Adsorption of cupric, cadmium and cobalt ions from the aqueous stream using the composite of iron(II,III) oxide and zeolitic imidazole framework-8

With rapid industrialization and urbanization, partially treated/untreated water discharges have considerably polluted freshwater resources. These contaminated waters alter the surface water quality by releasing toxic, pathogenic substances, organic and inorganic materials, toxic metals (Zhang et al. 2021), pharmaceuticals and personal care products, fertilizers and radioactive compounds, and so on, all of which have direct and indirect adverse impacts on human and marine life (Arif 2020; Bangari et al. 2021). Among various organic and inorganic pollutants, potentially toxic metal ions are of high threat (Tchounwou et al. 2012; Masindi & Muedi 2018). Municipal waste (Ishchenko 2019), fertilizers (Cheraghi et al. 2013), fossil fuel (Kamran et al. 2013), mining and smelting of metallic ferrous ores (Fashola et al. 2016), pesticides (Defarge et al. 2018), sewage sludge (Shamuyarira & Gumbo 2014), battery processing (Khan et al. 2020), textile printing (Velusamy et al. 2021), paper industries (Reza et al. 2013) and leather industries (Abbas et al. 2012) are the primary contributors to potentially toxic metals in freshwater resources. Potentially toxic metals such as Cu, Zn, Hg, Ni, Pb, Cd, As and Cr are non-biodegradable and accumulate in living beings (Fu & Wang 2011; Chowdhury et al. 2018, 2021; Gupta et al. 2021a). Potentially toxic metals have long term health effects, including kidney damage, bone damage, fracture, skin cancer, neurological damage, skin lesions such as hyperkeratosis and pigmentation change, lung cancer and neurotoxic effects (Mishra et al. 2019). As a result, they must be eliminated from water to protect public health.

This study considered three potentially toxic metals at high concentrations, namely copper, cobalt, and cadmium. Copper is a toxic metal with adverse health and metabolic effects on living things and can enter the environment through natural and anthropogenic sources (Gandhi et al. 2010). Health effects of copper include mucosal irritation, capillary damage, hepatic, renal, and lung cancer (Briffa et al. 2020). Cobalt is widely spread in the environment and can be inhaled, ingested, or absorbed via the skin. Cadmium is a harmful trace metal found in low concentrations in most rocks, coal, and oil. In low levels, cadmium causes coughing, headaches, nausea, and vomiting; it accumulates in the liver and kidneys in high doses, causing severe bone issues, renal failure, and hypertension. Chronic anaemia can arise from consuming water contaminated with Cd (Burke et al. 2016). As per the guideline values for drinking water, Cu, Co and Cd concentrations should not exceed 2 mg L⁻¹, 0.05 mg L⁻¹ and 0.003 mg L⁻¹, respectively (Sayato 1989)_.

Several technologies exist for the removal of potentially toxic metal ions, such as chemical precipitation (Pohl 2020), ion exchange (Al-Enezi et al. 2004; Dabrowski et al. 2004), membrane filtration (Qdais & Moussa 2004; Khulbe & Matsuura 2018), adsorption (Gupta et al. 2021a), electrochemical treatment technologies (Gupta et al. 2012), to name a few. Among all these, the most convenient, cost-efficient and scalable technology for removing potentially toxic metals is adsorption (Park et al. 2016; Qasem et al. 2021). Activated carbon (Abdulrazak et al. 2017), fly ash (Nguyen et al. 2018), metal oxides (Ming et al. 2012), zeolites (Elboughdiri), biomass (Nemeş & Bulgariu 2016), and carbon nanotubes (Tran et al. 2017) are the most widely used adsorbents for the removal of potentially toxic metal ions. MOFs are a type of porous crystalline material made by combining metal ions with organic linker molecules that have shown to be excellent adsorbents due to their large surface areas, consistent pore size distributions, and tunable pore sizes (Song et al. 2012). MOFs have recently been indicated to be efficient at adsorbing toxic contaminants, including lethal metal ions and organic compounds, from an aqueous stream (Chen et al. 2020; Yadav & Indurkar 2021). Zeolitic imidazole frameworks (ZIFs), a crystalline porous framework family member, have frequently been used to adsorb potentially toxic metal. ZIF-8 is a network molecule that is generated by coordination bonding by zinc and 2-methylimidazole (Hmim). It has been broadly examined in many fields such as H₂ storage, CO₂ capture, gas separations, and photocatalysis (Lee et al. 2015).

