Bottle Gourd (Lagenaria vulgaris) Shell as a Natural, Biodegradable, Highly Available, Cheap, Agricultural by-Product, Miscellaneous Biomass, Ion Exchanger, Biosorbent and Fertilizer

Abstract

The increased interest in natural, renewable, biodegradable, easily available, low-cost materials makes agricultural residues, such as lignocellulosic biomass, attractive raw materials for the preparation of effective biosorbents for various pollutants (metal ions, anions, dyes, pharmaceutical degradation products, metabolites, organic macromolecules) for the wastewater treatment. Various covalent and non-covalent modification approaches significantly improve the sorption properties of these lignocellulosic functional particles, even improving their dispersion in hydrophobic polymer matrices, associative properties in water, rheological properties, surface-active properties, which can control the sorption of various ionic pollutants both in batch and in flow mode. Advantages over commercial sorbents (techno-economic aspect, no secondary pollution, usability as fertilizers), easy separation from the sorption medium, microstructural properties (strength, porosity, interactivity, stability), as a promising and sustainable biosorbent highlight the environmentally friendly bottle gourd shell. On the example of this biosorbent, the conventional approach to the pollutant sorption process (comparative kinetic, thermodynamic and equilibrium tests) was improved, as well as its shortcomings in predicting optimal process parameters. To fill the gaps of the already unnecessary numerous experiments, a design study involving OVAT experimental approaches integrated with DoE methodology was conducted. This integrated experimental design was implemented in the optimization of the pollutant sorption process.

1. Introduction

The generation of wastewater on a large scale is an inevitable consequence of today’s modern society. One of the reasons for this environmental problem is the rapid growth of the human population, which has a significant impact on the natural waters pollution by municipal waste. On the other hand, the progressive development of industry has led to increased production of industrial wastewater and its discharge into natural waterways. The generation of municipal and industrial wastewater, polluted with organic and inorganic pollutants, is becoming an increasing problem for human health and the environment.

Various conventional methods are used for wastewater treatment (chemical precipitation, oxidation-reduction, coagulation and flocculation, sedimentation). More efficient, but also more expensive procedures have been developed, such as: ion exchange, reverse osmosis, electrodialysis, ultrafiltration, membrane separation and adsorption with activated carbon. Due to its advantages, such as ease of use and the efficiency of removing various pollutants, adsorption with activated carbon has become one of the most commonly applied wastewater treatments. However, due to the high costs of obtaining and the complexity of the regeneration process, which causes secondary pollution, the application of activated carbon is limited today.

In order to overcome the shortcomings of this adsorbent, numerous studies have shown that various agro-industrial waste biomasses can be used alternatively for these purposes [1, 2]. There are well-known examples of sorbents based on: tree bark, fruit peels (banana, orang, lemon, grapefruit), woody shell (walnut, peanut, almond), pits (plum, cherry, peach), seed husks (rice, wheat, pumpkin, sunflower), straw, algae and others. Of particular importance is the pretreatment of biomass, for which numerous technologies based on physical, chemical and biological methods have been developed [3].

Apart from native biomass as cheap natural lignocellulosic polymers, sorbents that can be obtained through a series of simple physical and chemical modifications of waste biomass are of significant interest for these purposes [4, 5, 6]. Today, biomass modification is increasingly the subject of scientific research due to improved characteristics and their potential application in numerous fields. Modified sorbents based on lignocellulosic material have an advantage compared to synthetic polymers, considering that they are significantly cheaper, renewable and less toxic [7]. In addition, they represent a functional ecological material due to unique properties, such as: hydrophilicity, adsorptivity, biocompatibility and biodegradability.

Modification of lignocellulosic material, by introducing specific functional groups, is of interest for obtaining sorbents with improved characteristics. For example, cationic functional groups (from quaternary ammonium salts) increase the sorption capacity for anions, primarily the ion exchange capacity [8]. Sorbents with cationic active centers are of particular importance for the treatment of wastewater containing large amounts of various anionic pollutants (nitrates, phosphates, sulfates, cyanides, etc.).

It is characteristic that municipal and agro-industrial waste waters increasingly contain excessive concentrations of phosphates and nitrates. These anionic species represent a serious problem of endangering the environment (they are directly responsible for the occurrence of eutrophication), and can also lead to serious health problems. Considering that these anionic species are present in water in the form of soluble salts, conventional procedures for water treatment (coagulation, sedimentation) have not proven to be effective for their removal. Developed physical-chemical (reverse osmosis, electro-dialysis) and biological denitrification methods [9], as well as the application of commercial ion-exchange resins for the removal of nitrates and phosphates from wastewater [10], are not sustainable from the economy aspect. Therefore, further research was focused on sorption processes, using cheap agricultural waste as ecological sorption material. Various studies have investigated the cationic modification of biomass via quaternization of lignocellulosic material with amino [1, 2] or ammonium compounds [11], as well as surfactants [12]. In this sense, various agricultural by-products were examined: wheat straw, sugar cane, rice husk, corn grain, sawdust, etc. [13, 14, 15, 16].

Bearing in mind the mentioned facts and insufficiently solved current ecological problems, the development of new cationic sorbents will always be an open topic of research. First of all, these researches are based on the discovery of new usable sorbents of natural origin, as well as the improvement of their sorption characteristics. In this sense, the attention of our research group during the last decade was attracted by the fact that in the literature there was no data on the use of the Lagenaria vulgaris shell (LVS) as a sorbent of anionic pollutants. Therefore, as a contribution to this field, our research was focused on the development of new cationic sorbents through the lignocellulosic biomass modification of LVS, a plant by-product that had not been used for these purposes until then.

Namely, due to its nutritional and medicinal properties, the bottle gourd (Lagenaria vulgaris species) is increasingly cultivated in many countries of the world. It is a good prerequisite for the availability of a significant amount of LVS as a potentially valuable agro-waste [17, 18, 19]. The bottle gourd shell itself is characterized by a lignocellulosic composition, an extremely compact and solid structure in the mechanical sense, as well as a porous structure, with low ash content and a high volumetric mass. Tests have shown that LVS biomass, as well as its modified products, do not swell in water and are easily separated from the water phase after treatment [20]. These are very important properties of a precursor for the sorbents synthesis, which are not found in most other lignocellulosic biomasses. Additionally, LVS lignocellulosic biomass contains hydroxyl groups on the glucopyranose units of cellulose as the main functional groups, which are easily susceptible to various chemical reactions. In order to synthesize cationic sorbents, research into the chemical modification of LVS biomass was focused on the ammonolysis of alkaline treated or etherified cellulose, as well as the grafting of a cationic surfactant onto the lignocellulosic skeleton of biomass [21]. The obtained cationically modified products were expected, first of all, to have a high sorption capacity for various anionic species. In this sense, the optimization of the synthesis process parameters and the definition of the product sorption capacity towards nitrates and phosphates from the aqueous solution are of particular importance. Feedback on the influence of process parameters on the formation, amount and type of active sorption centers is important for improving the efficiency and selectivity of the anionic species sorption process. Sorption tests of nitrates and phosphates in a mixture with other anionic species can be helpful in the application of these cationic sorbents in real conditions of municipal or industrial wastewater treatment.

From all of the above, potential sorbents based on LVS have become the main subject of our research. Initially, LVS in its native form [20] or in some chemically modified forms (xanthated, sulfonated, and methylsulfonated) [22, 23, 24, 25] was investigated as a sorbent for the removal of heavy metals from water. Later, research was focused on the removal of other pollutants from aqueous solutions, such as dyes [26], herbicides and pharmaceutical substances [27]. More recent research is related to the modification of sorbents to solve the problem of eutrophication, primarily the removal of phosphates and nitrates, which is the subject of this book chapter.

Considering the development and improvement of new sorbent materials, in addition to the expected scientific contribution, the research presented in this chapter has great practical importance from both the technological and ecological aspects. Due to changes in the structure and chemical composition of the starting raw material, as well as the predicted ion exchange mechanism, the advantage of the obtained sorbents is the possibility of multiple regeneration and reuse. On the other hand, it solves the problem of disposal of anion-saturated biomass, which can be composted and used for fertilizing agricultural land with renewable nutrients (phosphates and nitrates).

