Project topics and materials
Get Complete Project Material File(s) Now! »
Theoretical background
Cellulose and nanocellulose
Cellulose is a chain polymer built of (1→4)-β-glucose monomers. Cellulose nanofibril (CNF) consists of cellulose chains packed parallel to each other in a specific manner and are linked through van der Waals forces and hydrogen bonds. (Ek et. al., 2011) The crystallinity of the CNF is dependant on the amassed interface of the fibrils and the degree of order of single chains. A non-crystalline region forms in between fibrils and is made of amorph, unaligned cellulose. (Daicho et. al.,2018)
Derivatization, substitution of at least one hydroxyl group into another functional group, changes the features of the cellulose. As the functional groups are changed, the force of the intermolecular bonds also changes through derivatization. (Ek et. al., 2011)
Through high-pressure homogenization of pulp, wood based CNFs can be produced, as pulp in general has a high weight percentage (wt%) of cellulose. At the same time mechanical extraction of CNFs from wood pulp may result in a significant decrease of cellulose crystallinity. In order to decrease clogging, which often occurs in this process and preserves native cellulose crystallinity, hydrophilic cellulose polymers and other swelling agents can be added to pulp prior to treatment or charged groups can be introduced within the wood fibers by chemical pre-treatments. (Klemm et. al., 2011)
TEMPO-oxidized Cellulose (TCNF)
When oxidizing cellulose with 2,2,6,6-tetramethylpiperidine-1-oxyl radical (TEMPO), various systems can be used. For the TEMPO/NaClO/NaBr system under alkaline conditions, the primary hydroxyl of C6 is selectively converted into a carboxylate group. TEMPO and NaBr act as catalysts while NaClO acts as an oxidant. The mechanism is illustrated in Fig.1 below. Further, selective reduction of aldehyde groups that have not been converted into carboxylates can be done using
NaBH. (Saito et. al., 2006)
It has been found that using a higher concentration of NaClO during oxidation results in CNFs with higher fibrillation yield and increased carboxylate content. The reason for the higher fibrillation yield is, among others, the high content of carboxylate. Carboxylates cause the CNFs to repel each other, through electrostatic repulsion. It has also been observed that usage of higher concentrations of NaClO result in stronger networks of CNFs as it increases the specific area of the cellulose which results in entanglements of CNF. (Bettaieb et. al., 2015)
TEMPO-oxidation does not affect the crystallinity of the cellulose as the substitution of functional groups does not occur inside the cellulose crystallites but rather on the surface of them. (Saito et. al., 2006)
Carboxymethylated Cellulose (CMC)
Under alkaline conditions, cellulose dispersed with NaOH in an organic liquid, the hydroxyl groups of cellulose can undergo etherfication with monochloroacetic acid (C2H3ClO2) or sodium monocholoacetic acid (SMAC) as described below. This process can be repeated with the product of the foregoing step. (Aguir et. al., 2005) CellONa + C2H3ClO2 → CellOCH2COONa
Several carboxymethyl groups can be substituted onto the alkaline cellulose depending on the the conditions of the reaction. The determining factors of the degree of substitution (DS) are reagent concentrations, type of solvent and amount of times carboxymethylation has occurred. (Aguir et. al., 2005) DS defines how many functional groups per repeating unit that are substituted within a polymer. (No author, 2011) For the reagent concentrations, the amounts of NaOH and C2H3ClO2 are relevant. An optimum DS can be reached with increasing concentration of NaOH and the higher the concentration of C2H3ClO2 the higher the DS. For the solvent, the DS increases with decreasing polarity of the organic solvent. The more times that the carboxymethylation is done, the higher the DS. (Aguir et. al., 2005)
When NaOH is introduced during carboxymethylation, it affects the crystallinity of the cellulose as it reacts in the amorphous regions of it. The reactions between NaOH and cellulose results in cellulose-II replacing some of the cellulose-I. It has been found that the overall crystallinity increases when increasing the amount of NaOH used. (Bhandari et. al.,2011)
Rheology of CNF Dispersions
