Effect of alkali modification on the powder flowability of rapeseed straw cellulose fibers

ABSTRACT Powder flowability of natural reinforced fibers has a significant effect on the processing and properties of the composite. Sodium hydroxide modified rapeseed straw reinforced fibers were prepared in four fiber sizes and the powder flowability before and after modification was tested. The results showed that the alkali modification reduced the total flowability index of the fibers by 6–11 within the test range, with the effect on each flowability index in the order of compression > angle of spatula > angle of repose > miformity. This effect was mainly attributed to morphological, chemical and surface characteristics. Moreover, the slimming effect of the alkali modification reduced the average particle size by 9.4% to 30.8%, exacerbating the anisotropy of the morphological structure. Fourier transform infrared (FTIR) spectroscopy and scanning electron microscopy (SEM) analyses showed that pectin, lipids, hemicellulose and impurities were removed, exposing the clean rough fiber surface. Box plot analysis showed that the modification resulted in a uniform and regular distribution of shape factors. This work has valuable implications for the industrial application of straw cellulose fibers in the field of natural fiber composites.


Introduction
The non-biodegradability of petroleum products has prompted researchers to develop environmentally friendly and sustainable natural fiber composites (Singh et al. 2019). Natural fibers have unique properties as reinforcing fibers such as wide source, ecological soundness and low cost in comparison to synthetic fibers (Farhad et al. 2021). However, the presence of some strongly hydrophilic groups such as hydroxyl and carboxyl groups on the surface of these natural fibers makes their interfacial compatibility with the matrix poor, and their mechanical properties also need to be improved (Amroune et al. 2021). Surface modification is therefore used to regenerate natural fibers in order to improve their mechanical properties and interfacial compatibility properties. In fact, surface modification has become a necessary process step for natural fibers in composite applications (Matykiewicz et al. 2021).
The powder flowability of natural reinforcing fibers has a significant impact on production processes in composite applications. The effect of surface modification on flowability forces some processes to be redesigned. Due to changes in angle of repose, for example, we have had to redesign storage sizes and conveying volumes. In fact, powder flowability has been studied for a long time. Researchers have proposed several evaluation methods to characterize powder flowability, such as the angle of repose method, the compression characteristics method and the Carr flowability index evaluation method (Carr 1965;Saker et al. 2019). Several studies have reported on factors that can affect the flowability of powders, such as shape, micromorphology and chemical structure ). However, research on powder flowability is still mainly focused on the energy, food and pharmaceutical industries (Durán-Olivencia et al. 2021), and very little has been reported on reinforced cellulose fibers.
Rapeseed straw fiber is abundantly produced and it is estimated that approximately 20 million tonnes of rapeseed straw are produced annually as an agricultural by-product of rapeseed oil extraction in China (Cong et al. 2019). In recent years, some reports have demonstrated the potential of rapeseed straw cellulose fibers as reinforcing materials for composites (Mirski, Banaszak, and Bekhta 2021). More complete data on the structure-property of reinforced cellulose fibers are urgently needed to facilitate the industrial application. Therefore, this study confirms the effect of alkali modification on the powder flowability and structure of rapeseed straw fibers, providing standardized powder flowability data of the fibers. The mechanisms of the effect were analyzed by chemical, surface characteristics and morphology. FTIR was used to analyze the molecular structure and chemical, SEM was used to characterize the surface micromorphology, and Image J software and box plots were used to analyze the morphological characteristics of the fibers.

Preparation of cellulose fibers from rapeseed straw
Rapeseed straw was obtained from local farms in Shandong Province, China, as shown in Figure 1a. After removing the roots and leaves, the rapeseed straw was washed using distilled water and naturally dried, then pulverized using a pulverizer at 500 mm/r for 2 minutes and prepared into rapeseed straw cellulose fiber (RSF). RSF was classified according to particle size using a standard sieve (according to GB/T 6005-1997). The classification and code are shown in Table 1. RSFs in the 710-250 μm range were tested as plant fibers in this range are often used as reinforcing fibers. On the basis of ensuring that the chemical structure of the cellulose fiber was not significantly damaged, which will be substantiated in the FT-IR analysis, some RSF were chemically modified using a 5% NaOH solution at 50°C for 30 minutes, followed by repeated rinses in distilled water to pH 7. Modified RSF (MRSF) is shown in Figure 1b. The modification process of cellulose fibers by NaOH is considered to be that of graft modification (Kumar et al. 2020). In this process, the sodium cations (Na+) in the NaOH molecule are first separated from the hydroxide anions (OH-) and then grafted to the cellulose components in covalent bonding connections, a diagram of this process is shown in Figure 2. All fibers in the test range, both unmodified (RSFs) and chemically modified (MRSFs), were dried to a constant weight using a drying oven to exclude the effect of moisture content on the flowability test.

