A novel application of cellulose aerogel composites from pineapple leaf fibers and cotton waste: Removal of dyes and oil in wastewater

In a world where demands for freshwater are ever-growing, wastewater remediation becomes a global concern. Especially, water, which is contaminated by oil, dyes, poses challenges to the management of water resources. The development of innovative processes for wastewater treatment is still a major obstacle. With regard to its fast removal rate and environmental compatibility, cellulose aerogel composites are recently considered as a potential contributor for water remediation. In this study, cellulose aerogel composites are fabricated using the sol–gel method from two-agroindustrial wastes: pineapple leaf fibers and cotton waste fibers in alkali-urea solution followed by freeze-drying. The prepared cellulose aerogel composites are extremely lightweight with a low density (0.053 − 0.069 g cm−3) and high porosity of nearly 95%. It is worth noting that the mechanical strength of the cellulose aerogel composites is remarkably improved with their Young’s modulus increasing by 5–9 times compared to that of the previous aerogel composites using polyvinyl alcohol as a binder. The as-synthesized aerogel composites are directly applied to adsorb cationic methylene blue and exhibit a maximum adsorption uptake of 34.01 g g−1. The methyltrimethoxysilane-coated cellulose aerogel composites also show their ability to deal with oil pollution with a maximum oil adsorption capacity of 15.8 g g−1 within only 20 s. Besides the oil removal, our developed cellulose aerogel composites have demonstrated their capability in treating dye-contaminated wastewater for the first time based on their evidenced ability to eliminate methylene blue.

The average true density of components C PF The mass concentration PF C CF The mass concentration CF

C o Initial concentrations of adsorbate C t
Concentrations of adsorbate at various time t V Solution's volume m Mass of component q t Adsorption capacity at various time t k 1 The kinetic rate constant of the pseudo − first order model of adsorption k 2 The kinetic rate constant of the pseudo − second order model of adsorption K L The Langmuir constant q e The equilibrium adsorption capacity q m The maximum adsorption capacity R L Equilibrium parameter K F The Freundlich constant n The empirical constant C e The equilibrium adsorbate concentration CMC Carboxymethyl cellulose GO Graphite oxide RCE Regenerated cellulose

