Physico-Chemical, Mechanical and Thermal Properties of Novel Cellulosic Fiber Extracted from the Bark of Tithonia Diversifolia

ABSTRACT The global demand for plastics has been continuously increasing at a rapid rate and their disposal largely pollutes the atmosphere and oceans. Natural fibers possess the inherent characteristics required for plastic making and their disposal causes less harm to the environment. In this paper, physicochemical, thermal, mechanical and morphological characteristics of novel cellulosic fiber extracted from the bark of Tithonia diversifolia (TD) investigated through various characterization tests have been reported. TDFs were found to possess higher cellulose (63.9 wt.%) and high lignin content (19.3 wt.%) which make them stronger and more stable. Lower density (1440 kg/m3), higher crystallinity index (51.84%), higher tensile strength (243–386 MPa) and higher thermal stability (up to 337.17°C) make TDFs more suitable for making green composites. The morphology of the TD fiber surface observed through field emission scanning electron microscopy (FESEM) was found to be rougher with some serrations which facilitate better bonding characteristics with the polymer matrix.


Introduction
From 1950 to 2015, 8.3 billion tonnes of plastics were produced globally, out of which 4.9 billion tonnes of plastic waste were generated and disposed of through landfills (Barnes 2019). About 8 million tonnes of plastic waste are dumped yearly into the ocean, and this results in the release of toxic chemicals such as polystyrene and Bisphenol-A during plastic degradation which adversely affects the life of marine habitats. The global demand for plastics has been increasing continuously over years, by about twentyfold times since 1964, and this huge demand for plastics is attributed mainly to the growing population and tremendous increase in the demand for automobiles (MacArthur, Waughray, and Stuchtey 2016). The rapid depletion of fossil fuel resources and the lack of environmentally safe disposal procedures for conventional petroleum-based plastics have

Fiber extraction
Tithonia diversifolia identified in Arakkonam town, Ranipet district, TamilNadu, India, was taken for the present investigation. The tree has grown to a height of 2 m approximately and was found to bear daisy-shaped yellow flowers as shown in Figure 1(a,b). The water retting technique (Ilaiya Perumal and Sarala 2020) was adopted for the extraction of Tithonia diversifolia fibers. The stem of the tree was cut to short lengths using a sharp knife as shown in Figure 1(c,d) and soaked in water maintained at room temperature for a period of 14 days as shown in Figure 1(e). During this period, microbes act upon the bark of the stem and facilitate easy separation of the inner fibrous layer from the outer layer through manual peeling. The wet fibers thus extracted were allowed to dry completely by placing them under direct sunlight for a period of 24hrs as shown in Figure 1 (f).

Chemical analysis
The chemical constituents of TDFs were determined through standard test methods. Kurschner and Hoffer's method was adopted to assess the quantity of cellulose present in TDFs (Kurschner and Hoffer 1993). A thoroughly washed and oven-dried fiber sample weighing approximately 1 g was immersed in a mixture of 20 mL of 95% ethanol and 80 mL of 69% nitric acid and heated up to 110°C for 1 hour. The fibers were placed in an oven maintained at 60°C until constant weight is attained and the insoluble fraction thus left out represents the cellulose present in TDFs. The hemicellulose content in TDFs was determined through the NFT standard 12-008 method (Manivel et al. 2021). The powdered fiber samples were heated in hydrobromic acid, and the hemicellulose present in the sample was transformed to furfural and was separated out and measured using distillation and by applying the spectrophotometric technique. The lignin content in TDFs was determined by applying APPITA P11S-78 method. The crushed fiber sample was hydrolyzed using 72% of H 2 SO 4 for one hour and at 30°C. The sample was then mixed with methylene chloride and placed in an autoclave maintained at 125°C for 1 hour. The lignin present in the sample was then filtered out through retentate and weighed. The wax content in TDFs was quantified through the Soxhlet extraction process (Conrad 1944).
The wax content present in TDFs was then determined using the equation (1) Waxð%Þ (1) W f -Weight of oven dried sample (g) W w -Weight of wax extracted from the sample (g) The ash content of TDFs was determined as per ASTM E 1755-01 standard. The crushed fiber sample was completely burnt to ash by placing the sample inside a calorimeter maintained at 250°C. The quantity of ash finally left out as residue was then weighed.
The density of TDFs was determined by applying the Mettler Toledo xsz05 balance method (Gopinath et al. 2016). TDF sample was completely dried by placing them in an airtight non hygroscopic desiccator containing CaCl 2 . The microbubbles, if any, present in the fiber sample were completely removed by impregnating the sample in toluene for 2 hours. The fibers were chopped to 10 mm size and placed in a pycnometer. The density of TDFs was then determined by using equation (2) where m 1 -mass of empty pycnometer (g) m 2 -mass of pycnometer with fiber sample (g) m 3 -mass of pycnometer with toluene (g) m 4 -mass of pycnometer with fiber sample and toluene(g) ρ T -density of toluene (0.8669 g/cm 3 ). The moisture content present in TDF was determined through the normal weight loss method. A fiber sample weighing 20 mg was placed in an oven maintained at 100°C for 4 hours 30 minutes until constant weight is attained. The percentage of moisture content in TDFs is determined using equation (

