Utilization of water hyacinth (Eichhornia crassipes) biomass as eco-friendly sorbent for petroleum oil spill cleanup

The convergence of global rapid industrialization and consumer requirements for energy resources has created an enormous demand for fossil fuels in both consumer and industrial sectors. Nonetheless, this creates environmental concerns, given the potential for oil spills in the ocean or land while transporting petroleum products (Ali et al. 2012; Liu et al. 2017). These incidents can also arise at various stages of the petroleum industry, including drilling, extraction, refining, and unintentional discharges (Gurav et al. 2017, 2021). Oil spills in the ocean have the potential to result in significant water contamination and damage to marine ecosystems (Ramirez Leyva et al. 2018; Li et al. 2020; Guo et al. 2022). The enduring consequences of such incidents may persist for many years as sediments hold residual oils (Wang et al. 2021). Hence, the pollution arising from these spills can inflict harmful consequences on the natural ecology of the aquatic environment and its living resources (Ramdhani 2023). The primary constituents of petroleum oil, namely n-alkanes, cycloalkanes, and aromatic hydrocarbons, possess mutagenic and carcinogenic potential to humans and are highly toxic to marine life and the overall ecosystem (Zhang et al. 2016; Alzahrani & Rajendran 2019). Exposure to these compounds can cause health risks affecting several systems within the human body, including the respiratory, nervous, circulatory, reproductive, immune, and endocrine systems (Udgire et al. 2015). Moreover, petroleum hydrocarbons are harmful to marine organisms, causing impacts such as habitat damage, impaired body functions, and the accumulation of these substances in food chains (Lin & Tjeerdema 2008).

In 2020, crude petroleum was ranked as the third most traded commodity, valued at $640 billion. Saudi Arabia was the main exporter, followed by Russia, the United States, Canada, and Iraq. On the import side, China was the largest buyer, with the United States, India, South Korea, and Japan being major importers. Over 50 years, from 1970 to 2020, close to 5.86 million tons of oil was spilled during shipping accidents. In 2021, the industry saw one major spill in Asia (more than 700 tons) and five medium spills across Africa, Asia, and North America (7–700 tons) (Upadhyay 2022). It has been assessed that 630 billion gallons of oil-contaminated wastewater are produced annually in the oil industries of the United States (Cherukupally et al. 2020). The oil spill of the Deepwater Horizon in 2010 stands as the largest oil spill in the history of the United States, discharging around five million barrels of oil into the Gulf of Mexico over 87 days (Follett et al. 2014). The spill had immediate environmental consequences, including the death of thousands of animals, such as turtles, birds, whales, and dolphins (McGee 2010). Additionally, it caused high concentrations of polycyclic aromatic hydrocarbons along the coast, negatively affecting wildlife (Ramseur & Hagerty 2014). Again, a study indicated that phototoxicity may still be occurring due to the presence of remaining oil (Evers et al. 2019). Another study considered the impact of an oil spill in Rhode Island, which resulted in the death of numerous birds and the need for restoration efforts (Samuelson et al. 2021). The Dalian New Port oil spill in 2010, one of China's most severe, released 35,000 tons of crude oil, resulting in long-lasting damage to the coastal and marine ecosystems. Oil slicks covered over 100 km² of the sea and 20 km of shoreline, affecting water quality, marine life, and coastal ecosystems with persistent contamination evident even years later (Guo et al. 2022). Also, the oil spills near Sori Island and Yeo-Cheon in South Korea in 1995 released thousands of tons of crude oil, affecting over 230 km of coastline and necessitating extended cleanup operations. The Hebei Spirit spill in 2007, the country's largest, resulted in the release of about 10,800 tons of oil near the port of Daesan, impacting the same region analyzed for pollution risk in the study (Lee 2017). Several studies found that oil spills have a significant impact on global climate change. For example, research on the Deepwater Horizon oil spill revealed its impact on British Petroleum's climate change discourse, as seen in their annual reports before and after the spill (Kapranov 2017). The world is now struggling with a freshwater shortage to meet the global water demand due to climate change (Khondoker et al. 2023). Also, climate change poses challenges to the gas and oil infrastructures in coastal and offshore areas (Dong et al. 2022). Comprehending the chemical composition of petroleum hydrocarbons (PHs), their movement and deposit in the sea, and their harmful effects on marine life and ecosystems is essential to implementing efficient emergency responses and regulatory actions (Boehm 1964; Gad & Gad 2014).