Introduction
Environmental pollution, particularly water pollution, is currently a major global issue imposing a serious threat to both human and ecosystem health. According to statistics, water pollution is responsible for roughly 14.000 fatalities each day [1]. Amongst various pollution causes, organic pollutants comprising soluble dyes and insoluble oils are the main contaminants in wastewater [1,2]. The amount of these pollutants have been significantly increasing due to an increase in industrial activities and oil spill incidents [3].
Oily wastewater devastates both aquatic and terrestrial lifeforms in the marine areas and has even detrimental economic impacts on tourism and fishing. Interaction with spilled oils has not only killed innumerable aquatic organisms (such as sea birds, mammals, algae, and so on) but the sand on seashore beaches has also been badly polluted [4,5]. Deepwater Horizon, the biggest oil spill in U.S. history, released an estimated 4.9 million barrels of oil into the ocean resulting in severe damage to marine life, coastal wetlands, the fishing industry, tourism, and so on [4][5][6]. Additionally, the toxic effects of dye-contaminated water relate to the risk of dyes degrading into carcinogenic amine substances whilst the dark hue of certain dyes interferes with the photosynthetic activity of marine plants [7]. Therefore, an effective technique for the removal of dyes and oils from wastewater is urgently required.
Despite many attempts in treatment of the contaminated wastewater, adsorption has plenty of advantages such as high efficiency, cost-effectiveness, and ease of use for dye and oil remediation [2]. However, conventional adsorbents such as inorganic absorbents, natural fibers, and synthetic organic polymers have several detrimental aspects involving secondary pollution, poor selectivity, and limited sorption capacity [8]. Consequently, studies to eliminate the limitations of previous adsorbents and develop a novel material with superior properties should be urgently performed. Furthermore, the process of synthesizing adsorbents should be improved to meet the criteria of green chemistry and avoid the release of waste into the environment.
Cellulose-based aerogels owning outstanding features such as low density, high porosity, large specific surface area, and excellent flexibility have attracted worldwide attention over the past decade [5]. Various studies have revealed that it could be an excellent material to remove oil and dye effectively in contaminated water [6,9,10]. The sources to produce cellulose aerogels are diverse, but cellulose-originated wastes, especially biomass and textile waste, are of primary interest since they are appropriate for developing eco-friendly production technologies with the use of renewable feedstocks [11]. Aerogels prepared from cellulose-rich waste are biodegradable, easy to be functionalized for different uses, and inexpensive due to the abundance of raw materials [5]. Do et al. developed cellulose aerogels from pineapple agricultural waste that had excellent properties for applications such as sound insulation, thermal insulation, as well as oil and organic solvent adsorption [12,13]. Later generations of cellulose aerogel have been developed with the addition of reinforcements, for example, polymers, metals, and inorganic compounds to obtain novel aerogel composites that not only keep aerogel-like properties but also have improved compressive strength, elasticity, toughness, and high module − volume ratio [14,15]. Cheng et al. fabricated cotton aerogels and cotton − cellulose aerogels using recycled fibers for oil-spill cleaning up [16]. The adsorption capacity for various oils and organic solvents could range from 19.8 to 41.5 g per gram of cotton fiber-derived aerogel fabricated by Wang and Liu [17]. Lim et al. synthesized pineapple aerogels functionalized with activated carbon and diethylenetriamine for respective ethylene adsorption in fruit preservation and removal of nickel ions in water treatment. The modified aerogels exhibit significant adsorption uptake of 1.08 and 0.835 mmol g -1 for ethylene and nickel ion adsorption, respectively [18]. Do et al. invented heat-insulating aerogel composites by the consolidation of pineapple leaf fibers (PFs) and discarded cotton fibers (CFs) and the use of polyvinyl alcohol as a cross − linker. The obtained aerogel composite showed an increase in their compressive modulus from 11.33 to 44.63 kPa with increasing the total fiber content [15]. A simple alkali-urea method without consumption of any crosslinkers was developed by Fauziyah et al. in the synthesis of cellulose aerogels from coir fibers. Briefly, a mixture of sodium hydroxide (NaOH) and urea at a fixed concentration is applied to cause cellulose fibers to swell and disperse uniformly without agglomeration, followed by sol-gel with absolute ethanol at room temperature and freeze-drying [19,20].
To the best of our knowledge, there has been no study on the fabrication of pineapple aerogels via the advanced alkaliurea method as well as their application of dye removal in wastewater. In this work, novel cellulose aerogel composites from a combination of PFs and CFs are synthesized via NaOH/urea method with a modification of using 60 °C absolute ethanol to promote gelation faster and shorten the synthesis time. The obtained PF/CF aerogel composites are characterized in terms of their density, porosity, surface morphology, mechanical properties, and thermal stability. The PF/CF aerogel composites are also evaluated for their capability to remove aqueous methylene blue which is a chemical extensively used in the disciplines of biology and medicine, as well as in the textile industry, coloring fabric, silk, and cotton [21]. Because the molecular size of MB is approximately 1.38 nm [22] smaller than the pore size of the analyzed sample points so it is appropriate for this investigation. Besides, the adsorption of engine oil for oil absorption application was also investigated. The effect of total fiber content on some characteristics such as physical-mechanical properties, maximum adsorption capacity, adsorption kinetics, adsorption isotherms, and thermodynamics of the aerogel composite is comprehensively studied.

Aerogel composite synthesis
NaOH/urea/water solution was firstly prepared in a mass ratio of 7:12:81 at room temperature. PFs were milled to a fine powder (ca. 100 µm) before being mixed with CFs in a different solid mass proportion of 1:1, 2:1, and 4:1, respectively. Each mixture of PFs and CFs was then dispersed into the NaOH/urea/water solution, refrigerated to 0 °C for 15 min, and magnetically stirred to achieve a homogeneous dispersion containing the total fiber content of 4 wt%. The system was frozen for 24 h and then defrosted to create strong hydrogen bonding between the fibers. The suspension was poured into a cylinder mold and incubated with 60 °C ethanol to promote gelation before being washed with DW to remove excessive NaOH and urea. Finally, the whole mixture was frozen at − 50 °C for 4 h and freeze-dried for roughly 50 h to yield PF/CF aerogel composites. The hydrophobic aerogel composites were obtained via a simple chemical vapour deposition with MTMS as described in our previous study [12].

