Two-Dimensional Aggregation and Semidilute Ordering in Cellulose

Dec 18, 2015 - Kevin J. De France , Kevin G. Yager , Todd Hoare , and Emily D. Cranston ... Gilles Lenfant , Marie-Claude Heuzey , Theo G. M. van de V...
0 downloads 0 Views 7MB Size
Subscriber access provided by GAZI UNIV

Article

Two-Dimensional Aggregation and SemiDilute Ordering in Cellulose Nanocrystals Martin Uhlig, Andreas B. Fall, Stefan Wellert, Maren Lehmann, Sylvain Prévost, Lars Wagberg, Regine von Klitzing, and Gustav Nyström Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.5b04008 • Publication Date (Web): 18 Dec 2015 Downloaded from http://pubs.acs.org on December 24, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Two-Dimensional Aggregation and Semi-Dilute Ordering in Cellulose Nanocrystals Martin Uhlig,



Andreas Fall,

Lars Wågberg,





Stefan Wellert,



Regine v. Klitzing,

Maren Lehmann,





Sylvain Prévost,

and Gustav Nyström



∗, ‡

†Stranski-Laboratorium, Department of Chemistry, Technische Universität Berlin, Strasse

des 17.Juni 124 D-10623 Berlin, Germany ‡Department of Fibre and Polymer Technology, and Wallenberg Wood Science Centre, KTH

Royal Institute of Technology, School of Chemical Science and Engineering Teknikringen 56, 10044 Stockholm, Sweden ¶ESRF - The European Synchrotron, 71, avenue des Martyrs, 38000, Grenoble, France E-mail: [email protected]

Phone: +41 44 632 03 60. Fax: +41 44 632 16 03

Abstract The structural properties and aggregation behavior of carboxymethylated cellulose nanocrystals (CNC-COOH), was analyzed with small angle neutron scattering (SANS), transmission electron microscopy (TEM), atomic force microscopy (AFM) and dynamic light scattering (DLS) and compared to sulfuric acid hydrolyzed cellulose nanocrystals (CNC-SO3 H). The CNC-COOH system, prepared from single carboxymethylated cellulose nanobrils, was shown to laterally aggregate into 2D-stacks that were stable both in bulk solution and when adsorbed to surfaces. CNC-SO3 H also showed a 2D aggregate structure with similar cross sectional dimensions (a width to height ratio of 8) as 1

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

CNC-COOH, but a factor of 2 shorter length. SANS and DLS revealed a reversible ordering of the 2D aggregates under semi-dilute conditions and a structure peak was observed for both systems. This indicates an early stage of liquid crystalline arrangement of the crystal aggregates, at concentrations below those assessed using birefringence or polarized microscopy.

Introduction Nanocellulose represents, alongside DNA, virus and protein based materials, an emerging class of biological nanomaterials. 1,2 Over the last two decades careful structural investigations have revealed the cellulose crystal structure 3 as well as particle organization and properties of macroscopic materials. 4,5 So far, the mesoscopic length scale, bridging the internal cellulose chain organizations in the cellulose nanoparticles and the nal cellulose conguration in materials, has not been fully elucidated. This leaves room for further studies with the potential to create new material designs from self-assembly of cellulose nanoparticles on the mesoscale. Cellulose nanocrystals (CNCs) extracted from natural sources, e.g. wood are a promising renewable building block material. Their high stiness, high aspect ratio, dispersibility, biocompatibility and low cost 1,6 make them interesting for a lot of applications e.g. in nanocomposites, 7,8 food packaging 9,10 or biomedicine. 11,12 Thus, controlling the surface and size parameters and hence the colloidal behavior of CNC suspensions is desirable. This can be done by varying the raw material 13,14 or by changing the preparation route. 15,16 The most common type of CNC preparation is based on sulfuric acid hydrolysis of micrometer sized cellulose particles. 17,18 This preparation route, has been very successful to prepare CNC based liquid crystals and nanomaterials. It has, however, some drawbacks in that the amount and alkaline stability of the sulfate ester charged groups is limited. 16,19 An alternative preparation route is to combine hydrochloric/hydrobromic acid hydrolysis with oxidative carboxylation. 16,20 This leads to carboxyl instead of sulfate groups on the 2

ACS Paragon Plus Environment

Page 2 of 31

Page 3 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

CNC surface. The carboxyl groups are more stable under alkaline conditions than the sulfate ester groups, leading to a broader application range of CNC. 16,19 The number density of carboxyl groups can be tuned by the oxidation conditions and a higher charge density can be reached, compared to the sulfuric acid preparation, 21 thus improving the particle's dispersion capability and colloidal stability. The dierent reaction conditions, compared to sulfuric acid hydrolysis, also lead to higher aspect ratios of the carboxylated CNC in comparison to sulfated CNC. 16 Moreover, the carboxylic groups, more amenable to surface modication than the hydroxyl group typically used to modify sulfuric hydrolyzed CNC, can be easily chemically modied to further change the properties of the CNC surface. 7,19,20 Dierent techniques, such as atomic force microscopy (AFM), 14,19 transmission electron microscopy (TEM) 15,17,20 and small angle scattering (SAS) 13,14,18,2224 have been used to characterize the size parameters of CNC. Small angle scattering with neutrons (SANS) showed that CNC has a rectangular cross-sectional shape 14,22,24 and that, depending on the cellulose source, size parameters can dier signicantly. 13,14 Small angle scattering with X-Rays (SAXS) showed that also carboxylated cellulose nanobrils (CNF) can have a rectangular cross section. 23 These studies are mainly focused on the properties of the single cellulose particle. Structural studies of the interaction between particles and aggregation behavior exist 13,24,25 but are more rare. CNCs are known to self-assemble into chiral nematic liquid crystals above a threshold particle concentration, 17 forming locally well-ordered CNC arrangements. Physical templates 26 or external elds 18,27 can increase the scale of local organization from the micrometer towards the centimeter length scale. So far, however, the mechanism behind the isotropic to chiral nematic transition is not fully understood. 7,21 High resolution microscopy recently revealed detailed structural information about the right-handed chirality of individual cellulose brils. 28 While this observation gives important structural information, it is clear that more data of CNC arrangements, on the length scale between a single cellulose particle and that of the inter-particle spacing in the liquid crystal state, is needed to understand the