The noticeable methods used for cleaning the oil spills from water surfaces are based on physical/mechanical methods (skimmers, barriers, booms, hydrophobic meshes), chemical (dispersants or surfactants), and biological methods (microbial) (Liu et al. 2016; Doshi et al. 2018; Debs et al. 2019; Abdullah et al. 2022; Hethnawi et al. 2023). Although these methods are partly successful, searching for plant biomass-based green adsorbents is of the most significant interest from economic and ecological standpoints. Over recent years, there has been widespread utilization of advanced materials such as meshes, foam membranes, aerogels, and fabrics with modified surfaces for the effective separation of oil from water. Raffia fiber, a natural fiber abundant in eastern Africa, has shown ability as an absorbent for hydrocarbons, with higher sorption capacities observed for larger particle sizes (Silva et al. 2022). Additionally, for the removal of petroleum oil hydrocarbons, different particle sizes of activated charcoals have also been studied, resulting in the smallest particle offering the highest adsorption capacity (Ji et al. 2020). Also, inexpensive sorption materials based on raw materials derived from plants, such as vegetable waste and fruit peels, have been explored for the removal of heavy metals and petroleum hydrocarbons from water (Smyatskaya et al. 2018).

Water Hyacinth (WH) (Eichhornia crassipes), an invasive plant native to the Amazon basin, is considered one of the 10 most problematic plants because of its rapid growth and detrimental effects on aquatic ecosystems (Babatunde et al. 2023; Namasivayam et al. 2023). Furthermore, it has been found that invasive plants, including WH, enhance the methane emissions on decomposition, contributing to global warming (Zhou et al. 2022). In recent years, WH biomass (WHB) and WH biochar have been attracting researchers to utilize them as eco-friendly sorbents to remove heavy metals and other pollutants from wastewater (Priya & Selvan 2017; Gaurav et al. 2020; Amalina et al. 2022; Carneiro et al. 2022). Also, very few studies were carried out to remove petroleum oil using WHB. For instance, time-dependent capillary rise experiments were conducted to remove diesel, lubricant, and castor oil using different parts of the WH plant (leaf, stem, and root), and the stem proved to be a highly efficient sorbent among them (Rani et al. 2014). Another study was found utilizing free-floating WH plants as a sorbent for oil removal (Yang et al. 2014). The cellulose-based aerogel produced from WHB proved to be an effective sorbent for oil removal (Yin et al. 2017).

Therefore, the present research focuses on the utilization of WHB as an eco-friendly and cost-effective sorbent for the effective removal of petroleum oil spills from the water surface. To further enhance the petroleum oil sorption performance and hydrophobicity, WHB was modified using extra virgin coconut oil. Subsequently, the modified water hyacinth biomass (MWHB) was characterized and applied to remove simulated petroleum oil spills. Various isotherm and kinetic models, and analytical methods were comprehensively evaluated to deduce potential insights into the sorption process. This biomass can be used independently or with mechanical methods such as skimmers, barriers, and booms to clean the oil spill sites.

The WH plants were collected from the San Marcos River (29°53′34.94″ N, 97°55′53.49″ W), San Marcos, TX, United States. After collection, the plant material was washed several times with running tap water, and the hairy roots were separated. The remaining plant biomass was dried for 24 h in active sunlight (4 days for 6 h each). Finally, the plant biomass was dried in an oven at 70 °C for 24 h. Spent engine oil (SEO) was collected from a local car oil changing service station. The density and viscosity of SEO were 0.88 g/cm³ at 25 °C and 1.27 Poise at 40 °C. The extra virgin coconut oil was purchased from a local general store in San Marcos, TX.

