Removal of pollutants from landfill leachate by adsorption with nano zero-valent iron particles: adsorption isotherms and kinetic studies

Adsorption can be defined as the absorption of solutes by liquid or solid surfaces as part of the intermass transport mechanism (Cabooter et al. 2021). Adsorption is a significant treatment technology when used singly or in a combined treatment process to remove organic and inorganic substances (Rodriguez et al. 2004; Kurniawan & Lo 2009; Kulikowska et al. 2016; Lee & Hur 2016; Oloibiri et al. 2017). In recent years, nanoparticles (NPs) have been proposed as efficient adsorbents in the remediation of water and soil because of their unique atomic properties, higher reduction capacity, higher removal efficiency, magnetic property, nontoxic, molecular, and chemical properties compared to traditional adsorbents such as granulated-powdered and fiber-activated carbons Granular Activated Carbon (GAC), Powdered Activated Carbon (PAC) and Activated Fiber Carbon (AFC) (Hu et al. 2004; Chang & Chen 2005;,Yantasee et al. 2007; Liu et al. 2008; Yuan et al. 2010). All these properties can be used in the treatment of different types of wastewaters. There are many reports, where different categories of adsorbents had been used for organic and inorganic pollutants from wastewaters such as activated carbon (AC) (Ai et al. 2011), graphenes (Tamer & Berber 2022), ash (Liu et al. 2018), agricultural wastes (Crini 2006), biomass (Canizo et al. 2019), clay materials (Loqman et al. 2017), silica (Li et al. 2020a, 2020b), zeolites (Wu et al. 2021), acid-based hydrogels (Yuan et al. 2019), magnetic-coated biochars (Sun et al. 2015), activated alumina (Adak et al. 2005), magnetic oxide-based materials (Dutta et al. 2021), nonmagnetic iron oxide (Hamidzadeh et al. 2015), nanorods (Thattil & Leema Rose 2020), and magnetic calcium ferrite NPs (An et al. 2015). In recent years, adsorption processes using iron-based nanomaterials have attracted the attention of researchers due to their high surface area, cost-effectiveness, and properties to remove pollutants from wastewater (Amiri et al. 2018,Nisticò & Carlos 2019; Li et al. 2020a, 2020b; Yi et al. 2020). The commonly used NPs for wastewater treatment are nano zero-valent iron (nZVI), metal oxide NPs, carbon nanotubes, and nanocomposites. Among them, nZVI has achieved very interesting and promising results for various pollutant remediation such as pesticides, polychlorinated hydrocarbons, chlorobenzenes, and coloring agents (Ghasemzadeh et al. 2014; Galdames et al. 2020). nZVI has been reported to be an environmental friendly material with high reactivity, effective yield, controllable particle size, active site with a diameter ranging from 1 to 100 nm. nZVI also has small size, large specific surface area, and strong reducing capacity (Fan et al. 2020; Ken & Sinha 2020). In addition, nZVI has high aggregation due to its high surface area and magnetic properties. High surface area is preferred in the removal of pollutants in wastewater because nZVI has a high adsorption capacity. As a result, a small amount is needed to reduce contaminants in an aqueous solution, thus reducing its cost (Stefaniuk et al. 2016; Wang et al. 2019).