2. Bottle gourd (Lagenaria vulgaris)

The bottle gourd (or flask gourd) is one of the 6 known species of the gourd genus Lagenaria. The calabash or Lagenaria siceraria (synonym Lagenaria vulgaris Ser.) is the best-known species of aerial annual creeping plant from the kingdom Plantae (division Magnoliophyta), squash family Cucurbitaceae, genus Lagenaria (Cucurbita lagenaria L.) [28]. This species is one of the first cultivated plants in the world (indigenous species), mainly for its fruit. It grows on sandy soils and red loam, on flat surfaces and moderate slopes of the ground [17]. Given that Lagenaria vulgaris requires a lot of light and heat, it originates from Africa or Asia (morphologically different autochthonous species), from where it was transferred to other countries of the world. Apart from bottle gourd, other species are not cultivated.

2.1 Bottle gourd valorization

Young fruits of Lagenaria vulgaris (LV) have light-green smooth skin and spongy white pulp, in which the seeds are arranged (Figure 1). Harvested young gourds are used as vegetables. Mature gourd in the shape of a bottle reaches a size of about 50 × 20 cm, and a weight of up to 1.2 kg. The gourd shell is hard and woody, and very bitter in taste. After matures, the gourd dries completely naturally, so only the seed remains in the fruit. As the shell matures and dries, it acquires a yellow or light-brown color.

Figure 1.

Lagenaria vulgaris: Flower (a), young fruit (b), spongy tissue with seeds (c), dried pulp (d) and dried gourd shell (e).

The Lagenaria vulgaris is a plant of economic importance, considering that the pulp has nutritious and pharmaceutical properties [18]. The pulp contains 85–95% water and is low in calories (20 kcal/100 g). In addition to water, it also contains some carbohydrates, proteins, vitamins A, B and C, as well as minerals (calcium and iron). In traditional medicine, the pulp is known as a medicine for the treatment of skin diseases. The pulp can be used as a sedative and emetic purgative, diuretic, as well as for pectoral complaints. In addition, phytochemical and pharmacological studies have indicated the potential treatment of heart failure, hypertension, bilious disorders (jaundice), diabetes, ulcers, hemorrhoids, etc. [18]. Extracts of this plant have been proven to have antioxidant and antibiotic activity [19].

Lagenaria vulgaris is cultivated not only for its nutritional and medicinal properties but also for the practical use of the dried shell (LVS). After ripening and drying, the shell is used for various purposes, such as water storage vessels (the so-called “lejka” in Serbian), funnel, pipes, accessories (musical instruments, liquid storage containers).

2.2 Physico-chemical properties of bottle gourd shell

The bottle gourd shell does not retain a large amount of water in its structure (moisture content 3.8%), which is why the crushed material can be stored in the air without sticking particles or changing granulation. The mineral substances of the shell are mainly alkaline earth metals (Ca and Mg), which the plant accumulates during its growth. This inorganic fraction is mostly incorporated into the highly developed porous structure of biomass and is found in conjunction with surface functional groups. The gourd shell is characterized by relatively low values of density and bulk weight, which is important for application as a sorbent [20].

Based on Boehm’s method for determining the proportion of functional groups, the concentration of strongly acidic groups of 0.138 mmol/g and very weakly acidic groups of 0.104 mmol/g was determined in the LVS native form. The amount of weakly acidic groups is slightly higher and is 0.218 mmol/g [26]. The specific surface area and porosity of the LVS were analyzed using the isotherm of the nitrogen sorption/desorption process [29]. The analysis showed that nitrogen saturation is achieved at relatively low gas pressure values. The small value of the sorbed gas volume at the moment of equilibrium (0.28 cm³/g) indicated a small specific surface area of the material, without pronounced porosity. Using the BET method, a total specific surface of 1.075 m²/g was determined, whereby the outer surface of the shell (0.226 m²/g) has a relatively small share in the total surface. The shell surface is characterized mainly by micropores (1.048 m²/g) of small volume (4.96∙10⁻⁴ cm³/g), while the remaining specific surface is represented by macropores and surface elements of significantly larger dimensions than standard pores (cavities and channels). The presence of macropores and larger cavities can enable better movement of the aqueous phase through the structure of the material and promote internal diffusion [30].

The constitutional analysis of the LVS structural components showed that this lignocellulosic biomass mainly consists of holocellulose (cellulose and hemicellulose) and lignin [21, 29]. These are typical structural constituents of woody plants or the bark of many other similar plant products (pumpkin, corn on the cob, chestnut, hazelnut, peanut and others) [1]. Cellulose and hemicellulose are generally classified as polysaccharides, while lignin is a complex phenolic (aromatic) polymer. Cellulose does not dissolve in water, diluted acids and bases. It has a well-ordered (crystalline) and randomly ordered (amorphous) structure. This property is of particular importance, because it makes the cellulose material chemically and mechanically stable and resistant to aqueous solutions. Hemicellulose has a more complex structure (pentose-hexose composition) and bonds than cellulose. It is linked to the cellulose microfibrils by non-covalent bonds in an amorphous matrix, thus holding the rigid cellulose fibers in place. As a polysaccharide of amorphous nature (due to branching), hemicellulose is sensitive to acids and bases in aqueous solution and more susceptible to depolymerization. Lignin is a macromolecule built from phenylpropane units derived from p-coumaryl, coniferyl and sinapyl alcohol as monomers. The composition of macromolecules includes methoxyl, acetyl and formyl groups. As an amorphous polyaromatic substance, in addition to the basic phenylpropanoid units, lignin also contains aliphatic constituents that affect the increase in hydrophobicity. Of the 11 types of intermolecular bonds within the molecule, the most abundant is the β-O-4 ether bond (60%), which leads to linear stretching of the polymer. Lignin is stable in contact with water, weak bases and acids. However, it undergoes chemical changes under the influence of elevated temperature. The three-dimensional structure of lignin, interwoven with cellulose and hemicellulose molecules, gives LVS biomass distinct hardness and mechanical stability. Due to the strong mutual interaction of phenylpropanoid units with other lignin molecules and polysaccharides, difficulties arise in the isolation of lignin from biomass. Therefore, for applying any method of delignification or changing its structure, the inevitable breaking of strong covalent bonds must be taken into account.

Such structural properties of lignocellulosic biomass are of crucial importance for the direct application of LVS as a sorbent of various pollutants, primarily heavy metals or other cationic species from waste or contaminated natural waters [27]. Considering the structural characteristics, LVS in contact with water does not swell and does not change its macroscopic or microscopic form. Also, after separation from the aqueous phase, LVS is very easily washed and dried, which is important for its regeneration and reuse. Both structural components (cellulose and lignin) contain a large number of different oxygen functional groups, which are responsible for binding cationic pollutants [20].

In addition, during the physico-chemical treatment, the LVS retains the solid nature of the structure, built of cellulose and lignin, which is characterized by mechanical resistance and compactness as important characteristics for the LVS application as a sorbent.

2.3 Sorption characteristics of bottle gourd shell

Raw LVS biomass does not possess sorption properties or exhibits them to a very small extent (<1% of various sorbates), which is not of practical importance for these purposes [21]. However, modified LVS biomass can be a good sorbent for a wider range of pollutants. Like other lignocellulosic biomasses, LVS can be modified by chemical treatment or thermal carbonization. Biomass modified with various chemical agents is suitable for removing cationic pollutants from water, such as heavy metals and cationic dyes. Biomass modified by cationic grafting of cellulose chains is important for the purification of waters contaminated with anionic species. Activated carbon obtained by thermal carbonization of biomass and steam activation is a good sorbent of organic, non-polar and weakly polar pollutants, such as some pharmaceutical substances and pesticides.

