Mechanical Performance of Macrofibers of Cellulose and Chitin

Jun 20, 2014 - (11, 46) Iwamoto et al. showed that wet-spinning at high rates leads to a partial orientation of NFC nanofibrils in wet spun macrofiber...
1 downloads 13 Views 669KB Size
Download PDF
Article pubs.acs.org/Biomac

Mechanical Performance of Macrofibers of Cellulose and Chitin Nanofibrils Aligned by Wet-Stretching: A Critical Comparison Jose Guillermo Torres-Rendon, Felix H. Schacher,† Shinsuke Ifuku,‡ and Andreas Walther* DWI − Leibniz-Institute for Interactive Materials, Forckenbeckstr. 50, D-52074 Aachen, Germany † Laboratory of Organic and Macromolecular Chemistry (IOMC), Jena Center for Soft Matter (JCSM), Friedrich-Schiller-University Jena, Lessingstr. 8, D-07743 Jena, Germany ‡ Graduate School of Engineering, Tottori University, 101-4 Koyama-cho Minami, Tottori, 680-8502, Japan ABSTRACT: Renewable nanofibrillated cellulose (NFC) and nanofibrillated chitin (NFCh) are attractive fibrillar bionanoparticles due to their remarkable properties such as outstanding mechanical stiffness and strength, thermostability, barrier properties, and also for their global availability from renewable resources and food waste. One major bottleneck to maximize the mechanical properties of materials based on these bionanoparticles (e.g., nanopapers and macroscale fibers) is to find pathways to control their direction of alignment and understand how preferred alignment correlates with macroscale properties. Herein, we will demonstrate how strain-rate controlled wet-stretching of rehydrated macroscale fibers composed of nanofibrillated chitin and cellulose (NFCh, NFC) induces a high degree of orientation and how the degree of alignment scales with macroscale mechanical stiffness. We find similar degrees of alignment in both types of nanofibril-based macrofibers, yet substantially different macroscale stiffness, with the NFC-based fibers (ENFC = 33 GPa) outperforming the NFCh-based ones (ENFCh = 12 GPa) considerably. These differences can be correlated to the mechanical properties of the underlying cellulose I and α-chitin crystals and the degree of crystallinity of the nanofibrils, which both govern the stiffness of an individual nanofibril. Our study likely demonstrates the maximum performance in terms of stiffness of materials prepared by NFC and NFCh and reveals a critical difference in the performance of both classes of bionanoparticles.

■

INTRODUCTION Highly crystalline polysaccharide bionanoparticles based on cellulose, chitin and starch are emerging as renewable components for future sustainable high-performance mechanical and functional materials. They can be distinguished into short 1D chitin1−3 or cellulose nanocrystals (CNC),4,5 1D chitin6−9 and cellulose nanofibrils,10−13 and 2D starch nanoplatelets.14 Nanofibrillated chitin and cellulose (NFCh, NFC) are typically isolated by chemical or enzymatic pretreatment in combination with mechanical homogenization based on globally abundant resources and partly even waste products of the food industry.15−19 While the cellulose-based bionanoparticles have been subject of intense studies, comparably little is still known about the material properties found in, or able to create with chitin- or starch-based analogues.20,21 In general, the nanoscale dimensions, high stiffness and also crystalline character of these bionanoparticles have been optimized in biological selection for best fracture properties, as, for example, recently shown for cellulose nanocrystals.22 Hence, they need to be considered as superior entities in terms of mechanical properties, compared to recrystallized biomaterials, which mostly offer little control over crystal dimensions.23−27 One bottleneck in fully realizing the potential of these bionanoparticles in contrast to established processing based on molecularly dissolved cellulose or chitin is to master © XXXX American Chemical Society

their water-borne processing schemes and, similar as in natural biocomposites, find ways to align these anisotropic bionanoparticles inside a final material to maximize and control directional mechanical and functional properties. Nanopapers based on nano/microfibrillated cellulose are possibly the most thoroughly studied material shape and seminal work has demonstrated the influence of porosity28−33 and mesoscale alignments of the nanofibrils,31,34−36 avenues to alter mechanical property profiles using tailored polymers for nanostructured and bioinspired composites12,37−40 or the addition of clay or graphene,13,41−45 and how fracture proceeds depending on the humidity.10 One of the bottlenecks toward larger scale applications of nanopapers is still the rather time-consuming filtration process needed to dehydrate the aqueous dispersion, in particular, when one aims at best mechanical properties that are only obtained from highly stabilized nanofibrillar dispersions without the use of flocculants.10,33 Wet-spinning is a promising alternative for a fast and efficient dehydration of nanofibrillar dispersion gels of either NFC or NFCh rendering fibers that can find applications in highperformance, next generation biocomposites. This has been Received: April 15, 2014 Revised: June 18, 2014

A

dx.doi.org/10.1021/bm500566m | Biomacromolecules XXXX, XXX, XXX−XXX

Biomacromolecules

Article

polymer chains typically used for classical wet-spinning, are able to reorient. This requires a fundamental understanding on the reorientation kinetics of nanofibrils. Hence, we herein critically compare how macrofibers wetspun from nanofibrillated chitin and cellulose under standardized extrusion conditions of rather long nanofibrils in the micrometer regime perform in terms of mechanical performance as a function of the degree of alignment imposed during wet-stretching of as-prepared fibers. We correlate mechanical data with 2D X-ray diffraction and scanning electron microscopy to quantitatively understand the degree of orientation and deduce the difference between both highly crystalline bionanoparticles. The wet-stretching procedure mimics to some extent the crucial alignment step typically performed in real-life wet-spinning of materials, and its understanding is the next key step toward a continuous spinning process.