The WHB (including leaves, petioles, and stems) was chopped into small pieces with an X-ACTO wood guillotine trimmer and a kitchen knife. The chopped biomass was sieve analyzed using (1/4)″ (6.3 mm), #4 (4.75 mm), #10 (2 mm), #20 (850 μm), #40 (425 μm), #60 (250 μm), #100 (150 μm), and #200 (75 μm) size sieve screens and a pan. Particle sizes 4.75–2 mm and 2 mm–850 μm were selected to be used for the experiment in a 2:1 ratio, respectively. Preliminary sorption tests for SEO were done using varied sizes of WHB before finalizing the specific particle sizes mentioned in Section 3.1. The biomass particle sizes of 4.75–2 mm and 2 mm–0.85 μm were denoted as ‘type-1’ and ‘type-2’, respectively. Both sorbents have proven to be ‘type-II sorbent (loose)’ according to the ‘American Society for Testing and Materials, ASTM F726–17’ guideline for the dynamic degradation test and the oil sorption long and short test.

To enhance the hydrophobicity and oil sorption capacity, six types of modified biomass were designated as Mod1, Mod1′, Mod2, Mod2′, Mod3, and Mod3′. The Mod1 and Mod1′ were prepared using a 1:1 ratio of dry biomass and coconut oil, and 10 and 20% methanol in deionized (DI) water, respectively. Whereas Mod2 and Mod2′ were prepared with a 1:1.5 ratio of dry biomass and coconut oil, and 10 and 20% methanol in DI water, respectively. The Mod3 and Mod3′ were developed using a 1:2 ratio of dry biomass and coconut oil, with 10 and 20% methanol in DI water, respectively.

The extra virgin coconut oil contains about 50% lauric acid which is natural and biodegradable and helps to increase the hydrophobicity of the sorbent (Dankovich & Hsieh 2007; Zinjarde et al. 2008; Gurav et al. 2021). Therefore, six 150 ml glass beakers were added with the sieved and dried WHB along with the required concentration of coconut oil. These beakers were placed in an oven at a temperature of 70 °C for 5 h. Hand stirring was performed using a glass rod at 20-min intervals throughout this duration. Once the coconut oil was evenly dispersed, the required quantity of methanol solution was added to the beakers. The beakers were then kept in the oven at 70 °C for 1 h, with intermediate mixing. Methanol reduces surface tension and allows the coconut oil to disperse more consistently over the surface, promoting uniform coverage and minimizing clumping or uneven distribution (Navarathna et al. 2020). Then, the mixtures were filtered through a stainless-steel mesh filter and washed repeatedly with n-hexane after 20 min of soaking time. Finally, the biomass was dried in the oven at 70 °C for 24 h and preserved in air-tight containers. The SEO sorption experiments were performed with diverse types of MWHBs. To check the effect of organic solvents on the hydrophobicity of unmodified biomass (without coconut oil), the control biomass (C′) was prepared with 20% methanol solution followed by two n-hexane washes. The experimental conditions for all the types of MWHB and control biomass (C′) have been summarized in Table 1. To check the hydrophobicity, the water sorption test of raw biomass, Mod1′, and C′ was performed. Each sample weighing 0.1 g was placed in 100 ml of DI water for 1 h, and the weight difference of the biomass was considered. The Mod1′ was found to be more hydrophobic than other sorbents, hence, it was selected for further sorption experiments.

The surface morphology, microstructure, and texture of WHB and MWHB were investigated using scanning electron microscopy (SEM; JEOL JSM-6010PLUS/LA). Whereas the elemental composition of WHB and MWHB was determined using SEM–energy dispersive spectroscopy (EDS). To examine changes in the functional groups on MWHB before and after SEO sorption, Fourier transform infrared (FTIR) spectrometry was employed in the scanning range of 400–4,000 cm⁻¹ (Bruker ALPHA II FTIR Spectrometer). The contact angle of water was measured on both WHB and MWHB using the sessile drop technique. Pictures were captured by iPhone 13 Pro, and subsequently, these photographs were examined using the ImageJ software.