Landfill leachate (LFL) is very strong wastewater and causes environmental concern due to the wide range of polluting parameters it contains. LFL has a highly contaminated complex content of biodegradable and nonbiodegradable substances (Ziyang et al. 2009; de Godoy Leme & Miguel 2018). There are many factors affecting the quality of such leachates, i.e., age, precipitation, seasonal weather variation, waste type, composition, operating procedures, hydrogeological factors, climatic conditions of the region, garbage depth and permeability, storage method, and changes. LFL is dark brown colored and odorous wastewater containing many suspended and dissolved organic substances, pathogenic Fe element content was determined as Biochemical oxygen demand (BOD), chemical oxygen demand (COD), dissolved organic carbon (DOC), nitrogenous compounds (NH₃–N, organic nitrogen, NO₃–N, ⁠, etc.), heavy metals, and xenobiotics substances (Akkaya & Demir 2009; Atmaca 2009; Gotvajn et al. 2009; Lou et al. 2009). The age of LFL is classified as young, middle, and mature. Generally, storage age plays an important role in the quality of collected LFL (Renou et al. 2008). The age of LFL is determined by the BOD/COD ratio. The BOD/COD ratio for young, medium, and mature has value ranges of 0.5–1.0, 0.1–0.5, and <0.1, respectively. Mature LFL is characterized by organic substances resistant to biodegradation such as humic acid and high concentrations of NH₃–N. As the LFL age increases, the rate of easily decomposed organic substances decreases after completion of biological decomposition (Renou et al. 2008). Mature LFL (>10 years) is rich in ammonia nitrogen due to fermentation and hydrolysis of nitrogenous fractions of biodegradable substrates (Nordin 2006). As time progresses, leachate passes through the aerobic, acetogenic, methanogenic, and stabilization stages of organic waste degradation, and in these stages, the properties of leachate such as COD, BOD, BOD/COD ratio, NH₄⁺–N, and pH vary considerably (Kjeldsen et al. 2002). LFL treatment is very complicated, is expensive, and generally requires a combination of various treatment processes (Bashir et al. 2010). Although several methods of LFL treatment are available, adsorption has been widely used as pretreatment or post-treatment due to its simplicity and cost-effectiveness. Furthermore, the adsorption process of LFL by nZVI is preferred comparatively with other treatment methods due to its high efficiency, ease of operation, and ability to wastewater treatment. Also, many isotherm and kinetics models have been proposed for describing mechanisms of NP adsorption from the aqueous solution. However, the report provides practical guidance for choosing an appropriate method by comparing Langmuir and Freundlich isotherm models, and pseudo-first-order and pseudo-second-order models have been tested for modeling the adsorption kinetics of NPs from the aqueous solution (Ruíz-Baltazar et al. 2015). Although different isotherm models have been developed to describe the sorption of several pollutants on nZVI, there has been very limited information on adsorption performance toward LFL. Kashitarash et al. (2012) reported fast removal efficiency of 47.94% for COD in 10 min at optimal conditions; however, sorption interactions onto nZVI particles have not been included in this study. Thus, for a deeper understanding of the LFL treatment and the related sorption interactions onto nZVI particles, the Langmuir and Freundlich isotherms have been considered in this study. Moreover, pseudo-first-order and pseudo-second-order kinetics models were applied to the sorption data for a deeper insight into the removal mechanism. Conde-González et al. (2016) reported that AgNPs, using copper-based metal NP material, fit well with pseudo-second-order and Langmuir models compared to adsorption from aqueous solutions, but that the adsorption did not continue beyond the monolayer, describing the equilibrium behavior better than the Freundlich isotherm. In another study, experimental data for the adsorption of AgNPs on commercial AC showed that the Freundlich isotherm can describe the equilibrium behavior better than the Langmuir isotherm because the adsorption continues beyond the monolayer (Gicheva & Yordanov 2013). Results from different studies have shown that the usefulness of statistical tests in model validation for the adsorption of NPs onto the surface of a material is very limited.

The main aim of the present research was to investigate the removal of DOC, COD, ⁠, ⁠, and color of LFL by applying nZVI as an efficient adsorbent. The suitability of the selected adsorption isotherms (Langmuir and Freundlich) and kinetics (pseduo-first-order and second-order equations) were determined. Moreover, the adsorption mechanism of LFL on the surface of nZVI was investigated in a batch adsorption process. Finally, DOC, COD, ⁠, and of LFL and the effect of pH, reaction time, and nZVI concentration variation were investigated to determine the optimum conditions.