The magnetic Fe₃O₄ is chosen to overcome separation after the adsorption experiment. Due to the magnetic Fe₃O₄, the composite becomes magnetic, which solves the issue of unloading adsorbent post-experiment with the assistance of an external magnetic field (Jiang et al. 2021). Few researchers have studied adsorption with the composite of Fe₃O₄ and ZIF-8 (Fe₃O₄@ZIF-8) for removal of potentially toxic metals such as As(III) (Huo et al. 2018), U(IV) and Eu(III) (Wu et al. 2019), Pb(II) and Cu(II) (Jiang et al. 2021).

In this study, we synthesized Fe₃O₄@ZIF-8 using a two-step method. XRD, SEM, TGA, XPS and FTIR were used to characterize the synthesized samples. For the first time, we report removal of potentially toxic metals ions (Cu²⁺, Co²⁺ and Cd²⁺) through adsorption using Fe₃O₄@ZIF-8 composites. The effect of adsorbent dose, contact time, pH and concentration influence on the adsorption performance of Fe₃O₄@ZIF-8 for the removal of Cu²⁺, Co²⁺ and Cd²⁺ was studied. To further study the adsorption process, kinetic and thermodynamic studies were also carried out. Regeneration of the composite is done to study its reusable capability for up to three cycles. The effect of counter-ions on the adsorption capacity of Fe₃O₄@ZIF-8 was also investigated.

Ferric chloride, ethylene glycol, sodium citrate, sodium acetate (NaAc), zinc nitrate hexahydrate and 2-methylimidazole were purchased from Rankem chemicals. Ethylene glycol and ethanol were purchased from Sigma Aldrich. Copper chloride, cobalt chloride, and cadmium chloride salts were purchased from Spectrum Chemicals Manufacturer Corporation. All the reagents were of analytical grade and used without any additional purification. During all experiments, deionized (ultrapure: 18.2 MΩ cm) water was used.

The sample's surface morphology was studied using a scanning electron microscope (JEOL JSM 7100F). The polydispersity index and crystal size of the resultant micrographs were manually analyzed using the software ImageJ. The surface area of the adsorbent was assessed from a surface area analyzer (ASAP 2020, Micromeritics, USA). A Eutech PC2700 multi-parameter device was used to quantify the pH. Perkin Elmer Fourier transform spectroscopy (FTIR) spectrometer (USA) was used to perform an FTIR spectral study of the adsorbent samples. The zeta potential and particle size distribution of PTS in the aqueous phase was determined using a Zetasizer Nano-S (Malvern Instruments Ltd, Malvern, UK). A Philips X'pert MPD system was used to analyze the samples for X-ray diffraction with Ni-filtered Cu Ka radiation, k = 1.5404 Å. The metal ion concentrations before and after adsorption were measured using a UV-Vis spectrophotometer (Shimadzu, Japan) by building a calibration curve.

The solvothermal technique was used for synthesized Fe₃O₄ nanoparticles (Wu et al. 2015). FeCl₃.6H₂O (0.8 g), sodium acetate (1.2 g), sodium oleate (1.83 g) and ethylene glycol (25 mL) were added to a beaker and stirred on a magnetic stirrer for half an hour at room temperature. The resultant solution was placed in a Teflon autoclave with a capacity of 50 mL. The autoclave was then placed in a 200 °C oven for 10 hours. The autoclave was turned off and allowed to cool to ambient temperature. The obtained black precipitate (Fe₃O₄) was washed with DI water and ethanol. Fe₃O₄ particles were dried in a vacuum oven at 50 °C for 24 hours before use.

The synthesis procedure of Fe₃O₄@ZIF-8 is illustrated in Figure S1 (Supplementary Information). Fe₃O₄ (0.5 g) and Zn(NO₃)₂.6H₂O (0.6 g) were added to 50 mL methanol and sonicated for 20 minutes. In a separated beaker, 5 mL DI water, 2-methylimidazole (11.5 g) was dissolved. Both the solutions were mixed and vigorously stirred for 10 minutes to reach the shell growth of ZIF-8. The resulting composite was retrieved by applying a magnetic field through a strong magnet, then washed with DI water and ethanol. The composite was dried in a vacuum oven at 50 °C for 24 hours before use.

