Article
pubs.acs.org/Biomac
Cilia-Mimetic Hairy Surfaces Based on End-Immobilized
Nanocellulose Colloidal Rods
Arcot R. Lokanathan,*,† Antti Nykan̈ en,§ Jani Seitsonen,‡ Leena-Sisko Johansson,† Joseph Campbell,†
Orlando J. Rojas,†,∥ Olli Ikkala,*,§ and Janne Laine†
†
Department of Forest Products Technology, Aalto University, P.O. Box 16300, FIN-00076 Espoo, Finland
Nanomicroscopy Center and §Department of Applied Physics, Aalto University, P.O. Box 15100, FIN-00076 Espoo, Finland
∥
Departments of Forest Biomaterials and Chemical and Biomolecular Engineering, North Carolina State University, Raleigh, North
Carolina 27695, United States
‡
S Supporting Information
*
ABSTRACT: We show a simple method toward nanoscale
cilia-like structures, i.e., functional hairy surfaces, upon
topochemically functionalizing nanorods of cellulose nanocrystals (CNCs) with thiol end groups (CNC-SHs), which
leads to their immobilization onto a gold surface from one end,
still allowing their orientational mobility. CNCs having a
lateral dimension of 3−5 nm and length of 50−500 nm
incorporate the native crystalline structure with hydrogenbonded cellulose chains in the parallel configuration. This
facilitates asymmetric, selective chemical modification of the
reducing ends through reductive amination. Successful thiol
functionalization is demonstrated using cryo transmission
electron microscopy based on selective attachment of silver nanoparticles to the CNC-SH ends to form Janus-like colloidal rod−
sphere adducts. The extent of thiol modification of CNC-SHs is quantified using X-ray photoelectron spectroscopy. The
promoted binding of CNC-SHs on gold surfaces is shown by atomic force microscopy and quartz crystal microbalance, where the
high dissipation suggests pronounced orientational mobility due to flexible joints at one rod end onto the surfaces. That the
joints are flexible is also shown by the bending and realignment of the CNC-SH rods using a receding triple-phase evaporation
front of a drying drop of water. The ability of the hairy surface to size-selectively resist particle binding was also investigated. As
the CNCs are piezoelectric and allow magnetic functionalization by nanoparticles, we foresee a general platform for nanosized
artificial cilia for fluid manipulation and controlled adsorption/desorption.
■
INTRODUCTION
length scale of cilia have remained challenging, especially using
biological materials.
On the other hand, plant biomass is one of the most
abundant renewable natural resources for polymers.11 It mainly
consists of cellulose assembled in crystalline and amorphous
domains as well as heteropolysaccharides and lignin.12
Cellulosics might one day reduce mankind’s dependence on
nonrenewable polymeric materials.11 Beyond structural applications, cellulosics can allow various types of functionalities,
such as piezoelectricity,13,14 electro-active paper,15,16 haptics,17
wireless communication,18 and damage detection.19 Importantly, native nanocelluloses involving the cellulose I crystal
structure have attracted considerable recent attention to allow
mechanically strong materials.12,20,21 As they involve parallel
hydrogen-bonded cellulose chains, they cannot be obtained if
dissolution steps are used, but they can be directly cleaved from
Static or actuating hairy and bristle-like surface structures at
different length scales have aroused considerable interest
toward various surface functionalities, as inspired by control
of, e.g., wetting, adhesion, material transport, and insulation in
various biological systems.1 Cilia in biological systems have
several functions: they keep lung airways clean from debris due
to reduced adsorption, they are able to transfer fluid and
materials due to their capability for actuation, and they allow
signaling and sensing.2,3 The typical lateral dimension of a
cilium is 250 nm, and the length can be up to a few
micrometers. As biological cilia are complex, involving
“molecular motors”, various routes toward artificial cilia have
recently been introduced based on simpler forms of actuation,
notably magnetically.4−8 At a larger length scale, on the other
hand, hairy or bristle-like surfaces have attracted interest to
allow patterns and complexity.9,10 Therefore, static and
dynamic hairy surfaces are inherently interesting, where facile
routes for nanoscale/colloidal level structures approaching the
© 2013 American Chemical Society
Received: May 2, 2013
Revised: June 24, 2013
Published: June 25, 2013
2807
dx.doi.org/10.1021/bm400633r | Biomacromolecules 2013, 14, 2807−2813
Biomacromolecules
Article
Figure 1. Schematic illustration of reductive amination (upper scheme) at the reducing ends of a CNC (prepared by sulfuric acid hydrolysis) using
sodium cyanoborohydride (Cn), sodium triacetoxyborohydride (Ac), or 2-picoline−borane complex (Pc) under three different pH conditions (5, 7,
and 9.2). The lower scheme shows a cryo-TEM image of thiolated CNC rods (Ac, pH 7) functionalized with silver nanoparticles at their reducing
ends.