mastered in discontinuous processes by extrusion of NFC and NFCh gels through a syringe needle and subsequent coagulation in a solvent. Investigations by Walther et al. and Iwamoto et al. demonstrated the principal feasibility of the approach and reported mechanical properties exceeding those of typical, nonoriented nanopapers.11,46 Iwamoto et al. showed that wet-spinning at high rates leads to a partial orientation of NFC nanofibrils in wet spun macrofibers, when very small nanofibrils in the range of 200−500 nm were used.46 We earlier demonstrated that NFCh can be similarly spun into comparably extensible and tough fibers with good mechanical properties with about 12% in strain and high work-of-fracture of 10 MJ m−3, albeit with significantly lower stiffness.6 Beyond mechanical properties, we and others47 showed how to add catalytic and magnetic properties, achieve transparency and conductivity, and find ways toward stability in wet conditions.6,11 One of the starting points for the present investigations is the observation of rather significant differences in the mechanical performance of macrofibers based on NFC and NFCh, as for instance visible in the fact that the first reported NFC-based macrofibers demonstrate an order of magnitude greater stiffness as compared to NFCh analogues (ENFC = 22.5 GPa vs ENFCh = 3.0 GPa).6,11 A similar behavior is found for nanopapers based on both materials.48−50 This is somewhat surprising, as both materials are constituted of highly crystalline bionanoparticles. This difference may, in principal, relate to unlike stiffness of the underlying crystalline nanomaterial (cellulose I vs α-chitin), differences in crystallinity of the nanofibrils (containing amorphous and crystalline parts), unlike porosities in the resulting materials and different interfibrillar interactions. Ogawa et al. reported the stiffness of the α-chitin crystal structure from the tendons of a snow crab using a synchrotron X-ray diffraction analysis to around 60 GPa,51 which is slightly higher than the value reported by Nishino et al. (41 GPa) for chitin fibers from crab shells.52 Both values are considerably lower than the reported cellulose I crystalline structure modulus (E = 138 GPa).53 The higher stiffness of cellulose Iβ is attributed to the smaller cross sectional area of a single molecule lattice of cellulose Iβ (31.7 Å2) compared to α-chitin (44.9 Å2).51,54−56 Additionally, the ratio of the force needed to extend a single molecule of α-chitin with respect to cellulose Iβ is 1:1.6. This is due to the shorter fiber repeat of α-chitin, which leads to smaller bond angles of the glucosidic linkage and consequently makes the expansion along the fiber axis easier in α-chitin.55,56 Aside the mechanical properties of the crystalline material, it is important to consider that the true stiffness of the actual nanofibril on the next length scale depends on its degree of crystallinity, which differs depending on source and preparation.57−59 Typical degrees of crystallinity for NFCh are above 80%.15,60 In case of NFC, the values differ widely depending on the starting material and preparation conditions and the type of measurement (e.g., XRD, 13C CPMAS), as well as method of peak evaluation in XRD and sample preparation. Values are reported ranging from 5−85%.26,57−59,61−65 These differences in macroscale and nanoscale mechanical properties motivate to better understand how interfibrillar interactions and degree of alignment influence macroscale properties. Additionally, to establish a desirable continuous wet-spinning process, the ability for wet-stretching of the fibers needs to be understood to proceed to maximum alignment. This involves considerable subtleness, as the wet-stretching occurs on time scales in which the nanofibrils, that are larger than ordinary

■

EXPERIMENTAL SECTION

Preparation of Nanofibrillated Cellulose Suspension. A TEMPO-mediated oxidation of a softwood sulfite pulp was performed under alkaline conditions (pH = 10.5) for 30 min using a TEMPO/ NaClO/NaBr system according to Isogai et al.17 After oxidation, the underlying cellulose has an apparent viscosity average degree of polymerization (DPv) of 376 using copper ethylendiamine as solvent and the content of carboxyl groups is 0.54 mmol/g. Recent investigations showed that (i) elimination of remaining aldehyde functions prior viscosity measurement66 or (ii) oxidation of aldehyde functions and their methylation followed by subsequent light scattering combined with size exclusion chromatography can increase the accuracy of molecular weight determination67 of partly oxidized celluloses and typically yields higher degrees of polymerization. The NFC suspension was prepared by mechanical homogenization of the oxidized pulp fibers after extensive washing, dilution to 1.3 wt % and adjustment of the pH to 8.5 with NaOH. A high pressure microfluidizer MRT model CR5 was used for homogenization by applying three passes at 500, 1400, and 1400 bar. The final pH of the dispersion was readjusted to pH = 8.5 by addition of small amounts of NaOH, yielding a 1.3 wt % suspension. We used X-ray diffraction to determine the apparent degree of crystallinity, Xapp,NFC, of the NFC after freeze-drying (to prevent a preferential orientation). Deconvolution of the obtained diffractogram to separate amorphous and crystalline contributions yields a Xapp,NFC of 69%, being in range with previous XRD characterization.59,65,68,69 Prior to extrusion, the dispersion was diluted with deionized water to 1 wt % and mixed using shaking. Preparation of Nanofibrillated Chitin Suspension. A 1.3 wt % suspension in water of chitin nanofibrils with deacetylated surface was prepared according to the procedure reported by Ifuku et al.15,70 First, chitin powder was suspended in an aqueous NaOH bath and refluxed for 6 h. After partial deacetylation, the product was washed by centrifugation several times. The nanofibrils were obtained using a stone grinder model MKCA6−3 (Masuko Sangyo Co., Ltd.) at a speed rotation of 1500 rpm and a gap height of −1.5 from the zero position (meaning a 0.15 mm shift). The apparent degree of crystallinity, as determined by XRD, is close to 84%,60,70 and the degree of deacetylation is 10.6 mol %.6 Following a simple geometrical model in our previous publication,6 this allows to conclude that 50% of the surface groups are in fact hydrolyzed (exp. section original publication). Acetic acid was added to facilitate disintegration during homogenization. The final suspension (pH = 4) was diluted to 1 wt % with water and homogenized with shaking. Atomic Force Microscopy (AFM). Atomic force micrographs of NFC and NFCh nanofibrils were obtained in tapping mode using a NanoScope V (Digital Instruments Veeco Instruments Santa Barbara, CA) of dilute samples in water (0.005 wt %) after deposition onto freshly cleaved mica. B