Fourier transform infrared spectroscopy
FTIR spectra of TDFs were obtained using an FTIR spectrometer (Model: FTIR-8400S Spectrum, Make: SHIMADZU, Japan). The fiber sample was powdered by using a mortar and pestle tools and mixed with potassium bromide. The mixture was made in pellet form by applying adequate pressure using a hydraulic press machine. Infrared radiation was allowed to an incident on the pelletized sample and FTIR spectra over the wavenumber region 4000 cm −1 -500 cm −1 were generated by maintaining the scan rate of 32 scans/min and resolution of 4 cm −1 .

X-ray diffraction
The molecular structure, percentage of crystallinity and crystallite size of TDFs were analyzed through X-ray diffraction using an X'pert Pro PANalytical system diffractometer over the range of 2θ from 10° to 80° and at step size (2θ) of 0.05°. The generator settings of 30 mA and 45kV and a scan rate of 5°/ min were maintained during the investigation. A monochromatic beam of CuKα radiation of wavelength 0.1542 nm was generated by bombarding copper target using high-energy electrons and allowed to an incident on the sample placed on a rotating platform. The diffracted radiation was then captured by the detector which records the counts of X-rays scattered by the sample. The relative amount of crystalline cellulose out of the total quantity of cellulose which includes both crystalline and amorphous fractions known as crystallinity index(CI) was evaluated through the peak height method proposed by the Segal et al. (1959) (Segal et al. 1959). The value of CI of TDF was thus determined using Segal equation (4): The size of cellulose crystals significantly influences the chemical reactivity and water absorption behavior of the fiber. Scherer's empirical equation given in (5) (Indran and Edwin Raj 2015) was used to determine the crystallite size (CS) of TDFs.
where K = 0.89 is Scherer's constant, β is the peak's full width at half maximum and λ is the wavelength of radiation.

Thermal properties of TDFs
The thermal degradation behavior of TDFs was studied through various techniques such as thermogravimetric (TG), differential thermogravimetric (DTG) and differential thermal analysis (DTA). For this purpose, Jupiter simultaneous thermal analyzer (Model: STA 449F3, NETZSCH, Germany) was used. The procedure involves heating up of powdered fiber sample gradually from room temperature to 1000°C with the rate of temperature maintained at 10°C/min. The combustion process was performed in a nitrogen atmosphere so as to avoid charging effects due to oxidation and highpurity nitrogen gas was continuously passed into the furnace at a flow rate of 20 mL/min.

Tensile properties
The tensile resistance offered by TDFs was determined by performing a single fiber tensile test on twenty numbers of randomly selected TDF samples as per ASTM 3822-07. The fiber samples with a gauge length of 75 mm were subjected to tensile load using the INSTRON Universal testing machine of 5kN capacity load cell. During the test, crosshead speed was maintained at 0.1 mm/min.

Surface morphology
The characteristics of the fiber surface were studied through high-resolution scanning electron microscope images obtained using Schottky Emitter field emission scanning electron microscope (FESEM) (Model:Thermo Scientific Apreo S). The fiber samples were coated with a thin layer of gold to overcome the charging effects of the electron beam. An accelerating voltage of 30kV was maintained during the course of the examination.

Energy-dispersive X-ray spectroscopy (EDX)
The elements present at the fiber surface were analyzed by using an energy-dispersive spectrometer (EDS) analyzer (BRUKER(German), Model: Nano X flash detector with a magnification of 1000× and other settings such as accelerating voltage and working distance maintained at 10.0kV and 6.2 mm.

Atomic force microscopy (AFM)
Atomic force microscopy (AFM) analysis was used to assess the roughness of the TDF. The AFM analysis was performed in the Park XE-70 Model AFM (Korea Make) with XEI image processing software to investigate the data. The parameters such as average surface roughness (Ra), root-meansquare roughness (Rq or Rrms),10-point average roughness (Rz), skewness (Rsk), kurtosis (Rku) and maximum peak-to-valley height (Rt) can be found out from AFM analysis. The working range of the scanner scale in x, y and z-direction is 10 μm x 10 μm x 70 μm resolutions.