SEM characterization
The surface morphological characteristics of the fibers before and after modification were studied by scanning electron microscopy (SEM). The experiments were carried out on a JSM6390A SEM (JEOL, Japan) and the morphological changes of the fibers were observed under high vacuum at an accelerating voltage of 10 kV.

FT-IR measurements
The test was carried out using a FTIR spectrometer (6800-50/NEXUS, ThermoNicolet, USA). The fibers were mixed and ground with KBr and then compacted with a press. During the measurements, the scanning range was 400 cm −1 to 4000 cm −1 and the spectral resolution was 2 cm −1 .

Characterization of morphological features
Morphological characteristics of the fiber specimens were characterized and analyzed by optical microscopy, Image J analysis software and box plot tools. 200 particles of each fiber were randomly selected, and digital images of these fibers magnified 20-50 times were obtained using a digital camera in conjunction with an optical microscope, as shown in Figure 3a. These images were then analyzed using Image J software following the steps below. First, the images were binarised to gray-scale values 8. This reduces the amount of data in the image considerably and no longer involves multiple levels of pixel values, thus highlighting the fiber profile and improving the accuracy of fiber selection and flowability analysis. Then, the fibers were precisely selected by adjusting the upper and lower thresholds in the 0-255 gray-scale range until they were converted to red, as shown in Figure 3b. Finally, the morphological parameters of the fibers are automatically measured by the particle analysis function. The aspect ratio (R) and circularity (C) are calculated using equations 1 and equations 2 based on the measurements and are used to characterize the morphological properties (Vaško 2016;Wiwart et al. 2012). Of these, circularity is used to analyze the distribution characteristics of the fibers using the box plot tool.
Where Major L is the length of the long axis of the best-fit ellipse of the fiber, Minor L is the length of the short axis of the best-fit ellipse of the fiber, A is the projected area of the fiber and P is the perimeter of the fiber profile.

Characterization of flowability
The angle of repose (α), angle of spatula (θ), compression (C f ) and miformity (M f ) of the fibers were measured and assessed to characterize the flowability of the fibers in accordance with Chinese National Standard (CNS) GB/T 31,057.2-2018. The angle of repose was obtained by measuring the cone formed by fibers falling freely and continuously from a multi-stage funnel with a taper of 60° ± 0.5° and an outflow diameter of 10 mm onto a disc with a diameter D α of 80 mm according to Equation 3. The angle of spatula, which reflects the friction between the fibers, was obtained by measuring the average value of the angle between the side of the pile on the spatula and the spatula before and after vibration according to Equation 4. The compression was calculated based on apparent density and tap density according to Equation 5. To measure tap density, 10 g of each fiber sample was packed in a measuring cylinder with a measurement accuracy of ±0.5 cm 2 and shaken for 2 minutes at a vibration frequency of 250 vibrations/min on a compaction densitometer (BT-301, China). The miformity was calculated based on the circularity characteristics of the fibers using equation 6. All tests were averaged five times to reduce errors.
where H α is the height of the fiber cone and D α is the diameter of the bottom surface of the fiber cone.
Where θ 1 and θ 2 are the angles between the side of the pile and the spatula before and after vibration respectively.
where ρ a is the apparent density and ρ t is the tap density.
Where C 60 and C 10 are the circularity values of the fibers arranged from smallest to largest at 60% and 10% respectively.

FT-IR analysis
The chemical structures of the raw (RSF) and modified rapeseed straw fiber (MRSF) samples are similar and their FTIR spectra are shown in Figure 4. It can be seen that although the chemical functional groups of the two samples are not identical (which we will explore further), their patterns are similar overall. The broad absorption peaks at 3419 cm −1 in RSF and 3422 cm −1 in MRSF are associated with the stretching vibration of O-H in the cellulose (Ganan et al. 2008). The C-H stretching vibration absorption peak at 2922 cm −1 in RSF and 2915 cm −1 in MRSF; the C-O stretching vibration peak at 1054 cm −1 in RSF and 1060 cm −1 in MRSF are characteristic bands for cellulose and hemicellulose (Hu et al. 2018). The absorption peaks at 1596 cm −1 and 1423 cm −1 in RSF and 1594 cm −1 and 1419 cm −1 in MRSF are from the stretching vibrations of c=c of the aromatic bone ring in lignin (Maache et al. 2017). The FTIR showed that the fiber samples before and after modification were consistent with celluloseIand had the same chemical structure.
Thus, the effect of chemical modification on the interactions between the fibers, the flowability of the fibers, appears to be more subtle. What is understandable, however, is that morphological factors can affect the flowability of powders, as we introduced in the introduction, so the possible effects of the differences in the chemical functional groups of the two fibers on their morphological characteristics are further explored.
The FTIR patterns of the two fibers differ in at least two places, however, these differences may have opposite effects on the shape and flowability of the fibers from the known mechanism. On the one hand, the characteristic peak of RSF near 1721 cm −1 disappears after modification. This characteristic peak is formed by the stretching vibrations of carbonyl groups (C=O) in substances such as lipids and pectin in the chemical structure of the fiber (Guerra et al. 2021). Its disappearance indicates that these substances are removed from the fiber surface by the alkali solution. This may give a slimming effect to the fibers, making them smaller in diameter. On the other hand, the characteristic peak of RSF near 1240 cm −1 , which originates from the C-O stretching vibration in lignin and hemicellulose (Shanmugasundaram, Rajendran, and Ramkumar 2018), is weakened by the modification, indicating that some of the lignin and hemicellulose between the cellulose molecules is removed by the lye. This causes the fibers to appear protofibrillated and loosely swell (Udomkichdecha, Chiarakorn, and Potiyaraj 2002), possibly making them larger in diameter. It is interesting that two opposite possible effects of slimming and swelling are obtained. Which effect is the dominant one? We will keep this question in mind and discuss it during the particle size analysis.

