Magnetic fly-ash geopolymer for heavy metal removal in aqueous solutions Open Access

Heavy metals in industrial wastewaters are of special concern since they can accumulate in different components of the environment (Nordberg et al. 2015). Removal of heavy metals from wastewater and contaminated surface and groundwater before they are discharged is desired for the protection of both human health and aquatic ecosystems (Waleeittikul et al. 2019). Conventional technologies for heavy metals removal from aqueous solutions involve physical–chemical treatments, such as precipitation, adsorption, membrane filtration, and ion exchange (Chotpantarat et al. 2011; Shipley et al. 2013; Ong 2018). Some challenges with these technologies include, for instance, generating a large amount of sludge from precipitation which would require disposal costs. Membrane system has high treatment cost owing to the high energy cost and frequent replacement of membranes. Adsorption is the most popular method among the above-mentioned technologies (Bereket et al. 1997) and may be an economical and effective treatment technology for industrial wastewaters. Adsorbents made from natural materials or waste products have always attracted widespread attention due to their low cost and sustainability properties as a renewal resource. Some examples of base materials used for the manufacture of adsorbents include locally available agriculture bio-based materials such as coconut shells and husk (Senanu et al. 2023), marine-based polymers such as chitosan (Romal & Ong 2023) or natural geomaterials such as zeolite (Motsi et al. 2009) or waste products such as sewage sludge ash (Pan et al. 2003). Suitable selection of the adsorbent base material is one of the factors to consider in developing a suitable adsorption system as a specific type of adsorbent may play a dominant role in removing specific pollutants in the adsorption process (Benettayeb et al. 2024).

In the last two decades, fly ash, a waste product from coal-fired power plants, has shown promise as an adsorbent base material. Fly ash can be used with other common chemical ingredients such as sodium hydroxide to produce geopolymer (Davidovits 1991). Geopolymer can be categorized as an economical and environmentally friendly material due to its low manufacturing cost (US$0.0001/g), its low-temperature manufacturing process and low CO₂ emission as compared to producing standard cementitious materials (Duxson et al. 2007; Liang et al. 2022). Interests in the applications of geopolymer were initially focused on producing cementitious materials for construction purposes but this interest has expanded to environmental applications such as adsorption and immobilization of heavy metals in wastes (Liang et al. 2022; Liu et al. 2023; Ren et al. 2024). Specific examples of heavy metals removed by geopolymers include copper removal from wastewater (Mužek et al. 2014; Al-Harahsheh et al. 2015) and zinc from water (Tang et al. 2023).

Researchers have studied the use of granular geopolymer particles (size: 0.6–2 mm) for environmental applications such as adsorption or precipitation of pollutants. The granular particles can be used in a column treatment system or as a media placed in the path of the waste streams. The use of fine geopolymer particles has some advantages in that they have large specific surface areas and can be easily dispersed to maximize contact with the pollutants in an aqueous solution, such as in a slurry form. However, this treatment suffers from some drawbacks such as the need to separate out the geopolymer, if it needs to be regenerated and, as in precipitation, can produce large quantities of residual sludge. To overcome these drawbacks, a magnetic signature can be applied to the geopolymer to generate a novel magnetic reactive/adsorbent material that can be separated from the aqueous phase. Rapid separation of the materials from the solution by using an external magnetic field without filtration or centrifugation, allows the magnetic materials to be recovered and reused for the removal and recovery of pollutants (Maleki et al. 2019; Hua et al. 2020; Rossatto et al. 2020; Maranhão et al. 2021; Salah et al. 2024). Maleki et al. 2019 used bentonite clay to make the geopolymer and embedded Fe₃O₄ nanoparticles to magnetize the materials. The prepared material was shown to have good removal efficiency for copper, lead and nickel. Rossatto et al. 2020 prepared geopolymer using metakaolin, rice husk and Fe₃O₄ and the material was found to be a good adsorbent for acid green 16 dye. Similar results were obtained by Hua et al. 2020 using a metakaolin/Fe₂O₃ geoploymer for the adsorption of acid green 16 dye. Maranhão et al. 2021 synthesized a geopolymer using metakaolin and maghemite nanoparticles and found the material to adsorb cadmium, chromium and lead. Salah et al. 2024 used a magnetitie-enriched copper ore and glauconite as the aluminosilicate precursor to prepare the geopolymer. The material was then used to demonstrate the adsorption capabilities of crystal violet dye. Buzukashvili et al. (2024) synthesized a magnetic material by binding magnetite nanoparticles to a zeolite material (prepared by high temperature fusion of coal fly ash) in a colloidal polyvinyl alcohol (PVA) solution. The material was found to adsorb various heavy materials and separated well using a low field magnet. However, this study used a waste material such as fly ash and non-magnetic iron oxides as the precursor materials for the preparation of the magnetic geopolymer. With a magnetic signature, the spent geopolymer can be easily separated magnetically and recovered and if needed the heavy metals can be desorbed and recovered and the magnetic geopolymer reused, offering further economic benefits as a low-cost adsorbent with a low greenhouse gas footprint. This material has the potential to compete with the more traditional and expensive chemical precipitation/sedimentation methods, adsorption processes using synthetic ion exchange or the use of energy-intensive membrane filtration systems.

