Removal of water pollutants using plant-based nanoscale zero-valent iron: A review

In the last few decades, the use of nanotechnology for the treatment of different waste streams has increased, and nanoparticles have been widely investigated for the treatment of a variety of organic and inorganic pollutants in water/wastewater (Jamkhande et al. 2019). Metallic nanoparticles, because of their small size ranging between 1 and 100 nm and high reactivity, have shown high efficiency in removing different contaminants (Priya et al. 2021). Various metals such as gold, silver, palladium, platinum, iron, nickel, and copper have been used to synthesize nanomaterials for pollutant removal (Lu & Astruc 2020).

Among the different nanoparticles, nanoscale zerovalent iron (nZVI) is one of the most commonly investigated materials. nZVI is a magnetic material and has gained popularity due to its size, low energy requirement for synthesis, low toxicity levels, high pollutant removal efficiency, easy accessibility, and minimal secondary pollution (Zhou et al. 2022). It is highly regarded due to its high specific surface area which increases the number of active sites, even if a small quantity of nanoparticles is used for adsorption in the process of pollutant removal (Li et al. 2017a). The reducing power of nZVI in solution is strong (standard reduction potential = −0.44 V), empowering it to reduce several different kinds of contaminants such as dyes, pharmaceutical compounds, pesticides, metals, and nutrients from wastewater (Eljamal et al. 2020).

Nanoparticles can be synthesized using two different approaches: top-down and bottom-up. Top-down methods for nZVI production involve converting bulk particles of iron to smaller nanosized particles by milling processes, pulsed laser ablation, or noble gas sputtering, while the bottom-up approach involves aggregating smaller particles, molecule by molecule, and thus, chemically synthesizing nZVI from either dissolved iron salts, nanosized iron oxides, or iron-containing molecules, by methods such as vapor deposition method, hydrothermal method, sol-gel method, spray pyrolysis, and laser pyrolysis (Saravanan et al. 2021). Both these approaches have various disadvantages. The use of the top-down method generates the risk of combustion hazards and requires heavy equipment which makes the process expensive and highly energy consuming. The bottom-up approach generates toxic by-products and is expensive because of the use of high amount of chemicals as reducing agents. It requires an inert atmosphere and, hence, is done on a small scale (Pasinszki & Krebsz 2020).

At the nanoscale, it is easy for Fe⁰ particles to agglomerate and get oxidized to Fe⁺² or Fe⁺³ when exposed to air or water, thereby weakening its reducing capacity due to an increase in size (Tang et al. 2021). To minimize aggregation, increase stability and mobility, and reduce leaching, a wide range of inorganic and organic support and capping materials are used for nZVI (Adusei-Gyamfi & Acha 2016). These common modifications are conducted via chemical methods and involve the release of toxic by-products, the use of an expensive reducing reagent like sodium borohydride (NaBH₄), and high energy consumption (Yang et al. 2021).

Conventional methods used for the removal of water contaminants, in turn, cause other environmental problems due to the use of excessive energy and toxic chemicals (Venkatas et al. 2022). A sustainable solution to overcome these limitations is the green synthesis of nZVI which is economic and nontoxic, involves minimum by-product formation, and most of the time, does not require a separate capping agent, thereby reducing energy consumption and the use of toxic reducing agents (Bharti et al. 2021). Green materials contain polyphenols and proteins that can replace chemical reagents as reducing agents to reduce multivalent iron ions into their zerovalent states and also can act as capping agents, preventing agglomeration of particles (Mahajan et al. 2021). Green synthesis of nZVI can be done using microorganisms and plant extracts from flower, leaf, fruit, and stem (Ying et al. 2022).

Due to the limitations of maintaining the media culture and specific conditions required in the use of microorganisms, arguably the most interesting and most environmentally friendly form of green synthesis is by utilizing plant or food scraps, to create nZVI as plant extracts having various ketones, aldehydes, polyphenols, flavonoids, caffeine, and carbohydrates, making them excellent reducing agents (Huston et al. 2021). However, each plant used as a raw material has unique phytochemicals, and so, the nanoparticles produced have different physicochemical properties, which is an advantage of plant-mediated synthesis (Ebrahiminezhad et al. 2018).

Several papers recently reviewed the application of nZVI for the removal of different contaminants from water (Kalra & Gupta 2019; Li et al. 2021; Tarekegn et al. 2021). However, no review is available on the use of nZVI synthesized from plant-based materials. This paper first summarizes the synthesis of plant-based nZVI and their characteristics. Removal of different classes of pollutants such as dyes, heavy metals, nutrients, and trace organic pollutants by these nZVI are then discussed. The advantages and drawbacks of this technology are highlighted along with future research needs.

Plants are rich in bioactive components which act as reducing agents for the synthesis of iron nanoparticles by reducing the precursor iron compounds (Manquián-Cerda et al. 2017). Various plant parts such as leaf, root, stem, seed, flower, and fruit can be used to prepare the plant extract for synthesis (Poguberović et al. 2016). The bioactive components present in these plant parts include polyphenols, cyclic peptides, enzymes, proteins, amino acids, and other substances that do not let the particles agglomerate, thereby also acting as capping agents while working as reducing agents (Sharma et al. 2019). These phytochemicals present in plants are responsible for reducing Fe⁺³/Fe⁺² to Fe⁰. Since different plants have different compositions of biomolecules present in them, the nanoparticles produced also show unique morphological characteristics making them suitable for various potential uses (Pattanayak et al. 2021).

The reduction capacity of the plant extract is mainly dependent on the number of phenolic groups (polyphenol content) present. The major class of polyphenols is flavonoids that consist of various functional groups, encouraging the formation of nanoparticles (Makarov et al. 2014). Based on the combination of iron precursor and plant extract pair used, reactivity is determined and nanoparticles are formed within a few minutes to a day, after plant extract is mixed in the iron precursor solution. The phenolic group of biomolecules gets complexed with iron and the biomolecules assume their oxidized ascorbic form (Kozma et al. 2016). The donation of electrons to free radicals from the hydroxyl group due to the conjugated π electron system of polyphenols and flavonoids enables high antioxidant activity, and so, it will prolong the life of produced nZVI (Speisky et al. 2022). The minimal use of chemicals, ambient room temperature and pressure, as well as easy handling and requirement of no heavy equipment for synthesis make the process economically viable for large-scale production (Nasrollahzadeh et al. 2021).

The mechanism of nZVI formation includes three phases namely activation phase, growth phase, and termination phase (Dikshit et al. 2021). In the activation phase, the metal precursor is ionized and the free positive metal ions combine with the negative biomolecule functional groups present in the plant extract. Nanoparticles are formed due to chelation and involve the reduction of multivalent iron precursor to its zerovalent form, followed by the nucleation of reduced metal atoms (Singh et al. 2021). The properties of nanomaterial formed depend on pH, temperature, incubation period, bioactive components in the plant and their concentration, and most importantly, the number of phenolic groups present (Shafey 2020).

The activation phase is followed by the growth phase in which the reduced and nucleated Fe ions come together to form metal nanoparticles. In this stage, the agglomeration of metal atoms takes place and the particles gain their morphological shape, while other metal atoms are still getting reduced (Khandel et al. 2018). Due to extended nucleation, agglomeration may lead to large particle formation, so thermodynamic energy needs to be minimized for particle stabilization to take place (Zikalala et al. 2018).