Viscosity of CNF dispersions are in direct correlation to shear rate, which is the propagation of movement through a liquid. (Naderi et. al., 2016) A CNF dispersion is a shear thinning liquid, meaning it is non-Newtonian and the viscosity of the liquid decreases during shear strain. (Meng et. al., 2016) The CNF dispersions typically obtain some elastic properties above a certain overlap concentration, typically around 0.01-0.05 wt%, due to increasing fibril-fibril interactions. (Geng et. al., 2018) The overlap concentration thus signifies the critical concentration separating the semi-dilute and dilute regions of the dispersion. (Onyianta et. al., 2017) A semi-dilute CNF dispersion is also known to be thixotropic, meaning that the viscosity is not only dependent on the instantaneous shear rate, but also on the shear history, for example if dispersion has been pre-sheared or not. (Naderi et. al., 2016) At higher concentration, such as 0.3 wt% for a dispersion of 980 μmol/g charge, the connectivity between nanofibrils is so high that the dispersion transitions to a volume-spanning arrested gel-like state. (Geng et. al., 2018)
It has been proven that for CMC an increase of the system’s ionic strength leads to a decrease in viscosity as well as gel stiffness. (Naderi et. al., 2016) The viscosity of TCNF follows the same trend. (Moberg et. al., 2017)
Sonication and centrifugation
Sonication and centrifugation are two processes that can be used to homogenize dispersions. The sonication process disrupts aggregates in a dispersion by applying high energy ultrasonic frequencies through a liquid sample. The process agitates the nanofibrils through rapid compression and disturbs clotting. (Thanu et. al., 2019) A probe is inserted in the sample and emits focused acoustic energy evenly throughout the sample. (Covaris, 2020) Centrifugation can, in turn, be performed to separate fibrils of different lengths and aggregates to prevent abnormally big objects from entering the machinery or threads, and thereby causing clogging or defects within filaments, respectively. (Yang, et. al., 2019)
Flow-focusing
Cellulose nanofibrils (CNFs) can be organized and spun into strong macroscale fibers by ensuring fibril alignment in the structure, which can be achieved using a method called flow-focusing spinning. (Mittal et. al., 2018)
Flow-focusing spinning is performed using a double flow-focusing channel geometry, which consists of six different channels: one for the CNF dispersion (marked by 1 in Fig.2), two each for deionized water (marked by 2 in Fig.2) and acid at low pH (marked by 3 in Fig.2), and one that serves as outlet (marked by 4 in Fig.2). The CNF dispersion is injected in the core flow channel, while sheath flows of water and acid are injected in channels perpendicular to the flow of the CNF dispersion. The principle for the double flow-focusing spinning method is shown in Fig.2. (Mittal et. al., 2018)
Figure 2. Illustration of the principle structure of a double flow-focusing channel that is used in order to align cellulose nanofibrils (CNFs) into macroscale fibers. A CNF dispersion is injected in the core flow (marked by 1), while deionized water (marked by 2) and acid at low pH (marked by 3) are injected perpendicular to the CNF dispersion flow. This figure is adapted from Mittal et. al.,2018
When the nanocellulose dispersion comes in contact with the sheath flows of water (marked by 2 in Fig.2), hydrodynamically induced fibril alignment is obtained in the direction of the flow. (Håkansson et. al., 2014) As a direct effect of introducing acid at low pH, the alignment is then locked in, what is described by Mittal et. al. (2018) as, a metastable colloidal glass structure. When the nanocellulose dispersion comes in contact with the acid, the carboxylate (COO-) groups become protonated and the electrostatic repulsions are reduced and overcome by van der Waals and hydrophobic forces. The resulting metastable structure prevents loss of alignment, that would otherwise occur due to Brownian motion. (Mittal et. al., 2018)
The deionized water flowing along the walls of the flow-focusing channel prevents transition of the CNF dispersion into the glass state in contact with the walls, which could otherwise lead to clogging of the cell. (Mittal et. al., 2018)
Analytical methods
Flow-stop
A similar flow-focusing setup to the one explained in section 2.4 above can be combined with a flow-stop procedure, in order to measure birefringence in cellulose nanofibril (CNF) dispersions and thereby provide information on the fibril alignment in relation to Brownian motion. The single flow-focusing setup is in this case consisting of four main channels as can be seen in Fig.3. The CNF dispersion is injected in the core channel and deionized water is injected in the two sheath flow channels perpendicular to the flow of CNF. The fourth channel serves as an outlet. This setup ensures an extensional flow of the CNF dispersion, which causes the CNFs to align in the direction of the flow. The flow cell is mounted in between two cross polarizers, and a high-speed camera is installed to measure red laser light passing through the setup. The birefringence properties of the CNF dispersions allow for measuring fibril alignment in the system by observing changes in birefringence of the red laser light. Once the flows of CNF dispersion and deionized water reach a steady state, the flows are rapidly stopped by closing the valves (marked with V in Fig.3) connected to the pumps. Brownian motion then causes the system to de-align, which can be observed as a decay of birefringence. The collected data is compared to a reference, whereupon comparisons can be made between different dispersions. (Brouzet et. al., 2019)