The adsorption experiments were conducted to determine the requisite condition for maximum removal of Cu²⁺, Co²⁺ and Cd²⁺ from water using Fe₃O₄@ZIF-8 as adsorbent. The batch adsorption mode was adopted to evaluate the performance of Fe₃O₄@ZIF-8 with simulated solutions with fixed experimental conditions. Stock solutions for removing Cu²⁺, Co²⁺ and Cd²⁺ were prepared from the salt of copper chloride, cobalt chloride and cadmium chloride, respectively, in DI water (electrical conductivity: 0.054 μS cm⁻¹, resistivity: 18.2 MΩ). A predetermined amount of adsorbent was added to Cu²⁺, Co²⁺ and Cd²⁺ solution separately of known concentration. The samples were collected at a fixed time interval and filtered before analyzing the residual Cu²⁺, Co²⁺ and Cd²⁺. The optimum dose was selected for further study under variable control parameters such as contact time, pH, feed concentration, and the presence of co-ions. To understand the effect of coexisting ions, cations (K⁺, Na⁺ and Ca⁺) and anions (Cl⁻, NO³⁻ and SO₄²⁻) solutions (10 mg L⁻¹) of KCl, NaNO₃, CaCl₂, Na₂SO₄ were prepared. The variations in all four parameters were considered to find a suitable environment for removing Cu²⁺, Co²⁺ and Cd²⁺. Therefore, the parameters were varied as contact time (5, 15, 30, 45, 60, 90, 120, 150, 180, 240 and 300 minutes), pH (2, 3, 4, 5, 6 and 7), adsorbent dose (0.1, 0.2, 0.4, 0.5, 0.8, 1, 1.5 and 2 g) and feed concentration (5, 10, 20, 30, 40, 60, 80 and 100 mg L⁻¹). At room temperature (28 °C), all adsorption tests were conducted. Each experiment was performed in triplicate to ascertain the accuracy of the results. Further, the adsorbent was also tested for desorption and regeneration to verify the long-term reusability. This study used a 0.2 M NaOH solution for regeneration purposes (Ke et al. 2017). Because of this, we were able to regenerate the highly effective Fe₃O₄@ZIF-8 many times by ultrasonically cleaning the post-adsorption adsorbent with NaOH. After that, the adsorbent was dried in a vacuum oven at 200 °C overnight. Using a spectrophotometric method, the removal efficiency of the adsorbent was determined before and after regeneration. A UV-Vis spectrophotometer was used to determine concentrations of Cu²⁺, Co²⁺ and Cd²⁺; calibration was done using various stock solutions concentrations (0, 1, 5, 10, 15, 20 and 30 ppm).

q_e and q_t amount of M²⁺ ions adsorbed at the adsorption equilibrium and at time t, respectively, k₁ and k₂ rate constant for pseudo-first-order and pseudo-second-order.

q_eadsorption capacity at equilibrium (mg g⁻¹)

b Langmuir isotherm constant (L mg⁻¹), q_m maximum adsorption capacity (mg g⁻¹), C_eequilibrium concentration of potentially toxic metals in solution (mg L⁻¹).

K_f and n Freundlich constants represent coefficient and intensity, respectively.

FE-SEM images were used to study the morphological properties of the prepared Fe₃O₄ and ZIF-8 particles, shown in Figure 1. The Fe₃O₄ particles were spherical, and their surface was rough. Moreover, particles were in the form of agglomerates. The magnetic dipole moment interaction between the Fe₃O₄ particles was responsible for such an aggregate formation (Raghava Reddy et al. 2019). Large rhombic dodecahedron-shaped crystals with a wide range of crystal sizes in the range of 60–240 nm can be seen in SEM images ZIF-8 (Figure 1(b)). The SEM image of the Fe₃O₄@ZIF-8 confirmed the effective encapsulation of ZIF-8 into Fe₃O₄. These images of Fe₃O₄@ZIF-8 composites (Figure 2) indicate that the structure of both the particles remained unchanged and was found to be comparable with pristine ZIF-8 (Schejn et al. 2015). To describe the uniformity of the particle size distribution and the term polydispersity index (PDI) is used. From the histogram (Figures 1 and 2), the particle distribution of Fe₃O₄, ZIF-8 and Fe₃O₄@ZIF-8 can be seen. The mean particle size of Fe₃O₄, ZIF-8 and Fe₃O₄@ZIF-8 was 2.5 μm, 0.18 μm and 2.80 μm. The standard deviation of particle size of Fe₃O₄, ZIF-8 and Fe₃O₄@ZIF-8 was found to be 0.99, 0.04 and 0.70. From the mean particle diameter and standard deviation values, the PDI of Fe₃O₄, ZIF-8 and Fe₃O₄@ZIF-8 was calculated to be 0.039, 0.012 and 0.016. The values of PDI of Fe₃O₄, ZIF-8 and Fe₃O₄@ZIF-8 are less than 0.1, establishing the monodispersive nature of these particles (Danaei et al. 2018). The Brunauer–Emmett–Teller (BET) method was used to determine the surface area of the synthesized ZIF-8 and Fe₃O₄@ZIF-8. The surface area of ZIF-8 and Fe₃O₄@ZIF-8 was 71.6 and 82.7 m² g⁻¹, respectively. The surface area of Fe₃O₄@ZIF-8 rose when ZIF-8 was added onto the surface. Fe₃O₄@ZIF-8 was able to offer more active sites for adsorption of potentially toxic metals as a result of this phenomenon.