Density and diameter of TDFs
The density of TDF (1440 kg/m 3 ) was found to be appreciably lower than other plant fibers such as jute (1460 kg/m 3 ), hemp (1500 kg/m 3 ) but higher than Sida cordifolia (1330 kg/m 3 ), kenaf (1400 kg/m 3 ), Acacia leucophloea (1385 kg/m 3 ), Sida rhombifolia (1320 kg/m 3 ), Cissus vitiginea (1287 kg/m 3 ), Carica papaya (943 kg/m 3 ) and Indian mallow (1330 kg/m 3 ) (Gopinath, Billigraham, and Sathishkumar 2021). Java-based image processing image J software was utilized to find the average TDF diameter on the FESEM micrographs of individual TDF as shown in Figure 8(d). FESEM examination was done on 25 different TDF specimens of each 40 mm in order to determine the diameter along the width. Fifty readings were recorded on each FESEM micrographs, and it was repeated on 25 samples and the mean, standard deviation and standard error of TDF diameter were calculated (Mwaikambo and Ansell 2006). The average diameter of the TDF specimens using image analysis software was found to be 320 µm, and this value is comparable with other natural fibers such as Petiole bark (250-650 µm), palm (400-490 µm), Coccinia grandis(543-621 µm), Phoenix dactylifera (577 ± 83 µm) and Cissus quadrangularis stem (770-870 µm). The correlation between the diameter and density of TDF is that the density tends to increase when the diameter values become smaller. An increase in density, which is directly related to the decrease in diameter, can be attributed to fewer defects in the TDF. With appropriate surface treatments, the diameter can further be reduced to have fewer defects in the fiber which directly influences the mechanical properties of the TDF (Indran and Edwin Raj 2015). The density of TDF is 1440 kg/m 3 , which confers that the fiber can be reinforced with matrix for making lightweight polymer and hybrid composites.

X-ray diffraction
The X-ray diffractogram of the TD fiber sample is shown in Figure 2. The peak at 2θ = 22.28° corresponds to the crystallographic plane (002) of cellulose and that observed at 2θ = 18.73° corresponds to the amorphous fraction. The crystallinity index of Tithonia diversifolia fiber determined

FTIR analysis
The FTIR spectrum of TDFs is shown in Figure 3. The broad absorbance peak at 3400-3600 cm −1 range is attributed to OH stretching vibration of hydroxyl groups present in cellulose (Vijay et al. 2020), and it is absorbed at a peak of 3590 cm −1 .Similar OH groups were absorbed in Erythrina variegata and Sida acuta fibers (Gopinath, Billigraham, and Sathishkumar 2021) at peaks of 3327 and 3341 cm −1 . The prominent peak at 2918 cm −1 is attributed to the C-H stretching vibration of CH and CH 2 groups present in cellulose and hemicelluloses (Kathirselvam et al. 2019). The similar CH and CH2 groups were absorbed in Erythrina variegata, Sida acuta and Sansevieria ehrenbergii fibers (Gopinath, Billigraham, and Sathishkumar 2021;Sathishkumar et al. 2013) at peak of 2918, 2924 and 2924 cm −1 . The sharp peak at 2162 cm −1 corresponds to C≡C stretching and C≡N stretching of nitriles and alkynes (Tamanna et al. 2021). The sharp peak at 1653 cm −1 is attributed to carbonyl (C=O) stretching vibration of Hemicelluloses (Sahoo et al. 2021). The minor peak at 1246 cm −1 corresponds to the C-O stretching vibration of acetyl groups of lignin (Cai et al. 2016). It was absorbed from Erythrina variegata and Sida acuta fibers (Gopinath, Billigraham, and Sathishkumar 2021) at peaks of 1246, and 1246 cm −1 . The peak at 1121 cm −1 corresponds to symmetric stretching vibration of β(1,4)-glycosidic links of cellulose (Alves et al. 2016). It was absorbed from Erythrina variegata, Sida acuta and Sansevieria ehrenbergii fibers (Gopinath, Billigraham, and Sathishkumar 2021;Sathishkumar et al. 2013) at a peak of 1042, 1028 and 1122 cm −1 . The band at 660 cm −1 is attributed to the out-of-plane vibration of the ring structure and the band at 551 cm −1 indicates C-OH bending vibration (Bezazi et al. 2014). It was absorbed from Erythrina variegata, Sida acuta and Sansevieria ehrenbergii fibers (Gopinath, Billigraham, and Sathishkumar 2021;Sathishkumar et al. 2013) at a peak of 677, 604 and 618 cm-1.