Based on numerous methods of chemical modification, such as acid-base activation [20], sulfonation [22], methyl-sulfonation [23], xanthation [24], several procedures have been developed to obtain new, more efficient and economical LVS sorbents. The obtained materials were characterized in detail by various physico-chemical methods and applied as sorbents for the removal of various cationic species from aqueous solutions in a batch mode. The results of our research confirmed that sorbents based on modified bottle gourd shell can be used for effective removal of heavy metals, primarily Pb(II), Cd(II), Zn(II), Cu(II), Ni(II) [20, 25, 31]. Also, we confirmed the sorption efficiency of other cationic pollutants from natural or wastewaters, such as the cationic dye methylene blue [26]. These sorbents represent an alternative to more expensive commercial sorbents based on activated carbons, as well as other sophisticated technologies.

The possibility of using LVS biomass as a sorbent for some drugs was investigated [30, 32], given their significant presence in municipal waters via the sewage system. Namely, drugs and their metabolites are excreted from the body through urine and feces, so it is necessary to remove them from wastewater before discharge into natural waterways (rivers and lakes). One such drug is ranitidine hydrochloride (RH), which is classified as a drug with hazardous effects on aquatic organisms. Additionally, RH is partially biodegradable and susceptible to structural change under the influence of sunlight, creating toxic photoproducts. In this sense, the sorption removal of ranitidine hydrochloride from an aqueous solution was monitored using Lagenaria activated carbon (LAC). Tests were performed in batch mode, in the presence and absence of ultrasound. LAC is characterized as a porous spongy material with an amorphous structure, high surface area (665 m²/g) and pH_PZC (point of zero charge) of 7.2. The process of RH removal is best described by the pseudo-second order kinetic model and the Langmuir isotherm, which indicated monolayer sorption and significantly faster equilibrium establishment in the presence of ultrasound. The maximum RH sorption capacity of 425 mg/g for ultrasound power of 50 W was achieved at ambient temperature. Thermodynamic data indicated that RH sorption is an exothermic and spontaneous process, implying a physical mechanism of RH removal from aqueous solutions by LAC. The obtained data were used to design a sorption process in order to remove other organic and weakly polar pollutants from aqueous solutions, such as herbicide 2,4-dichlorophenoxyacetic acid [27].

From an economical aspect, LVS as a precursor for the production of a superior carbon sorbent can be considered an extremely economical resource because the costs of its production are minimal [30]. Namely, the LV plant is grown without special requirements, without the use of pesticides and expensive agricultural procedures. On the other hand, the thermochemical conversion to the final carbon product consists only of the costs of a few cheap chemicals and the electricity used for heating. The final cost of the obtained LVS carbon is estimated at less than 500 USD/t, in contrast to commercial activated carbons present on the market in the form of powders and granules whose average value ranges between 600 and 2000 USD/t. Therefore, LVS activated carbon is a very interesting sorption material in terms of its commercialization and future practical use.

However, in the scientific literature at the time, there were no known methods of modifying LVS in order to obtain sorbents for the removal of anionic species from aqueous solutions, which opened a new field of research and wider possibilities of application of this economic and environmentally acceptable agro-waste material.

3. Eco-sorbents based on cationic LVS

Further research was carried out in the direction of the development of an original cationic sorbent based on lignocellulosic plant material, with the property of efficient sorption of anions from aqueous solutions (primarily phosphates and nitrates). In this sense, as a precursor for the synthesis of a new eco-sorbent, LVS biomass was investigated as a valuable agro-waste that had not been used for this purpose until then. However, the high mechanical stability of LVS biomass, which originates from the cross-linked lignocellulosic structure, does not favor the process of further modification in order to develop new or improve existing sorbents for the removal of anionic pollutants from contaminated waters. In this sense, the destruction of LVS biomass and partial delignification by the pretreatment process, before the synthesis of an adequate cationic type sorbent, is desirable. Bearing in mind that lignin is very stable in contact with water, weak bases and acids, as well as the fact that it is subject to chemical changes under the influence of elevated temperature, these properties must be taken into account when choosing the appropriate method and designing the biomass pretreatment procedure.

3.1 Pretreatment of lignocellulosic biomass

Pretreatment of lignocellulosic biomass involves partial delignification, and then depolymerization of hemicellulose, using appropriate reagents and optimal pretreatment conditions [33]. Delignification and depolymerization of hemicellulose increases biomass porosity, which during sorbent synthesis will enable better accessibility of agents for more efficient conversion of cellulose to the desired product. At the same time, the pretreatment enables the reduction of the required modifying agent amount, which represents an important economic and ecological effect. Pretreatment includes the separation of cellulose microfibrils from other biomass components (lignin, hemicellulose, pectin). This significantly increases the reaction surface, i.e. the number of reaction sites (primarily on cellulose) for the synthesis of the targeted sorbent. It should be kept in mind that the variability of the chemical composition and structure of the lignocellulosic material requires an appropriate pretreatment design. Therefore, numerous physical, chemical and biological methods have been developed for biomass pretreatment [3]. The choice of method should be in accordance with the expected outcome. In practical application, there are pretreatments of biomass: under mild alkaline conditions (for partial delignification while preserving cellulose), in an acidic solution (preferably for complete depolymerization of polysaccharides to monosaccharides), using organic solvents, by oxidative delignification (hydrogen peroxide or ozonolysis, preferably for lignin degradation), by biological methods (microbes) or microwave radiation [34].

In order to choose and design the pretreatment method, the physico-chemical characterization of LVS biomass was performed. The most abundant component in raw LVS biomass is lignin (41.90 ± 0.51%). The high content of holocellulose (57.80 ± 0.63%) indicates a significant presence of OH groups as active centers for further modification reactions. Holocellulose consists of cellulose (39.58 ± 0.42%) and hemicellulose (18.22 ± 0.14%). The protein content is extremely low (<0.1%), while volatile substances were not detected since the LVS was already naturally dried. The low ash content (0.28 ± 0.04%) mainly consists of bio-accumulated metals during plant growth, such as Ca, Zn, Cu and Mn (≈20 ppm). The biomass is characterized by a moisture content of 3.80 ± 0.14%, density of 0.46 ± 0.07 g/cm³ and a bulk weight of 103.03 ± 0.87 kg/m³. These data are consistent with the characteristics of other plant materials from the genus Lagenaria [20, 35].

In accordance with the established composition and purpose of the biomass, the alkaline pretreatment of LVS was chosen as the most suitable method. Considering that strongly alkaline reagents can lead to a significant loss of important polysaccharides, a green procedure of pretreatment with a green liquid as weakly alkaline reagent (1% Na₂CO₃ + 1% NaCl) was applied at a temperature of 23 ± 0.5°C, atmospheric pressure, for 2 hours with stirring (150 rpm). The resulting alkaline treated LVS biomass (LVSAT) was subjected to further characterization.

Namely, the most important parameters of pretreatment with green liquids are the biomass particle size (<10 mm, preferably about 1 mm), concentration of alkaline reagent, pH value, time and temperature [36]. In our case, for pretreatment purposes, LVS biomass was crushed and ground to a particle size in the range of 400–800 μm. The concentration of the applied reagent affects the yield of the solid phase of the biomass. During the pretreatment, a pH of around 9–10 is reached, which is significantly lower than in the case of treatment with alkaline cooking of biomass (pH 12–13). A lower pH value can effectively reduce the alkaline degradation and secondary reactions of polysaccharides during pretreatment. At the same time, a certain degree of swelling of cellulose fibers is achieved, which can be useful for further biomass modification. On the other hand, increasing the temperature shortens the pretreatment time, but affects the increased consumption of alkali and the formation of degradation products. Higher temperature contributes to the alkaline degradation of biomass, due to the breaking of ester bonds between lignin, hemicellulose and cellulose. Substituted uronic and acetyl groups on hemicellulose can be hydrolyzed to the form of uronic acid and acetic acid with more intensive alkaline degradation, which results in the formation of a larger proportion of acidic functional groups. Apart from the mentioned effects, alkaline pretreatment changes the macroscopic and microscopic properties of lignocellulosic biomass, including the size, structure and chemical composition of the particles.

In our case, the applied green procedure leads to a slight degradation of lignocellulosic biomass, while maintaining a higher cellulose content and increased delignification selectivity (Figure 2).