dx.doi.org/10.1021/bm500566m | Biomacromolecules XXXX, XXX, XXX−XXX

Biomacromolecules

Article

Scheme 1. Overview of the Wet-Spinning Process and Characterization of Starting Materialsa

a

(a) Wet-spinning set-up using controlled extrusion with a syringe pump into a THF coagulation bath using NFC and NFCh dispersions, containing surface modifications as depicted in (b) and (c). (d−g) AFM height images for NFC (d) and NFCh (g), as well as their corresponding image analysis concerning the average height of 200 nanofibrils for each class of nanoparticle (e = NFC; f = NFCh). Fiber Preparation. Following Walther et al. and Iwamoto et al.,6,11,46 NFC and NFCh macrofibers were prepared by wet extrusion, coagulation, and drying at room temperature. An amount of NFC or NFCh suspension (diluted to 1 wt % with deionized water) was centrifuged to remove entrapped air and then spun into a coagulation bath of THF from a needle with a diameter of 2 mm at an extrusion rate of 126 mL/min. After extrusion, the macrofibers were taken out from the coagulation bath, dried at room conditions, and stored until testing. Field-Emission Scanning Electron Microscopy (FE-SEM). Cross sections of NFC and NFCh macrofibers before and after stretching were observed by SEM using a Hitachi S4800. All fibers were fractured by hand and sputtered with Au/Pd for 15 s. Mechanical Properties. Tensile tests were performed on stretched and unstretched macrofibers in a DEBEN minitester with a 20 N load cell and a strain rate of 1 mm/min at a gauge length of 1 cm. Cross section areas were calculated by relating the weight, length and a density of 1.5 g/mL. All samples were preconditioned in a chamber at 55%RH for at least 24 h and measured at the same conditions at room temperature. Wet Stretching. We built a wet-stretching device based on aluminum profiles and a computer-controlled linear actuator (Haydon Kerk stepper motor 21000 series size 8) able to perform straincontrolled stretching within solvents (see Scheme 2). NFC and NFCh macrofibers with gauge lengths of 10 cm were clamped in the device and submerged in water. After relaxation, the macrofibers were subjected to a two-step stretching routine. The first step consisted of applying a strain velocity of 5 mm/min until 10% of elongation. After this point, a strain velocity of 0.5 mm/min was utilized until the desired final length is reached. Wide Angle X-ray Scattering (WAXS). Wide angle X-ray scattering measurements were performed on a Bruker AXS Nanostar-U instrument equipped with a microfocus X-ray source (Incoatec ImSCu E025) operating at λ = 1.54 Å. A pinhole setup with 750, 400, and 1000 mm (from source to sample) was used and the sample-todetector distance was 6 cm. The scattering patterns were corrected for the beam stop and background (Scotch tape) prior to evaluations. Intensity distribution profiles in the azimuthal angle (ϕ) were used to

Scheme 2. Schematic Drawing for the Computer-Controlled Wet-Stretching Device into Which a Fiber Can Be Clamped and Immersed into a Liquid and Stretched Using Controlled Strain Rates; The Right Hand Side Shows the Various States during Wet-Stretching

calculate the orientation index (π) and the order parameter (S) according to the equations: π=

180° − fwhm 180°

S=

3 1 cos2 ϕ⟩ − 2 2

(I) (II)

with π /2

cos2 ϕ⟩ =

∑0 I(ϕ) sin ϕ cos2 ϕ π /2

∑0 I(ϕ) sin ϕ

where fwhm is the full width of the half-maximum of the azimuthal profiles from the selected equatorial reflection and I(ϕ) is the intensity distribution along the Debye−Scherrer ring.

■

RESULTS AND DISCUSSION The NFC- and NFCh-based macrofibers are prepared by standardized wet-extrusion of aqueous dispersions of TEMPOoxidized NFC (1 wt %, pH = 8.5) and surface-deacetylated NFCh (1 wt %, pH = 4) at an extrusion speed of 126 mL/min C

dx.doi.org/10.1021/bm500566m | Biomacromolecules XXXX, XXX, XXX−XXX

Biomacromolecules

Article

limiting structural features on the mesoscale, as shown in our previous publication.10 To allow for constant conditions during the wet-stretching, we built a computer-controlled device, which allows strain-rate controlled stretching while a sample is immersed into a water bath (Scheme 2). The system works as follows: Initially, a macrofiber with a length L0 of about 10 cm is fixed to the clamps and allowed to relax in water for 10 min, whereupon residual stress imparted during previous drying is lost. When the macrofiber is totally relaxed, it is stretched without force into a straight position L1. Then the sample is stretched up to the desired length L2, which defines the stretching ratio, SR = L2/L1 − 1. The stretching routine consists of two steps. The strain velocity in the first period is set to 5 mm/min until SR = 0.10. After this point, the strain velocity is lowered to 0.5 mm/min to decrease mechanical stress imparted on the macrofiber and give sufficient time for efficient nanofibrillar/colloidal flow and rearrangement until the desired SR is reached. Nanofibrillar flow is reminiscent of polymer disentanglement during inelastic deformation, but occurs on much slower time scales and requires reduction of the interfibrillar forces and slow stretching rates. Thereafter, the macrofiber and the stretching device are removed from water and immediately immersed into acetone for coagulation. After 30 min, the system is taken out and dried while still being clamped in the device. This procedure allows stretching ratios of up to 30%, SR = 0.30. Higher stretching ratios turned out to be unreliable and typically lead to fracture. Hence, this SR can be considered as the maximum rearrangement possible for both kinds of fibers, also bearing in mind that the strain velocity was set to 0.5 mm/ min, which is extremely low and the lowest possible value for the used motor. Interestingly, the maximum SR is similar for both materials, NFC and NFCh, pointing to some intrinsic limitations for rearrangements in these nanofibrillar macrofibers. The stretching ratios are smaller compared to the ones reported for the alignment of NFC nanopapers (SR = 0.60), which of course can be understood considering the fact that unstretched nanopapers exhibit an isotropic in-plane order, while the macrofibers already have a flow-induced order due to the extrusion (as quantified below).34 A direct comparison of the mechanical properties for both NFCh- and NFC-based macrofibers as a function of the stretching ratio is displayed in Figure 2 and summarized in Table 1. Both plots clearly reveal a mechanical stiffening and strengthening with increasing stretching ratio, and a loss in

