Characterization of Natural Cellulosic Fiber Obtained from the Flower Heads of Milk Thistle (Silybum marianum) as a Potential Polymer Reinforcement Material

ABSTRACT

Natural fibers are being investigated as alternatives to synthetic fibers in composite materials. In this paper, the physical, chemical, thermal, and morphological characteristics of fibers derived from milk thistle flower heads were characterized as potential polymer reinforcement material using Fourier transform infrared (FTIR) spectroscopy, X-ray diffraction (XRD), thermogravimetric analysis (TGA), and scanning electron microscopy (SEM). The fibers were found to contain 42.9% cellulose, 21.2% hemicellulose, and a negligible lignin content. The fiber constituent functional groups were similar to those observed for other natural fibers. The crystallite size and the crystallinity index were found to be 2.23 nm and 84 ± 3%, respectively. The thermal stability of the fibers was found to be 220°C. The fibers start rapidly degrading above 290°C. Morphologically, the fibers appeared to have a nonporous surface with well-structured longitudinal parallel grooves. Cross sectionally, the fibers consisted of multiple hollow lumens interconnected with cell walls. The hollowed nature of the fibers explains the low value (0.33 ± 0.04 g mL⁻¹) of their density. The tensile strength of the fibers was determined to be 22 ± 1 MPa. In conclusion, the fibers of the flower heads of milk thistle can be used as effective reinforcements in polymer matrices.

摘要

天然纤维作为合成纤维在复合材料中的替代品正在被研究中. 本文利用傅里叶变换红外光谱（FTIR）、X射线衍射（XRD）、热重分析（TGA）和扫描电子显微镜（SEM），对乳蓟花头纤维的物理、化学、热学和形态特征进行了表征，认为其是一种潜在的聚合物增强材料. 发现纤维含有42.9%的纤维素、21.2%的半纤维素和可忽略不计的木质素含量. 纤维组成官能团与观察到的其他天然纤维的组成官能团相似. 晶粒尺寸和结晶度指数分别为2.23 nm和84 ± 3%. 发现纤维的热稳定性为220°C. 纤维在290°C以上开始迅速降解. 从形态上看，纤维表面无孔，具有结构良好的纵向平行凹槽. 横截面上，纤维由多个与细胞壁互连的中空管腔组成. 纤维的中空特性解释了其密度的低值（0.33 ± 0.04 gmL − 1）. 纤维的抗拉强度测定为22 ± 1MPa. 总之，乳蓟花头的纤维可以作为聚合物基质中的有效增强剂.

KEYWORDS:

• Natural fiber
• cellulose
• polymer enforcement
• characterization
• chemical analysis
• Silybum marianum

关键词:

