Applicability of ferric(III) hydroxide as a phosphate-selective adsorbent for sewage treatment

Phosphorus together with nitrogen is an essential nutrient in aquatic ecosystems as well as agriculture. However, excessive input of phosphate in water bodies usually triggers eutrophication, which threatens the aquatic ecosystem (Genkai-Kato & Carpenter 2005; Conley et al. 2009; Lewis et al. 2011). Thus, methods of sewage treatment focus on the efficiency improvement of phosphorus compound reduction (Schindler et al. 2016). In Korea, the discharge limits of T-P in sewage treatment plants has been greatly strengthened from 2.0 mg/L to 0.2 mg/L, to prevent eutrophication. China has also strengthened the maximum P allowed for secondary effluents from wastewater treatment plants from 1.0 mg/L to 0.5 mg/L (Wu et al. 2019). As discharge limits of nutrients become more stringent, the demands for an introduction of advanced treatment techniques continue to increase in sewage treatment for the appropriate removal of phosphorus and nitrogen. Various advanced treatment techniques, as the form of a hybrid process integrated into existing biological treatment process, including chemical precipitation, biological phosphorus removal, and adsorption, etc., have been developed for removal of phosphate from sewage and wastewater to control eutrophication (Gogate & Pandit 2004; Park & Na 2009; Park et al. 2016; Zhao et al. 2016; Omwene et al. 2018; Jiang et al. 2019).

In recent years, chemical phosphate removal facilities have been widely adopted as advanced treatment processes for sewage due to their relatively simple treatment process, manageable and high reliability of treatment efficiency, although their chemical and waste disposal costs are very high. However, chemical and biological methods commonly used for phosphorus removal are not economically and cost-effective in small treatment plants (Gajewska et al. 2011). Moreover, although it is well demonstrated that the application of commercial coagulants such as polyaluminum chloride (PAC) is very effective in removing phosphate from sewage, and the presence of aluminum is essential to plants growth, but excess can be toxic to humans as well as plants (Kochian & Jones 1997). It is known that potable water highly contaminated with aluminum (320 mg/L) can affect the digestive system, and cause skin rashes and memory loss (Kasprzyk & Gajewska 2019). The impact of aluminum on Parkinson and Alzheimer disease is also the subject of many research studies (Bondy 2014). Thus, these facts suggest that another efficient technology needs to be found.

Together with chemical precipitation method, it is well known that adsorption is a promising method to remove phosphate due to its advantages of operational simplicity, environmental friendliness, and higher effectiveness even at low phosphate concentrations (Méndez-Álvarez et al. 2002). Accordingly, many adsorbents, such as zeolite, activated carbon, industrial solid wastes such as red mud, fly ash, and slag, ion-exchange resin, and polymeric materials have been investigated for phosphate removal (Sendrowski & Boyer 2013; Park et al. 2015; Wang et al. 2016). Recently, metal-based adsorbents have been extensively investigated to selectively adsorb phosphate anions in wastewater, such as magnesium oxide-enriched biochar, iron-modified biochars, hydrous zirconia-modified zeolite, hydroxy-iron-aluminum pillared bentonite, and lanthanum-modified diatomite (Yan et al. 2010; Zhang et al. 2012; Fan et al. 2017; Yang et al. 2018; Wu et al. 2019). Most of these adsorbents exhibit high adsorption capacity and selectivity for phosphate. However, there is a growing demand for more efficient, cost-effective and effective regenerative adsorbents.

It is known that the specific adsorption of phosphate with iron oxyhydroxides is important in controlling phosphate in industrial and engineering processes, as well as in the environment (Krom & Berner 1980). Iron oxyhydroxide is known to occur commonly as goethite (α-FeOOH), akaganeite (β-FeOOH), and lepidocrocite (γ-FeOOH). Among them, goethite is the most widespread iron oxide in the natural environment and has been used in various adsorption studies for phosphate, fluoride, arsenate, selenite and etc. Conversely, akaganeite is a rare mineral in nature, and has semiconductor, catalytic, and ion-exchange properties (Villalba et al. 2010).