SEM analysis
The surface morphology of the raw and modified rapeseed straw fibers is shown in Figure 5. As shown in Figure 5a, the raw fiber surface was tightly wrapped by components such as pectin and wax, smooth and with some impurity particles. After the alkali modification, as shown in Figure 5b, the surface of the fibers showed a significant reduction in impurities, an increase in surface roughness and more pronounced and uniform longitudinal grooves. The main reason for these changes is the removal of waxes, pectin, hemicellulose and lignin from the fiber surface by the chemical solution (Farhad et al. 2021). The roughness and cleanliness of the natural fiber surface is desired by the chemical modification treatment, as this may lead to mechanical interlocking and thus significantly improve the interfacial compatibility between the natural fiber and the material matrix during the preparation of the composite (Ma et al. 2019), while the increased friction between the fibers due to these surface features may reduce the flowability.

Particle size analysis
Particle size have an impact on the force transfer processes between fibers such as flow and accumulation (Xiu et al. 2020). Major L, Minor L and Aspect Ratio were used to visually describe the fiber particle size and shape. More than 200 measurements of each fiber size were averaged to eliminate as much as possible the measurement error caused by the anisotropy of the spatial morphological structure of individual fibers, as shown in Figure 6.
As can be seen in Figure 6, the chemical modification resulted in a significant change in the particle size. The results show that the lye modification reduced the particle size and that the slimming effect was greater than the swelling effect. The change is attributed to the removal of pectin, wax, hemicellulose and lignin from the fiber surface by the chemical modification treatment. As shown in Figure 6a, the average value of Minor L for the modified fibers in each particle size range was smaller compared to the raw fibers. Of the four particle size ranges, MRSF35 (0.307 mm) had the largest reduction of 30.8% compared to RSF35 (0.443 mm) and the smallest reduction was 9.4% for MRSF25 (0.236 mm) compared to RSF25 (0.261 mm). Figure 6b follows a similar trend to Figure 6a, with the average value of Major L for the fibers also being smaller after modification, with the fibers shrinking in size in chemical modification. Moveover, the slimming effect caused an increase in the aspect ratio of the fibers as shown in Figure 6c. The results suggest that chemical modification exacerbates the anisotropy of the spatial morphological structure of the fibers. In general, the greater the anisotropy, the greater the difference from the spherical shape, the lower the flowability of the fiber particles (Chen et al. 2019), so the flowability of the modified fibers is perhaps lower than before the modification.

Box plot analysis
The distribution characteristics of the shape factor are an important indicator for assessing the powder flowability (Sousa and Ferreira 2019). The circularity distribution characteristics of the fibers were therefore analyzed by means of box plots. The values of these circularity are arranged in ascending order, with the maximum value, the three quartile values (Q3, Q2, Q1) and the minimum value being used to produce the box plot as shown in Figure 7. The box length depends on the interquartile distance (IQR, the difference between Q3 and Q1), with shorter boxes indicating a more concentrated distribution.
The results show that the chemical modification treatment resulted in a concentrated and regular distribution of the morphological factors. Figure 7 shows that most of the boxes were significantly shorter after the modification. 35.3% of the box lengths were shortened for MRSF60 compared to RSF60, 17.9% for MRSF35 compared to RSF35 and 18% for MRSF25 compared to RSF25. Moveover, the relative position and length difference of these boxes has changed after the modification. The height of the top edge (Q3), bottom edge (Q1) and median (Q2) of the modified boxes showed a gradual increase in height with decreasing particle size, and the length difference between the boxes decreased. These changes are probably the result of repeated stirring and rinsing during the modification treatment. Some of the fibers that were agglomerated by van der Waals forces were dispersed and those of them whose morphological characteristics were misclassified and mismeasured were reevaluated (Ma et al. 2018). In conclusion, the concentrated distribution resulting from the modification treatment may have improved the flowability of the fibers due to the increased miformity, but the extent of the improvement is to be determined by flowability tests.