• 天然纤维
• 纤维素
• 聚合物强制执行
• 刻画
• 化学分析
• 水飞蓟

Introduction

Natural fibers have recently gained popularity over synthetic nonbiodegradable, hazardous fibers (Thakur, Thakur, and Gupta 2014). Natural fibers have several desirable properties, including renewability, sustainability, and eco-friendliness, that have made them an excellent alternative for synthetic fibers in many applications, such as automobiles and bio-medical applications (Indran et al. 2022). Natural fibers are divided based on their origin to mineral, animal, or plant-based fibers. Cellulosic and lignocellulosic plant-based fibers include different classes such as bast, straw, seed, grass, leaf, and wood fibers (Thakur, Thakur, and Gupta 2014). Cellulosic fibers are known for their low density, low tool abrasiveness, safety, low production cost, wide availability, high specific strength, and stiffness, which make these fibers candidates for reinforcing polymers in the manufacture of fiber-reinforced composites (Indran et al. 2022). However, the characteristics of natural fibers derived from plants may differ based on the plant type, the geographical location, and the environmental conditions (Jawaid and Abdul Khalil 2011; Raghunathan et al. 2022). Therefore, to determine the suitability of natural fibers for certain applications, their physical and chemical characteristics need to be assessed (Indran et al. 2022). For polymer reinforcement matrices, the presence of cellulose, the crystallinity and thermal properties of the fibers are the most important factors that determine the suitability of natural fibers for such applications (Jawaid and Abdul Khalil 2011; Raghunathan et al. 2022). Different characterization approaches have previously been used to study natural fibers derived from jute, sisal, flax, hemp, coconut (Karimah et al. 2021), Calotropis gigantea (Ganeshan et al. 2018), pampas (A. Khan et al. 2021) and many more. The high demand for eco-friendly fibers for use in polymer reinforcement applications urged many researchers to investigate the physical and chemical properties of chemically treated natural fibers. The chemical treatment of natural fibers may enhance their specifications for certain applications. (Raghunathan et al. 2022), investigated the physical and chemical properties of untreated, NaOH-treated and HCl-treated bark fibers of Vachellia farnesiana. The results showed that, the NaOH treatment of Vachellia farnesiana fibers enhanced their properties for use as a reinforcement material in light to medium weight applications. In the same context, (Vijay et al. 2021) found that silane-treated Citrullus lanatus fibers can be effectively used for the reinforcement of brake pads. To meet the growing need for natural fibers, other research teams carry out investigations to find novel natural fibers with desirable properties for broad fields of application (Alshammari et al. 2019).

Milk thistle (Silybum marianum L. Gaertn., Asteraceae) is a wild thorny annual plant widely distributed in the Mediterranean (Sulas, Ventura, and Murgia 2008). Although milk thistle is of medicinal and economical value, it has been considered a highly invasive weed (N. Khan et al. 2022). Moreover, milk thistle has been reported to be a valuable source of biofuels and hence, energy (Ledda et al. 2013).

The flower heads of milk thistle house the seeds. They have long pappus that serve as parachutes to support seed dispersal by wind. The seeds usually shed during February and March of every year (M. A. Khan, Blackshaw, and Marwat 2009). When the seeds are blown or removed, the flower heads are left with empty hairy seed beds in the flower bulb as shown in Figure 1. When milk thistle dries, in June, it is removed from the ground by tillage or chemical treatment and is disposed as waste or incinerated after drying (M. A. Khan, Blackshaw, and Marwat 2009). Thus, dry milk thistle could be a valuable economical source of natural fibers.

Figure 1. a) a dry flower head of milk thistle with a bulb filled with hairy fibers in the seed bed surrounded by spine-tipped bracts at the base. b) the flower bulbs after removal of the stems and bracts. c) Hairy fibers of the flower bulb.

Materials and methods

The materials and methods section are reported as a supplementary online resource.

Results and discussion

Fourier transform infrared (FTIR) spectroscopy

The characteristic functional groups present in the fibers obtained from the flower heads of milk thistle were identified by the absorption bands in the FTIR spectrum (Figure 2 and Table 1). The broad peak centered near 3395 cm⁻¹ in the region between 3560 and 3200 cm⁻¹ (peak 1; Figure 2) is a characteristic feature of natural fibers. This signal represents the – OH stretching vibration of the hydroxyl groups present in biopolymers, such as cellulose, hemicelluloses, lignin, and other hydrogen-bonded hydroxyl groups of α-cellulose in plant-derived fibers (NagarajaGanesh and Muralikannan 2016). The peak at 2912 cm⁻¹ (peak 2; Figure 2) corresponds to C – H stretching vibration (Sain and Panthapulakkal 2006). The small peak at 2135 cm⁻¹ (peak 3; Figure 2) is due to the asymmetrical C – N stretching vibrations that may indicate the presence of non-cellulosic materials such as waxes (Guo et al. 2006). The peak centered at 1737 cm⁻¹ (peak 4; Figure 2) is attributed to the C=O stretching vibration of the acetyl groups and ester groups of the hemicellulose or the ester linkage of the carboxylic acid group in the ferulic and p-coumaric acids of hemicellulose (Guo et al. 2006). This peak has been reported for other natural fibers, such as luffa and Calotropis gigantea fibers, to confirm the presence of hemicellulose in the fiber sample (Ganeshan et al. 2018; YWang and Shen 2012). The peak at 1602 cm⁻¹ (peak 5; Figure 2) was previously detected for raw natural fibers. It corresponds to the C=O stretching vibration of the amido group in hemicellulose (Y. Wang and Shen 2012). Peaks at 1504 cm⁻¹ and 1427 cm⁻¹ (peaks 6 and 7; Figure 2) correspond to the C=C stretching frequency of an aromatic ring of lignin (Zhang et al. 2016). The peak at 1247 cm⁻¹ (peak 8; Figure 2) represents the C – O stretching vibration of an acetyl group in lignin and/or hemicellulose (Santhanam et al. 2016). Signals for cellulose C – O–C bridges and C – O bond stretching were detected at 1165 and 1118 cm⁻¹ (peaks 9 and 10; Figure 2), respectively (Ávila Ramírez et al. 2014). Finally, the band detected at 899 cm⁻¹ (peak 11; Figure 2) has been typically reported for β-glycosidic linkages between two monosaccharides (Ávila Ramírez et al. 2014).