Figure 2.

Schematic presentation of LVS biomass pretreatment (a) and corresponding SEM micrographs (b) with images of raw LVS biomass before treatment (left) and brown-red filtrate after alkaline treatment of LVS biomass (right).

The advantage of applied alkaline pretreatment is the mild destruction of the lignostic part of the biomass, during which the ester bonds associated with hemicellulose xylan and lignin are hydrolyzed. Mild reaction conditions prevent the condensation of the released lignin fragments, which affects their greater solubility and their better removal from the biomass. The released hemicellulose dissolves in water and undergoes hydrolysis under the alkali action, which leads to its conversion into water-soluble monomers. Research has shown that this process removes uronic (through uronic acid) and acetyl groups (through acetic acid) present in hemicellulose. In this way, the availability of various chemical agents to the cellulose chain increases. These findings were confirmed by FTIR spectroscopic and SEM-EDS morphological analysis of biomass samples before (LVS) and after (LVSAT) the pretreatment procedure [37].

Under the specified pretreatment conditions, the mass loss of the initial LVS sample (15.2%), color change (from colorless to brown-red) and pH value of the filtrate after washing the LVSAT have been found. A significant decrease in the content of hemicellulose (71.4%) and partially lignin (5.2%) was registered, which directly affected the enrichment of biomass with cellulose (from 40 to 47%). Intense coloring of the reaction solution (from colorless to brown-red) during alkaline pretreatment of biomass originates from released pigments, hetero-saccharides and resulting degradation products (organic resin compounds, oxidized phenols and organic acids). Also, pectin (polymer of polygalacturonic acid) is released, whereby only that part of the pectin network where the glucosidic bond was previously broken can be separated. As a result of hemicellulose and pectin depolymerization, cinnamic, glucuronic and galacturonic acids are released. On that occasion, metal ions (such as Ca(II)) that kept the pectin structure more stable are also removed. The high pH value of the initial solution (about 10.1) during biomass pretreatment leads to breaking of ester bonds between lignin, hemicellulose and cellulose, as well as secondary reactions of polysaccharides (mainly released hemicellulose, which is much more sensitive to hydrolysis). Also, substituted uronic and acetyl groups on hemicellulose hydrolyze to the form of uronic and acetic acids, which results in the formation of more acidic substances. It is these organic acids that lead to a decrease in the pH value (from 10.1 to 8.6 during pretreatment). Used alkalis and released mineral substances are converted during the process into non-renewable salts. In order to remove undesirable salts and released oligosaccharide or monomer constituents, the biomass is filtered and subsequently washed with deionized water to a colorless filtrate (pH 6.8). After the applied pretreatment, it was determined that about 78% of the total polysaccharides are retained in the treated LVSAT biomass, which is important for further application as a sorbent synthesis precursor.

3.2 Functionalization of lignocellulosic biomass by cationic grafting

The increased interest in natural, renewable and biodegradable materials makes lignocellulosic biomass (primarily agricultural by-products or industrial waste biomass) important raw materials for the preparation of eco-sorbents, especially in the field of purification of polluted natural and wastewater [7]. Biomasses with cellulose as the main constituent are generally hydrophilic in nature. The presence of polar functional groups improves the stability of biomass in aqueous solutions, which is an advantage of these materials. Additionally, biomass as relatively porous materials is characterized by a high surface/volume ratio, which means that they have a highly reactive surface that can be easily functionalized [26]. The primary reactive sites for surface modification are mainly accessible hydroxyl and carboxyl groups. Various non-covalent and covalent modification approaches, such as radical polymerization and graft reactions, have been developed for the functionalization of biomass particles, as well as for improving their dispersibility in hydrophobic environments [38]. Alternatively, surface modification of lignocellulosic biomass can be achieved by sorption of the appropriate surfactant type (cationic or anionic), which is considered a non-covalent surface modification [12].

Surfactant-modified lignocellulosic biomasses have attracted a lot of attention as sorbents of various pollutants, considering that they are characterized by good performances, such as: low cost, biodegradability, suitable rheological properties, associative properties in water and surface-active properties that enable the sorption of various ionic pollutants matter [15, 16]. By the way, surfactants are defined as surface-active substances that greatly reduce the surface tension of water when used in very low concentrations [39]. However, one of the disadvantages of using surfactants for surface modification of the porous solid phase is that a large amount of surfactant is required due to the high specific surface area of the material. In this regard, some methods have been developed that are based on the use of a reduced amount of cationic surfactant [40]. For the surface modification of lignocellulosic biomass, the following cationic surfactants were mainly used: octadecyltrimethylammonium bromide (C18, OTAB), cetyltrimethylammonium bromide (C16, CTAB) or chloride (C16, CTAC), cetylpyridinium bromide (C16, CPB) or chloride (C16, CPC), tetradecyltrimethylammonium bromide (C14, TTAB) and dodecylpyridinium chloride (C12, DPC) [12]. For the purpose of producing sorbents, numerous biomasses or other substrates were modified: zeolite with CTAB for phosphate removal [41]; coconut kernel with CTAB to remove chromate, sulfate and thiocyanate [42]; yeast and lichen biomass with CTAB for chromate removal [43]; barley straw with CPC to remove food and mineral oils [44]; silkworm skeleton biomass with CTAB to remove anionic methyl orange dye [45]; cellulose fibers with CTAB [46]; wheat straw with CPB to remove anionic dyes [47]; tea waste with CTAB and CPB to remove the anionic Congo red dye [48].

In our case, prepared LVSAT biomass was modified with CTAC surfactant and investigated as a phosphate/nitrate sorbent in order to solve the problem of the eutrophication phenomenon [21]. Biomass modification was carried out at a temperature of 23 ± 0.5°C, during 8 h with mixing (150 rpm). To evaluate the sorption effect, two samples were prepared depending on the concentration of the surfactant solution: sample LVSAT-CTAC1 (0.9 × CMC) and sample LVSAT-CTAC2 (2 × CMC), where CMC is the critical micellar concentration of surfactant (1.28 mmol/dm³).

In order to evaluate the added cationic groups in the surfactant-modified lignocellulosic biomass, elemental analysis (CHNS/O) was performed first. Changes in the content and type of elements of the initial LVSAT biomass and the obtained cationic product are shown in Table 1. As a function of the nitrogen (N) content, other characteristic parameters were also determined, such as: reaction efficiency (RE), degree of substitution (DS), amount of added cationic groups (N_ad) and product yield (Y_pr).

Samples	Elements (%)				Nad (mmol/g)	RE (%)	DS	Yield (%)
	C	H	O	N
LVS	45.46	5.98	48.55	0.01	—	—	—	—
LVSAT	45.51	5.84	48.65	0.01	—	—	—	84.4
LVSAT-CTAC1	52.33	6.67	40.20	0.79	0.56	97.39	0.11	97.4
LVSAT-CTAC2	60.09	6.95	31.52	1.43	1.01	78.91	0.25	97.9

Table 1.

Elemental analysis and reaction efficiency of initial (LVS), alkali-treated (LVSAT) and its surfactant-modified (LVSAT-CTAC) lignocellulosic biomass.

N_ad = (N_product–N_biomass)⋅0.714; DS = 162⋅N_product/(100⋅14−M⋅N_product); RE = 100⋅N_ad/Q_surfactant; and Y_pr = 100⋅(m_product−m_biomas).

The elemental analysis of the samples determined that the value of CHO content in the biomass (H/C ratio 0.13) did not change significantly after the alkaline pretreatment. The presence of sulfur was not identified in the tested samples. Very similar results were observed in the case of research on other biomass [1, 20]. Significant changes in the nitrogen content were found in both samples of surfactant-modified biomass, which confirms the effectiveness of the cationization reaction. A higher N content in the LVSAT-CTAC2 sample was registered as a result of a more concentrated CTAC reagent solution (>CMC surfactant). The increase in N content was proportional to the concentrations of the solutions used. This fact indicates a possible mechanism of two-layer aggregation of surfactants on the biomass surface [42]. Based on the amount of added cationic groups in the biomass after modification, the theoretical ion exchange capacity of the LVAT-CTAC2 sorbent of 1.01 mmol N/g was calculated.