Figure 1. Tensile mechanical properties of nonstretched NFC and NFCh macrofibers after rehydration in water.

using THF as coagulation solvent and using a needle with an internal diameter of 2 mm. These nanofibrils are stabilized by anionic or cationic electrostatic repulsion, respectively. Such repulsions take place between the carboxylic entities (COO−Na+) present on the surface of the TEMPO-oxidized NFC and the ammonium groups (NH3+CH3COO−) that exist on the surface of the deacetylated NFCh. No additional salt was added beyond acid and base used to adjust the pH. We refer to the Experimental Section for charge densities, degrees of crystallinity, and the molecular weight of the underlying cellulose. Atomic force microscopy (AFM) depicts nicely defined nanofibrils with average diameters, as deduced by a statistical height analysis of 2.5 ± 2 and 3.2 ± 1.1 nm for NFC and NFCh, respectively, and lengths ranging up to several micrometers (Scheme 1d−g). Extrusion and subsequent drying typically yields fibers with diameters in the range of 250 to 300 μm and slightly irregular noncircular cross sections. Following drying, we performed initial screening experiments to understand which solvent is most suitable for performing the wet-stretching to allow longest plastic deformation, needed for an efficient alignment of the nanofibrils during stretching. The most efficient plasticization was observed in water, for which strain-to-failures of up to 20% (NFC) and 27% (NFCh) are observed (Figure 1). The pronounced inelastic deformations, typically being 2−5× larger than for dried NFC and NFCh materials,6,10,11,49 are due to an efficient disengagement of the interfibrillar hydrogen bonds. The tensile curves show multiple yielding phenomena, which correspond to overcoming strain-

Figure 2. Direct comparison of tensile mechanical properties of wet-spun NFCh and NFC macrofibers as a function of stretching ratio (SR), as indicated within the figure. Stress/strain curves of (a) NFCh and (b) NFC macrofibers at different SR. (c) Young’s modulus, E, and ultimate tensile strength, σUTS, as a function of SR for both materials, as indicated within the figure. Ten samples were tested and averaged. D

dx.doi.org/10.1021/bm500566m | Biomacromolecules XXXX, XXX, XXX−XXX

Biomacromolecules

Article

which is frozen during the second coagulation process. The mesoscale layers originate from floc/aggregate formation during the solvent exchange in the coagulation bath, happening from outside to inside, and is possibly assisted further by the concentric shear profile in the extrusion die. Looking at the higher magnification images (Figure 3b,d), one can appreciate a qualitatively higher alignment of the nanofibrils in stretched NFCh macrofibers (Figure 3d). Similarly, the NFC nanofibrils in stretched macrofibers are better oriented in the same direction (Figure 3h), meaning a better alignment compared to the unstretched sample (Figure 3f) in which the nanofibrils are less visible within the network structure of the layers. To go beyond this qualitative understanding of the alignment procedure, we performed 2D WAXS and monitored how azimuthal intensity profiles of selected crystal reflections change during stretching. The setup consists of irradiating the fibers perpendicular to the main axis with the X-ray beam. Looking at the series of 2D detector images (Figure 4a−c and d−f), it becomes obvious that the reflections are increasingly confined into arcs, thereby suggesting an increased alignment of the crystallites along the fiber axis. We traced this orientation more closely for both materials using the crystal (200) reflection for NFC macrofibers and the crystal (110) reflection for NFCh macrofibers. Both reflections (200) and (110) have clear confinement of their arc patterns revealing a preference of orientation in the direction of stretching. Figure 4 displays the azimuthal intensity profiles of the mentioned reflections with their corresponding 2D diffractograms of both types of macrofibers at three different stretching ratios, as spun, as well as with intermediate and maximum stretching ratios. The narrower and more defined reflexes at higher SR clearly confirm increasing degrees of alignment, which can be quantified using the orientation index defined as π = (180° − fwhm)/180° and the order parameter defined as S = (3/2) ⟨cos2 ϕ⟩ − (1/2). Table 2 summarizes the orientation index and the order parameter values for all unstretched and stretched macrofibers. Both parameters, π and S, range from 0 to 1, with unity corresponding to perfect alignment parallel to the stretching direction, while zero corresponds to a random orientation of the nanofibrils. Both methods show a consistent increase for rising stretching ratio, SR, due to the gradual orientation of the nanofibrils induced by the wet stretching procedure. In the following, we mainly use the orientation index for discussion, because these values can be directly related to previous work on stretching of NFC nanopapers, and macrofibers prepared at higher extrusion speed.34,46 The relationship between the orientation index, stretching ratio and stiffness is close to linear, establishing a direct correlation of these parameters, and enabling a direct link between alignment of crystallites and increased mechanical stiffness (Figure 5a,b). The dependency of the stiffness on the alignment of crystallites is a consequence of the anisotropic mechanical properties of such crystal structures within the nanofibrils (α-chitin and cellulose I). The difficulty to align the nanofibrils to an extent larger than 0.85 (breakage above SR > 0.30) is seen in the onset of asymptotic behavior of the orientation index observed in Figure 5b. In terms of pure values, it is possible to compare the data of the NFC macrofibers to the mechanical properties published for aligned nanopapers and NFC macrofibers prepared by extrusion at different speeds.34,46 NFC macrofibers can be aligned to a similar degree as cold-drawn nanopapers prepared

