Review on nanocellulose composites and CNFs assembled microfiber toward automotive applications

2.1 Nanocellulose structure

Cellulose is a highly abundant green substance on our Earth and possesses a hierarchical structure. Figure 1 illustrates the hierarchical structure of cellulose in plant-based biomass. In terms of various retraction strategies, a diverse range of nanocellulosic materials can be obtained with at least one dimension in the nanometer scale. The molecular structure depicts that nanocellulose is mostly made up of repeating D-glucopyranosyl units (also known as dehydrated glucose units, AGUs), arranged in a ⁴C₁ chain arrangement with each monomeric unit helixed at an angle of 180° from its neighboring units [25]. This molecular chain structure enables the nanocellulose surface to contain a large number of hydroxyl functional groups, resulting in the crystalline regions having a much higher number and strength of intermolecular and intermolecular hydrogen bonds than the amorphous regions. Therefore, the crystalline form of nanocellulose cannot be disrupted easily, while the amorphous regions can be disrupted relatively easily.

Figure 1

Structure and three types of nanocellulose: (a) CNFs, (b) CNCs, and (c) BC. Reproduced with permission from ref. [29–32].

In general, these nanocellulosic materials can be categorized into three main groups: Cellulose nanocrystals (CNCs), Cellulose nanofibers or nanofibrils (CNFs), and Bacterial nanocellulose (BC). Figure 1(a)–(c) demonstrates the SEM images of CNFs, CNCs, and BC. CNFs consist of highly entangled networks of nanofibrils made of alternating crystalline and amorphous domains. CNFs have the characteristics of high mechanical qualities, high surface area, high aspect ratio, and a profusion of surface functional groups. Contrarily, CNCs are mostly made up of rod-like crystalline regions because the majority of amorphous regions of cellulose have been removed using strong acids (Figure 1). As a result, CNCs have stronger mechanical properties than CNFs but is smaller, more spherical in shape, and has a lower aspect ratio [26]. Unlike CNFs and CNCs, BC is a cellulose nanomaterial processed by various species of the Acetobacteriaceae family.

BC can be distinguished from plant-derived CNFs and CNCs by its high purity and distinct crystal structure. These characteristics make it more suitable for applications in the biomedical field [27,28]. Meanwhile, BC is processed by bacteria instead of direct chemical or mechanical processes. The process to produce BC is less efficient and more expensive in comparison with the process to make CNCs and CNFs. Since the cost of nanocellulose-based automotive components and the efficiency to make them are essential for automotive companies, CNFs and CNCs are more preferred. BC-based composites for automotive applications are also extremely rare, and this study only covers contents related to CNFs and CNCs.

2.2 Preparation technology

In 1974, nanocellulose was first chemically extracted from cellulose by the acid hydrolysis process [33]. Since then, various methods have been developed by numerous researchers to obtain nanocellulosic materials, including CNCs and CNFs. Generally, these methods can be divided into three categories, including chemical treatment technique, physical treatment technique, and combined physical and chemical treatment technique. Table 1 summarizes the commonly used preparation techniques of nanocellulose.

With the assistance of chemical treatment technique, Kishimoto et al. [35] effectively prepared CNCs and enhanced its crystallinity by acid hydrolysis of micron-sized cellulose using various ionic solutions (FeSO₄, Fe₂(SO₄)₃, and FeCl₃) as catalysts in conjunction with sulfuric acid solutions. Although high crystallinity nanocellulose materials can be prepared chemically, the inorganic strong acid hydrolysis method is highly corrosive to equipment and difficult to recycle, leaving a large amount of strong acid to be treated at the end of the preparation and causing severe environmental pollution. In enzymatic hydrolysis, enzymes are more expensive, the reaction time is lengthier and less efficient, and the comprehensive chemical reaction results in a significantly lower yield compared to the majority of chemical treatment methods.