The last stage is the termination phase where the nanoparticles stabilize and the desired morphology of the nanoparticle is maintained owing to the formation of a layer of capping agent around the nanoparticles by improved thermodynamic stability. This capping agent is also formed from the bioactive components present in the plant and it does not allow the particles to agglomerate any further, by reducing the surface energy (Stefaniuk et al. 2016).

Particle size, morphology, size distribution, functional groups present, composition, agglomeration, reactivity, and structure are important parameters to be evaluated before deploying nZVI in any pollution control application. These characteristics are obtained with the help of microscopic and spectroscopic techniques such as transmission electron microscopy (TEM), energy-dispersive X-ray spectroscopy (EDX), X-ray photoelectron spectrum (XPS), Fourier transform infrared spectroscopy (FTIR), dynamic light scattering (DLS), and scanning electron microscopy (SEM) (Zhou & Li 2022). A summary of the different characteristics of plant-based nZVI reported in the literature is presented in Table 1.

The images and data from TEM and SEM give information about the morphology and surface details of synthesized nZVI (Jain et al. 2021). The agglomeration properties of the nanoscale particles can also be interpreted from the images (Shi et al. 2015). It is observed that the size of plant-based nZVI is generally larger than the size of nZVI synthesized from chemical reducing agents, because the plant extract components, in addition to acting as reducing agents, also act as capping agents, forming an outer coating of biomolecule components in nZVI synthesized using plant extracts. The leaf extract of Mentha piperita was used to synthesize green nZVI (Shad et al. 2020). The nanoparticles produced were of the size 5–10 nm with a spherical shape and showed some agglomeration. Ebrahiminezhad et al. (2017) used Urtica dioica leaves to synthesize nZVI. The TEM images showed diameters in the range of 21–71 nm with a mean diameter of 46 nm. The shape was mostly spherical, while certain oval particles were also observed. The micrographs also revealed the presence of a biological matrix around the particles because of the leaf extract. Karam et al. (2020) compared the characteristics of chemically synthesized nZVI and green nZVI synthesized from Camelia sinensis (black tea) extracts. The images of both chemically and green synthesized nZVI showed regular and irregular surfaces, and the average diameter was 40 and 80 nm, respectively. The verification of size was done using EDX, which showed a diameter of 36 nm for chemically synthesized nZVI, depicting the major content of iron in it, while the diameter of green nZVI was obtained as 83.7 nm, depicting the coating of leaf extract because of the presence of carbon and oxygen layers surrounding the particle. However, some researchers reported a smaller size for plant-based nZVI. The average diameter of chemically synthesized nZVI and nZVI synthesized from Vaccinium corymbosum (blueberry) extract were found to be 80.2 and 52.4 nm, respectively, by Manquián-Cerda et al. (2017). The conditions impacting the size and shape of synthesized nZVI include composition and concentration of extract, concentration of reagent, synthesis temperature, and reaction time (Machado et al. 2014).

EDX micrograph showed the chemical composition of nZVI and the surface atomic distribution (Yusmartini et al. 2015). Plant-based nZVI is found to have more concentrations of carbon and oxygen and a lesser concentration of iron, as compared to chemically synthesized nZVI, owing to the presence of organic substances in the biomolecules of plant extract. nZVI synthesized from green tea was characterized by EDX to determine its atomic composition (Shahwan et al. 2011). The spectrum peaks confirmed the presence of C (47.9%), Fe (27.8%), Na (12.5%), O (6.3%), and Cl (5.5%). The presence of Na, O, and Cl might be because of the use of NaCl and FeCl₂ during synthesis, while C might be because of polyphenol and some other biomolecule groups in the tea extract. Vilando et al. (2019) showed the use of Euphorbia hirta linn (Tawa-tawa) extracts for nZVI synthesis. Two types of precursors, FeSO₄ and FeCl₃, were used. The composition interpreted using EDX peaks showed the presence of Fe (43.77%) and O (56.07%) when FeSO₄ was the precursor, and Fe (65.97%), O (22.58%), and Cl (11.45%) were present during the synthesis with FeCl₃ as the precursor. The presence of Fe was attributed to the precursor used, O to phenolic compounds in plants, and Cl was found only when FeCl₃ was used as the precursor.

X-ray powder diffraction (XRD) is used to understand the crystalline or amorphous structure of iron nanoparticles based on the sharpness of the peaks obtained and the type of iron nanoparticles (iron oxide, iron hydroxide, zerovalent iron) based on the peaks obtained at 2θ value (Huang et al. 2015). Jeyasundari et al. (2017) used Psidium guajava to prepare nZVI. Using XRD, the structure was concluded to be hexagonal, based on the peaks obtained, and the size of the nanoparticles calculated, based on the width of the peak came out to be 27 nm. Machado et al. (2015) showed the use of various leaf extracts for nZVI synthesis. It was interpreted that nZVI produced was amorphous in nature and the heterogeneous constitution of leaves was considered to be a probable reason for it.

FTIR analysis is performed to determine the functional groups present in synthesized nZVI and those are responsible for the reduction of iron precursor. Plant-based nZVI has more carbon and oxygen functional groups owing to the biomolecules that are absent in chemically synthesized nZVI. Jha & Chakraborty (2020) studied the FTIR analysis of nZVI synthesized using Shorea robusta leaf extract. The functional groups found were alcohol, aldehyde, ketone, polyphenol, and alkyl groups. Alcohol and polyphenols were responsible for reducing FeSO₄ and acting as capping agents. The peaks obtained at 1,045 and 1,230 cm⁻¹ were attributed to C–O stretching from primary alcohol, while the peak at 1,654 cm⁻¹ signified the presence of C = O stretching in aldehyde and ketones. Peaks at 2,854 and 2,924 cm⁻¹ showed C–H stretching of alkyl groups while peak at 3,286 cm⁻¹ was owed to O–H in polyphenols. Lin et al. (2021) showed a comparative analysis of chemically synthesized nZVI and green nZVI synthesized from Eucalyptus leaves. The FTIR studies showed that chemically synthesized nZVI had very weak peaks as compared to green nZVI which had definitive peaks at 3,388 cm⁻¹ showing the presence of O–H vibration, 1,636 cm⁻¹ signifying C = O vibration, and 1,039 cm⁻¹ stretching to the vibration of C–O–C.

Several authors used UV–visible spectral analysis to verify the reduction of Fe⁺³/Fe⁺² to Fe⁰ with excitation at a particular wavelength of surface plasmon resonance (SPR) in nZVI (Sunardi et al. 2017). However, conclusive evidence on the formation of Fe(0) from Fe⁺³/Fe⁺² is not always obtained. nZVI synthesized from Syzygium aromaticum extract was shown to have its absorption peak in the range of 216–265 nm, and the synthesis was complete as shown by high band intensity with no other apparent evidence (Pattanayak et al. 2013). The reduction of Fe⁺² and formation of nZVI synthesized from tea extracts was observed in the intensity of peaks at 205 and 272 nm by the excitation of surface plasmon vibrations in Fe nanoparticles, and the reduction in the intensity of peaks corresponding to tea polyphenols after the reaction with Fe⁺² was the only proof of formation of nZVI (Huang et al. 2014). In another study, Sravanthi et al. (2018) used Calotropis gigantea extract for the synthesis of nZVI, and the UV–vis absorption peaks of flower extract and nZVI synthesized from the extract were compared to understand nZVI formation. A wide range of 200–800 nm was selected for the analysis, and the peaks obtained with flower extract analysis at 300–350 nm disappeared during nZVI analysis, and broader absorption emerged at higher wavelengths which was the only evidence of nZVI formation. The characteristic UV–vis peaks were not observed for nZVI due to high iron reactivity, thereby giving no further evidence of nZVI formation (Sravanthi et al. 2018).