Table 1. Comparison of Mechanical Properties of NFC- and NFCh-Based Macrofibers as a Function of the Stretching Ratio, SR stretching ratio SR 0 0.16 0.22 0.28 0 0.10 0.20 0.30

Young’s modulus E (GPa)

tensile strength σUTS (MPa)

NFC Macrofibers 8.2 ± 2.4 118 26.1 ± 3.1 239 29.2 ± 3.0 289 33.7 ± 4.0 289 NFCh Macrofibers 5.3 ± 0.9 132 8.3 ± 1.4 170 9.2 ± 0.7 175 12.6 ± 1.3 223

elongation εmax (%)

± ± ± ±

12 26 17 34

8.3 2.4 1.9 1.6

± ± ± ±

2.9 1.2 0.2 0.3

± ± ± ±

21 8.2 15 13

9.5 5.5 4.5 3.5

± ± ± ±

1.8 1.5 1.1 1.0

elongation at break, which intuitively is clear due to an increased degree of alignment of the nanofibrils. A direct comparison of both materials shows that the stretching of the NFC-based materials leads to much larger changes in the elastic properties as compared to NFCh-based ones (Figure 2c). The Young’s modulus, ENFC, rises to four times its original value, while ENFCh only doubles. Similarly, the tensile strength shows a larger increase for NFC-based macrofibers and a more drastic lowering of the maximum elongation at break can be observed. This is an interesting phenomenon, as both materials are composed of rather similar, highly crystalline polysaccharide nanoparticles with anisotropic mechanical properties, yet wetstretching has a significantly different impact on the overall changes in mechanical properties. Several factors contribute to the increased mechanical stiffness and strength. First of all, the rapid coagulation during the initial preparation of the fibers in THF can lead to some porosity. In the case of the nonstretched NFC macrofiber, this can explain the rather high elongation at break, low stiffness, strength, and yield point.33 The effect of porosity on mechanical properties of NFC-based materials has been discussed in various previous studies.28−31 For comparison, an ordinary nanopaper based on the same NFC leads to values commonly found in literature for TEMPO-NFC (E = 13 GPa, σUTS = 232 MPa, εmax = 6%).10 Despite efforts using N2 and Xe adsorption measurements, we were not able to determine the porosities, which is due to yet insufficient surface area in the materials. During initial stretching, this residual porosity will be diminished and furnishes a mechanically more cohesive fibrillar network with higher stiffness, strength, and lower elongation. Second, and more importantly, the applied force induces alignment and disentanglement of nanofibrils due to slippage against each other allowed by the plasticizing water. These changes can be traced by electron microscopy and most efficiently by 2D wide-angle X-ray diffraction (WAXS). We performed scanning electron microscopy (SEM) of the fibers at different degrees of stretching and report images for the cross sections of nonstretched (SR = 0) and highly stretched fibers, that is, SR = 0.28 and SR = 0.30 for NFC and NFCh macrofibers, respectively. The most important difference between nonstretched and stretched fibers is the more pronounced appearance of aligned layers on the mesoscale in the stretched fibers, while slightly more random and undulated layers of fibrillar networks can be observed in the unstretched conditions. The more pronounced layers suggest some reduction in residual porosity and also increased alignment, E

dx.doi.org/10.1021/bm500566m | Biomacromolecules XXXX, XXX, XXX−XXX

Biomacromolecules

Article

Figure 3. SEM images of cross sections of (a, b) unstretched and (c, d) stretched (SR = 0.30) NFCh macrofibers and (e, f) unstretched and (g, h) stretched (SR = 0.28) NFC macrofibers.

increasingly controlled by flaws imparted during sample preparation for strongly oriented materials with predominantly linear elastic fracture behavior. More important is the comparison to the maximum performance that can be achieved by increasing the extrusion speed during fiber preparation. All our stretched NFC macrofibers have higher values of orientation index and stiffness than previous NFC macrofibers extruded even at a ten times higher extrusion velocity (E = 23 GPa, π = 0.72) by Iwamoto et al.46 This is an important advantage because, herein, we also use much longer nanofibrils, which are anyway

by Sehaqui et al.34 The aligned nanopapers display a degree of orientation of 82% (π = 0.82), similar to our wet-stretched fibers (0.83), which correspondingly leads to a similar elastic modulus: 33.7 GPa for stretched macrofibers versus 33.3 GPa for drawn nanopapers. We also point to results by Baez et al.35 who used hyperbolic tangential fitting targeting to maximize the slope by extrapolation to zero extension and reported higher values for stretched nanopapers compared to Sehaqui et al.34 The tensile strength of the NFC macrofibers is slightly lower, which is however likely caused by a larger number of surface defects in the macrofibers, because tensile strength is F

dx.doi.org/10.1021/bm500566m | Biomacromolecules XXXX, XXX, XXX−XXX

Biomacromolecules

Article

Figure 4. Azimuthal intensity profiles at the reflection (110) for NFCh macrofibers with (a) SR = 0, (b) SR = 0.20, and (c) SR = 0.30 and at the reflection (200) for NFC macrofibers with (d) SR = 0, (e) SR = 0.16, and (f) SR = 0.28. 2D detector images are shown as inset with the corresponding lattice planes used for the analysis highlighted by the arrows in (a) and (d).