Here, Q_m denotes the highest achievable sorption capacity (g/g), while K_l, K_f, and K_t represent the constants linked with the Langmuir, Freundlich, and Temkin isotherm models, respectively. C_e stands for the equilibrium concentration of SEO (g/l), 1/n depicts the intensity of sorption, and A signifies an additional Temkin constant. The appropriateness of these models was ascertained by comparing the coefficient of determination (R²) against the actual experimental data, which served as a benchmark for validation. The constants and parameters within these isotherm models, such as the maximum sorption capacity and the constants associated with each model, were determined by fitting the batch sorption data to the respective equations using the Origin software.

To check the reusability of sorbents, sorption–desorption tests were conducted to assess the ability to regenerate MWHB. Ten milliliter of n-hexane was used to wash the sorbed SEO from 0.2 g of MWHB. After three repetitive washings, the used MWHB was dried and reused to remove 1 g of SEO. This sorption–desorption process was performed for four repetitive cycles.

The result justified that WHB with different particle sizes ranging 6.3 mm to 425 µm could remove more than 90% of SEO. However, on using large particle size (6.3–4.75 mm) WHB, only the lower surface was exposed to SEO slick, whereas the upper part of WHB remained unexposed due to large particle size and floating nature. As a result, 100% SEO removal with a 6.3–4.75 mm particle size of WHB was not achieved. On the other hand, particles ranging from 850 to 425 μm were extremely small, making their separation after SEO sorption more challenging.

Based on the particle distribution diagram (Figure 1), type-1 and type-2 WHB were found in greater quantity and the SEO sorption capacity was more than 90% for each. Therefore, these two types of WHB were used in a 2:1 ratio. Also, a sorbent size of 1–2 mm can perform better removal of oil (Boleydei et al. 2018). A blend of one portion of type-2 and two portions of type-1 WHB samples was prepared for further experiments. Two grams per liter of blended WHB was used to remove 5, 10, 15, and 20 g/l of SEO as reported in Figure 2(b). The removal of 5, 10, 15, and 20 g/l of SEO was 90% (2.21 g/g), 95% (4.66 g/g), 94% (6.95 g/g), and 88% (8.73 g/g) with 2 g/l of blended WHB. Due to the high amount of SEO (15 and 20 g/l) used in the experiment responsible for increasing the thickness of the oil spill in the beaker containing water, a good amount of SEO removal was observed due to the availability of more surface area for interaction between WHB and SEO. However, high SEO concentration in the reaction beaker could also attach more SEO on the inner wall beaker resulting in a huge loss of SEO. Hence, 10 g/l was used for further experiments and analysis.

The lauric acid and other fatty acids in coconut oil demonstrated higher hydrophobic properties and the ability to form molecular interactions with the sorbent surface, contributing to an increased contact angle and enhancing the hydrophobicity of the sorbent (Navarathna et al. 2020; Gurav et al. 2021). The materials with contact angles beyond 90° were considered more hydrophobic as reported in lauric acid-modified Douglas fir biochar (Navarathna et al. 2020). Enhanced hydrophobicity helps the sorbent to float which could help the easy recovery of sorbent from the water surface after petroleum oil sorption. Besides, the carboxylic acid groups present in fatty acids could interact with oleophilic alkyl chains in petroleum oil that help higher SEO sorption.