dynamical properties of the adlayer were understood using
the dissipation-frequency analysis of the QCM-D.
plant or wood cell walls. This leads to nanofibrillar objects of a
high modulus of ca. 140 GPa and high strength.22−24 Two
forms are obtained, both having a lateral dimension of a few
nanometers: rodlike cellulose nanocrystals (CNCs) and long
and entangled nanofibrillated cellulose.12,20,21 They have been
much pursued to reinforce nanocomposites, to construct
functional ductile aerogels and selective absorbents, and to
make coatings, films, aerogels, and fibers.23,25−28 Several studies
have successfully demonstrated various chemical routes to
modify CNCs to make them suitable for a variety of
applications, such as ferrocene-based electron conductors,29
fluorescently tagged bioimaging agents,30 surfactant-modified
drug delivery systems,31 imidazolium-functionalized ion exchangers,32 and thermal responsive materials,33,34 and allowing
controlled polymerization on CNCs.35
Here we report a facile method to introduce thiol groups
selectively at one end of CNC rods, as made possible by their
native crystalline internal structure consisting of parallel
cellulose chains. The reducing ends of CNCs are modified by
reductive amination (Figure 1) using sodium triacetoxyborohydride (Ac), sodium cyanoborohydride (Cn), or 2-picoline−
borane complex (Pc) under three different pH conditions. In
comparison to the previous reports on reducing end
functionalization of cellulose,36−38 the route presented here
allows a greater control over the way CNCs interact with each
other and also with components of composite materials using
the versatile thiol group dependent chemistry.39 In fact, the
synthesized CNC-SHs (thiolated cellulose nanocrystals) can be
regarded as colloidal level objects corresponding to the classic
molecular level low molecular weight alkane chains with thiol
end groups forming self-assembled monolayers (SAMs) on
noble metal surfaces.40 The analogy suggests modifying
surfaces, which in our case leads to cilia-mimetic materials,
taken the colloidal size of the CNC-SHs. The thiolation
reaction was followed by X-ray photoelectron spectroscopy
(XPS) and cryo transmission electron microscopy (cryo-TEM).
The adsorption on a gold surface was characterized using a
surface-sensitive quartz crystal microbalance with dissipation
(QCM-D) and atomic force microscopy (AFM). The
■
MATERIALS AND METHODS
Chemicals. Cn, Ac, Pc, potassium chloride, 6-amino-1-hexanethiol
hydrochloride (NH2(CH2)6SH·HCl), sodium phosphate dibasic,
sodium phosphate monobasic, sodium carbonate, sodium bicarbonate,
sodium acetate, sodium borohydride, silver nitrate, and glacial acetic
acid were purchased from Sigma-Aldrich Finland Oy, Helsinki,
Finland. Hardened ashless filter paper was bought from Whatman
GmbH, Dassel, Germany. Ultracel 3K centrifugal filters were procured
from Millipore Oy, Espoo, Finland. Spectra/por dialysis membrane,
MWCO 6000−8000, was purchased from Spectrum Laboratories Inc.,
Rancho Dominguez, CA.
Functionalizing CNCs by Reductive Amination. CNCs made
by sulfuric acid hydrolysis of ashless Whatman filter paper were used in
the reductive amination procedure. The powdered filter paper (10 g)
was added to 175 mL of sulfuric acid (64%, w/w) maintained at 45 °C,
and acid hydrolysis was stopped after 45 min by adding 1800 mL of
Milli-Q (MQ) water. The resulting dispersion was centrifuged, and the
washed sediment was dialyzed against MQ water, followed by a 24 h
incubation with 1 g of mixed bed resin. The resulting dispersion was
filtered to remove the resin particles, hence leaving a filtrate with
CNCs. The CNC aqueous dispersions (10 mg mL−1) of pH 5, 7, and
9.2 were buffered with 0.1 M sodium acetate, 0.1 M sodium phosphate,
and 0.1 M sodium carbonate buffers, respectively. The buffered CNC
dispersions (10 mL) were placed in a water bath maintained at 70 °C.
A total of 150 μmol of NH2(CH2)6SH·HCl and 22.5 mmol of
reducing agent (Cn, Ac, or Pc) were added in three steps over a period
of 3 days accompanied by stirring (50 μmol of NH2(CH2)6SH·HCl +
7.5 mmol of reducing agent day−1 for 3 days). After 3 days the reaction
mixture was cooled to room temperature, followed by addition of 10
mL of 3 M HCl to neutralize the excess reducing agent. (Caution:
Neutralization of sodium cyanobrohydride produces poisonous cyanide and
therefore this step should be carried out in a fume hood!) The neutralized
reaction mixture was dialyzed against MQ water, followed by addition
of appropriate amounts of KCl salt so that the final dispersion was 2 M
in KCl, and the resulting mixture stirred overnight. After incubation
with KCl, the reaction mixture was dialyzed against MQ water to
remove electrolytes. The CNC-SH dispersion obtained from the final
dialysis step was used for various characterization procedures.