Characterization of cellulose aerogel composite
The bulk density of each aerogel composite is determined via its weight and volume. Its porosity ( ) is calculated based on its bulk density ( a ) and the average density of components ( b ) as indicated by Eq. (1) [15,23]: where PF = 1.07 g cm −3 and CF = 1.56 g cm −3 are the density of PF and CF, respectively. C PF and C CF are the mass concentration of PF and CF, respectively.
Field − Emission Scanning Electron Microscope (FE − SEM Hitachi S4800) is applied to investigate the porous structure of samples at an acceleration voltage of 10 kV [24]. All samples are coated with a thin layer of Pt for 30 s before measurement [25].
Fourier Transform Infrared spectrometer (FTIR) Perki-nElmer FTIR device (USA) is used to analyze the chemical structure of the PF/CF aerogel composites with and without coating with MTMS. The samples are characterized in the wide range of wavenumber from 4000 to 550 cm −1 at a resolution of 4 cm −1 using the attenuated total reflection testing.
Thermogravimetric analysis (TGA) curves are recorded by a TGA analyzer (LINSEIS DSC PT 1600). All surveyed samples are heated from 40 to 800 °C at a heating rate of 10 °C min −1 .
Ultraviolet − visible (UV − VIS) spectrometer Shimadzu's UV − 1800 is utilized to analyze the absorption of light intensity of incoming beam through sample solution placed in a cuvette. The wavelength of the incoming beam is 640 nm corresponding to the color wavelength of MB.
The compressive property of PF/CF aerogel composites samples is evaluated by using an Instron 5500 Universal Testing Machine with a compression rate of 10 mm min −1 .
The obtained data are plotted as a curve of stress and strain. The Young's modulus of each aerogel composite is determined from the fit linear at less than 10% strain.

Study on oil removal of hydrophobic PF/CF aerogel composites
In this study, 5w30 motor oil is used for evaluating oil removal of the aerogel composites. The initial mass m o of the sample is firstly weighed by a 4 − digit analytical balance. The sample is then placed in the beaker containing 5w30 motor oil for more than 1 min. The material sample is submerged in a beaker of 5w30 motor oil for 1 h to guarantee that the most oil could penetrate the pore structure of the sample to evaluate the maximum adsorption. Finally, the sample is placed on a sieve, allowed to drain for 2 min, and then the adsorbed mass is reweighed. The experiment is repeated three times to determine the average value, which is the maximum oil adsorption capacity of the aerogel composite.
Oil adsorption capacity q t is measured at 5, 10, 15, 20, 30, 40, 50 (seconds) to identify the most suitable adsorption kinetic model for the aerogel composite.
The adsorption capacity of each sample at the time t is calculated using the following Eq. (3): where m t (g) is the composite weight at the time t, and m o (g) is the composite weight at the initial time.
In the adsorption mechanism, kinetic prediction of the adsorption capacity and adsorption time play an important role. Kinetic data are analyzed using pseudo − first order and pseudo − second order models to characterize the dependency of the oil adsorption on time.
The pseudo − first order model can be expressed as Eq. (4) [26]: where q e and q t (g g −1 ) are the amounts of 5w30 oil (g g −1 ) adsorbed on adsorbent at. equilibrium and various time t (sec), respectively. k 1 (sec −1 ) is the kinetic rate constant of the pseudo-first order model of adsorption.
The pseudo − second order model is another kinetic model that includes several types of adsorptions such as external film diffusion, adsorption, and internal particle diffusion ts equation can be given as Eq. (5) [26]: where q e and q t (g g −1 ) are the amounts of 5w30 oil (g g −1 ) adsorbed on adsorbent at equilibrium and various time t (sec). k 2 (g g −1 .sec −1 ) represents the rate constant of the pseudo-second order model of adsorption.