In this method, 1.7868 g FeSO₄•7H₂O was dissolved in 90 mL 4/1 (v/v) ethanol/deionized water mixture in a 500-mL bottle and then 0.3 g PEG-4000 was added into the aforementioned solution. The temperature was kept at 20°C. The solution was stirred at 220 rpm for 30 min to ensure that PEG-4000 was completely dissolved. Before reductant addition, the solution pH was adjusted to about 6.5 with 1 M NaOH. Then, 50 mL of 1.3883 g KBH₄ aqueous solution was added dropwise into the mixture at 220 rpm. The solution was stirred for another 30 min after the addition of all KBH₄. Then the resulting black solid particles were washed with deoxygenated water three times and with deoxygenated absolute ethanol two times and were collected by magnetic separation. Finally, the nZVI particles were dried at 70°C and stored in N₂ gas to prevent nZVI oxidation from atmospheric oxygen. In our previous work, synthesized nZVI characterization was carried out using X-ray diffraction (XRD), scanning electron microscopy with energy dispersive X-ray spectroscopy (SEM-EDX), and Fourier-transform infrared spectroscopy (FTIR) techniques (Göçer et al. 2019). According to XRD, the 2θ peak at 44.8096°, representing 100% intensity, indicates the presence of nZVI NPs in the sample. According to this result, it was observed that nZVI was synthesized successfully and that nZVI showed superiority against oxidation. According to SEM and EDX, the surface roughness of nZVI is indicated to have a core–shell structure, where the shell represents the oxidized part surrounding the Fe core and protects it from further oxidation. Also, based on EDX results of nZVI, the Fe element content was determined to be 98%. The Brunauer-Emmett-Teller (BET) surface area of nZVI was determined to be 36.8063 m²/g. The peaks indicating the presence of nZVI were detected in FTIR analysis to be between 500 and 1,200 cm⁻¹ (Göçer et al. 2019).

Raw LFL was collected from a municipal sanitary landfill located in Kahramanmaras, Turkey. The total amounts of LFL deposited daily were 815–830 tons. LFL was collected from the equilibration tank and stored at 4°C until used. The characteristics of raw LFL are summarized in Table 1. The analyses show the typical characteristics of medium LFLs, with high COD, high ammonium concentration, and BOD/COD ratio 0.10. In general, the pH value in medium LFL is determined as 6.5–7.5 (Table 1). The pH value of experimental sets varied between 3 and 8, by adding 1 M NaOH and 1 M HCl to adsorption experiments.

Table 1

Landfill leachate characterization

Parameters (this study) .	Concentration (mg/L) (this study) .	Concentration (mg/L) .	Reference .
pH	6.5–7.5	7.7	Lau et al. (2001)
DOC	7058 ± 400	1400–11,000 ± 400	Duyar et al. (2021)
COD	16,000 ± 1500	10,000–18,000 ± 1500	Çeçen & Aktaş (2004)
BOD	1500 ± 300	2300 ± 300	Lopez et al. (2004)
	2120 ± 200	2020 ± 200	Ozturk et al. (2003)
BOD/COD	0.10 (medium age)	0.19	Li & Zhao (2001)
	670 ± 40	187 ± 40	Duyar et al. (2021)
Pt-Co (color unit)	6380 ± 300	Dark brown	Duyar et al. (2021)
	78 ± 10	32 ± 10	Lopez et al. (2004)

Parameters (this study) .	Concentration (mg/L) (this study) .	Concentration (mg/L) .	Reference .
pH	6.5–7.5	7.7	Lau et al. (2001)
DOC	7058 ± 400	1400–11,000 ± 400	Duyar et al. (2021)
COD	16,000 ± 1500	10,000–18,000 ± 1500	Çeçen & Aktaş (2004)
BOD	1500 ± 300	2300 ± 300	Lopez et al. (2004)
	2120 ± 200	2020 ± 200	Ozturk et al. (2003)
BOD/COD	0.10 (medium age)	0.19	Li & Zhao (2001)
	670 ± 40	187 ± 40	Duyar et al. (2021)
Pt-Co (color unit)	6380 ± 300	Dark brown	Duyar et al. (2021)
	78 ± 10	32 ± 10	Lopez et al. (2004)