Figure 2. FTIR spectrum of the fibers obtained from the flower heads of milk thistle showing the % transmittance vs. wavenumber in cm⁻¹. Identified peaks are numbered 1 through 11.

Table 1. The FTIR-obtained absorption band wavenumbers (cm⁻¹), with the assigned functional groups and fiber components.

Wavenumber (cm^−1)	Functional group	Fiber component	References
3395	–OH	Cellulose, hemicellulose, lignin and α-cellulose	NagarajaGanesh and Muralikannan (2016)
2912	C – H	Cellulose and hemicellulose	Sain and Panthapulakkal (2006)
2135	C – N	Non-cellulosic materials	Guo et al. (2006)
1737	C=O	Hemicellulose	Ghali et al. (2012)
1602	C=O	Hemicellulose	Madhu et al. (2020)
1504	C=C	Lignin	Zhang et al. (2016)
1427	C=C	Lignin	Zhang et al. (2016)
1247	C – O	lignin and/or hemicellulose	Santhanam et al. (2016)
1165	C-O-C	Cellulose	Ávila Ramírez et al. (2014)
1118	C – O	Cellulose	Ávila Ramírez et al. (2014)
899	β-glycosidic linkages	Cellulose and hemicellulose	Ávila Ramírez et al. (2014)

Fiber constituents

The mass percentages of hemicellulose, cellulose, and lignin content of the fibers were determined by a standard raw fiber determination method (Goering 1970). The experiments were performed in triplicates. The result of this determination is expressed as g of dry content of hemicellulose, cellulose, and lignin in a 100 g fiber sample. The average percentage of crude neutral detergent fiber (NDF) that is the hemicellulose, cellulose, and lignin in the dry matter of the analyzed fiber samples was 64.0 ± 0.3%. The average content of cellulose was 42.9 ± 0.6%. Although some of the peaks detected by FTIR were attributed to the presence of lignin, no weighable amount of lignin was found in the fiber samples. The hemicellulose content was on average 21.2 ± 0.3%. The average cellulose content in the fibers obtained from the flower heads of milk thistle in the current study is similar to values previously reported for natural fibers for potential use in polymer reinforcement, such as Abutilon indicum and Leucas aspera (56.12% and 50.7%, respectively) (ArunRamnath et al. 2023; Vijay et al. 2021). The strength of cellulosic fibers is determined by the cellulose content in the polymer. The other components, such as hemicellulose, starch, lignin, and proteins, are amorphous polymers and lack the strength of cellulose (Hossain et al. 2013). Moreover, moisture containment and degradation of polymer composites was found to be related to lignin content (Indran et al. 2022). Therefore, the low lignin content has been considered advantageous in natural fibers for applications in polymer reinforcement such as sisal, Sida rhumbifolia, flax and hemp-obtained fibers (12.8%, 7.48%, 3% and 4%, respectively) (Indran et al. 2022; Moshi et al. 2019).