For the as-synthesized magnetic Fe₃O₄ nanoparticles, six characteristic diffraction peaks at diffraction angles 2θ ∼ 30.2°, 35.3°, 43.2°, 53.8°, 57.0° and 62.8° corresponding to planes (220), (311), (400), (422), (511) and (440), respectively, were observed (Figure 3). This revealed that the Fe₃O₄ particles were pure and crystalline and matched well with ICDD: 00-001-1111 (Shagholani et al. 2015; Chaves et al. 2021). In the case of ZIF-8, the XRD peaks at 2θ = 7.4°, 10.4°, 12.7°, 14.7°, 16.5°, 18.0°, 22.3°, 24.5°, 26.8° and 29.8° were observed, which corresponded to planes (011), (002), (112), (022), (013), (222), (114), (233), (134) and (044), respectively. The noticeable reflections were in line with the previous results (Zhao et al. 2014; Nordin et al. 2017), conforming to the typical sodalite structure of ZIF-8. Furthermore, the diffraction patterns of Fe₃O₄@ZIF-8 demonstrated in Figure 3 illustrate that the phase composition of Fe₃O₄@ZIF-8 indeed consists of Fe₃O₄ and ZIF-8. The synthesized Fe₃O₄@ZIF-8 composites show identical peaks with ZIF-8 and Fe₃O₄, indicating that their crystal structure has not changed despite the adjustments that have been created in their compositions.

Figure S2 (Supplementary Information) represents the FTIR spectra of synthesized Fe₃O₄ particles. The characteristic absorption band of Fe₃O₄ at wavelength 580 cm⁻¹ can be accredited to the Fe-O bond stretching vibrations. Another prominent adsorption band at 3,398 cm⁻¹ was because of -OH group stretching vibrations (Ozkaya et al. 2009). Figure 4 shows the FTIR spectra of synthesized ZIF-8 and Fe₃O₄@ZIF-8 composite. The FTIR spectra of ZIF-8 revealed a band in the range 3,200–3,600 cm⁻¹ corresponding to N-H and O-H (Zhang et al. 2016). The broad, strong band in the region 2,500–3,000 cm⁻¹ represents the C-H, N-H and O-H stretching vibrations of methyl, hydroxyl and amine group in ZIF-8. The bands at 1,849 cm⁻¹ and 2,300–2,500 cm⁻¹ represent the C = N-H stretching vibration. The bands at 1,681 cm⁻¹ and 1,591 cm⁻¹ were attributed to N-H's bending and stretching vibration in the imidazole ring. The band between 1,350 and 1,500 cm⁻¹ was due to the ring stretching, and 421 cm⁻¹ was due to the stretching vibration of Zn-H. The characteristic band of Fe₃O₄ at 580 cm⁻¹ represents the Fe-O bond, and the small band at 421 cm⁻¹ representing the Zn-H bonding from ZIF-8 can be observed. The imidazole ring's in-plane stretching and bending vibrations were attributed to the band at 1,700–1,500 cm⁻¹ and 500–1,350 cm⁻¹. The stretching vibrations of N-H of the imidazole ring are visible in the band between 3,300 and 3,450 cm⁻¹. The absorption band at 1,584 cm⁻¹ shows the C = N stretching vibration of the imidazole ring, while the band at 2,929 cm⁻¹ shows the aromatic stretching of the aromatic ring (Zhang Zhang et al. 2013).