3

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

self-assembly of CNC. In the present report, we examined carboxymethylated CNC by using SANS, TEM, AFM and DLS to elucidate the CNC aggregation behavior on the microscale and the ordering of CNC aggregates into liquid crystal like arrangements on the mesoscale. By using a wellcharacterized system of single carboxymethylated CNFs as starting material for the CNC hydrolysis, we discovered an unexpected lateral organization of the CNC into 2D aggregates, that were stable in dispersion and resisted drying. With increasing concentration the 2D aggregates further showed a liquid crystal like ordering and dilution of the 2D aggregates demonstrated that this ordering was reversible. The results from the carboxymethylated CNCs were compared to the standard system of sulfuric acid hydrolyzed CNC. Both systems were found to show bulk structuring at concentrations in the lower range of the semi-dilute regime. This indicates the onset of a liquid crystalline arrangement at concentrations below those where bulk-ordering features such as birefringence or tactoids can be observed using polarized optical microscopy. The data presented here demonstrates that specic structuring occurs between single CNC particles, and the observation that the formed 2D aggregates order in liquid crystal like phases indicates that the 2D aggregation may be an intermediate structure in the formation of the CNC chiral nematic phase.

Experimental Section CNC-COOH-Preparation

Innventia AB kindly provided carboxymethylated cellulose nanobrils (CNF), where the pulp bers had been carboxymethylated according to an earlier described procedure. 29 In this procedure the bers were solvent exchanged to ethanol and impregnated with a solution of 10 grams of monochloroacetic acid in 500 mL isopropanol. They were then added in portions to a solution of 16.2 g of NaOH in 500 mL of methanol mixed with 2 L of isopropanol. The so prepared mixture was heated just below its boiling temperature and the reaction was 4

ACS Paragon Plus Environment

Page 4 of 31

Page 5 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

continued for 1h. The so prepared bers were then washed in 20 L of deionized water before all the carboxylic groups were converted to the sodium form. 29,30 For experimental details the readers are referred to these earlier publications. The charge content, i.e. the amount of acetic acid moieties grafted to the hydroxyls of the cellulose backbone was 0.49 mmol/g, determined by polyelectrolyte titration (PET) against Polydiallyldimethylammonium chloride (PDADMAC) (Stabino 2.0, Particle Metrix GmbH). The CNF was converted to CNC by hydrolysis, following procedures from Salajkova et al. . 19 Briey, the nanobrils were mixed with 2.5 M hydrochloric acid and heated to 105 ◦ C under reux of water vapor for 7 h. The reaction was quenched by a 10-fold dilution with deionized water. Thereafter, the resulting dispersion was cleaned by centrifugation and then extensively dialyzed (Spectrapor 1, 6000-8000 MWCO membranes, Spectrum Labs) against deionized water (5-7 days). The cleaned suspension was sonicated at 80% amplitude for 2 x 5 min (Sonics Vibracell, VCX 750, Sonic Materials, USA) and centrifuged afterwards at 4700 g for 60 min. The supernatant was collected, while the pellet was discarded. The charge content of the CNC was 0.36 mmol/g, determined by PET as above. No additional ions were added to the samples and a pH of 7 was measured after converting the samples to D 2 O by dialysis.

CNC-SO 3 H-Preparation

Commercial micro crystalline cellulose (MCC), prepared from wood pulp (Avicel PH-200, FMC Biopolymer), was hydrolyzed to cellulose nanocrystals (CNC) using sulfuric acid according to a previously developed protocol. 31 Briey, 175 mL of 64 wt% (10 M) sulfuric acid was added to 20 g of MCC and kept at 45 ◦ C for 1 h. Then 2 L of water were added to quench the reaction, followed by washing by centrifugation and extensive dialysis against deionized water, until the surrounding water had a pH of 6. The washed suspension was sonicated at 80% amplitude for 2 x 5 min (VCX 750) and centrifuged afterwards at 4700 g for 60 min. The supernatent was collected, while the pellet was discarded. The process generates electrostatically stabilized CNCs due to the introduction of sulfate esters. On 5

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the CNC surface a surface charge density of 0.35 mmol/g was measured by polyelectrolyte titration against PDADMAC (Stabino 2.0, Particle Metrix GmbH). No additional ions were added to the samples and the pH varied with the concentration of the samples in the range of 2 to 3.

Atomic Force Microscopy

Atomic force microscopy was performed in tapping mode using both, a Multimode 8 microscope (Bruker) and a Cypher (Asylum Reseach, Inc.). The samples were prepared on APTES ((3-Aminopropyl)triethoxysilane) modied mica by allowing 20 µL of 0.3 g/L dispersion to adsorb for 2 minutes before rinsing with Milli-Q water and blow drying within an air stream. The images were treated with the attening function in the Nanoscope Analysis software (Bruker) and the respective attening function in the Cypher software (Asylum Reseach, Inc.). The CNC height distributions were analyzed with FiberApp. 32

Transmission Electron Microscopy

Transmission electron microscopy (TEM) images were recorded with a JEOL JEM-2100 (JEOL gmbH, Eching, Germany). A drop of the sample solution (0.1 g/L) was placed on a carbon-coated TEM copper grid (Electron Microscopy Sciences, Germany) and the samples were stained with uranyl acetate. Excess liquid was blotted with a lter paper and the TEM grid was then air-dried at room temperature. The specimen was inserted into a sample holder (EM21010, JEOL gmbH, Eching, Germany) and transferred into the microscope. The TEM was operated at an acceleration voltage of 200 kV. All images were recorded digitally by a bottom-mounted 4k CMOS camera system (TemCam-F416, TVIPS, Gauting, Germany) and processed with a digital imaging processing system (EM-Menu 4.0, TVIPS, Gauting, Germany). The obtained TEM images were evaluated with ImageJ.

6

ACS Paragon Plus Environment

Page 6 of 31

Page 7 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Small Angle Neutron Scattering

Small angle neutron scattering (SANS) measurements were carried out at the V4 instrument 33 at the Helmholtz-Zentrum-Berlin (Germany). Samples were placed in quartz cuvettes (Hellma, Germany) of 2 mm neutron pathway and studied at three detector distances (1.35 m, 6 m and 15.75 m with collimation at 2, 8 and 16 m, respectively). For short and intermediate distances a wavelength of 4 Å was used, while for the long distance a wavelength of 8 Å was used. This conguration of the instrument provides a Q-range of 0.03 nm −1

≤ Q ≤ 6.4 nm−1 . The scattering of water was used to correct for the detector eciency and as a secondary standard bringing the data to absolute scale. To provide a bulk contrast of the samples, H 2 O was exchanged against D 2 O through a membrane (Slide-A-Lyzer dialysis cassettes 10k MWCO). The residual volume fraction of light water was determined by comparing the incoherent backgrounds of samples to those of D 2 O/H2 O mixtures with known composition. The resulting solvent scattering length density (SLD) was calculated accordingly. Data reduction and tting of the data was performed with the BerSANS 34 and SASt software, 35 respectively.