Study on cationic dye removal of PF/CF aerogel composites
To process the experiment, 0.1 g of solid MB is dissolved in distilled water in a 1000 mL cylinder to create a 1000 ppm MB standard sample solution. The standard sample is diluted into MB solutions with concentrations of 25, 50, 75, 100, 125 (ppm) for kinetic and isothermal adsorption studies, respectively. PF/CF aerogel composites with different proportions of fibers (PC41, PC21, PC11) are weighed to 0.05 g, placed in 25 mL of MB color solution at room temperature, and shaken well under time conditions. After adsorption, the sample is removed, and UV − VIS analyzes the MB color solution adsorption at 640 nm; the result obtained is the solution's absorbance. Each composition is performed three times to determine the average value and standard deviation.
The adsorption capacity at the t time ( q t , mg g −1 ) is calculated by the Eq. (6): where C o , C t (ppm) are the concentrations of MB in the initial solution and at the t time, respectively; V (L) is the volume of solution, and m(g) is the initial aerogel composite mass.
To conduct a kinetic adsorption experiment, the investigation time for the experiment is 15, 20, 25, 30, 45, 60, and 75 (min) with MB color solution with the initial concentration of 50 ppm. The adsorption parameters of the research experiment according to the first and second − order apparent kinetic models of adsorption is the same as those of the oil adsorption experiment. Experiment for isothermal studies with MB color solution starting concentrations of 25, 50, 75, 100, and 125 (ppm).
Finally, the removal efficiency of MB (H%) refer to the total amount of adsorbed with respect to the total initial concentration of adsorbent is calculated according to Eq. (7): where C o and C e refer to the initial and equilibrium concentration of MB (ppm), respectively.
The kinetics study of MB adsorption is done by using the pseudo-first order model and pseudo-second order model followed by Eq. (4) and Eq. (5), as mentioned above.
The Langmuir isotherm model is suitable for the monolayer adsorption occurring on a homogeneous surface without subsequent interaction among the adsorbed species. The model can be expressed as Eq. (8) [26]: where K L (L.mg −1 ) is the Langmuir constant. q e , q m (mg g −1 ) is the equilibrium adsorption and the maximum adsorption capacity of adsorbate, respectively. C e (ppm) is the equilibrium concentration of the adsorbate.
The constant K L characterizes the interaction force between the adsorbent and the adsorbate at a certain temperature, or in other words, K L describes the adsorption center's selectivity. Construct a graph C e ∕q e based on C e to determine the adsorption constants K L and R 2 as well as the maximum adsorption q m . From the K L value, it is possible to assess whether the adsorption process is suitable or not in the investigated concentration range of the adsorbent through the equilibrium parameter R L determined as Eq. (9) [27]: where R L is the equilibrium parameter, K L is the Langmuir constant, and C o (ppm) is the initial concentration of the adsorbate. The Freundlich isotherm model can be described as Eq. (10) [28]: where K F is the Freundlich adsorption constant, and n is the empirical constant, indicating the quantity that evaluates the degree of heterogeneity of the adsorbent surface and describing the appropriateness of adsorbent molecule distribution on the surface of the adsorbent. A value of n higher than 1 implies that molecules are adsorbing well onto (10) ln q e = ln K F + ln C e n the adsorbent surface. The higher the value n , the higher the adsorption strength. Construct a graph of lnq e dependent on lnC e , thereby determining R 2 , the Freundlich constant K F and the empirical constant n.