The thermographs of synthesized ZIF-8 and Fe₃O₄@ZIF-8 are illustrated in Figure S3. Starting at 30 °C and rising to 60 °C, the TGA plot of the ZIF-8 revealed a 98% residual mass (2% weight loss). The decreased residual mass was attributed to the existence of guest molecules or residual solvent. ZIF-8 decomposed between 300 and 800 °C. At 400 °C, the ZIF-8's weight loss slowed dramatically. The TGA plots proved the thermal stability of ZIF-8 up to 300 °C, which is in line with earlier studies (Awadallah et al. 2019). The TGA curves for Fe₃O₄@ZIF-8 showed only a mild weight loss of less than 0.5% from 20 to 250 °C, corresponding to the elimination of guest molecules (Hmim or methanol). After creating the guest-free phase Zn(mim)₂, a long plateau was seen until 300 °C, indicating that the Fe₃O₄@ZIF-8 composites have good thermal stability. Since no weight loss was observed until 450 °C, it is evident that the unreacted species or solvents have escaped the intra-crystalline cavities. When the temperature was raised from 500 to 650 °C, all nanocrystals lost a significant amount of weight, showing that ZIF-8 was thermally decomposed in that range (Pan et al. 2011). In a nutshell, it can be deduced that the Fe₃O₄@ZIF-8 composite is thermally more stable than the precursor ZIF-8 itself.

Figure 5 shows the removal efficiency for Cu²⁺, Co²⁺ and Cd²⁺ removal with respect to time. The effect of contact time on the Fe₃O₄@ZIF-8 adsorption characteristic was investigated in a solution at neutral pH at 28 °C. The removal efficiency of Cu²⁺ by Fe₃O₄@ZIF-8 reached 85% in 30 minutes and 95% in 2 hours (Figure 5(a)). The high Fe₃O₄@ZIF-8 adsorption rate was due to the high specific ZIF-8 surface area (Yamamoto et al. 2013). It is evident from Figure 4(a) that the contact time of 2 hours ensured the adsorption of Cu²⁺ ions was completed. The consumption of Co²⁺ by Fe₃O₄@ZIF-8 reached 80% in the first 30 minutes and 95% in 2 hours and remained constant afterwards (Figure 5(b)). The removal efficiency of Cd²⁺ by Fe₃O₄@ZIF-8 reached 85% in 30 minutes and 95% in 1.5 hours and remained constant later Figure 5(c). From this, it can be inferred that 2 hours was the optimum contact time for the complete adsorption of potentially toxic metals Cu²⁺, Co²⁺ and Cd²⁺ on the adsorbent surface.

Figure 6 shows the effect of pH on the removal efficiencies for removing Cu²⁺, Co²⁺ and Cd²⁺. As the pH of the aqueous solution affects the adsorbent and the ionic conditions and influences the state of the metal ions, the pH of the solution of a potentially toxic metal has a significant impact. Potentially toxic metal ions change their oxidation state along with the pH variation. Under acidic conditions, Fe₃O₄@ZIF-8 was unstable and dissolved, which reduced its removal efficiency. Hence, the effects of the Fe₃O₄@ZIF-8 have been investigated in the 2.0–7.0 pH range. Figure 6(a) indicates that at pH 2.0–3.0, Fe₃O₄@ZIF-8 adsorbed Cu²⁺ scarcely in water, perhaps as the ZIF-8 shell of Fe₃O₄@ZIF-8 was unstable under acidic conditions. The adsorption of Cu²⁺ in the solution increased rapidly in the pH range of 3.0–4.0. In acidic solution, the stability of Fe₃O₄@ZIF-8 was analogous to that of pure ZIF-8 in the acidic solution; therefore, Fe₃O₄@ZIF-8 may be utilized to uptake Cu²⁺ at a pH of more than 3.0. At a pH range of 4.0–7.0, the uptake capabilities did not increase with pH change, indicating that coordination reactions might drive adsorption (Zhang et al. 2016). For Co²⁺, as shown in Figure 6(b), a linear increase in the adsorption in the pH range of 2.0–6.0 became stable in the range of 6.0–7.0. The adsorption capacity of Cd²⁺ shows a similar type of trend as of Co²⁺ (Figure 6(c)).