Silver Nanoparticle Tagging and Cryo-TEM. The CNC-SH
aqueous dispersions obtained from reductive amination using different
reducing agents and pH conditions were diluted using MQ water for a
2808
dx.doi.org/10.1021/bm400633r | Biomacromolecules 2013, 14, 2807−2813
Biomacromolecules
Article
Figure 2. Cryo-TEM images of silver nanoparticles synthesized on the CNC-SH templates. Images a, c, and e represent CNC-SH samples prepared
at pH 9.2 using the reducing agents sodium triacetoxyborohydride, sodium cyanoborohydride, and 2-picoline−borane complex, respectively. The
scale bar in all images is 100 nm. Images b, d, and f are magnified 100 nm × 100 nm images of the areas marked by square boxes in images a, c, and e,
respectively.
final CNC-SH concentration of 1 mg mL−1. To 2 mL of diluted CNCSH dispersion was added 200 μL of aqueous silver nitrate solution (10
mM), followed by addition of 2 mL of sodium borohydride (5 mM),
and this final dispersion was used in the sample preparation procedure
for cryo-TEM. The cryo-TEM samples were prepared on holey carbon
grids using an FEI Vitrobot Mark IV. The solutions were pipetted (3
μL drops) onto the grids, blotted, and vitrified using a mixture of
ethane and propane. Zero-loss imaging of the vitrified samples was
carried out with a JEOL JEM 3200FSC 300 kV transmission electron
microscope operated at liquid nitrogen temperature.
X-ray Photoelectron Spectroscopy. CNC-SH films for XPS
measurement were made by drop-casting concentrated CNC-SH
dispersions on silicon wafers. XPS spectra were recorded with a Kratos
AXIS 165 electron spectrometer using an aluminum anode (mono Al
Kα) operated at 100 W, and a charge neutralizer was used to neutralize
the positive charge developed due to photoelectric emission. Highresolution S2p spectra were recorded using a 20 eV pass energy. The
doublet centered at 169 eV binding energy (BE) was assigned to S2p
of sulfate groups,41 and the peak centered at 163−165 eV BE was
assigned to S2p of thiol groups (Figure 3a).42 Since the concentration
of sulfate groups in CNC-SHs is known,43 it is possible to estimate the
concentration of thiol groups using the ratio of the areas of the XPS
peaks in the S2p spectrum corresponding to sulfate and thiol groups.
Quartz Crystal Microbalance with Dissipation. Adsorption
studies on gold surfaces were performed using a QCM-D E4
instrument (Q-Sense, Sweden). The quartz crystals coated with gold
were cleaned by UV ozone treatment for 30 min followed by
sonication in acetone for 30 min. Experiments were performed with 1
and 0.1 mg mL−1 CNC or CNC-SH dispersions. All measurements
were performed at 23 °C, with a flow rate of 100 μL min−1. First, CNC
or CNC-SH dispersions were flowed through the chambers for 30
min, followed by a rinse with MQ water. Once the QCM
measurement was finished, the crystals were dried using a jet of
nitrogen, followed by analysis using AFM.
Atomic Force Microscopy. AFM images (2 × 2 μm2) were
recorded in air using a Nanoscope IIIa multimode scanning probe
microscope from Digital Instruments (Santa Barbara, CA). Imaging
was done in the tapping mode using silicon cantilevers. The recorded
images were flattened, and the height scale was adjusted using
NanoScope Analysis 1.2 software. The NanoScope Analysis software
was also used to count the number of particles in each image frame
using the particle analysis feature.
Subjecting Chemisorbed CNC-SHs to Convective Flow. First,
chemisorption was performed by placing a 0.1 mL drop of 1 mg mL−1
CNC-SH or CNC dispersion on cleaned gold QCM crystals and
allowing them to stand for 5 min. The drop of CNC or CNC-SH
dispersion was removed after 5 min, followed by a rinse with MQ
water to remove loosely bound crystals. After the rinse, a 0.1 mL MQ
water drop was pipetted onto the gold surface, followed by placing a
glass coverslip over the drop of water in such a way that the drop was
sandwiched between the two slides. Two spacer slides between the
glass coverslip and gold crystal helped maintain the height of the
sandwiched drop. The spacer slides were placed at the edge such that
the periphery of the sandwiched drop did not touch the spacer slides.
The sandwiched drop of MQ water was allowed to evaporate at room
temperature, followed by AFM imaging. The AFM images were
obtained and processed in the same way as described in the previous
section.
Adsorbing Polystyrene Particles onto a CNC-SH-Functionalized Surface. Polystyrene (PS) particles of sizes 50 nm, 200 nm,
and 1 μm (poly bead microspheres, with the surface covered with
sulfate ligands) were purchased from Polysciences Gmbh, Germany.