The aim of this study was to develop phosphate-selective absorbents that are not affected by co-existence anions, especially sulfuric acid and chlorine ions, to apply the adsorption process as a phosphate removal facility in urban sewage treatment plants. The phosphate adsorption and adsorption selectivity of akaganeite (β-FeOOH) were evaluated by a batch-type adsorption experiment and a fixed bed column experiment. The akaganeite was processed into a spherical bead-type adsorbent combined with a polymer material and used in the adsorption experiments. The experiments were conducted with synthetic pure phosphate solutions and with discharge water from a public sewage treatment plant.

Ferric chloride anhydrous (FeCl₃, 97%) was used as source of Fe(III) to prepare iron (oxy)hydroxide adsorbent. NaOH (98%) was used to precipitate akaganeite (β-FeOOH) from the FeCl₃ solution. Polyacrylonitrile (PAN, MW = 150,000) as a binder, polyvinyl alcohol (PVA, MW = 500) as a porosity controller, and dimethyl sulfoxide (DMSO) as a solvent were used to make akaganeite powder into a spherical bead. The discharge water was collected from a sewage treatment plant located in Mokpo city, Korea and filtered through a GF-C filter paper and then used as a real wastewater for adsorption experiments. The discharge water used in experiments had a pH of 6.8, 15.1 mg/L nitrate, 3.4 mg/L phosphate, 442 mg/L chloride and 69 mg/L sulfate ions.

Extra pure chemicals, NaH₂PO₄, Na₂SO₄, and NaCl were used to prepare anion solution for sorption experiments. The working solutions were prepared by dissolving the chemicals in deionized water (DIW) or in a pH buffer solution. The pH buffer solution was composed of CH₃COOH and CH₃COONa, with a total acetate concentration of 0.01 M (pH 4.7). Commercial anion-exchange resin, PA308 resin (–N⁺(CH₃)₃Cl⁻, Samyang Co.) was used as a comparative adsorbent.

Akaganeite (β-FeOOH) was synthesized using a procedure reported previously (Chitrakar et al. 2006). First, 0.1 M NaOH solution was slowly added to a 0.1 M FeCl₃ solution at room temperature until the pH of the mixture was attained around 10, the mixture was stirred mechanically during the addition of the NaOH solution. The suspension was further stirred for 12 h. The precipitate was separated by centrifugation, washed with DIW until neutral, and finally dried in air at room temperature. Dried precipitate was pulverized with a particle size range of 150–200 mesh and then used as an adsorption functional inorganic powder to prepare a spherical bead-type adsorbent combined with polymeric materials. The spherical bead-type adsorbent was prepared as follows: 0.3 g of PAN and 0.1 g of PVA were dissolved in 6 mL DMSO solvent, and then 2 g of akaganeite powders was added to this solution with continuously mixing to obtain homogeneous composite solution. The composite solution was dropped into DIW stirring at 120 rpm and room temperature by using a syringe with an air spray. The suspension was further stirred for more than 12 h to dissolve the PVA component in the spherical beads. The solidified beads were separated and washed with DIW, and air dried for more than 24 h at room temperature, and then dried completely at 60 °C to remove the solvent. The spherical beads used as an adsorbent have a grain size range of 10–20 mesh.

The effect of different pH values on phosphate adsorption was studied by adjusting the pH of solution using either 0.1 M HCl or 0.1 M NaOH solutions to the required pH range of 3–12 after the adsorption equilibrium.

The effect of competing anions such as Cl⁻ and SO₄²⁻ ion co-existing in the sewage with phosphate ion on the phosphate adsorption onto adsorbent was evaluated based on the binary system by referring to the chemical composition of discharge water from sewage treatment plant.

To evaluate the regeneration efficiency of adsorbent, repeated adsorption–regeneration experiments were conducted using an alkaline washing process, where phosphate-sorbed adsorbent was immersed in 0.1 N NaOH solution and shaken for 1 h at room temperature. After complete desorption, the adsorbent was washed with DIW until the pH was neutral and dried at 60 °C. The adsorption–regeneration cycle was repeated 10 times and the phosphate adsorption capacity was measured for each cycle.

