Pre-Fibrillation of Pulps to Manufacture Cellulose Nano ber Reinforced High-Density Polyethylene using the Dry-Pulp Direct Kneading Method


 The dry-pulp direct-kneading method is an industrially viable, low-energy process to manufacture cellulose nanofiber (CNF) reinforced polymer composites, where chemically modified pulps can be nanofibrillated and dispersed uniformly in the polymer matrix during melt-compounding. In this study, cellulose fibers with different sizes, ranging from surface-fibrillated pulps with 20 µm in width to fine CNFs with 20 nm in width were prepared from softwood bleached kraft pulps (NBKPs) using a refiner and high-pressure homogenizer (HPH). These cellulose fibers were modified with alkenyl succinic anhydride (ASA), and then dried. The dried ASA-treated cellulose fibers were used as a feed material for melt-compounding in the dry-pulp direct kneading method to fabricate CNF reinforced high-density polyethylene (HDPE). When surface-fibrillated pulps were employed as a feed material, the pulps were nanofibrillated and dispersed uniformly in the HDPE matrix during the melt-compounding, and the composites had much better properties (i.e., much higher tensile modulus and strength and much lower coefficient of thermal expansion) than the composites produced using the pulps without pre-fibrillation. However, when CNFs were used as a feed material, the CNFs were shortened and agglomerated during the melt-compounding, thus deteriorating the properties of the composites. The study concludes that the pre-fibrillation of pulps had a significant impact on the morphology and properties of the composites. Unexpectedly, the surface-fibrillated pulp, which can be produced cost-effectively using a refiner at an industry scale, was a more suitable form than the CNF as a feed material for melt-compounding in the dry-pulp direct kneading method.


Introduction
composites have yet to expand in industrial application. One of the main reasons is a low a nity between hydrophilic CNFs and hydrophobic polymers. Due to hydrogen bonding between OH groups in CNFs, CNFs aggregate in hydrophobic polymers during melt-compounding. Therefore, many researchers have been trying to improve the low a nity by modifying CNF surface chemically (Hassan et (Igarashi et al. 2018) to manufacture CNF reinforced thermoplastic polymer composites. In this method, pulps are modi ed with alkenyl succinic anhydride (ASA), and then dried. The dried, ASA-treated pulps are used as a feed material for melt-compounding with thermoplastic polymers, and the ASA-treated pulps are nano brillated and dispersed uniformly in the polymer matrix during the melt-compounding. This process enables to combine the two processes (i.e., the production of CNFs and melt-compounding of CNFs and thermoplastic polymers), thus reducing the manufacturing time, energy and cost signi cantly. Furthermore, this process increases the aspect ratio of ber (length/width) by changing its width from micro-scale (pulp) to nano-scale (CNF) during the meltcompounding, and enhances mechanical properties of the composites. Therefore, if CNFs, rather than pulps, are used as a feed material for melt-compounding with thermoplastic polymers in the dry-pulp direct kneading method, the CNFs may be further brillated to ner CNFs, thus further improving mechanical properties of CNF reinforced thermoplastic polymer composites. However, effects of prebrillation of pulps on morphology and mechanical property of the composites have not been reported before.
In this study, ASA-treated cellulose bers with different sizes (from pulps with 20 µm in width to ne CNFs with 20 nm in width) were used as a feed material for melt-compounding with high-density polyethylene (HDPE) in the dry-pulp direct kneading method, and their effects on morphology, mechanical properties and coe cient of thermal expansion of the composites were investigated.

Pre-brillation of pulps
Cellulose bers with different sizes, ranging from surface-brillated pulps to ne CNFs with 20 nm in width, were prepared from NBKPs using a re ner and high-pressure homogenizer (HPH). The NBKPs were treated using a re ner until CSF was less than 100 mL. The degree of polymerization was 940, calculated from the value of relative viscosities, as per TAPPI standard T230om-99. The lignin content was 0wt% according to the Klason lignin method (TAPPI standard T230om-02). The re ner-treated pulps in water (0.5wt%, 2L) was disintegrated mechanically using a high-pressure homogenizer (HPH) (Star Burst 10, HJP-25008 K, Sugino Machine CO., Ltd., Japan) at 15°C. The nozzle size was 0.17 mm and pressure was ca. 200 MPa. The HPH treatments were repeated up to ten times.
In this study, 5 types of cellulose bers were prepared: pulps without pre-brillation, pulps treated by the re ner and CNFs treated by the re ner and HPH (1 pass, 3 passes and 10 passes).