View Large

The effect of various parameters such as pH (3–8), contact time (15–330 min), and nZVI concentration (50–500 mgFe⁰/L) was tested on ⁠, ⁠, DOC, and COD removal in batch adsorption experiments, for the evaluation of the optimum process conditions (Table 2). The adsorption tests were performed in ‘Jar test’ equipment (VELP scientific, JLT 6, Italy), where six simultaneous tests could be performed. The jars tests (500 mL capacity) were submitted to a constant rotation speed of 200 rpm and the total test duration of 330 min. All the experiments were carried out at a laboratory scale using 500 mL Erlenmeyer flasks as batch reactors (Table 2). Batch adsorption tests were conducted to evaluate the isotherms and kinetics of adsorption onto the nZVI adsorbents at 25°C (room temperature). Stock solutions of nZVI were prepared by dissolving 1.7868 g FeSO₄.7H₂O in 90 mL 4/1 (v/v) ethanol/deionized water mixture in a 500-mL bottle. A calibration curve was then created based on 10 concentrations (range, 0–500 mg/L) at 25°C. Five concentrations (50–50 mg/L) of the solutions described earlier were prepared in 500-mL bottle. Using 50–100 mg/L nZVI, 500 mL of each LFL wastewater was added, followed by the jar test experiment stirring at pH 3–8 and 200 rpm for 330 min at room temperature (25°C). After the adsorption experiment, 20 mL of the solutions were centrifuged for 10 min at 3,500 rpm using a centrifuge.

In Equations (2) and (3), V is the volume of the leachate (L), W is the amount of adsorbent (g), and C_o and C_e are the initial and equilibrium concentrations of pollutants (mg L⁻¹) in the leachate, respectively.

All samples were centrifuged at 4,000 rpm for 5 min (Eppendorf Centrifuge 5415R, Hamburg, Germany) and then were filtered using a sterile syringe 0.45 μm filter (Sartorius AG, Gottingen, Germany). Dissolved oxygen carbon (DOC) and Total nitrogen (TN) DOC and TN concentrations were analyzed using a TOC instrument coupled with TN (Shimadzu TOC-VCPN, Kyoto, Japan). The pH was measured by a pH meter (Thermo, Orion 4 Star, Indonesia). The ionic composition of influent and effluent samples (ammonium, nitrate) was measured by ion chromatography (Dionex ICS-3000, Sunnyvale, CA, USA). The COD measurements were carried out according to the dichromate-closed reflux colorimetric method described in Standard Methods (Standard Methods, 5220 D). The color was analyzed as Pt-Co units. Pt-Co color measurements were performed spectrophotometrically at 465 nm during lab-scale studies.

The Langmuir and Freundlich isotherms and pseudo-first-order and pseudo-second-order kinetics models were selected to simulate the isotherm adsorption of nZVI in this work.

where q_e expresses the adsorption capacity (mg g⁻¹) and C₀ and C_t are pollutant concentrations (such as DOC, COD, ⁠, and ⁠) (mg L⁻¹) at time 0 and t, respectively. V indicates the volume of solution (mL) and m is the mass of nZVI (g). Adsorption kinetic parameters are divided into two: pseudo-first-order equations (Ho 2004) and pseudo-second-order equations (Azizian 2004; Ho 2006).

Adsorption of pollutants may decrease at lower and higher pH values (Zou et al. 2016). At lower pH values, the adsorption of metal ions is low due to increased nZVI corrosion. According to our study, it was observed that adsorption was effective at high pH. However, at lower pH, the nZVI surface will be positively charged, and this will negatively affect the adsorption of iron ions (Zou et al. 2016). It is clear that the removal efficiency increases with increasing nZVI concentrations. Removal efficiencies and adsorption capacities ranging from 50 to 500 mg/L nZVI concentrations are quite low (Figure 1). According to adsorption experiments, it is observed that the efficiency and adsorption capacity remain constant after 15 min of contact time. The results were similar to the previous studies (Efecan et al. 2009; Poguberovic et al. 2016; Rathor et al. 2017; Ulucan-Altuntas et al. 2017). While the removal efficiency of increasing nZVI concentration up to 500 mg/L within the same contact time increases almost linearly, it continues at a low angle after 50 mg/L nZVI concentration. On the other hand, it was observed that as the amount of adsorbent increased, the adsorption capacity decreased. A literature review of studies using AC reported that it is necessary to select high adsorbent concentrations in order to have high removal efficiency (Kadirvelu et al. 2001; Yavuz et al. 2003; Onundi et al. 2010; Wang et al. 2010a, 2010b). However, in our study, effective results were observed at a concentration of 50 mg/L nZVI.