The objective of this study is to produce magnetic geopolymer using two synthesis methods to incorporate different amounts of environmentally friendly magnetic Fe₃O₄ particles in fly ash-based geopolymer (FAG). This study is probably the first to incorporate economical and environmentally friendly magnetic Fe₃O₄ particles in fly-ash geopolymer. Based on the initial characterization of synthesized materials for their magnetic and physical–chemical properties, a synthesis method was selected to produce magnetic geopolymer for further investigation on its adsorption capacity for a heavy metal in wastewater and its separation properties in a magnetic field. This investigation is expected to provide further information on incorporating magnetic Fe₃O₄ particles in geopolymer to create novel magnetic geopolymer/Fe₃O₄ composite materials for various applications including environmental applications such as heavy metal removal and magnetic separation from wastewater.

Magnetic iron oxide particles (Fe₃O₄) were synthesized by using iron(III) nitrate nonahydrate (Fe(NO₃)₃·9H₂O, 98–102%, Fisher Chemical) as the starting material. A solution of 1 g·mL⁻¹ of iron nitrate was prepared by dissolving iron nitrate. Batches of magnetic iron oxide were prepared by placing 4 mL of iron nitrate solution in quartz crucibles and the solution was calcined at a temperature of 773 K for 20 min in a flowing forming gas with 95% Argon and 5% H₂. The samples were then cooled in the same gas stream inside the furnace to room temperature. About 0.6 g of pure magnetic Fe₃O₄ particles (size of 0.5–5 μm) were collected from each crucible. Preliminary calcining temperatures at 723, 773, and 823 K were tested and 773 K was found to be the preferred calcine temperature to get the highest saturation magnetization values.

The geopolymer was made of Class-F (as defined in ASTM C618-19 (2019)) fly ash and an alkaline activator solution. The alkaline activator consisted of sodium silicate solution (water glass, 3Na₂O·3SiO₂), solid NaOH, and water in a mass ratio of 0.49:0.11:0.40. The mass ratio of the alkaline activator to the fly ash and the activator was 0.33:1 (Ling et al. 2019; Shi 2019). The geopolymer was prepared by mixing the fly ash and the activator for 5–10 min and placing the paste in a mold to cure for 24 h at room temperature. It was then removed from the mold and cured for 7 days at 323 K. The geopolymer was then washed three times with water and crushed to the appropriate particle sizes. Particle sizes used for the preparation of magnetic FAG ranged from 0.6 to 2 mm.

Two methods were used in this study to prepare a novel magnetic geopolymer. The advantage of a magnetic geopolymer is that it can be magnetically separated easily in an aqueous environment.