Table 2. Orientation Indices, π, and Order Parameters, S, of NFC and NFCh Macrofibers at Different Stretching Ratios, SR stretching ratio SR 0 0.16 0.22 0.28 0 0.10 0.20 0.30

orientation index π NFC Macrofibers 0.74 0.80 0.82 0.83 NFCh Macrofibers 0.65 0.77 0.79 0.84

orientation of nanofibrils in this kind of materials. One of the reasons for enhanced orientation is the fact that the coagulation in the coagulation bath is not instantaneous and nanofibrils closer to the center of the extrusion die may in parts reorient and loose shear-imposed orientation. However, during stretching of an equally swollen fiber, the full bulk/cross section will be exposed to the same force. Clearly, this shows that wet-stretching needs to be incorporated into real up-scaled spinning procedures to achieve maximum performance. The highest values for the stiffness of NFC macrofibers are in the range of the reported Young’s modulus of single NFC nanofibrils by Tanpichai et al. (E = 29−36 GPa).71 This demonstrates that close to the full potential of what NFC nanofibrils have to offer can be realized by wet-spinning and subsequent alignment. Moreover, the direct comparison between the properties obtained for NFC and NFCh macrofibers reveals subtle differences. XRD analysis clearly shows that the NFCh macrofibers are being aligned to an even higher degree during stretching as compared to NFC macrofibers (Table 2). The

order parameter S 0.48 0.53 0.56 0.57 0.43 0.52 0.53 0.59

more difficult to align with higher extrusion speeds due to persisting colloidal entanglements. Moreover, a gain of additional 10 GPa in stiffness due to wet-stretching demonstrates it to be a more effective process to induce

Figure 5. Variation of the orientation index, π, and the Young’s modulus, E, as a function of the stretching ratio for (a) NFCh and (b) NFC macrofibers. G

dx.doi.org/10.1021/bm500566m | Biomacromolecules XXXX, XXX, XXX−XXX

Biomacromolecules

Article

between macroscopic and nanofibrillar mechanical properties with the crystallinity of the nanofibrils also motivates to identify new plant sources, or genetically evolve plants with intrinsically higher crystallinity of plant microfibrils. This goes along with optimizing isolation pathways to prepare longest nanofibrils with highly maintained crystallinity, which are needed to allow enhanced mechanical performance passing the herein-presented maximum values based on present-day wood-based NFC materials.

orientation index increases from 0.65 to 0.84 for the NFCh macrofibers, while the NFC-based ones only show an increase from 0.74 to 0.83. Nonetheless, this even higher change in alignment only leads to a comparably smaller increase in mechanical stiffness, that is, a doubling of the Young’s modulus (NFCh: ESR=0 = 5.3 GPa; ESR=0.30 = 12.6 GPa), while the Young’s modulus of the NFC macrofibers quadruples (NFC: ESR=0 = 8.2 GPa; ESR=0.28 = 33.7 GPa). This is a surprising difference and may be caused by differences in the relative anisotropy in the mechanical properties of the underlying crystals, that is, the longitudinal against the transversal elastic modulus, E versus ET. Exact anisotropy of cellulose I and αchitin crystalline structures are still unknown since precise data of ET for both materials are not reported yet. The final Young’s modulus, 33.7 and 12.6 GPa, can be correlated roughly to the differences in the elastic modulus of the underlying materials, Ecellulose I = 138 GPa versus Eα‑chitin = 41−59 GPa,51−53 however, bearing in mind that (i) the stiffness of the actually used nanofibrils depends on its degree of crystallinity and (ii) the interfacial cohesion of the network depends on the extent of interfibrillar interactions. A comparison with classical spinning procedures of chitin materials (e.g., by dissolution in strongly solvating solvents and ionic liquids, and subsequent recrystallization) reveals the advantages of using the native nanocrystalline nanofibrils. The herein found elastic modulus of our NFCh-macrofibers significantly exceeds the ones prepared in classical procedures (E = 5−10 GPa),72 and has the additional benefit of being prepared using a water-borne and environmentally friendly route.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Baochun Wang for the AFM measurements, ́ for his assistances in the schemes and Alejandro J. Benitez figures, and also Khosrow Rahimi for XRD discussions. We acknowledge the BMBF in the framework of the AQUAMAT research group and the “ERANET Woodwisdom Program” financed by the BMELV. This work was performed in part at the Center for Chemical Polymer Technology CPT, which is supported by the EU and the federal state of North RhineWestphalia. F.H.S. acknowledges the Thuringian Ministry for Education, Science and Culture (Grants #B514-09051, NanoConSens, and #B515-11028, SWAXS-JCSM) for financial support of this study.

■

■

CONCLUSIONS We established a controlled wet-stretching procedure to induce high orientation in macroscale fibers based on NFC and NFCh nanofibrils. Alignment of the nanofibrils is allowed due to nanofibrillar flow in a strongly disengaged network swollen in water. The Young’s moduli of both types of macrofibers have a close to linear relationship with respect to the stretching ratio and the orientation index used to quantify the alignment. The mechanical stiffness doubles in stretched NFCh macrofibers (from 5.3 to 12.6 GPa), while in stretched NFC macrofibers it quadruples (from 8.2 to 33.7 GPa). The differences in ultimate mechanical performance are attributed to the higher axial modulus of cellulose I over α-chitin. Although this demonstrates some limitations in using nanofibrillar chitin for highest performance fibers, the yet robust mechanical properties combined with the known biocompatibility and bioactivity raise interest for their application in the biomedical field as, for example, surgical sutures or for medical textiles. In the case of NFC macrofibers, aligned samples have a close to maximum mechanical performance, even allowing to unlock another 10 GPa in stiffness as compared to previous attempts using higher extrusion speeds to force alignment. Our results provide important implications for macroscale continuous fiber manufacturing processes based on nanofibrillar gels, which have to become reality to use such sustainable high performance fibers in real-life advanced biocomposites. It has become evident that in situ wet-stretching needs to be included into the process as it allows to increase alignment and performance beyond what higher extrusion rates can offer. One bottleneck to progress in this direction is to increase the gel strength to have sufficiently robust gel fibers sustaining direct and fast stretching procedures. Moreover, the relationship