Additionally, an identified peak at 1,742 cm⁻¹ can be linked to the stretching of the C=O bond, typically found in carbonyl groups. Moreover, a peak that appeared at 1,598 cm⁻¹ was associated with the C=C bond stretching, a characteristic often seen in aromatic compounds present in both the sorbent and the SEO. Nevertheless, the MWHB displayed further intense peaks at 2,922 and 2,852 cm⁻¹, which are associated with the stretching of C−H bonds in CH₂ and CH₃ groups, signifying the integration of fatty acids from coconut oil into WHB (Navarathna et al. 2020; Gurav et al. 2021, 2023). Furthermore, the appearance of a new peak at 1,742 cm⁻¹, associated with the stretching of the carbonyl groups present in fatty acids, was noted, indicating an enhancement in hydrophobic properties of MWHB (Sidik et al. 2012; Gurav et al. 2021). Furthermore, wide peaks were observed at 1,153 and 1,006 cm⁻¹. These are believed to be due to the stretching of the C−O−C and C−O within lignin, and those within cellulose and hemicellulose, respectively (Pandey & Pitman 2003). Additional peaks in MWHB were identified at 1,317 cm⁻¹ (related to CH₃ bending) and 1,153 cm⁻¹ (indicating C=O stretching). Hence, the presence of saturated C−H, C−C, CH₂, CH₃, and C=O groups on MWHB suggested the biomass possessed increased hydrophobic and oleophilic properties (Yin et al. 2017; Ribeiro et al. 2000; Oliveira et al. 2020; Gurav et al. 2021, 2023).

Nevertheless, the peak intensity at 2,923 and 2,853 cm⁻¹ for C–H justifies asymmetric stretching in CH₃ and CH₂ groups in alkane hydrocarbons was increased after SEO sorption on MWHB. Two new peaks appeared at 1,461 and 1,377 cm⁻¹ after the SEO sorption was detected. The peak at 1,461 cm⁻¹ can be attributed to the asymmetric bending mode of CH₃ groups, which are abundant in petroleum hydrocarbons. Similarly, the peak at 1,377 cm⁻¹ revealed the presence of symmetric C–H bending in SEO (Sidik et al. 2012).

This rapid initial sorption is more likely attributed to the accessibility of numerous available surfaces, functional groups, and vacant pores on the surface of MWHB. However, as time progressed, the sorption rate gradually decreased, reaching equilibrium after 75 min and achieving a maximum SEO removal of 96.3% (4.77 g/g). It was observed that as the sorption time was extended from 2 to 75 min, the amount of SEO uptake increased. Beyond 75 min, the rate of sorption remained constant, suggesting that the surface of the MWHB was completely covered with SEO, therefore equilibrium was reached. To ensure that a complete equilibrium condition was achieved, the observation was continued for 120 min. It was noted that approximately 60 min was the effective time for the sorption of SEO, hence, 60 min was the ideal contact time set for later experiments. The R² values and the linear-fit graphs for all the kinetic models are shown in Figure 8(b)–(d) and Table 2. The sorption process of SEO on MWHB was effectively described by the pseudo-second-order model (R² > 0.999), which aligns with the findings from previous research on petroleum oil sorption (Sidik et al. 2012). This model suggests that the step that limits the rate of sorption involves a chemical interaction between the hydrophobic tails on MWHB and the sorbate surface (Cheu et al. 2016; Boleydei et al. 2018; Khamizov 2020; Yasid et al. 2022).

Table 2

Kinetic and isotherm parameters for the sorption of SEO on MWHB

.	Model .	Parameters .	Value .
Kinetic	Pseudo-first order	Qe	0.361
		K1	−0.0006
		R2	0.908
	Pseudo-second order	Qe	4.774
			22.793
		K2	1.144
		R2	0.999
	Elovich	α	21.182
		β	96.535
		R2	0.940
Isotherm	Langmuir	Qm	4.998
		K1	5.622
		R2	0.992
	Freundlich	1/n	0.723
		Kf	8.816
		R2	0.999
	Temkin	A	1.359
		Kt	50.948
		R2	0.865