The second method is an in-situ method for synthesizing magnetic fly ash geopolymer (identified as MFAG) and is illustrated in Figure 1(b). In this method, various amounts (3, 6, and 10 g) of prepared magnetic Fe₃O₄ were evenly mixed with the fly ash (30 g). The mixed materials were then reacted with 12 mL of alkaline solution (6 mL of sodium silicate solution +6 mL of 6.68 M of NaOH solution). The ratio of liquid to solid was fixed at about 0.4. The obtained paste was mixed for 5 min and was subjected to ultrasound for 5 min to enhance dispersion. The obtained paste mixture was poured into a 25 mm × 25 mm × 25 mm (1″ × 1″ × 1″) cubic mold and pre-cured for 24 h at room temperature (294–296 K). The samples were then de-molded, placed in a cylindrical glass container that was closed, and cured at a temperature of 773 K in an oven for 7 days. The MFAG cubes were crushed and grounded to obtain an MFAG sample with a particle size of less than 0.177 mm. The MFAG samples were washed several times with distilled water to remove surface alkalinity, and oven-dried overnight. A magnet was used to separate the MFAG and any non-magnetic fine particles. Table 2 presents the different samples prepared and the mixture proportions of the raw material.

Initial screening of all the samples was conducted by measuring the magnetic properties, the pH change in the aqueous solution and the single-point batch experiments of the samples for metal removal. The magnetic properties of the samples were measured at 300 K by a vibrating sample magnetometer (VSM) on a Quantum Design DynaCool physical properties measurement system (PPMS) with a maximum field of μ₀H = 3 or 9 T, which is high enough to saturate the samples.

The metal concentrations in the filtrate were determined by an inductively coupled plasma (ICP) atomic emission spectrometer (ICPE-9800 Series, Shimadzu Corporation, Kyoto, Japan).

Based on the initial screening, a magnetic FAG was selected for further testing. This geopolymer was characterized by its physical–chemical properties. The specific surface area of the magnetic FAG was determined using the nitrogen BET method. Compressive strength was determined by using the method from ASTM C109 (2016). SEM images were taken using a scanning electron microscope (Quanta FEG 250, Hillsboro, Oregon, USA).

Batch adsorption experiments were conducted using 50 mL of metal (Cu²⁺) solution and various masses of magnetic geopolymer powder in 60-mL centrifuge tubes and capped. The masses of MFAG ranged from 0.01 to 0.1 g and the initial Cu²⁺ concentrations ranged from 0 to 200 mg·L⁻¹. The initial pHs of the solutions for all the tests were adjusted to 3.0 ± 0.1 by adding 1 M of H₂SO₄ solution. The tubes were agitated on a rotary shaker at a speed of 30 rpm at room temperature (294–296 K) for 24 h. Preliminary experiments indicated that steady-state conditions were achieved within 24 h (data not shown). The solution was then filtered and the final pH of filtrate was measured by a pH meter and the final equilibrium concentration of the metal was analyzed using the ICP atomic emission spectrometer. All experiments were carried out in duplicate, and the average concentration was reported. The experimental data were fitted using two isotherm models, i.e., Langmuir and Freundlich models, to obtain the maximum uptake capacity.

Preliminary magnetic separation tests were conducted using 0.1 g of fine sample mixed with 50 mL copper solution for 5 min. The turbidity (Nephelometric Turbidity Unit (NTU)) and suspended solids of the mixture were measured with and without a magnetic field of 0.48 T.

Further tests were conducted using a standard settling test (White 1975). A 1-L measuring cylinder was used with different concentrations of the magnetic particles (100, 500, and 1,000 mg L⁻¹) uniformly distributed. A permanent magnet was placed at the bottom of the cylinder producing a magnetic field of 0.11 T and a permanent magnet, producing magnetic field of 0.48 T, which was applied to the side of the cylinder for 10 seconds in a descending position (see Supplementary material, Figure S1). The turbidity and the suspended solids of the suspensions were measured with and without a magnetic field.

The saturation magnetization (M_s) and intrinsic coercivity (H_ci) values of all prepared samples are presented in Table 3. The saturation magnetization of Fe₃O₄ at a high field yield of 9 T was found to be between 89 and 92 emu g⁻¹ for the three different calcine temperatures. The calcining temperature of 773 K seemed to give a slightly higher M_s value.