Additionally, high values of reaction efficiency and product yield indicate maximum utilization of the CTAC reagent used. However, the small degree of substitution and the predicted two-layer mechanism suggest that the obtained data cannot be relevant for the sorption efficiency evaluation of anionic species from the solution [49]. This can be explained by the fact that at least half of the total amount of bound surfactants is occupied by interaction with biomass active centers, while the rest is available for the sorption process of anionic pollutants (Figure 3). Accordingly, for LVAT-CTAC2, we should expect an exchange capacity that is half the theoretically calculated one.

Figure 3.

Model of the CTAC surfactants organization on the LVSAT biomass surface and sorption of anionic species from aqueous solution (modified from Ref. [50]).

The effectiveness of the surfactant-modified LVSAT biomass as a cationic sorbent was evaluated by means of the sorption capacity for the tested anionic species (phosphates and nitrates). The mechanism of anions sorption involves ion exchange of weakly stable Cl⁻ counterions with ions of higher affinity (H₂PO₄⁻ or NO₃⁻) towards active (−NR₄⁺) centers. As expected, low efficiency of anion sorption was registered with both sorbents (about 40% for phosphates and 22% for nitrates). Namely, the formation of two-layer surfactant aggregates increases the number of cationic added groups, but not the number of available active centers for interaction with the anions presents in the solution. In addition, the sorption efficiency is influenced by the pronounced steric effect of surfactant molecules. The layer of long hydrocarbon chains limits the access of anions to the less accessible active centers, distributed in the macroporous structure of the biomass. Also, the rheological properties of the solution due to the increase in surfactant concentration directly affect the difficult diffusion of anions through the solution during sorption, and thus lower sorption efficiency.

3.3 Synthesis of cationic biosorbents

Numerous cationization methods have been investigated for the chemical modification of lignocellulosic structures, mainly with secondary and tertiary amino reagents [2], as well as with quaternary ammonium reagents [11, 14]. The most commonly applied method of biomass cationization involves the cross-linking of epichlorohydrin (ECH) with appropriate amino reagents, such as trimethylamine (TMA), dimethylamine (DMA), diethylenetriamine (DETA), ethylenediamine (EDA), triethylamine (TEA), and others [1, 13]. Although the cationic sorbent synthesis method with ECH and TMA reagents is one of the most effective, with a high yield of the desired product as an anion exchanger, it should be noted that very toxic and harmful reagents are used, such as ECH, TMA and pyridine as a catalyst [51]. Therefore, less dangerous reactions for the modification of lignocellulosic biomass were recommended as an alternative [52]. Alternative methods for the synthesis of cationic sorbents include less toxic reagents, such as N-(3-chloro-2-hydroxypropyl)-trimethylammonium chloride (CHMAC) and glycidyl-trimethylammonium chloride (GTMAC). The application of these quaternary ammonium reagents is considered a safer and better option in terms of safer handling and environmental protection [14, 53].

In our case, following both synthesis procedures, two cationic biosorbents based on the prepared LVSAT precursor were synthesized in order to compare the efficiency of lignocellulosic biomass cationization, as well as their sorption characteristics. The LVSAT-TMA biosorbent was synthesized according to the first ECH-TMA method, while the less toxic CHMAC reagent was used for the synthesis of LVSAT-CHMAC biosorbent according to the second method [21].

According to the first ECH-TMA method, the reaction between ECH and cellulose is promoted by prior activation of primary and secondary OH groups of cellulose with an alkaline reagent (NaOH), resulting in alkali-cellulose with more reactive −ONa groups. The obtained hydroxy-cellulose ether is then cyclized in an alkaline medium using pyridine as a catalyst. Furthermore, epoxy-cellulose ether is used as an intermediate in the ammonolysis reaction. It is assumed that the ammonolysis reaction with TMA occurs after ring opening of the epoxy group, by condensation via the epoxy chloromethyl group in excess ECH [6]. ECH bound to the cellulose skeleton represents the active site of attack by amino reagents. Higher reaction efficiency and higher yield of the cationic product are achieved in the presence of N,N-dimethylformamide (DMF) as a medium and pyridine (PIR) as a catalyst. As a polar (hydrophilic) solvent with a high boiling point (152°C), DMF enables reactions that follow polar mechanisms to be carried out. On the other hand, as an aprotic solvent (it does not form hydrogen bonds, it solvates ions more weakly), it accelerates SN2 reactions given that nucleophiles are not strongly solvated. Although the reaction conditions (temperature and pressure) are not so drastic, in the case of interaction with alkali (NaOH) a part of DMF can be converted into dimethylamine, which increases the influence of the amino reagent on the cationic sorbent synthesis. Pyridine as a base has chemical properties similar to tertiary amines, so it is used as a catalyst for organic reactions. Apart from its presence in the solution accelerating the degree of wetting and causing the swelling of lignocellulosic biomass, pyridine is very easily protonated by available H⁺ ions, forming a pyridinium cation. The use of DMF medium and pyridine as a catalyst increases the sensitivity of the epoxy ring to react with the –OH groups of the glucopyranose unit of cellulose. For this purpose, the Lewis acid leads to the appearance of a lowering of the activation energy for ring opening and allows the subsequent attachment of epichlorohydrin to the cellulose skeleton, as the target molecule for the initiation of the reaction [54]. The flow of successive chemical reactions of the cationic LVSAT-TMA biosorbent synthesis is shown in Figure 4.

Figure 4.

Chemical reactions flow of the LVSAT-TMA biosorbent synthesis: Activation of LVS-cellulose (I), interaction of alkali-cellulose and ECH (II), conversion of hydroxy-cellulose to epoxy-cellulose ether (III), reaction of epoxy-cellulose ether with TMA to a cationic product (IV) [21].

In the second case, a procedure involving CHMAC reagent for the chemical modification of LVSAT precursor into LVSAT-CHMAC biosorbent was used. This sorbent synthesis procedure is simpler and more comprehensive compared to the previous ECH-TMA method (Figure 5). However, this synthesis process itself requires optimization of the reaction conditions which are peculiar to the nature of initial lignocellulosic biomass.

Figure 5.

Reaction path of the cationic LVSAT-CHMAC sorbent synthesis: Activation of LVS-cellulose monomer (I), reaction of alkali-cellulose with CHMAC reagent (II); potential active monomer groups are marked with asterisks [21].

Many studies have dealt with the optimization of cationic sorbent synthesis using the CHMAC reagent, alternately changing one of the important parameters, such as temperature (60–100°C), time (2–24 h), medium (ethanol, water) and the quantitative NaOH/CHMAC ratio (0.5–2.0) [13, 14]. In our case, after a series of experiments in the function of preliminary phosphate and nitrate sorption, the following optimal conditions for the synthesis of LVSAT-CHMAC biosorbent were defined: aqueous medium, 5 M NaOH solution (25 mmol NaOH/g biomass) temperature 80°C, time 10 h, and the quantitative NaOH/CHMAC ratio of 1:1 (20 mmol CHMAC/g biomass) [21].

The synthesized cationic biosorbents were characterized by elemental analysis and pH_PZC, including FTIR spectroscopy. Elemental (CHNS/O) analysis was applied in order to evaluate the added cationic groups in chemically modified lignocellulosic biomass. Changes in element content, reaction efficiency, and sorption properties of the synthesized biosorbents compared to LVSAT biomass are shown in Table 2.

Samples	Elements (%)				Nad (mmol/g)	RE (%)	DS	Yield (%)
	C	H	O	N
LVSAT	45.51	5.84	48.65	0.01	—	—	—	84.4
LVAT-TMA	47.67	6.34	44.25	1.74	1.24	4.86	0.22	94.6
LVAT-CHMAC	47.22	6.12	44.71	1.95	1.39	6.95	0.31	98.8
Biosorbents	Sorption efficiency
	Phosphates (10 mgP/dm^3)				Nitrates (10 mgN/dm^3)
	Qe (mg/g)		%		Qe (mg/g)		%
LVSAT	0.05 ± 0.03		1.0		0.04 ± 0.02		0.8
LVAT-TMA	3.94 ± 0.09		78.8		3.43 ± 0.14		68.6
LVAT-CHMAC	4.41 ± 0.13		88.2		3.95 ± 0.10		79.0

Table 2.