Brunauer–Emmett–Teller isotherm (BET) analysis is used to estimate the specific surface area (SSA) of synthesized nZVI. The SSA of plant-based nZVI is generally less, as compared to chemically synthesized nZVI. This might be due to the increased size of plant-based nZVI because of the biomolecule coating around it. Francy et al. (2020) showed the SSA of nZVI synthesized from neem and mint leaves to be 6.2 and 13 m²/g, respectively, which was less than chemically synthesized nZVI with a SSA of 15.2 m²/g. Raman et al. (2021) found the SSA of chemically synthesized nZVI and nZVI synthesized from grape leaf to be 31.04 and 15.08 m²/g, respectively, depicting the decreased surface area of plant-based nZVI. Nasiri et al. (2019), however, found a higher SSA for nZVI synthesized from Ferula persica, as compared to chemically synthesized nZVI. The BET surface area of chemically synthesized nZVI and nZVI from plant extract were 34.38 and 46.68 m²/g, respectively. An increased surface area generally results in better adsorption efficiency (Huang et al. 2014). A comparison of the properties of chemically synthesized nZVI and plant-based nZVI is presented in Table 2.

Plant-based nZVI has been used for the removal of a variety of contaminants such as dyes, heavy metals, nutrients, and trace organics from different waste streams. This section describes these studies, highlighting the mechanisms of contaminant removal by nZVI.

Dyes are majorly classified as natural and synthetic dyes, based on their source material. Synthetic dyes can be divided into ionic and non-ionic categories. Ionic dyes are further classified into cationic dyes and anionic dyes which include acid, reactive, and direct dyes, while non-ionic dyes are classified into vat and dispersive dyes (Kadhom et al. 2020). According to their chemical structures, dyes are categorized as azo, nitro, indigoid, anthraquinone, phthalein, and triphenyl methyl dyes, based on the attached chromatograph group (Benkhaya et al. 2017). Reactive, direct, vat, dispersive, acid, basic, and solvent are all subgroups of dyes based on the method of application (Pandey et al. 2020). It is reported that discoloration and removal of dyes can be achieved by nZVI as it is capable of destroying their chromatograph group (Pourabadeh et al. 2020). Plant-based nZVI shows high removal efficiency of cationic and anionic dyes consisting of attached chromatograph groups at optimum conditions. Methylene blue (MB), methyl orange (MO), and crystal violet (CV) are the most commonly studied dyes for removal by plant-based nZVI. A summary of these studies is presented in Table 3.

The adsorption capacity of plant-based nZVI is promising and is also generally higher than the conventional nZVI which is chemically synthesized. Abdelfatah et al. (2021b) compared the adsorption capacity of nZVI synthesized from chemical reducing agents and various plant extracts for MB removal. The adsorption capacity of chemically synthesized nZVI was 5.53 mg/g, which was much lower than those of nZVI synthesized from Riccinus communis and Platanus occidentalis with an adsorption capacity of 64.94 and 126.60 mg/g, respectively.

Various factors such as pH, contact time, adsorbent and adsorbate dose, as well as temperature, affect the adsorption capacity and removal efficiency. Hence, determining optimum conditions is important for the application of this process.

The pH at which optimum adsorption takes place depends on the properties of the dye to be adsorbed. It is known that at higher pH, the adsorbent surface is negatively charged, helping better removal of cationic dyes, and at lower pH, anionic dyes are better removed owing to the positive surface charge of the adsorbent (Bao & Zhang 2012). Shaibu et al. (2014) used bamboo, which is a waste agricultural product, to synthesize plant-based nZVI for the removal of MB, which is a cationic dye. The removal was less at acidic pH values. The highest removal of 79.6% was observed at a higher pH value of 11. The higher removal was attributed to the effect of alkaline pH on the charge of reactive groups in adsorbents, which is responsible for the ionization of dyes. Electrostatic attraction of opposite charges on the surface was responsible for removal.

Khashij et al. (2020) used Thymus vulgaris plant-based nZVI for the degradation of an anionic dye Reactive Black 5. Dye removal was the highest at pH 4 and the reason could be attributed to electrostatic attraction. The efficiency decreased by 15% when the pH was increased to 10. At a lower pH, the formation of H⁺ ions led to its reaction with nitrogen atoms of azo dye, which is favorable for Reactive Black 5 degradation by the reduction of azo bond (Quan et al. 2018).

The contact time for which adsorbent and adsorbate interact is an important parameter in the process design. Initially, the number of vacant sites available for adsorption on the surface of an adsorbent is more and, hence, the uptake is faster. The removal depends on the adsorbent's potential. Since with time the number of active sites saturates, the rate at which sorption takes place decreases. Thus, as time passes, adsorption still takes place until equilibrium, but the rate of removal of dye from wastewater decreases. This mechanism of higher initial uptake and subsequent decrease, until equilibrium is achieved, was shown by Parlayıcı & Pehlivan (2019) who used nZVI synthesized from Platanus occidentalis tree seed pod fibers for the removal of MB, malachite green oxalate (MG) and methyl violet 2B (MV) dyes. The quickest uptake took place in the first 30 min, followed by 30 min of slow uptake, and the equilibrium was reached at 70 min.

The initial concentration of dye affects its removal. Increased dye concentration results in a longer time for removal and reduction in removal efficiency. With the increment of initial dye concentration, the adsorption capacity increases until saturation due to the enhanced driving forces that surpass mass transfer resistance. Abd El-Monaem et al. (2022) showed a higher adsorption capacity of MB using lemon-derived biochar-based nZVI, when the initial concentration of MB was increased. The adsorption capacity increased from 495.65 to 1,864.41 mg/g when the initial MB concentration was increased from 250 to 1,000 mg/L. Sahu et al. (2019) used sweet-lime pulp-derived nanoparticles to remove MB. Increasing MB concentration from 10 to 40 mg/L increased the adsorption capacity from 8.2 to 32.6 mg/g.

An increase in nZVI concentration improves the dye removal efficiency with the availability of more adsorption sites and surface area, and hence, the time required decreases until saturation. After saturation, an increase in dosage does not influence the removal rate. Sahu et al. (2019) reported the removal of MB using sweet-lime pulp-based nZVI and showed that the removal percentage increased from 71 to 86% with the increase in adsorbent dosage from 0.4 to 1 g/L. A further increase in nZVI concentration to 1.2 g/L did not affect the removal efficiency.

Temperature is another factor that influences the adsorption process. The increase in temperature during the adsorption process increases the diffusion rate of molecules. Abd El-Monaem et al. (2022) used lemon-derived biochar in combination with nZVI to remove MB. The temperature was varied from 25 to 55 °C and the results clearly showed an increase in adsorption capacity and removal percentage from 957.71 mg/g and 95.58% to 997.51 mg/g and 99.74%, respectively, as the temperature increased from 25 to 55 °C. The increase in temperature leads to the increased movement of MB molecules toward the nanocomposite.