REFERENCES

(1) Gopalan Nair, K.; Dufresne, A. Biomacromolecules 2003, 4, 666− 674. (2) Saralegi, A.; Fernandes, S. C. M.; Alonso-Varona, A.; Palomares, T.; Foster, E. J.; Weder, C.; Eceiza, A.; Corcuera, M. A. Biomacromolecules 2013, 14, 4475−4482. (3) Villares, A.; Moreau, C.; Capron, I.; Cathala, B. Biomacromolecules 2014, 15, 188−194. (4) de Mesquita, J. P.; Donnici, C. L.; Pereira, F. V. Biomacromolecules 2010, 11, 473−480. (5) Way, A. E.; Hsu, L.; Shanmuganathan, K.; Weder, C.; Rowan, S. J. ACS Macro Lett. 2012, 1, 1001−1006. (6) Das, P.; Heuser, T.; Wolf, A.; Zhu, B.; Demco, D. E.; Ifuku, S.; Walther, A. Biomacromolecules 2012, 13, 4205−4212. (7) Ifuku, S.; Morooka, S.; Morimoto, M.; Saimoto, H. Biomacromolecules 2010, 11, 1326−1330. (8) Kittle, J. D.; Wang, C.; Qian, C.; Zhang, Y.; Zhang, M.; Roman, M.; Morris, J. R.; Moore, R. B.; Esker, A. R. Biomacromolecules 2012, 13, 714−718. (9) Shams, M. I.; Ifuku, S.; Nogi, M.; Oku, T.; Yano, H. Appl. Phys. A: Mater. Sci. Process. 2011, 102, 325−331. (10) Benítez, A. J.; Torres-Rendon, J.; Poutanen, M.; Walther, A. Biomacromolecules 2013, 14, 4497−4506. (11) Walther, A.; Timonen, J. V. I; Díez, I.; Laukkanen, A.; Ikkala, O. Adv. Mater. 2011, 23, 2924−2928. (12) Wang, M.; Olszewska, A.; Walther, A.; Malho, J.-M.; Schacher, F. H.; Ruokolainen, J.; Ankerfors, M.; Laine, J.; Berglund, L. A.; Ö sterberg, M.; Ikkala, O. Biomacromolecules 2011, 12, 2074−2081. (13) Wu, C.-N.; Saito, T.; Fujisawa, S.; Fukuzumi, H.; Isogai, A. Biomacromolecules 2012, 13, 1927−1932. (14) Angellier, H.; Molina-Boisseau, S.; Lebrun, L.; Dufresne, A. Macromolecules 2005, 38, 3783−3792. (15) Ifuku, S.; Nogi, M.; Abe, K.; Yoshioka, M.; Morimoto, M.; Saimoto, H.; Yano, H. Biomacromolecules 2009, 10, 1584−1588. H

dx.doi.org/10.1021/bm500566m | Biomacromolecules XXXX, XXX, XXX−XXX

Biomacromolecules

Article

(16) Goodrich, J. D.; Winter, W. T. Biomacromolecules 2007, 8, 252− 257. (17) Saito, T.; Nishiyama, Y.; Putaux, J.-L.; Vignon, M.; Isogai, A. Biomacromolecules 2006, 7, 1687−1691. (18) Päak̈ kö, M.; Ankerfors, M.; Kosonen, H.; Nykänen, A.; Ahola, S.; Ö sterberg, M.; Ruokolainen, J.; Laine, J.; Larsson, P. T.; Ikkala, O.; Lindström, T. Biomacromolecules 2007, 8, 1934−1941. (19) Henriksson, M.; Henriksson, G.; Berglund, L. A.; Lindström, T. Eur. Polym. J. 2007, 43, 3434−3441. (20) Ifuku, S.; Saimoto, H. Nanoscale 2012, 4, 3308−3318. (21) Le Corre, D.; Bras, J.; Dufresne, A. Biomacromolecules 2010, 11, 1139−1153. (22) Sinko, R.; Mishra, S.; Ruiz, L.; Brandis, N.; Keten, S. ACS Macro Lett. 2014, 3, 64−69. (23) Pillai, C. K. S.; Paul, W.; Sharma, C. P. Prog. Polym. Sci. 2009, 34, 641−678. (24) Kurita, K.; Sannan, T.; Iwakura, Y. Makromol. Chem. 1977, 178, 3197−3202. (25) Cho, Y.-W.; Jang, J.; Park, C. R.; Ko, S.-W. Biomacromolecules 2000, 1, 609−614. (26) Moon, R. J.; Martini, A.; Nairn, J.; Simonsen, J.; Youngblood, J. Chem. Soc. Rev. 2011, 40, 3941−3994. (27) Eichhorn, S. J.; Young, R. J.; Davies, G. R. Biomacromolecules 2005, 6, 507−513. (28) Cervin, N. T.; Aulin, C.; Larsson, P. T.; Wågberg, L. Cellulose 2012, 19, 401−410. (29) Liimatainen, H.; Ezekiel, N.; Sliz, R.; Ohenoja, K.; Sirviö, J. A.; Berglund, L.; Hormi, O.; Niinimäki, J. ACS Appl. Mater. Interfaces 2013, 5, 13412−13418. (30) Sehaqui, H.; Zhou, Q.; Ikkala, O.; Berglund, L. A. Biomacromolecules 2011, 12, 3638−3644. (31) Sehaqui, H.; Morimune, S.; Nishino, T.; Berglund, L. A. Biomacromolecules 2012, 13, 3661−3667. (32) Yano, H.; Nakahara, S. J. Mater. Sci. 2004, 39, 1635−1638. (33) Henriksson, M.; Berglund, L. A.; Isaksson, P.; Lindström, T.; Nishino, T. Biomacromolecules 2008, 9, 1579−1585. (34) Sehaqui, H.; Ezekiel Mushi, N.; Morimune, S.; Salajkova, M.; Nishino, T.; Berglund, L. A. ACS Appl. Mater. Interfaces 2012, 4, 1043−1049. (35) Baez, C.; Considine, J.; Rowlands, R. Cellulose 2014, 21, 347− 356. (36) Håkansson, K. M. O.; Fall, A. B.; Lundell, F.; Yu, S.; Krywka, C.; Roth, S. V.; Santoro, G.; Kvick, M.; Wittberg, L. P.; Wågberg, L.; Söderberg, L. D. Nat. Commun. 2014, 5, 4018. (37) Svagan, A. J.; Azizi Samir, M. A. S.; Berglund, L. A. Biomacromolecules 2007, 8, 2556−2563. (38) Martikainen, L.; Walther, A.; Seitsonen, J.; Berglund, L.; Ikkala, O. Biomacromolecules 2013, 14, 2531−2535. (39) Das, P.; Walther, A. Nanoscale 2013, 5, 9348−9356. (40) Walther, A.; Bjurhager, I.; Malho, J.-M.; Pere, J.; Ruokolainen, J.; Berglund, L. A.; Ikkala, O. Nano Lett. 2010, 10, 2742−2748. (41) Liu, A.; Walther, A.; Ikkala, O.; Belova, L.; Berglund, L. A. Biomacromolecules 2011, 12, 633−641. (42) Malho, J.-M.; Laaksonen, P.; Walther, A.; Ikkala, O.; Linder, M. B. Biomacromolecules 2012, 13, 1093−1099. (43) Laaksonen, P.; Walther, A.; Malho, J.-M.; Kainlauri, M.; Ikkala, O.; Linder, M. B. Angew. Chem., Int. Ed. 2011, 50, 8688−8691. (44) Sehaqui, H.; Liu, A.; Zhou, Q.; Berglund, L. A. Biomacromolecules 2010, 11, 2195−2198. (45) Liu, A.; Berglund, L. A. Carbohydr. Polym. 2012, 87, 53−60. (46) Iwamoto, S.; Isogai, A.; Iwata, T. Biomacromolecules 2011, 12, 831−836. (47) Hamedi, M. M.; Hajian, A.; Fall, A. B.; Håkansson, K.; Salajkova, M.; Lundell, F.; Wågberg, L.; Berglund, L. A. ACS Nano 2014, 8, 2467−2476. (48) Nogi, M.; Iwamoto, S.; Nakagaito, A. N.; Yano, H. Adv. Mater. 2009, 21, 1595−1598. (49) Ifuku, S.; Morooka, S.; Nakagaito, A. N.; Morimoto, M.; Saimoto, H. Green Chem. 2011, 13, 1708−1711.