For the Fe₃O₄-coated samples prepared from Method 1, the saturation magnetization was found to be between 0.93 and 1.52 emu·g⁻¹ at a high field yield of 3 T. The relative level of magnetization for samples prepared using Method 1 (geopolymer coated with Fe₃O₄ were in the range of 1–2%). As a comparison, the zeolite materials were found to have a relative level of magnetization of (4–6%). Although the amount of Fe₃O₄ was not measured, it can be assumed that the relative level of magnetization reflects the level of Fe₃O₄ coating on the surface of the geopolymer assuming that the Fe₃O₄ did not lose its magnetization during the preparation process.

For Method 2, the magnetic FAG showed high saturation magnetization ranging from 6.8 to 15.81 emu·g⁻¹ (Table 3). Assuming no reduction in the magnetization of the Fe₃O₄ (92.47 emu·g⁻¹), the estimated masses of Fe₃O₄ based on magnetization fraction was 7–17% (Table 3). These were close to the 6.34–18.4% by weight of Fe₃O₄ added to prepare the MFAG samples.

The FAG itself had a saturation magnetization of 0.94 emu·g⁻¹ and an H_ci value of 61 Oe while raw material Class-F fly ash had a saturation magnetization of 1.07 emu·g⁻¹ and an H_ci value of 107 Oe. This indicates that the fly ash contained small amounts of magnetic iron oxide which may come from the 5.43% by weight of Fe as per the chemical composition provided by the manufacturer (Ash Grove Technical Center, Overland Park, Kansas, USA).

Samples synthesized by Method 2 exhibited a higher saturation magnetization which was mainly attributed to the more uniform dispersion of Fe₃O₄ particles during the synthesis process, allowing for better alignment of magnetic domains and thus enhanced magnetic performance. On the contrary, Method 1 resulted in a relatively lower saturation magnetization, likely due to the aggregation of Fe₃O₄ particles on geopolymer surface during synthesis, which disrupted the continuous magnetic path within the geopolymer.

The results indicated that the synthesized MFAG12 from Method 2 had the highest uptake of Cu²⁺ (49.81 mg·g⁻¹) which was similar to that of non-magnetic geopolymer (49.79 mg·g⁻¹). MFAG18 had a slightly lower uptake of Cu²⁺ (45.08 mg·g⁻¹) than MFAG12 but the uptake of Cu²⁺ (49.68 mg·g⁻¹) for sample MFAG6 was similar to that of MFAG12 and the non-magnetic geopolymer. The raw fly ash removed about 21.40 mg·g⁻¹ of Cu²⁺. A possible reason is that the added prepared magnetic Fe₃O₄ (18.4% wt) into the geopolymer may have impacted the surface adsorption sites for removal of the metal as compared to MFAG12 which had 11.2% wt of magnetic Fe₃O₄ added. The results of the single-point adsorption showed that non-magnetic polymer can be magnetized at about 9% level of magnetization or slightly higher (by the addition of 11.9% wt of magnetic Fe₃O₄) without affecting its adsorption capacity.

Except for MC-FAG6, the magnetic-coated geopolymer particles from Method 1 gave a lower uptake of Cu²⁺ than that of Method 2. The Cu²⁺ removed by the MC-FAG6 was 48 mg·g⁻¹, which was comparable to MFAG9. Since the FAG was saturated with iron nitrate and calcined (Method 1), possible reasons for the reduced adsorption at higher iron content include interaction of FAG adsorption sites with iron during calcination and possible iron precipitation on the FAG which may have reduced active surface area on the geopolymer.

All the synthesized magnetic geopolymers showed improved metal uptake behavior compared to the magnetic Fe₃O₄ itself which had limited uptake of Cu²⁺ ion (1.40 mg·g⁻¹). As a comparison, zeolite showed a Cu²⁺ uptake of 2.55 mg·g⁻¹ which was higher than the magnetic-coated zeolite (MC-Z17: 0.85 mg·g⁻¹ and MC-Z9: 1.35 mg·g⁻¹). The decreased metal uptake was probably due to the low metal uptake affinity of the magnetic iron oxide which reduced the metal uptake affinity of the zeolite.

Figure 2(b) shows the pH change when the samples were added to an acidic solution. The magnetic FAG all gave pH changes that were less than the FAG.