.	Model .	Parameters .	Value .
Kinetic	Pseudo-first order	Qe	0.361
		K1	−0.0006
		R2	0.908
	Pseudo-second order	Qe	4.774
			22.793
		K2	1.144
		R2	0.999
	Elovich	α	21.182
		β	96.535
		R2	0.940
Isotherm	Langmuir	Qm	4.998
		K1	5.622
		R2	0.992
	Freundlich	1/n	0.723
		Kf	8.816
		R2	0.999
	Temkin	A	1.359
		Kt	50.948
		R2	0.865

View Large

Among these models, both Langmuir and Freundlich isotherms proved to be the most suitable for describing the sorption of SEO, as evidenced by higher regression coefficient values (R²) of 0.992 and 0.999 (Table 2). The Langmuir isotherm model suggests that the sorption of SEO onto the MWHB surface occurs as a homogeneous monolayer process. This implies that the SEO molecules were evenly sorbed on the active surfaces of MWHB susceptible to holding only one kind of sorbent on each side, and all these sites possess equipotential energy levels. According to the Langmuir analysis, the maximum sorption capacity (Q_m) reached 4.99 g/g for MWHB, whereas the experimental value was 4.75 g/g. Furthermore, the Freundlich model justified that the sorption process might have occurred in a heterogenous multi-layer after the monolayer sorption mechanism (El-Hassouni et al. 2013; Senturk et al. 2016; Boleydei et al. 2018). However, a higher precision, with an R² value (>0.999), suggests that the Freundlich isotherm model is a more accurate description of how SEO attaches to the MWHB. MWHB sorbent might have a heterogeneous surface with a non-uniform distribution of active sites, capable of sorbing SEO droplets in a multi-layer fashion. With a K_f value of 8.816, it is understood that higher K_f values correlate with higher sorption capacity. The literature supports that attachment is effective when the n value is between 1 and 10, and a higher n value means higher sorption intensity. Here, the n value is 1.38, indicating a favorable attachment of oil to the MWHB. On the other hand, the Temkin model showed a lower R² value of 0.865, which suggests that it does not fit the experimental data as well, pointing to a less satisfactory model for describing the isotherm model (Sidik et al. 2012).

Evaluating the reusability of sorbents was an essential step in the development of a cost-effective and environmentally friendly technology to prevent further pollution when disposing of spent biomass (Gurav et al. 2021). While the sorption efficiency of MWHB decreased with each cycle of sorption and desorption, it is noteworthy that the sorbent still maintained an excellent performance, with a sorption efficiency of over 90% even after the third sorption–desorption cycle as shown in Figure S1 (Supplementary Data). This indicates effective reusability characteristics for the sorbent. The decrease in the sorption capacity with each cycle might be attributed to SEO components that were not fully removed during the treatment with n-hexane (Kim et al. 2020). Similarly, the decline in performance could be linked to a loss in the fatty acid content on MWHB. The carboxylic acid groups on fatty acids actively participate with the oleophilic C–H (alkyl) chains, therefore loss of fatty acid composition could be another reason for the decrease in sorption performance (Cai et al. 2019; Ji et al. 2020; Gurav et al. 2021).

Among the different MWHBs, the MWHB produced using an equal portion of extra virgin coconut oil and WHB followed by treatment with 10% methanol solution demonstrated a higher hydrophobicity and SEO sorption. The best-fit sorption kinetics and isotherms indicated a chemisorption mechanism involving initial monolayer coverage of the MWHB surface followed by multi-layer coverage with SEO. Due to the increased hydrophobicity of MWHB, a higher sorption capacity was observed as compared to raw WHB. Furthermore, reusability testing of MWHB demonstrated over 90% sorption efficiency even after the third sorption–desorption cycle, highlighting its excellent reutilization properties. Therefore, modifying WHB with extra virgin coconut oil could offer an inexpensive and sustainable approach for treating water contaminated with petroleum hydrocarbons. In real-scale applications, floating MWHB could be retrieved using other mechanical techniques such as fishing nets (smaller pore size than biomass), booms, or skimmers.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.