Effective phosphate removal from sewage water using zerovalent iron nanomaterial as an adsorbent

Safe drinking water is the most pivotal commodity for the nourishment of life. Potable water, which is used widely for numerous purposes, is getting increasingly polluted today on account of rapid industrialization, man-made activities, and geogenic disturbances. Thus, it is our primary responsibility and a challenge to keep water sources pollutant-free and also keep mineral ion concentration within the permissible levels of toxicity so that water sources can be consumable for human activities (Fu et al. 2014; Brandl et al. 2015; Morillo et al. 2015). Phosphorus (primarily in the form of phosphates) serves as an important nutrient for every existing organism, which, in turn, acts as a major resource for the synthesis of energy carriers, proteins, and nucleic acids in any living cell (Rotzetter et al. 2013).

Nowadays, phosphate (⁠⁠) in different forms is one of the vital components that is used in agriculture and industries (fertilizer, domestic detergent, animal feed, food and beverage industries, etc.) (Cordell et al. 2009). The application of phosphorous-containing fertilizers is an integral part of conventional farming practices that produce food for billions of people (UNEP 2011). Besides, the available quantity of this nutrient controls the growth of algae and aquatic plants and helps sustain aquatic ecosystems. But an exponential growth of population and rapid industrialization leads and is leading to an increase in the amount of phosphate used in both soil and water. Thus, this industrial and agricultural runoff leads to excessive liberation and the accumulation of phosphate in water bodies. This may induce several detrimental environmental effects, including eutrophication (anomalous flourishment of seaweeds and marine plants) (Chouyyok et al. 2010). It is observed that the eutrophication process causes a gradual depletion of dissolved oxygen content, which results in the death of several aquatic species (Zhao et al. 2009). The algae growing in long strands often twine around boat propellers and hamper transportation (Ansari et al. 2011). Eutrophic waters tend to be scummy, cloudy, or even soupy green. The rapidly growing aquatic plants may wash onto the shores in storms or high winds, where these plants die, decay, and produce a foul smell all around such water bodies. Enhanced eutrophication also indirectly hampers the economic development of communities that depend on aquatic food and other resources (Cleary et al. 2009). The United States Environmental Protection Agency has endorsed the permissible limit of phosphorus in water to be less than 0.1 mg/L to prevent eutrophication (USEPA 1986). Additionally, the Florida Everglades Forever Act endorsed an original acclamation regarding phosphorus concentration in water, which is to be limited within 10 ppb (Florida State Legislature 1994; Arshadi et al. 2015a, 2015b).

In this sense, phosphate recovery and recycling that can be done from waste materials becomes a fundamental element for sustainable management (Jupp et al. 2021). However, researchers have gradually focused on the importance of phosphate recovery from wastewater (Kumar & Pal 2015). Agriculture can recycle the P-rich effluents produced in the food chain, directing them to production systems and concomitantly reducing supply risks and ensuring an economically P-stable value (Roy 2017; Jansson et al. 2019). Besides, livestock waste and manure contain high concentrations of phosphate (Nancharaiah et al. 2016; Tao et al. 2016), while the availability of such resources is inadequate. However, by comparison, municipal wastewater possesses the greatest potential for phosphate recovery (Mehta et al. 2015). Specifically, it has been reported that municipal wastewater flows contain a rich amount of phosphorus, with about 60,000–70,000 t P/a in Germany (Adam 2011). However, in practice, in developed and developing countries, the process of P recovery is still a challenge and requires greater attention.

In this context, different treatment techniques have been employed to date for the removal of phosphate, which include physio-chemical processes (Mishra et al. 2010), chemical precipitation (de-Bashan & Bashan 2004), and biological phosphate removal (Gouider et al. 2011). Although these techniques have been and are effective, they have disadvantages. For example, the biological treatment process has a major drawback, in that the transfer of phosphate from liquid to the sludge phase is essential, leading to a decrease in removal efficiency (RE) up to 30% (Stensel 1991). Physico-chemical treatment processes are associated with certain problems (such as high chemical expenses and operational costs as well as complex sludge generation), which make these processes less desirable (Ghoreishi & Haghighi 2003; Sirianuntapiboon et al. 2006). The chemical precipitation method has limited applicability in the case of real water samples due to the high disposal cost of the formed precipitate and the use of expensive chemicals (Wang et al. 2005).