The interaction between adsorbate and adsorbent at equilibrium is determined by adsorption isotherms. The reported studies show that Langmuir and Freundlich isotherms are, generally, well-fitted for various dyes and plant-based nZVI combinations, while in certain cases, Temkin isotherms are used to model adsorption using batch tests. Abdelfatah et al. (2021b) reported the removal of MB dye using nZVI for Ricinus communis and showed that adsorption was well-fitted into Freundlich isotherm (R² = 0.990), indicating that the dye molecules adsorbed were following multilayer adsorption. The n value derived from the Freundlich isotherm equation was 1.838 indicating moderate adsorption capacity. The maximum adsorption capacity calculated from Langmuir isotherm was 61.73 mg/g. The linear plot for the Temkin adsorption isotherm suggested physical adsorption. Ahmad et al. (2020) showed the removal of MB dye using date palm waste. The Redlich–Peterson isotherm was the best-fitted isotherm followed by the Langmuir isotherm, indicating both heterogeneous and homogeneous adsorption. The Langmuir model predicted the maximum sorption capacity as 1,182.9 mg/g. The hydrogen bond that existed between –OH groups of sorbent and –N– groups of MB contributed to the removal. The major mechanisms for MB removal were electrostatic interactions between cationic MB and anionic surface functional group, high surface area through π-π electron donor–acceptor interactions between aromatic sorbents and cationic dye, as well as diffusion. The binding energy was very low, indicating that adsorption did not follow the ion-exchange process.

Kinetic studies have shown that the adsorption of various dyes by plant-based nZVI is, generally, well-fitted by a pseudo-second-order kinetic model. However, dye removal kinetics following pseudo-first-order and Elovich models have also been reported. Ahmad et al. (2020) synthesized nZVI from date palm waste and used it for the removal of MB. They found that adsorption followed pseudo-second-order (R² = 0.93–0.99) and Elovich model (R² = 0.86–0.97). The higher value of rate constants indicated chemisorption. Abdelfatah et al. (2021b) showed that the removal of MB, a cationic thiazine dye, using nZVI prepared from Ricinus communis seed extract followed the pseudo-first-order kinetic model (R² > 0.97). The values of the kinetic constants found in the pseudo-first-order equation decreased with the increase in MB concentration indicating a physical adsorption reaction taking place. The Elovich model describes effectively the solid–liquid interaction. The value of the Elovich constants for the initial adsorption rate was higher than the desorption rate constant, indicating low desorption and a higher adsorption tendency of MB.

Several studies compared the performance of plant-based nZVI with chemically synthesized nZVI. Studies have shown plant-based nZVI as a better material for dye degradation than chemically synthesized nZVI. It is also seen that the efficiency of plant-based nZVI increases when used with some added chemicals. Nasiri et al. (2019) compared the efficiencies of chemically synthesized nZVI, green synthesized nZVI from Ferula persica extracts (G-nZVI), and green synthesized nZVI functionalized with β-cyclodextrin (β-CD) for the removal of CV dye. Chemically synthesized nZVI had a larger size as compared to the other two, which had smaller average size, higher surface area, and pore volume. It was also inferred that the removal of dye by both green synthesized nZVI particles was much more pH-dependent than chemically synthesized nZVI. The removal efficiency of 500 mg/L CV solution by chemically synthesized, green synthesized, green synthesized and functionalized with β-cyclodextrin nZVI were 7.6, 35.6, and 68.5%, respectively, at pH 9. This result validates that the use of plant root extract, instead of NaBH₄, results in higher dye removal because of a smaller particle size and a higher surface area. The higher removal efficiency by βCD functionalized G-nZVI compared to G-nZVI was due to the fact that βCD coating provided more adsorption sites owing to very little aggregation taking place and the presence of more than one hydroxyl group. CV inclusion via βCD hydrophobic cavity instead of just electrostatic interaction among hydroxyl groups on the surface in G-nZVI also resulted in higher dye removal.

Hydrogen peroxide acts as a Fenton reagent and increases the dye removal efficiency of plant-based nZVI when added along with it. Lim et al. (2018) compared nZVI synthesized using Mulberry leaves and nZVI synthesized using Mulberry leaves with added H₂O₂ for the removal of MB and MO. For a time period of 180 min and a dye dose of 5 mg/L, the removal efficiencies for MB and MO by nZVI synthesized from mulberry leaves were 39.94 and 57.32%, respectively. The removal efficiency increased to 98.77 and 98.94%, respectively, when plant-based nZVI was used along with H₂O_2, nZVI without H₂O₂ removed MO more than MB, owing to the electrostatic attraction. Hydrogen peroxide acted as a heterogeneous Fenton catalyst, producing OH radicals and accelerating the reaction rate. This shows that plant-based nZVI along with hydrogen peroxide can result in a much higher dye removal.

Industries such as pesticides, paints, mining, electroplating, ceramics, electronic products, and metal alloys use heavy metals in huge quantities and the effluent generated from these industries is considered as the main source of heavy metal generation (Li et al. 2017b). Because of industrial discharges, elevated levels of heavy metals are found in most groundwater and surface water (Angaru et al. 2021). Metals such as iron, zinc, chromium, lead, and copper are essential to a certain limit in the human body for biochemical and physiological functioning, but in excess quantities, they produce various adverse health impacts (Latif et al. 2020). These heavy metals impact the environment because of their toxicity and non-biodegradability. For example, Pb(II) affects the kidneys and reproductive system in humans (Ali et al. 2019). Cr(VI) is highly soluble, carcinogenic, and teratogenic (Samuel et al. 2021). Considering the threat of such heavy metals in the ecosystem, their removal is inevitable. Plant-based nZVI has shown the capability of removing various heavy metals such as lead, mercury, arsenic, chromium, copper, nickel, and cadmium. This section discusses these studies and a summary of the reported studies is presented in Table 4.

High adsorption capacity has been observed for plant-based nZVI for different heavy metals. Mulberry-based nZVI was found to have an As(III) adsorption capacity of 1,329 mg/g and cherry leaves-based nZVI had an adsorption capacity of 904.1 mg/g for Cr(VI) (Poguberović et al. 2016). The adsorption capacity of plant-based nZVI for metals has been observed to be higher than that of chemically synthesized nZVI. Du et al. (2023) showed that nZVI synthesized from green tea extracts had an adsorption capacity seven times higher for Pb(II) compared to chemically synthesized nZVI. They also found better removal efficiency when Fe₂(SO₄)₃ was used as the precursor iron in plant-based nZVI, compared to FeSO₄. The adsorption capacity was 377.3 mg/g when Fe₂(SO₄)₃ was used, which was higher than 327.6 mg/g for nZVI synthesized using plant extract and FeSO₄, owing to easier agglomeration of FeSO₄ particles. Arshadi et al. (2017) reported that Azolla filiculoides-based nZVI had an adsorption capacity of 459.3 mg/g for Hg(II) and 462.7 mg/g for Pb(II).

Studies have been conducted on the effect of various factors such as pH, contact time, temperature, adsorbent and adsorbate dosages that affect the adsorption process of heavy metals by plant-based nZVI. The optimum pH is dependent on parameters like functional groups present in nZVI, which in turn, is dependent on the type of plant selected, charge on the pollutant to be removed, and electrostatic interaction.