The physical treatment technique relies on mechanical process. Initially, 2% wood pulp was treated with a high-pressure homogenizer to produce CNFs with diameters below 100 nm [46], and subsequently, microfluidic injection [31], grinding [47], ultrasonic [48], emulsification [49], steam blasting [50], and ball milling [51] were developed. Figure 2 depicts the most widely used methods for the preparation of nanocellulose based on physical treatment techniques, including homogenization (100-KF3-8BS Manton Gaulin homogenizer, USA and M-110Y Microfluidizer, USA), grinding (Masuko Sangyo Co. Ltd, Japan) and other non-conventional processes. The purely mechanical methods are currently regarded as having significant limitations, such as high energy consumption, severe damage to the cellulose structure, uneven particle size and crystallinity [52], and poor dispersion during application process [53].

Figure 2

Mechanical processes for CNFs production. Reproduced with permission from ref. [45].

To address the drawbacks of the pure chemical and mechanical methods, combined chemical treatments and physical techniques are developed. For instance, the typical process to process wood residues involves two steps. First, wood residues are pre-treated with chemicals, such as acids and certain enzymes, to remove lignin and hemicellulose and then cellulose is isolated and oxidized by a TEMPO-mediated oxidation system under alkaline conditions. The second step is to conduct a mild homogenization process as a physical treatment to obtain nanocellulose. Vallejos et al. [42] successfully obtained nanocellulose from wood chips by using sodium hydroxide to remove lignin, fractionation, TEMPO oxidation, washing, neutralization, and defibrillation of the resulting cellulose suspension using a homogenization in a five-step sequence, resulting in the CNFs with high surface area. In addition to physical homogenization combined with chemical methods, ultrasonic methods coupled with chemical treatments are also recommended. Xiao et al. [43] created pine needle nanofibers by acidifying sodium chlorite treatment (75°C for 1 h), followed by alkali treatment (KOH), acid hydrolysis (HCl, 80°C for 2 h), and finally coupled with ultrasonic treatment (60 kHz for 30 min). The resulting CNFs had a width between 30 and 70 nm, a crystallinity of 66.19%, which shows great promise of usage in thermoplastic composites.

Therefore, chemical pretreatment can be initially used to dissociate the nanocellulose raw material and to remove some of the indeterminate regions. Then, physical treatment can be used to further stabilize the dimensional stability of the prepared nanocellulose and disperse it. Compared to the pure chemical and physical treatment methods, the combination of chemical and physical treatment methods not only significantly reduces energy consumption (pure physical treatment, 3,000–30,000 kW h/t [54]) and the usage of strong acids and alkalis (much lower concentrations than usual), contributing to the environmental sustainability, but also obtains high-quality CNFs with excellent mechanical performance (high crystallinity, dimensionally stable, and thermal stability). The combined chemical and physical treatment method has provided a realistic technique of nanocellulose mass-production for the automotive industry.

2.3 Mechanical properties

As a polymeric and dense nanomaterial, nanocellulose has impressive mechanical properties, which are further enhanced by the presence of intermolecular and intramolecular hydrogen bonds [55]. The key factors such as size, morphological structure, and cross-linking of nanocellulose vary due to inconsistent sources and different preparation processes, which also bring about significant differences in mechanical properties between the nanocellulose materials. The rod-like CNCs have the best mechanical properties and their overall theoretical modulus is about 100–200 GPa (approximately 63–125 GPa cm³/g), with tensile strengths around 4.9–7.5 GPa (approximately 3.0–4.7 GPa cm³/g) [4]. CNFs also possess exceptional mechanical properties such as outstanding tensile strength (0.3–22 GPa) and Young’s modulus (58–180 GPa). The extraordinary mechanical features of nanocellulose naturally offer a great promise for designing and manufacturing green and lightweight composite materials in the automotive industry.

Table 2 summarizes the mechanical properties of nanocellulose materials and the reinforcement fibers of commonly used automotive composites. While carbon fibers (CF) still have a significant advantage over weaker fibers like glass and aramid, nanocellulose even has a higher specific strength than CF.

3.1 Nanocellulose directly reinforced polymers

Nanocellulose directly reinforced polymer is one of the most well-studied nanocellulose composites. In this type of composite, CNCs or CNFs are dispersed directly as fillers in various polymers such as polypropylene (PP), high-density polyethylene (HDPE), Acrylonitrile butadiene styrene (ABS), epoxy (EP), and even in polylactic acid (PLA) and thermo-plastic starch (TPS), which are very promising biodegradable polymers. CNCs and CNFs can not only reinforce polymers and directly improve their mechanical properties, but also reduce the consumption of fossil fuel-based polymers. This type of composite possesses a broad potential in making automotive interior and exterior non-structural components, such as door panels, trim components in dashboards, parcel shelves, etc. [66,67,68].