Against this background, it has been found that among the aforementioned methods, adsorption of phosphate from wastewater by using a suitable non-toxic environment-friendly highly porous adsorbent is the most desirable to control pollution. It is considered to be suitable in terms of less sludge generation, easy operation, and high removal efficacy (Ryu et al. 2011; Novillo et al. 2014; Tran et al. 2015). So far, typical adsorbents, e.g., zeolite, activated carbon, and diatomite, were chosen for the process. However, they come with certain drawbacks such as high production cost, recovery, and separation difficulties from aqueous solutions, hindering the reusability property of adsorbents (Quesada et al. 2020). For ameliorated and effective adsorption of phosphate, the use of nanoparticles as adsorbents has emerged as a highly effective strategy because of their higher specific surface area, higher effective surface-to-volume ratio, and chemical as well as biological stability (Wen et al. 2014; Ghosh et al. 2017).

Recently, nZVI has been broadly utilized for the adsorption of inorganic anions (Wen et al. 2014), metalloids (Kanel et al. 2005), and metals because of its alluring properties like the higher surface area-to-volume ratio, resulting in the formation of more active sites. In our work, pure nZVI was selected for remediation of phosphate due to its above-mentioned properties, along with the added advantage of improved magnetic behavior, which helps in a facile separation of adsorbents from water samples after treatment processes. Several research groups have utilized suitable nanoscale adsorbents for phosphate removal and the treatment of sewage water. Some literature reports of such groups have been documented.

Arshadi et al. observed phosphate adsorption by nitric acid–treated ostrich bone waste–supported nZVI (OBW-HNO₃-nZVI) from both synthetic and real (Persian Gulf) water samples. The chain-like structure of OBW-HNO₃-nZVI (0.05 g/L dosage) having a Brunauer–Emmett–Teller (BET) surface area of 41.4 m²g⁻¹ shows maximum phosphate RE (100%) at pH 5.0. The removal percentage of phosphate from the waters of the Persian Gulf was calculated to be 99.9% after 30 min (using 0.05 g/L dosage of OBW-HNO₃-nZVI) (Arshadi et al. 2015a, 2015b). Arshadi et al. investigated phosphate uptake by ferrocene-functionalized aluminum–silicate nanoparticles (Si/Al@Fe) from synthetic and real water samples (Persian Gulf). A total of 96.6% adsorption efficiency was found at pH 8 with an initial phosphate concentration of 850 mg/L (0.05 g/L Si/Al@Fe dosage). The phosphate removal percentage from the Persian Gulf waters was found to be 99.9% by utilizing the synthesized adsorbent after 120 min. The results confirm that Si/Al@Fe is a potential adsorbent to uptake phosphate ions from real water at low concentrations (Arshadi et al. 2015a, 2015b). Ge et al. synthesized 3D flower-like hierarchical iron containing MnO₂ hollow microspheres for efficient phosphate removal from synthetic water and eutrophic lake water (Nanfei river located in the Hefei city of China). A total of 98% phosphate removal was obtained when the initial phosphate concentration was 10 mg/L, with the initial solution at pH 7.0 and a fixed adsorbent dosage of 0.5 g/L. However, almost complete phosphate removal from lake water (total phosphate concentration 1.48 mg/L) was achieved with 0.25 g/L of adsorbent dosage (Ge et al. 2016). Riahi et al. reported fluoride adsorption by synthesized Fe₃O₄ nanoparticles modified with zirconia (Fe₃O₄@ZrO₂). The fluoride adsorption capacity was found to be 123.9 mg g⁻¹ at pH 2.5 (adsorbent dosage 1 mg mL⁻¹). The fluoride adsorption was also carried out on tap water, leading to an appreciable removal of fluoride below its permissible limit (Riahi et al. 2015). Sun et al. performed a simultaneous removal of nitrate and phosphate by a synthesized solid carbon source/zerovalent iron (SCS/ZVI) composite. Herein, nitrate and phosphate removal rates of 1.1±0.1 mg L⁻¹ h⁻¹ and 0.21±0.07 mg L⁻¹ h⁻¹ were achieved, respectively, at pH 7 (Sun et al. 2021).