Rana et al. (2018) studied the oxidation of As(III) using nZVI synthesized from guava leaves at different pH values. At pH 3, in the first 2 min itself, 70% of As(III) was oxidized to As(V) owing to a Fenton-like reaction at acidic pH and prevention of the formation of coating of an oxide layer, which could otherwise hinder adsorption. At neutral pH, 30% oxidation took place in 2 min and 74% in 10 min. After the first 2 min, no significant oxidation was observed owing to the competition of As(III) and As(V) for the available active binding sites. At pH 9, the removal through oxidation decreased drastically to 25% as ferrous and ferric hydroxides were formed on the surface, thereby decreasing the active sites for adsorption. They also suggested that at high pH, polyphenols in the plant extract get ionized, thereby causing the capping agents to be unavailable. Due to this, ZVI reacts with atmospheric oxygen forming hydroxides. Hence, a lower pH gives better efficiency.

Higher removal of heavy metals at higher pH has also been reported. Arshadi et al. (2017) found a higher removal of heavy metals such as Pb(II) and Hg(II) in the pH range of 5–9. At a higher pH, the surface of the adsorbent became negatively charged causing precipitation of Pb(OH)₂ and protonation/deprotonation of hydroxyl species on the nZVI surface. Although uptake was high at alkaline pH, the formation of iron hydroxides hindered the uptake.

The removal of pollutants increases with the increase in contact time, until an equilibrium is achieved. After that point, the uptake is negligible due to the filling of all vacant active sites on the adsorbent. As the adsorbent dose increases, the uptake increases, owing to the availability of more active sites for adsorption. However, increasing it above a certain limit could cause agglomeration of nZVI particles, thereby decreasing efficiency. At an increased pollutant concentration, more time for removal is required, and this would also cause less removal, as the pollutant would occupy the vacant sites on the adsorbent faster (Yan et al. 2012).

The adsorption isotherm that better fits the adsorption by plant-based nZVI was mostly Langmuir or Freundlich. Poguberović et al. (2016) showed the use of oak, mulberry, and cherry leaves nZVI for the removal of As(III) and Cr(VI). For isotherm studies, the coefficients indicated that the data better fitted Freundlich isotherm for the removal of both As(III) and Cr(VI), indicating multilayer adsorption. Weng et al. (2016) used Eucalyptus leaf-based nZVI for the removal of Cr(VI) and Cu(II). The adsorption isotherm coefficients calculated from experimental data showed that the data better fits the Freundlich model. Since the value of coefficient n exceeded 1, the adsorption was regarded as chemisorption, showing a high affinity between the adsorbate and the adsorbent. The highest adsorption capacity obtained at 308 K for Cr(VI) and Cu(II) were found to be 149.3 and 163.9 mg/g, respectively. They also showed that Eucalyptus-based nZVI had the potential to remove Zn(II) and Pb(II), along with Cr(VI) and Cu(II) from electroplating wastewater. The use of Eucalyptus-based iron nanoparticles for the removal of various heavy metals has been reported by various studies. Kamath et al. (2020) used Eucalyptus leaves and other plants' leaves such as black tea, oak, green tea, and pomegranate leaves for the removal of arsenic from water. Eucalyptus leaves showed the fastest adsorption with an adsorption capacity of 39.84 mg/g. The Langmuir isotherm fitted into the data well, indicating monolayer adsorption.

It has been known from various studies that the adsorption of heavy metals by plant-based nZVI generally follows pseudo-second-order kinetics, indicating the rate-limiting step to be chemisorption. However, certain exceptions are also observed. Poguberović et al. (2016) showed the removal of As(III) and Cr(VI) using oak, mulberry, and cherry leaf-based nZVI. The pseudo-second-order kinetic model was proved to be a better fit and the mechanisms involved were adsorption, reduction, metal-nZVI interactions like oxidation/re-oxidation and co-precipitation of As(III). The use of Eucalyptus leaf for the synthesis of iron nanoparticles for the removal of mixed contaminants was shown by Weng et al. (2016). Since the removal mechanism was not only attributed to adsorption but also reduction, the kinetics for both adsorption and reduction were studied separately. The study showed Cr(VI) and Cu(II) removal efficiency of 74.2 and 45.2% in separate solutions which decreased to 58.9 and 33%, respectively, owing to the competition in occupying active sites during adsorption and reduction. The adsorption kinetics study showed that the adsorption fitted well to the pseudo-second-order, while the reduction kinetics study showed that the removal due to functional iron nanoparticles was well-fitted to the pseudo-first-order model. Hence, sorption kinetics followed pseudo-second-order, while reduction followed pseudo-first-order, in this case.

The mechanisms involved in the removal of heavy metals with nZVI include adsorption, absorption, precipitation, co-precipitation, oxidation, and reduction. The predominant mechanism depends on the redox nature of the pollutant and solution pH, as well as ionic strength and dosage of the adsorbent (Yan et al. 2012). The removal of arsenate using blueberry leaves-supported nZVI also showed the formation of transitional compounds during removal, along with different removal mechanisms as mentioned above, due to the various oxidation states of iron (Manquián-Cerda et al. 2017).

Wastewater discharges containing inorganic nutrients like nitrogen and phosphorus disposed in aquatic bodies cause eutrophication, which is one of the most prevalent global water pollution problems (Jiang et al. 2019). Increased use of mineral and organic fertilizers and generation of higher quantities of industrial and municipal sludge are factors responsible for nitrogen and phosphorus in the wastewater discharged (Kończak & Huber 2022). Eutrophication promotes algal bloom which results in increased oxygen consumption, which in turn, causes aquatic and ecosystem imbalance (Govarthanan et al. 2020). It also causes loss of biodiversity, reduced visual amenity, unpleasant odor, toxin production, and impaired habitat for fish species (Abell et al. 2022). Water with an excess amount of nutrients, if consumed, poses threats to human health, along with ecosystem imbalance; hence, nutrient removal is necessary. Plant-based nZVI is capable of removing nutrients like nitrogen and phosphorous in their various forms. Table 5 presents the summary of some of the recent studies on nutrient removal by plant-based nZVI.

Table 5

Summary of the studies on the removal of nutrients using plant-based nZVI

Plant material .	Nutrient .	qm (mg/g) .	Removal (%) .	Optimum conditions .	Kinetics model .	Isotherm model .	Added reagent .	Reference .
Green tea		13.06	60%	pH: 8.45, nZVI dose: 1 g/L, Initial nutrient concentration: 20 mg/L, Contact time: 90 min	Pseudo-second-order	-		Wang et al. (2014)
Eucalyptus		9.698	41.40%	pH: 8.45, nZVI dose: 1 g/L, Initial nutrient concentration: 20 mg/L, Contact time: 90 min	Pseudo-second-order	−		Wang et al. (2014)
Eucalyptus		18.69	97.30%	pH: 1–9, nZVI dose: 2 g/L, Initial nutrient concentration: 20 mg/L, Contact time: 90 min	Pseudo-second-order	Langmuir	CTAB	Cao et al. (2016)
Eucalyptus		3.47	43.30%	pH: 4, nZVI dose: 7 g/L, Initial nutrient concentration: 10 mg/L, Contact time: 30 min	Pseudo-second-order	Langmuir	Zeolite	Xu et al. (2020)
Eucalyptus		38.91	99.80%	pH: 8, nZVI dose: 7 g/L, Initial nutrient concentration: 10 mg/L, Contact time: 30 min	Pseudo-second-order	Langmuir	Zeolite	Xu et al. (2020)
Azadiracta indica		−	98.08%	pH: 7.5, Initial nutrient concentration: 92.5 mg/L, Contact time: 13 days	−	−	−	Devatha et al. (2016)
Azadiracta indica		−	84.32%	pH: 6.5–8, Initial nutrient concentration: 44.46 mg/L, Contact time: 13 days	−	−	−	Devatha et al. (2016)
Eichhornia crassipes		–	74.52%	Initial nutrient concentration: 14.77 mg/L, Contact time: 5 days	−	−	−	Prabhakar et al. (2017)
Mimosa Pudica		–	65.23%	Initial nutrient concentration: 14.77 mg/L, Contact time: 5 days	−	−	−	Prabhakar et al. (2017)
Lantana camara		−	53.68%	Initial nutrient concentration: 23.5 mg/L, Contact time: 5 days	−	−	−	Prabhakar et al. (2017)
Eucaluptus	N	−	71.70%	Initial nutrient concentration: 25.57 mg/L, Contact time: 21 days	−	−	−	Wang et al. (2014)
Eucaluptus	P	−	30%	Initial nutrient concentration: 10.55 mg/L, Contact time: 5 days	−	−	−	Wang et al. (2014)
Banana		–	88.00%	nZVI dose: 9.5 mg/L, Initial nutrient concentration: 20 mg/L, Contact time: 5 min	Pseudo-second-order	Langmuir	−	Abida et al. (2022)
Mango		−	78.00%	nZVI dose: 9.5 mg/L, Initial nutrient concentration: 20 mg/L, Contact time: 4 h	Pseudo-second-order	Langmuir	−	Abida et al. (2022)
Pomegranate		−	88.00%	nZVI dose: 9.5 mg/L, Initial nutrient concentration: 20 mg/L, Contact time: 4 h	Pseudo-second-order	Langmuir	−	Abida et al. (2022)