Characterization of PF/CF aerogel composite
Lyophilization is the preferred technique in the synthesis of aerogel composites from PF and CF in order to avoid structural damage. As the result, the porous structure of developed aerogel composites with high porosity is clearly observed through SEM images in Fig. 1. The pores inside the structure of aerogel composites created by PFs (straight, sizes 40 − 60 µm) and CFs (twisted, sizes 10 − 20 µm) were primarily macropores, ranging from 30 to 90 µm. Furthermore, the increase of CF concentration causes a more packed structure, resulting in a remarkable decrease in the diameter of the pores [16]. The results are consistent with previous studies on PF/CF aerogel composites [15]. At the same time, the arrangement and linkage between the fibers are disordered as they are cast on the mold randomly before adding the NaOH/urea/water solution, which make the cellulose chains swelling and dissolving into solution, then gelation caused by physical cross-linking of cellulose chains to form the structure of the material [29].  The physical properties of the aerogel composite samples in terms of density and porosity were measured and are shown in Table 1. The results reveal that the obtained aerogel composites possess a low density in the range of 0.053 − 0.069 g cm −3 and high porosity of 94.7 − 95.2%, indicating their highly porous structure after synthesis. Increasing the CF content whilst remaining the total fiber content unchanged (4%) results in an increase in the density of the aerogel composite from 0.053 to 0.069 g cm −3 . This can be explained by the fact that the density of CF (1.56 g cm −3 ) is greater than that of PF (1.07 g cm −3 ) [15]. Compared to the previous work, our PF/CF aerogel composite have a higher density and lower porosity than that of PF/ CF aerogel composite synthesized by Do et al. (density of 0.019-0.046 g cm −3 and porosity 96.1-98.4%) [15].
The correlation between compressive stress and corresponding strain of the PF/CF aerogel composite with different mixing ratios of the two fiber compositions is shown in Fig. 2.a, and their Young's modulus parameter is tabulated in Table 1. From the results, it can be seen that the synthesized aerogel composites exhibit a high compressive modulus of 90.97 kPa (PC41), 145.28 kPa (PC21), and 203.72 kPa (PC11). These parameters are approximately 5 − 9 times as much as those of PF/CF aerogel composite produced via the crosslinking method with PVA (11.33-44.63 kPa) [15] and are much higher than those of PF only aerogel (1.64-5.34 kPa) [13]. The higher compressive strength of the material can be explained by the fact that when the fibers are dispersed with NaOH/Urea/H 2 O aqueous solution and then gelation with ethanol, the cellulose fibers are bonded together by intramolecular and intermolecular hydrogen bonds which make the structure become tight aggregates than when using PVA as a cross-linking [29]. The stress − strain curves of aerogel composites show a linear line at strains less than 10%; this linear elastic region shows the ability to fully recover to the original state before deformation. A smooth plastic yielding region witnessed at the compression strain of 10 − 60% demonstrates the collapse of the porous structure. Finally, a densification area is observed for compression strains greater than 60%, where stress is dramatically increased. As in Fig. 2a, when CF content increases, the mechanical strength under the same pressure point is greater. This proves that CF plays a role in the reinforcement of the hollow structure of the PF/CF aerogel composite, and increasing CF content makes the fabricated composites stiffer.
The graph of TGA analysis of PF/CF aerogel composite is shown in Fig. 2b. From the graph, it can be seen that changing the mixing PF/CF ratios has no significant effect on the mass loss of aerogel composites by thermal decomposition. The PF/CF aerogel composites exhibit a typical thermogram with mass degradation in three phases by the temperature as follows: (i) 60-100 °C, (ii) 270-350 °C, and (iii) 360-460 °C. At the first step, the sample weight decreased from 8 to 10%, attributed to the evaporation of moisture. The reason for this phenomenon is because the PF/ CF aerogel composites after drying still have a small amount of moisture inside and contain many hydroxyl groups on cellulose chains of two types of constituent fibers that absorb water in the air. At the next stage from 270 to 350 °C, a drastic decline of about 67-70% is witnessed in the mass of all composite samples caused by oxidative decomposition of organic compounds including cellulose, hemicellulose, lignin [13]. Eventually, the mass degradation of aerogel composites occurs rapidly from 360 °C until the material is completely decomposed at 460 °C, and the rest is ash. As a result, the highest stable temperature range for PF/CF aerogel composites is around 230 − 280 °C, compared with PF/CF aerogel composite synthesized by Do et al., there is