Halim et al. (2010) reported that adsorption capacities of composite medium, zeolite, and AC were 32.89, 17.45, and 6.08 mg/g, respectively, which is in good agreement with our study. Eljamal et al. (2022) reported that NPs have great potential for the practical applications of pollutant removal from wastewater. Unlike our study, according to the study by Boparai et al. (2011), nZVI particles reported removal of Cd²⁺ in the concentration range of 25–450 mg/L. The maximum adsorption capacity of nZVI for Cd²⁺ was determined as 769.2 mg/g. They reported that nZVI can be used as an efficient adsorbent for the removal of cadmium from polluted water sources. Ashrafi et al. (2017) reported in a similar study using Fe₃O₄@MnO₂ core-shell magnetic NPs that optimum times for adsorption were reached in 60 min. Altuntas et al. (2018) reported in their study that while they achieved a removal efficiency of 24% by using AC as an adsorbent for nickel removal, they achieved a removal efficiency of approximately 80% with 200 mg/L nZVI. On the other hand, in the study where AC–nZVI NPs containing certain amounts of AC were used, it was observed that the removal efficiency increased up to 99% (Altuntas et al. 2018). A range of biodegradable, nonbiodegradable, and humic substances are detected in LFL (Lai et al. 2007; Jun et al. 2009) and contribute to the COD pollutants. Wang et al. (2010a, 2010b) obtained approximately 56 and 50% COD removal efficiency, respectively, by using mango peel nano zero valent iron (MP-nZVI) and starch nano zero valent iron (S-nZVI) modifications as adsorbents in their study. According to the results obtained, it is thought that the removal efficiency of COD increases with the adsorption of the organic compound and its precipitation with iron corrosion products. Also, nitrogen-based compounds are also pollutants found in LFL and cause serious environmental consequences. The removal efficiency of magnetic adsorbents varies depending on the adsorbent structure and pollutants used. In addition, temperature, pH, ionic strength, reaction time, adsorbent concentration, the presence of pollutants, and some other factors also play important roles. In this study, the effects of nZVI adsorbent material on LFL are given with examples. However, very few studies have been conducted using magnetic adsorbents to treat real LFL (Zhang et al. 2016, 2018). One study observed that approximately 30% of COD was removed from real LFL by adsorption onto magnetite (Augusto et al. 2019). In another study, a combination of oxidation and adsorption of mature LFL using magnetic CuFe₂O₄/Reduce Graphene Oxide (RGO) nanocatalyst was also reported (Karimipourfard et al. 2019). Unlike our study, Wang et al. (2023) obtained an adsorption capacity of 210.9 mg/g by using PS-HQ-HCP (phenolic hydroxyl-functionalized hyper-cross-linked polymers) as an adsorbent in their study.

The slope and intercept of plots of C_e/q_e versus C_e, at adsorbent concentration, were used to calculate q_m and K_L (Figure 2(a)). The size of q_max and K_L indicates high adsorption capacity (50 mg/L adsorption concentration, Table 3). Adsorption is low if and high if ⁠. Table 3 shows that K_LC_e is and the amount of adsorption is high. The adsorption process is inconvenient if R_L is greater than 1, it is linear if R_L is equal to 1, it is convenient if R_L is between 0 and 1, and it is irreversible if R_L is 0 (Equation (7)). When the Langmuir isotherm is examined, it is observed that the R_L value is between 0 and 1 (Table 3). The Freundlich isotherm constants K_F and n are determined from the intercept and slope of a plot of Lnq_e versus LnC_e (Figure 2(b)). It is 1/n from the slope of the LnC_e graph against Lnq_e and LnK_F from the cutting point of the y-axis (Equation (6)). The high LnK_F and n values show that sorbent has a high tendency to adsorption and adsorption capacity. The value of 1/n ranging from 0 to 1 is expressed as a measure of an adsorption tendency, and heterogeneity becomes heterogeneous as it approaches zero. 1/n < 1 indicates compatibility with Langmuir isotherm, and 1/n > 1 indicates compatibility with adsorption condition. As a result, it is observed that the DOC removal adapts to the Langmuir isotherm. The best adsorption capacity suitable for the Langmuir isotherm occurred at a concentration of 50 mg/L nZVI (R² > 0.9931). Langmuir isotherm parameter fits (Table 3) for DOC removal efficiency adsorption on nZVI yielded isotherms that were in good agreement with observed behavior (R² ≥ 0.99). Maamoun et al. (2021) investigated the suitability of various adsorption isotherms and kinetic models to explain the removal of phosphorus (P) from aqueous solutions by nZVI. They found that the Langmuir isotherm and the pseudo-second-order kinetic model are the best models with the highest linear and nonlinear correlation (R²). Similarly, in another study, the adsorption capacities of synthesized modified adsorbent (nZVI/AC) for arsenite and arsenate at pH 6.5 calculated from Langmuir adsorption isotherms in batch experiments were 18.2 and 12.0 mg/g, respectively. It can be concluded that nZVI without the use of support material reached the similar results due to its positive effect (Zhu et al. 2009). Adsorption of by nZVI particles was modeled using the Freundlich and Langmuir isotherms with the quality of the fit assessed using the correlation coefficient (Figure 3(a) and 3(b)). Isotherm parameters in terms of removal efficiency are listed in Table 4. Altuntas et al. (2018) reported in their study that the adsorption capacity was increased from 125 and 820 mg/g for AC and nZVI, respectively, to 1,190 mg/g for 50% AC–nZVI. They also reported that they found the Freundlich model to be the model that best represents isotherm modeling for nZVI. In another study, the adsorption isotherm and kinetic equations using aminated chitosan and graphene oxide-modified adsorbent are similar to our study (Yi et al. 2020).