Dynamic Light Scattering

Dynamic light scattering was performed with a Zetasizer ZEN3600 (Malvern Instruments Ltd., U.K.), to investigate the onset of CNC cluster formation and increase in CNC cluster size upon increasing CNC concentration. The CNC concentration was increased by evaporating the solvent (water) from cuvettes sealed with perforated lids and placed in an oven at 45 ◦ C. For each CNC concentration 5 repeated 100 s measurements were performed. In order to obtain good statistics, the CNC concentration was increased in several cuvettes and these were analyzed in parallel. The data was analyzed using the cumulants method. Here a single exponential function is tted to the correlation function to obtain the mean CNC cluster size.

7

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Viscosity measurements

Shear rate ramp measurements where performed using the instrument MCR 301 (AntonPaar GmbH, Graz, Austria) with a cone and plate geometry (CP25). 200 µl of sample was deposited on the bottom plate and pre-sheared for 5 min by oscillation, 3 % amplitude at 1 Hz. The ramp was then performed by increasing the shear rate constantly from 0.1 to 10 s−1 within 10 min. Values from the shear rate vs. viscosity plot were taken from a linear region (0.3 - 0.6 s−1 ) and the zero shear viscosity was obtained by extrapolation.

Results Carboxymethylated CNC (CNC-COOH) was prepared from HCl hydrolysis of carboxymethylated cellulose nanobrils (CNF-COOH) 19 and compared to sulfuric acid hydrolyzed CNC (CNC-SO3 H). 31

TEM analysis of CNC

For both CNC-COOH and CNC-SO 3 H, TEM images were recorded (Fig. 1), where the rod like shape of both types of CNCs is clearly visible.

Figure 1: TEM images of a) CNC-COOH and b) CNC-SO 3 H. CNC-COOH crystals are longer and show kinks while CNC-SO 3 H crystals are shorter and kinks were not seen for this sample. For both types of CNC, aggregates of nanocrystals can be seen. The TEM images were used to extract the length of the nanocrystals. For 8

ACS Paragon Plus Environment

Page 8 of 31

Page 9 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

the length evaluation 250 single crystals of each type were evaluated. The resulting length distribution can be seen in Fig. 2.

Figure 2: Distribution of crystal lengths extracted from TEM images for a) CNC-COOH and b) CNC-SO3 H. By tting the extracted length data to a log normal distribution function ( µ is the location parameter and σ is the scale parameter, details given in the Supporting Information) the following t parameters were found: for CNC-COOH µ = 5.705 and σ = 0.42 and for CNCSO3 H µ = 5.003 and σ = 0.426 corresponding to average lengths of 328 nm for CNC-COOH and 163 nm for CNC-SO 3 H.

SANS from CNC

SANS was used to get ensemble average structural information about the individual crystal morphology (from semi-dilute dispersions) as well as their arrangements in liquid crystal assemblies (from concentrated dispersions). SANS was performed for the CNC-COOH samples in the concentration range 1 - 30 g/L and for CNC-SO 3 H samples in the concentration range 2 - 46 g/L. A two-step tting process was used to analyze the scattering data. At rst the data were tted for high

Q values (Q ≥ 0.7 nm−1 ) with the long cylinder model 36 (see

Supporting Information): 9

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 31

  ω(2QR) sin(QL) 2 2 F (Q, R, L) = (δηR L) · Si π2 (QL)Λ1 (QR) − − QL QL (QL)2 2

2

(1)

with

Si Π = (Si(x) + 2

cosx sinx x→∞ π + 2 ) → x x 2

(2)

and

ω(x) = where

8 (3J2 (X) + J0 (x) − 1) x2

(3)

Jn (X) is the cylindrical Bessel functions of order n. R is the radius of the rod,

L the length of the rod, δ the dierence between scattering length density (SLD) of the solvent and CNC and

Q the scattering vector. This model is applicable, as the length of the

cylinder is more than twice the radius of the cylinder ( L≥ 2R ). The extracted length from the TEM analysis was used in the calculations, given that the maximum observable dimension in these SANS data, limited by the lowest

Q

value, is around 200 nm (ca. 2 Π/Qmin ). 37

Thus, at least the CNC-COOH length would not be extractable. SLDs can be found in the Supporting Information. The tting resulted in a radius

R of 1.3 nm for CNC-COOH and

1.6 nm for CNC-SO 3 H. To describe the size distributions log-normal functions were used. For CNC-COOH µ = 0.15 and σ = 0.51 and for CNC-SO 3 H µ = 0.44 and σ = 0.22 were found. The resulting size distributions are shown in Fig. 3. For CNC-COOH the size distribution of the radius is broader than for CNC-SO 3 H. Moreover, this model was used to investigate the aggregation behavior of the nanocrystals. As mentioned, in the TEM images, aggregation of single nanocrystals to larger fractal structures was observed (Fig. 1) for both nanocrystals types. Thus, for high

Q values the structure

factor of fractal aggregates was added, 35 making it possible to extract the dimension of the aggregates. Here, a value of 2 would correspond to 2D-aggregation and a value of 3 would correspond to 3D-aggregation. The tting resulted in a dimension of 2.05 for CNC-COOH 10

ACS Paragon Plus Environment

Page 11 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 3: Distribution of crystal radii extracted from SANS curves for a) CNC-COOH and b) CNC-SO3 H. and 2.1 for CNC-SO 3 H (mainly 2D aggregation), which is also in accordance with the observations from the TEM images. To get a deeper insight into the shape of both CNC-COOH and CNC-SO 3 H, for low and intermediate

Q the parallelepiped model was used. 38 High Q values were not included in the

t, as due to the polydispersity the model fails to describe this range. Earlier reports have shown the applicability of the model for CNCs. 24,39,40



sin(Qx a/2) F (Q) = Qx a/2



sin(Qy b/2) Qy b/2



sin(Qz c/2) Qz c/2



The parallelepiped model uses the following parameters: the height a, the width

(4)

b and

the length c. Again, the extracted length from TEM was used and as a starting parameter for the aggregate height a,

2R from the rst model was used. In this analysis, the tted

value b corresponds not to the width of a single crystal, but the width of a crystal aggregate. In Fig. 4 the scattering data and corresponding ts for CNC-COOH and CNC-SO 3 H at 11

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 31

dierent CNC concentrations can be seen (additional curves for CNC-COOH can be found in the Supporting Information). In Table 1 the results from the ts are listed.

