Lanthanum and magnesium activated palygorskite for adsorption of phosphate in piggery wastewater

Phosphorus widely exists in animal and plant tissues. However, when the phosphate content in the water body accumulates too much, it can cause the rapid growth of algae and other plankton. The dissolved oxygen content in the water decreases and the water quality deteriorates. This ultimately leads to eutrophication of the water (Chouyyok et al. 2010). Human activities such as mining, agricultural pollution, industrial wastewater and domestic sewage may cause eutrophication of water bodies. In particular, the overuse of phosphorus products such as phosphate fertilizers brings a large amount of phosphorus to water bodies, resulting in serious eutrophication problems (Lei et al. 2019). Hereby, it is essential to remove phosphorus in wastewater and reduce the phosphate content of water.

At present, the methods of removing phosphate from water generally include crystallization, chemical precipitation, ion exchange, adsorption and biological methods. Chemical precipitation is done by adding metal salts, such as magnesium (Mg) and aluminum (Al), to form precipitation and separate it from the aqueous phase. However, traditional chemical precipitation methods require strict pH control, which not only consumes a lot of cationic salt, but also produces sludge (Lei et al. 2017). Although the biological method has high phosphorus removal efficiency and low operating costs, it requires strict control of anaerobic and aerobic conditions. In addition, the biological method has a poor effect on wastewater treatment with trace phosphate (Huang et al. 2017). The adsorption method has the characteristics of low cost, high selectivity and less waste generation. As opposed to other methods, the adsorption method is suitable for the advanced treatment of wastewater with low phosphorus concentration before effluent (Liu et al. 2023).

Although traditional clay minerals and industrial wastes are cheap and readily available, they do not have a strong affinity for phosphorus (Li et al. 2016). Lanthanides have a strong affinity for phosphates (Dong et al. 2017), so they have received a lot of attention. Among them, lanthanum is relatively abundant in nature. It is a promising adsorbent for phosphorus removal. To date, a multitude of lanthanum-modified materials have been used for phosphate adsorption (Wei et al. 2021). In addition to this, there have been various studies on La-based gel materials as phosphate scavengers.

Palygorskite is a chain structure hydrated magnesium aluminum silicate clay mineral with a large specific surface area and strong adsorption capacity. Due to its special structure, it has been widely used as an adsorption carrier (Kong et al. 2018). Proper heat treatment can change the structure of palygorskite and enhance its capture of phosphate. However, there are few studies on the modification of high-temperature treated palygorskite by doped lanthanum and other metals.

Oxides or hydroxides of multi-component metals have complementary advantages in removing phosphates, and have more efficient phosphorus removal properties than one-component metals. In addition, the use of aluminum salts in some studies is harmful to the environment and humans, while magnesium salts are more environmentally friendly. The amalgamation of lanthanum and magnesium with palygorskite can not only improve the zero-charge point of the material, but also enhance the dispersion of lanthanum, thereby improving the phosphorus removal performance of the material (Shi et al. 2019). Therefore, magnesium and lanthanum were selected for the secondary modification of high-temperature treated palygorskite. Then its adsorption capacity and mechanism of phosphate were investigated, along with its recyclability and reproducibility. On this basis, the material's adsorption effect on phosphate in piggery wastewater was further studied.

The original palygorskite (PAL) was procured from Xuyi County Sinoma Gravite Clay Co., Ltd (Jiangsu, China). Potassium dihydrogen phosphate (KH₂PO₄), lanthanum chloride heptahydrate (LaCl₃⋅7H₂O) and magnesium chloride hexahydrate (MgCl₂⋅6H₂O) were from Maclean Biochemical Technology Co., Ltd (Shanghai, China). The chemical reagents used in the experiment reached analytical purity. The piggery wastewater was from a pig farm in Hubei Province. The collected wastewater was filtered with gauze and stored in a refrigerator for later use.

The original palygorskite (PAL) was roasted in a muffle furnace with a set temperature range (500–900 °C) for a certain period of time (2–5 h). Then, the sample with the best phosphate adsorption performance was selected. The palygorskite sample after high-temperature treatment was named HPAL. Then, lanthanum and magnesium were loaded onto HPAL by coprecipitation. Firstly, a certain amount of LaCl₃⋅7H₂O, MgCl₂⋅6H₂O and HPAL were added to water, and shaken in a thermostatic water bath shaker at 200 rpm for 4 h. Then, the pH value of the mixture was adjusted to 10 with sodium hydroxide solution. The mixture was then oscillated for another 8 h and left to rest at room temperature for 12 h. After standing, the samples were filtered, washed with deionized water and dried in a drying oven at 65 °C. Eventually, the dried samples were roasted in an electric furnace with a set temperature of 500 °C for 2 h. The material obtained through the above-mentioned lanthanum and magnesium modification process was named LM-HPAL.