Dispersions of particles with 0.01% (w/v) concentration were made in
MQ water and used for adsorption experiments. The ζ potential values
of the three different PS particles measured in MQ are presented in the
Supporting Information, Table T2. A QCM-D E4 instrument (QSense, Sweden) was used to perform the adsorption experiments.
QCM gold crystals were cleaned as mentioned before. All adsorption
experiments were performed using a flow rate of 0.1 mL min−1 at 23
°C. First, CNC-SH dispersion (0.1 mg mL−1) was introduced and
allowed to flow for 15 min, followed by a rinse with MQ water for 15
min. After the MQ water rinse, the flow of polystyrene dispersion was
started and continued for 2 h, followed by a rinse with MQ water for
30 min. The final MQ water rinse was followed by drying with a jet of
nitrogen gas. The dried crystals were analyzed using XPS, with the
same parameters mentioned earlier. The results and discussion related
to QCM-D and XPS studies of PS particle adsorption on CNC-SH
SAMs are included in the Supporting Information (Figures S2 and S3
and Table T1).
■
RESULTS AND DISCUSSION
Evidence for Topochemical Thiolation of CNCs. The
presence of the parallel configuration of cellulose I crystals
implies that any aldehyde-based reaction would result in
anisotropic distribution of the reacting species at one of the
2809
dx.doi.org/10.1021/bm400633r | Biomacromolecules 2013, 14, 2807−2813
Biomacromolecules
Article
reducing agents, and there was only a small difference in yields
between the reducing agents. The pKa of NH2(CH2)6SH is
close to 9.0, and hence, at pH 9.2 the amine group of the
molecule is deprotonated, which enables higher efficiency of
the reductive amination reaction.39 Since the thiol concentration obtained using reducing agents Ac and Pc was close to
that achieved using Cn at basic pH, it can be concluded that
both Ac and Pc could be used as safer alternatives to Cn in
aqueous conditions without compromising the extent of the
thiolation reaction.
Chemisorption of CNC-SHs on a Gold Surface. The
CNC-SH synthesized using Ac, pH 9.2, was chosen for detailed
adsorption studies. Adsorption of CNCs and CNC-SHs on a
gold surface was quantitatively studied using QCM-D and
AFM. The results of AFM studies are shown in Table 1 and
CNC ends. The reactivity of thiols with noble metals is utilized
here to prove the existence of an anisotropic distribution of
thiol groups in CNCs (CNC-SH). Silver nanoparticles
(AgNp’s) were synthesized in a dispersion of CNC-SHs, and
cryo-TEM was used to identify the location of AgNp’s with
respect to the ends of the CNCs. The TEM images of AgNptagged CNC-SHs synthesized using three different reducing
agents at pH 9.2 are shown in Figure 2. The 100 nm × 100 nm
images clearly show that the AgNp’s are bound to only one end
of the CNC-SHs. A control sample made by synthesizing
AgNp’s in a dispersion of unmodified CNCs had negligible
binding of AgNp’s to one end of the CNCs (data not shown).
This result indirectly confirms that the CNCs have parallel
cellulose chains and shows successful thiol functionalization
through reductive amination. Thus, we demonstrate that
thiolation followed by AgNp tagging is a reliable way to
qualitatively identify the configuration of CNCs and an
alternative to the conventional procedure involving chloritebased oxidation of aldehyde groups to carboxyl groups,
followed by AgNp tagging.38
Extent of CNC Thiol Functionalization. The concentration of thiol groups on CNC-SHs was quantified using XPS.
The XPS analyses were performed using high-resolution S2p
spectra (Figure 3a), and these results show that the thiol
Table 1. Surface Number Density (via AFM) and Shift in
Frequency (QCM) upon Adsorption of CNCs and CNC-SHs
on Gold Surfaces Using 0.1 and 1 mg mL−1 Dispersions (23
°C, 100 μL min−1 Flow Rate)
sample
CNC-SHs,
0.1 mg mL−1
CNC-SHs, 1 mg mL−1
CNCs, 0.1 mg mL−1
CNCs, 1 mg mL−1
AFM particle density
(μm−2)
QCM-D frequency
(Hz)
29.9 ± 1.6
−37.4 ± 2.8
42.0 ± 3.7
0.0 ± 0.0
12.6 ± 0.4
−44.7 ± 4.8
−4.3 ± 1.6
−8.9 ± 3.9
Figure 3. S2p high-resolution XPS spectrum of CNC-SHs (Ac, pH
9.2) showing two peaks corresponding to sulfate (binding energy 169
eV) and thiol (binding energy 164.5 eV) groups (a). Concentration of
sulfur (excluding sulfate) determined by XPS analysis of CNCs
modified by reductive amination using Ac, Cn, and Pc as reducing
agents under three different pH (5, 7, and 9.2) conditions (b).