Figure 4: SANS curves recorded for a) CNC-COOH and b) CNC-SO 3 H. For better visibility the curves are incrementally shifted by a factor of ten. The lowest concentration curve is in absolute scale. As predicted the ts describe the recorded data very well for low and intermediate values, while for high

Q

Q values the t to the data is poor. All the used t parameters can

be found in the Supporting Information. Table 1: Extracted size parameters from the ts (length Nanocrystal type CNC-COOH CNC-SO3 H

aggregate height 2.7 nm 3 nm

c from TEM).

a aggregate width b length c [b /a ] aggregation number 21.7 nm 24.4 nm

328 nm 163 nm

8 8

AFM analysis of CNC

To investigate the 2D aggregates found in TEM and SANS further, AFM was performed (Fig. 5). In the obtained images, individual CNC as well as CNC aggregates were found. Based on image analysis 32 of 300 individual CNCs, the CNC height distributions of CNCCOOH and CNC-SO 3 H were extracted (Fig. 6), where all height coordinates (points) along the analyzed CNCs have been plotted. Similarly as for the length distributions, the height 12

ACS Paragon Plus Environment

Page 13 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 5: AFM images of a) CNC-COOH and b) CNC-SO 3 H.

Figure 6: AFM height distributions from individual CNCs of a) CNC-COOH b) CNC-SO 3 H tted with a log-normal distribution function. From the t an average CNC height of 3.2 nm was found for CNC-COOH and 4.0 nm for CNC-SO 3 H.

13

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 31

data can be well described using a log-normal distribution function. For CNC-COOH µ = 1.1 and σ = 0.35 and for CNC-SO 3 H µ = 1.33 and σ = 0.34 was found. This gives an average CNC height of 3.2 nm for CNC-COOH and 4.0 nm for CNC-SO 3 H. These values are in good agreement with the aggregate thicknesses obtained in the SANS analysis. In the AFM data, CNC aggregates as well as clusters of CNC aggregates are also visible (see Supporting Information). Height proles across these aggregates reveal a at conformation of the CNC in the aggregates. However, since the presence of artifacts related to AFM tip broadening 41 cannot be ruled out, we refrain from quantifying the exact lateral dimensions of these aggregates.

Interactions between CNC aggregates

For samples with a higher concentration a peak in the scattering data was visible (Fig. 4). For fully randomly oriented anisotropic CNC aggregates (as in the dilute scattering described above) no peak would be expected, since the peak results from a high probability of a given distance

d (equation 5) between the scatterers. d=

2π Qmax

(5)

As the CNC aggregate concentration is increased, the aggregates start to interact and a peak is developed at the scattering vector

Qmax , corresponding to the most probable inter-

aggregate distance d. The sharper this peak is, the better dened is the order between the aggregates and very sharp peaks are indicative of liquid-crystal-like ordering. So far, there is no model to describe the structure contribution of long rigid rods. Thus, the modied Caillé model 35 was used to describe the structure contribution at higher concentrations (for the ts in Fig. 4). The modied Caillé model is normally used to describe disorder due to uctuations in multilamellar structures. Thus, it can be only used to qualitatively extend the t to include the structure contribution. No quantitative information could be extracted

14

ACS Paragon Plus Environment

Page 15 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

from this contribution to the t. Therefore, the peak was evaluated by dividing the recorded curve (equation 6) by the form factor (up-scaled from lower concentrated curves which do not show a structure contribution).

I(Q) F (Q)

S(Q) =

(6)

Thus, only the structure factor remained (see Supporting Information), which was evaluated with a Lorentzian t (equation 7). 42

I(Q) = Here

(Q −

Il + I0 + (∆Q/22 )

Q2max )

(7)

I0 is the incoherent background, I1 is the incident beam intensity, Qmax is the peak

position, and ∆Q is the full width at half-maximum of the peak. An example for such a Lorentz t can be found in the Supporting Information. Fig. 7 shows the evolution of the inter-aggregate distance (calculated with equation 5) as a function of increasing CNC aggregate concentration and aggregate number density. CNC-SO3 H already shows structuring for low concentrations (above 2 g/L) while CNCCOOH starts to show structuring at slightly higher concentrations (above 4 g/L), see Fig. 4 and Fig. 7a. The intermediate distance between aggregates is smaller for CNC-COOH than for CNC-SO3 H. For higher concentrations the intermediate distance become more similar. Rescaling the concentrations to aggregate number densities (Fig. 7b) shifts the CNC-COOH data points relative to the CNC-SO 3 H data and demonstrates that the onset of structuring occur at the same aggregate number density for both samples. At the same time, the dierence in inter-aggregate distance becomes larger at similar aggregate number densities. The plotted inter-aggregate distances versus aggregate number density satises a power law (d ∝ nα ) with exponential decay factors α = -0.37 (CNC-COOH) and α = -0.4 (CNC-SO3 H). For comparison a line corresponding to α = -1/3 as expected from space lling arguments (see Supporting Information and the Discussion section) has been added.

15

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 7: Inter-aggregate distance d extracted from Qmax of the structure peaks plotted as a function of a) aggregate concentration and b) aggregate number density. The solid lines are ts to the data. The dotted lines represent all measured concentrations, thus concentrations which did not show a structure contribution could be seen.

16

ACS Paragon Plus Environment

Page 16 of 31

Page 17 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Dynamic Light Scattering from CNC-COOH

As a complement to the SANS measurements, dynamic light scattering was performed for the CNC-COOH samples in the concentration range 5 - 23 g/L. Fig. 8 shows the relative hydrodynamic diameter

Dh extracted from the measurements through the translational dif-

fusion coecient, Dt , (which is in turn obtained from the scattering autocorrelation function) using the Stoke-Einstein equation:

Dh =

kB · T 3π · η · Dt

(8)

Figure 8: Relative change in apparent hydrodynyamic diameter Dh and viscosity η as a function of concentration obtained from DLS measurements of CNC-COOH. and, for simplicity, assuming a spherical CNC aggregate shape. It can be seen that the apparent hydrodynamic diameter is relatively unchanged until a concentration of 10 g/L, where it starts to increase. As a control experiment, the viscosity of the CNC-COOH dispersion was also measured and the result was plotted as relative values in Fig. 8. These results show a small increase of the dispersion viscosity. As the CNC aggregate concentration is increased to 20 g/L, the relative viscosity increases by a factor of 1.54 compared to the viscosity at 5 g/L.