The surface structure and element distribution of the materials were analyzed by a scanning electron microscope with an energy-dispersive spectrometer (SEM-EDS; MIRA3, Czech Tescan). The specific surface area of the material was analyzed by Brunauer–Emmett–Teller (BET; ASAP 2020 HD88, Micromeritics). The crystal structure of the material was analyzed by an X-ray powder diffractometer (XRD; Empyream, PANalytical B.V.). The functional groups of the material were analyzed using Fourier transform infrared spectroscopy (FT-IR; Nicolet 6700, Thermo Fisher Scientific). The surface elements and their chemical states were analyzed by X-ray photoelectron spectroscopy (XPS; ESCALAB 250Xi, Thermo Fisher Scientific).

The adsorption behavior of the LM-HPAL material to phosphate under different conditions was studied by batch experiments. In this study, KH₂PO₄ was used to prepare a 50 mg/L phosphorus stock solution to simulate phosphorus wastewater. Typically, 1 g/L of adsorbent was added to a reaction flask and mixed with 100 mL of 10 mg P/L simulated wastewater. The mixture was then placed in a thermostatic water bath shaker and continuously oscillated at 200 rpm. Then, a certain amount of mixed solution was extracted at a predetermined time and filtered using a 0.45 μm drainage filter. The phosphate content was analyzed using ammonium molybdate spectrophotometry. In addition, the initial pH of the simulated wastewater was adjusted by adding 0.5 mol/L HCl solution and 1 mol/L NaOH solution. Unless otherwise stated, all experiments were conducted at room temperature.

The effect of the initial pH value on phosphate adsorption was investigated in the solution pH range of 2–11. The influence of concentration on phosphate adsorption was studied by adjusting different initial concentrations of phosphate (1–40 mg P/L). Since the pH range of natural water bodies is about 6.5–8.5, the adsorption kinetic behavior of the material under neutral conditions was also investigated. The adsorption isotherm of the material in the range of 1–50 mg P/L phosphate was investigated. In addition, the effects of competitive anions (Cl⁻, ⁠, ⁠, and ⁠) with ionic concentrations of 2 and 10 mmol/L on adsorption were also investigated. Different concentrations of NaCl solution were used to simulate three water bodies with different salinities, so as to study the impact of salinity on phosphate adsorption. The three simulated waters were fresh water (8.5 mM NaCl), brackish water (85 mM NaCl) and seawater (600 mM NaCl) (Huo et al. 2021). Additionally, 1 mol/L sodium hydroxide solution was chosen as the eluent to study the reproducibility of the adsorbent. In the adsorption–desorption experiment, the simulated solution was 40 mg P/L. Otherwise, only circulating adsorption experiments were performed on the dried adsorbent to investigate its reuse when the desorption agent was not used. Finally, the treatment effect of the LM-HPAL material on the piggery wastewater was investigated.

The data from all experiments were obtained as the average of three sets of parallel experimental data. The formulas for calculating phosphate removal rate (e, %) and adsorption capacity (q_e, mg/g) are shown in Supplementary Eqs. (S1) and (S2). The experimental data processing and analysis methods are shown in Supplementary Text S1.

The effect of roasting temperature and retention time of the material on phosphate adsorption was investigated under different conditions (Supplementary Figures S1 and S2). The results indicated that the best adsorption effect of palygorskite on phosphate was achieved when the roasting temperature reached 700 °C. Furthermore, increasing the retention time from 2 to 4 h led to a synchronous increase in the adsorption capacity. However, when the roasting temperature exceeded 700 °C and the holding time exceeded 4 h, the adsorption effect would deteriorate. As the temperature increased, the physically adsorbed water on the palygorskite was gradually removed, which enhance its affinity for phosphate. However, if the critical conditions were exceeded, the crystal structure would collapse and the pore structure would be destroyed. Therefore, the palygorskite material held at 700 °C for 4 h was named HPAL, with an adsorption capacity of 1.65 mg P/g. It was used for subsequent experiments.