SEM image (Figure 1(a)) and XRD pattern (Figure 1(b)) of β-FeOOH, and morphology (Figure 1(a)) of cross-section of β-FeOOH/PAN bead are shown. The XRD pattern indicates that the β-FeOOH used in this study was a two-line amorphous akaganeite with poor crystallinity, which is consistent with previous studies (Fu et al. 2018; Zhang et al. 2019). The point of zero charge (PZC) of β-FeOOH estimated from zeta potential measurements (Otsuka ELS-Z1, Japan) at different pHs is about 8.0, similar to the results reported by other studies (Chitrakar et al. 2006). The BET specific surface area and pore volume of the β-FeOOH measured by a surface area and porosity analyzer (Micrometrics ASAP2010, USA) are 173.5 m²/g and 0.38 cm³/g, respectively, suggested that the β-FeOOH can be used as a good adsorbent for phosphate.

Figure 2 shows the phosphate adsorption capacity of ferric hydroxide (Fe(OH)₃, 90%), akaganeite (β-FeOOH) and commercial anion-exchange resin (PA308), based on the adsorption amount obtained by a batch mode adsorption test, in which 0.1 g of each adsorbent is added to 100 mL of 1 mM NaH₂PO₄ solution and stirred at 25 °C for 12 hours. The phosphate adsorption capacity was increased to β-FeOOH > PA308 > Fe(OH)₃. This indicates that, when processing Fe³⁺ in the form of β-FeOOH, the adsorption capacity for phosphate ion was higher than that of a commercial anion-exchange resin as well as ferric hydroxide.

Also shown in Figure 2, the difference in adsorption capacity between the composite bead and the powder was about 15%. This can be explained by the effect of the PAN content used as binder in the composite beads, which means that the adsorption capacity of the inorganic adsorbent has not been seriously decreased in the process of manufacturing the powder into spherical composite beads.

The pH of a solution is an important parameter in the adsorption process because the speciation of anions and the surface characteristics of the adsorbent are influenced by the pH value. To study the effect of solution pH on phosphate adsorption by β-FeOOH, the experiment was conducted with 0.5 mM NaH₂PO₄ solution at different pH values (2.5–12.0). Figure 3 shows the results of the phosphate adsorption on β-FeOOH as a function of the equilibrium pH of solution. The phosphate adsorption on the β-FeOOH remained almost constant until the pH of the solution was increased up to around 6, and then sharply decreased with increase in pH. Similar findings have been observed in the phosphate adsorption by other metal-based adsorbents (Yan et al. 2010; Fu et al. 2018). The removal of phosphate from the solution by β-FeOOH was nearly 100% at pH < 6. A possible explanation for the high adsorption capacity of β-FeOOH at pH < 6 is the comprehensive effect of the formation of outer-sphere complexes by electrostatic force and inner-sphere complexes by ligand exchange. The decrease of adsorption capacity in alkaline region may be interpreted by change in adsorbent surface charge rather than by ion competition. As the pH of solution was increased, the number of positively charged surface sites decreased, which is unfavorable to the sorption of negatively charged anions. The point of zero charge (pzc) of β-FeOOH is determined at about pH 8. When pH approached pH_pzc, the positively charged surface sites, which served as phosphate-binding sites, significantly decreased with the increase in pH. The pH effect on phosphate adsorption of conventional anion-exchange resin (PA308) was different from that of β-FeOOH, as shown in Figure 3. The phosphate adsorption capacity of the anion-exchange resin tends to increase continuously with increasing pH of solution up to 7, and then maintained constant values. At pH > 7, the phosphate ion was almost removed (≥95%) from the solution by adsorption of the anion-exchange resin, which indicates probably that it was affected by changes in ionic specification of phosphate. As the pH increases, the phosphates such as H₂PO₄⁻, HPO₄²⁻, and PO₄³⁻ become the major species, which are more favorable to adsorption by anion-exchange resin, but not by β-FeOOH due to greater electrostatic repulsion between the more negatively charged surfaces and phosphate anions.

The phosphate adsorption capacities of various adsorbents reported in the literature are summarized in Table 2.