Considering the above-mentioned studies/discussion, it is evident that excessive phosphate in water bodies and agricultural runoff are very harmful and need the urgent attention of the research community. It is in this context that we decided to focus our attention on the analysis and treatment of real water samples from a neighboring lake (Saheb bandh lake), situated in the north-western part of Purulia district (23.33 °N, 86.37 °E), West Bengal. The soluble reactive phosphorus or orthophosphate (⁠⁠) present in lake water generally ranges between 0.223 and 0.95 mg/L season wise, which is much higher than the permissible level of phosphate in drinking water (Siddiqi & Chandrasekhar 2010). Hence, according to the Trophic State Index, eutrophy is observed in this lake water (Dutta et al. 2019). Besides, lake water constitutes natural organic matter (NOM) of great quantity, and NOM consists of a complex heterogeneous continuum of high- to low-molecular-weight species exhibiting different water solubilities and reactivities. Generally, different organic acids such as citric acid, humic acid, oxalic acid, and chlorophyll-a persist in lake water as NOMs in very small amounts. Therefore, the competition between the anion forms of acids and phosphate for the active sites of the studied adsorbent becomes irrelevant and negligible. Furthermore, in the presence of appreciable quantities of NOM, nZVI is the most suitable and potential adsorbent for the removal of phosphate from the studied lake water. NOM enhances the mobility of nZVI (Johnson et al. 2009). Thus, the aggregation of nZVI reduces markedly, which, in turn, improves the efficiency of phosphate removal by the studied adsorbent (Aiken et al. 2011). The presence of NOM not only enhances the mobility of nZVI but also reduces the aggregation of nZVI. This leads to an increase in the stability of nZVI in the aqueous medium, which makes it a better adsorbent (Domingos et al. 2009).

In the present study, we have synthesized porous spherical-shaped zerovalent iron nanoparticles via a simple cost-effective reduction method for phosphate removal by varying different reaction parameters. Herein, it is shown that nZVI can be reused for phosphate adsorption up to the fifth cycle. Furthermore, the probable phosphate adsorption mechanism by synthesized nZVI has been elucidated. Finally, the phosphate uptake capacity of the developed adsorbent for practical application in lake water samples has been studied.

Iron (III) chloride hexahydrate (FeCl₃·6H₂O, 98%), the precursor of iron, and sodium borohydride (NaBH₄, 98%), reducing agent, were purchased from Merck (India). Absolute ethanol (>99.9%), glycerol (>99.5% Emparta, AR grade), sodium hydroxide, and hydrochloric acid used for pH adjustment were purchased from Merck (India). Milli-ultrapure water (18.2 MΩ cm) was collected for making all solutions and for performing all other tests. Stock solutions having a concentration of 1,000 mg/L were prepared from oven-dried potassium phosphate (K₃PO₄) purchased from Merck, India, and used to calibrate estimates. Ammonium molybdate [(NH₄)₆Mo₇O₂₄·4H₂O], sulphuric acid (H₂SO₄), and stannous chloride (SnCl₂·2H₂O) were used as reagents.

nZVI was synthesized as reported earlier (Ghosh et al. 2019) according to Scheme S1.

Different techniques were employed to characterize the synthesized nZVI particles after their preparation. An X-ray diffraction (XRD) study was performed using the PANalytical Xpert PRO XRD unit to investigate the crystalline structure of the obtained particles. During the process, the operating voltage was set at 40 kV. Also, a Cu-Kα radiation source having a wavelength of 1.54 Å was used to scan in the range of 10°–75°. A field emission scanning electron microscope (FESEM) study was performed using a Zeiss Sigma field emission scanning electron microscope (SE2 model, 17 kV) in order to examine the surface morphology of nZVI. High-resolution transmission electron microscope (HRTEM) imaging was performed with Tecnai TF20G2ST HRTEM operating at 220 kV voltages. BET surface area experiments were performed at a temperature of −196 °C using a volumetric apparatus. (Micrometrics ASAP 2020 is an automated gas sorption analyzer.) The degassing of all samples was performed under vacuum at 120 °C for 16 h before the adsorption experiments. The BET surface areas of the adsorbents and the pore size were calculated using the BET method and the Barrett–Joyner–Halenda (BJH) method, respectively. Fourier transform infra-red (FT-IR) spectrum was recorded in the range of 4,000–400 cm⁻¹ as KBr pellets using the Nicolet 5700 spectrometer to track changes on the nZVI surface before and after the adsorption experiment.