Field emission scanning electron microscopy
The cellulose bers were observed using a eld emission scanning electron microscope (FE-SEM, JSM-6700F; JEOL, Ltd., Tokyo, Japan) at an acceleration voltage of 1.5 kV. Before the observation, the cellulose bers were coated with platinum using an ion sputter coater (Auto-Fine Coater JFC-1600; JEOL, Ltd., Tokyo, Japan).

X-ray diffraction
X-ray diffraction patterns of the cellulose bers were obtained using an X-ray diffractometer (UltraX

Dewatering time test of cellulose ber suspension
A specially designed vacuum lter was used to characterize ease of removing water from suspensions of the cellulose bers with different sizes. The ltration system was composed of a stainless-steel mesh sieve having a large aperture size and lter papers. Three sheets of quantitative ashless lter paper with a diameter of 185 mm (5A, Advantec Toyo Kaisha Ltd., Japan) were wetted with distilled water prior to the test. The lter papers were sandwiched between O-rings and put on the sieve. The ltration system was placed between two hollow cylinders. Six hundred grams of 0.1wt% cellulose ber suspension was poured into the upper cylinder at 20°C. The suspension was then gently stirred before applying a vacuum of -30.3 kPa (± 0.4) to the lower cylinder. The ltration time of the cellulose bers through the system was recorded.

Surface modi cation of cellulose bers
All the cellulose bers (i.e., the pulps without pre-brillation, pulps treated by the re ner and CNFs treated by the re ner and HPH) were modi ed with ASA in NMP. Since ASA reacts with water before reacting with OH groups of the cellulose bers, water was replaced with an aprotic organic solvent, NMP, before the ASA treatment. Five hundred grams of wet cellulose bers (solids content of 20-25wt%) and 450 g of NMP were mixed (Tri-mix TX-5; INOUE MFG., Inc. Kanagawa, Japan), followed by evaporation of water under reduced pressure at 20-60°C. ASA (80 g) in 50 g of NMP and 25 g of K 2 CO 3 were added to the cellulose ber/NMP, and the mixture was stirred for 1 hour at 70-80°C. The ASA-treated cellulose bers were washed with a series of acetone, ethanol, aqueous acetic acid, distilled water, and isopropanol to have ASA-treated cellulose bers in isopropanol. Acetone and ethanol were used to gradually change polarity of the solvent into hydrophilic nature. Acetic acid was used to replace the ASA functional end structure COOK with COOH. Further information about the ASA treatment can be found elsewhere (Sato et al. 2016(Sato et al. , 2019. The degree of substitution (DS) of the cellulose bers was calculated from the area of the 1740 cm − 1 absorbance peak in Fourier transform infrared (FT-IR) spectrum. Attenuated total re ection infrared (ATR-IR) spectra were recorded using an FT-IR spectrometer (Spectrum One, Perkin Elmer) in the range of 600 to 4000 cm − 1 at a resolution of 2 cm − 1 . A spectrum was obtained by accumulating 16 scans and normalized using the 1315 cm − 1 peak of the cellulose CH 2 vibration. The band of the C = O stretching vibration mode of the acyl group (from 1735 to 1750 cm − 1 ) increased signi cantly for the modi ed cellulose bers, indicating that esteri cation occurred. The DS and the area of the 1740 cm − 1 peak were correlated by titration of the cellulose bers in advance. The DS of all the cellulose bers prepared in this study was 0.4 ± 0.03. Dumbbell-shaped specimens were prepared from the pellets using an injection molding machine (NPX7-1F; Nissei-Plastic Industrial Co., Ltd., Nagano, Japan) at an injection temperature of 160°C, an injection pressure of 100 MPa, an injection speed of 80 mm/s and a mold temperature of 40°C. The specimens had a length of 60 mm, a width of 5 mm at the neck and a thickness of 1 mm. Figure 2 summarizes the procedure for fabricating the composites described above.