Elemental analysis and sorption characteristics of the synthesized cationic biosorbents compared to LVSAT biomass.

Based on the elemental analysis of the tested samples, it was determined that 17.36 mgN/g in the form of a cationic −N⁺R₃ group was bound on the biomass surface treated with the ECH-TMA reagent, which indicates the theoretical value of the LVAT-TMA sorbent ion exchange capacity of 1.24 mEq/g. In the case of the LVAT-CHMAC sorbent, it was observed that the quaternary ammonium CHMAC reagent contributes to a higher ion exchange capacity of the sorbent (1.39 mEq/g). Based on the DS and RE values, one gets the impression that substitution takes place on every fifth glucopyranose unit of the cellulose in the case of the TMA reagent, or on every third unit in the case of the CHMAC reagent. Assuming that RE is a property of the biomass rather than the amount of reagent, the initial concentration of the cationic reagent can be reduced. This is very important in a semi-industrial (scale-up) process, because it is one of the factors that determine the costs of the final product. This assumption can be explained by the fact that the reactive sites availability of lignocellulosic biomass is the limiting factor of reaction efficiency. In this sense, reaction barriers can be biomass density and the lignin amount. The bulk density of the biomass may be the result of the degree of lignification in this case, since the bulk density shows an inverse linear correspondence with the amount of added cationic groups [1]. Consequently, there may be diffuse limitations of the reaction between the cationic reagent and the reactive cellulose as a result of the bulk density (0.46 g/cm³ in the case of LVS biomass). Similar findings that delignification of lignocellulosic materials can increase the efficiency of the quaternization reaction have been reported for other highly lignified biomasses [55].

The results of testing the change in surface charge of LVS biomass and LVSAT-CHMAC biosorbent as a function of pH value (from 3 to 11) are shown in Figure 6. The higher value of pH_PZC biosorbent (7.04) compared to pH_PZC of biomass (6.37) indicates the presence of cationic (–N⁺R₃) functional groups that lead to an increase in the surface potential of the biomass. Also, the trend of decreasing potential of both biomass and sorbent with increasing pH (from 3 to 11) can be attributed to the influence of pH-dependent functional groups of biomass with negative zeta potential (OH group of cellulose and OCH₃ group of lignin skeleton). These groups show a higher negative charge with increasing pH, which results in a decrease in the positive charge of the tested samples [56]. Certainly, the analysis of the PZC confirmed the effectiveness of the biomass chemical modification by introducing quaternary ammonium groups that contribute to a more positive potential of the sorbent.

Figure 6.

The influence of pH on the surface charge of LVS biomass and LVSAT-CHMAC biosorbent.

FTIR spectroscopic characterization of alkaline modified biomass (LVSAT) and synthesized biosorbent (LVSAT-CHMAC) was aimed at identifying structural changes of the precursor during synthesis and confirming the incorporation of quaternary ammonium groups into the biomass structure [37]. Corresponding segments of FTIR spectra in the areas of valence and deformation vibrations of functional groups with characteristic changes are shown in Figure 7. FTIR spectra of both analyzed LVS samples in the region 1800–500 cm⁻¹ are typical for lignocellulosic biomass. In the FTIR spectrum of LVSAT biomass, the IR band at 1735 cm⁻¹ originates from saturated ester C=O bonds of hemicellulose. The decrease in the intensity of this band indicates that during the alkaline pretreatment of biomass, the hydrolysis of glucosidic C-O-C bonds occurs, i.e. depolymerization of the released hemicellulose and its partial removal from the biomass. In the spectral region 1700–1560 cm⁻¹ there is a complex IR band, with centroid at about 1650 cm⁻¹ and a shoulder at about 1600 cm⁻¹. This broad band indicates the overlap of the O-H vibrations of the crystal hydrates (water molecules in the lignocellulosic structure) with the present phenolic and carboxylic OH groups, as well as the skeletal C=C vibrations of the conjugated rings of lignin. The band at 1508 cm⁻¹ originates from the skeletal vibration of the lignin aromatic ring. The doublet at around 1550 cm⁻¹ indicates different types of aromatic C=C skeletal vibration, i.e. the presence of lignin fragments of different nature in the biomass. The subunits of lignin incorporated into the polymer can be guaiacyl, syringyl and p-hydroxyphenyl, which include methoxyl, acetyl and formyl groups.

Figure 7.

FTIR spectra of LVSAT biomass and LVSAT-CHMAC biosorbent (significant changes are highlighted with asterisks).

This difference in composition has a great influence on delignification, as well as on biomass destruction. It is typical that guaiacyl units are more often cross-linked at the C₅ position of the aromatic ring, so that these cross C-C bonds make the delignification of biomass difficult, given that they cannot be hydrolyzed by either acid or base. In contrast, with syringyl units, this C₅ position is substituted, so it does not participate in further substitution reactions. The weak IR band at 1327 cm⁻¹ represents the vibrations of the condensed syringyl G-ring, i.e. coniferyl alcohol substituted in the C₅ position. The splitting of this band into two peaks (1336 and 1320 cm⁻¹), as well as the appearance of a slightly more intense band at 1260 cm⁻¹, can be explained by the increased number of C-OH bonds due to the presence of ferulic acid in the lignin polymer. The IR band at about 1460 cm⁻¹ originates from the deformation C-H vibrations. IR bands at 1423 cm⁻¹ (O-H bending vibrations) and 1260 cm⁻¹ (vibrations of cellulose ester group) are correlated with the crystalline structure of cellulose. Small changes in the intensity of these IR bands may indicate a partial transition from the crystalline to the amorphous form of cellulose in the treated LVSAT sample, which favors a more efficient synthesis of sorbents based on such a precursor. The broad IR band in the region 1100–900 cm⁻¹ indicates a significant overlap of C-OH and C-C-O vibrations. The IR band at 897 cm⁻¹ and the shoulder at 990 cm⁻¹ originate from the C-H bending vibrations, which are characteristic of the anomeric β-glucosidic (C-O-CH) bonds of cellulose. These bands remain unchanged during the treatment of biomass, which indicates the structural stability of the cellulose chain under the applied reaction conditions. The spectral region below 800 cm⁻¹ corresponds to the deformation vibrations of the C-H and O-H partners [57].

In the FTIR spectrum of LVSAT-CHMAC biosorbent (Figure 7), IR bands at 1030, 1066, 1113 and 1157 cm⁻¹ typical for O-H, C-OH, C-O and C-H vibrations of OH and CH₂ groups of cellulose glucopyranose units were identified [57]. The IR bands at higher wavenumbers (1266, 1328, 1371, 1421, 1475, 1509 and 1594 cm⁻¹) correspond to the vibrations of OH, CH₂, O-CH₃, CO and C=C functional groups typical of lignin [4]. The weakened IR band at 1718 cm⁻¹ originates from the C=O vibration of the remaining hemicellulose (trapped in the biomass structure). The assigned absorption bands were used to detect structural changes during the chemical treatment of LVSAT biomass.