At the initial stage of adsorption, the uptake capacity of the adsorbent increases with an increase in adsorbent dosage because the active adsorption sites increase. Figure 7 shows the effect of the dosage of Fe₃O₄@ZIF-8 on the adsorption capacity for metal ions (Cu²⁺, Co²⁺ and Cd²⁺). For Cu²⁺, the removal efficiency increases up to a dose of 1.5 g L⁻¹ of Fe₃O₄@ZIF-8. Up to 1.5 g L⁻¹ of adsorbent, the residual percentage of Co²⁺ in water was only about 2%. Similarly, in the case of Cd²⁺ maximum removal efficiency is shown at 1.5 g L⁻¹. After a particular dose, the constant removal efficiency was due to the accumulation of adsorbent and limiting active adsorption sites.

15 mg Fe₃O₄@ZIF-8 was added to 10 mL of Cu²⁺ solutions with concentrations ranging from 5 to 100 mg L⁻¹ to evaluate the maximum adsorption capacity of Fe₃O₄@ZIF-8. As the initial concentration of Cu²⁺ rises, the amount of Cu²⁺ adsorbed per unit mass of Fe₃O₄@ZIF-8 (q, mg g⁻¹) increases, as illustrated in Figure 8(a). The adsorption capacity is a minimum when the feed concentration of the Cu²⁺ solution is less than 20 mg L⁻¹. On further increase in the adsorbate concentration, the removal adsorption capacity increases. Figure 8(b) represents the adsorption capacity of Co²⁺, which shows the adsorption capacity is maximum when the solution concentration is more than 80 mg L⁻¹. Similarly, in Figure 8(c), the adsorption capacity is highest for concentrations more than 90 mg L⁻¹. From the adsorption capacity data, it can be predicted that a feed solution of concentration more than 80 mg L⁻¹ is appropriate to study adsorption.

The adsorption capacities of different MOFs and related adsorbent materials used for adsorption of potentially toxic ions (Cu²⁺, Cd²⁺ and Co²⁺) are listed in Table S1 (Supplementary Information). The adsorption capacity, q_e (mg g⁻¹), measured in this work for Fe₃O₄@ZIF-8, was in good agreement with available literature.

To further confirm the successful synthesis of Fe₃O₄@ZIF-8, XPS examination was employed. As demonstrated in Figure S4, the XPS analysis of Fe₃O₄@ZIF-8 showed characteristic peaks of C1s (285.8 eV), N1s (401.98 eV), which belongs to C = N and C-N bonds, O-1s (531.35 eV), which show the presence of O in the form of Zn-O, and has two Fe peaks, viz Fe-2p3/2 and Fe-2p1/2 oxidation states at 712.5 and 726.0, which also belong to the Fe-O bond in Fe₃O₄ and Zn (1,021.58 eV), which denoted that the imidazole group and Zn ions exist in the adsorbent. The Zn-2p shows two strong characteristic peaks at 1,021.58 and 1,044.5 eV, which indicate the Zn is present in the +2 oxidation state (Liu Liu et al. 2017).

Figure 9 demonstrates the XPS spectra of the composite before and after adsorption of Cu²⁺, Co²⁺, and Cd²⁺. Figure 9(a) shows the binding energy peaks of potentially toxic metals with Zn-2p, at Zn-2p_3/2-Cu(1,022 eV), Zn-2p_3/2-Co(1,022.3 eV), Zn-2p_3/2-Cd(1,022.5 eV) and for Zn-2p_1/2, Zn-2p_1/2-Cu (1,044 eV), Zn-2p_1/2-Co(1,045.5 eV), Zn-2p_1/2-Cd(1,045.1 eV). The binding energy of the ZIF-8's original coordination centre Zn ions did not appear to have changed in any noticeable way after adsorption, as Zn ions remained unchanged. The strength of the binding energy, on the other hand, was diminished. In this case, it is possible that the newly introduced ions competed with Zn ions to form bonds, resulting in a drop in the binding energy strength. Figure 9(b) demonstrate the binding energies of N-1s with potentially toxic metals are N-1s-Cu (398.99 eV), N-1s-Co (399.6 eV) and N-1s-Cd (399.2 eV), which vary from the binding energy of N-1s (398.9 eV). The variation in N-M²⁺ (M = Cu, Cd, Co) binding shows bond formation with the N of the imidazole ring in the ZIF-8 shell of Fe₃O₄@ZIF-8. The binding energies of C-1s with potentially toxic metals are C-1s-Cu (284.8 eV), C-1s-Co (284.7 eV) and C-1s-Cd (284.8 eV), which are shown in Figure 9(c). In this case, the binding energy of C with potentially toxic metal ions was similar, but with different peak area and intensity. This signifies that the composite structure was partially preserved. However, the decrease in the intensity of binding energy showed interference of ions in the molecule. The binding energies of O-1s, in Figure 9(d) with potentially toxic metals are O-1s-Cu (530.35 eV), O-1s-Co (532.45 eV) and O-1s-Cd (532.18 eV) and show the formation of new bonds with metal as Zn-O-Cu, Zn-O-Co and Zn-O-Cd. The binding energy changes due to the formation of a coordinate bond with N, C, and O. The change in binding energy shows the successful capture of potentially toxic metal on the surface of the adsorbent through binding with adsorbent atoms like C, O and N.