Tensile test
Tensile properties of the composites were measured using a universal testing machine (Model 3655; Instron Corp., Canton, MA, USA) with a crosshead speed of 10 mm/min at 23°C. The load cell of the machine was 5 kN. Strain at failure was measured using a video camera. Tensile modulus was calculated from the stress-strain curve. Five specimens were tested and the average values were reported.

Thermomechanical test
Coe cient of thermal expansion (CTE) of the composites was acquired using a thermomechanical analyzer (TMA/SS6100, SII Nanotechnology Inc., Japan). A specimen was 30 mm long, 1 mm thick and 5 mm wide. Before measurement, the specimen was heated at 110°C for 48 h to remove moisture and then allowed to cool in a desiccator. The original grip distance was 20 mm and the specimen was tested under a tensile load of 29.4 mN. Measurement was performed in the range of -5° to 130°C at a heating rate of 5°C/min under a nitrogen atmosphere (60 ml/min). CTE was evaluated as the fractional change in length per degree of temperature change in the range of 0°C to 60°C.

X-ray computed tomography
Dispersibility and degree of brillation of cellulose bers in the composites was observed using a X-ray computed tomography (CT) (SKY Scan 1172 instrument, Bruker-Micro CT, Kontich, Belgium).
Observations were performed at the center of the dumbbell-shaped specimen.

Extraction of cellulose bers from composites
To remove the HDPE matrix from composites, the composite sample was placed in a stainless-steel mesh container, and the container was immersed and stirred in boiling p-xylene (160°C). The sheet-like cellulose bers left in the stainless-steel mesh were used to observe degree of brillation of cellulose bers in the composites using the FE-SEM. In addition, to measure length of cellulose bers in the composites, the cellulose bers left in the stainless-streel mesh were dispersed into ethanol (0.001 wt%). The dilute solution was placed and dried on a glass plate for FE-SEM observation. Lengths of 100 cellulose bers were measured from the FE-SEM images.

Results And Discussion
3.1 Pre-brillation of pulps Figure 3 and Fig. 4 show, respectively, optical micrographs and FE-SEM micrographs of the cellulose bers treated using the re ner and HPH. It is noted that in the FE-SEM micrographs (Fig. 4), the scale bars for the pulps without pre-brillation and re ner-treated pulps are 100 µm while those for the HPH-treated pulps are 1 µm. Both gures indicated that the degree of brillation increased through the re ner and subsequent HPH treatments. Most of the re ner-treated pulps were not brillated to CNFs but their surfaces were brillated, which suggests that external brillation of the pulps occurred. The external brillation is cutting or removal of the primary wall and the outer layer of the secondary wall (S 1 layer) of a pulp, which allows u ng the pulp surface and accelerates brillation of the inner walls (S 2 and S 3 layers) of a pulp (Uetani and Yano 2011).
A high-pressure homogenizer (HPH) is often used to manufacture CNFs from pulps (Baati et al. 2018;Phanthong et al. 2018). Unlike a bead mill, HPH does not cause contamination during the brillation process. In this study, HPH was used to brillate the re ner-treated pulps. After the re ner-treated pulps passed through the HPH (1 passage of the bers through the HPH), bers with a wide range of width were produced. Some surface brillated pulps were brillated to CNFs with 20-100 nm in width, suggesting that internal brillation (i.e., delamination and brillation of the inner S 2 and S 3 layers) occurred, while some surface-brillated pulps with about 20 µm in width were remained. After 3 passages of the bers through the HPH, most of the bers were CNFs with 20-100 nm in width, but some micro-sized bers with several µm in width (named CMFs hereafter) were observed. After 10 passages of the bers through the HPH, the bers became ne CNFs with around 20 nm in width and either pulps or CMFs were not observed.
It is known that brillation of pulps under high shear stress can decrease crystallinity of cellulose bers, which may affect reinforcing e ciency of cellulose bers (Iwamoto et al. 2007;Ho et al. 2015). X-ray diffraction patterns and crystallinities of the re ner-treated pulps and HPH-treated (1 pass, 3 passes and 10 passes) CNFs are shown in Fig. 5 and Table 1, respectively. All the cellulose bers had similar diffraction patterns and crystallinities, which indicates that the HPH treatments had little effect on the crystal structure and crystallinity of the cellulose bers prepared in this study.
It is not an easy task to quantify the degree of brillation of cellulose bers with a wide range of sizes (from pulps to ne CNFs) using a single experimental method. The Canadian Standard Freeness (CSF) test was conducted for the pulps and CNFs produced in this study. However, the HPH-treated CNFs were too small to measure their freeness. Therefore, viscosities of water solution of the cellulose bers as well as the speci c surface area of the cellulose bers were measured. These two test methods re ected difference in the degree of brillation between the CNFs and ne CNFs but could not detect the difference between the pulps without pre-brillation and surface-brillated pulps. Consequently, the dewatering time test was used to characterize the degree of brillation of the cellulose bers used in this study. The dewatering time test successfully measured the time required for ltration of cellulose ber-water slurry for all the cellulose bers, ranging from the pulps without pre-brillation to the ne CNFs. As listed in Table 1, the dewatering time increased with increasing the degree of brillation. In particular, the dewatering time of the HPH-treated (1 pass) CNFs was about 8 times longer than that of the re nertreated pulps, which mirrors the signi cant difference in the degree of brillation between the HPH-treated (1 pass) CNFs and the re ner-treated pulps (see Figs. 3 and 4). It should also be noted that such a longer dewatering time causes a large increase in the cost to produce dried HPH-treated (1 pass) CNFs, which is not favorable to industrial setting.