According to Wei et al. (2021), the molar ratio of 1:1 for lanthanum and magnesium was selected, and the ratio of metal salts to palygorskite was explored under this condition. The analysis results are displayed in Supplementary Figure S3. As can be observed from the figure, the higher the proportion of metal salts, the better the adsorption. When the content of La–Mg exceeded 0.003 mol/g, the adsorption capacity of the material was stabilized at around 10 mg P/g, and its adsorption capacity for phosphate increased by 83.5% compared to HPAL. Supplementary Table S1 shows that the specific surface area of the material reached its maximum at this ratio. Therefore, this study selected a sample with a metal salt content of 0.003 mol/g for follow-up research and named it LM-HPAL. In addition, the calcined palygorskite material loaded with 0.003 mol/g of lanthanum was prepared and named La-HPAL.

After the modification of lanthanum and magnesium, the strength of most peaks of palygorskite was reduced. This may be because NaOH, added during the preparation of the material, interacted with the mineral and changed the structure of the mineral (Kong et al. 2018). The new peaks at 27.1, 29.4, 43.0 and 50.3° can correspond to (0 1 1), (0 0 2), (−2 0 2) and (2 1 0) plane reflections of the LaOOH (PDF# 19-0656). Peaks occurred at 23.8, 26.2 and 36.6° can be assigned to the reflection of LaCO₃OH (PDF# 49-0981) in the (1 1 1), (0 1 2) and (1 2 2) planes. This indicated that lanthanum was successfully loaded onto the palygorskite in different forms. After the adsorption of phosphate, the peaks of the different forms of lanthanum mentioned above were significantly weakened. The reflections of LaPO₄ (PDF# 32-0493) in the (2 0 0), (−2 1 2) and (−2 1 4) planes can correspond to the peaks at 26.8, 36.6 and 60.0°. In addition, diffraction peaks of lanthanum oxide phosphate (La₅P₆O_22.5, PDF# 21-0046) were present at 42.5 and 80.8°. These findings suggested that lanthanum in the LM-HPAL material plays a major role in the process of adsorption. The captured phosphate was mainly deposited on the LM-HPAL material in the form of P–La complexes. In addition, the analysis of magnesium can be viewed in Supplementary Text S3 and Figure S4.

Through the FT-IR data map of the material, the conversion of its surface functional groups and the adsorption mechanism were analyzed (Figure 3(b)). In the spectrum, the frequency bands around 3,450, 3,160, and 1,634 cm⁻¹ correspond to the O–H structure of the material and the physically adsorbed water (zeolite water) (Mahmoud et al. 2013). The band at 1,400 cm⁻¹ was the vibrational mode of anion (Zhang et al. 2012). The band value at 3,450 cm⁻¹ was the tensile fluctuation of O–H, which was formed by the surface hydroxyl group (Mahmoud et al. 2013). This surface hydroxyl group facilitated phosphate adsorption by the material. In the LM-HPAL spectrum, this peak became stronger, which may be caused by lanthanum loaded onto the material to form a complex.

The band at 1,035 cm⁻¹ was generated by the in-plane tensile fluctuation of Si–O (Wang et al. 2008). The band at 471 cm⁻¹ was formed by the bending fluctuation of O–Si–O (Li et al. 2016). The band strength of these two places remained unchanged after calcination, indicating that calcination had little effect on the crystallinity of quartz. The band at 881 cm⁻¹ was the deformed fluctuation of Si–O–H (Li et al. 2016), but the band disappeared after calcination. This indicated that calcination had an effect on the structure of the palygorskite. In the LM-HPAL spectrum, the peak at 471 cm⁻¹ became stronger, which may be due to the formation of MgO after the modification of the palygorskite, so the tensile vibration of Mg–O was superimposed here (Li et al. 2019).

The frequency band at 1,490 cm⁻¹ may be the offset fluctuations of carbonate impurities (Morsch et al. 2018). In the preparation of the LM-HPAL material, during the oscillating reaction phase, CO₂ may dissolve in solution and convert to and react with lanthanum in solution under alkaline conditions (Kang et al. 2015). During the calcination stage, the La (OH)₃ precipitation in the material may react with CO₂ in the air. Lanthanum carbonate complexes may be formed at both stages. Nonetheless, the peak intensity at 1,490 cm⁻¹ diminished after adsorption, indicating that lanthanum carbonate complexes were involved in phosphate capture. The band situated at 694 cm⁻¹ may be the La–OH structure in lanthanum complexes (Aghazadeh et al. 2014). After the material adsorbed phosphate, the peak strength at 1,490 and 694 cm⁻¹ decreased, indicating that La in the material provided the main adsorption sites.