Table 1 lists theoretically computed values of adsorption capacities at equilibrium (q_e) and kinetic coefficients. The PFO model's nonlinear coefficient of determination (R²) was poor compared to the second-order model. The high k₂ value for pseudo-second-order illustrated the increase in adsorption rate. The low k₁ value indicated that the adsorption process in the PFO is slow. The adsorption of Cu²⁺, Co²⁺ and Cd²⁺ metal ions on Fe₃O₄@ZIF-8 does not appear to be a first-order reaction, based on these findings and the curve fitting as shown in Figure S5, since the regression coefficient and adsorption capacities are comparatively low in this case. The R² for Cu²⁺, Co²⁺ and Cd²⁺ give greater values. The experimental and theoretical q_e values are in good agreement, as shown by the results of the PSO model. The low value for the residual sum of the square represents the minimum difference between the experimental and theoretical values for PSO than PFO reaction (Moussout et al. 2018). These findings imply that the adsorption of Cd²⁺, Co²⁺, and Cu²⁺ ions on Fe₃O₄@ZIF-8 in single and binary systems follows a second-order kinetic mechanism.

Figure S6 and Table 2 show that the Freundlich model is not suitable for studying the adsorption isotherm for Cu²⁺, Co²⁺ and Cd²⁺. The Langmuir model appears to be suitable for modelling Cu²⁺, Co²⁺ and Cd²⁺ adsorption onto Fe₃O₄@ZIF-8, based on the R² value. When the Freundlich isotherm model and the Langmuir isotherm model were compared, the R² value was observed to be far from unity. The maximum Fe₃O₄@ZIF-8 elimination capacity from the associated isotherm model is 106.7 mg g⁻¹, close to the experiment results. The value for the residual sum of the square in the Langmuir isotherm is comparatively less than the value in the case of the Freundlich isotherm, which also favours the Langmuir since the lesser the difference between theoretical value and experimental value the more it favours a particular isotherm. The conditional removal capacity can be calculated from the equilibrium adsorption isotherm, allowing comparisons between different adsorbents. Several factors influence removal capacity, including contact time, pH, competition ions, and ion concentrations. The latter substantially impacts adsorption capacity since a higher adsorbate concentration implies a more significant driving force and, therefore, a better interaction process (Lou et al. 2018).

Different ions are also present in water with potentially toxic metals; therefore, it is necessary to study the impact of these ions on the adsorption capacity of Fe₃O₄@ZIF-8. To study the effect of coexisting ions on the adsorption capacity of Fe₃O₄@ZIF-8, cations (K⁺, Na⁺ and Ca²⁺) and anions (Cl⁻, NO³⁻ and SO₄²⁻) were separately added to the Cu²⁺, Co²⁺ and Cd²⁺ solution. The uptake capacity of Cu²⁺, Co²⁺ and Cd²⁺ solution by Fe₃O₄@ZIF-8 was slightly affected by Cl⁻, NO³⁻ and SO₄²⁻ because these anions have a modest electrostatic attraction (Cl⁻, NO³⁻ and SO₄²⁻) and Cu²⁺, Co²⁺ and Cd²⁺. The adsorption capacity of Fe₃O₄@ZIF-8 was not much affected by the cation (K⁺, Na⁺ and Ca²⁺). It was found that there is no competition between these cations (K⁺, Na⁺ and Ca²⁺) and the metals ions (Cu²⁺, Co²⁺ and Cd²⁺) for the adsorption sites on the adsorbent Fe₃O₄@ZIF-8. Therefore, it can be concluded that the Fe₃O₄@ZIF-8 has great potential and selectivity for capturing Cu²⁺, Co²⁺ and Cd²⁺ ions from water even in the presence of competitive ions.