Flowability analysis
Powder flow is a process with complex force transfer and is related to many factors such as particle size, shape and surface morphology, so powder flowability is generally expressed through multiple physical property measurements. Therefore, according to the flowability index table in GB/T31057.2-2018, the flowability of fibers is evaluated by converting four physical property measurements, including angle of repose (F1), compression (F2), angle of spatula (F3) and miformity (F4), into flowability indices. These four measurements are shown in Table 2 and the flowability indices are shown in Figure 8. The higher the index, the better the flowability.
The results show that the total flowability index of the modified fibers is lower than that of the raw fibers in all particle size ranges. This can be attributed to the changes in chemical and surface morphology and morphological structure caused by the alkali modification treatment. Surface roughening, removal of surface lubricants (waxes, pectins) and increase in anisotropy of the spatial morphological structure due to the modification are the main reasons for the decrease in the flowability. MRSF50 showed the greatest decrease (11) in total flowability index with a decrease of 16.9% and MRSF25 showed the least decrease (6) with a decrease of 8.5%. Moreover, the flowability indices of rape straw fiber were compared with those of other reported non-chemically modified crop straw fibers as follows. The flowability index of RSF60 (60) was similar to that of corn straw (61.5); MRSF60 (52) was lower than corn straw, similar of soybean straw (53.5) and higher than wheat straw  (44.5). The angle of repose of RSF35 (54.0°) was lower than that of corn straw (58.9°); MRSF35 (58.5°) was similar of corn straw (Guo, Chen, and Liu 2012). Furthermore, it can be seen from Figure 8 that the four indices F1, F2, F3 and F4 contribute differently to the total flowability index, due to the different focus of these indices on the evaluation of flowability and the different force fields in which the indices were measured. Further, these indices have a different focus on the evaluation of the effect of fiber flowability on actual processing (Hao 2015). Therefore, the difference between the pre-modification (RSFs) and post-modification (MRSFs) values of these four indices are presented in the last bar (Rs-MRs) and are used to assess the extent to which NaOH modification affects these flowability indices.
The results show that the alkali modifications affected the flowability indices of the fibers in the order of F2 > F3 > F1 > F4. The F2 values decreased by a total of 13.5, the largest decrease of the four indices, indicating that the modification reduced the compressive flow and cohesiveness of the fibers. This change may have a significant effect on the pressing process of the composite in practice, for example, the density of the finished pressed product under the original pressing conditions may change as a result, and a deeper cavity in the press mold may be required to accommodate the same mass fraction of the modified fiber as before the modification. Both F3 and F1 are affected by the modification to a lesser extent than F2, with a decrease of 12 and 10 respectively. F3 focuses on the friction between fibers, reflecting the shear resistance of the fiber fluid, while F1 focuses on the flow in a gravitational field (Penkavova et al. 2021). Finally, it is interesting to note that the modification has no significant effect on the value of F4 (both 23), although there was a difference in the miformity measurements before and after modification, this difference was not of sufficient magnitude to cause a change in the flowability index of the fibers, as the high miformity due to sieving and grading was the more important factor affecting the F4 values compared to the modification.

Conclusion
Four different particle sizes of rapeseed straw reinforced cellulose fibers were modified with sodium hydroxide and standardized powder flowability data were provided to investigate the effect of modification on the powder flowability of the fibers. The following conclusions were drawn. The mechanisms by which modification affects the powder flowability of cellulose fibers include mainly chemical, surface characteristics and morphology. The alkali modification reduced the total flowability index of the fibers by 6-11, with the effect of modification on each flowability index was in the order of F2(compression) > F3(angle of spatula) > F1(angle of repose) > F4(miformity). Moreover, Morphological analysis showed that the slimming effect of the alkali modification reduced the average particle size by 9.4% to 30.8% and increased the anisotropy of the morphological structure. FTIR analysis showed that this slimming effect was attributed to the removal of pectin, lipids, and hemicellulose from the fiber surface by the alkali solution. Furthermore, SEM images showed that the rough surface of the modified fibers reduced the powder flowability. The box plot analysis showed that the chemical modification treatment resulted in a concentrated and regular distribution of the shape factors.
The results of this study only apply fully to rapeseed straw fibers, due to the large differences in the flowability of the various species of natural fibers. Even for rapeseed straw fibers, the effect of surface chemistry cannot be completely ruled out, although this effect appears to be more subtle than fiber morphology, as there is no guarantee that the surface chemistry will be uniform.