X-ray diffraction (XRD)

The X-ray diffraction pattern of the fibers obtained from the flower heads of milk thistle confirms the presence of both amorphous and crystalline content in the sample (Figure 3) (Park et al. 2010). In the present study, the 2θ for the amorphous peak was found to be near 15.74°, whereas the 2θ peak at 22.49°Corresponded to the crystalline content of cellulose in the fiber. Similar results have been reported in the literature for cellulosic fibers (Park et al. 2010; Rocky and Thompson 2021). To mention a few, the 2θ peaks that correspond to the crystalline cellulose in fibers obtained from Cocos nucifera, Calotropis gigantea, and rice husk were 22.19°, 22.8, and 22.17°, respectively (Ganeshan et al. 2018). The crystallinity index of the fibers studied herein was found to be 84 ± 3%, whereas the crystallite size of the fibers was found to be 2.23 nm (from Scherrer’s formula). Similar crystallite sizes have been reported for natural fibers. For example, Calotropis gigantea, Cocos nucifera, cotton, and ramie were reported to have crystallite size of 2.05, 4.46, 5.5, and 3.31 nm, respectively (Ganeshan et al. 2018; Poletto, Ornaghi, and Zattera 2014). The crystallinity indices of natural fibers vary widely depending on the fiber source. Some values reported in the literature include those of flax, buriti, and sisal that were found to equal 80.0%, 71.2%, and 57.3%, respectively (Poletto, Ornaghi, and Zattera 2014; Tamanna et al. 2021; Yueping et al. 2010).

Figure 3. The XRD pattern of the raw fibers obtained from the flower heads of milk thistle. The peak at 2θ 15.74°Corresponds to the amorphous constituents of the fibers, whereas the 2θ peak at 22.49°Corresponds to the crystalline content of cellulose in the fiber. I₀₀₂ is the intensity of the 002-lattice diffraction peak corresponding to cellulose and I_am is the peak intensity of the amorphous part of the sample.

Thermogravimetric analysis (TGA)

The TGA and differential thermogravimetry (DTG) curves are shown in Figure 4. DTG is the derivative of the TGA curve, and it allows better interpretation of the data. Thermogravimetric analysis is used to study the thermal degradation behavior of natural fibers and their constituents. The thermal stability of a fiber and its composites is highly influenced by thermal expansion or contraction, which affects the rate and volume of moisture absorption leading to fiber swelling. This determines its mechanical properties (Monteiro et al. 2012). The process of degradation includes dehydration combined with the emission of volatile components along with rapid weight loss due to oxidative decomposition. Oxidative decomposition corresponding to the formation of char as the temperature is increased. An initial 6.5% weight loss occurred when the temperature was raised from room temperature to 110°C due to the evaporation of the water content and other volatile extractives present in the fiber. Maximum weight loss due to dehydration was detected at 60°C. This is a common phenomenon that is detected for almost all plant fibers on account of their hydrophilic nature (A. Khan et al. 2021). At about 220°C, a second mass loss was initiated due to the degradation of hemicellulose. The maximum weight loss due to the degradation of hemicellulose was detected at 274°C. In agreement, the initial decomposition temperature of Calotropis gigantea, Dichrostachys cinerea, Lygeum spartum, Heteropogon contortus, bagasse, kenaf, cotton stalk, rice husk, and wood staple, after dehydration, is 220, 284, 220, 240, 223, 219, 221, 223, and 220°C, respectively (Ganeshan et al. 2018; Moshi et al. 2019). Hemicellulose degradation was followed by the degradation of cellulose at a rapid pace from 290°C up to 390°C. The maximum weight loss attributed to cellulose occurred at 340°C. Similarly, maximum degradation temperatures of 340, 345, 359, 363, and 365°C have been reported for cellulose in sisal, flax, okra, jute, and curaua fibers, respectively (Poletto, Ornaghi, and Zattera 2014; Tamanna et al. 2021). About 63% of the thermal decomposition took place in the temperature region between 220°C and 390°C. This percentage agrees with the combined mass percentages of cellulose and hemicellulose obtained using the AOAC method shown in Table 2. The degradation of cellulose was followed by another mass loss at about 400°C, which may have been due to the thermal degradation of lignin and possibly other constituents in the fiber, such as wax.