3.1.1 Surface modification of nanocellulose in nanocellulose directly reinforced polymers

In contrast with synthetic inorganic fibers such as GF or CF, nanocellulose possesses a large number of hydroxyl functional groups on its surface, allowing for the formation of a large number of hydrogen bonds. On the one hand, this feature provides great potential to form strong interfacial bonding between nanocellulose and its adjacent matrix. On the other hand, hydrophilic nanocellulose is extremely susceptible to agglomeration [69], which causes poor interfacial compatibility between nanocellulose and matrix, voids inside the composites, and in sequel, the poor mechanical properties of the composites. Therefore, it is critically important to modify the nanocellulose and/or the polymer matrix to improve their interfacial compatibility.

Saba et al. [70] used the hand-paste technique to fabricate CNFs/EP composites with different filler levels (0.5, 7.5, and 10%). They characterized the micro-morphology of the material by scanning electron microscopy (SEM) and transmission electron microscopy (TEM), combined with tensile, impact, and flexural mechanical properties tests, and then discovered that 7.5% of CNFs filler had the best mechanical and structural properties. However, agglomerations and voids continued to be visible in the SEM images. Figure 3 depicts significant voids and agglomerations in the CNFs/EP composites as well as the large fracture of the material during tensile testing, which may be the primary cause of the degradation of the composite properties at high content. This is further emphasized in Palange et al.’s study [71], who created CNFs/PPPE (PP-polyethylene copolymer) composites with various volume fractions, tested their mechanical properties like tensile strength and impact properties, and characterized the fiber distribution. They discovered that the effect of CNFs agglomeration at high content on the mechanical properties of the composites, particularly the impact properties, was very significant. According to the research conducted by Chu et al. [13], the dispersibility and interfacial compatibility of CNFs are greatly influenced by the surface hydrophilicity/hydrophobicity, charge repulsion, and spatial site resistance. In order to create high performance green lightweight structural materials, it is imperative to address the issue of poor interfacial compatibility and aggregation of nanocellulose caused by hydroxyl groups.

Figure 3

SEM images of the CNFs/EP composites showing (a) agglomerations, (b) void, and (c) deep fracture in 1% CNFs filler loading. Reproduced with permission from ref. [70].

To enhance the mechanical characteristics of the composites by adding nanocellulose, several research works have been conducted to obtain good dispersion and compatibility of nanocellulose/polymer composites using various methods. One method for enhancing the mechanical characteristics of composites and improving dispersion is to reduce the hydrophilicity of nanocellulose by surface modification. Satio et al. [36] carried out TEMPO-mediated systematic oxidation of natural CNFs to enhance the dispersion and transparency of CNFs in dispersions and water. This method is currently used to improve the dispersion of CNFs by oxidizing hydroxyl groups to carboxyl groups and reducing the formation of hydrogen bonds to improve negative charge rejection. In the research conducted by Noguchi et al. [72], TEMPO-CNFs/PE demonstrated good dispersion and enhancement of the processed CNFs.

In addition to carboxylation, there are some other effective hydrophobic modification methods, such as asesterification [73], silylation [74], and graft copolymerization [75] to introduce a variety of hydrophobic surface functional groups to enhance the electrostatic repulsion between nanocellulose, as shown in Figure 4. Yano et al. [76] accomplished diverse esterification modifications of the hydroxyl groups on the surface of CNFs filaments by adding six aliphatic ester groups with linear, cyclic, and branched chains. The modified CNFs were compounded with HDPE, a hydrophobic polymer commonly used in automobiles, by melt extrusion and injection molding, and X-ray CT analysis revealed that the esterified hydrophobic surface modification of CNFs was effective in enhancing the dispersion of CNFs in hydrophobic materials such as HDPE. In terms of tensile modulus (3.32 GPa), tensile strength (51.2 MPa), and linear swelling of CNFs composites, the incorporation of branched chains such as pentane-oil groups exhibited the highest enhancement efficacy. Figure 5 demonstrates that the incorporation of branched chains, such as pentane-oil groups, enhances both the dispersion of CNFs in HDPE materials and the interfacial compatibility of CNFs/HDPE, thereby enhancing their interfacial properties. Kahawita et al. [77] surface modified CNFs and CNCs with a methylsilane coupling agent (Si-69) and compounded it with PP material. They discovered that the methylsilylated CNFs had good dispersion in the PP material, and compared to the pure PP material, the mechanical properties were enhanced.