Figure 4. AFM topography images of QCM gold crystals after CNCSH and CNC adsorption in QCM-D experiments. [CNC-SH] = 0.1
mg mL−1 (a) and 1 mg mL−1 (b). [CNC] = 0.1 mg mL−1 (c) and 1
mg mL−1 (d). The range of the height scale in all images is −2 to +12
nm, and the image size is 2 μm × 2 μm.
concentration was higher in CNC-SH samples functionalized at
pH 9.2 compared to those functionalized at neutral and acidic
pH conditions. For instance, with the reducing agent Ac, the
thiol concentration was estimated to be 26 ± 4.5, 20.4 ± 2, and
45.2 ± 5 μmol g−1 for pH 5, 7, and 9.2, respectively (Figure
3b). A higher yield was observed at basic pH for all three
Figure 4. The coverage measured in units of particles per square
micrometer clearly indicated higher amounts of adsorption by
CNC-SHs compared to CNCs. For instance, with a dispersion
having a 1 mg mL−1 concentration of CNCs or CNC-SHs, a
coverage value of 42.0 ± 3.7 particles μm−2 was observed for
CNC-SHs, while the CNC-adsorbed surface had a coverage of
only 12.6 ± 0.4 particles μm−2. In the case of adsorption with a
2810
dx.doi.org/10.1021/bm400633r | Biomacromolecules 2013, 14, 2807−2813
Biomacromolecules
Article
0.1 mg mL−1 dispersion, the unmodified CNCs did not
essentially adsorb on the gold surface. Representative AFM
images corresponding to the coverage values in Table 1 are
shown in Figure 4, and in these images one can clearly visualize
the higher coverage in the case of CNC-SHs (Figure 4a,b)
relative to CNCs (Figure 4c,d). The results of QCM-D-based
adsorption studies showed a trend similar to that observed with
AFM studies. Representative QCM-D plots of adsorption
involving CNCs/CNC-SHs on gold are shown in the
Supporting Information (Figure S1). The frequency decrease
was higher in the case of CNC-SH adsorption when compared
to CNC adsorption. For example, with 0.1 mg mL −1
dispersions, the frequency decrease in the case of CNC-SHs
was greater than that for CNCs by a factor of 9. The higher
extent of adsorption of CNC-SHs was attributed to the
presence of thiol groups at the reducing ends of CNCs, thus
resulting in chemisorption due to chemical affinity between
thiol and gold.
Besides the adsorbed amount, the QCM-D technique
provides information on the viscoelastic properties of the
adlayer.44 The slope of the dissipation vs frequency plot
(ΔD−ΔF plot) is inversely related to the rigidity of the
adsorbed film.45 Representative QCM ΔD−ΔF plots for
adsorption of CNC-SHs and CNCs on gold surfaces are
shown in Figure 5. Significantly higher slopes are observed for
Figure 6. Schematic representation of CNC-SHs chemisorbed on a
gold surface to form a hairy surface, immersed in water (a), and
chemisorbed nanorods aligned by an evaporating water front, thus
showing that the CNC-SH rods on the gold surface have flexible joints,
allowing realignment (b).
Flexible Immobilization of the Chemisorbed CNC-SHs
from One End on a Gold Substrate and Alignment
Using Convective Flow. AFM images of gold surfaces that
were exposed to CNC-SH (synthesized using Ac, pH 9.2) and
CNC dispersions followed by convective flow are presented in
Figure 7, as locally imposed by the receding water droplet upon
Figure 5. ΔD−ΔF plot, change of dissipation as a function of the
change in frequency, for CNC-SHs and CNCs adsorbed on QCM
gold substrates based on QCM-D studies. Note that the gray and
green lines are overlapping.
CNC-SHs in comparison to CNCs, thereby indicating that the
chemisorbed thiolated CNCs are far more flexible relative to
unmodified CNCs. The CNC-SH rods chemisorb onto the
gold surface using just one end and repel each other
electrostatically due to the presence of sulfate groups along
the rods. On the basis of these two considerations, it can be
proposed that the CNC-SH rods upon chemisorption onto the
gold surface exist primarily in an upright orientation with the
nonreducing end facing up as shown in Figure 6 a. A SAM of
CNC-SH rods in the upright orientation would be expected to
have lower rigidity compared to horizontally adsorbed rods and
hence a higher slope for teh ΔD−ΔF plot, as seen
experimentally in our plots. If the adsorbed CNC-SH rods
were indeed flexible as depicted in the schematic in Figure 6 a,
then a simple interesting implication would be the ability to
align the adsorbed crystals using a receding triple-phase
evaporation front (Figure 6b), similar to the well-known
technique used to create two-dimensional colloidal crystals
through convective self-assembly.46
Figure 7. AFM topography images of CNCs (a) and CNC-SHs (b)
adsorbed on a gold surface followed by convective flow from an
evaporating drop of water. The range of the height scale in both
images is −2 to +12 nm, and the image size is 2 μm × 2 μm.
drying. It is obvious from Figure 7 that a significantly greater
proportion of crystals are aligned in the case of the CNC-SHadsorbed surface compared to what is observed with pure
CNCs. The results of quantitative analysis of the alignment of
nanocrystals in the AFM images are presented in Figure 8.