It was observed that removal is fitted to the Freundlich isotherm (Figure 3(b)). In addition, K_LC_e is ⁠, and the amount of adsorption is high. According to the Langmuir isotherm, it was observed that the R_L value is between 0 and 1 (Table 4) confirming that the Langmuir isotherm is suitable for removal. The high LnK_F and n values show that sorbent has a high tendency for high adsorption capacity. Unlike our study, Kanel et al. (2005) investigated arsenic removal efficiency in groundwater by using nZVI as an adsorbent and observed 3.5 mg As(III)/g adsorption capacity with Freundlich isotherm. Other studies using new composite materials for LFL treatment found different results. Langmuir and Freundlich reported that the isotherm's regression coefficients (R²) for COD and ammonia nitrogen were 0.9971 and 0.9914, respectively (Detho et al. 2021). Hashemi et al. (2021) investigated ammonium removal efficiency from LFL (LL) using montmorillonite/hematite nanocomposite (M/HNC). They reported that the ammonium adsorption data on the NP material agreed with the Langmuir isotherm models. Isotherm parameters in terms of removal efficiency are presented in Table 5.

In terms of removal efficiency and zero-valued iron concentration, it was observed that it fits the Freundlich isotherm (Figure 4(b)). The size of q_max and K_L in Langmuir isotherm indicates high adsorption capacity (Table 5). When all these results were examined, Langmuir and Freundlich isotherm equations are created upon ammonium removal and with increasing adsorbent concentrations. These results showed that the optimum condition for adsorption capacity and isotherms was 50 mg/L nZVI (R² > 0.9). Mittal (2006) used chicken feathers as an adsorbent for the removal of toxic substances containing dye waters. Additionally, it was reported that Langmuir-Freundlich fits adsorption isotherm models. According to our study, similar results were obtained in terms of both removal efficiency and adsorption isotherms. In another study, the adsorption capacities of As(III) and As(V) were 35.83 and 29.04 mg g⁻¹, respectively, as determined from Langmuir adsorption isotherms in batch experiments (Wang et al. 2014). When compared with our study, it is seen that the adsorption capacity is high. Similarly, in another study, lead and cadmium adsorption efficiency from leachate by the natural zeolite clinoptilolite was investigated, and they found that best isotherm model for lead adsorption fits the Freundlich model and that for cadmium fits the Langmuir model (Faraji et al. 2012).