17

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Discussion Shape of CNC aggregates

We notice only minor dierences in the CNC-COOH and CNC-SO 3 H cross sections (see Table 1). The CNC-COOH nanocrystals are slightly thinner and their size distribution is broader compared to CNC-SO 3 H. The two types, however, strongly dier in nanocrystal length (as shown in Fig. 1 and 2). The observed dierences could possibly be traced back to the dierent cellulose sources used for the CNCs or the less aggressive reaction conditions used for CNC-COOH preparation compared to CNC-SO 3 H preparation. 16,43 While the chiral nematic ordering of the shorter CNC-SO 3 H crystals can be used for photonic materials, 5 the resulting increase in aspect ratio for CNC-COOH may be desirable for applications in biocomposites. 44,45 The CNC-COOH height and length parameters extracted in this study are comparable to earlier reported values, using similar reaction conditions and measured by AFM. 19 Carboxylated CNC and sulfuric acid hydrolyzed CNC were also recently compared with a focus on the resulting properties in nanopaper. 45 Here, an average length/diameter of 200.7/5.8 nm was found for carboxylated CNC and 163.0/15.6 nm for sulfuric acid hydrolyzed CNC. The longer length of carboxylated CNC was also observed in our work (see Table 1). We, however, observed quite similar cross sections for both CNC types (2.7 · 21.7 nm for CNC-COOH vs. 3 · 24.4 nm for CNC-SO 3 H, see Table 1) and a smaller width distribution for CNC-SO3 H than for CNC-COOH (see Fig. 2). While this presents a relative comparison between the two systems, it is important to note that in general for CNCs the absolute size parameters are hardly comparable for dierent cellulose sources, reaction conditions and measurement techniques. 46 With TEM, or low-resolution AFM images, aggregates of CNCs can also easily be mistaken for single CNCs. Here, we observe lateral CNC aggregation based on both sample averaging scattering data (CNCs in dispersion) and AFM analysis (CNCs adsorbed to a surface). The formed 18

ACS Paragon Plus Environment

Page 18 of 31

Page 19 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

aggregates are at and do not possess a detectable twist, as might be expected based on the observation of a right-handed twist of individual cellulose brils. 28 From the AFM images it is clear that there is furthermore a tendency for the CNC aggregates to arrange laterally when adsorbed on a surface (see Supporting Information). To minimize the risk that drying eects inuences the lateral organization, the CNCs were electrostatically immobilized on the surface (APTES-modied mica). The lateral arrangement indicates that the surface properties of the sides of the individual CNCs could be dierent. Cellulose has a crystal structure where the polymer chains are oriented in a at planar organization along the particles long direction. 3 This suggests that the hydroxyl groups, which are all oriented parallel to the crystalline cellulose planes, can create two sides that are more polar than the two sides of the particle that are oriented perpendicular to the crystalline planes. 47,48 Thus, it is possible, that the less polar sides of two neighboring particles orient towards each other, thereby exposing more of their hydrophilic sides to the surrounding water and minimizing their free energy. SAXS of CNF and SANS of CNC have shown that the width of the cellulose nanoparticles can be ∼4 times larger than their height. 23,24 This should be compared to the factor ∼8 larger width to height of the 2D CNC aggregates in the present study. Aggregation of single nanocrystals into larger aggregates has also been previously observed. 13,24 Two ways to reduce CNC aggregation can be to use more strongly charged cellulose or to use longer ultra sonication times, but this also has been shown to damage the nanocrystals. 15 In the case of CNC-COOH, we start from a well characterized system of carboxymethylated cellulose nanobrils (CNF-COOH), 29 with a mean height of 2.5 nm in good agreement to the height parameter of the CNC-COOH aggregates (see Table 1 and Supporting Information for an AFM image and crystal height distribution of this material) as well as previously reported values reported height values of 3.5-4 nm. 25,49 The known nanometer structure of the CNF-COOH starting material oers, unlike the CNC-SO3 H synthesis where the source material is several micrometers in all dimensions, a

19

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

direct link between the structure of the source material and the nal CNC-COOH structure. As a rst approximation the interaction between the individual CNCs can be described, using the DLVO theory, 50,51 as the sum of a repulsive electrostatic potential and an attractive van der Waals potential. During the HCl hydrolysis, the pH is drastically lowered eliminating the electrostatic repulsion between the brils. Without the electrostatic repulsion the brils are forced to close proximity by van der Waals attraction. Simultaneously, the acid hydrolyses the brils and thereby decreases their length. Since the van der Waals attraction is larger for parallel than for perpendicular arrangements of rods, 52,53 a lateral particle aggregation is energetically favorable. Following hydrolysis, the pH is again increased and the CNCs have regained their surface charge. Most likely, the mechanical energy (sonication) that is subsequently used to liberate the CNCs is not sucient to fully separate the aggregates into free CNCs, and rather aggregates of CNCs are dispersed. This explanation model could, combined with the previously mentioned crystal structural argument, oer a qualitative understanding of the observed lateral aggregation behavior.

Interactions between CNC aggregates

At the lowest CNC concentrations the SANS data reveals no structuring in the samples and the scattering is only based on the form factor of the CNC aggregates. As the CNC concentration increases, however, both CNC samples show structure peaks indicating an ordering (see Fig. 4). For CNC-SO 3 H structure formation starts at a lower crystal mass concentration. Furthermore, for CNC-COOH the distances at similar concentrations are smaller (Fig. 7). At the onset of ordering the CNC aggregates are at average distances of

∼ 90 nm (CNC-COOH) and ∼ 130 nm (CNC-SO3 H) from each other. Based on the total ionic concentration, estimated from the counterion concentrations, (1.4-11 mM for CNC-COOH and 0.7-14 mM for CNC-SO 3 H at low and high aggregate concentration respectively) we notice that both crystals are at distances farther away than their Debye lengths ( κ−1 ∼ 3-8 nm for CNC-COOH and κ−1 ∼ 3-12 nm for CNC-SO 3 H, see 20