Figure 4

Common surface modification reactions of nanocellulose (CNC as an example). Reproduced with permission from ref. [55].

Figure 5

Enhancement of CNFs in injection-molded HDPE: (a) and (c) the X-ray CT and TEM images of neat HPDE, (b) and (d) the X-ray CT and TEM images of CNFs/HPDE. Reproduced with permission from ref. [76].

The addition of nanocellulose improves the mechanical properties of a variety of polymers, but the dispersion of nanocellulose as a reinforcement in the matrix greatly affects the stress transfer within the composite. Therefore, surface modification to improve hydrophobicity to reduce agglomeration is the key to resolving this issue and enhancing the use of nanocellulose materials in automotive applications.

3.2 Hybrid fiber-cellulose composites with nanocellulose-enhanced interface

CF and GF reinforced polymers exhibit high strength, high stiffness, excellent fatigue performance, and corrosion resistance, making them widely utilized in automotive structures. However, CF or GF are inorganic and lack chemical functional groups, often, resulting in weak interfacial properties leading to fiber/matrix debonding and delamination. As a high-performance nanoscale filler, nanocellulose can enhance the interfacial properties between the filaments and matrix, thereby improving the composite’s mechanical properties [93].

Azhary et al. [94] added 1% CNFs to a GF/EP composite and hand-glued and vacuum-bagged standard specimens. Mechanical and thermal property experiments revealed that the addition of CNFs increased the tensile strength and modulus by 9 and 10%, respectively, while the flexural strength and modulus are increased by 16 and 6%, respectively. Moreover, the interlaminar tensile strength of the composite increased by 11% due to the uniform distribution of CNFs at the GF/EP interfacial layer. FT-IR analysis also revealed that the CNFs formed a local strengthening of the matrix around the GF, improving stress transfer and enhancing fracture resistance to delamination. A consistent conclusion was reached in a study by Kumar et al. [95]. They used the preparation method shown in Figure 6(a)–(c) to enable the uniform distribution of CNCs around the GF and then utilized vacuum-assisted resin transfer molding to prepare composite laminates. The addition of CNCs increased the interlaminar shear strength (ILSS) by approximately 30%, the flexural strength by 43%, and the flexural modulus by 40%. The enhancement mechanisms of CNCs in the CNCs/GF/EP composite to resist fracture failure are illustrated in Figure 6(d–f). The strong nanoscale interactions and high modulus of the interfacial layer at the GF/EP interface caused by the CNCs significantly improves the energy absorption of interlaminar fracture (Figure 6(d–f)), as evidenced by the SEM images of the interlaminar fracture (Figure 6(g) and (h)).

Figure 6

Reinforcement mechanism of hybrid fiber-cellulose composites: (a)–(c) the preparation process of CNCs/GF/EP composites; (d)–(f) CNCs enhancement mechanisms in fracture failure; (g) and (h) SEM images of the GF/EP composites and CNCs/GF/EP composites. Reproduced with permission from ref. [95].

In addition to CF, nanocellulose has also been utilized in composites containing CF and natural fibers. Wang et al. [96] incorporated CNCs in CF/EP laminates and significantly enhanced the fracture energy, resulting in mode I and II interlayer fracture energies of 310.0 kJ/m² and 2.843 kJ/m², respectively, which are highly desirable for the design of high strength, high fracture toughness composites. Qiu et al. [97] even added two nanomaterials, graphene (GO), and CNFs to CF/EP composites to improve the mechanical interlocking and chemical bonding between the fibers and the matrix. These two nanofillers improved the surface roughness and wettability of the CF and increased the interfacial bonding strength.