Figure 4. TGA plot (represented by a solid line) and DTG curve (represented by a dashed line). Three main stages were detected: The maximum mass loss due to dehydration occurred at 60°C, the maximum mass loss due to degradation of hemicellulose took place at 274°C, and the maximum mass loss attributed to the degradation of cellulose occurred at 340°C.

Table 2. The mass percentages of fiber constituents obtained via neutral detergent solution, acid detergent solution, and acid detergent lignin analyses. Values from three experiments, averages and standard deviations are provided.

Experiment #	NDF%	ADF%	ADL%	%Cellulose	%Hemicellulose
1	63.71	42.22	–	42.22	21.49
2	64.40	43.44	–	43.44	20.96
3	64.02	42.97	–	42.97	21.05
Average	64.04	42.88	–	42.88	21.17
Standard deviation	0.34	0.62	–	0.62	0.28
NDF: neutral detergent fiber; ADF: acid detergent fiber; ADL: acid detergent lignin

Scanning electron microscopy (SEM)

Figure 5 shows SEM images of the fibers obtained from the flower heads of milk thistle. Figure 5A is an image showing the outer surface of a single fiber. Unlike many other natural fibers, the outer surface of the fiber appeared smooth with parallel lines of longitudinal grooves. It is worth mentioning that longitudinally grooved micro- and nano-fibers are currently attracting attention because of their unique morphology and properties (Wu, Xue, and Xia 2020); however, they are currently made synthetically from polymers such as polycaprolactone, polylactic acid, polystyrene, and cellulose acetate butyrate by utilizing methods such as electrospinning (Huang et al. 2011). Grooved fibers have been synthesized for valuable applications including tissue engineering and peripheral nerve repair (Huang et al. 2011). In addition, the grooves have been found to increase the roughness of synthetic fiber surfaces, which was found to improve the electrical properties of friction nanogenerators and piezoelectric materials (Fan et al. 2012). Hence, finding natural longitudinally grooved fibers would attract the interest of scientists in different fields.

Figure 5. SEM images of a single raw fiber obtained from the flower heads of milk thistle. a) a single fiber outer surface at a magnification of 600× showing the longitudinal grooves on the outer surface and the thorny structures on the sides of the fiber. b) a single fiber outer surface at a magnification of 1200× showing impurities on the surface in addition to the longitudinal grooves on the outer surface and the thorny structures on the sides of the fiber. c) a cross-sectional view of a cryo-fractured fiber at a magnification of 2400× showing the hollow lumens interconnected by cell walls. d) a greater magnification of the hollow lumens, 5000×, showing the irregular polygonal internal lumen shapes.

A more magnified image of the surface is shown in Figure 5B, in which impurities or debris were observed on the surface and might have remained on the surface despite washing the fibers. A very interesting finding that should be highlighted here is the presence of thorny structures on the sides of the fibers. The presence of these thorns can improve the mechanical anchoring and thus improve the quality of the fiber/matrix interface when using the fibers in composite materials.

The cross-sectional view of the fiber shows that it is in the form of a bundle of fibrils combined by means of middle lamella (Figure 5C). All fibers showed similar structures, comprising a lumen in the center that is surrounded by cell walls. This is similar to many other natural fibers including those obtained from jute, flax, sisal, and ramie (Hamad et al. 2017); however, the shape, the cell wall thickness, and the internal lumen size and shape vary substantially between natural fibers from different plants (Hamad et al. 2017). Similar to flax and ramie fibers, the fibers obtained from the flower heads of milk thistle have irregular polygonal internal lumen shapes (Hamad et al. 2017). The hollow lumen structure decreases the density of the fiber and is responsible for other favorable properties of natural fibers including acoustic insulation (K. Liu et al. 2012). Furthermore, these cavities can improve the mechanical anchoring when fibers are used in polymer composites (K. Liu et al. 2012). The average diameter and length of 60 random fibers were 70 ± 5 μm and 1.2 ± 0.2 cm, respectively.