ACS Paragon Plus Environment

Page 20 of 31

Page 21 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Supporting Information for calculation details). We therefore suggest that the electrostatic interaction between the CNC aggregates is not primarily controlling this structuring, and instead use steric arguments to explain the observed dierences in inter-aggregate distances. Both 2D aggregate types are in the semi-dilute concentration range, i.e. above the overlap concentration, and at average distances shorter than their length (see Supporting Information for details). They will therefore have a steric inuence on their surrounding aggregates and there will be an entropic driving force favoring a parallel orientation of the aggregates (minimizing the excluded volume). Based on the larger aggregate size of CNC-COOH, compared to CNC-SO 3 H, and by taking into account the concentration dierences at the onset of ordering, a smaller average inter-aggregate distance could therefore be expected. This corresponds well to the experimental observations. The eect is even more pronounced when the aggregate concentration is re-scaled to CNC aggregate number densities (Fig. 7b). From the larger aspect ratio, of CNC-COOH compared to CNC-SO 3 H, one would actually also expect the onset of structuring to occur at a lower CNC aggregate concentration for the CNC-COOH. This is not observed in the experiments. A possible reason could be the larger polydispersity observed for the CNC-COOH crystals (Fig. 2 and 3). A larger polydispersity decreases the driving force which assembles the CNC into ordered structures. 54 The interCNC aggregate distances reported here at higher CNC concentrations ( Fig. 7, above 20 g/L ) are also in good agreement with those previously reported for similar CNC concentrations. 40 So far previous reports in literature have only shown weak 18 or no structuring 22,24 at CNC concentrations in the range 7.8-13.7 g/L. While a direct comparison to previous experimental studies is dicult because of the many experimental dierences, we note that our results may be the rst report of structure formation in the low range of the semi-dilute regime for cellulose nanocrystals. The power law decay in inter-aggregate distance is slightly steeper for CNC-SO 3 H (α=0,4) than for CNC-COOH ( α=-0,37). In general a scaling of -1/3 is characteristic for particles using simple space lling arguments in three dimensions (see Supporting Information for

21

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 31

details). A deviation from this slope would indicate a clustering of the aggregates. For instance a slope of -1/2 is characteristic for entangled polyelectrolyte systems. 55 Both scalings are close to the theoretical value (see Fig. 7) indicating that we mainly have space lling packing of aggregates and not clustering. Finally, we used DLS to monitor the relative aggregate size, based on the apparent hydrodynamic diameter

Dh , as a function of increasing CNC aggregate concentration for the

CNC-COOH system (Fig. 8). From these measurements, we conclude that the aggregate size remain relatively unchanged in the concentration range 5 - 10 g/L. This suggests that the observed early structuring is a bulk structuring of the aggregates and not a result of the arrangement of the aggregates into clusters. Above 10 g/L

Dh increases, indicating formation

of clusters of the aggregates. The rheology of CNC-COOH was investigated, as a control experiment, and veried that the observed increase in

Dh was not an eect of an increase in

viscosity. The measurement showed only a factor 1.54 increase in viscosity compared to a 5-8 times observed increase in

Dh . This supports the conclusion that we do observe a relative

increase in aggregate cluster size for concentrations above 10 g/L. The concentration range of the onset of cluster formation from DLS also corresponds well to the more pronounced structure peak observed in the SANS data for similar concentrations (see Fig. 4, 8.5 g/L curve). The interpretation of a clustering of 2D aggregates is in apparent contrast to the observed scaling behavior of Fig. 7. One possible explanation for this could be that the clusters observed in the DLS are not tightly bound clusters, but small liquid crystal domains of 2D aggregates. This would seem plausible since at 10 g/L we are approaching the concentration range where liquid crystalline tactoids are routinely observed. 17,18,22,27,40 Dilution of the concentrated samples showed (see Supporting Information) that the original apparent hydrodynamic diameter could be recovered indicating that the formation of clusters is reversible. This supports the interpretation that we observe a bulk structuring of the 2D aggregates at low concentrations and above 10 g/L small liquid crystal like domains are formed that grow

22

ACS Paragon Plus Environment

Page 23 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

with size for increasing concentration. The liquid crystal domains are also reversible which would not be expected for tightly bound clusters. This explanation is in agreement with the previously discussed scaling behavior.

Conclusion SANS was used in combination with DLS, TEM and AFM to probe the structural properties of, carboxymethylated and sulfuric acid hydrolyzed, cellulose nanocrystals and their organization into ordered phases. The experiments showed a tendency of the CNCs to aggregate into 2D-stacks. The SANS data could be well described by a parallelepiped model (with the addition of the modied Caillé model to qualitatively describe the structure contribution). From this analysis a similar cross-section of the 2D aggregates was found for both CNC systems with a width to height aggregation number of 8. The length of the CNC-COOH aggregates was a factor 2 longer than those of the CNC-SO 3 H. Using SANS an ordering of both types of aggregates in the low range of the semidilute concentration regime was detected. This is interpreted as an early stage of liquid crystalline ordering at concentrations below those where macroscopic liquid crystalline CNC phases are observed. The inter-aggregate spacing extracted from the SANS data followed a power law of the number density of aggregates with exponents -0.37 and -0.4 for CNC-COOH and CNC-SO 3 H respectively. These exponents indicate that the packing of aggregates occurs mainly through space lling rather than tightly bound clustering. For the CNC-COOH aggregates, DLS revealed an increase in the apparent hydrodynamic diameter above a threshold concentration of 10 g/L. This increase in hydrodynamic diameter was reversible and indicates the development of small liquid crystal domains of 2D aggregates.

23

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 31

Acknowledgement The authors thank HZB for the allocation of neutron radiation beamtime. Martin Uhlig thanks the DFG for nancial support via Kl1165-10. Martin Uhlig and Stefan Wellert thankfully acknowledge the nancial support from HZB. This project has received funding from the European Union's Seventh Framework Programme for research, technological development and demonstration under the NMI3-II Grant number 283883. Furthermore, the authors want to thank Sebastian Schön and Dr. Jozef Adamcik for help with the AFM images. Andreas Fall, Gustav Nyström and Lars Wågberg acknowledge the nancial support from the Wallenberg Wood Science Centre at KTH.

Supporting Information Available Additional data regarding the SANS tting procedure, descriptions of the statistical analysis of the samples, additional AFM images, DLS data and calculations of ionic strength and overlap concentrations can be found in the Supporting Information. This material is available free of charge via the Internet at http://pubs.acs.org/ .

References (1) Klemm, D.; Kramer, F.; Moritz, S.; Lindström, T.; Ankerfors, M.; Gray, D.; Dorris, A. Nanocelluloses: A New Family of Nature-Based Materials.