3.3 ACCs

In contrast with nanocellulose-reinforced polymers and nanocellulose-enhanced fiber reinforced polymers (FRP), ACCs use a single nanocellulose as both reinforcement and matrix. This material type is anticipated to address the problem of fiber/matrix adhesion by employing chemically similar or identical cellulose in both the matrix and reinforcement [98] and enable a desirable green material with full biodegradability.

As depicted in Figure 7, the primary methods to prepare ACCs are the CIM, (dissolve nanocellulose completely) and the PDM (dissolve the surface of nanocellulose) [99]. Nishino et al. [100] created the first all-cellulose-composites with tensile strengths of 480 MPa at 25°C and dynamic energy storage moduli of up to 20 GPa at 300°C by partially dissolving different sizes of CNFs/CNCs in solvents (CIM). The mechanical and thermal properties of these materials are comparable or even superior to those of conventional composites reinforced with GF. Gindl et al. [101] also prepared ACCs films by PDM, and depending on the controlled CNFs/CNCs ratio, the final material exhibited excellent strength and resilience with a tensile strength of up to 240 MPa, an elastic modulus of 13.1 GPa, and an elongation at break of 8.6%. Due to the lower solubility, dimensional instability, and residual stress, PDM seems to be more suitable for industrial production than CIM.

Figure 7

Schematic of processing approaches of ACCs: (a) the conventional impregnation method (CIM) and (b) the partial dissolution method (PDM).

In recent years, the self-assembly technique (assembled into one macroscopic material by self-bonding in the non-dissolved state) has emerged as an effective method to prepare high-performance macro or micro-scale material with nanocellulose. Nanopaper, a typical material prepared via the self-assembly technique, is produced using only nanocellulose through processes such as thermosetting or filtration. Henriksson et al. [102] fabricated nanopaper by filtration of uniformly distributed nanocellulose in suspension. The resulting nanopaper exhibited porous properties in microscopic images, but still exhibited excellent mechanical properties, with a Young’s modulus and tensile strength of 13.2 GPa and 214 MPa, respectively, and excellent fracture toughness, which may be caused by fiber breaks and hydrogen bonds slippage. Mittal et al. [22] also assembled nanocellulose by flow-assisted wet spinning based on the concept of multiscale design, and the assembled micro-size fibers exhibited dense, ordered, and interlinked properties, which led to a Young’s modulus and tensile strength of 86 and 1.57 GPa, achieving a microscopic-to-macroscopic transformation of nanocellulose with high performance. Meng and Wang [103] and Rosén et al. [104] used simulation to disclose the nanocellulose strengthening mechanism. The results demonstrated that the increase in intermolecular interactions such as van der Waals forces and hydrogen bonds between chemical functional groups on the surface of nanocellulose improved the mechanical properties of the material, and that properties such as fracture recombination of surface hydrogen bonds also made the assembled CNFs material more ductile, which is the foundation to realize the design and assembly of nanosphere cellulose materials into the next layer.

4.1 CNFs assembled continuous microfiber

The primary challenge in fabricating structural materials is to translate the exceptional mechanical properties (such as tensile strength and Young’s modulus, etc.) of nanoscale building blocks into macroscopic blocks. Aligning nanocellulose and assembling them to form continuous macroscopic fibers through bottom-up synthesis strategies is an effective method to fully exploit the properties of nanomaterials. Nanocellulose alignment is a highly effective material design strategy to reduce defects such as pores and improve interfacial bonding between nanocellulose. The size and the mechanical properties of assembled macroscopic fibers are comparable to conventional GF and CF, which are widely utilized in the automotive industry. Since CNFs have a 3D network morphology and are much longer than CNCs, these features facilitate a better interfacial bonding for assembled microfibers and the majority of current research is focused on macroscopic fibers made with CNFs. In our review, CNFs assembled microfiber is utilized to refer to this type of assembled fiber. Various techniques have been developed to prepare CNFs assembled microfiber and Figure 8 depicts the major assembly techniques for nanocellulose assembled microfibers, which include moist wet spinning, self-assembly, surface molecule or group induced assembly, flow-assisted assembly, interfacial complexation, and other techniques. The preparation of dense, highly oriented, and high connection strength assembled fibers is the key to successfully prepare high performance assembled fibers, as determined by microstructure and assembly mechanisms.