Thermal analysis
Thermal analysis was conducted to identify the temperature at which each chemical constituent of TDFs degrades thermally. The thermal degradation behavior of TDFs was evaluated by using TG, DTG and DTA curves as shown in Figure 4(a,b). The peak in the TG curve at 77.31°C attributed to the loss of moisture content present in TDFs through evaporation. The peak at 274.69°C observed in the DTG curve associated with a mass loss of 14.26% corresponds to the thermal depolymerization of hemicellulose. The prominent peak in DTG curve at 337.17°C with a mass loss of 51.60% is attributed to the thermal degradation of cellulose. Similar peaks of cellulose degradation were observed in other natural fibers such as kenaf (307.2°C), hemp (308.2°C), Cissus quadrangularis (328.9°C), Prosopis juliflora (331.1°C) and Acacia leucophloea (346.8°C) (Arthanarieswaran, Kumaravel, and Saravanakumar 2015;Indran and Edwin Raj 2015;Saravanakumar et al. 2013). After the thermal decomposition of cellulose, the thermal degradation of constituents such as lignin, wax and other impurities occurs over a wide range of temperatures. The thermal stability of TDFs can be assessed by evaluating the kinetic activation energy which represents the amount of energy required to decompose one mole of sample thermally. The activation energy (Ea) of TDFs obtained using the slope of the Broido plot as shown in Figure 4(c) and using equation (6) was found to be 94.87 kJ/mol which is comparatively higher than that of other natural fibers such as Saharan aloevera cactus leaf (60.2 kJ/ mol), Cissus quadrangularis (65.23 kJ/mol), Lygeum spartum (68.77 kJ/mol) and Prosopis Juliflora (76.72 kJ/mol) (Manimaran et al. 2018;Sudhir Chakravarthy et al. 2020).
The range of temperature over which the thermal decomposition of each chemical constituent occurs is analyzed quantitatively through differential thermal analysis (DTA). In the DTA curve of TDF given in Figure 4(b) four endothermic peaks at 242.03°C, 364.86°C, 522.48°C and 663.77°C were observed. The first major endotherm ranging between 171.03°C and 364.86°C corresponds to thermal depolymerization of hemicellulose, and the second major endotherm existing within 333.62°C and 476.33°C attributed to the thermal degradation of cellulose. The minor endotherm which ranges between 476.33°C and 616.20°C corresponds to the thermal degradation of lignin and the peak at 663.77°C indicates the temperature at which thermal degradation of wax and other impurities occurs (Manivel et al. 2021).

Tensile properties of TDFs
The single fiber tensile test setup and stress vs. strain plot of twenty samples are shown in Figures 5 and  6 respectively. Further, the mechanical properties of raw TDFs and various natural fibers are provided in Table 3. The tensile strength of TDF was found to be 243-386 N/mm 2 with failure strain 4.07-5.52%. TDF was found to offer comparatively higher tensile strength than other fibers such as Juncus effusus (113 ± 36 MPa), Piassava (134-143 MPa), Cyperus pangorei (196 ± 56 MPa), Arundo donax (248 MPa), Ficus racemosa (270 MPa) and Sansevieria ehrenbergi (278.82 MPa) (Manivel et al. 2021;Sudhir Chakravarthy et al. 2020). The tensile modulus determined using the slope of tensile stress vs. strain plot of each sample was found to be 5.87 to 7.02 GPa. The deformation characteristics of TDFs can be evaluated by determining the microfibrillar angle (α) using an empirical equation (7). The value of α for TDF was found to be 16.23°-18.86°. The microfibrillar angle of natural fibers could range from 2° to 83° with 2° for kenaf and hemp fibers and 83° for ramie fibers (Chokshi et al. 2020). The natural fibers which possess lower microfibrillar angle offer high tensile strength and those which possess higher microfibrillar angle offer high ductility (Djafari Petroudy 2017). Weibull statistical analysis was used to make the relation between the various tensile strength and strains. It predicates the Weibull modulus of TDFs. Based on Weibull distribution analysis, the fiber strength in terms of survival probability is given in equation 8 (Sathishkumar et al. 2013).
where σ, m, and σ 0 is the tensile strength of fiber, Weibull modulus and characteristic strength of the fiber. N is the number of fibers taken for analysis (N = 20 samples). The following relation was used to calculate the Weibull modulus of the Sansevieria ehrenbergii fiber (Sathishkumar et al. 2013).
The Weibull modulus of the TDFs was calculated using equation 9 and it is found to be around 2.16. Probability curve in Figure 7 was used to calculate the Weibull modulus (m) with help of an estimator (Equation 10) (Sathishkumar et al. 2013). It shows that the ratio of the vertical axis to the horizontal axis is confirmed by the Weibull modulus and R values which is equal to 0.99. So, this model value is equal to Sansevieria ehrenbergii fiber (Sathishkumar et al. 2013).