Mechanical properties and CTE of composites
Typical stress-strain curves of the composites from tensile tests are shown in Fig. 6, and the mechanical properties (tensile modulus, tensile strength, and strain at failure) are summarized in Table 2. Meanwhile, elongation-temperature curves of the composites from thermomechanical tests are shown in Fig. 7, and the CTE values are summarized in Table 2. Since the HPH treatments facilitated nano brillation of the re ner-treated pulps, it may be expected that the HPH-treated CNFs have higher reinforcing e ciency than the re ner-treated pulps. However, the composites produced using the re ner-treated pulps and the HPHtreated (1 pass) CNFs had similar tensile modulus, tensile strength and CTE, as shown in Table 2.
Interestingly, further increase of the number of passages through the HPH caused deterioration of the properties (i.e., decrease in tensile modulus and strength and increase in CTE). It should be noted that the properties of the composites produced using the HPH-treated (10 passes) CNFs are similar to those of the composites produced using the pulps without pre-brillation.

Morphology of composites
Cellulose bers in the HDPE matrix were observed using X-ray CT (Fig. 8) and FE-SEM (Fig. 9). The X-ray CT image shows density distribution in three-dimensional space with special resolution of 700 nm. In Fig. 8, the HDPE matrix, which has a lower density than cellulose bers, is shown in blue and cellulose bers larger than 700 nm are shown in white. It is noted that the cellulose bers observed in the X-ray CT images were pulps and CMFs while CNFs could not be detected because their width was beyond the spatial resolution of the X-ray CT. Both gures indicated that the composites produced using the pulps without pre-brillation had many large bers with about 10 µm in width, which suggests that nano brillation of the pulps without pre-brillation did not occur during the melt-compounding. However, the composites produced using the re ner-treated pulps had many CNFs with 20-100 nm in width and some CMFs with several µm in width, and these bers were dispersed uniformly in the HDPE matrix. The result suggests that the surface-brillation of pulps using the re ner promoted nano brillation of the pulps as well as uniform dispersion of the CNFs during the melt-compounding.
When the HPH-treated (1 pass and 3 passes) CNFs were used as a feed material, the composites had many CNFs with 20-100 nm in width and some CMFs with several µm in width. Their morphologies were similar to that of the composites produced using the re ner-treated pulps. In contrast, the composites produced using the HPH-treated (10 passes) CNFs had agglomeration of CNFs (see the increase of spherical white regions in the X-ray CT image (Fig. 9)). Figure 10 shows length frequency distribution of the cellulose bers. The composites produced using the re ner-treated pulps and the HPH-treated (1 pass) CNFs had longer bers than those produced using the HPH-treated (3 passes and 10 passes) CNFs. It is noted that the composites produced using the pulps without pre-brillation had long bers (i.e., lengths of all the measured bers were greater than 0.15 mm) though the results are not shown in the gure. Table 3 summarizes relationships between the pre-brillation of pulps, morphology of the composites and properties of the composites. The surface-brillated pulps were nano brillated during the meltcompounding while maintaining a high aspect ratio (i.e., length/width) and uniform dispersion of CNFs in the HDPE matrix, thus giving good performance. The results suggested that the surface-brillated pulp was a suitable form as a feed material for the melt-compounding, and CNF is not a necessary form as a feed material for the melt-compounding. This enables to remove a preliminary step to prepare CNFs from pulps using a time consuming and energy intensive process, such as HPH, grinder and bead mill (Ho et al. 2015). Furthermore, the excessive use (i.e., multiple passes) of the HPH deteriorated the properties of the composites due to the agglomeration of ne CNFs during the melt-compounding. The excessive use of the HPH also reduced length of CNFs. It is speculated that the shorter CNFs moved more easily to agglomerate during the melt-compounding.