Figure 8

Preparation process of nanocellulose assembled fibers: (a) Wet spinning, (b) self-assembly, (c) and (d) flow assistance, (e) flow assistance and interfacial complexation, (f) and (g) microscopic images of assembled fibers, (h) interaction between enhanced interfacing properties by introducing chitosan. Reproduced with permission from ref. [22,105–107].

Initially, Iwamoto et al. [108] spun a CNFs suspension into an acetone coagulation bath by a wet spinning process, and obtained nanocellulose assembled microfibers with an attractive Young’s modulus of 23.6 GPa, a tensile strength of 321 MPa, and an elongation at break of 2%. In comparison to other plant fibers, the outcomes are very appealing. According to the findings of the present study, wet spinning is one of the most appropriate procedures for the preparation of high-performance nanocellulose assembled fibers, and the resulting fibers are dense, highly oriented, and aligned. In addition to wet spinning, processes such as dry spinning [109] and electrostatic spinning [110] have been utilized in the fabrication of CNF assembled filaments, but these methods continue to be plagued by nanoscale defects and porosity issues. However, this does not fully exploit the full potential of nanocellulose, and it is known that in many high-performance nanocellulose assembled fibers, controlling the degree of orientation is important for the mechanical properties of microfibrils and functionalized materials [111]. For example, nanocellulose exhibits high tensile strength and fracture toughness following highly uniform orientation. Current methods for obtaining highly oriented assembled microfibers rely on external forces, such as shear forces and focused flow channels, to enhance the orientation of microfibers [112]. Table 4 lists the main assembly methods of high-performance nanocellulose microfibers, external forces during preparation, interactions in microfibers, and fiber diameters and properties.

In a recent study, Mao et al. [114] used a simple wet spinning process with a syringe combined with a stretch drying process to prepare assembled filaments with different concentrations of CNFs spinning suspensions. The final dried CNFs filaments had excellent mechanical properties, including a Young’s modulus of 35 GPa and a tensile strength of 567 MPa. Kim et al. [115] also investigated the stretch drying process following wet spinning and discovered that the alignment of the stretched CNFs assembled fibers was more consistent, which increased the number of hydrogen bonds between the CNFs and consequently enhanced the mechanical properties of the CNFs assembled fibers. Flow-assisted techniques have also been applied to the preparation of CNFs assembled microfiber in an effort to enhance their mechanical properties. Wang et al. [31] designed a self-twisting microfluidic chip to achieve shear forces applied to CNCs assembled fibers during the preparation process, yielding assembled fibers with an orientation of 0.77, a torsion angle of 5.5° on the fiber surface, a tensile strength of 400 MPa, and a Young’s modulus of 37.3 GPa. Mittal et al. [22] used a dual-flow focusing method to put together CNFs. The fibers were arranged in a near-perfect unidirectional dense arrangement, which was also linked to the effect of fiber size. This gave the macroscopically assembled fibers a Young’s modulus of 86 GPa and a tensile strength of 1.57 GPa, which is higher than the known mechanical properties of natural or synthetic biopolymer [122]. During the preparation process, they first introduced deionized water and then low pH acid to reduce the electrostatic repulsion between the fibers caused by the carboxyl group. This has a very positive effect on the dense and isotropic fibers, as there is no significant build-up of defects or pores in the fiber cross-section, and the highly oriented and dense fiber structure results in very high mechanical properties. Electrostatic forces are also used as an important aid in the preparation of dense, highly oriented fibers. Panicker et al. [117] even used AC electric field assisted wet spinning to create nanocellulose assembled fibers, and they discovered that under AC assistance, the fibers’ Young’s modulus, tensile strength, yield strength, strain at break, and toughness all increased because the nanocellulose’s orientation increased with voltage and the fibers were densely stacked.