Figure 8a shows the percentage vs angle plot for three different
AFM images corresponding to CNCs, and it can be observed
that the distribution of orientation is quite random as seen from
the nonexistence of distinct large peaks. By contrast, Figure 8b
2811
dx.doi.org/10.1021/bm400633r | Biomacromolecules 2013, 14, 2807−2813
Biomacromolecules
Article
on surfaces has been reported previously; for example, Yokota
et al. end-modified cellulose chains with semithiocarbazone,
thereby enabling the formation of a chemisorbed cellulose I
film on a gold surface.50 To our knowledge the formation of
CNC-based flexible hairy SAMs has not been reported, and
given the fact that CNC crystals have chemical versatility to
allow a modification conferring magnetic or electric susceptibility, it becomes realistic to design reversibly switchable
nanocilia with interesting applications. In the case of studies
related to PS particle adsorption on hairy surfaces, the XPS
results (Supporting Information, Figure S3) do not give a
complete picture of changes happening to eh CNC-SH SAM
upon exposure to PS particles; it may be suggested that the
smaller particles could be affecting the hairy surface more than
the larger particles, and this speculation needs to be backed by
further studies.
■
CONCLUSIONS
The reducing ends of CNCs were functionalized with thiol
groups using reductive amination reaction in an aqueous
medium. Environmentally benign reducing agents, including
sodium triacetoxyborohydride and 2-picoline−borane, were
found to be good alternatives to toxic sodium cyanoborohydride for the reductive amination reaction. The pH of the
reaction medium was found to be a critical factor in
determining the functionalization efficiency. The anisotropic
thiol modification at the reducing ends was characterized
qualitatively and quantitatively using TEM and XPS,
respectively. A QCM-D- and AFM-based adsorption study on
gold clearly showed that the thiolation of CNCs enabled
significantly higher adsorption relative to unmodified CNCs,
thus suggesting that the CNC-SHs chemisorb on the gold
surface. A detailed analysis of QCM-D ΔD−ΔF hinted at an
upright orientation of the chemisorbed CNC-SHs. It was also
demonstrated that the chemisorbed CNC-SH rods could be
easily aligned using convective shear of an evaporating drop of
water, thus further consolidating the QCM-D-based inference
that the chemisorbed thiolated rods form a hairy colloidal level
surface. Additionally, particle binding studies on a selfassembled layer of CNC-SHs using polystyrene particles of
varying sizes hinted that the colloidal scale hairy surface may
reject particle binding on a size-selective basis.
Figure 8. Quantitative orientation analysis of AFM topography images
of CNCs (a) and CNC-SHs (b) adsorbed on a gold surface followed
by convective flow from an evaporating drop of water. Each panel has
data corresponding to three different AFM images (blue, image 1; red,
image 2; green, image 3). Note that in all cases the curves are compiled
from AFM pictures taken from different sites, thus explaining why the
pronounced peaks are in different angular positions.
corresponding to CNC-SHs shows distinct peaks indicating
preferred alignments. That the positions of these peaks are at
different angles in the three different AFM images is due to the
fact that the receding water droplets upon drying lead only to
local alignment. Still the concept unambiguously shows the
capability of the adsorbed CNC-SH rods to realign as driven by
external conditions.
As AFM shows that CNC-SHs bind on gold surfaces, TEM
shows that there is thiolation only on one end, QCM-D reveals
that, in spite of the high adsorption on gold, CNC-SH
nanorods show a “loose” dissipative structure, suggesting the
capability to “swing”, and AFM shows that the adsorbed rods
can be realigned. We conclude that a colloidal level “hairy
surface” has been created, structurally allowing a route toward
artificial cilia.