The nding of the agglomeration of the ne CNFs indicates that when ne CNFs are used as a feed material, properties of the composites can be improved by mitigating the agglomeration. Meltcompounding is capable of dispersing and distributing CNFs in the polymer matrix by mixing CNFs and molten polymer. However, when CNFs, including chemically modi ed CNFs, have low a nity for the polymer matrix such as HDPE, the mixing action would facilitate separation of the CNFs from the polymer matrix (i.e., agglomeration of the CNFs) rather than dispersion and distribution of the CNFs in the polymer matrix. Furthermore, when ne CNFs such as the HPH-treated (3 passes and 10 passes) CNFs are mixed with molten polymer, the mixing action would promote breakage of the ne CNFs, which would encourage agglomeration of the ne CNFs. Consequently, when ne CNFs are used as a feed material, agglomeration of the ne CNFs during melt-compounding may be prevented (1) by further improving the a nity between the ne CNF and HDPE and/or (2) by avoiding breakage of the ne CNFs during the meltcompounding. Further research is needed to solve the problem of the agglomeration of the ne CNFs.

Conclusion
The dried, ASA-treated cellulose bers with different sizes, ranging from pulps with 20 µm in width to ne CNFs with 20 nm in width, were used as a feed material for melt-compounding in the dry-pulp direct kneading method to fabricate CNF reinforced HDPE. When the pulps without pre-brillation were fed into the twin-screw extruder, the pulps were not nano brillated during the melt-compounding, and the resulting composites had low tensile modulus and strength and high coe cient of thermal expansion. On the other hand, when the pulps were pre-brillated using the re ner, only surface of the pulps was brillated without sacri cing their internal structure. The surface-brillated pulps were nano brillated during the meltcompounding and the CNFs were dispersed uniformly in the HDPE matrix. The composites had much better properties (i.e., much higher tensile modulus and strength, as well as much lower coe cient of thermal expansion) than the composites produced using the pulps without pre-brillation. The 1-pass HPH treatment of the re ner-treated pulps changed the surface-brillated pulps into CNFs. Nevertheless, when the HPH-treated (1 pass) CNFs were used as a feed material for the melt-compounding, the composites had similar morphology and properties to the composites produced using the surfacebrillated pulps (i.e., re ner-treated pulps). The further HPH treatments (i.e., increase in the number of passages of CNFs through the HPH) had the CNFs ner. However, the ner CNFs were shortened and agglomerated during the melt-compounding and thus deteriorated the properties of the composites. The study concludes that the pre-brillation process of the pulps signi cantly affected the morphology and properties of the composites. The form of the cellulose ber suitable for melt-compounding in the drypulp direct kneading method was the surface-brillated pulp, which can be produced cost-effectively using a re ner at an industry scale, rather than the CNF. The study also indicates that when ne CNFs are used as a feed material, properties of the composites can be improved by mitigating agglomeration of the ne CNFs, and suggests that agglomeration of the ne CNFs during melt-compounding may be prevented (1) by further improving the a nity between the ne CNF and HDPE and/or (2) by avoiding breakage of the ne CNFs during the melt-compounding.

Declarations
The authors declare no con icts of interest associated with this manuscript.    X-ray diffraction patterns of cellulose bers.

Figure 6
Typical stress-strain curves of composites from tensile tests.