Morphological analysis of TDFs
TDF surface morphology obtained using high-resolution FESEM at different magnifications (400×, 500× and 1.5K ×) in order to investigate the reinforcement and bonding ability is represented in Figure 8(a-e). The cross-sectional surface of TDF as shown in Figure 8(a,b) reveals the multicellular surface structure with a higher amount of cellulose and rough surface with a little amount of wax. Small cracks and voids were also visible on the surface of the fiber. Similarly fiber external surface provided in Figure 8(e) shows a smaller quantity of impurities, pores and pits. It should be noted that pores contribute to surface roughness which significantly influences the tensile properties of fiber due to this a strong interfacial bonding between the fiber and matrix can occur which will further improve the mechanical properties of the fiber reinforced polymer matrix composites. Further Figure 8(e) surface image clearly reveals that the fiber should undergo chemical treatments in order to remove the undesirable debris and amorphous elements from the surface which will enhance the bonding between the fiber and the matrix.

Energy-dispersive X-ray Spectroscopy (EDX)
The EDS spectrum showing the distribution of various elements on the surface of TDFs is given in Figure 8(f). The atomic and weight percentage of elements present at the TDF surface is summarized in Table 4. The existence of high-intensity carbon and oxygen peaks at 0.25 and 0.52 keV in the spectrum corresponds to cellulose which is primarily built up of carbon and oxygen. The presence of relatively higher percentage of carbon than oxygen confirms the existence of amorphous constituents  such as hemicellulose and lignin in higher fractions (Gopi Krishna, Kailasanathan, and NagarajaGanesh 2020).

Atomic force microscopy (AFM) analysis
AFM images of TDF (both 2D and 3D high resolution) with surface roughness, prominent hills and valleys are represented in Figure. 9 with indicated surface roughness values such as Ra (547 nm), Rq (655 nm) and Rz (2577 nm). The parameter roughness skewness (Rsk) computes the symmetrical profile of surface variations about the centerline (Manimaran et al. 2018). It should be known that the negative value of Rsk designates that the nature of the fibers is porous. The value obtained for TDF was −0.296, which means the fibers are slightly porous in nature. The Roughness kurtosis (Rku) parameter provides information about the roughness of the fibers. The obtained value of Rku is 2.13 which indicates that the surface of TDF is spiky.

Conclusion
The suitability of novel cellulosic fiber extracted from the bark of Tithonia diversifolia for a composite application was assessed through various characterization tests. The chemical tests on TDFs revealed the presence of higher cellulose (63.9 wt.%), higher lignin (19.3 wt.%) and lower wax content (0.61 wt. %) which results in high strength and rigidity of fiber and facilitates better bonding characteristics when reinforced with polymer matrices while making composites. Through X-ray diffraction, the crystallinity index and crystallite size of TDFs were estimated to be 51.84% and 2.93 nm respectively. Higher crystallinity index and higher crystallite size lead to superior mechanical properties, lower water absorption and greater resistance against chemical attack. TDFs were found to exhibit higher kinetic activation energy (94.87 kJ/mol) and withstand higher temperature (up to 337.17°C). Higher tensile strength (243-386 MPa), lower density (1440 kg/m 3 ) and better thermal stability validate TDFs as a promising reinforcement with matrix in the making of polymer composites for various nonstructural applications such as partition boards in buildings, interior body panels in automobiles and for making sports goods. The scope for the future is that the identified novel fiber through characterization has shown good mechanical and structural property that is suitable for making sports and nonstructural automotive components. This characterization focused only on microstructure and mechanical properties; however, an intuition for the same will be carried out in the future to evaluate the thermal and tribological characteristics. However, more research must be done on processing techniques in a more sustainable way, which is of current scientific interest with appropriate chemical treatments which will eliminate the amorphous compounds found on the surface of TDFs toward making effective polymer composite reinforcements.