(50) Isogai, A.; Saito, T.; Fukuzumi, H. Nanoscale 2011, 3, 71−85. (51) Ogawa, Y.; Hori, R.; Kim, U.-J.; Wada, M. Carbohydr. Polym. 2011, 83, 1213−1217. (52) Nishino, T.; Matsui, R.; Nakamae, K. J. Polym. Sci., Part B: Polym. Phys. 1999, 37, 1191−1196. (53) Sakurada, I.; Nukushina, Y.; Ito, T. J. Polym. Sci. 1962, 57, 651− 660. (54) Yamamoto, H.; Horii, F. Cellulose 1994, 1, 57−66. (55) Nishiyama, Y.; Langan, P.; Chanzy, H. J. Am. Chem. Soc. 2002, 124, 9074−9082. (56) Sikorski, P.; Hori, R.; Wada, M. Biomacromolecules 2009, 10, 1100−1105. (57) Eichhorn, S. J. ACS Macro Lett. 2012, 1, 1237−1239. (58) Aulin, C.; Gällstedt, M.; Lindström, T. Cellulose 2010, 17, 559− 574. (59) Aulin, C.; Ahola, S.; Josefsson, P.; Nishino, T.; Hirose, Y.; Ö sterberg, M.; Wågberg, L. Langmuir 2009, 25, 7675−7685. (60) Ifuku, S.; Yamada, K.; Morimoto, M.; Saimoto, H. J. Nanomater. 2012, 2012, 645624. (61) Park, S.; Baker, J. O.; Himmel, M. E.; Parilla, P.; Johnson, D. K. Biotechnol. Biofuels 2010, 3, 10. (62) He, J.; Cui, S.; Wang, S.-Y. J. Appl. Polym. Sci. 2008, 107, 1029− 1038. (63) Lavoine, N.; Desloges, I.; Dufresne, A.; Bras, J. Carbohydr. Polym. 2012, 90, 735−764. (64) Sacui, I. A.; Nieuwendaal, R. C.; Burnett, D. J.; Stranick, S. J.; Jorfi, M.; Weder, C.; Foster, E. J.; Olsson, R. T.; Gilman, J. W. ACS Appl. Mater. Interfaces 2014, 6, 6127−6138. (65) Saito, T.; Isogai, A. Biomacromolecules 2004, 5, 1983−1989. (66) Shinoda, R.; Saito, T.; Okita, Y.; Isogai, A. Biomacromolecules 2012, 13, 842−849. (67) Hiraoki, R.; Fukuzumi, H.; Ono, Y.; Saito, T.; Isogai, A. Cellulose 2014, 21, 167−176. (68) Saito, T.; Kimura, S.; Nishiyama, Y.; Isogai, A. Biomacromolecules 2007, 8, 2485−2491. (69) Peng, Y.; Gardner, D. J.; Han, Y.; Kiziltas, A.; Cai, Z.; Tshabalala, M. A. Cellulose 2013, 20, 2379−2392. (70) Ifuku, S.; Nogi, M.; Yoshioka, M.; Morimoto, M.; Yano, H.; Saimoto, H. Carbohydr. Polym. 2010, 81, 134−139. (71) Tanpichai, S.; Quero, F.; Nogi, M.; Yano, H.; Young, R. J.; Lindström, T.; Sampson, W. W.; Eichhorn, S. J. Biomacromolecules 2012, 13, 1340−1349. (72) Qin, Y.; Lu, X.; Sun, N.; Rogers, R. D. Green Chem. 2010, 12, 968−971.

I

dx.doi.org/10.1021/bm500566m | Biomacromolecules XXXX, XXX, XXX−XXX