In batch bottles, 50 mL of phosphate solutions (phosphate concentration 25 mg/L) and the nZVI adsorbent (560 mg/L dosage) were mixed. The pH value of the mixed solutions was changed to the preferred values from 2.0 to 12.0 due to the addition of HCl or NaOH solution. To ensure equilibrium, the sealed bottles were positioned in a thermostatic shaker (500 rpm) for 2 h at room temperature.

To monitor the effect of ionic strength and coexisting anions on phosphate RE, the reactions were performed in the same manner as that of the above experiment. The addition of nZVI (560 mg/L) to 50 mL (25 mg/L) sample phosphate solution changing concentrations of NaCl (0.005–0.1 M) was done for establishing ionic strength dependence on phosphate removal. Likewise, coexisting anions such as carbonate and sulphate, which are commonly present in surface water and groundwater, were used for preparing 50 mL of phosphate solution. The concentrations of (0.05–0.2 M) and (0.05–0.2 M) were varied in the solution. The reaction was allowed for 2 h on a mechanical stirrer (500 rpm) at 25 °C, and the equilibrium phosphate concentrations were resolved at regular intervals.

The nZVI adsorbent was collected after the adsorption experiment by centrifugation, washed thoroughly with double distilled water several times, and dried under vacuum. The dried nZVI adsorbent was further used for the removal of phosphates from the water samples when it was considered as the first reusability cycle. Here, the reusability of the adsorbent was checked by performing the experiment up to five cycles.

Real water samples were collected from Saheb bandh, Purulia district, in a polyethylene bottle of 5 L to perform the phosphate removal experiment. The water sample was well shaken before performing the reaction. The concentration in the real sample was determined by the ammonium molybdate method.

Fig. S1(A) shows the nitrogen adsorption–desorption isotherms of synthesized nZVI particles. According to the IUPAC classification, the curve shows type IV isotherm, which indicates meso-porosity in nZVI. In this isotherm, the H3 type hysteresis loop reveals the existence of an irregular orientation of pores as well as interconnected slit-shaped pores in nZVI. The BJH pore size distributions (PSDs) derived from isotherm desorption data are shown in Fig. S1(B). The curve shows relatively wide PSDs in the mesoporous area. The textural properties of the synthesized nZVI are tabulated (Table S1).

Different factors such as initial phosphate concentration, the dosage of adsorbent (nZVI), pH, ionic strength, and coexisting anions influencing the phosphate RE from laboratory-based synthesized water were studied efficiently and elaborated in this section.

Several coexisting anions present in wastewater along with phosphate and having their influence on phosphate removal were studied. The salts of chloride, sulphate, and carbonate with their varying concentrations in phosphate solution were set for the batch reaction discussed earlier, and the phosphate removal was investigated as shown in Fig. S3. The results from Fig S3(A) show that the presence of 0.005 M NaCl yields a phosphate removal of 95.53% at lower pH, while the presence of 0.01 moles of Na₂SO₄ shows a phosphate removal of 94.53% (Fig S3(B)). This shows that the existence of these common anions does not disturb the removal of phosphate by nZVI; however, strongly affects adsorption, and the % of removed in our experiment was 31.29, 25.09, and 17.06% at 0.0025, 0.005, and 0.01 mole concentrations, respectively (Fig S3(C)) (Wu et al. 2013a, 2013b; Zhang et al. 2017). Herein, the anions forming the outer sphere complexes are strongly dependent on ionic strength, which may be due to the electrostatic forces. In this case, phosphate probably forms inner sphere complexes via ionic bonding at the solid–liquid boundary, which is quite stable compared to outer sphere complexes (Liu et al. 2007).

