolecules 11 (2010) 1060–1066.
3358–3393. [33] X.Z.M. Pan, Bioresources 8 (2013) 933–943.
[2] S.M. Mukherjee, H.J. Woods, Biochim. Biophys. Acta 10 (1953) 499. [34] A.P. Mathew, K. Oksman, Z. Karim, P. Liu, S.A. Khan, N. Naseri, Ind. Crop. Prod.
[3] B. Wang, M. Sain, K. Oksman, Appl. Compos. Mater. 14 (2007) 89–103. 58 (2014) 212–219.
[4] J.N. Saddler, H.H. Brownell, L.P. Clermont, N. Levitin, Biotechnol. Bioeng. 24 [35] S. Beck-Candanedo, M. Roman, D.G. Gray, Biomacromolecules 6 (2005) 1048–
(1982) 1389–1402. 1054.
[5] M.M. De Souza Lima, J.T. Wong, M. Paillet, R. Borsali, R. Pecora, Langmuir 19 [36] G. Siqueira, J. Bras, A. Dufresne, Biomacromolecules 10 (2009) 425–432.
(2002) 24–29. [37] J. Araki, S. Kuga, Langmuir 17 (2001) 4493–4496.
[6] M.M. de Souza Lima, R. Borsali, Langmuir 18 (2002) 992–996. [38] M.N. Anglès, A. Dufresne, Macromolecules 33 (2000) 8344–8353.
[7] G.W.D. Fengel, Wood, Chemistry, Ultrastructure, Reactions, Walter de Gruyter, [39] D. Bondeson, A. Mathew, K. Oksman, Cellulose 13 (2006) 171–180.
New York, 1983. [40] J.R. Capadona, K. Shanmuganathan, S. Trittschuh, S. Seidel, S.J. Rowan, C.
[8] Y. Habibi, Chem. Soc. Rev. 43 (2014) 1519–1542. Weder, Biomacromolecules 10 (2009) 712–716.
[9] M. FitzPatrick, P. Champagne, M.F. Cunningham, R.A. Whitney, Bioresour. [41] D.P.A.N. Frone, D. Donescu, C.I. Spataru, C. Radovici, R. Trusca, R. Somoghi,
Technol. 101 (2010) 8915–8922. Bioresources 6 (2011) 487–512.
[10] H.J. Prado, M.C. Matulewicz, Eur. Pol. J. 52 (2014) 53–75. [42] L.S.S.O.M. Jayan, P.B. Brown, Nanocomposite Science and Technology (2003).
[11] A.G.M.N. Belgacem, Production, chemistry and properties of cellulose-based Wiley VCH, Weinheim.
materials, in: D. Plackett (Ed.), New Materials for Sustainable Films and [43] A. Šturcová, G.R. Davies, S.J. Eichhorn, Biomacromolecules 6 (2005) 1055–
Coatings, John Wiley & Sons Ltd., Chichester (UK), 2011, pp. 151–178. 1061.
[12] S.J. Eichhorn, A. Dufresne, M. Aranguren, N.E. Marcovich, J.R. Capadona, S.J. [44] Y.C. Hsieh, H. Yano, M. Nogi, S.J. Eichhorn, Cellulose 15 (2008) 507–513.
Rowan, C. Weder, W. Thielemans, M. Roman, S. Renneckar, W. Gindl, S. Veigel, [45] V. Favier, H. Chanzy, J.Y. Cavaille, Macromolecules 28 (1995) 6365–6367.
J. Keckes, H. Yano, K. Abe, M. Nogi, A.N. Nakagaito, A. Mangalam, J. Simonsen, [46] A. Morin, A. Dufresne, Macromolecules 35 (2002) 2190–2199.
A.S. Benight, A. Bismarck, L.A. Berglund, T. Peijs, J. Mater. Sci. 45 (2010) 1–33. [47] K. Shanmuganathan, J.R. Capadona, S.J. Rowan, C. Weder, J. Mater. Chem. 20
[13] Y. Habibi, L.A. Lucia, O.J. Rojas, Chem. Rev. 110 (2010) 3479–3500. (2010) 180–186.
[14] S.J. Eichhorn, Soft Matter 7 (2011) 303–315. [48] T. Zimmermann, E. Pöhler, P. Schwaller, Adv. Eng. Mater. 7 (2005) 1156–1161.
[15] A.F. Turback, F.W. Snyder, K.R. Sandberg, J. Appl. Polym. Sci. Appl. Polym. [49] S.-Y. Lee, S.-J. Chun, I.-A. Kang, J.-Y. Park, J. Ind. Eng. Chem. 15 (2009) 50–55.
Symp. 37 (1983) 797–813. [50] R.I.N. Terinte, R. Ibbett, K.C. Schuster, Lenzinger Ber. 89 (2011) 118–131.
[16] K. Spence, R. Venditti, O. Rojas, Y. Habibi, J. Pawlak, Cellulose 18 (2011) 1097– [51] J.W.Q. Cheng, J.F. McNeel, P. Jacobson, Bioresources 5 (2010) 1945–1954.
1111. [52] L. Tang, C. Weder, ACS Appl. Mater. Interfaces 2 (2010) 1073–1080.
[17] T. Zimmermann, E. Pöhler, T. Geiger, Adv. Eng. Mater. 6 (2004) 754–761. [53] S. Kumar, M. Hofmann, B. Steinmann, E.J. Foster, C. Weder, ACS Appl. Mater.
[18] P.O.S. Josset, G. Siqueira, A. Tejado, P. Tingaut, T. Zimmermann, Nord. Pulp Pap. Interfaces 4 (2012) 5399–5407.
Res. J. 29 (2014) 167–175. [54] M. Jorfi, M.N. Roberts, E.J. Foster, C. Weder, ACS Appl. Mater. Interfaces 5
[19] K. Uetani, H. Yano, Biomacromolecules 12 (2010) 348–353. (2013) 1517–1526.
[20] W. Chen, H. Yu, Y. Liu, P. Chen, M. Zhang, Y. Hai, Carbohyd. Polym. 83 (2011) [55] Q. Wu, M. Henriksson, X. Liu, L.A. Berglund, Biomacromolecules 8 (2007)
1804–1811. 3687–3692.

Please cite this article in press as: D. Bandera et al., React. Funct. Polym. (2014), http://dx.doi.org/10.1016/j.reactfunctpolym.2014.09.009

You might also like