The higher adsorption of CNC-SHs on gold surfaces as
observed in our studies shows for the first time in the case of
CNCs that the topochemical modification is capable of
significantly altering their adsorption behavior, thereby
resulting in a colloidal level hairy structure. This is a proof of
concept demonstrating the potential behind topochemical
modification to enable a good control over the supramolecular
self-assembly behavior of CNCs either in pure films or in
composite materials. Convective shear-driven alignment of
CNCs has been investigated in earlier studies, for instance, by
Hoeger et al.47 The uniqueness of the alignment technique
investigated in our study lies in the fact that the sparsely spaced
CNC-SHs being aligned were originally chemisorbed onto the
substrate and also the crystals were in an upright orientation. In
principle, the inter CNC-SH crystal distance could be
controlled by systematically changing the extent of electrostatic
interactions arising from the sulfate groups surrounding the
crystals. The upright-oriented CNC-SHs under aqueous
conditions are similar to the well-known carbon nanotube
forest48 and are capable of being manipulated by applying an
electric field.49 Self-assembly of end-modified cellulosic chains
■
ASSOCIATED CONTENT
S Supporting Information
*
QCM-D micrographs of various adsorption studies, relative
atomic percentages of polystyrene-adsorbed CNC-SH-functionalized surfaces, and ζ potentials of polystyrene particles. This
material is available free of charge via the Internet at http://
pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: lokanthan.arcot@aalto.fi (A.R.L.); olli.ikkala@aalto.fi
(I.O.). Phone: +358 505924209 (A.R.L.); +358 504100454
(I.O.).
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
We thank the Aalto School of Chemical Technology, the
Academy of Finland, the Finnish Funding Agency for
2812
dx.doi.org/10.1021/bm400633r | Biomacromolecules 2013, 14, 2807−2813
Biomacromolecules
Article
(31) Jackson, J.; Letchford, K.; Wasserman, B.; Ye, L.; Hamad, W.;
Burt, H. Int. J. Nanomed. 2011, 6, 321−330.
(32) Eyley, S.; Thielemans, W. Chem. Commun. 2011, 47, 4177−
4179.
(33) Zoppe, J. O.; Ö sterberg, M.; Venditti, R. A.; Laine, J.; Rojas, O.
J. Biomacromolecules 2011, 12, 2788−2796.
(34) Zoppe, J. O.; Venditti, R. A.; Rojas, O. J. J. Colloid Interface Sci.
2012, 369, 202−209.
(35) Majoinen, J.; Walther, A.; McKee, J. R.; Kontturi, E.; Aseyev, V.;
Malho, M. J.; Ruokolainen, J.; Ikkala, O. Biomacromolecules 2011, 12,
2997−3006.
(36) Velleste, R.; Teugjas, H.; Väljamäe, P. Cellulose 2010, 17, 125−
138.
(37) Koyama, M.; Helbert, W.; Imai, T.; Sugiyama, J.; Henrissat, B.
Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 9091−9095.
(38) Hieta, K.; Kuga, S.; Usuda, M. Biopolymers 1984, 23, 1807−
1810.
(39) Hermanson, G. T. Bioconjugate Techniques; Academic Press:
Amsterdam, Boston, 2008.
(40) Love, C. J.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.;
Whitesides, G. M. Chem. Rev. 2005, 105, 1103−1169.
(41) Littlejohn, D.; Chang, S. J. Electron Spectrosc. 1995, 71, 47−50.
(42) Castner, D. G.; Hinds, K.; Grainger, D. W. Langmuir 1996, 12,
5083−5086.
(43) Li, Y.; Ragauskas, A. J. Cellulose Nano Whiskers as Reinforcing
Filler in Polyurethanes. In Advances in Diverse Industrial Application of
Nanocomposites; Boreddy, R., Ed.; InTech: New York, 2011; pp 17-36.
(44) Dixon, M. C. J. Biomol. Tech. 2008, 19, 151−158.
(45) Höök, F.; Rodahl, M.; Brzezinski, P.; Kasemo, B. Langmuir
1998, 14, 729−734.
(46) Dushkin, C. D.; Lazarov, G. S.; Kotsev, S. N.; Yoshimura, H.;
Nagayama, K. Colloid Polym. Sci. 1999, 227, 914−930.
(47) Hoeger, I.; Rojas, O. J.; Efimenko, K.; Velev, O. D.; Kelley, S. S.
Soft Matter 2011, 7, 1957.
(48) Yucheng, L.; Yang, W.; Ren, Z. F. Adv. Phys. 2011, 60, 553.
(49) Habibi, Y.; Heim, T.; Douillard, R. J. Polym. Sci., Part B: Polym.
Phys. 2008, 46, 1430−1436.
(50) Yokota, S.; Kitaoka, T.; Sugiyama, J.; Wariishi, H. Adv. Mater.
2007, 19, 3368−3370.
Technology and Innovation, and ERC for funding this project.
We thank Prof. Ö sterberg Monika and Anna Olszewska for the
useful discussions related to QCM-D studies. Prof. Janne
Ruokolainen and the Aalto Nanomicroscopy Center are
acknowledged for use of the electron microscopy facilities.
■
REFERENCES
(1) Bhushan, B. Philos. Trans. R. Soc., A 2009, 367, 1443−1444.
(2) Satir, P.; Pedersen, L. B.; Christensen, S. T. J. Cell Sci. 2010, 123,
499−503.
(3) Button, B.; Cai, L.; Ehre, C.; Kesimer, M.; Hill, D. B.; Sheehan, J.
K.; Boucher, R. C.; Rubinstein, M. Science 2012, 337, 937−941.