Plant material .	Nutrient .	qm (mg/g) .	Removal (%) .	Optimum conditions .	Kinetics model .	Isotherm model .	Added reagent .	Reference .
Green tea		13.06	60%	pH: 8.45, nZVI dose: 1 g/L, Initial nutrient concentration: 20 mg/L, Contact time: 90 min	Pseudo-second-order	-		Wang et al. (2014)
Eucalyptus		9.698	41.40%	pH: 8.45, nZVI dose: 1 g/L, Initial nutrient concentration: 20 mg/L, Contact time: 90 min	Pseudo-second-order	−		Wang et al. (2014)
Eucalyptus		18.69	97.30%	pH: 1–9, nZVI dose: 2 g/L, Initial nutrient concentration: 20 mg/L, Contact time: 90 min	Pseudo-second-order	Langmuir	CTAB	Cao et al. (2016)
Eucalyptus		3.47	43.30%	pH: 4, nZVI dose: 7 g/L, Initial nutrient concentration: 10 mg/L, Contact time: 30 min	Pseudo-second-order	Langmuir	Zeolite	Xu et al. (2020)
Eucalyptus		38.91	99.80%	pH: 8, nZVI dose: 7 g/L, Initial nutrient concentration: 10 mg/L, Contact time: 30 min	Pseudo-second-order	Langmuir	Zeolite	Xu et al. (2020)
Azadiracta indica		−	98.08%	pH: 7.5, Initial nutrient concentration: 92.5 mg/L, Contact time: 13 days	−	−	−	Devatha et al. (2016)
Azadiracta indica		−	84.32%	pH: 6.5–8, Initial nutrient concentration: 44.46 mg/L, Contact time: 13 days	−	−	−	Devatha et al. (2016)
Eichhornia crassipes		–	74.52%	Initial nutrient concentration: 14.77 mg/L, Contact time: 5 days	−	−	−	Prabhakar et al. (2017)
Mimosa Pudica		–	65.23%	Initial nutrient concentration: 14.77 mg/L, Contact time: 5 days	−	−	−	Prabhakar et al. (2017)
Lantana camara		−	53.68%	Initial nutrient concentration: 23.5 mg/L, Contact time: 5 days	−	−	−	Prabhakar et al. (2017)
Eucaluptus	N	−	71.70%	Initial nutrient concentration: 25.57 mg/L, Contact time: 21 days	−	−	−	Wang et al. (2014)
Eucaluptus	P	−	30%	Initial nutrient concentration: 10.55 mg/L, Contact time: 5 days	−	−	−	Wang et al. (2014)
Banana		–	88.00%	nZVI dose: 9.5 mg/L, Initial nutrient concentration: 20 mg/L, Contact time: 5 min	Pseudo-second-order	Langmuir	−	Abida et al. (2022)
Mango		−	78.00%	nZVI dose: 9.5 mg/L, Initial nutrient concentration: 20 mg/L, Contact time: 4 h	Pseudo-second-order	Langmuir	−	Abida et al. (2022)
Pomegranate		−	88.00%	nZVI dose: 9.5 mg/L, Initial nutrient concentration: 20 mg/L, Contact time: 4 h	Pseudo-second-order	Langmuir	−	Abida et al. (2022)

View Large

The adsorption capacity of chemically synthesized nZVI for nutrients is generally higher than plant-based nZVI. Wang et al. (2014) showed the equilibrium adsorption capacity of chemical nZVI to be 20.95 mg/g and that of green tea and Eucalyptus-based nZVI to be 13.06 and 9.69 mg/g, respectively, for nitrate. On the other hand, plant-based nanoparticles are seen to be less effective for the removal of nutrients as compared to chemically synthesized nanoparticles. Prabhakar et al. (2017) showed the removal of nitrate and phosphate by chemically synthesized nZVI to be 85.27 and 67.98%, respectively, while Eichhornia crassipes-based nZVI was able to remove 74.52% nitrate and 55.39% phosphate.

The adsorption capacity changes with variation in pH are based on the chemistry between the functional groups and the reaction behavior of the adsorbent and the adsorbate. Devatha et al. (2016) showed the removal of total phosphate and nitrate using leaf extracts of Mangifera indica, Murraya koenigii, Azadiracta indica, and Magnolia champaca. For phosphate, which is in anionic form, the removal was better at a lower pH, as in that condition iron nanoparticles are in the cationic form, making the removal effective by electrostatic attraction. Hence, in the pH range of 3.5–5.5, the removal efficiency of phosphate was 90% by Mangifera indica, Murraya koenigii, and Magnolia champaca-based nZVI. However, for the Azadiracta indica plant-based nZVI, at a pH of 7.5, high phosphate removal efficiency (98%) was observed, which may be due to the presence of various polyphenols, making the adsorption more efficient. Nitrate removal was observed to be better by all the aforementioned plant-based nZVI at a pH of 6.5–8.0, owing to the presence of more polyphenols, and the removal mechanism was concluded to be adsorption and co-precipitation.

Most reported studies show better removal as temperature increases, but there are exceptions depending on the nutrient to be removed and the bond between the adsorbent and the adsorbate. If the removal is improved with increasing temperature, the reaction is endothermic. Cao et al. (2016) showed the removal of phosphate using nanoparticles synthesized by Eucalyptus leaf extract, along with the use of CTAB (cetyltrimethylammonium bromide) as a stabilizer in the green synthesis process. Based on batch experiments with 20 mg/L of phosphate at different temperatures of 288, 298, and 308 K, the phosphate uptake values were 5.64, 7.94, and 9.41 mg/g, respectively, which indicated that an increase in the temperature led to an increased adsorption capacity, within a few minutes itself. The increased adsorption uptake is supported with a thermodynamic study. The results showed a positive value of enthalpy (ΔH = 119.15 kJ/mol) indicating an endothermic process. The value of ΔG (Gibbs free energy) decreased from −2.81 to −11.3 kJ/mol with the increase in temperature (288–308 K), showing spontaneity in the adsorption process. The positive value of entropy (ΔS = 422.55 J/k mol) showed increased randomness at the solid–liquid interface during adsorption. These phenomena are because of the increase in activation energy, causing an increased reaction velocity, making the uptake easier with temperature increase. The mechanism associated with adsorption here was an inner-sphere complexing mechanism and electrostatic attraction.