Angew. Chem. Int. Edit.

2011, 50, 54385466. (2) Berglund, L. A.; Peijs, T. Cellulose Biocomposites - From Bulk Moldings to Nanostructured Systems.

MRS Bulletin 2010, 35, 201207.

(3) Nishiyama, Y.; Langan, P.; Chanzy, H. Crystal Structure and Hydrogen-Bonding Sys-

24

ACS Paragon Plus Environment

Page 25 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

tem in Cellulose Iβ from Synchrotron X-ray and Neutron Fiber Diraction.

J. Am.

Chem. Soc. 2002, 124, 90749082. (4) Saito, T.; Uematsu, T.; Kimura, S.; Enomae, T.; Isogai, A. Self-aligned integration of native cellulose nanobrils towards producing diverse bulk materials.

Soft Matter 2011,

7, 88048809. (5) Kelly, J. A.; Manchee, C. P. K.; Cheng, S.; Ahn, J. M.; Shopsowitz, K. E.; Hamad, W. Y.; MacLachlan, M. J. Evaluation of form birefringence in chiral nematic mesoporous materials.

J. Mater. Chem. C. 2014, 2, 50935097.

(6) Siró, I.; Plackett, D. Microbrillated cellulose and new nanocomposite materials: a review.

Cellulose 2010, 17, 459494.

(7) Habibi, Y.; Lucia, L. A.; Rojas, O. J. Cellulose Nanocrystals: Chemistry, Self-Assembly, and Applications.

Chem. Rev. 2010, 110, 3479500.

(8) Moon, R. J.; Martini, A.; Nairn, J.; Simonsen, J.; Youngblood, J. Cellulose nanomaterials review: structure, properties and nanocomposites.

Chem. Soc. Rev. 2011, 40,

39413994. (9) Mihindukulasuriya, S.; Lim, L.-T. Nanotechnology development in food packaging: A review.

Trends in Food Science & Technology 2014, 40, 149167.

(10) Ahmadzadeh, S.; Nasirpour, A.; Keramat, J.; Hamdami, N.; Behzad, T.; Desobry, S. Nanoporous cellulose nanocomposite foams as high insulated food packaging materials.

Colloids Surface A 2015, 468, 201210. (11) Plackett, D. V.; Letchford, K.; Jackson, J. K.; Burt, H. M.; Jackson, J.; Burt, H. A review of nanocellulose as a novel vehicle for drug delivery.

2014, 29, 105118.

25

ACS Paragon Plus Environment

Nord. Pulp. Pap. Res. J.

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 31

(12) Jor, M.; Foster, E. J. Recent advances in nanocellulose for biomedical applications.

J.

Appl. Polym. Sci. 2015, 132, 41719. (13) Elazzouzi-Hafraoui, S.; Nishiyama, Y.; Putaux, J.-L.; Heux, L.; Dubreuil, F.; Rochas, C. The Shape and Size Distribution of Crystalline Nanoparticles Prepared by Acid Hydrolysis of Native Cellulose.

Biomacromolecules 2008, 9, 5765.

(14) Khandelwal, M.; Windle, A. H. Small angle X-ray study of cellulose macromolecules produced by tunicates and bacteria.

Int. J. Biol. Marcomol. 2014, 68, 215217.

(15) Dong, X.; Revol, J.; Gray, D. Eect of microcrystallite preparation conditions on the formation of colloid crystals of cellulose.

Cellulose 1998, 2, 1932.

(16) Habibi, Y.; Chanzy, H.; Vignon, M. R. TEMPO-mediated surface oxidation of cellulose whiskers.

Cellulose 2006, 13, 679687.

(17) Revol, J. F.; Bradford, H.; Giasson, J.; Marchessault, R. H.; Gray, D. G. Helicoidal selfordering of cellulose microbrils in aqueous suspension.

Int. J. Biol. Marcomol. 1992,

14, 170172. (18) Orts, W.; Godbout, L.; Marchessault, R. H.; Revol, J. F. Enhanced ordering of liquid crystalline suspensions of cellulose microbrils: A small angle neutron scattering study.

Macromolecules 1998, 31, 57175725. (19) Salajková, M.; Berglund, L. A.; Zhou, Q. Hydrophobic cellulose nanocrystals modied with quaternary ammonium salts.

J. Mater. Chem. 2012, 22, 1979819805.

(20) Araki, J.; Wada, M.; Kuga, S. Steric Stabilization of a Cellulose Microcrystal Suspension by Poly(ethylene glycol) Grafting.

Langmuir 2001, 17, 2127.

(21) Lagerwall, J. P. F.; Schütz, C.; Salajkova, M.; Noh, J.; Hyun Park, J.; Scalia, G.; Bergström, L. Cellulose nanocrystal-based materials: from liquid crystal self-assembly and glass formation to multifunctional thin lms. 26

NPG Asia Materials 2014, 6, e80.

ACS Paragon Plus Environment

Page 27 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

(22) Terech, P.; Chazeau, L.; Cavaille, J. Y. A Small-Angle Scattering Study of Cellulose Whiskers in Aqueous Suspensions.

Macromolecules 1999, 32, 18721875.

(23) Su, Y.; Burger, C.; Hsiao, B. S.; Chu, B. Characterization of TEMPO-oxidized cellulose nanobers in aqueous suspension by small-angle X-ray scattering.

J. Appl. Crystallogr.

2014, 47, 788798. (24) Cherhal, F.; Cousin, F.; Capron, I. Inuence of Charge Density and Ionic Strength on the Aggregation Process of Cellulose Nanocrystals in Aqueous Suspension, as Revealed by Small-Angle Neutron Scattering.

Langmuir 2015, 31, 55965602.

(25) Fall, A. B.; Lindström, S. B.; Sundman, O.; Ödberg, L.; Wågberg, L. Colloidal stability of aqueous nanobrillated cellulose dispersions.

Langmuir 2011, 27, 1133211338.

(26) Nyström, G.; Fall, A. B.; Carlsson, L.; Wågberg, L. Aligned cellulose nanocrystals and directed nanoscale deposition of colloidal spheres.

Cellulose 2014, 21, 15911599.

(27) Park, J. H.; Noh, J.; Schütz, C.; Salazar-Alvarez, G.; Scalia, G.; Bergström, L.; Lagerwall, J. P. F. Macroscopic control of helix orientation in lms dried from cholesteric liquid-crystalline cellulose nanocrystal suspensions.