Based on spectral-structure correlation and comparative analysis of FTIR spectra shown in Figure 7, it is evident that chemical modification of biomass with CHMAC reagent induces significant changes in the number, position and intensity of some IR bands. This confirms the decrease in the intensity of the IR bands originating from the vibrations of all types of OH groups (1266, 1030 and 608 cm⁻¹), the hemicellulose C=O group (1718 cm⁻¹), and–CH groups of the glucosidic bonds (911 cm⁻¹). Also, there is a shift in the positions of the IR bands characteristic of cellulose (from 1260 to 1266 cm⁻¹, from 1040 to 1066 cm⁻¹ and from 897 to 911 cm⁻¹) and hemicellulose (from 1738 to 1718 cm⁻¹). These changes clearly indicate a partial disruption of the biomass lignocellulosic structure and the destruction of the hemicellulose glycosidic structure. The decrease in the intensity of the band at 600 cm⁻¹ (from C=C), with a simultaneous shift to 1594 cm⁻¹, as well as the change in the position of the band from 1462 to 1475 cm⁻¹ originating from lignin -OCH₃ groups, confirm the partial disruption of the lignin structure. Additionally, the IR band at 1260 cm₋₁ (from the C-OH group) splits into two new bands (at 1266 and 1227 cm⁻¹), indicating the appearance of OH groups of different origin, probably from the glucopyranose unit of cellulose and the trimethylammonium- hydroxypropyl agent. A similar phenomenon was observed in the region of valence C-O vibrations, where the band at about 1040 cm⁻¹ doubles into two bands (1066 and 1029 cm⁻¹), which confirms the previous statement.

The most important proof of the biosorbent synthesis success is the appearance of a new band of lower intensity (due to overlapping) at 1490 cm⁻¹, which corresponds to asymmetric C-H vibrations of the introduced quaternary –N⁺(CH₃)₃ functional group. The partner of this IR band (from the asymmetric C-N bond) is located at about 980 cm⁻¹, suggesting the presence of new –CH₃ groups. All these observations clearly indicate the incorporation of the cationic –N⁺R₃ group on the LVSAT-CHMAC biosorbent surface during biomass modification [4, 6, 37].

3.4 Biosorbent efficiency and mechanism of sorption process

The efficiency of the synthesized LVSAT-CHMAC biosorbent was tested through the processes of removing phosphate and nitrate anions from aqueous solutions, which appear more and more often in natural and especially in wastewater. After a series of sorption experiments in batch mode, an analysis of the process parameters influence on the biosorbent sorption capacity was performed. The following optimal parameters were determined for the sorption of phosphate ions: sorbent dose (2 g/dm³), sorbent/sorbate contact time (30–40 min), medium pH (6.0 ± 0.2), temperature (20.0 ± 0.5°C), and mixing speed (150 rpm). Based on the conducted tests (Figure 8), the sorption capacity of the LVSAT-CHMAC biosorbent for phosphate ions was determined to be 17.8 mg P/g at optimal conditions.

Figure 8.

Phosphate removal efficiency from aqueous solutions of different initial concentrations using LVSAT-CHMAC biosorbent as a function of sorption time, according to optimal sorption parameters.

In the case of nitrate sorption, the following optimal parameters are defined: sorbent dose (2 g/dm³), sorbent/sorbate contact time (20–30 min), medium pH (6.5 ± 0.2), temperature (20.0 ± 0.5°C), and mixing speed (150 rpm). Based on the conducted tests (Figure 9), the sorption capacity of the LVSAT-CHMAC biosorbent for nitrate ions was determined to be 15.7 mg N/g at optimal conditions.

Figure 9.

Nitrate removal efficiency from aqueous solutions of different initial concentrations using LVSAT-CHMAC biosorbent as a function of sorption time, according to optimal sorption parameters.

In order to examine the qualitative and semi-quantitative elemental composition of both biomass and biosorbent surface, the Energy-dispersive X-ray spectroscopy (EDS) was used. This method combined with SEM microscopy, in addition to elemental composition, also provided data on the elements location on the analyzed biosorbent surface after phosphate and nitrate sorption. The results of SEM-EDS analysis of the samples before and after anion sorption are presented in Figure 10. It should be noted here that the Cr signals originate from the electroconductive thin layer of chromium used during the preparation of samples for analysis. SEM-EDS analysis of the examined samples provided direct evidence of the incorporation of quaternary ammonium groups (NR₄⁺Cl⁻) on the biosorbent surface after the biomass chemical modification, by detecting new N signals. The sorption efficiency of phosphate and nitrate anions on the LVSAT-CHMAC biosorbent surface was confirmed by the detection of new P and K signals (from K⁺H₂PO₄⁻), as well as more intense N signals (from NO₃⁻), respectively. Characteristic changes of the Cl signal in the spectrum of the biosorbent, whose intensity significantly decreases after sorption, indicated a possible ion exchange mechanism of chloride ions with phosphate or nitrate ions in this process. Other studies also point to ion exchange as the dominant sorption mechanism of these anionic species by different agricultural wastes [13].

Figure 10.

EDS spectrum (left) and corresponding SEM micrograph (right) of LVSAT-CHMAC biosorbent before sorption (a), after phosphate sorption (b) and after nitrate sorption (c).

Phosphate and nitrate sorption is best described by a pseudo-first-order nonlinear kinetic model (R² in the range 0.990–0.998). The complex nature of the sorption process, based on both surface sorption and diffusion within the biosorbent particles, was confirmed by the Weber-Morris model. This model indicated the dominant effect of the boundary layer, which has a direct influence in limiting the overall speed of the sorption process [58, 59, 60]. Phosphate sorption equilibrium is subject to laws of Freundlich and Sips isotherms, which indicates the complex nature of the sorption process (physical sorption and ion exchange). The mutual interaction of sorbed anions represents the accompanying mechanism, which leads to the formation of a multimolecular layer of phosphate on the biosorbent surface. On the other hand, the equilibrium of nitrate sorption is subject to the laws of the Sips and Langmir isotherms, which means that nitrate sorption takes place in a monomolecular layer, on an energetically uniform surface of the biosorbent with a finite number of active binding centers, without mutual interaction and trans-migration of ions on the biosorbent surface [58, 59, 60]. The corresponding model of phosphate and nitrate sorption on the biosorbent surface is shown in Figure 11. Thermodynamic studies of the investigated anions sorption indicated a spontaneous exothermic process in the temperature range of 20–40°C. At higher temperatures, the process of anions desorption from the biosorbent surface is favored. This fact and a slight decrease in the disorder of the system at the biosorbent/solution interface, confirm ion exchange as the most likely sorption mechanism [58, 59, 60].

Figure 11.

Structural model of sorption of dominant phosphate (H₂PO₄⁻ and HPO₄²⁻) and nitrate (NO₃⁻) anions from aqueous solutions using LVAT-CHMAC biosorbent.

4. Experimental design and sorption process optimization

The fact is that the efficiency of pollutant removal from contaminated aqueous solutions, as well as the sorption capacity of the biosorbent, depends on the choice and way of varying numerous reaction conditions during the sorption process (pH, temperature, sorbent dose, solution concentration, mixing speed, contact time, etc.). In this sense, in order to determine the key process parameters and their optimization, it is necessary to conduct a very large number of experiments (Figure 12). At the same time, it should be taken into account that in each series of experiments only one parameter varies, while the others are kept constant. It requires a lot of time, patience and persistence, and it also creates additional costs. Despite this, the conventional approach to such research (including kinetic, equilibrium and thermodynamic studies) leaves a lot of experimental gaps important for predicting optimal sorption parameters.

Figure 12.

Example of experiment series on phosphate removal efficiency as a function of two variables (initial solution concentration and sorbent-sorbate contact time).

Therefore, in order to optimize the sorption process and predict optimal factors (experimental conditions), it is desirable to apply the experimental design methodology. This study involves the integration of the OVAT methodology (One-Variable-At-a-Time) with a factorial design (Design of Experiments - DoE). The advantage of the DoE methodology is a significantly smaller number of experiments that should be conducted with the aim of evaluating the sorption capacity of the sorbent, the interaction effects of independent variables on the sorption efficiency, as well as the clarification of the sorption mechanism.

In our study on the phosphate anions sorption using LVAT-CHMAC biosorbent, the critical factors of the sorption process were determined by the OVAT method [61]. In order to optimize and evaluate the factors interaction effects, the Central Composite Design (CCD) was used within the Response Surface Methodology (RSM) analysis. Using the OVAT approach, the four most important input variables were determined (A - initial concentration of the phosphate solution, B - pH value, C – temperature, and D - time), as well as their individual trends that significantly affect the phosphate sorption process. Within the planning of experiment, a CCD matrix with the specified 4 factors on three levels (−1, 0, +1) was randomly generated using adequate statistical software (in our case, JMP-Pro 16 by SAS). The estimated values of critical variables represent the zero coded values for the CCD design. Based on the set data, the plan included 26 experiments, where eight axial points will be used to evaluate the square terms, 16 factor points to evaluate the main effects and two-way interactions, and two central points to evaluate the model adequacy. The outcome of the combined variables influence is the response with predicted values of sorbed phosphate amount.