Density

The density of the fibers was found to be 0.33 ± 0.04 g mL⁻¹. This value is lower than most densities reported in the literature for natural fibers. However, the density is closest to that of kapok fibers (0.29 g mL⁻¹), previously reported as potential fibers for polymer reinforcement (Komuraiah, Shyam Kumar, and Durga Prasad 2014). The very low density of kapok fibers was previously attributed to the huge empty lumen structure of the fibers (Meiwu, Hong, and Weidong 2010). Posidonia fibers have been found to have a low density of 0.42 g mL⁻¹ for similar reasons (Allegue, Zidi, and Sghaier 2014). The large hollow lumen of the fibers is believed to be the reason behind the low density of the fibers obtained from the flower heads of milk thistle in this study.

Single fiber tensile strength

The average tensile strength of the fibers obtained from the flower heads of milk thistle was 22 ± 1 MPa. Although the tensile strength of the fibers obtained in the herein study is lower than that previously reported for sisal, kenaf bast, jute and flax ((274–526), (427–519), (400–773) and (500–1500) MPa, respectively) (Moshi et al. 2019). The value has previously been reported acceptable for natural fibers used in polymer reinforcement applications, and is comparable to the tensile strength of pampas, Ziziphus mauritiana and Catharanthus roseus (20, 25.89 and 27.02 MPa, respectively) previously reported suitable for such applications (Khan et al. 2021; Vinod et al. 2022).

Overall, the novel fibers from the flower heads of milk thistle were easily obtained without the need for retting or chemical treatment. The properties of the fibers showed the feasibility of their use as an alternative to expensive natural fibers and synthetic fibers in polymer reinforcement. This work will be initiative for researchers who work on fiber reinforced polymer composites to utilize the fibers obtained from the flower heads of milk thistle for the manufacture of greener composite materials for sustainable and cleaner production in structural applications.

Conclusions

This study discloses the physical, chemical, thermal, and morphological characteristics of the fibers obtained from the flower heads of milk thistle. The FTIR results were consistent with what is expected for natural fibers containing cellulose, hemicellulose, and lignin. The XRD results showed that the crystalline index and the crystallite size were similar to those of fibers obtained from other plant sources that are used for polymer reinforcement. The thermal stability of the fibers and the thermal degradation characteristics were also similar to other natural fibers commonly used in polymer enforcement applications. Hence, it can be concluded that the fibers derived from the flower heads of milk thistle are suitable for use in polymer composites. SEM images of the fibers revealed some favorable morphological properties of the fibers, including having a longitudinally grooved surface, the presence of thorny anchors on the fiber sides, and the hollowed structure of the fibers that causes the fibers to have a low density that makes it suitable for lightweight applications. The tensile strength was found to be acceptable for fibers in use for polymer reinforcement applications.

Highlights

• For the first time, fibers from the flower heads of milk thistle were characterized and identified as a potential polymer reinforcement material.

• The fibers obtained from the flower heads of milk thistle were found to have some favorable morphological properties, including having a longitudinally grooved surface, the presence of thorny anchors on the fiber sides, and the hollowed structure of the fibers that would enhance the anchoring of the fibers in polymer composites.

• The fibers obtained from the flower heads of milk thistle were found to have a low density, making it suitable for light weight applications.

Acknowledgements

The authors would like to thank Eng. Rawan Hayajneh from the Nanotechnology Institute at JUST for her help in acquiring the SEM images. The authors are also grateful to Ramzi Munaiem for conducting the FTIR experiments.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This research received funding from the Deanship of Scientific Research at the Jordan University of Science and Technology under grant number 20220605.