Considering the adsorbent concentration and DOC removal efficiency, the R² value is very close to 1 in the pseudo-second-order adsorption kinetic equation (R² = 0.9853, 400 mg/L Fe⁰) (Table 6). However, DOC was complied with pseudo-second-order adsorption kinetics (Figure 5(b) and Table 6). These results show that pseudo-second-order kinetics are adsorbed onto the nZVI surface via chemical interaction. Foul et al. (2009) investigated the removal efficiency of LFL using two different adsorbents (AC and limestone). Adsorption kinetics, on the other hand, determined their suitability to the pseudo-second-order kinetic model. Similarly, Boparai et al. (2011) used nZVI particles to investigate the removal of Cd²⁺ (25–450 mg L⁻¹⁾. They observed that the adsorption kinetics was well adapted using a pseudo-second-order kinetic model. In another study, they aimed to achieve both anionic and cationic dye removal efficiency by using a zero-valent iron-loaded composite NP material. The adsorption process of both dyes was well fitted with the Langmuir isotherm model and pseudo-second-order kinetic model (Eltaweil et al. 2021).

Yoon et al. (2014) reported in their study that the pseudo-second-order model and Freundlich isotherms were the most appropriate model. Sun et al. (2020) reported in their study that the DOC, ⁠, and Total Phosphorus (TP) removal efficiencies using electrodes from LFL were 72, 98, and 98%, respectively. They also reported fit to pseudo-first-order and pseudo-second-order kinetic models for DOC, ⁠, and TP pollutants. It is similar to our study. Mortazavian et al. (2018) reported in their study that it was suitable for pseudo-second-order kinetics and Freundlich isotherm for both AC and AC/nZVI adsorbents. In another study, they reported that the adsorption of Cr(VI) on AC/Fe₂O₃ composite conformed to the Langmuir isotherm model (R² = 0.98) and pseudo-second-order kinetic model (R² = 0.99) (Sahu et al. 2023). In another study, adsorption of nZVI reported that it was in accordance with second-order kinetic equations and Langmuir isotherm (Delnavaz & Kazemimofrad 2020). Khadim et al. (2022) in their study, the adsorption experiment continued for 1 h, while in our study, it was allowed to continue for 5.5 h. Mahdi et al. (2023) reported that their study was similar to our study in terms of kinetics and isotherm, but they used different adsorbents as mesoporous material MCM-48. As a result, the adsorption process has been widely used in the removal of organic and inorganic substances. For this reason, it has been reported that it occurs by adsorption to the surface of a highly porous structure through physical and chemical bonds (Foo & Hameed 2009). NPs, which are easy to operate, widely available, and have low synthesis costs, have recently been used in adsorption processes. One of the main advantages of iron and iron-based metal oxide compounds is that they do not lose their magnetic properties during the thermal regeneration process (Peng et al. 2006). It has been reported that treatment of LFL using the adsorption process gives successful results, especially in the removal of organic compounds and ammonia nitrogen (Foo & Hameed 2010).

nZVI can be used as an effective adsorbent for removing COD, DOC, ⁠, and from LFL. Equilibrium isotherms in this study were analyzed using Langmuir and Freundlich models. Kinetic data were obtained and analyzed using pseudo-first-order and pseudo-second-order equations. Optimum conditions were determined as pH 8, reaction time 120 min, and adsorbent dose 50 mg/L nZVI. COD, DOC, ⁠, and removal efficiency was approximately 75, 60, 57, and 33%, respectively. DOC, ⁠, and adsorption capacities, respectively, were 29.62, 21.01, and 3.12 mg/g. The experimental results were applied to isotherms commonly used in LFL, and the most suitable isotherm was found to be Langmuir. As a result of administering the experimental data to the kinetic equation, the mechanism controlling the rate was determined, and the most suitable model was found to be the pseudo-second-order equation. Sorption kinetics is investigated to develop an understanding of controlling reaction pathways and the mechanisms of sorption reactions. In the near future, NPs may turn out to be the essential and indispensable components of water and wastewater treatment systems and facilities. The use of nanoparticulate substances, especially nZVI, can partially reduce the concentration of pollutants to meet the final discharge. Therefore, natural nZVI adsorbent may be a suitable alternative in the significant treatment of LFL and pre-treatment or post-treatment. This study demonstrated that nZVI could be applied as adsorbents for pollutant treatment from aqueous solutions.

This article was supported by the Scientific Research Unit of the Çukurova University. Project No: FDK-2019-11782.

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

The authors declare there is no conflict.