Oil adsorption of the PF/CF aerogel composites
The PF/CF aerogel composites after MTMS coating are hydrophobic because the -OH groups of cellulose are displaced by -O-Si-(CH 3 ) 3 groups from MTMS, indicating by not absorbing water droplets on the surface as shown in Fig. 3a. A piece of this hydrophobic material was placed in a petri dish already containing 5w30 oil (Fig. 3b) to evaluate the oil adsorption of the produced aerogel composite. The results showed that the aerogel composite sample had almost completely absorbed all of the oil on the petri dish after only 2 min (Fig. 3c).
The maximum oil adsorption of the material is 15.8 g g −1 with sample PC41, and the lowest adsorption is 11.3 g g −1 with sample PC11 (Fig. 4a). This shows that the higher the porosity, the better the adsorption capacity. The aerogel composites have higher oil adsorption when compared with coir-based aerogels prepared by the same alkali-urea method (10 g g −1 ) [30].
Oil adsorption capacity gradually increases with the adsorption time of PF/CF aerogel composite (Fig. 4b). The material exhibits oil adsorption fastest in the first 10 s then reaches equilibrium after about 20 s. PF/CF aerogel composite can quickly adsorb oil because of its porous structure and 3D space structure so that oil can penetrate into the structure of material through hydrophobic interaction and capillary force [18,31].
The R 2 values for both models are calculated and shown in Table 2. As a result, the 5w30 motor oil adsorption process of PF/CF aerogel is consistent with the pseudo − second order model with R 2 values closer to 1 than the pseudo − first

Adsorption kinetic
It can be seen from Fig. 5 that all three materials with different fiber ratios have a fast adsorption time in the first 20 min due to the diffusion process to the large surface without obstruction, as well as the number of empty adsorption centers. At the same time, it can be observed that the lower the porosity of the material, the lower the adsorption capacity due to the decrease of empty adsorption centers in the material. Adsorption equilibrium was reached about 30 min from when the material was placed in the MB solution. The explanation for the short − time adsorption is because (1) the material has a porous structure and high surface area, (2) there is a hydrogen interaction between the hydroxide functional group (− OH) of cellulose and nitrogen atoms on MB [32], (3) capillary force bring MB from the surface to the interior of the material. It can be concluded that the adsorption undergoes rapid adsorption stages on the surface of the material, so the adsorption increases the first time very quickly. The diffusion process into the porous structure inside the material and this process continue until equilibrium is reached.
Both common model pseudo-first-order and pseudosecond-order are also studied. The R 2 values for both models are calculated and shown in Table 3. All PF/CF aerogel composites fit better to the pseudo-second-order model than the pseudo-first-order model. This demonstrates that both adsorbent and adsorbate molecules have an influence on adsorption [7]. Figure 6 shows that the adsorption capacity increases with increasing initial MB concentration. With a fixed amount of material, there is a constant number of adsorption centers. Therefore, when the concentration of MB is increased to a certain level, a saturated adsorption state will be reached which followed by the drop of removal efficiency on MB absorption. Two models isotherms of PF/CF aerogel composite are investigated (Fig. 7), the adsorption data according to Langmuir and Freundlich adsorption isotherm models ( Table 4) gives results that fitted Langmuir's model with higher R 2 , which indicates that the adsorption process is  monolayer on the surface of the material [28]. The R L values are all in the range between 0 and 1, representing the suitable initial concentrations chosen to investigate the Langmuir isotherm adsorption equation [33]. Through Langmuir's adsorption equation, the maximum adsorption capacity q m of PF/CF aerogel composite materials was determined as PC41 (34.01 mg g −1 ), PC21 (33.22 mg g −1 ), and PC11 (23.20 mg g −1 ). Table 5 shows the comparison of MB adsorption of materials from various studies. Even though the material porosity is large, the specific surface area is low due to the gelation method, hence the MB adsorption of PF/CF aerogel composite is still rather poor, lower than that of most materials

Conclusion
The PF/CF aerogel composite with outstanding mechanical strength has been successfully synthesized from pineapple leaf fibers and cotton waste fibers by using the sol − gel method with sodium hydroxide NaOH/urea solution as a solvent and followed by freeze − drying. The fabricated aerogel composite exhibits low density and high porosity, and robust compressive modulus of 0.053 − 0.069 g cm −3 , 94.7-95.2%, and 90.97-203.72 kPa, respectively. The motor oil adsorption of PF/CF aerogel composites occurs effectively, with adsorption capacity nearly 16 times their weight. Whereas, the maximum MB adsorption capacity of 34.01 mg g −1 is determined for PF/CF aerogel composite having the highest cotton fiber content. This proves that PF/CF aerogel composite fabricated in this study could be a promising material to be used to deal with the removal of organic pollutants in contaminated water treatment.