According to previous research (Kong et al. 2019), there are three chemical states of O in the LM-HPAL material. Therefore, the O 1s spectrum can be divided into three distinct peaks. They corresponded to zeolitic water, M–OH (hydroxyl group bound to metal) and M–O (oxygen bound to metal), respectively. After adsorption, the relative peak area of M–OH diminished from 62.74 to 44.06%, while the relative peak area of M–O raised from 22.79 to 28.87%. Moreover, the relative area of physically adsorbed water increased from 14.47 to 27.07%. These findings indicated that ligand exchange reactions occurred during the adsorption, and hydroxyl groups in the material were replaced by phosphates and released into the solution. In addition, the ratio of the relative area of M–OH (surface hydroxyl group) before and after adsorption was about 1.42, which is within the range of 0.5–2. This suggested that the complexes formed during adsorption may be monodentate, bidentate mononuclear or bidentate binuclear (He et al. 2016). Combined with the change of O–H peak in the FT-IR pattern, it can be concluded that ligand exchange is the main mechanism by which the materials adsorb phosphate.

Phosphate in water has several forms, such as ⁠, and and the ratio between them is being affected by the pH value (Supplementary Figure S6). The phosphorus removal efficiency of the LM-HPAL material under various pH conditions is demonstrated in Figure 5(b). The pH_pzc of the LM-HPAL material is 7.1 (Supplementary Figure S7). Within the pH range of less than 7.1, the surface of the LM-HPAL material is positively charged. Therefore, within the pH range of 2–7, the LM-HPAL material can effectively adsorb phosphate. Overall, the phosphate removal rates were maintained above 95% in the pH range of 2–11. Most surface water has a pH value between 6.5 and 8.5. Therefore, combined with the experimental results of Figure 5, the pH of phosphate-simulated wastewater was held between 7 and 8 for subsequent experiments.

Figure 5(c) displays the effect of the LM-HPAL material on phosphate removal when several common interfering ions coexist separately with phosphate. The concentrations of coexisting ions were chosen as 2 and 10 mM. The experimental results of coexisting anion indicated that Cl⁻, and had an inconspicuous impact on phosphate by the LM-HPAL material. When they coexist with phosphate, the phosphate removal rate of the material remains above 80%. However, and had obvious inhibitory impacts on phosphate removal by the LM-HPAL material. In particular, when was 10 mM, the phosphate removal rate of the LM-HPAL material was only about 70%. The adsorption capacity of the LM-HPAL material to phosphate was reduced to 7.58 mg P/g. This may be due to the fact that the Ksp (3.98 × 10⁻³⁴) of La₂(CO₃)₃ is smaller than the Ksp (3.7 × 10⁻²³) of LaPO₄, resulting in competing with phosphate for adsorption sites (Zhang et al. 2017b). The pH changed during the reaction, so both and can directly or indirectly affect the adsorption efficiency of phosphate. The effect of salt ion concentration on phosphate adsorption by the LM-HPAL material was investigated by adding different contents of NaCl in simulated wastewater. Figure 5(d) displays the effect of salt ionic concentration of different water bodies simulated by NaCl solution on the phosphorus removal performance of the LM-HPAL material. The data of salt concentration experiment implied that the phosphate removal rate of the LM-HPAL material remained above 90% in the range of 0–85 mM salt ionic concentration. The adsorption of phosphate by the LM-HPAL material was only slightly inhibited at a salinity of 600 mM. At this point, the adsorption capacity of the LM-HPAL material decreased by 12.92% compared to 0 mM salt concentration. This indicated that the adsorption phosphate LM-HPAL material is almost unaffected by salt ionic concentration. Previous research (Zhang et al. 2017a) had shown that the adsorption effect of rare-earth element zirconium-modified materials on phosphate did not change due to an increase in salt ionic strength, likely because of the inner-sphere surface precipitations formed with phosphate. If the adsorption effect is significantly reduced with an increase of salt ionic strength, the outer-sphere surface precipitations are formed. The above results showed that the LM-HPAL material and phosphate-generated inner-sphere surface precipitations.

The isotherm of the LM-HPAL material to phosphate was further discussed. Both the Langmuir (Supplementary Eq. (S6)) and Freundlich (Supplementary Eq. (S7)) adsorption isotherm models were used to analyze the isotherm experimental results of LM-HPAL material (Supplementary Table S5). The adsorption isotherm plot is displayed in Figure 6(b). The fitted results showed that the Freundlich model (R² = 0.973) provided a better description of the phosphate adsorption characteristics of the LM-HPAL material compared to the Langmuir model (R² = 0.813). This indicated that phosphate had multi-layer adsorption in the heterogeneous system of the LM-HPAL material. Based on the fitted results, the theoretical maximum adsorption capacity was 20.4 mg P/g. The R_L values (R_L = 1/(1 + K₁C₀)) obtained from the Langmuir model ranged from 0.00146 to 0.06793, all of which were close to zero. This indicated that the LM-HPAL material had a strong affinity for phosphate. Meanwhile, the Freundlich model fitted results showed that the adsorption affinity constant was greater than 1, indicating that the LM-HPAL material can effectively adsorb phosphate.