Another assembly method is to encourage self-assembly and cross-linking of nano fibers. By altering the nanocellulose before spinning it or by adding suitable reinforcement materials, nanocellulose interactions can be enhanced. In reference to the structure of natural spider silk, Zhao et al. [120] assembled protein-nanocellulose fibers with increased mechanical strength and resilience. In the protein-nanocellulose system, it was discovered that the covalent bond bridging fibers, such as the dynamic imine bonds and acetal bonds, required more energy than hydrogen bond bridged fibers to break, achieving higher tensile strength, stretchability, and toughness. This was done by adjusting the interactions between the nanocelluloses. By spinning CNFs/CNCs with different anions and cations, Zhang and Liimatainen [123] successfully created nanocellulose assembled fibers, but the fibers’ poor Young’s modulus and tensile strength were largely attributable to the interfacial complexation’s use of opposite charges, which had a greater influence on the microstructure. Compared to densely aligned fibers prepared by other methods, this fiber has a lower number of hydrogen bonds and weaker mechanical properties.

Currently, the best way to manufacture CNFs assembled fibers is probably to employ interfacial complexation to enhance the interfacial characteristics and flow-assisted control to orient CNFs. In the study by Cai et al. [105], they used a convergent microfluidic spinning technique to apply external forces to the CNFs to improve the orientation of the fibers, as well as to improve the interfacial interaction between the CNFs through chitosan molecules. Finally, through Ca⁺ synergistic cross-linking of the filaments, the tensile strength was further improved, and the tensile strength was increased to 1,627 MPa, which is already higher than the current mechanical properties of GFs for automotive applications. They also revealed that the mechanism of the high toughness of CNF-assembled fibers is due to the breaking and recombination of hydrogen and interfacial ionic bonds during CNFs slippage through molecular dynamics analysis.

In addition to the above preparation processes, the mechanical properties of assembled microfiber can be improved by adding superior nanomaterials such as graphene [124] or carbon nanotubes [125]. However, in the automotive industry, the automotive parts are cost sensitive. The addition of superior nanomaterials will significantly increase the cost of the corresponding automotive components and hinder their mass production. Therefore, the use of flow-assisted preparation processes and improved interactions between nanocellulose hold promise for the production of green, low-cost, high-performance assembled microfiber in the automotive industry.

4.2 Multiscale modeling of CNFs assembled continuous microfiber

In contrast with conventional CF and GF, CNFs assembled microfiber possesses a distinctive hierarchical microstructure, adding a new layer to the multiscale modeling of CNFs assembled microfiber-based composite structures. Current research is mainly focused on the micromechanics study of CNFs assembled microfiber. These investigations are largely split into two categories: molecular dynamics and theoretical methods. In the study of CNFs assembled fiber by Cai et al. [105], a molecular dynamics model was used to simulate the tensile fracture mechanism of CNFs assembled microfiber. It was discovered that the toughness of the assembled microfiber was primarily due to the energy dissipated by the fracture and reorganization of the ionic and hydrogen bonds at the interface between CNFs. Wan et al. [126] also carried out molecular dynamics simulations on CNFs/single-walled carbon nanotubes (SWNT) wet-spun filaments and discovered that the high mechanical characteristics were due to the high number of hydrogen bonds inside the macroscopic structure of the one-dimensional constructed fiber. However, the molecular dynamics method can only predict the molecular behavior of ionic and hydrogen bonds and it is very time consuming to generate the homogenized responses of CNFs assembled microfiber, which contains numerous CNFs. In contrast to the molecular approach, Meng and Shi [127] developed a multiscale tension-shear model for nanopaper and CNFs assembled microfiber. The theoretical multiscale model not only can consider the nanoscale parameters such as self-healing of hydrogen bonds, but also can predict the macroscopic anisotropic nonlinear behavior of bulk materials. However, the microstructure of CNFs in the theoretical method is simplified to be parallel with each other and the typical curviness of CNFs and entanglement between CNFs, as well as van der Waals force interaction are ignored in their model.

From our perspective, the finite element [128] or finite volume-based micromechanics approach [129–132] could be effective to consider the real CNFs spatial distributions and their interaction inside CNFs. In contrast with the molecular dynamics method, the finite element or finite volume-based approach will be much more efficient since it deals with a limited number of CNFs in a representative volume element (RVE) at the nanoscale instead of coping with numerous molecules directly. To create a high-fidelity RVE of CNFs assembled microfiber, there are two key aspects that need to be considered. First, the 3D spatial distribution of CNFs needs to be characterized and the distribution can be obtained by integrating the SEM images of CNFs assembled microfiber sections processed at different depths via focused ion beam [133]. Second, the current multiscale cohesive law describing the interactions between CNFs, developed by Meng et al. [134], needs to be expanded to consider van der Waals force interaction and the variable distance between adjacent CNFs, as shown in Figure 9.