Figure 3. Illustration of the setup used for flow-stop analysis of CNF dispersions, where a flow cell with a single flow-focusing geometry is mounted in between two cross polarizers. The CNF dispersion (grey) flows in the center channel of the cell, while deionized water (blue) flows in the two channels perpendicular to the flow of CNF. The stopping of the flows is controlled by three-way valves, marked by V. This figure is adapted from Brouzet et. al., 2019 (Supporting information).
The definition for birefringence used when performing flow-stop analysis is presented in equation 1 below. Birefringence = √ I (1) texp where I is the intensity of the collected red laser light and texp is the exposure time used when performing the analysis.
Tensile test
To examine the mechanical properties, attributes and, by extension, the real-world applications of finished filaments stress-strain (tensile) testing can be performed. Stress-strain testing is the procedure of dislocating two points of a sample’s length from each other to the point of full fracture. The main values sought are the overall load withstood before fracture, ultimate stress, and the force withstood without altering the filament geometry. In addition, the extruded force for a deformation to occur may be relevant, if noticeably high for a sample, depending on the filaments intended use, and should be taken into account. Tensile testing plots the Engineering Stress ( ) with respect to the Engineering Strain ( ). The implemented equations are listed below (equations 2 and 3), along with their completions. (Tu, S. et. al., 2020) Engineering Strain ( ), displacement of the sample ( δ ) over the original length of the sample (L), is calculated as seen in equation 2.
At lower loads applied, the filament thread will act according to Hooke’s law, as a brittle, elastic material, which is reflected in the linear part of the graph in Fig.4, referred to as the elastic modulus (E). At higher loads however, the sample takes on a ductile property, irreversibly stretching with the displacement until fracturing. The strain required for fracture is referred as strain at break. The elastic modulus is the constant proportionality between strain and stress at lower loads and can be calculated as follows in equation 4. (Roylance, 2001) σe = Eεe (4)
The area below the graph represent the total energy requirement for fracture, namely, the toughness of the sample. (Roylance, 2001)
Chemical cross-linking has proven to be a, to some degree, replication of the natural cross-linking between cellulose and lignin/hemicellulose by introducing bonds between fibrils. This leads to an improvement in connectivity and stress transfer. (Mittal et. al., 2018)
Scanning electron Microscopy
In order to assess the topography and alignment of individual filaments, scanning electron microscopy (SEM) can be utilized. This method allows for analysis of the electron patterns emitted of microscopic regions on a sample when focusing an electron beam on it. The microscope can determine so called secondary, back-scattered, electrons diffracted in a thin specimen such as nanocellulose filament. These electron-diffractions along with emitted X-rays show dislocations, defects, interfaces and second phase particles in the sample. (Clarke et. al., 2002)
Wide-angle X-ray scattering
Wide-angle X-ray scattering (WAXS) is a method that could be used to analyze the structure of a filament, as well as the alignment and crystallinity of the fibrils inside the filament. (Clarkson et. al., 2018) Alignment refers to the orientation of crystalline regions, amorphous regions or both. (Bunsell, et. al., 2018) Crystallinity refers to the periodic ordering of the cellulose chains inside the fibrils. (Fahlman, 2018) The process involves scattering of a sample. First an X-ray beam is fired towards the sample. A collimator narrows the beam further towards the sample in several layers to parallelize the beam of the X-rays while removing the excess X-rays. Any and all density irregularities of the sample, including the sample in its entirety will scatter the primary beam away from its source. The detector then picks up the scattered X-rays while a physical object, called beamstop, between the sample and detector prevents the unscattered parts of the primary beam from damaging the detector. (Pauw, 2007) This characterization method can reveal information on crystallinity, orientation and alignment. (Clarksonet. al., 2018)
Figure 5. Illustration of WAXS analysis. Figure a.) shows a schematic of the basics for WAXS analysis and b.) shows a WAXS analysis instrument.