Here, we present Table S2 containing comparative data for the removal of contaminant from water samples by various research groups.

In Equation (8), K_T and b signify the equilibrium-binding constant (L mg⁻¹) and adsorption energy (J mol⁻¹), respectively, and R and T indicate the universal gas constant (J mol⁻¹ K⁻¹) and temperature (K), respectively. Here, K_T and b can be calculated by plotting q_e vs. ln C_e in a graph.

Then, the soluble (sorbate) is diffused from the bulk solution to the outer plane around nZVI (sorbent) via liquid–liquid diffusion. A feasible adsorption of the sorbate solute on to the active sites on the outer surface of nZVI occurs. The liquid–solid step occurs mainly through both physisorption and chemisorption to control this step.

• (1)

  Physisorption: The electrostatic attraction between the negatively charged phosphate (⁠⁠) species and the positively charged nZVI surface results in the deposition of iron phosphate.

Water samples from the Saheb Bandh Lake were tested for various parameters such as pH, Chemical oxygen demand (COD), Biological Oxygen Demand (BOD), Electrical Conductivity (EC), temperature, Dissolved oxygen (DO), Total Dissolved Solids (TDS), and total hardness, and their values are tabulated (Table S4).

To evaluate the potentiality of the synthesized adsorbent, i.e., nZVI for practical application, real water samples were collected from Saheb Bandh and treated using the synthesized nZVI. Before treating the collected lake water, several important characteristics were determined, which are listed below (Table 2).

Herein, we utilized a laboratory-based cost-effective method to synthesize nanoscale zerovalent iron (nZVI) and evaluated its feasibility as an adsorbent toward achieving clean and hygienic water by removing phosphate contaminants present in real water samples.

At first, phosphate removal study from Bandh water was conducted at room temperature by keeping the nZVI dosage at 560 mg/L and at a neutral pH. From Fig S4, it can be understood that with increasing the time of reaction ⁠, the removal percentage increases rapidly. The phosphate RE reaches about 86.5% at 120 min reaction time. The scope of to be adsorbed on the nZVI surface gradually increases with increasing the reaction time as nZVI took more time to adsorb in the Bandh (real) water solution. But above the 120 min reaction time, phosphate RE is unaltered because all the adsorbed sites or pores of nZVI become full with ⁠.

The influence of pH on removal from Saheb Bandh water was carried out by keeping the pH range from 2 to 10 and the nZVI dosage at 560 mg/L. From Fig. S5, it is clear that the % RE becomes lower at pH 10 and reaches about only 50%. But the % RE is maximum (99%) at pH 2. Finally, the % RE reaches nearly 90% at pH 6. At a lower pH, phosphate is mainly present in the form of and ⁠, while at a pH above 10, phosphates are mainly present in the form of ⁠. These results can be studied on the basis of the IEP of the nZVI particles, which is ∼8.0. Therefore, the nZVI surface possesses a more positive charge when the pH of the solution is less than that of the IEP. Hence, the absorption of negatively charged species is better at a lower pH compared with that in highly basic pH conditions. The affinity of nZVI toward phosphate ions decreases at a higher pH because of the presence of ample OH⁻ ions, which compete for the adsorption sites along with ions.

The cost-effectiveness factor of the remediation process is a crucial parameter to establish the field application feasibility of synthesized nZVI. The total cost calculation for the laboratory preparation of 1 g of nZVI is represented in Table S4. From Table S4, it is clear that the cost incurred for the preparation of the adsorbent is approximately INR 125.478 g⁻¹.