In the fifth adsorption–desorption experiment, the LM-HPAL material exhibited a maximum adsorption capacity of 32.2 mg P/g, whereas the maximum adsorption amount in the adsorption cycle without NaOH solution treatment was only 17.7 mg P/g. This suggested that the alkaline condition modified the crystal structure of the palygorskite, thereby increasing its exchange capacity. During the washing and drying processes, the specific surface area of the LM-HPAL material increased (Supplementary Table S1). Therefore, in the first few cycles, the adsorption capacity of the LM-HPAL material showed an upward trend. However, as the number of cycles increased, more and more phosphorus accumulated in the material. When the additional adsorption site after desorption or washing is less than the site occupied by phosphate, the utilization rate of the adsorption site will gradually become saturated. Consequently, the adsorption site gradually decreased, leading to a reduction in the adsorption capacity of the material. Based on the above results, it can be concluded that the LM-HPAL material has the potential for reuse and demonstrates satisfactory renewable performance.

The practical application effect of the LM-HPAL material was further discussed through the adsorption experiments on the piggery wastewater after preliminary filtration (Supplementary Table S6). The soluble phosphate content in the piggery wastewater was measured at 7.03 mg/L, and the total phosphorus content was 7.61 mg/L. The adsorption treatment was carried out at the original pH (8.2) of the wastewater. The total phosphorus removal rate reached 95.3% (Figure 6(d)). The adsorption amounts of orthophosphate and total phosphorus at equilibrium were 6.95 and 7.25 mg P/g, respectively. The adsorption process of piggery wastewater also followed the pseudo-second-order kinetic model, but the presence of other organic or inorganic pollutants in the piggery wastewater disrupted the adsorption process. As a result, the adsorption effect of the LM-HPAL material on orthophosphate was reduced by 38.62% compared to simulated wastewater. Nevertheless, the LM-HPAL material still showed great potential for practical applications (Supplementary Table S7).

In this exploration, the LM-HPAL material was prepared by loading lanthanum and magnesium onto the thermally modified palygorskite through coprecipitation-roasting. The SEM and element mapping showed that the surface of the LM-HPAL material exhibited a multi-layer network structure, with La being relatively evenly dispersed in the material. Under the same conditions, LM-HPAL exhibited a 100% phosphate removal rate, whereas La-HPAL showed a phosphate removal rate of only 55.39% (Supplementary Table S8). The results of zeta potential showed that the main function of Mg was to increase the zero-charge point of the LM-HPAL material and enhance the surface charge of the material. Moreover, the LM-HPAL material was found to be effective in adsorbing low-concentration phosphate within a pH range of 2–7. The kinetic curves displayed that the LM-HPAL material can adsorb phosphate rapidly in the early stage. The adsorption can reach the equilibrium state within 4–6 h. The isotherm results were consistent with the Freundlich model, indicating that phosphate existed in the form of multi-layer coverage in the LM-HPAL material. The coexistence anion studies showed that and had a certain inhibitory effect on phosphate adsorption. In addition, the LM-HPAL material exhibited nice renewable properties and could be reused at least five times. In the adsorption–desorption cycle, the maximum phosphate adsorption capacity of the LM-HPAL material could achieve 32.2 mg P/g after elution treatment with NaOH solution. The equilibrium adsorption capacity of the LM-HPAL material for total phosphorus in piggery wastewater was 7.25 mg P/g, and the removal rate of total phosphorus reached 95.3%. Based on the XRD, FT-IR and XPS results of the LM-HPAL material, phosphate mainly exchanged ligands with La–OH in the material to form LaPO₄ complexes. The adsorption mechanisms of the LM-HPAL material for phosphate include ligand exchange and inner-sphere complexation. In summary, the LM-HPAL material has great potential for piggery wastewater treatment.

We thank the editors and anonymous reviewers for their comments and suggestions to improve the quality of this paper.

This work was funded by the Hubei New Rural Development Research Institute (Yangtze University) Open Fund.

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

The authors declare there is no conflict.