ChemPhysChem 2014, 15, 1477

84. (28) Usov, I.; Nyström, G.; Adamcik, J.; Handschin, S.; Schütz, C.; Fall, A.; Bergström, L.; Mezzenga, R. Understanding nanocellulose chirality and structure-properties relationship at the single bril level.

Nature Communications 2015, 6, 7564.

(29) Wågberg, L.; Decher, G.; Norgren, M.; Lindström, T.; Ankerfors, M.; Axnäs, K. The build-up of polyelectrolyte multilayers of microbrillated cellulose and cationic polyelectrolytes.

Langmuir 2008, 24, 784795.

(30) Wågberg, L.; Winter, L.; Ödberg, L.; Lindstrom, T. On the Charge Stoichiometry upon

27

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 31

Colloid Surface 1987,

Adsorption of a Cationic Polyelectrolyte on Cellulosic Materials.

27, 163173. (31) Beck-Candanedo, S.; Roman, M.; Gray, D. Eect of conditions on the properties behavior of wood cellulose nanocrystals suspensions.

Biomacromolecules 2005, 6, 10481054.

(32) Usov, I.; Mezzenga, R. FiberApp: An Open-Source Software for Tracking and Analyzing Polymers, Filaments, Biomacromolecules, and Fibrous Objects.

Macromolecules 2015,

48, 12691280. (33) Keiderling, U.; Wiedenmann, A. New SANS instrument at the BER II reactor in Berlin, Germany.

Physica B 1995, 213-214, 895897.

(34) Keiderling, U. The new 'BerSANS-PC' software for reduction and treatment of small angle neutron scattering data.

Appl. Phys. A 2002, 74, s1455s1457.

(35) Kohlbrecher, J.; Bressler, I. http://sast.ingobressler.net/. (36) Porod, G. Die Abhängigkeit der Röntgenkleinwinkelstreuung von Form und Grösse der Kolloiden Teilchen in verdünnten Systemen. (37) Glatter, O.; Kratky, O. (38) Guinier, A.

Acta Physica Austriaca 1948, 2, 255292.

Small Angle X-ray Scattering ; 1982.

X-ray Diraction in Crystals, Imperfect Crystals, and Amorphous Bodies ;

1963; pp 3334. (39) Bonini, C.; Heux, L.; Cavaillé, J.; Lindner, P. Rodlike cellulose whiskers coated with surfactant: A small-angle neutron scattering characterization.

Langmuir 2002, 18,

33113314. (40) Schütz, C.; Agthe, M.; Fall, A. B.; Gordeyeva, K.; Guccini, V.; Salajková, M.; Plivelic, T. S.; Lagerwall, J. P. F.; Salazar-Alvarez, G.; Bergström, L. Rod Packing in Chiral Nematic Cellulose Nanocrystal Dispersions Studied by Small-Angle X-ray Scattering and Laser Diraction.

Langmuir 2015, 31, 65076513. 28

ACS Paragon Plus Environment

Page 29 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

(41) Wilson, D. L.; Kump, K. S.; Eppell, S. J.; Marchant, R. E. Morphological Restoration of Atomic Force Microscopy Images.

Langmuir 1995, 11, 265272.

(42) Qu, D.; Pedersen, J. S.; Garnier, S.; Laschewsky, A.; Möhwald, H.; von Klitzing, R. Eect of Polymer Charge and Geometrical Connement on Ion Distribution and the Structuring in Semidilute Polyelectrolyte Solutions: Comparison between AFM and SAXS.

Macromolecules 2006, 39, 73647371.

(43) Ren, S.; Sun, X.; Lei, T.; Wu, Q. The Eect of Chemical and High-Pressure Homogenization Treatment Conditions on the Morphology of Cellulose Nanoparticles.

Journal

of Nanomaterials 2014, 2014, ID 582913. (44) Yu, H. Y.; Qin, Z. Y.; Liu, L.; Yang, X. G.; Zhou, Y.; Yao, J. M. Comparison of the reinforcing eects for cellulose nanocrystals obtained by sulfuric and hydrochloric acid hydrolysis on the mechanical and thermal properties of bacterial polyester.

Compos.

Sci. Technol. 2013, 87, 2228. (45) Sun, X.; Wu, Q.; Ren, S.; Lei, T. Comparison of highly transparent all-cellulose nanopaper prepared using sulfuric acid and TEMPO-mediated oxidation methods.

Cellulose

2015, 22, 11231133. (46) Jonoobi, M.; Oladi, R.; Davoudpour, Y.; Oksman, K.; Dufresne, A.; Hamzeh, Y.; Davoodi, R. Dierent preparation methods and properties of nanostructured cellulose from various natural resources and residues: a review. (47) Jarvis, M. Cellulose stacks up.

Cellulose 2015, 22, 935969.

Nature 2003, 426, 611612.

(48) Li, Q.; Renneckar, S. Supramolecular Structure Characterization of Molecularly Thin Cellulose I Nanoparticles.

Biomacromolecules 2011, 12, 650659.

(49) Fall, A. B.; Lindström, S. B.; Sprakel, J.; Wågberg, L. A physical cross-linking process

29

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 31

of cellulose nanobril gels with shear-controlled bril orientation.

Soft Matter 2013, 9,

18521863. (50) Derjaguin, B.; Landau, L. Theory of the Stability of Strongly Charged Lyophobic Sols and of the Adhesion of Strongly Charged Particles in Solutions of Electrolytes.

Acta

Physicochim. URSS 1941, 14, 633662. (51) Verwey, E. J. W.; Overbeek, J. T. G.

Theory of the Stability of Lyophobic Colloids ;

Elsevier Publishing Co., 1948. (52) Israelachvili, J. N.

Intermolecular and Surface Forces , 3rd ed.; 2011.

(53) Maeda, H.; Maeda, Y. Orientation-Dependent London-van der Waals Interaction Energy between Macroscopic Bodies.

Langmuir 2015, 31, 72517263.

(54) Odijk, T. Theory of lyotropic polymer liquid crystals.

Macromolecules 1986, 19, 2313

2329. (55) Üzüm, C.; Christau, S.; von Klitzing, R. Structuring of Polyelectrolyte (NaPSS) Solutions in Bulk and under Connement as a Function of Concentration and Molecular Weight.

Macromolecules 2011, 44, 77827791.

30

ACS Paragon Plus Environment

Page 31 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Graphical TOC Entry

31

ACS Paragon Plus Environment