Based on the single-point adsorption results and the level of saturation magnetization, magnetic geopolymer MFAG12 from Method 2 which showed the highest uptake of Cu²⁺ (49.81 mg·g⁻¹) and a 9% level of saturation magnetization, was selected for further physical and chemical characterization and adsorption studies.

The K_L, Q_m, n, K_F values and the regression correlation coefficients (R²) for Langmuir and Freundlich isotherms are given in Table 4. The R² correlation indicates that experimental data was better fitted using the Langmuir model (R² = 0.981–0.999) as compared to the Freundlich model (R² = 0.901–0.981). The maximum uptake capacity (Q_m) for Cu²⁺ was 111.11 mg·g⁻¹ for MFAG12 (<0.177 mm) which was slightly higher than the uptake capacity (102.04 mg·g⁻¹) for FAG (<0.177 mm). A possible reason may be due to a slightly larger surface area available as shown by the higher BET surface area (36.46 mg·g⁻¹) of MFAG12 as compared to that (32.39 mg·g⁻¹) of the FAG. In addition, both fine geopolymer particles showed much higher uptake capacity as compared to 29.41 mg·g⁻¹ for gravel-sized FAG (0.6–2 mm) at initial pH 3.0. This also showed that incorporation of the Fe₃O₄ particles did not affect the adsorption capacity of the fly-ash geopolymer. Magnetic geopolymer adsorption results were found to be of the same magnitude as that of Cu²⁺ adsorption capacity (87 mg·g⁻¹) of a magnetic composite based on zeolite (Oliveira et al. 2004). Similarly, the Cu²⁺ adsorption capacity of magnetic chitosan chelating resin was reported as 103.16 mg·g⁻¹ (Monier et al. 2010) which was similar to that of the magnetic geopolymer in this study.

Table 5 shows the solution turbidities and suspended solids concentrations with and without the magnetic settling. The solution with magnetic settling showed significantly lower turbidities from 68.3 to 0.66 NTU within 5 min of settling for 100 mg·L⁻¹ of fine MFAG12 solids. For gravitational settling, the solution turbidity after 120 min settling was 6.2 NTU with a starting turbidity of 68.3 NTU. The solution suspended solids for the magnetic settling were 35.1 mg·L⁻¹ after 5 min for a starting suspended solids of 1,000 mg·L⁻¹. In contrast, the solution suspended solids were 158.3 mg·L⁻¹ after 120 min of gravitational settling. The results from Table 5 showed the magnetic geopolymer particles can be rapidly separated from water solution by a magnetic field.

Magnetic geopolymers were made from two inexpensive and widely available raw materials, i.e., fly ash, an industrial waste, and iron nitrate. In Method 1, Fe₃O₄ magnetic particles were synthesized on the surface of the prepared geopolymer while in Method 2 pre-synthesized magnetic Fe₃O₄ particles were incorporated into and during the manufacture of the geopolymer. With Method 2, the magnetic property of the Fe₃O₄ particles was found to be preserved in the synthesis process and the adsorption capacity for copper was similar to that of the non-magnetic fly-ash geopolymer indicating that the Fe₃O₄ did not affect the adsorption capacity of the fly-ash geopolymer. The magnetic geopolymer was effectively separated with a magnetic field within minutes as compared to gravitational settling which requires about 2 h to obtain comparable effluent water quality results. In this study, about 12% by weight of magnetic Fe₃O₄ was found to be sufficient to produce magnetic fly ash-based geopolymer (MFAG12) without affecting the adsorption capacity and magnetization properties. The MFAG can be applied to treat industrial acidic wastewaters or acid mine drainage by neutralizing the pH of these wastewaters and at the same time removing heavy metals. Further study will focus on optimizing the magnetic Fe₃O₄ content on its physical–chemical properties such as specific surface area and surface functional groups, competitive adsorption properties, long-term stability of the magnetic Fe₃O₄ and adsorption properties of the geopolymer, and long-term stability of geopolymer in treatment of industrial wastewater with complex pollutants.

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

The authors declare there is no conflict.