The RSM-CCD design methodology was useful in forming a complete picture of the phosphate sorption process by revealing significant effects of variables and determining the effects of factor interactions on the response. In order to predict the sorption capacity, a second-order polynomial model was developed and validated (Eq. (1)). This model reveals the influence of the combined process factors (A, B and C) on the response (Y). Initial phosphate concentration (A) was identified as the main process factor with a positive impact (synergism). A negative influence of the temperature (C) was found, while the sorbent-sorbate contact time (D) was not a statistically significant factor (given that the equilibrium of the process is established after 40 minutes).

$$\text{Y}=23.52+5.98\text{A}-2.88\text{C}-0.97AB+2.74{\text{A}}^2-6.13{\text{B}}^2$$E1

Also, the interaction between initial phosphate concentration and pH value (AB) was found to have a significant effect on the predicted response (Figure 13). In addition to identifying and testing significance of the main process factors, the RSM metodology indicated the importance of pH value for the phosphate removal from aqueous solution. Namely, different forms of phosphate anions (H₂PO₄⁻, HPO₄²⁻, and PO₄³⁻) are represented in aqueous solutions. As a critical factor, pH can affect their sorption differently, primarily depending on the LVAT-CHMAC biosorbent pH_PZC value. Thus, although increased anion sorption is expected at pH < pH_PZC (due to the positively charged surface of the biosorbent), it was observed that the sorption efficiency decreases with decreasing pH. The decrease in sorption efficiency (PO₄³⁻ > HPO₄²⁻ > H₂PO₄⁻ > H₃PO₄) can be explained by the pronounced protonation effect of anions to the H₃PO₄ molecules in more acidic solutions. The maximum phosphate sorption efficiency is achieved at pH between 5 and 7, by the ion exchange mechanism of the dominant HPO₄²⁻ and H₂PO₄⁻ forms. In alkaline medium (pH > pH_PZC), the sorption of PO₄³⁻ form decreases due to competition with OH⁻ ions, which indicates the mechanism of electrostatic interaction between the sorbate and the biosorbent surface.

Figure 13.

Profiles of the factors interaction effect on the response Q (sorbed phosphate amount in mg).

In accordance with the applied experiment design methodology, the sorption capacity average value of 17.6 mgP/g at optimal process parameters (C_o = 70 mg/L, t = 40 min, pH = 5.8, and T = 20°C) was predicted (with 95% certainty). The close match between the predicted and experimental value (17.8 mgP/g) of the sorption capacity confirmed the applicability of the developed regression model.

5. Biosorbent valorization

Investigations of the potential interference of coexisting anions in the solution on the sorption capacity of LVAT-CHMAC biosorbents showed that the presence of nitrates and sulfates leads to a decrease in phosphate sorption capacity by about 35%, while chlorides did not show a significant effect. Phosphate sorption interference follows the order: NO₃⁻ > SO₄²⁻ > Cl⁻. Phosphates were more competitive than nitrates in the sorption process of anions from the binary mixture [62].

The possibility of LVAT-CHMAC biosorbent regeneration was monitored through five consecutive cycles of the sorption/desorption process, as a function of the anions initial concentration (5–100 mg/dm³). In the first cycle, desorption degree of anions from 99 to 95% was achieved. This indicates a reversible process, as well as the fact that there is no strong interaction between the biosorbent and the sorbed anions. The biosorbent effectiveness was also confirmed in the fifth cycle, when a desorption rate of about 50% was achieved.

Our study indicated that the sorption capacities of LVAT-CHMAC biosorbent for phosphates (0.58 mmol P/g) and nitrates (1.18 mmol N/g) are comparable to the results of most other cationic sorbents tested for anions removal under similar conditions [21]. Thus, the phosphate removal ability by modified LVS biomass was higher or very similar to walnut shell, almond shell, peanut shell, rice husk, corn cob, oak chip, palm kernel, wheat straw (0.45–0.71 mmol/g), even in the case of some commercial exchange resins (such as quaternary ammonium cellulose QA52). In the case of nitrates, a higher sorption capacity of LVS biosorbent was determined in relation to cationically modified biomasses such as coconut husk, wheat or rice straw, coconut fibers, sugar cane and various sawdust, even compared to some mineral clays such as halloysite and bentonite.

From the techno-economic aspect, the production cost analysis of any adsorbent is an important factor in the profitability of its practical application in the water purification process. Therefore, the production total cost of the cationic LVS biosorbent was estimated. Production costs are influenced by numerous factors such as: availability of raw material (LVS biomass), processing and treatment requirements (synthesis), as well as reuse of biosorbent. The total cost includes the price of each individual technological operation (rinsing, drying, grinding, sieving, activation, neutralization, synthesis, heating, and mixing). Here, it is taken into account that the bottle gourd shell was collected as agricultural waste (free of cost) and dried naturally (which was not included in the processing cost). Based on the reagents costs and the price of electricity, the approximate total price for the biosorbent production is estimated at 11.5 EUR/kg. This sum is comparable to the production costs of other sorbents based on similarly modified lignocellulosic biomasses (10–25 USD/kg). Therefore, LVAT-CHMAC biosorbent can be considered an economical resource in water treatment technology [60].

In addition to the established properties (economical production, efficient sorption of anionic species, possibility of regeneration), a significant advantage of using LVAT-CHMAC biosorbent is ecological utilization at the end of the life cycle. As a biodegradable material, it is not harmful to the environment. On the other hand, the multiple sorptions-desorption process makes it possible to obtain a concentrated phosphate-nitrate solution or to accumulate these anions in the biosorbent. This solution, as well as the used biosorbent enriched with nutrient elements available for plants (P, K, N, and minerals) can be used as compost for feeding plants or making artificial fertilizers for acidic soils in dry climate areas. In this way, the ecological problem of disposing of the used sorbent or the removed sorbate is also solved.

6. Conclusion

The study indicates that sorbents with significantly improved sorption characteristics can be obtained by simple and accessible procedures of chemical-thermal modification of bottle gourd shell as a cheap and available agro-waste material. The LVS sorbent in form of activated carbon has a high potential for removing various cationic pollutants from aqueous solutions, primarily toxic metals, as well as organic non-polar and weakly polar compounds, in a wide range of sorption parameters (initial concentration of pollutants, pH value and temperature).

Conducted research validates the superiority of LVS biomass as a precursor for cationic modification with quaternary ammonium reagents compared to other agricultural by-products. The sorption efficiency of the cationic LVS biosorbent is comparable to more expensive commercial anion exchange resins tested for the removal of anionic species from wastewater, mainly phosphate and nitrate as the cause of eutrophication. In accordance with the technological, economic and ecological requirements of contaminated water purification, the following process parameters are recommended for effective anions removal using LVS biosorbent: temperature 18–22°C, slightly acidic solution (pH 5–7), initial concentration of anion solution 30–50 mg/dm³ (removal rate 90–60%, respectively), sorbent dose 2 g/dm³ for low level polluted or 5 g/dm³ for more contaminated solutions, contact time 30–40 min.

High sorption capacities, rapid achievement of sorption equilibrium, relatively simple and low cost production, as well as reuse (by multiple sorption/desorption process) indicated that LVS sorbents can be competitive on the market of materials applicable in water purification technology. In order to solve the problem of waste disposal, although it is biodegradable and harmless to the environment, the biosorbent and preconcentrated solution after desorption can be utilize to fertilizer or compost produce.

Acknowledgments

This research was supported by the Ministry of Science, Tehnological Development and Innovation of the Republic of Serbia within the Program of scientific work (No. 451-03-65/2024-03/ 200133; 451-03-65/2024-03/200124; 451-03-66/2024-03/200124).

Conflict of interest

The authors declare no conflict of interest.