The removal efficiency for the metal ions was assessed for three regenerative cycles (Figure S7). Cu²⁺ had removal effectiveness of about 99% after the first cycle, which fell gradually to 94.1, and 87% in the second and third cycles. Similar trends were observed for Co²⁺ and Cd²⁺. In the case of Co²⁺, the removal efficiencies were 99.2%, 95.1, and 89% for the respective three regenerative cycles. For Cd²⁺, the removal efficiencies were 99.4%, 94.3 and 88% for the respective three regenerative cycles. Hence, Fe₃O₄@ZIF-8 exhibited superior regenerative properties for Cu²⁺, Co²⁺ and Cd²⁺ ions from water. The significant loss in the removal efficiencies was due to the reduction in the number of active adsorption sites (Jiang et al. 2021).

To determine the adsorption mechanism of Fe₃O₄@ZIF-8, it is necessary to consider ion exchange and coordination reactions. Since the Fe₃O₄@ZIF-8 is negatively charged (zeta potential -1.52 mV) and the metal ions are positively charged, the ion exchange and coordination reaction could be explained. In the process of Mn⁺ adsorption in Fe₃O₄@ZIF-8 (Cu²⁺, Co²⁺, and Cd²⁺), Mn⁺ replaces Zn²⁺ to form a new complex with Zn²⁺. The following are the reasons for forming the new complex: first, because Cu²⁺, Co²⁺, and Cd²⁺ have atomic radii that are comparable to those of Zn²⁺, they may be easily substituted by Zn²⁺. Secondly, Zn²⁺ has a better coordination ability than the three metal ions Cu²⁺, Co²⁺ and Cd²⁺ (Jiang et al. 2021). This is because the valence e⁻ layer structure of Zn²⁺ is more stable than the valence e⁻ layer structure of the three metal ions Cu²⁺, Co²⁺ and Cd²⁺. Because of the coordination between Cu²⁺, Co²⁺ and Cd²⁺ and the N atom on 2-methylimidazole, the adsorption on Fe₃O₄@ZIF-8 continues to increase per unit mass after Zn²⁺ has been replaced by Cu²⁺, Co²⁺ and Cd²⁺.

The Fe₃O₄@ZIF-8 nanocomposite was synthesized in this study, and it had a larger absorption capacity and faster adsorption kinetics for Cu²⁺, Cd²⁺ and Co²⁺ from an aqueous stream. The magnetic Fe₃O₄ nanoparticle was prepared through the solvothermal method, and Fe₃O₄@ZIF-8 was synthesized by hydrothermal method and further characterized by SEM, XRD, XPS, FTIR and TGA. Characterization of both nanoparticles and composite showed the successful synthesis of material without impurities. The magnetic Fe₃O₄@ZIF-8 composite facilitated the easy loading and unloading of the adsorbent that too with the increased removal efficiency. The adsorption process obeyed the Langmuir model, and the adsorption capacities of Fe₃O₄@ZIF-8 were calculated to be 46.82 mg g⁻¹, 71.29 mg g⁻¹ and 54.49 mg g⁻¹ for Cu²⁺, Co²⁺ and Cd²⁺, respectively. The PSO model gave the best-fit curve from the adsorption kinetics. Exchange and coordination bond formation could be the main mechanisms responsible for removing Cu²⁺, Cd²⁺ and Co²⁺ by Fe₃O₄@-ZIF-8 from water. The uptake capacity of Cu²⁺, Co²⁺ and Cd²⁺ solution by Fe₃O₄@ZIF-8 were not significantly affected by the presence of competitive cations and anions. The material also exhibited superior regenerative properties for Cu²⁺, Co²⁺ and Cd²⁺ ions from the water up to three cycles. Due to these merits, the current study shows that Fe₃O₄@ZIF-8 is an excellent adsorbent for the uptake of Cu²⁺, Co²⁺ and Cd²⁺ in water containing potentially toxic metals.

The CSIR-CSMCRI PRIS number for this manuscript is 174/2021. The authors are grateful for partial funding support from the Council of Scientific and Industrial Research, India (MLP-0043). The authors also acknowledge AED&CIF, CSMCRI for providing instrumental facilities. The comments from anonymous reviewers and the editor have greatly improved the content.

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