Xu et al. (2020) showed ammonia and phosphate removal using nanoparticles synthesized from Eucalyptus leaves with zeolite as a supporter. Batch experiments for 10 mg/L concentration for simultaneous removal of and were performed at various temperatures of 20, 30, and 40 °C. Ammonia removal decreased with an increase in temperature, suggesting the process to be exothermic. The increase in temperature did not directly contribute to the adsorption of phosphate, but reduced the equilibrium time. For both and ⁠, ΔH was negative with values of −3.40 and −10.51 kJ/mol, respectively, indicating the reaction to be exothermic. The positive values of ΔS, 0.0096 for ⁠, and 0.0049 for showed that the solid–liquid interface degree of disorder decreased during the reaction. ΔG values were negative at all temperatures, showing the adsorption to be spontaneous and favorable. The results of experimental data showed that the mechanism of adsorption of was physical adsorption via electrostatic attraction, while that of was chemisorption, wherein removal takes place by an exchange between Fe–OH formed and phosphate.

Studies show that the ability of nanoparticles synthesized from plant extract for nutrient uptake decreases as they age. The freshly synthesized plant-based nano-zerovalent iron shows a better removal efficiency as compared to aged nanoparticles. Prabhakar et al. (2017) showed the use of aquatic and terrestrial weed-based plants for nZVI synthesis. The synthesized nanoparticles showed high efficiency for nitrate and phosphate removal. Eichhornia crassipes, Lantana camara, Mimosa pudica, and chemically synthesized nZVI showed a nitrate removal efficiency of 74.52, 71.12, 65.23, and 85.27% and a phosphate removal efficiency of 55.39, 53.68, 46.23, and 67.98%, respectively. Green synthesized nZVI was left open for 15 days, and then, its efficiency was slightly lower with nitrate removal of 67.21, 66.29, 57.53 and 63.21% and phosphate removal of 43.67, 40.53, 38.17 and 39.32%, respectively.

The kinetics followed by plant-based nZVI for the removal of nutrients is seen to fit well in the pseudo-second-order model. Wang et al. (2014) used green tea and Eucalyptus leaves to synthesize nZVI for the removal of nitrates. The use of these plant-based nZVI was successful as green tea-based nanoparticles removed 59.7% and Eucalyptus leaves-based nanoparticles removed 41.4% of nitrates present in wastewater. They also compared the removal with chemically synthesized nZVI, whose removal efficiency was 87.6%, which was more than green nZVI. The adsorption capacity at equilibrium was 13.06 and 9.698 mg/g for green tea and Eucalyptus nanoparticles, respectively. The major mechanisms responsible here were reduction by nZVI along with adsorption and co-precipitation.

Batch adsorption experiments showed the Langmuir adsorption isotherm in the majority of analyses. Abida et al. (2022) studied the removal of nitrates and phosphates using green synthesized nZVI from peels of mango, banana, and pomegranate. The Langmuir isotherm was proved to be best fitted for the adsorption of phosphates at various concentrations, with an equilibrium time of 60 min. The maximum adsorption capacity was found to be higher for nZVI prepared from pomegranate peels (68.45 mg/g), followed by banana and mango peels, with values of 35.71 and 24.39 mg/g, respectively. The adsorption kinetics followed pseudo-second-order and the small value of rate coefficient k showed the process to be slow.

The removal of nutrients from wastewater using plant-based nZVI involves electrostatic attraction, adsorption, and co-precipitation (Cao et al. 2016). Xu et al. (2020) showed interactions between and ⁠, and Eucalyptus-based nZVI supported by Zeolite, with electron sharing or ligand exchange mechanisms. Prabhakar et al. (2017) showed that the removal of nitrate and phosphate from chemically synthesized nZVI involved a reduction mechanism, making it better than Eichhornia crassipes, Lantana camara, and Mimosa pudica-based nZVI, which involved adsorption and co-precipitation mechanisms because of the presence of polyphenols as the capping layer.

Trace organic compounds are organic compounds that are natural or synthetic, mostly non-biodegradable, and cause harm as they are hazardous even at low concentrations (Tufail et al. 2020). They include pesticides, endocrine-disrupting chemicals, personal care products, pharmaceutical chemicals, disinfection, and industrial wastewater by-products (Shin et al. 2022). The presence of these trace organics is known to disrupt the aquatic ecosystem along with having a highly negative impact on humans and wildlife (Alexander et al. 2012). Trace organics have various functional groups like electron donating or electron accepting that have the property of being toxic, recalcitrant, and resistant to the basic physicochemical and biological processes, hence, requiring advanced processes for their removal (Li et al. 2022). Plant-based nZVI has been studied for removing major pharmaceutical compounds among the different classes of trace organics, and these studies are summarized in Table 6.

As observed from the literature, low pH gives better removal of pharmaceutical compounds. Demirezen et al. (2019) showed the removal of amoxicillin using green nanoparticles synthesized from Ceratonia siliqua (Carob pod). The removal was tested at various pH values. The results showed the highest removal efficiency of 99% at pH 2. At higher pH, inactivation of active sites takes place and a passive oxide layer is formed on the surface, decreasing the adsorption capacity. It was also observed that for a time period of 80 min, the removal was more at pH 5, owing to electrostatic interactions, but at a longer contact time, pH 2 also gave high removal efficiency. Jha & Chakraborty (2023) showed the removal of tetracycline and ciprofloxacin antibiotics using Shorea robusta leaf extract. Removal efficiencies of 82.3 and 77.9% were observed at pH 6 for tetracycline and ciprofloxacin, respectively. A high removal efficiency was observed in the pH range of 4–8.

Batch studies are performed to understand the removal process. Since the trace organics group includes major chemicals with varying functional groups, the adsorption isotherm also varies with each individual case, depending on the chemicals, functional groups, and electrostatic nature of the adsorbent and the adsorbate. Leili et al. (2018) prepared nZVI using nettle and thyme leaf extracts for the removal of antibiotic cephalexin. The experimental data fitted well in the Freundlich model (R² = 0.992) for nettle leaf with a maximum adsorption capacity of 1,667 mg/g, indicating a heterogeneous adsorbent surface and multilayer adsorption. The adsorption isotherm for removal by thyme-based nZVI fitted better to the Langmuir model (R² = 0.994), indicating monolayer adsorption with a maximum adsorption capacity of 1,428 mg/g. Abdelfatah et al. (2021b) used nZVI synthesized from Riccinus communis for the removal of tetracycline. The value of R² was highest for the Langmuir model, suggesting it to be the best fit. The maximum adsorption capacity of 72.64 mg/g was achieved with 98% removal in 45 min. Adsorption and reduction were the mechanisms of removal.