Figure 9

Multiscale modeling of cellulose nanopaper: (a) and (b) schematic of the interfacial stretch and shear interaction due to hydrogen bonds, and the cohesive law of the interface; (c) schematic of the crack-bridging model for cellulose nanopaper. Reproduced with permission from ref. [134].

4.3 Multiscale modeling and design of CNFs assembled continuous microfiber-based composite structures

Molecular dynamics and finite element method-based homogenization theory have been substantially utilized for the multiscale modeling [135–138] of nanocellulose directly reinforced polymers. Uetsuji et al. [135] conducted a multiscale analysis in terms of a finite element-based asymptotic homogenization theory toward polymers reinforced by randomly distributed nanocellulose. The effects of fiber aspect ratio, aggregation, and fiber volume fraction on the homogenized properties were quantitated. In their recent research [136], a two-step multiscale homogenization scheme is proposed to homogenize PP/Talc/CNFs three-phase composites. The method facilitates the design of new PP composites by finding a combination of the fillers and CNFs contents to tailor the mechanical properties of the composites. Mishnaevsky et al. [137] also developed a code to generate a 3D RVE model of nanocellulose composites reinforced with “snake”-shaped nanocellulose fibrils and conducted a study toward the structure-mechanical properties relationships of composites. Ren et al. [138] conducted a molecular dynamics study on cellulose composites and discussed the bonding mechanism between cellulose and PLA at the molecular level. However, those multiscale methods are focused on the homogenization of nanocellulose randomly reinforced polymers and there is a lack of a modeling strategy toward the nanocellulose assembled microfiber and the relevant structures.

Carbon/glass FRPs have been extensively used in transportation instruments, such as airplanes, and automobiles. The bottom-up multiscale modeling strategy [139–141] has proven to be a highly effective tool to analyze composite materials and structures and facilitate their design. Homogenization techniques are utilized to establish the link between different length scales. The homogenized stress-strain responses are generated and are fed into a constitutive model at the upper scale. To apply the bottom-up multiscale modeling technique to CNFs assembled microfiber composite structures, the conventional bottom-up multiscale modeling strategy needs to be expanded to the nanoscale by incorporating the micromechanics model. As illustrated in Figure 10, from nanoscale to microscale, the anisotropic macroscopic responses of the CNFs assembled microfiber can be predicted via a finite element or finite volume-based micromechanics model and the corresponding RVE contains the real microstructures of CNFs assembled microfibers and the interfacial behaviors between CNFs. Next from microscale to ply level (mm), computational micromechanics are utilized to predict the ply properties from the properties of constituents and geometric information of the microstructure including fiber volume fraction and spatial distribution. With the homogenized ply properties and laminate lay-up information, as well as interplay properties, computational mesomechanics is then used to determine the homogenized relations of laminates. These results are finally utilized within the framework of computational mechanics to obtain the overall response of the composite structures. The detailed process to conduct multiscale modeling is presented in previous studies [139,140].

Figure 10

The bottom-up multiscale modeling strategy of nanocellulose assembled continuous microfiber-based automotive structures.

With the establishment of a multiscale modeling method for CNFs assembled microfiber composite structures, the structural performance can be optimized by adjusting the design parameters at different scales. For instance, at the nanoscale, the diameter and length of CNFs can affect the macroscopic properties of CNFs assembled microfiber. At the microscale, given the CNFs properties, the ply properties are affected by the volume fraction of CNFs, interfacial properties between the fiber and matrix and matrix properties. At the laminate level, the laminate properties are affected by the lamination sequence, and ply thickness. At the structural level, the automotive structures are strongly affected by the type and thickness of the cross-section of automotive components. The CNFs assembled microfiber-based automotive composite structures possess a very large design space from the nanometer scale to the meter scale. This allows for effective utilization of the exceptional properties of CNFs and opens the door to the development of green, high-performance, lightweight structural materials made of nanocellulose.