One of the major problems which limit the wide application of adsorption method to eliminate pollutants is the ion selectivity for an adsorbent. Figure 6 shows the effects of competing anions such as Cl⁻ and SO₄²⁻ on the uptake of PO₄-P ion onto the β-FeOOH and the commercial polyamide (PA) resin depending on the dose of adsorbent. The experiments were conducted under a binary system, which contains 5 mg/L of PO₄³⁻ ion paired with 400 mg/L of Cl⁻ ions or 50 mg/L of SO₄²⁻ ion. In which the concentrations of Cl⁻ and SO₄²⁻ ions are the same values as the concentration of the discharged water used as the real wastewater in this study. As shown in Figure 6, the effects of co-existing competitive anions on PO₄-P adsorption capacity of β-FeOOH were almost negligible in all adsorbent dosage conditions (0.2–1.0 g/L). In contrast, the PA resin showed a significant decrease in adsorption of PO₄-P ions as the presence of competing anions. Furthermore, it is notable that the effects of co-existing anions on PO₄-P adsorption of PA resin were greatly increased as adsorbent doses decreased, because this shows that adsorption competition can take place more intensely under saturation conditions.

To investigate more the interference of co-existing competitive anions, phosphate adsorption was also carried out in the fixed bed column experiments. As shown in Figure 7, the breakthrough curve for phosphate solution (C₀ = 5 mg/L as PO₄³⁻ ion) prepared with deionized water for β-FeOOH was similar to the case prepared with the discharge water from sewage treatment plant. Both curves breakthrough at almost the same time although they are slightly different in the shapes that are probably related to the packing properties. Thus it is clear that the ions in the discharge water had a negligible effect on the adsorption of phosphate by β-FeOOH. The adsorption capacities calculated from the ideal breakthrough (C/C₀ = 0.5) points were 21.6 mg/g for DIW and 21.3 mg/g for discharge water, which are much lower than that obtained from the batch experiment. This difference is assumed to be caused by the slow adsorption rate of β-FeOOH/PAN bead and the pH difference of the solutions (see Figure 3), although further experiments are needed to determine the exact cause. In contrast, PA resin showed a significant difference in the breakthrough time depending on the blank solution. The results indicated that PA resin hardly adsorbs phosphate from sewage discharge water due to competition with co-existing anions.

From these results, it is obvious that β-FeOOH has a high affinity for the PO₄-P anion and there is no significant influence of competing anions on its PO₄-P adsorption capacity, and hence that β-FeOOH can be used effectively as a phosphate-selective adsorbent.

In order to assess the possibility of recycling the adsorbent for reuse in multiple adsorption cycles, repeated adsorption–regeneration experiments were performed. Phosphate desorption experiments for the used β-FeOOH were conducted using 0.1 M NaOH solution, where the 100 mL fraction of NaOH solution was added per gram of phosphate-loaded β-FeOOH and stirred for 1 h at 25 °C. As shown in Figure 8, the adsorption capacity remained almost constant regardless of the number of regeneration cycles. As a result, β-FeOOH could be regenerated by a simple alkali washing process without lowering the adsorption capacity and thus, could be used more than 10 times maintaining the same adsorbent quality.

The phosphate adsorption capacity was increased to β-FeOOH > PA resin > Fe(OH)₃ and showed high adsorption capacity when the ferric iron was processed in the form of β-FeOOH. The adsorption behavior of β-FeOOH for phosphate anions could be explained by both the Langmuir isotherm model and the Freundlich isotherm model, but it showed a better match for the Langmuir model. The maximum adsorption capacity of β-FeOOH for phosphate was found to be 60.98 mg/g from the Langmuir model. The adsorption process was found to follow a pseudo-second-order kinetic model. The phosphate adsorption capacity of β-FeOOH was high at acidic pH values below 6, but sharply decreased with increase in pH higher than 7. The phosphate adsorption capacity of β-FeOOH was not affected at all by the co-existence of competing anions, such as Cl⁻ ion or SO₄²⁻ ions contained in large quantities in the sewage. β-FeOOH could be regenerated by a simple washing process using NaOH solution and could be used more than 10 times while retaining the same adsorption capacity. The results of this present investigation suggest that β-FeOOH can be used as a phosphate-selective adsorbent and the adsorption process using β-FeOOH can be efficiently utilized to the phosphate removal systems of sewage treatment plants.

This research was supported by the National Research Foundation of Korea (NRF-2020R1I1A1A01070853).

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