In summary, spherical-shaped nZVI was successfully synthesized by the reduction method, and this nZVI was employed for the removal of phosphate from laboratory-based synthetic water as well as real water collected from Saheb Bandh, Purulia, West Bengal, Eastern India. XRD study confirmed the presence of metallic iron. A higher BET surface area (143.163 m²/g) indicated that nZVI might act as a potential adsorbent for phosphate and could efficiently adsorb phosphate from wastewater. Various parameters, e.g., initial phosphate concentration, adsorbent dosage, pH, and ionic strength, were optimized. A maximum removal rate of ∼96% was achieved at pH 2. Ionic strength did not affect the removal of in the presence of 0.01 moles of NaCl. The coexisting anions chloride and sulphate did not affect the phosphate removal process. However, carbonate anions strongly affected phosphate removal. Equilibrium and kinetic studies depicted that the adsorption process was a good fit with the Temkin isotherm model and the pseudo-2nd-order kinetic model, respectively. Some thermodynamic parameters, e.g., ΔG⁰, ΔS⁰, and ΔH⁰, were estimated, which suggested that the adsorption of phosphate on the nZVI surface is endothermic in nature and spontaneous. The reusability test of nZVI proved that it could be reused up to the fifth cycle and be applied in large-scale industries. Also, the phosphate RE from Saheb Bandh water was about 86.5% up to a reaction time of 120 min, and a lower pH value (pH 2) facilitated achieving maximum RE. Furthermore, the real water (Saheb Bandh water) sample results demonstrate that the synthesized and cost-effective nZVI is a promising nanomaterial for the removal of from contaminated wastewater. From all these perspectives, the nZVI system can be prepared as a bed for industrial application. Then real water samples can be passed through this bed to obtain clean and hygienic water. Moreover, the amount of clean water produced can be determined. Also, how many times the same nZVI bed can be used to make phosphate-free water can be evaluated. Ultimately, in this study, it was proved that nZVI was a suitable adsorbent for effective removal of ⁠, a dangerous pollutant found in aqueous bodies. A cost estimation analysis was done for lab-scale preparation of nZVI. Table S4 (Supplementary file) reveals that only 125.478 INR has been spent on the synthesis of 1 g nZVI. Thus, only 70.267 INR is needed for the complete removal of -containing wastewater of 1 L volume.

Although, in our work, synthesized nZVI has been proved to be a feasible nanomaterial for efficient removal of from wastewater, some problems associated with this application should be fixed.

Rapid corrosion is one of the most serious problems that limit its application. nZVI produces Fe(II), Fe(III), and hydrogen while in contact with polluted water. These generated nZVI corrosion by-products block the active site on the nZVI surface, which leads to surface passivation. As a result, the removal rate of contaminants is reduced. In this context, research focus should be on the development of polymer, carbon-supported nZVI-based nanomaterial to prevent its corrosion during reaction. In addition, the selectivity of nZVI for removing specific contaminants and multicomponents in the complex system is poor, which should be improved. Moreover, it is essential to prove the practicability of nZVI in large-scale applications, and its effectiveness, safety, and economy should also be guaranteed. Today, knowledge about the probable hazards caused by the utilization of nZVI is inadequate. The toxicity of nZVI materials must be taken into account before applying them for environmental purification. Hence, efforts should be taken to elucidate the toxicity mechanism of nZVI and evaluate its environmental hazards for prescribing a suitable standard.

We express our sincere thanks to the National Institute of Technology (NIT), Durgapur, and Burdwan University for providing us the necessary infrastructure to perform our work successfully. Finally, we greatly acknowledge the Department of Earth and Environmental Studies, NIT Durgapur, for offering M.Tech dissertation work.

The authors confirm that the final version of the manuscript has been reviewed, approved, and consented for publication by all authors.

Material preparation (zerovalent iron preparation), laboratory work, manuscript draft writing, data analysis, data interpretation, editing, and formatting of the manuscript were done by I.S. Laboratory work and data collection were performed by S.S. Planning and design of work, analysis, interpretation of data, and revision of the manuscript were done by M.R. Real water sample collection and the design of the project were done by S.G. Oversight and leadership responsibility for the project planning and execution including mentorship to the core team and funding acquisition were taken up by R.S. Finally, all authors read and approved the final manuscript.

This work was funded by the Department of Higher Education, Science & Technology and Biotechnology, Govt. of West Bengal (Project Memo No. 45(Sanc.)/ST/P/S&T/1G-60/2017 dt. 11.07.2018) and the sponsored project (No. PDF/2017/000390) supported by Department of Science and Technology under the Science and Engineering Research Board, Government of India.

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

The authors state that there is no conflict to declare.