(4) Evans, B. A.; Shields, A. R.; Lloyd, C. R.; Washburn, S.; Falvo, M.
R.; Superfine, R. Nano Lett. 2007, 7, 1428−1434.
(5) Shields, A. R.; Fiser, B. L.; Evans, B. A.; Falvo, M. R.; Washburn,
S.; Superfine, R. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 15670−15675.
(6) Timonen, J. V. I.; Johans, C.; Kontturi, K.; Walther, A.; Ikkala, O.;
Ras, R. H. A. ACS Appl. Mater. Interfaces 2010, 2, 2226−2230.
(7) Breidenich, J. L.; Wei, M. C.; Clatterbaugh, G. V.; Benkoski, J. J.;
Keng, P. Y.; Pyun, J. Soft Matter 2012, 8, 5334−5341.
(8) Den,Toonder, J. M. J.; Onck, P. R. Trends Biotechnol. 2013, 31,
85−91.
(9) Sidorenko, A.; Krupenkin, T.; Taylor, A.; Fratzl, P.; Aizenberg, J.
Science 2007, 315, 487−490.
(10) Pokroy, B.; Kang, S. H.; Mahadevan, L.; Aizenberg, J. Science
2009, 323, 237−240.
(11) Klemm, D.; Heublein, B.; Fink, H.; Bohn, A. Angew. Chem., Int.
Ed. 2005, 44, 3358−3393.
(12) Habibi, Y.; Lucia, L. A.; Rojas, O. J. Chem. Rev. 2010, 110,
3479−3500.
(13) Lee, S. W.; Kim, J. H.; Kim, J.; Kim, S. H. Chin. Sci. Bull. 2009,
54, 2703−2707.
(14) Levente, C.; Ingrid, H. C.; Rojas, O. J.; Ilona, P.; Pawlak, J. J.;
Perry, P. N. ACS Macro Lett. 2012, 1, 867−870.
(15) Kim, J.; Lee, H.; Kim, H. S. Int. J. Precis. Eng. Manuf. 2010, 11,
823−827.
(16) Jin, H.; Kettunen, M.; Laiho, A.; Pynnolnen, H.; Paltakari, J.;
Marmur, A.; Ikkala, O.; Ras, R. H. A. Langmuir 2011, 27, 1930−1934.
(17) Yun, G.; Kim, J.; Kim, J.; Kim, S. Sens. Actuators, A 2010, 164,
68−73.
(18) Kim, J. H.; Kang, K.; Yun, S.; Yang, S.; Lee, H. M.; Kim, J. H.;
Kim, J. Cellulose Electroactive Paper (EAPap): The Potential for a
Novel Electronic Material. MRS Proceedings; Materials Research
Society: Warrendale, PA, 2008; Vol. 1129, Paper V05-02.
(19) Wandowski, T.; Malinowski, P.; Ostachowicz, W. M. Smart
Mater. Struct. 2011, 20, 025002−025016.
(20) Moon, R., J.; Ashlie, M.; Nairn, J.; Simonsen, J.; Youngblood, J.
Chem. Soc. Rev. 2011, 40, 3941−3994.
(21) Klemm, D.; Kramer, F.; Moritz, S.; Lindström, T.; Ankerfors,
M.; Gray, D.; Dorris, A. Angew. Chem., Int. Ed. 2011, 50, 5438−5466.
(22) Iwamoto, S.; Kai, W.; Isogai, A.; Iwata, T. Biomacromolecules
2009, 10, 2571−2576.
(23) Nogi, M.; Iwamoto, S.; Nakagaito, A. N.; Yano, H. Adv. Mater.
2009, 21, 1595−1598.
(24) Saito, T.; Kuramae, R.; Wohlert, J.; Berglund, L. A.; Isogai, A.
Biomacromolecules 2013, 14, 248−253.
(25) Lin, N.; Huang, J.; Dufresne, A. Nanoscale 2012, 4, 3274−3294.
(26) Korhonen, J. T.; Kettunen, M.; Ras, R. H.A.; Ikkala, O. ACS
Appl. Mater. Interfaces 2011, 3, 1813−1816.
(27) Walther, A.; Timonen, J. V. I; Díez, I.; Laukkanen, A.; Ikkala, O.
Adv. Mater. 2011, 23, 2924−2928.
(28) Iwamoto, S.; Isogai, A.; Iwata, T. Biomacromolecules 2011, 12,
831−836.
(29) Eyley, S.; Shariki, S.; Dale, S. E. C.; Bending, S.; Marken, F.;
Thielemans, W. Langmuir 2012, 28, 6514−6519.
(30) Dong, S.; Roman, M. J. Am. Chem. Soc. 2007, 129, 13810−
13811.
2813
dx.doi.org/10.1021/bm400633r | Biomacromolecules 2013, 14, 2807−2813