The scattering intensity, I, is analyzed at various scattering vectors, q, defined as follows in equation 5. (Björn, 2018)
q = 4π • sin(θ) (5) λ
where 2θ is the scattering angle between primary beam and scattered X-rays, see Fig. 5 a.), and λ is the wavelength of the X-ray beam. (Björn, 2018)
Furthermore, if the scattering intensity pattern is displaying ordered properties it may be seen as dependant of the Azimuthal angle χ, see Fig. 5 a.). (Björn, 2018)
WAXS analysis of CNF filaments
A crystalline material or region can be represented by viewing it as a lattice of atoms intersected by its crystal planes and corresponding Miller indices. The indices represent the geometric orientation of the crystal planes by three numbers, correlating to the axis coordinates intersected by the plane and the amount of planes per crystal facete. (Björn, 2018)
Three Miller indices for cellulose are (1 1 0), (110), and (200) which translate to angles where q is approximately 1.02 nm-1, 1.16 nm-1 and 1.59 nm-1, respectively. These numbers correlate to known crystal planes for cellulose. (Han et. al. 2013)
Equations for WAXS analysis
From the data given through WAXS analysis of a sample a crystallinity index and an orientation index can be calculated. The crystallinity index is calculated using the intensity data calculated as a function of the scattering vector, q, and gives an idea of the crystallinity of a sample. Crystallinity indices can be used to compare the crystallinity of different samples, providing that the crystallinity index is computed the same way for all of them. A crystallinity index of one hundred percent indicates complete crystallinity of the sample, while a crystallinity index of zero percent indicates that the structure is completely amorphous. In this report equation 7 below is used as the definition for crystallinity index.
Table of contents :
1. Introduction
2. Theoretical background
2.1 Cellulose and nanocellulose
2.1.1 TEMPO-oxidized Cellulose (TCNF)
2.1.2 Carboxymethylated Cellulose (CMC)
2.2 Rheology of CNF Dispersions
2.3 Sonication and centrifugation
2.4 Flow-focusing
2.5 Analytical methods
2.5.1 Flow-stop
2.5.2 Tensile test
2.5.3 Scanning electron Microscopy
2.5.4 Wide-angle X-ray scattering
2.5.4.1 WAXS analysis of CNF filaments
2.5.4.2 Equations for WAXS analysis
3. Experimental setup
4. Method
4.1 Dispersions tested for spinning
4.2 Spinning
4.2 Optical Microscopy
4.3 Flow stop
4.4 Tensile testing
5. Difficulties resulting from COVID-19
6. Results
6.1 Spinning
6.2 WAXS-analysis
6.3 SEM
6.4 Flow-stop
6.5 Tensile testing and microscopy
7. Discussion
7.1 Spinning
7.2 WAXS
7.2.1 Crystallinity
7.2.2 Alignment
7.3 SEM
7.4 Flow-stop
7.5 Tensile testing and microscopy
7.6 Combined discussion
8. Conclusions
9. References
10. Appendix
10.1 Calculations
10.1.1 Calculation of crystallinity index
10.1.2 Calculation of orientation index
10.2 Figures
10.2.1 Flow-stop results
10.2.1.1 Flow-stop results for dispersions tested for spinning
10.2.1.2 Flow-stop results for dispersions not tested for spinning
10.2.2 Tensile results
10.3 Tables