Leili et al. (2018) used nettle and thyme-based nZVI for the removal of cephalexin. The pseudo-first-order kinetics model was a better fit, based on the experimental data for both the adsorbents. The coefficients suggested that diffusion through the boundary liquid layer around the adsorbent was the main controlling step because the R² value of film diffusion was higher than that of intra-particle diffusion. Adsorption was favorable as values of constant n in the Freundlich equation were 0.501 and 0.505 for nettle and thyme nZVI, respectively. Abdelfatah et al. (2021b) used Riccinus communis-based nZVI for the removal of tetracycline. The experimental data of removal well-fitted to the pseudo-second-order kinetic model. The decreasing value of the kinetic coefficient showed that the removal was a physical adsorption process. Generally, van der Waal's forces are responsible for physical adsorption requiring very less adsorption energy.

The two major mechanisms for the removal of trace organics are adsorption or adsorption followed by reduction. Conde-Cid et al. (2021) showed the use of black tea leaves and vineyard pruning residue to synthesize nZVI for the removal of sulfadiazine, which is a sulfonamide antibiotic. They aimed at quantifying the removal, owing to adsorption alone and a combination of adsorption and reduction. These two experiments were distinguished by preparing two types of green nZVI. The first nZVI was made unreactive by exposing it to atmospheric oxygen so that all the uptake would be due to adsorption and the second type was reactive nZVI capable of reacting with the adsorbate, and hence, the removal of antibiotic would be done because of adsorption and reduction. Hydrolysis and biodegradation were not the removal mechanisms here, as there was no removal of sulfadiazine without the presence of green nZVI. For unreactive nZVI, the highest removal efficiency for maximum dose was 58% showing removal via adsorption mechanism, while it was 69% when reactive nZVI was used showing reductive degradation as the dominant mechanism in the initial phase, followed by adsorption. A long contact time of 192 h was required. They also showed the increase in removal efficiency to be 90% and the drastic decrease in contact time to be 1 h, when H₂O₂ was used as the catalyst for the Fenton reaction, and further showed that the use of green nZVI in simulated sunlight would highly enhance the removal process by giving a complete removal in 5 min.

Regeneration and reusability of synthesized nZVI are important to have a sustainable life and to make it environment friendly and cost-effective (Li et al. 2023). The regeneration method generally used is chemical regeneration, wherein the removal of contaminants from exhausted nZVI is done by the ion-exchange process when washed with a chemical solvent (Rashtbari et al. 2020). The different solvents commonly used to remove the contaminants from the surface of plant-based nZVI include NaOH, HCl, EDTA, H₂SO_4, HNO₃, NaNO_3, and Ca (NO₃)₂ (Hassan et al. 2020). Once treated with the solvent, the adsorbent is ready to be reused, although the efficiency decreases after a few cycles because of the reduction in active sites of the adsorbent and the change in their physicochemical properties (Pourali et al. 2022).

Lin et al. (2020) studied the removal of heavy metal Pb(II) and trace organic rifampicin using nZVI synthesized from tea extracts. Regeneration using three different solutions, 0.1 M HCl, absolute ethanol, and deionized water was attempted. The removal efficiency was high even after five cycles. The removal of Pb(II) and rifampicin was 52.3 and 64.9%, respectively, after five cycles, as compared to 97.5 and 68.8% in the first cycle. The use of kaolin-supported green nZVI prepared from Ruellia tuberosa leaves for decolorization of azo dye was shown by Khunjan & Kasikamphaiboon (2021). The removal efficiency in the first cycle was 99.7% which decreased to 88.6% in the fifth cycle when used after regeneration. Fatimah et al. (2020) showed the synthesis of nanoparticles using Parkia speciosa Hassk pod for the degradation of bromophenol blue dye. The reusability studies showed promising values as not much change in degradation efficiency and rate was observed. The reduction in the degradation rate was less than 5% between the first cycle and the third cycle. Moringa oliefera leaf extract was used to prepare iron nanoparticles for fluoride removal by Silveira et al. (2018). The regeneration was done by washing the adsorbent with 20% ethanol. The results showed little reduction in adsorption efficiency from 80.94 to 80.44% after the first wash. The decrease in later cycles was up to 75% and it was possible to reuse it three times. Pan et al. (2019) synthesized nZVI using red peanut skin extract for the removal of Cr(VI). Hydrochloric acid (16 M) was used as a regenerant and caused desorption of 70.2% Cr(VI). The removal efficiency of desorbed iron nanoparticles decreased significantly to 59.9% in the second cycle from complete removal in the first cycle. Ahmad et al. (2020) used date palm waste to synthesize nZVI for MB removal. The regeneration of nZVI was done using HCl (0.1 M). The removal efficiency decreased to 29.78% after four cycles of regeneration, while it was more than 80% in the first cycle.

While the potential of chemically synthesized nZVI particles has been known for their efficiency in removing different classes of pollutants from water, their large-scale commercial application is limited due to several drawbacks such as high cost of manufacturing, usage of toxic chemicals, and contamination from precursor chemicals. This has resulted in studies on the development of plant-based nZVI, in recent years. As described in this paper, plant-based nZVI has shown improved material characteristics resulting in increased adsorption capacity for removing a variety of contaminants, such as dyes, heavy metals, nutrients and trace organic pollutants. Furthermore, plant-based nZVI particles have shown much less aggregation compared to chemically synthesized particles, resulting in more colloidal stability in solutions.

The studies show the use of different plant parts by researchers for the synthesis of nZVI, resulting in varied characteristics for the nZVI formed. The characteristics play an important role in the usefulness and efficiency of the material synthesized. Efforts should be made to identify the best plant materials suitable for the synthesis, based on local conditions. Plant-based synthesis using plants with high polyphenol and flavonoid contents should be attempted. Utilizing abundantly available weed plants for nZVI synthesis would be a sustainable approach. This approach would help to reduce waste and pollution, while simultaneously providing cost-effective and eco-friendly solutions for wastewater treatment. Functional groups responsible for the reduction of Fe⁺³ and their presence in synthesized nZVI should be correlated by conducting suitably detailed chemical analyses of plant extracts and nZVI. This can help in selecting the best plant parts for synthesizing nZVI and in optimizing the synthesis procedure. Furthermore, various chemical groups in plants have different reducing capacities and it is necessary to ensure uniformity in the synthesized nanoparticles. Thus, more studies should be conducted to characterize synthesized nZVI.

Most of the studies reported in the literature utilized small-scale laboratory batch tests under ideal conditions. More tests should be conducted to evaluate the performance of the medium in the presence of competing and interfering ions and compounds. A comparative long-term evaluation of the chemically synthesized and plant-based nanoparticles should be conducted on a large scale using real wastewater or natural water. This can reveal the true potential of plant-based nZVI.

Increasing the reusability of plant-based nZVI would not only reduce sludge production, but also make the process cost-effective. A few studies have shown that plant-based nZVI can be regenerated. More research should be conducted on different methods of regenerating the spent nZVI particles, such as solvent treatment and surface modification, and on evaluating their effectiveness in restoring the effectiveness of the original nZVI.

Several innovative applications of nZVI have been suggested recently, and several classes of pollutants have been shown to be effectively removed by plant-based nZVI particles. Efforts should be made to optimize the process parameters. Very few studies have been reported on the use of these particles for removing/inactivating pathogenic organisms from water, and more studies should be conducted on this aspect because nZVI can potentially be used as a medium for small-scale/household drinking water treatment due to its proven ability to remove various classes of chemical contaminants from water.

In summary, continued research on plant-based nZVI particles to fully understand their potential in wastewater treatment, especially for the removal of a wider variety of pollutants, and for improving sustainability and reducing the cost and environmental impact of the process, is necessary.

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

The authors declare there is no conflict.